Technical Field
[0001] The present invention is in the field of security elements and authentication methods,
as used for verifying the origin, genuineness and/or authenticity of items such as
products or documents of value.
Background of the Invention
[0002] Many products of commercial value need to be protected against counterfeiting, forging
and copying. To this end, products of high value, such as perfumes or watches, as
well as documents of value, such as banknotes, tax stamps, credit cards etc., are
typically provided with security elements.
[0003] Typical security elements include for instance holograms, markings with luminescent
dyes or pigments emitting in the visible spectrum upon excitation by e.g. UV radiation
(one example of fluorescent inks that may be used for security applications is described
in
US2005/0120907A1), watermarks, or graphical elements using a specific kind of pigment that is not
easily available and/or which provides an optical impression by a specific orientation
of the pigment that is difficult to achieve with commercially available equipment.
An example of the latter is e.g. the so-called "rolling bar" effects that can be provided
by orienting magnetic non-spherical particles.
[0004] A drawback of such security elements is that they may be relatively easy to reproduce
and/or are not machine-readable. Further, the security elements are typically provided
in an identical manner on products of the same type, so that they cannot be used to
identify a specific product and cannot be used to distinguish between different products
of the same type. This is however desirable in many commercial fields, as this allows
to track a product over the distribution chain to identify stolen goods.
[0005] In order to solve these problems, the prior art suggests using specific kinds of
codes, such as alphanumeric product verification codes, barcodes or QR codes. However,
the problem of such codes is that the information contained therein can be easily
decoded. It is further possible for counterfeiters to predict, within certain boundaries,
a code that could be considered authentic, as the algorithms used for the production
of such codes are in the public domain or can be obtained by analysing the information
provided on a series of authentic products.
[0006] In order to address these problems of pre-designed codes, the prior art suggests
using random distribution features that are unique for each product and stored in
a database, to thereby allow to identify a product as genuine by comparison with database
entries. The random distribution forms a security feature that cannot be predicted,
as it is not a pre-designed code.
[0007] One document describing such a technology is
GB 2 324 065 A, which describes an identification code for banknotes or credit cards comprising
a pattern of random beads in a plastic matrix. The position of the beads inside the
plastic matrix is unique for each product, such as credit card or banknote, and e.g.
the position of a sequence of beads above or below a line represents the ones and
zeroes in a binary code that is used for identifying the product.
[0008] A similar technology is described in
EP 1 953 684 A1, which describes an authentication means including a random arrangement of stains.
This document describes that such a unique arrangement of stains can be obtained e.g.
by spraying an ink, which can be conventional or covert such as to be detectable only
under specific illumination conditions. The random arrangement of stains forms a machine-readable
code that can be read out by image processing, forming a descriptor or data set that
corresponds to the stains arrangement. A related technology using invisible taggants
that are randomly positioned inside a material as matrix, forming a random pattern
that can be used for authentication by comparison with a database, is described in
US 7,687,271 B2.
[0009] A common problem of all the technologies described above is that the formation of
the arrangement of stains or taggants, and subsequently the registration of the resulting
pattern, is performed in a database at a single place, i.e. a place of manufacture.
Nowadays many commercial goods are however prepared in multi-step processes performed
at different sites or by different manufacturers. One example is the production of
a jet engine where the blades of the turbine may be prepared from a specific highly
resistant material produced at a first site, and the engine is assembled at another
site by qualified mechanics. In such a case, it would be desirable to have available
a means for securing that the right material has been used, and that the assembly
has been correctly conducted. A security element providing authentication for either
one of these will not be sufficient to prove both, and a single security element clearly
identifying a specific produced item and capable of tracing the course of manufacture
in a simple, yet unique manner is desired.
[0010] Another example of prior art security elements is disclosed in
US2009/0074231A1 relating to a secure article, notably a security and/or valuable document, comprising:
at least one substrate, at least one visible authenticating structure, at least one
inspection zone defined at least partly by the authenticating structure, the authenticating
structure delimiting at least partly the outline of the inspection zone and/or serving
to locate the inspection zone, in the inspection zone, at least one identifying element
different from the authenticating structure, the inspection zone enabling an identification
information to be supplied with at least one feature, notably spatial or physical,
of said at least identifying element.
[0011] One more example of prior art security elements is disclosed in
WO 2009/010714 relating to a security device comprising a substrate having a viewing region. A first
pattern of a first colour and a second pattern of a second colour are printed on one
side of the viewing region. A third pattern of the first colour and a fourth pattern
of the second colour are printed on the other side of the viewing region. The first
and second patterns on the one side of the viewing region overlap the fourth and third
patterns respectively on the other side of the viewing region whereby a) when each
side of the viewing region is viewed in reflected light, the patterns on that side
can be distinguished by at least their colours, and b) when the viewing region is
viewed in transmitted light, from either side of the substrate, the viewing region
is sufficiently transparent that subtractive colour mixing between the overlapped
different colours results in a perceived single resultant colour being visible.
[0012] WO2005/104008 A1 describes one of methods for verifying authenticity of an article, which involves
producing a representation of a distribution pattern of randomly distributed taggant
particles on the article and producing a match signal when the representation at least
partially matches a pre-defined representation of the distribution pattern of the
randomly distributed taggant on the article.
Problems solved by the present invention
[0013] The present invention generally aims at providing a new security element capable
of improving the security level provided by prior art security elements.
[0014] It is a particular object of the present invention to provide a security element
that provides a high level of security in that it is extremely difficult to counterfeit
or reproduce and that cannot be predicted from a series of genuine products, yet that
can be prepared at low costs with uncomplicated equipment. It is a further object
of the present application to provide such a security element that provides several
levels of security in that it contains different patterns that are revealed under
different conditions, such as different viewing conditions, to thereby increase the
security level with respect to prior art random distribution security features.
[0015] It is a further object of the present invention to provide a security element that
provides a means for securing the genuineness and authenticity of a product or item
along the production or distribution chain.
Summary of the Invention
[0016] The present invention can be summarized by the following aspects. Further aspects
and preferred embodiments will become more apparent from the following detailed description.
[0017] A security element is provided as defined in claim 1.
[0018] The security element may be such that the first material (INK1) comprises one or
both of a first dye (DYE1) and a first pigment, and the second material (INK2) comprises
one or both of a second dye (DYE2) and a second pigment.
[0019] The security element according to another embodiment is such that the discrete elements
of at least one of the first and second patterns are not visually distinguishable
from the substrate. Preferably, the discrete elements of one of the first pattern
and the second pattern are not visually distinguishable from the substrate, and the
discrete elements of the other of the first pattern and the second pattern are visually
distinguishable from the substrate. Herein, "visually distinguishable" and "not visually
distinguishable" typically denote the distinguishability, respectively the lack thereof,
by the naked eye of a healthy human observer under typical illumination conditions,
such as under artificial light of an incandescent lamp.
[0020] The security element may be such that the second emission wavelength range λ2e does
or does not overlap with the first emission wavelength range λ1e. Furthermore, the
security element may be such that λ1a-max < λ1e-max < λ2e-max, wherein λ1a-max, λ1e-max,
and λ2e-max denote the wavelengths of the excitation and emission peaks in the respective
excitation and emission wavelength regions of the first dye (DYE1) or pigment and
the second dye (DYE2) or pigment, respectively.
[0021] The security element according to this and other embodiments is preferably such that
the randomly distributed discrete elements are obtainable by spraying ink.
[0022] According to another embodiment, a commercial good or value document is provided
that comprises a security element according to any one of the above embodiments.
[0023] According to another embodiment, a security arrangement is provided that comprises
a security element according to one of the above embodiments, and a data record of
an index for identifying said third pattern. The security arrangement can be such
that it also comprises a data record of a first index for identifying said first pattern,
a data record of a second index for identifying said second pattern and a data record
of a third index for identifying said third pattern.
[0024] According to the invention, a process for producing a security arrangement is defined
in claim 11.
[0025] According to the invention, a method for authenticating a security element is defined
in claim 13.
[0026] The method may comprise the additional steps of obtaining an image of said first
pattern and applying a predetermined first indexing routine to said obtained image,
for generating a first index of said first pattern, accessing said repository of said
data records, and comparing said generated first index with content from said repository,
wherein said authentication decision is also based on said comparing of said generated
first index with said content from said repository.
[0027] The method may furthermore comprise the additional steps of obtaining an image of
said second pattern and applying a predetermined second indexing routine to said obtained
image, for generating a second index of said second pattern, accessing said repository
of said data records, and
comparing said generated second index with content from said repository, wherein said
authentication decision is also based on said comparing of said generated second index
with said content from said repository.
[0028] In the present invention, the discrete elements of at least one of the first and
second patterns are distributed randomly. In one embodiment, the discrete elements
of one of the first and second patterns are distributed randomly, and the discrete
elements of the other of the first and second pattern are distributed non-randomly.
In another embodiment, the discrete elements of both the first and the second pattern
are distributed randomly.
Brief description of Figures
[0029]
Fig. 1 shows an example of a security element according to an embodiment;
Fig. 2 shows an example of a side view of a security element according to an embodiment;
Fig. 3 shows an example of a schematic view of a security arrangement according to
an embodiment;
Fig. 4 shows an example of a method for producing a security arrangement;
Fig. 5 shows an example of a method for authenticating a security element and a security
arrangement;
Fig. 6 shows embodiments of the invention in a system configuration relating to the
generation of an inventive security element and security arrangement, and to the authentication
of such a security element;
Fig. 7 shows another example of a security element according to an embodiment;
Fig. 8 shows examples of spray spatter patterns for explaining aspects of the present
invention;
Fig. 9 shows further examples of spray spatter patterns for explaining aspects of
the present invention;
Fig. 10 schematically illustrates the wavelength relationships between two dyes or
pigments that can interact energetically;
Fig. 11 shows luminescence emission and excitation spectra for two example dyes;
Fig. 12 shows a flow diagram of an example process of signature enrollment in a database;
Fig. 13 shows a flow diagram of an example of an authentication/identification process;
Fig. 14 shows a flow diagram of another example of an authentication/identification
process;
Fig. 15 shows a flow diagram of another example of an authentication/identification
method;
Fig. 16 shows an example of an indexing method;
Fig. 17 schematically illustrates a simplified working principle of a concentric nebulizer;
Fig. 18 schematically illustrates an example of a microspray system;
Fig. 19 schematically illustrates details of a microspray system;
Fig. 20 shows graphs of color components in two different color spaces according to
a color blending embodiment;
Fig. 21 shows contours extracted using a color de-convolution operation;
Fig. 22 shows contours extracted using color thresholding; and
Fig. 23 shows histograms related to the color thresholding of Fig. 22.
Figure 24, in particular Figures 24a to 24e, show specific embodiments of a spectral
overlap as schematically illustrated in Figure 10. Herein, MAT1 and MAT2 are used
as synonyms for INK1 and INK2, respectively.
Detailed Description of the Invention
[0030] In the present application, the terms used in the specification are given their normal
understanding in the art, unless specified otherwise. Irrespective of this, the following
provides a list of definitions of the terms used in the present invention:
The term "comprising" is used open-endedly, and allows for the presence of further
components that are not specifically recited. However, the term also encompasses the
more restrictive meanings "consisting of' and "consisting essentially of', so that
the term also encompasses the possibility that non-mandatory and not recited components
are absent.
[0031] The term "at least" is used to denote that mandatorily not less than the recited
amount is present. For instance, the term "at least two" requires the presence of
two or more of the recited components, yet also allows for the possibility of further
components. The term defines as such no upper limit, but often the amount of the recited
species or elements is limited in practice, as well known to the skilled person. In
many instances, a practicable upper limit is ten-fold the recited amount (e.g. twenty
if "at least two" are recited), preferably five-fold, and more preferably twofold
or threefold. However, a similar manner as outlined above for the term "comprising",
the term "at least" also encompasses the possibility that no more than the recited
amount is present, e.g. an amount of exactly two if "at least two" is recited.
[0032] Similarly, the term "x or more", such as "two or more", is used to denote that at
least the recited amount is present, but the term also encompasses the possibility
that more than the recited components are present. While in many cases the recited
minimum amount will be sufficient, a higher amount or number may be preferable for
certain applications. Again, the term per se does not define an upper limit, but often
the amount of the recited species or elements is limited in practice, as well known
to the skilled person. The term "one or more" preferably means one, two, three, four,
five, six or seven, more preferably one, two, three, four or five, and even more preferably
one, two or three, most preferably one or two, and the term "two or more" preferably
means two, three, four, five, six or seven, more preferably two, three, four, or five,
and even more preferably two or three.
[0033] The term "wavelength range", such as in the expressions "emission wavelength range"
and "excitation wavelength range", generally denotes the range around a peak at a
wavelength λmax in which excitation or emission, respectively, is observed. More precisely,
it defines the area around a peak value λmax in a optionally normalized and background-subtracted
spectrum, as measured on a transparent substrate such as a plastic (e.g. polyester)
film or carrier, including the respective peak and the shoulders thereof up to the
points where the line of the optionally normalized and background-subtracted spectrum
crosses the baseline (i.e. the reading in the optionally normalized and background-subtracted
spectrum where the observed value becomes zero). This range is centered around the
respective peak λmax. A wavelength range may thus also be regarded as the breadth
of the respective peak in an excitation or emission spectrum. As one example, if a
given first dye exhibits a peak in an excitation spectrum at 450 nm, and breadth of
this peak extends to wavelengths of 440 and 460 nm, respectively, the excitation wavelength
range is from 440 to 460 nm.
[0034] The term "ink" shall encompass any composition that can be used for image or mark
formation in an image forming or printing process. It includes water-based and organic
solvent-based inks used for generally known printing processes, such as inkjet printing,
offset printing, screen printing, lithographic printing, planographic printing or
intaglio printing. In a preferred meaning, the term "ink" denotes a material that
can be applied to a substrate by a spraying process and that retains on a substrate.
The term ink thus also includes e.g. lacquer.
[0035] The term "pattern" is used to denote an assembly or arrangement of a multiplicity
of discrete elements of identical composition. In the present invention, a pattern
is formed on or in a suitable substrate.
[0036] The area occupied by a pattern (also referred to as pattern area) can be of regular
form, such as a square, rectangle, triangle or circle, but can also be of irregular
form without clear boundaries. The pattern area can be visibly restricted, such as
by a solid line, but can also be unrestricted without clear outer boundaries. Preferably,
the pattern area defines a regular form selected from rectangle, square, triangle,
and indicia, such as logos, characters/letters and numbers.
[0037] The terms "randomly distributed" and "random distribution" denote a distribution
of discrete elements that does not follow certain construction criteria or a certain
scheme. Also, the distribution does not form a regular, repeating pattern. A random
distribution of discrete elements is obtained from a random process for the formation
of discrete elements on a substrate. One example of a random process for the formation
of discrete elements is the spraying of an ink composition, leading to a random distribution
of the ink stains (discrete elements) on the substrate. Another example of a random
distribution of discrete elements is the random distribution of small particles in
a paper substrate obtainable by adding a quantity of the particles to the bulk material
forming the substrate during the papermaking process.
[0038] The term "discrete elements" is used to denote an entity that forms and defines a
part of a pattern. An example of a discrete element is an ink stain obtained from
a spraying process. Another example is a particle of certain material.
[0039] The outer shape of the discrete elements is not particularly limited, and can be
regular or irregular. An example of a discrete element of regular form is an ink stain
in the form of a disk or solid ellipse. Depending on the method of manufacture, the
discrete elements will however typically have a slightly irregular form in that there
is no perfect point or mirror symmetry present in a single discrete element. A discrete
element may thus have any shape between a perfect symmetric shape and a fully irregular
shape. Also, for instance in the case of an ink spray spatter, the discrete elements
may be formed from single droplets of the ink, but may also be formed by multiple
droplets forming partially overlapping stains, together forming a discrete element
in the sense of the present invention.
[0040] The size of the discrete elements can be suitably chosen and is typically such that
a discrete element can be easily detected by available detection technology. In view
of the intended end purpose of the security element of the present invention and in
consideration of the necessity of forming a pattern by a multiplicity of discrete
elements in a relatively small area, the size of the discrete elements, expressed
in term of their equivalent diameter determined by a microscopic method and as the
number median value Dn50 of the discrete elements diameter distribution, is preferably
small, such as 200 micrometer or less.
[0041] The term "visible range" means from 400 to 700nm, "UV range" from 40 to less than
400nm and "IR range" more than 700nm to 2400nm.
[0042] "Fluorescence" denotes the emission of electromagnetic radiation from an excited
state of a material having a lifetime τ of less than 10
-5 seconds in terms of exponential decay according to

where t denotes time in seconds.
[0043] "Phosphorescence" denotes the emission of electromagnetic radiation from an excited
state of a material having a lifetime τ of 10
-5 seconds or longer in terms of exponential decay according to

where t denotes time in seconds.
[0044] A partial spatial overlap is characterized by an area in or on a substrate wherein,
when seen from an axis extending perpendicular to the plane of the substrate, there
are at least three areas recognizable under certain viewing conditions: An area wherein
discrete elements of the first pattern, but no discrete elements of the second pattern
are provided, an area wherein discrete elements of the second pattern, but no discrete
elements of the first pattern are provided, and an area wherein both discrete elements
of the first pattern and discrete elements of the second pattern are provided (overlapping
discrete elements). The certain viewing conditions may in some embodiments include
only wavelengths of the visible range, but may in other embodiments also include or
consist of wavelengths in the UV and/or IR range.
[0045] If, in the present description, an embodiment, feature, aspect or mode of the invention
is stated to be preferred, it should be understood that it is preferred to combine
the same with other preferred embodiments, features, aspects or modes of the invention,
unless there are evident incompatibilities. The resulting combinations of preferred
embodiments, features, aspects or modes are part of the disclosure of the present
description.
[0046] The present invention also relates to a method for authenticating an article, and
articles carrying a security element of the invention. The term "article" is to be
understood in a broad sense and includes, but is not limited to, banknotes, value
papers, identity documents, cards, tickets, labels, security foils, security threads,
products of value, such as perfume, and product packages.
[0047] Figure 1 shows a first embodiment of the present invention. A substrate 100 is provided,
and comprises a first region 101 and a second region 102. The first and the second
regions overlap at least partially. The two regions can also be identical, or one
of the regions can enclose the other. In the first region 101, a plurality of discrete
elements 105a to 105g is provided, which form a first pattern 105. In the example
of figure 1 the discrete elements are shown as ink stains, but this is only a non-limiting
example. The second region 102 contains a second pattern 106 formed by discrete elements
106a-106i. In the shown example the pattern 106 is a grid pattern of horizontal and
vertical lines, but this is again only a non-limiting example.
[0048] A pattern in the meaning of the present description is an arrangement that is distinctive
from its background, i.e. distinctive from the respective region of the substrate
on or in which it is provided. As such, the discrete elements of the first pattern
must not cover the entire first region of the substrate and the discrete elements
of the second pattern must not cover the entire second region of the substrate, as
in that case there would be no pattern. The coverage of the pattern expressed in terms
of the ratio of covered area of the region to total area of the region is preferably
less than 50%, more preferably less than 20%.
[0049] The discrete elements 105a-105g of the first pattern 105 are of a first material,
for example an ink of a first composition applied to the substrate 100. The discrete
elements 106a-106i of the second pattern are of a second material, for example an
ink of a second composition applied to the substrate 100.
[0050] In accordance with the concept of the present invention, the discrete elements of
at least one of the first and second patterns are distributed randomly. In the example
of figure 1, the discrete elements of the first pattern 105 are shown as a random
spatter pattern of ink stains. For example, such a random spatter pattern of ink stains
can be generated with a spray nozzle and a nebulizer. The second pattern 106 in the
shown example is a deterministic pattern, e.g. a regular pattern that can be applied
by a printing process, such as inkjet printing. However, this is only one example,
and the second pattern may also be provided in other ways, for example can also be
a random pattern.
[0051] The term "random pattern" expresses that the pattern is generated by a process that
comprises an aleatory characteristic so that a pattern is not systematically reproducible
and the distributions of the discrete elements over the region of the substrate for
individual applications of the generation process are random and uncorrelated, similar
to the term random number, which expresses that a number is generated by a random
process. Besides the formation of random spatter patterns by spraying ink droplets
on a substrate, a random distribution of discrete elements can also be obtained by
performing a random distribution of particles or fibres or threads when producing
a substrate or a part of a substrate, e.g. distributing such particles in the polymer
base of a plastic substrate (or a plastic layer of a composite substrate) while that
base is not yet hardened or cured, or distributing particles in a paper substrate
obtainable by adding a quantity of the particles to the bulk material forming the
substrate during the papermaking process.
[0052] At least a part of the discrete elements of the first pattern 105 and at least a
part of the discrete elements of the second pattern 106 are provided in such a way
that they overlap. In other words, there are specific sections identifiable in the
first and second region in which both discrete elements of the first pattern and discrete
elements of the second pattern are present. For the specific example of figure 1,
figure 1b shows an element 107b that corresponds to the overlap between discrete element
105b and discrete elements 106a and 106g, an element 107e that corresponds to the
overlap between the discrete elements 105e and discrete elements 106b and 106f, and
an element 107g that corresponds to the overlap between discrete element 105g and
discrete element 106h. The elements 107b, 107e and 107g thus form a third pattern
107.
[0053] Another close-up example is shown in Figure 7, where a first ink stain 701 and a
second ink stain 702 are provided on a common region of a substrate 710. The overlap
of the two ink stains 701 and 702 defines a section 703 that is an element in its
own right, i.e. an element of the third pattern.
[0054] In accordance with the concept of the present invention, a security element is defined
by the first pattern 105, the second pattern 106 and the third pattern 107 associated
with the overlap of the discrete elements of the first and second patterns. Consequently,
an authentication process using the security element of the invention may refer to
the third pattern for confirming the authenticity of the element.
[0055] The first and second region may be implicitly defined by the presence of the first
and second pattern, respectively. However, the respective areas may also be provided
at predetermined locations on the substrate. These predetermined locations have the
function of permitting a device for identifying the patterns to identify the regions,
such that the first and/or second and/or third pattern can then be readily located.
The locations can be indicated indirectly, e.g. as coordinates with respect to identifiable
locations of the substrate, such as the corners 110, 111 shown in Fig. 1, or they
can be indicated explicitly by appropriate markings that identify the first and second
regions. For example, the dotted lines 101 and 102, which in the general context of
the present invention are only for reference purposes and generally do not relate
to a visible marking, could in such a special embodiment relate to visible elements
that serve to identify the first and second region. Identifiable locations on the
substrate may also be constituted by the edges or other angles of the object or document
contour or boundaries.
[0056] The degree of overlap between the two patterns can be set in any suitable or desirable
way. In general, it is desirable to provide third patterns that have a small coverage
with respect to the overlap area of the first and second region of the substrate,
because in this way the number of distinctive states provided by the different third
patterns is large, so that the safety against counterfeiting is increased and the
possibility of a collision (i.e. that two independently generated third patterns are
by chance so similar that they lead to an identical index being generated in the ways
described further on) is made small. This means that the coverage with discrete elements
forming the third pattern 107b, 107e and 107g, i.e. the overlapping portions of discrete
elements 105 of the first and the discrete elements 106 of the second pattern, is
preferably small, such as 20% or less, preferably 10% or less, and further preferably
5% or less of the area defined by the overlap of the first region 101 and the second
region 102 of the first and second pattern, respectively.
[0057] In order for the invention to work, a minimum number of elements of the first pattern
should overlap with some elements of the second pattern in the Region Of Interest
(ROI). The ROI is defined as the intersection of the first region where the first
pattern is applied with the second region where the second pattern is applied. To
obtain a minimum number of overlapped elements within the ROI, the probability of
overlap should be higher than 90%, preferably higher than 95% and more preferably
higher than 98%. To reach such a high probability, the ROI surface coverage, defined
as the ratio of the area covered by discrete elements of the random pattern to the
area of the ROI, should be higher than 2%, preferably higher than 5% and more preferentially
higher than 20%. This applies to either one of both of the first and second pattern
if both patterns are formed by randomly distributed discrete elements. In one embodiment,
the corresponding degree of overlap of the discrete elements of the first pattern
with the discrete elements of said second pattern is such that more than 10%, preferably
more than 30%, more preferentially more than 50% of the discrete elements of the first
pattern overlap with discrete elements of the second pattern. These percentage values
refer to the number of discrete elements, not to the covered area.
[0058] The total coverage of the ROI by discrete elements of the first and second pattern
is typically 50% or less, preferably 40% or less, and more preferably 20% or less.
[0059] The inventive security element provides the advantage of increased safety against
forgery or counterfeiting. In a first aspect, the use of a pattern with random distribution
of discrete elements is much harder to reproduce than a regular pattern, because there
is no systematic process for regeneration of the random pattern, because the random
pattern is arbitrary and there is no correlation between patterns produced at different
instances. In a second aspect, the use of different materials for the discrete elements
of the first pattern and the discrete elements of the second pattern provides a further
layer of security, as a counterfeiter must analyse, obtain and be able to process
the respective materials when attempting to produce a successful forgery. In a third
aspect, the use of the third pattern for defining the security element provides a
further layer of security, as a straight forward analysis of the security element
at most shows the first and second patterns. Preferably, one or both of the first
and second marking are invisible to the human eye under ordinary daylight or conventional
room lighting, so recognizing the patterns is already a challenge for an uninitiated
counterfeiter. However, even if the first and second pattern are recognized, there
is no inherent indication that third pattern is being used for defining the security
element. As a consequence, a forger will not be aware of the importance of the precise
positional relationship of the first and second pattern, which defines the overlap
and ultimately the third pattern, so that even if the first and second pattern are
diligently reproduced, the precise positional relationship will probably not be, thereby
allowing identification of the counterfeit based on the inventive use of the third
pattern.
[0060] Expressed differently, the concept of the present invention allows adding a further
degree of safety in a security element without having to apply an additional marking
or pattern, as the third pattern is provided as the intersecting set of the first
and second patterns.
[0061] Figure 2 shows a side view of an embodiment of a security element. The substrate
200 comprises a layer 201 in which discrete elements 206a-206c are provided. It is
noted that the discrete elements 205 may be seen as corresponding to the discrete
elements 105 of figure 1, and equally the discrete elements 206 of figure 2 can be
seen as corresponding to the discrete elements 106 of figure 1. In the example of
figure 2, the discrete elements 205 are provided on the surface of the substrate,
while the discrete elements 206 are provided within the substrate beneath the surface.
Naturally, this is only an example, and both discrete elements 205 and 206 can be
provided on top of the surface, or both can be provided suitably within the substrate.
[0062] Figure 2 schematically represents the areas of overlap 207a to 207c that result when
viewing the substrate from a predefined angle, for example perpendicular to the surface
area of the substrate.
[0063] Figure 3 schematically shows an embodiment of the invention, in which a security
element of the above kind is combined with a data record of an index for identifying
the third pattern 107 or 207, in order to provide a security arrangement. More specifically,
as indicated in figure 3, features of the third pattern 207 are evaluated by an indexing
routine 300, in order to generate a data element 1(207) associated with the third
pattern 207 and arranged for identifying said third pattern. The stored index 1(207)
can then be used at a later time for authenticating the security element, as will
be explained in more detail further on.
[0064] Figure 4 shows an embodiment of the present invention for producing a security arrangement
as described above. In a first step 401 the first pattern is generated by distributing
the discrete elements of the first material over the first region of the substrate.
In step 402 the second pattern is formed by distributing discrete elements of the
second material over the second region of the substrate. The discrete elements of
at least one of the first and second patterns are distributed randomly. In one embodiment,
the discrete elements of both the first pattern and of the second pattern are distributed
randomly, e.g. by spraying ink droplets, while in another embodiment the discrete
elements forming one of the first and second pattern are randomly distributed (e.g.
by spraying ink) and the discrete elements forming the other pattern are distributed
non-randomly, such as to form a logo, symbol, indicia or other graphical element.
[0065] It is to be noted that the order of steps 401 and 402 could also be reversed, or
the two steps could also be performed in parallel. The first and second material are
different from one another, such that it is possible to identify an area of overlap
between the two patterns generated by the discrete elements of the first material
and the discrete elements of the second material. In principle the two materials can
be chosen in any suitable or desirable way to achieve this effect, for example can
comprise different dyes and/or pigments, such that the optical characteristics are
different under predetermined conditions and it becomes possible to distinguish the
first pattern from the second pattern, but also to identify the area of overlap of
the two patterns. For example, the two patterns can present respectively different
colored responses when illuminated by electromagnetic radiation having a predetermined
characteristic, and where the area of overlap provides a distinctive response distinguishable
from that of the first or second pattern. A simple example of this is the first pattern
presenting a first color under predetermined, e.g. standardized (in the meaning of
CIE) illumination, the second pattern presenting a second, different color, and the
area of overlap presenting a third color different from the first and second colors.
[0066] Returning to figure 4, the method further comprises a step 403 of obtaining an image
of the third pattern followed by step 404 of applying a predetermined indexing routine
to the obtained third pattern. Finally, in step 405 the generated index is stored
in a data record. The data record is preferably part of a data repository dedicated
to the storage of indices of third data patterns.
[0067] The step of obtaining an image of the third pattern can be done in any suitable or
desirable way. For example, it may comprise illuminating the first and second region
of the substrate with predetermined electromagnetic radiation and then performing
an imaging with an imager suitable for capturing the specific spectral response of
the overlap regions between the first and second pattern. In the above example, where
the overlap regions display a specific color response with a third color, the step
of obtaining an image of the third pattern may comprise performing an appropriate
color filtering operation, which can for example be done by setting specific color
filters in front of the lens of an imager, or can equally well be done by digital
analysis of the image data produced by the imager, in order to appropriately identify
pixels that fall into the desired color range associated with the third pattern.
[0068] An indexing routine within the meaning of the present description is any process
that is able to derive a storable data element from the third pattern obtained in
step 403. For example, the indexing operation can comprise taking the image data of
the third pattern (i.e. a set of pixel coordinates and associated intensity values,
e.g. color intensities on a scale of 0 to 255 in a three dimensional color space)
obtained from an imager as is and then simply arranging this data in a predetermined
format for storage. Preferably, the indexing operation allows a reduction in the amount
of information to be stored, such that is not necessary to store the entire image
information related to the third pattern. This can e.g. be done by using known pattern
recognition techniques. Another technique that can be used is to define a grid of
desired dimensions and to generate bit information in dependence on the presence or
absence of an image element in that grid. Figure 16 is an example of such a grid technique,
where and area of overlap between the first and second substrate regions shown on
the left hand side is analyzed using a N x N grid (N = 6 in the example of the figure),
and the different elements of the grid are assigned a value of 1 if a pattern element
is present and a value of 0 if no pattern element is present, and subsequently the
rows of the grid are simply concatenated to generate a bit string that can act as
an index or signature associated with the pattern being indexed.
[0069] The step of storing the index in the data record can be performed in any suitable
or desirable way, for example by registering the index in a suitable repository, such
as a standard database. Preferably, the stored indexes are associated in the repository
with further identifying elements of the object with which the security element is
associated. More specifically, the security element can be associated with a commercial
good, such as an item for sale (cigarette pack, beverage container, etc.), or can
be associated with a value document, such as a currency note or other form of payment
instrument. The commercial good or value document can itself constitute the substrate
on or in which the security element is provided, or the substrate can be a label for
attachment to the commercial good work value document. The commercial good or value
document may have specific identifying elements, such as serial numbers, a Stock Keeping
Unit (SKU) number or package IDs, which can appropriately be stored in association
with the index of the third pattern.
[0070] Figure 5 shows a further embodiment of the present invention. More specifically,
it shows an example of a method for authenticating a security element and security
arrangement of the above described kind. A situation is considered in which a commercial
good or value document as described above is present. A user, for example an inspector,
desires to determine whether or not the security element is authentic. In a first
step 501 an image of the third pattern is obtained. The method of obtaining the image
of the third pattern can be the same as described above in connection with the method
of figure 4. Furthermore, an indexing routine 502 is applied to the obtained image,
for generating an index of the third pattern. The indexing routine is preferably the
same as that used in the method of figure 4 for generating the index to be stored.
In step 503 a data repository of data records generated for third patterns as described
above in connection with figure 4 is accessed. Then, in step 504 the generated index
is compared with content from the repository, and in step 505 and genuineness decision
is made based on the comparison.
[0071] The individual steps can be performed in any suitable or desirable way. Especially,
the step 503 of accessing the repository of data records and the step 504 of comparing
the generated index with content can be implemented in different ways, depending on
for example how the repository is arranged.
[0072] In a first example, assuming that the repository only contains data records that
comprise indices of third patterns, but does not associate the recorded indices with
identification information of commercial goods or value documents, then the steps
of accessing the repository and comparing the generated index can be implemented by
comparing the generated index with some or all recorded indices, in order to determine
whether the generated index matches one of the recorded indices. If a match is found,
then the genuineness decision of step 505 results in a confirmation of genuineness.
If no match is found, then step 505 results in a finding of "non-genuineness".
[0073] In a second example, assuming that the repository contains data records in which
indices of third patterns are associated with identification information of items
associated with the security elements, then the step 503 of accessing the repository
of indices may comprise using an identification element obtained together with the
third pattern, in order to find a specific record in the repository that is associated
with the identification, in order to retrieve the index stored in association with
that identification in the repository, and compare that retrieved index with the generated
index. The genuineness decision of step 505 can then be such that genuineness is confirmed
if the generated index and the retrieved index match, and that genuineness is denied
in the event that the generated index and the retrieved index do not match.
[0074] Preferably, the method for authentication also takes into account one or both of
the first pattern and the second pattern. In other words, the method is described
in connection with figure 5 can be complemented by the step of obtaining an image
of the first pattern and applying a predetermined indexing routine associated with
the first pattern to the obtained image, for generating a corresponding first index
of that first pattern then the repository of data records can be accessed for comparing
the generated first index with content from that repository, similar to what was described
above with respect to the third pattern. The authentication decision can then also
be based on comparing the generated first index with the content from the repository.
As a complement or alternative to additionally taking the first pattern into account,
the method of figure 5 can also be amended by obtaining an image of the second pattern
and applying a predetermined indexing routine associated with the second pattern to
the obtained image, for generating a second index associated with the second pattern.
Then the repository of data records can be accessed for comparing the generated second
index with content from the repository, in order to base the authentication decision
also on the comparison of the generated second index with content from the repository.
[0075] Fig. 6 shows embodiments of the invention in a system configuration relating to the
generation of an inventive security element and security arrangement, and to the authentication
of such a security element.
[0076] Reference numeral 615 schematically indicates a production or packaging facility,
in which items 614 are prepared. Examples of such items can be the above mentioned
commercial goods or value documents. At the facility 615 a system 610, 611 is provided
for applying patterns to the items 614 by appropriately distributing the respective
discrete elements, and a system 612, 613 is provided for imaging the applied patterns.
For example, 611 may be a combination of an ink jet printer for applying a regular
pattern and a spray nozzle for applying a random ink spatter pattern, under the control
of a control module 610. Device 613 may be a digital camera suitable to capture an
image of the first and second patterns, from which an image of the third pattern associated
with the overlap of the discrete elements of the first and second patterns generated
by system 610, 611 can be obtained, under the illumination with electromagnetic radiation
suitably provided by an illumination element 617, where devices 613 and 617 operate
under control of a control module 612. However, as an alternative or in complement
to being able to capture images of the first and second patterns, device 613 may also
be arranged and controllable in such a way that the image of the third pattern is
captured directly based on the particular response of the third pattern to the irradiated
illumination. Thus, obtaining an image of the third pattern may be done by direct
image capture and/or indirectly by processing images of the first and second patterns.
The control modules 610, 612 can be provided in any suitable or desirable way in the
form of hardware, software or a combination of hardware and software. Preferably,
the control modules are provided as programmable computer devices, where both modules
610 and 612 may be provided separately, but can also be embodied by a single computer
unit.
[0077] The system of Fig. 6 also has a data repository 620, which is arranged such that
it can communicate at least with the control element 612 at the facility 615. The
repository 620 is preferably at a location remote from the facility 615, but facility
615 and repository 620 can also be at the same location. A control module 621 connected
to a storage unit 623 can exchange data with the control module 612, in order to perform
the above described operations of storing an index of a third pattern in a data record.
The connection can be by any suitable or desirable communication method, e.g. wire-bound
or wireless, via dedicated channels or via a general purpose communication network,
such as a telephone network or computer interconnection network. According to an embodiment,
the communication comprises using the Internet.
[0078] Referring to the above described method embodiment of Fig. 4, steps 401, 402 can
be performed by system 610, 611, whereas steps 403, 404 can be performed by system
612, 613, 617, and step 405 can be performed by module 621. However, it is noted that
the images of the first and second patterns captured by device 613 can also be processed
by module 621 at the data repository, in order to obtain the third pattern and subsequently
generate the index for storage.
[0079] The example of Fig. 6 furthermore schematically shows an inspection device 631 designed
for performing steps in a method of authenticating a security element and security
arrangement of the above described kind. The inspection device is preferably a mobile
device, more preferably a portable hand-held device, e.g. an appropriately programmed
mobile telephone. In the example, an item 614 is illuminated by a source 630 of suitable
electromagnetic radiation for identifying the third pattern. Inspection device 631
comprises an imager suitable to capture an image of the first and second patterns,
from which an image of the third pattern associated with the overlap of the discrete
elements of the first and second patterns can be obtained, and/or can be arranged
and controllable in such a way that the image of the third pattern is captured directly
based on the particular response of the third pattern to the irradiated illumination.
[0080] The data repository 620 comprises a communication module 622 that can communicate
appropriately with the inspection device 631. The communication can be by any suitable
or desirable method, e.g. wire-bound or wireless, via dedicated channels or via a
general purpose communication network, such as a telephone network or computer interconnection
network. According to the shown example, the communication involves a wireless connection
using antennas, e.g. through a mobile telephone system.
[0081] Referring to the above described method embodiment of Fig. 5, step 501 can be performed
in inspection device 631, but also can be performed by control module 621 in the data
repository 620. The same applies for step 502. The access step 503 can be performed
by control module 621, while steps 504 and 505 can again be performed in the control
module 621 and/or in the inspection device 631. According to a preferred embodiment,
the inspection device 631 captures under respectively suitable illumination images
of the first and second regions of the substrate, sends this information to the data
repository 620, and all of steps 501 to 505 are performed by control module 621, in
order to give back the authentication result of the authentication decision to the
inspection device 631, which can appropriately notify a user of the inspection device,
e.g. via information on a display of device 631. This embodiment has the advantage
of keeping the structure of the inspection device simple in terms of the authentication
functionality, as most authentication processing is done at the repository 620, and
the security of the authentication processing is enhanced, as none of the components
for performing the processing are accessible outside of the repository.
[0082] The control module 621 can be provided in any suitable or desirable way in the form
of hardware, software or a combination of hardware and software. Preferably, the control
module is provided as one or a set of programmable computer devices, especially as
a server accessible over a network, that comprises a database engine for managing
the data records held in the storage device 623, and with further processing modules
for performing one or more of the processing steps described above in connection with
Figures 4 and 5.
[0083] The present invention can also be embodied as software code parts designed to enable
the above described methods when loaded into and executed in a control system or control
module. The invention can also be embodied as a data carrier holding such software
code parts.
[0084] Now further embodiments and aspects of the present invention will be described.
[0085] The first and second material can be chosen in any suitable or desirable way. For
example, the first material may comprise one or both of a first dye and a first pigment,
and the second material may comprise one or both of a second dye and a second pigment.
Preferably, one or more of the dyes and pigments present in the first and second materials
is luminescent, i.e. displays one or both of fluorescence and phosphorescence. This
provides an added level of safety against counterfeiting, because the authentication
may be based not only on the appearance of the patterns, but also on the specific
spectral characteristics of the luminescent material or materials, i.e. the specific
excitation and emission wavelength ranges that are characteristic of the luminescence
of the material.
[0086] According to a further embodiment, the security element may be provided in such a
way that the discrete elements of at least one of the first and second patterns are
not visually distinguishable from the background, i.e. the substrate. In other words,
the respective pattern is designed to be invisible for a human observer under conventional
conditions, e.g. when looking at the security element under daylight or normal room
white lighting using only the naked eye. This increases the safety against forgery
and counterfeiting, as a counterfeiter cannot readily see the pattern, which only
appears under special conditions, e.g. when illuminated within a particular wavelength
range that leads to a predetermined spectral response, e.g. a luminescent response.
[0087] The determination of visual distinguishability can be done in any suitable or desirable
way. For example, it can be determined that a pattern is not visually distinguishable
when the contrast of the pattern to the background (i.e. the substrate) is below a
predetermined threshold. Contrast can be measured in any suitable or desirable way,
e.g. as the ratio of the luminance difference between pattern and background and the
average luminance of both pattern and background, or as the ratio of the luminance
difference between pattern and background and the luminance of the background. The
contrast can also be defined with respect to color, i.e. as color contrast, e.g. using
the chrominance in place of the luminance, or a contrast measure can be used that
takes both color and luminance into account. The threshold can be set to 5%, preferably
to 2 % and most preferably to 1%. Observation can also be measured in any suitable
or desirable way, e.g. by simple visual inspection by a human observer or also using
any predetermined standard observer. Illumination can be chosen in any suitable or
desirable way, e.g. an incandescent tungsten lamp or similar lighting system, or sunlight
can be used. It is also possible to use any suitable standard illuminant. A standard
illuminant can be chosen in any suitable or desirable way, and e.g. be a CIE Standard
Illuminant A, B, C or D. A standard observer can equally be chosen in any suitable
or desirable way, and e.g. be a CIE 1931 2° Standard Observer or a CIE 1964 10° Standard
Observer.
[0088] In particular, the security element may be provided such that the discrete elements
of one of the first pattern and the second pattern do not provide a contrast with
respect to the substrate that is visually perceivable, and the discrete elements of
the other of the first pattern and the second pattern provide a contrast with respect
to the substrate that is visually perceivable . In other words, one of the first and
second patterns is invisible while the other is visible. This enhances the security,
as it increases the tendency that a counterfeiter will concentrate on the visible
pattern and miss the invisible pattern, which in turn significantly lowers the probability
that a counterfeiter will correctly reproduce the invisible pattern not only in terms
of material and appearance, but also in correct spatial relationship to the visible
pattern. As a consequence, the risk of a counterfeiter reproducing the third pattern
is lowered.
[0089] In other words, such an arrangement comprising a first pattern formed by distributed
discrete elements of a first material that is visible by the naked eye under conventional
lighting (a "perceivable pattern"), and a second pattern formed by distributed discrete
elements of a second material that is not visible by the naked eye under the same
viewing conditions (a "non-perceivable pattern"), provides a combination of an overt
and a covert security feature. If as described in an embodiment of the invention the
non-perceivable pattern is provided on an area overlapping with the area in which
the perceivable pattern is provided, it is possible to create a unique security element
that, in a first aspect, is product-specific and can be checked and authenticated
by inexpensive and easily available image processing equipment for the visible pattern,
and that, in a second aspect, can be verified for its authenticity, production pathway
and origin using more sophisticated technology. For instance, taking the above example
of a turbine blade of a jet engine, a first visible pattern, e.g. a spray pattern
obtained using red ink, may be provided by the manufacturer of the turbine material,
to thereby authenticate the origin and quality of the turbine material. A second (invisible)
pattern, e.g. a spray pattern obtained from a clear, colorless ink containing a fluorescent
dye, which is provided on or overlapping with the region in which the first visible
pattern is provided, may then authenticate the site of manufacture. By the combination
of two patterns, e.g. by the pattern formed by the areas of (partial) overlap of the
discrete elements forming the visible and invisible pattern, respectively, a unique
security element is formed that simultaneously confirms the origin of the material
as well as its proper manufacture, wherein the latter can only be checked by someone
having knowledge of the invisible pattern (and having access to a database containing
information provided by the invisible pattern). The entire production pathway can
then further be verified by someone having access to the database in which information
extractable only from the combined patterns (e.g. the areas of overlap of discrete
elements of different patterns) is stored.
Random spray spatter (patterns)
[0090] According to an embodiment, the security element is provided such that the randomly
distributed discrete elements are obtainable by spraying ink, in order to generate
a random spatter pattern of ink stains. This can be done in any suitable or desirable
way. For example, a suitable spray nozzle configuration, which offers flexibility
and a wide parameter space of operation, in particular for scarce sprays, is a concentric
nebulizer. A simplified working principle of such a device is illustrated in Figure
17, which shows a cross-section diagram of a concentric nebulizer spray head, and
can be described as follows:
- Liquid (e.g. ink) 1702 is fed into a central capillary by a dosing valve;
- Gas (e.g. air) 1701 flows co-axially around the capillary and detaches the liquid meniscus at the tip
end by drag force and/or by Venturi effect (at higher gas flow speed);
- The gas flow nebulizes the liquid;
- The gas flow carries the nebulized droplets downstream.
[0091] An example of a suitable spraying device is the commercially sold Nordson MicroSpray
EFD787-MS spraying device, which may be operated in pulsed mode. The pulse mode allows
for very small volumes of liquid per spray mark, fast paces (up to 30 Hz) and a broad
range of variable nebulization gas flows.
[0092] More details of a microspray system are represented in Figure 18 and Figure 19. It
is composed of the microspray head, the liquid reservoir
1801, a controller
1803 and various pressure reducers, valves and gas pipes. The inlet circuit pressure may
be 6 bars, which is also the actuating pressure of the valve
1804 (
pvalve1805). The reservoir is pressurized with
pres 1806 (typically 1 bar) in order for the liquid to be pushed into the capillary
1902 when the valve is opened.
pres has a minor influence on the spray spatter features, but can be adjusted according
to liquid viscosity, e.g. can be increased as the liquid viscosity increases, in order
to preserve identical liquid volumes per spray puff at different viscosities.
[0093] Other spray parameters of interest may be the valve piston
1901 stroke, which is related to the liquid throughput during the valve opening time (
ΔT), and the nebulizing gas flow (
Qneb). The latter is determined by the nebulizing gas pressure
1807 (
pneb) and the nozzle conductance. The nozzle conductance is directly related to the annular
cross-section between the outer (capillary) needle
1902 diameter and the nozzle
1903 cap hole diameter (see Figure 19). The piston stroke can e.g. be adjusted using a
rotating knob
1802 with a scale. The creation of diffuse spray spatters is preferably done using very
small liquid volumes per puff. Consequently, the valve stroke may be set close to
its minimum value.
[0094] The inner diameter of the dispenser needle may be chosen in accordance with ink viscosity.
For inks of viscosity with typical values of 1 to 10 cp (centipoise - mPa.s), an inner
diameter range of 150 - 200 µm is adequate. This diameter may be increased for higher
viscosity fluids. The needle length may be ½ inch (12.7 mm), which makes the tip slightly
crop out of the nozzle end. The outer needle diameter may be 400 µm and the inner
nozzle diameter may be 0.9 mm, which leaves an annular cross-section of 0.5 mm
2 for the nebulizing gas. This gap may be varied for adjusting the atomization degree
and the spray mist aerodynamics towards the substrate (nebulization and velocity)
for a given gas flow, in order to generate patterns of a desired kind. A needle centering
ring can also be used to enforce coaxial geometry, resulting in an axially-symmetrical
spray spot. The nebulization gas may flow continuously, whereas the liquid may be
pulsed.
[0095] At a gas pressure of 1 PSI (pound per square inch), the nebulization gas flow was
estimated to be 2 to 3 l/min, which corresponds roughly to a local gas velocity at
the nozzle exit of 8-12 m/s. The Reynolds number
(Re) can be estimated by taking into account the hydraulic diameter of the annular nozzle
exit. It may be in the range 800 - 1200, which is characteristic of laminar flows.
The laminar flow results in a coarser nebulization when compared to the standard operation
of a spray device with 10 times more velocity and a
Re number in the turbulent regime.
Table 1
| Parameter |
Symbol |
Unit |
Range |
Preferred value |
| Dispenser needle Gauge |
G |
G |
32 - 23 |
27 (transparent colour code) |
| Inner needle diameter |
D |
µm |
100 - 330 |
200 |
| Outer needle diameter |
D |
µm |
230 - 640 |
400 |
| Valve stroke |
S |
Knob scale |
0-2.5 |
0.1 |
| Inlet/valve pressure |
pvalve |
bar |
5-7 |
6 |
| Reservoir pressure |
Pres |
bar |
0-6 |
1 |
| Nebuliser gas pressure |
Pneb |
PSI (bar) |
0 - 30 (0 - 2) |
3 (0.21) |
| Spray puff time |
ΔT |
ms |
10 - 10'000 |
20 |
| Spray distance |
L |
mm |
30 -100 |
60 |
[0096] Table 1 shows an example of spray working parameters with their ranges and preferred
values for obtaining a suitable spray spatter that produces a random pattern in accordance
with the present invention. The distance
(L) from the spray nozzle exit to the substrate is also a parameter. The larger the distance,
the broader and the more diffuse the spray spatter will be. Since nebulization also
occurs during the flight of the spray mist, a larger distance leads to smaller droplet
mean size. However, under Table 1 conditions, this effect is observable only at a
short spray distance (L < 20 mm), which is inadequate due to the spatial non-uniformity
of the spray spatter and too large droplet mean size. A suitable choice for L is between
30 - 60 mm for a spray spot diameter of 15 - 30 mm.
Substrate
[0097] The security element of the present invention comprises at least two patterns formed
by distributed discrete elements in or on a substrate. The substrate for this purpose
is mainly determined by the end use of the security element. If the security element
shall be used on e.g. a banknote, it is self-evident that then the substrate may be
the banknote substrate made mostly from paper, but in other instances plastics can
be used, or hybrid materials in which plastics and paper are combined.
[0098] It is further self-evident to the skilled person that the substrate and the material
forming the randomly distributed discrete elements thereon or therein may be chosen
such as to be compatible with each other, so that the presence, and in particular
the position and/or size of the discrete elements forming the pattern, can be inspected
for authentication purposes. For example, the randomly distributed discrete elements
may be formed by spraying two different ink compositions to thereby form two patterns,
and in this case the substrate may be a substrate on which an ink can be maintained
and fixed, such as paper or cardboard. In this case, the pattern (i.e. the position
and size of the randomly distributed discrete elements) can be inspected by conventional
image processing equipment, as the pattern is easily visible.
[0099] In one embodiment of the invention, the randomly distributed discrete elements are
present in the substrate. This can for instance be achieved by blending particles,
which represent the discrete particles forming a pattern, into the bulk of the substrate
during the manufacture thereof. For instance, a small amount of two different UV-fluorescent
particles can be included into the pulp for the papermaking process, and patterns
formed by the randomly distributed two kinds of particles can then be observed by
illuminating the substrate including the UV-fluorescent particles with UV light and
measuring the emission positions and/or intensities resulting from the randomly distributed
particles.
[0100] A similar effect can be obtained by mixing different materials (e.g. having different
color) into a translucent or transparent plastic raw material prior to formation of
a plastic substrate for e.g. a credit card. Again, the distribution pattern of the
different particles can be used for authenticating a security element.
[0101] A further possibility of including particles (as an example of discrete elements)
in a substrate is the use of detectable particles of different color in an adhesive
that is used for the formation of clear adhesive labels having translucent or transparent
plastic layer. Herein, two or more patterns can be formed by randomly distributed
particles of two or more kinds, which are present in an adhesive layer that is glued
to a material to be authenticated (e.g. a cardboard box containing high-value shoes).
The adhesive layer is then preferably of such a kind that a part of the adhesive containing
the particles of different kinds remains on the material to be authenticated, e.g.
by using an adhesive having low cohesive properties, or to which some of the material
(e.g. the top paper layer of a cardboard box) remains attached. In such manner, authenticating
labels that can be produced simply can be provided to a manufacturer, and these labels
cannot be detached from the authentic product without destroying the label.
[0102] As already extensively described, the pattern could also be formed by randomly distributed
discrete elements that are provided on the substrate. This can be achieved by providing
the discrete elements forming the pattern in such a way on the substrate that a random
and not predetermined distribution is achieved. Preferably, in this case the random
distribution of the discrete elements is achieved by a random application process,
such as spraying. Spraying allows achieving a random distribution of ink stains on
the substrate, each of the ink stains forming a discrete element that is part of a
pattern of ink stains.
[0103] Of course, it is also possible to use a combination of randomly distributed discrete
elements within the substrate, forming a first pattern, and randomly distributed discrete
elements on the substrate, forming a second pattern.
Discrete Elements forming a Pattern
[0104] The discrete elements forming a pattern are not particularly limited, as long as
they can be detected by available detection technology.
[0105] One preferred example are ink stains formed from two or more of different ink compositions
(formed from droplets of ink that are provided by a random process on the substrate,
such as by spraying), that form distinguishable discrete elements having certain detectable
intrinsic (e.g. color, luminescence, reflection) and extrinsic (position relative
to an orientation mark on the substrate, size, shape) properties. The ink stains may
be circular, but may also have an irregular shape as produced by overlaps between
ink droplets or by substrate irregularities or roughness.
[0106] Alternatively or additionally, the discrete elements may be provided in the substrate.
In principle, any particulate material that is conceivable after its incorporation
can be used for this purpose. Suitable materials include pigments and metal particles.
Pattern Regions
[0107] The area of a pattern, which corresponds to the above described first and second
region of the substrate, is generally defined by the randomly distributed discrete
elements, so that the entire area in which the respective kind of randomly distributed
discrete elements are present is to be considered as the pattern area. In the case
of an inclusion of particles into the substrate by addition of the particles during
the production process, to thereby form randomly distributed discrete elements within
the substrate, e.g. by addition of fluorescent pigments to the papermaking pulp, the
entire substrate forms the pattern area. If the randomly distributed discrete elements
are formed by ink droplets distributed randomly by a spraying process, the entire
area in which the stains are present on the substrate form the respective pattern
area.
[0108] It is immediately evident that the two or more patterns employed in the present invention
therefore define two or more pattern regions, i.e. one for each pattern. The pattern
regions of the two or more patterns overlap at least partially to allow the utilization
of the relative information between the discrete elements of the different patterns
and to allow generating the third pattern that is associated with the overlap of the
discrete elements of the first and second pattern. One example for such a preferred
embodiment is the provision of two patterns formed by spraying two inks, e.g. a yellow
ink and a cyan ink, to thereby form patterns of randomly distributed discrete elements
(ink stains), in the same or overlapping areas. In such a case, typically some of
the stains of a first pattern will overlap with stains of a second pattern, and the
area of overlap of the randomly distributed discrete elements forms a third pattern
(e.g. in green due to the combination of cyan and yellow by color subtraction) that
can be used for authentication purposes.
[0109] In a preferred embodiment, the pattern areas defined by the randomly distributed
discrete elements of the two or more patterns overlap by at least 50%, and more preferably
by at least 75%. In a further preferred embodiment, the pattern areas (e.g. the substrate
regions 101, 102) defined by all randomly distributed discrete elements of all patterns
are identical.
[0110] The pattern area of one of the two or more patterns, of two of the two or more patterns
or of all of the two or more patterns (in case of more than two patterns) may be restricted
by an outer marking, such as a line. It may be preferable to provide such a marking
or line in order to provide a basis for the assessment of the information encoded
by the random distribution (e.g. the position of a specific discrete element relative
to a crossing or corner of the line restricting the pattern area). The outer marking
may take any form, e.g. the form of a simple rectangular box, but may also take the
form of indicia, logo, numbers or characters.
[0111] The size of the pattern area is not particularly limited, but is preferably large
enough to allow proper detection. As one example, the pattern area is 0.1 cm
2 or greater.
Overlap of patterns
[0112] An example of two patterns each having discrete elements formed from different materials
is for instance a first spray pattern formed from discrete elements of a first colour
(e.g. yellow), and a second spray pattern formed from a different colour (e.g. cyan).
[0113] An advantage of the present invention, which uses two or more of patterns made from
different materials, is that a further level of information can be added, not only
including the size and position of the discrete elements (stains) of one pattern,
but also including information on the kind of material (e.g. the color of the discrete
element). An additional level of information can then be provided not only by the
kind of material forming the discrete elements of one of the two or more patterns,
but also by the nature of the surroundings.
[0114] In the concept of the invention as applied to the above example, it is not only possible
to attribute a first set of information to the position, size and/or shape of the
yellow stains (discrete elements) forming a first one of the at least two patterns,
but to attribute a second set of information to the position, size and/or shape of
the cyan stains (discrete elements) forming a second one of the two or more patterns.
By virtue of the third pattern, a third set of information is attributed to the relative
arrangement between discrete elements of the first and second pattern. Also, the relative
amount of the discrete elements may provide even further information that could be
analyzed e.g. by the spectral response under certain viewing conditions, representing
certain relative amounts of cyan and yellow stains.
[0115] Hence, the presence of a second pattern formed by discrete elements and the definition
of a security element by the third pattern not merely adds information provided by
the second pattern per se, but also allows defining certain relationships between
the elements of the first and second pattern, which also represents encodable information.
The presence of a second pattern formed by distributed discrete elements made from
a different material therefore not only doubles the amount of encodable information,
but multiplies it. This provides a higher level of security against counterfeiting,
and is extremely difficult to reproduce.
[0116] In an embodiment of the invention, discrete elements of different patterns are formed
in different colors, so that in the area of overlap a color impression different from
the color impression provided by each of the discrete elements per se is obtained.
In such a case, the areas of overlap and their arrangement can be detected and used
for authentication purposes in the form of the third pattern. As one example, if some
of the stains (discrete elements) of a spray pattern of cyan ink and some of the stains
(discrete elements) of a spray pattern of yellow ink overlap, the color impression
in the areas of overlap will neither be cyan nor yellow, but will be green. Correspondingly,
the distribution, position, color and/or shape of the areas of overlap provide a third
pattern of randomly distributed discrete elements.
EMBODIMENTS UTILIZING AN ENERGY TRANSFER (CASCADE EFFECT)
[0117] As outlined above, in a preferred embodiment of the present invention, the materials
forming the randomly distributed discrete elements of different patterns are able
to interact with each other in a specific way in the area of overlap of discrete elements
of different patterns. One example given above is the combination of cyan and yellow
discrete elements formed from an ink spray pattern, which provides a green impression
in the area of overlap.
[0118] According to the invention, the materials used in the preparation of different patterns
interact with each other in a specific way, namely by providing a so-called "cascade
effect" of two dyes present in the formulations for producing the respective patterns.
That is, the present invention provides a security element wherein the discrete elements
forming a first one of the two or more patterns are formed from a first ink INK1 comprising
a first dye or pigment DYE1, and the discrete elements forming a second one of the
two or more patterns are formed from another INK2 comprising a second dye or pigment
DYE2, wherein a part of the discrete elements forming the first pattern overlap with
a part of the discrete elements forming the second pattern, and wherein dye or pigment
DYE1 contained in INK1 is a fluorescent dye or pigment, which upon excitation by electromagnetic
radiation falling within an excitation wavelength range λ1a of the first fluorescent
dye or pigment DYE1 is capable of emitting electromagnetic radiation in at least one
first emission wavelength range λ1e, and the second dye or pigment DYE2 contained
in INK2 is a fluorescent dye or pigment, which upon excitation by electromagnetic
radiation falling within an excitation wavelength range λ2a of the second fluorescent
dye DYE2 is capable of emitting electromagnetic radiation in at least one second emission
wavelength range λ2e, wherein said first emission wavelength range λ1e of the first
fluorescent dye or pigment DYE1 contained in INK1 overlaps with the excitation wavelength
range λ2a of the second fluorescent dye or pigment DYE2 contained in INK2, so that
upon irradiation with electromagnetic radiation within the excitation wavelength range
λ1a of DYE1, DYE2 is excited, in the area of overlap of the discrete elements, to
emit electromagnetic radiation in the emission wavelength range λ2e. This embodiment
will in the following also be referred to as the "cascade embodiment". The respective
wavelength ranges are illustrated in Figure 10.
[0119] Regarding the cascade effect, we also refer to co-pending application
EP 14 184 924.0, which was filed by the same applicant.
[0120] By a suitable selection of the dyes and/or pigments in INK1 and INK2, it is possible
to obtain emission within the wavelength range λ2e by irradiating the security element
with exciting radiation in the wavelength range λ1a in the regions of overlap of the
discrete elements forming the respective patterns. The cascade effect is then only
observed in the area of overlap. If the fluorescent dyes are chosen such that λ2e
is outside the visible range, a covert security feature is formed by the pattern formed
by the overlapping areas of the discrete elements forming the respective patterns.
Additionally or alternatively, λ1a is chosen to be in the UV region, such as around
355 or 378 µm, so that the cascade effect is not observed under essentially UV-free
viewing conditions, such as under irradiation with an incandescent lamp. In this case,
the cascade effect is observed more clearly if a UV lamp emitting at the respective
wavelength is used for illumination/detecting the third pattern.
[0121] That is, by employing a combination of fluorescent dyes wherein the emission of the
first fluorescent dye is capable of exciting the second fluorescent dye, one can obtain
the emission of the second fluorescent dye by merely exciting the first fluorescent
dye, e.g. by irradiating the ink with electromagnetic radiation in a wavelength region
at which excitation of the first fluorescent dye occurs, and it is not necessary to
irradiate the ink with radiation capable of exciting the second fluorescent dye in
order to obtain the second dye's emission. If the proportion of the second dye or
pigment DYE2 (also referred to as acceptor), or the application amount of the respective
INK2 is significantly larger than that of the first dye or pigment DYE 1 (also referred
to as donor), the spectrum observed in the overlap regions upon excitation of the
donor will be dominated by the emission of the acceptor, with minor portions stemming
from the emission of the donor that is not utilized for excitation of the acceptor.
[0122] In this case, it is a decisive advantage of the invention that a counterfeiter will
be unable to detect, by analysis of the emission spectra of the ink mark, that two
dyes or inks are present in the area of overlap, because he will mainly (or exclusively)
measure the emission of the second dye (acceptor), since the emission of the donor
mainly excites the acceptor. The emission of the donor may thus not be detectable
at all in the area of overlap, or may be rather weak, depending on the relative amounts
of donor and acceptor.
[0123] A counterfeiter's analysis of the patterns, and in particular a clear assessment
regarding the areas of overlap of the discrete elements formed from INK1 and INK2
is rendered more difficult by the fact that relatively small amounts of donor dye,
e.g. 5 to 10% based on the total weight of acceptor and donor dye in the respective
areas of overlap, may in some cases suffice to produce the cascade effect. Such a
ratio can be adjusted by the concentrations of the dyes in the inks, for instance.
[0124] One further unique feature of the invention is that the authenticating method can
be used to observe the (possibly averaged) emission intensity response as tailored
by the cascade effect (i.e. depending on the choice and ratios of donor and acceptor)
but also by the specific illumination (i.e. the specific shape of the excitation spectrum,
e.g. intensity of radiation as a function of wavelength, which can for example be
adjusted by varying the specific illumination intensities of sources of radiation
that emit different wavelengths) used to generate the response, and thus the signature
generated in the areas of overlap of the discrete elements.
[0125] Therefore, the response, i.e. the observed emission, will be very sensitive to precise
compositions of the inks, which makes it more difficult to reproduce, leading to a
spectrum signature that provides a high reliability of authentication. Indeed, the
emission spectrum, observed over a spectral range that is not necessarily large but
identical for all the acquired images does not substantially vary in shape, but varies
in intensity level as the ratio of donor and acceptor is modified. As a consequence,
a counterfeiter is forced to reproduce very precisely the compositions of the inks
in order to generate an acceptable signature in response to the specific illumination.
Moreover, the inks' intensity responses strongly depend on the dry ink material layer
thickness, which requires to reproduce precisely the overall ink formulation (not
only the dye or pigment concentration) and to apply the inks with nearly the same
process. This represents an additional hurdle for a counterfeiter.
[0126] It is one particular advantage of the present invention that the spectral response
(also referred to as spectral signature) not only depends on the energy transfer from
the donor to the acceptor (the cascade effect), but also on the excitation wavelengths
used. Using the same fluorescent dyes producing the same cascade effect, but with
different excitation wavelengths, will change the observed spectral response (signature).
The preparation of very precise mixture compositions hence can produce distinctive
signatures that should only be reproducible at the same excitations and for the same
mixture ratios, as obtained by certain ink concentrations and coating thicknesses
of the inks. As a counterfeiter generally has no knowledge about the excitation wavelength(s)
used for authenticating, it is of utmost difficulty for him to mimic the signature
obtained by using the ink of the present invention by using a combination of dyes
that lead to a similar spectral response when excited within a broad wavelength range.
Rather, the counterfeiter would need to know which exact excitation conditions (such
as a combination of different excitation wavelengths in a certain intensity relationship)
are used for authentication, and would then need to adapt the spectral response so
as to mimic the signature of the ink of the present invention.
[0127] Therefore, the excitation and emission patterns can be made complex enough to carry
a significant and discriminant ink feature, and therefore, one can exploit these properties
by imaging the areas of overlap of the discrete elements formed from INK1 and INK2
with excitation light at different wavelengths.
[0128] Therefore, a problem solved by such an embodiment of this invention is that it allows
for a more robust authentication of luminescent marks than methods that simply image
a broader spectral range. Embodiments of the invention, which use an excitation spectrum
approach, can achieve these aims in a more compact and affordable way than other techniques
based on spectral emission analysis. In addition, embodiments of the present invention
also address the desire for tailoring the ink signature to certain requirements of
a user of the ink by using a combination of two specific fluorescent dyes.
[0129] These and other advantages of embodiments of the present invention over the prior
art can be summarized as follows:
- The cascade effect described above allows generating a unique excitation and/or emission
spectrum signature for an enhanced discrimination. The tailoring of the inks allows
quickly changing ink properties, addressing the case of an ink that is copied and
that requires a quick action to cure the problem. Ideally, the authenticating device
hardware does not need to be changed, as the change in ink properties can be taken
into account by simply updating the authentication criterion, i.e. whether a measured
response from a mark under examination shows the behavior expected of an authentic
make. This updating of the authentication of the criterion can be achieved by a simple
software update in a programmable authentication device.
- The emission spectra of commercial markers are often publicly available. Therefore,
counterfeiters may combine several dyes and/or pigments to mimic a certain signature.
However, excitation spectra are less straightforward to obtain, and therefore it poses
a more complex problem for a counterfeiter to select and combine known markers such
that a spectrum deemed authentic is obtained for a selected (predetermined) excitation
radiation.
- Embodiments of the present invention are compatible with laser illumination (quasi-monochromatic
excitation) for more detailed signatures and higher discrimination capability.
- The authentication method of an embodiment of the invention is also more suitable
for handheld devices (e.g. smartphone-based systems), where movable parts or bulky
components, which would be required for spectral emission analysis, represent a drawback.
- In accordance with an embodiment of the invention, it is more economical and technically
simpler to do a form of excitation analysis (e.g. with multi LED illumination or with
multi-laser illumination) with an imaging device than complex emission analysis by
hyperspectral imaging techniques (Fabry-Perot method; Custom Bayer; AOTF; tunable
band pass, etc.).
- Embodiments of the present invention are also easier to implement in authentication
systems that detect luminescence, since a modification of sensors to produce multispectral
emission imaging is not required.
- In addition, the spectral signature to be regarded as genuine may rely on relative
calculations (e.g. intensity ratios or correlations for different excitations). This
approach prevents problems posed by differing ink concentration or by ink aging.
- The authentication method of the invention can also be advantageously applied not
only to dot matrix codes or others, but also on some fine printed designs like logos
or images where security ink is printed on small areas hardly accessible with other
methods.
[0130] In accordance with the invention, the spectral signature of a genuine ink mark formed
by the areas of overlap of the patterns formed from INK1 and INK2 depends on the excitation
wavelength(s) used and the ink properties. Therefore, for a counterfeiter to forge
the ink, he needs not only to know the ink spectral emission properties as a function
of the excitation wavelength but also the excitation wavelengths used to generate
the signature, which requires reverse engineering the device used. This also means
that a given ink can have different signatures if different excitation wavelengths
are used. Further, the security of the solution can be enhanced by prescribing the
use of several excitation wavelengths for authentication, thus increasing the complexity
of the signature to match.
[0131] The cascade effect is illustrated in Figure 10, where λ
1a is an excitation range of INK1, λ
1e is an emission range of INK1, λ
2a is an excitation range of INK2, λ
2e is an emission range of INK2, λ
1a-max is a peak maximum of excitation of INK1, λ
1e-max is a peak maximum of emission of INK1, λ
2a-max is a peak maximum of excitation of INK2, and λ
2e-max is a peak maximum of emission of INK2. As shown in Figure 1, the degree of overlap
of the light emitted by the donor within the excitation wavelength range of the acceptor
(and the intensity) are chosen to be sufficient to excite the acceptor to emit electromagnetic
radiation. Therefore, the term "said first emission wavelength range λ1e of the first
fluorescent dye or pigment DYE1 overlaps with the excitation wavelength range λ2a
of the second fluorescent dye or pigment DYE2" denotes that there is an overlap in
the respective spectral ranges in the emission wavelength range of the fluorescent
dye or pigment DYE1 present in INK1 (donor) and the excitation wavelength range of
the fluorescent dye or pigment DYE2 present in INK2 (acceptor). Taking the example
of a fluorescent dye DYE 1 (donor) having a first excitation wavelength range λ1e
of 440 to 460 nm, a spectral overlap is given if an excitation wavelength range of
the fluorescent dye DYE2 (acceptor) in INK2, i.e. λ2a, includes the values of 440
nm or 460 nm, respectively.
[0132] As one example, an overlap is given if λ1e of the donor is from 440 to 460 nm, and
λ2a of the acceptor is from 450 to 470 nm. A spectral overlap in the sense of the
present invention is, however, not given if merely the end values of the ranges are
the same, such as in the case of λ1e = 440 to 460 nm and λ2a = 460 to 480 nm.
[0133] According to the above definition, a small overlap in the respective ranges λ1e and
λ2a suffices, as also then a cascade effect in the sense of the present invention
occurs. The occurrence of the cascade effect is, however, the more pronounced the
more there is a degree of overlap between an emission wavelength range λ1e of the
donor and an excitation wavelength range λ2a of the acceptor. In preferred embodiments
of the present invention, the "wavelength range" can be taken in a narrower fashion,
in order to ensure a stronger degree of overlap between λ1e and λ2a. Accordingly,
the term "wavelength range" preferably denotes the span of wavelength values , in
a normalized and background-subtracted emission or excitation spectrum, up to and
including the wavelengths where the line of the normalized and background-subtracted
peak falls to a value of n % (0 < x ≤ 100) of the peak value at the wavelength λmax,
e.g. 10%, more preferably 25%, further more preferably 50% of the peak value at the
wavelength λmax. Due to such a narrower "wavelength range", which includes only spectrum
values larger than n % (such as 10%, 25% or 50%) of the spectrum amplitude at the
maximum, the requirement of an overlap between the (narrower) wavelength ranges leads
to a greater overlap between the entire emission spectrum of the donor and the entire
excitation spectrum of the acceptor.
[0134] Considerations for choice of materials as donor and acceptor to obtain a sufficient
"spectral energy transfer" allowing a cascade effect can also be expressed as follows.
The spectral energy transfer ratio
SE of the cascade effect can be defined as the percentage of area under the normalized
(i.e. divided by the maximum spectral amplitude) excitation spectrum of the acceptor
A2(λ) that also falls under the normalized emission spectrum of the donor E1(λ).
[0135] According to a preferred example, the excitation spectral range of the acceptor fully
falls within the emission spectral range of the donor (CASE1, see Fig. 24a). In other
words, 100% of the excitation spectrum of the acceptor is included/comprised inside
the emission spectrum of the donor, and the above defined spectral energy transfer
ratio is 100%. Note that the donor is contained in MAT1 and that the acceptor is contained
in MAT2.
[0136] Even more preferably, compared to the situation represented in Figure 24a, the emission
spectrum of the donor could exactly match the excitation spectrum of the acceptor
so that the whole emission energy of the donor can potentially be transferred to the
acceptor. However this situation can rarely be achieved because only a few combinations
of materials (pigments and dyes) can satisfy it.
[0137] However, other choices of materials are also possible. For example CASE 2 in Figure
24b where a fraction of, but not the entire, excitation spectrum of the acceptor falls
within the emission spectrum of the donor. In this example, the dashed area under
A2(λ) that also falls under E1(λ) represents 50% of the total area under A2(λ), such
that SE is 50%. Preferably, SE should be larger than 50%, and more preferably larger
than 70%.
[0138] In CASE 2 according to Fig. 24b, a fraction of the acceptors cannot be excited by
the donor because no or little photons are emitted by the donor in a part of the excitation
wavelength range of the acceptor. In addition, a fraction of the emission of the donor
cannot be used to excite the acceptor because it falls within wavelengths that are
outside of the excitation spectrum of the acceptor.
[0139] An alternative case (CASE 3 in Figure 24c) can be envisioned where also 50% of the
area under A2(λ) is overlapped by the emission spectrum E1(λ) of the donor, but the
whole emission spectrum E1(λ) is enclosed within A2(λ). In this case, all the emitted
energy by the donor can potentially be transferred to the acceptor.
[0140] There are, however, other cases possible, such as CASE 4 illustrated as an example
in Figure 24d. Here, although the excitation spectral range of the acceptor is completely
overlapped by the donor emission spectrum, the spectral energy transfer ratio would
be rather low because the emission spectrum intensity is very low in the region of
overlap.
[0141] As a consequence, and as described above, the requirement of overlapping spectral
ranges can be chosen such that both spectral ranges may only include spectrum values
larger than n % of the spectrum amplitude at the maximum (e.g. 50 % in Fig. 24d).
Then a condition for the spectral energy transfer ratio to be sufficient could be
expressed in terms of the ratio of wavelength range where an overlap occurs, to the
excitation spectral range (as defined above) of the acceptor.
Preferably this ratio is 50 % or more, more preferably 70% or more and most preferably.
100%.
[0142] Figure 24e shows CASE 2b which is an alternative representation of CASE 2 of Figure
24b (same spectra), but using the criterion defined in terms of wavelength ranges
for intensity exceeding 50 % of maximum. In this example, the ratio of spectral range
overlap to the excitation spectral range of the acceptor is about 50 %, which is the
same as the criterion using the area of Figure 24b.
[0143] Note that in Figure 24, MAT1 and MAT2 denote INK1 and INK2, respectively, wherein
the excitation and emission spectra are determined as described later.
[0144] It is preferred that the first fluorescent dye DYE1 displays an excitation peak in
its excitation spectrum at a wavelength (λ1a-max) that is shorter than the wavelength
(λ2a-max) at which the second fluorescent dye DYE2 displays an excitation peak in
its excitation spectrum, i.e. that λ1a-max (nm) < λ2a-max (nm).
[0145] It is also preferred, in this and other embodiments of the invention, that the first
fluorescent dye DYE1 displays a maximum emission in its emission spectrum at a wavelength
(λ1e-max) that is shorter than the wavelength (λ2e-max) at which the second fluorescent
dye displays a maximum emission in its excitation spectrum, i.e. that λ1e-max (nm)
< λ2e-max (nm) .
[0146] It is further preferred that λ1a-max < λ1e-max < λ2a-max < λ2e-max, as illustrated
in Figure 10. This is however not mandatory, as an overlap between λ1e and λ2a can
also be realized if λ1e-max > λ2a-max. Accordingly, in one embodiment of the present
invention λ1a-max < λ2a-max < λ1e-max < λ2e-max
[0147] Typically, the emission peak wavelengths of the first and second dyes are located
at longer wavelengths than their respective excitation peak wavelengths, i.e. λ2a-max
< λ2e-max and λ1a-max < λ1e-max. In this case, the emission occurs at longer wavelengths
(at lower energy) as compared to the respective excitation. It is however also possible
to use, as a first fluorescent dye (donor), so-called anti-Stokes fluorescent dye
in the present invention, where the emission occurs at shorter wavelengths as compared
to the respective excitation, i.e. λ1a-max > λ1e-max. In such an embodiment, λ2a-max
may be at shorter or longer wavelengths as compared to λ1e-max.
[0148] The difference between the two excitation peaks of the first (donor) and second (acceptor)
fluorescent dye, respectively, i.e. (λ2a-max) - (λ1a-max), is for instance at least
5 nm, e.g. 5 to 500nm, 10 to 200nm, 20 to 80nm, 30 to 70nm, and preferably 50 to 200nm.
A difference of at least 20 nm is preferred in order to avoid excitation of the acceptor
dye by the irradiation of the electromagnetic radiation that is intended to excite
donor dye in an authentication method.
[0149] The absolute difference between the emission peak λ1e-max of the donor dye and the
excitation peak of the acceptor dye λ2a-max, i.e. ABS((λ2a-max) - (λ1e-max)) is for
instance at most 20 nm. A smaller difference is preferable, since then a greater overlap
between λ2a and λ1e can be ensured.
[0150] Due to possible interactions between the dyes, a potential overlap in the respective
peaks and ranges, and potential consequent difficulties in the spectral analysis,
the measurements are performed separately for each dye.
[0151] The wavelength at which a dye displays a peak in the excitation spectrum (λa-max)
or emission spectrum (λe-max), and the respective excitation and emission wavelength
ranges are measured as follows.
[0152] Notably, in the present invention all measurements are performed at room temperature
(20 °C), and consequently the peak wavelengths λ1a-max, λ1e-max, λ2a-max, and λ2e-max
as well as the respective ranges λ1a, λ1e, λ2a, and λ2e are those measured at room
temperature according to the following procedure:
First of all, a blank is prepared, which is ensured to be formulated such as not to
interfere with the fluorescence of the donor and acceptor dyes, both chemically and
optically. A composition that was found to serve this purpose well is composed of
87 wt.-% Methylethylketone, 10.3 wt.-% of a hydroxyl-containing copolymer made from
84 wt.-% vinyl chloride and 16 wt.-% of acrylic acid ester (commercially available
from Wacker Chemie under the tradename VINNOL E15/40 A) and 2% of a terpolymer made
from 84 wt.-% vinyl chloride, 15 wt.-% vinyl acetate and. 1 wt.-% dicarboxylic acid
(commercially available from Wacker Chemie under the tradename VINNOL E15/45 M). While
this system is preferably used for the present invention, also other systems can be
employed as long as it is ensured that there is no or very little interference with
the fluorescence of the donor and acceptor dyes, both chemically and optically.
[0153] Then, two separate inks INK1 and INK2 are prepared by dissolving 1.23 wt.-% of the
respective dye in the above blank. These inks are used for determining the wavelength
peaks and the wavelength ranges for both emission and excitation, separately for each
dye/ink.
[0154] Samples having 12 µm wet film deposit thickness are then prepared, using e.g. a K
Control Coater from RK Print Coat Instruments, for all mixtures, i.e. the two inks
and the blank using a coating bar, e.g. the HC2 coating bar, on a suitable white substrate
(e.g. the white part of LENETA N2C-2 substrates), followed by drying at room temperature.
Then, all drawdown samples are measured in emission and excitation mode using a commercial
Horiba Fluorolog III (FL-22) as further described below.
Horiba Fluorolog III measurement conditions:
[0155] The instrument used to perform emission and excitation spectra measurement is a commercial
twice double monochromator equipped with a continuous Xe arc lamp as illumination
source and a Hamamatsu R928P photomultiplier tube operated in photon counting mode
as detector. The flat sample is positioned so that its normal direction is at an angle
of 30 degrees with respect to the irradiation optical axis. The Fluorolog-III type
of light collection method used is "Front Face". In this collection mode, the emission
collection is performed at an angle of 22.5 degrees with respect to the irradiation
beam. By using this collection method and setup, it is ensured that collecting direct
specular reflection from the sample is avoided. Both excitation and emission monochromators
are double monochromators fitted with 1200 grid/mm holographic gratings blazed at
500 nm.
[0156] For excitation spectrum measurement, as shown for instance in the curves on the left
of both plots of Fig. 11, the following procedure is adopted: the emission monochromator
is set at a given wavelength (the one where the emission is to be measured, for example
530 nm in Figure 11, left graph) and the excitation monochromator is scanned at 1
nm increment, over the wavelength range where the excitation spectrum is to be measured
(e.g. 400 to 510 nm). At each excitation wavelength increment, a measurement of the
emission signal is recorded by the detector using a 100 ms integration time. As known
to the skilled person, since the irradiation source is not spectrally flat, a suitable
irradiation correction is applied onto the measured signal at every wavelength using
an appropriate spectral calibration. A spectral correction of the detector sensitivity
is also applied. The spectrally corrected excitation spectrum can hence be reconstructed.
[0157] For emission spectrum measurement, the excitation monochromator is set to the desired
excitation wavelength (e.g. at 480 nm for the left curve of the left plot of Figure
11) and the emission monochromator is scanned over the desired emission spectral range
(500 to 800 nm for the right curve of the left graph of Figure 11, for example) at
1 nm increment while recording the detector signal at each wavelength with a 100 ms
integration time. The emission spectrum is then constructed from all recorded data
points after having applied the suitable spectral sensitivity corrections of the instrument.
[0158] The spectral calibration of the Fluorolog III excitation channel is performed using
a procedure that is commonly applied by persons skilled in the art: the spectral irradiance
is measured using a calibrated detector (e.g. a reference photodiode) positioned at
the location of the sample. This is performed for all wavelengths by scanning the
excitation monochromators. This reference detector has a known spectral response (sensitivity
as a function of the wavelength of radiation impinging on it) previously determined
by measuring an irradiation standard (e.g. a calibrated tungsten ribbon lamp) in a
laboratory. An excitation spectral calibration curve is then calculated by dividing
the real spectral sensitivity of the used reference detector by the measured spectral
irradiance. This calibration curve can then be used to correct the spectral response
to excitation of subsequent measurements by simple multiplication.
[0159] The spectral sensitivity calibration of the emission measurement channel of the Fluorolog
III is performed in an analogue way by using a spectral irradiance standard (e.g.
a tungsten ribbon lamp, whose spectral irradiance has been determined in a laboratory).
This lamp is disposed at the location of the sample and spectral emission is recorded
by the Fluorolog III detector during the scanning of the emission monochromators.
An emission spectral sensitivity curve is obtained by dividing the spectral irradiance
curve of the standard irradiance source by the measured spectral curve. Subsequent
measurements are then corrected by multiplication by the spectral emission calibration
curve.
[0160] These calibration procedures are repeated regularly to ensure correction of any instrument
drift or detector/Xe lamp ageing. The overall spectral resolution of the instrument
for both emission and excitation measurements is 0.54 nm FWHM (Full Width at Half
Maximum), for the slits configuration used in the measurement conditions described
above.
[0161] The same above procedure is applied for the different sample measurements; only the
spectral ranges for the excitation and emission spectrum measurements, along with
the excitation and emission fixed wavelengths, may differ depending on the dye compositions
of the samples.
[0162] As derivable from the above, since the measurements shall serve to evaluate the spectral
properties of the final ink print, the donor or the acceptor dye is dissolved in a
blank composition at a concentration of 1.23 wt.-%. Then, emission and excitation
spectra are recorded separately for each ink under the same conditions as for the
blank. For each ink, the background is subtracted and the spectrum optionally normalized
(with the highest peak having an intensity of 1.0), and the peak wavelength(s) λmax
and the emission and excitation wavelength ranges λ1a, λ1e ,λ2a and λ2e are determined
by measuring the points where the spectrum returns to baseline (or to 10, 25 or 50%
above baseline, depending on the definition of the term "wavelength range" as discussed
above).
[0163] These measurements thus provide the wavelength ranges λ1a, λ1e ,λ2a and λ2e and the
respective wavelengths of the peaks λ1a-max, λ1e-max ,λ2a-max and λ2e-max: these are
then used to determine whether or not the requirements of the present invention are
satisfied. These measurements can also be used to identify suitable dyes as acceptor
and donor dyes for the purposes of this embodiment of the present invention.
[0164] In the above explanations, it was assumed that each dye exhibits only one excitation
peak (λ1a-max, λ2a-max) and one emission peak (λ1e-max, λ2e-max), and only one corresponding
excitation wavelength range (λ1a, λ2a) and one emission wavelength range (λ1e, λ2e).
While this is true for many dyes, a considerable number of dyes show multiple excitation
peaks and multiple emission peaks (see Figure 11). In such cases, each peak in the
optionally normalized spectrum reaching an intensity of 50% or more (preferably 75%
or more) may serve as emission peak (λ1e, λ2e) or absorption peak (λ1a, λ2a) for the
purposes of the present invention, so that there may be multiple λ1e and λ1a, or multiple
λ2e and λ2a.
[0165] The explanations above then apply to each of the peaks and wavelength ranges. For
instance, it goes without saying that it is sufficient that there is an overlap between
any λ1e and any λ2a, so that energy is transferred from the donor to the acceptor.
[0166] When the excitation or emission spectrum of a dye contemplated for use in the present
invention shows several overlapping peaks, the peaks and wavelength ranges are obtained
by fitting the obtained spectrum using a suitable software (least square method),
such as for instance OCTAVE. Herein, a spectrum of overlapping peaks can be satisfactorily
(Goodness of Fit <0.1) simulated by assuming an overlap of two (or rarely three) peaks,
and the simulated values are taken for the identification of the peak wavelengths
and for the identification of the wavelength ranges.
DYES
[0167] Generally speaking, both the first dye DYE1 and second dye DYE2 used in the respective
inks INK1 and INK2 employed in the cascade embodiment preferably show excitation bands
and emission bands in the range of 40 to 2400nm, in particular 300 to 1100nm. Preferably,
the donor dye DYE1 shows emission bands, in particular the maximum emission, in the
UV range or visible range (in particular 300 to 700nm), and the acceptor dye DYE2
shows excitation bands (to be excited by the donor), in particular the maximum excitation,
in the visible or IR range (in particular 400 to 1100nm). More specifically, the donor
dye shows preferably emission band(s) matching acceptor dye excitation band(s) in
the range 250-900 nm
[0168] Fluorescent dyes useful for preparing the inks according to the cascade embodiment
and for implementing the authentication method, can be suitably selected from commercially
available dyes. They can for instance be selected from the following substance classes:
Cyanines (polymethines) and the related cyanine-type chromophors, quinones and the
related quinone-type chromophors, porphines, phtalocyanines and the related macrocyclic
chromophors as well as polycyclic aromatic chromophors.
[0169] Cyanine (polymethine) dyes are known in the art and used as photographic sensitizers
(
D.M. Sturmer, The Chemistry of Heterocyclic Compounds, Vol 30, John Wiley, New York,
1977, pp 441-587; Eastman Kodak). In a more recent application, stable representatives of this compound
class, selected from the coumarins and rhodamines, were also used as laser dyes (
J.B. Marling, J.H. Hawley, E.M. Liston, W.B. Grant, Applied Optics, 13(10), 2317 (1974)). Known fluorescent Rhodamine dyes include e.g. Rhodamine 123, Rhodamine 6G, Sulforhodamine
101, or Sulforhodamine B.
[0170] Phthalocyanines and related dyes are the "industrial variant" of porphines and include
a greater number of well-known fluorescent dyes. They generally absorb at the long
wavelength end of the visible spectrum. The class of phthalocyanines at large comprises
as well the higher-conjugated analogs, such as the naphthalocyanines, which absorb
farther in the IR, as well as the heterosubstituted analogs of phtalocyanines; the
common point defining this compound class is that all of its members are derived from
aromatic ortho-dicarboxylic acids or from their derivatives.
[0171] Quinone dyes are known in the art and used for textile and related dying applications
(e.g. indigoid dyes, anthraquinone dyes, etc.). Electronegative groups or atoms along
the quinone skeleton can be present to enhance the intensity of the absorption band,
or to shift it to longer wavelengths.
[0172] Fluorescent aromatic polycyclic dyes include a rigid, planar molecular structure
(similar to the graphite lattice) which may carry substituents. Typically the planar
molecular structure comprises at least two fused aromatic benzene rings (e.g. 2 to
6 rings). In one of the fused aromatic rings, e.g. the central ring of three fused
six-membered aromatic rings, one or two carbon atoms may be replaced by C=O, O and/or
N. Fluorescent members of this class of dyes and pigments can be selected e.g. from
perylenes (e.g. Lumogen F Yellow 083, Lumogen F Orange 240, Lumogen F Red 300, all
available from BASF AG, Germany), naphtalimides (e.g. Lumogen F Violet 570, available
from BASF AG, Germany) quinacridones, acridines (e.g. Acridine orange, Acridine yellow),
oxazines, dioxazines, or fluorones (e.g. Indian Yellow) are examples of such dyes.
[0173] A suitable pair of donor and acceptor dyes can be properly selected from these and
other known fluorescent dyes based on their spectral properties, which, as a rule,
are published by the manufacturer and can be easily measured, as explained above.
However, the excitation effect on the dried printed ink composition is important,
so that published data generally should be validated by measuring the absorption and
emission spectrum in accordance with the method described above for a printed ink
on the final substrate. Indeed, published data may relate to solutions of the dyes
in a particular solvent (e.g. CH
2Cl
2) wherein the spectral properties may be different from the printed ink, e.g. due
to interaction with the substrate.
[0174] Even if only the excitation and emission maxima are available (before complete absorption
and emission spectra have been measured) an evaluation will be possible to what extent
the emission spectrum of the donor is likely to overlap with the excitation spectrum
of the acceptor dye thereby allowing a screening of suitable candidates.
[0175] The description above has been provided for patterns formed from two printing inks,
containing one donor dye or pigment (DYE1) or one acceptor dye or pigment (DYE2).
However, also more than one (e.g. two or three) donor dyes or pigments may be used
in a single ink INK1, and these may be used to excite more than one (e.g. two or three)
acceptor dyes or pigments in INK2. Also, it is possible to use a single donor dye
or pigment in an INK1, which emits in a wavelength range λ1e that overlaps with both
an absorption range λ2a of an acceptor dye or pigment present in ink INK2 to cause
emission in a wavelength region λ2e, and an absorption range λ3a of another acceptor
dye or pigment present in a further INK3 to cause emission in a wavelength region
λ3e. In this case, a first pattern produced by the cascade effect will be observed
in the areas of overlap between the discrete elements forming the first pattern (from
INK1) and second pattern (from INK2), and a second pattern produced by the cascade
effect will be observed in the areas of overlap between the discrete elements forming
the first pattern (from INK1) and the third pattern (from INK3). As the respective
first, second and third pattern are formed by randomly distributed discrete elements,
this will also apply to the patterns formed by the cascade effect in the areas of
overlap between the discrete elements forming the first, second and third pattern.
In consequence, the cascade effect produces two new patterns of randomly distributed
discrete elements, which have specific spectral properties due to the specific combinations
of the dyes in inks INK1, INK2 and INK3, respectively, and which can only be observed
under certain illumination conditions. This will be even more difficult to analyze
and counterfeit.
[0176] Also, in the case of an acceptor dye showing two different emission peaks in different
emission wavelength regions λ2e, λ2e', in response to excitation in two different
excitation wavelength regions λ2a, λ2a', the acceptor dye or pigment may be used in
combination with two donor dyes or pigments (in the same or different inks) having
respective emission ranges λ1e and λ3e. Preferably, the two donor dyes used in this
embodiment have overlapping excitation ranges X1a and λ3a, so that excitation with
a single wavelength (from e.g. a laser) is capable of exciting both donor dyes to
cause emission in both emission wavelength regions of the acceptor dye.
PRINTING INK COMPOSITIONS USED IN THE CASCADE EFFECT EMBODIMENT
[0177] The printing ink used in the cascade effect embodiment of the present invention comprises
at least one fluorescent dye or pigment acting as a donor, and at least one fluorescent
dye or pigment acting as an acceptor, as explained above. However, the printing ink
is typically not simply a solution or dispersion of one of these two dyes or pigments
in a solvent, but contains further components that render it suitable for use as an
ink for spraying or another printing process that causes a random distribution of
discrete elements formed by the ink. Also, the inks should remain on the substrate
and should be difficult to remove, which requires additional components. Such components
typically include at least a solvent and a binder, and optionally also a curing system
for fixing the discrete elements, as well known to the skilled person.
[0178] The solvent can be selected from solvents commonly used in the art of ink formulation
such as aliphatic or aromatic alcohols (e.g. ethanol, isopropanol or benzyl alcohol),
esters, (e.g. ethyl acetate, butyl acetate), ketones (e.g. acetone, methyl ethyl ketone),
carboxamides (e.g. diamethylformamide) or hydrocarbons including aliphatic and aromatic
hydrocarbons such as xylene or toluene and glycols.
[0179] The binder can also be selected from binders commonly used in the art of ink such
as polymeric binders of the resin type, e.g. alkyd resin, polyamide, acrylic, vinyl,
polystyrene, or silicone.
[0180] In one embodiment, the printing ink of the invention may also comprise other, in
particular, non-luminescent dyes or pigments. These other dyes or pigments are selected
such that they mask the presence of donor and/or acceptor dye or pigment, thereby
rendering their presence a covered secured feature. This masking is preferably effected
by using other dyes that do not strongly absorb in the emission wavelength range of
donor or acceptor λ1e, λ2e.
[0181] However, in order to avoid any interference with the cascade effect, the printing
inks preferably do not contain any further coloring additives, such as additional
dyes and pigments. In this case, nonetheless the presence of the printing ink may
be masked by printing the ink of the present invention on a region of a substrate
that is strongly colored, e.g. in black.
[0182] Depending on the type of ink to be formulated, the same may also include one or more
of the following optional additives: oils, diluents, plasticizers, waxes, fillers,
dryers, antioxidants, surfactants, defoaming agents, catalysts, UV-stabilizers, polymerizable
compounds and photoinitiators. When selecting suitable components for the printing
ink, the skilled person will consider that their properties, in particular their potential
capacity to absorb and/or emit light, does not adversely affect the energy transfer
(cascade effect) from the donor dye to the acceptor dye.
[0183] In the inks used for producing the cascade effect, the proportion of the fluorescent
dyes based on the total dry content of the ink is preferably 0.05 wt.-% to 20 wt.-%.
[0184] Without wishing to be bound by theory, it is believed that the cascade effect occurs
to a major degree at or close to the boundary between INK1 and INK2. In order to allow
for an efficient energy transfer from the donor in INK1 for exciting the acceptor
in INK2, donor and acceptor need to come reasonably close to each other. Accordingly,
it is preferred in the present invention that INK1 and INK2 are provided adjacent
to each other in z direction. It should be noted that a substrate is assumed to have
a predominantly two-dimensional extension, preferably planar, that is describable
by coordinates referred to as x,y, and that the third direction (third dimension)
perpendicular thereto and connecting the two opposing surfaces of the substrate is
referred to in this description as the z direction.
[0185] A more prominent effect can thus be achieved if INK1 and INK2 are provided over each
other, and wherein the solvent system of one of the inks (preferably the one applied
on top of the other, i.e. the one that is applied later) is capable of at least partially
dissolving the other ink, respectively the ink layer formed therefrom. In such a case,
the ink layers mix at the interface to some extent, thereby allowing the donor and
acceptor to come close to each other, thereby improving the efficiency of the cascade
effect.
[0186] In one embodiment, INK1 is applied first on the substrate, forming a first pattern.
That is, a substrate (e.g. paper or cardboard) is provided, INK1 comprising the donor
is provided first, and subsequently INK2 comprising the acceptor is provided on top
of INK1 to form the second pattern. In consequence, at least a part of the discrete
elements of the first pattern are provided in an overlapping manner on top (in z direction)
of discrete elements of the second pattern.
[0187] The present invention is however not limited to such an arrangement, as INK2 (comprising
the acceptor) may also be provided beneath INK1 (comprising the donor). However, since
in this case the fluorescence emission from the dye or pigment DYE2 caused by the
cascade effect, which may be used for authentication purposes, has to cross a layer
formed by INK1 in order to exit the security element and to reach a detector, generally
preferred is an arrangement wherein elements of the first pattern formed from INK1
comprising the donor are formed first (e.g. directly on a substrate), and wherein
discrete elements of the second pattern formed from INK2 comprising the acceptor are
formed over at least a part of the area occupied by the discrete elements of the first
pattern (ROI), such that at least a part of the discrete elements of the first and
second pattern overlap. If the alternative arrangement is chosen, i.e. wherein in
a region of overlap of discrete elements INK2 forms the lower layer closer to the
substrate and wherein a part of a discretele element formed from INK1 is provided
on top, it is preferred that INK1 is substantially translucent or transparent with
a light transmission at λ2e of 60% or more, preferably 80% or more, at the thickness
employed for the security element, in dry state.
[0188] For INK2, the material is preferably also translucent or transparent with a light
transmission at λ2e of 60% or more, preferably 80% or more, at the thickness employed
for the security element, in dry state, in order to avoid quenching of the emission
in the material. Further, in particular - but not exclusively - if INK2 is provided
above or on top of INK1, INK2 is preferably also translucent or transparent with a
light transmission at λ1a of 80% or more, preferably 90% or more, at the thickness
employed for the security element, in dry state, in order to allow efficient excitation
of the donor in INK1.
[0189] Also the printing methods have a significant impact on how the two ink layers in
a region of overlap of discrete elements will interact at the interface. In a particular
embodiment of the invention, two different application methods for the two distinct
patterns are used. It is an additional advantage of the invention that the response
- and hence a feature that may be used for authentication - will also depend on the
ink application methods employed.
[0190] With respect to the ink formulation, the combination of
- 1) the solvent used to print the secondly applied ink (SOLVENT T2) and
- 2) the type of resin or varnish or any other material which will form the solid dry
material from the formulation of the firstly applied ink (RESIN T1)
has a substantial impact on the efficiency of the cascade effect at the ink layers'
interface in the overlap region for the following reason. Here, and also in the following,
a component denoted with T1 or T2 represents a component of the material applied firstly
(T1) or secondly (T2). Yet, material applied firstly (T1) may be either of INK1 and
INK2. The material applied secondly (T2) it then the respective other material.
[0191] In case that RESIN T1 cannot be dissolved by SOLVENT T2, or in the case where RESIN
T1 is densely packed so it does not allow the secondly applied ink to diffuse into
the interface with the firstly applied ink, the interface will show an abrupt or sharp
transition from INK1 to INK2 respectively the dried layers obtained therefrom, and
the energy exchange between the donor and acceptor will not be favorable to the cascade
effect because only a few of them will be close to the others.
[0192] On the other hand, if SOLVENT T2 can dissolve partially RESIN T1 at the interface,
or if RESIN T1 is porous enough to let the secondly applied ink diffuse into the firstly
applied ink, an intermediate region will be formed. Here, DYE1 in INK1 and DYE2 in
INK2 will come closer to each other, so that the average distance between the two
is reduced for a significantly larger amount thereof. This enhances the efficiency
of the cascade effect. Note that it is advantageous that RESIN T1 is freely miscible
with RESIN T2, i.e. that no phase separation between the two occurs. This can be achieved
by using the same or chemically similar materials as RESIN T1 and RESIN T2.
[0193] Therefore, the formulations of INK T1 and INK T2 are preferably such that INK T1
is a solvent based ink providing a relatively porous print and that SOLVEN T2 is able
to dissolve RESIN T1 to a certain extent and diffuse within the ink layer formed first.
[0194] There is another factor that influences the proximity of the donor and acceptor dyes
or pigments and hence the efficiency of the cascade effect. In particular, in cases
where INK1 contains donor pigments, which are typically solid, isolated and non-soluble
grains, there is an additional challenge for the ink formulation to provide the acceptor
dye or pigment close to the donor pigments. For an efficient energy transfer, not
only the concentration of the pigment in INK1 within the dry ink layer should be sufficient,
but also the position of these (at the surface or uniformly distributed within the
layer) is critical and can be controlled by ink formulation. This can be achieved
by the skilled person by resorting to common knowledge in the field of ink formulations.
[0195] For an efficient energy transfer, not only the concentration of the donor pigment
in INK1 within the dry ink layer should be sufficient, but also the position of these
(at the surface or uniformly distributed within the layer) is critical and can be
controlled by ink formulation. This is an additional lever to tune the efficiency
of the cascade effect that can be exploited in the invention, as again a counterfeiter
would have to mimic not only the components employed, but also their interaction,
as influenced by the arrangement of the components at the interface. These effects
are also demonstrated in the Example provided at the end of the specification.
EXPERIMENTAL RESULTS
[0196] INK1 and INK2 as described above were formulated. The first ink INK1 was obtained
by adding 0.12 wt.-% of Lumogen® F Yellow 083 (lum1) to the blank preparation. Such
low weight % of dye guarantees that INK1 is practically invisible to the naked eye
when applied and dried. The second ink INK2 was obtained by adding to the blank 1.23
wt.-% of Lumogen® F Orange 240. The marks printed with INK2 were visible to the naked
eye and were imaged using a PMVS (Public Machine Vision System, such as a smart-phone,
tablet, smart-camera, camera/computer, scanner/computer, etc.), whereas the ones printed
with INK1 remained invisible.
[0197] Samples having a wet film deposit of 12 µm thickness were then prepared, using e.g.
a K Control Coater from RK Print Coat Instruments using, e.g., the HC2 coating bar,
on a suitable white substrate (e.g. the white part of LENETA N2C-2 substrates), followed
by drying at room temperature. Three samples were prepared:
- 1. Sample 1 with INK1 only
- 2. Sample2 with INK2 only
- 3. Sample3 with INK1 followed by INK2 further to ink 1 drying (the reverse was also
tested and gave similar results)
All three samples were then imaged using a PMVS in reflection mode and using an SMVS
(Secure Machine Vision System, i.e. a dedicated security device) sensitive in the
610 - 900 nm wavelength range, under blue illumination (around 450 nm) and green illumination
(around 530 nm). The corresponding measured intensities of the 3 samples were then
used to compose a spray spatter as illustrated in Figure 8.
[0199] On the second line of Figure 8, the picture
802 of the spray spatter, obtained using a SMVS with green illumination around 530 nm
shows, after proper image threshold operation, a contour
822 identical to the contour
811 from the PMVS. From this contour a signature
S1' is generated which is substantially equal to
S1. This is because the weak excitation of INK1 in green at 530 nm (see Figure 11) makes
it non-detectable, similarly to picture
801 where it is invisible.
[0200] At the bottom of Figure 8, a picture
803 of the mark using a SMVS with blue illumination around 450 nm exhibits the full complexity
of the synergic security element. Here both INK1 and INK2 stains are visible and their
combined contour
833 can be extracted to produce a signature
S2 with more complexity and resistance to copy than
S1=S1'. Moreover, picture
803 also reveals a higher intensity at the overlap of stains from both inks, whose contours
834 can be extracted to generate signature
S3 which exhibits the highest level of resistance to copy.
[0201] A further example is shown in Fig. 9, where both INK1 and INK2 were sprayed sequentially
on a substrate. It was ensured that both spray spatters had surface coverages that
favored sufficient stain overlap and that the first applied ink had dried before the
second ink was applied. The result is shown in figure 9 and can be interpreted the
same way as for Figure 8 by replacing
801 with
901, 811 with
911, and so forth.
Image processing
[0202] In order to obtain signatures from gray scale images of spray spatters, such as e.g.
901, 902, or
903 in Figure 9, several image processing and numerical operations can be performed.
First, regions of the spray stains and stain overlaps have to be detected on the image
using thresholding techniques. Thresholding is the simplest method of image segmentation.
From a grayscale image, thresholding can be used to create binary images (References:
L.G. Shapiro, G. Stockman, "Computer Vision," Prentice Hall, 2002;
R. C. Gonzalez, R. E Woods, "Digital Image Processing," Third Edition, Pearson-Prentice
Hall, 2008). There are several known methods for thresholding such as histogram shape-based
methods, where, for example, the peaks, valleys and curvatures of the smoothed image
histogram are analyzed, or clustering-based methods, where the gray-level samples
are clustered in two parts as background and foreground (object), or alternately are
modeled as a mixture of two Gaussians. Also entropy-based methods which use the entropy
of the foreground and background regions, or the cross-entropy between the original
and binarized image could be used. Another approach using object attribute-based methods
searching a measure of similarity between the gray-level and the binarized images,
such as shape similarity or edge coincidence could be applied.
[0203] A person skilled in the computer vision art could also make advantageous use of other
image processing techniques to extract other features of the stains such as intensity
gradients.
[0204] Once segmented or binarized, images such as
911, 922, 933 or
934 in Figure 9, representing the regions of stains are obtained; further numerical treatments
can be applied to extract a digital signature from the features of the stains. In
increasing order of complexity, non-limiting examples of features that can be extracted
are:
- The number of disjoint stains in the binarized image,
- A stain list of the coordinates of the stain centroids,
- Geometrical features such as the area (represented by a number of pixels), or the
diameter of the stain equivalent disc,
- More complex geometrical parameters such as convexity or eccentricity of each contour,
- Combinations of the above.
[0205] The above image processing and feature extraction operations may be performed in
the same way during the generation of the security arrangement (also referred to as
enrollment) and during authentication/identification as described below.
Example of producing signatures from a security arrangement
[0206] Figure 12 shows a flow diagram of a possible process of signature enrollment in a
database. The aim of this process is to extract from the patterns the three signatures
{S1'=S1, S2, S3} from a minimum number of pictures. This can e.g. be achieved using a SMVS operated
in fluorescence by the following steps.
[0207] First, a picture
802 of the mark is acquired using illumination in wavelength range λ
2a (step 1201), the spatter contour is extracted (step 1202), e.g. by image thresholding,
and signature
S1'=S1 is generated (step 1203) by applying the image processing and feature extraction
operation as described above. The same operation is performed from picture
803 using illumination in λ
1a wavelength range (step 1204) to get signature
S2 from contour
833 (steps 1205, 1206) and signature
S3 from contour
834 (steps 1207, 1208). More specifically, step 1204 comprises imaging picture
803 of spray spatter with illumination in λ
1a wavelength range, step 1205 comprises extracting spatter contour
833, e.g. using a first image threshold, step 1206 comprises generating a signature S2, step
1207 comprises extracting spatter contour
834 e.g. using a second image threshold, and step 1208 comprises generating signature S3.
[0208] All three signatures are then stored in a database (step 1209) with optionally a
product ID, an item serial number and/or product information.
Example of an authentication/identification process
[0209] The authentication/identification process having a confidence level L1 using a PMVS
by the public or occasional auditors/inspectors is described in Figure 13. It consists
of acquiring (step 1301) a picture
801 of the visible spray spatter, extracting (step 1302) contour
811, e.g. using an image threshold, and generating (step 1303) signature
S1 by applying the same image processing and feature extraction operation as the one
used for signature enrolment. The obtained signature is queried from the database
(steps 1304, 1305). For example, step 1304 may comprise sending the signature S1 to
the database, and step 1305 may comprise checking whether the signature can be found
in the database. If signature
S1 is found in the database, the security element is considered authentic or genuine
with confidence level
L1 (step 1307), if not, the security element is declared non-genuine (step 1306). If
the security element is authentic, other information on the product such as product
ID can optionally be retrieved from the database and sent to the PMVS (step 1308).
[0210] Authentication/identification with higher level of confidence can be achieved using
a dedicated inspection device (e.g. a SMVS) designed to be operated in luminescence
mode. Two possible implementations are presented in Figures 14 and 15.
[0211] Figure 14 presents a full-fledged authentication/identification method with confidence
level
L3. Step 1401 comprises imaging picture
802 of spray spatter with illumination in the λ
2a wavelength range. Step 1402 comprises extracting spatter contour 822, e.g. with an
image threshold. Step 1403 comprises imaging picture
803 of the spray spatter with illumination in the λ
1a wavelength range. Step 1404 comprises extracting spatter contour
833, e.g. with a first image threshold. Step 1405 comprises extracting spatter contour
834, e.g. with a second image threshold. Step 1406 comprises generating signatures {
S1',
S2, S3}. Step 1407 comprises sending the signatures {
S1',
S2, S3} to the database. Step 1408 comprises checking whether the signatures
{
S1',
S2, S3} can be found in the database. In the case of no, step 1409 is performed, which comprises
e.g. that the database returns a message
"Mark is NON-Genuine". In the case of yes, step 1410 is performed, which comprises e.g. that the database
returns a message
"Mark is Genuine with confidence level L3".
[0212] In summary, pictures
802 and
803 are obtained using a SMVS with illumination in the λ
2a and λ
1a wavelength ranges, respectively (steps 1401, 1403). From picture
802 the spatter contours
822 and signature
S1' are obtained (steps 1402, 1406), and from picture
803 the spatter contours
833 and
834 and signatures
S2 and
S3 are obtained (steps 1404, 1405, 1406), respectively by applying the same image processing
and feature extraction operation as the one used for signature enrollment. All three
signatures are queried from the database (steps 1407, 1408). If all three signatures
are found in the database, the mark is considered genuine with confidence level
L3 (step 1410). If the three signatures are not found all together in the database,
the mark is considered as non-genuine (step 1409). Optionally, the database may return
product ID or other information if the security element is authenticated at level
3, see step 1411.
[0213] Various authentication schemes using signature combinations could be implemented
without diverging from the scope of the invention. Figure 15 represents an example
of another signature scheme implementation where distinct and progressive confidence
levels
L3 >
L2 >
L1 can be obtained. More specifically, steps 1401-1406 are executed like in Fig. 14
and step 1507 is identical to step 1407, but query step 1408 is replaced by a series
of queries 1508, 1510 and 1512, each respectively querying whether S1, S2 or S3 can
be found individually in the database. In other words, step 1508 queries whether S1
is found in the database, step 1510 queries whether S2 is found in the database, and
step 1512 queries whether S3 is found in the database. If S1 is not found, then the
security element is identified as inauthentic, see step 1509, which could e.g. comprise
that the database returns a message
"Mark is NON-Genuine". If S1 is found in the database, but not S2, then e.g. a message of "authentic with
confidence level 1" is returned, see step 1511, which could comprise that the database
returns a message
"Mark is Genuine with confidence level L1". If S1 and S2 are found in the database, but not S3, then a message of "authentic
with confidence level 2" is returned, see step 1513, which could comprise that the
database returns a message
"Mark is Genuine with confidence level L2". If all three signatures or indices are found, then a message of "authentic with confidence
level 3" is returned, see step 1514, which could comprise that the database returns
a message
"Mark is Genuine with confidence level L3". Optionally, the database may return a product ID or other information if the security
element is authenticated at level 3, see step 1515.
EMBODIMENTS USING COLOR BLENDING
[0214] As already described above, although the use of the cascade effect is very useful
in connection with the concept of the present invention, the invention is by no means
limited thereto, and other mechanisms can be employed for generating security elements
in accordance with the present invention, in which a third pattern is used for defining
the security element.
[0215] With the wide spread of Smartphone devices in the public and the progress in their
imaging performance and computing capabilities, it is advantageous to implement an
embodiment of the present invention that does not rely on luminescence properties,
so that an authentication up to level 3 can be accomplished with conventional PMVS,
such as camera systems in smartphones or the like.
[0216] Such embodiments may make use of color blending in the overlap regions of the two
patterns. Color blending can be realized by depositing (by spraying, printing or other
methods) sequentially two patterns using two different inks having different colors,
ensuring minimum overlap of the two patterns. Color dyes or pigments are used to produce
the ink color. Color blending occurs by color subtraction in the overlay regions and
is intrinsically a non-linear process, the result of which is hardly predictable especially
for non-white substrate (see reference book
Industrial Color Physics, Klein, Georg A., Springer Series in Optical Sciences, Vol.
154, 2010, XIV).
[0217] A factor for color blending to occur through overlap of two sequentially deposited
patterns, as opposed to blending the pigments of different colors in a single ink,
resides in keeping the color saturation of the two inks at a relatively low level.
This prevents one ink layer to dominate the other by complete coverage (excess of
light absorption). Keeping the color saturation at a low level can easily be achieved
by diluting the dyes or pigments.
[0218] Reference is again made to Figure 7, which also serves to illustrate color blending
in a partial overlap of two spray stains (or of two patterns) obtained from two different
inks. INK1 of color1 produces pattern
701, INK2 of color2 produces pattern
702 and the overlap is represented by pattern
703 of color3, obtained from color blending of color1 and color2. Figure 20 shows color
components in two different color spaces, of the 3 regions of Figure 7, for the particular
inks described in the detailed example of this embodiment. Namely RGB color components
2001 and CIE L*a*b* color components
2002. This demonstrates that each region has a distinct color vector.
[0219] Figure 21 shows the contours corresponding to the 3 regions of Figure 10 extracted
using a color de-convolution operation. From the contour geometrical properties, 3
signatures can be built:
Sc1 2101 related to the properties of stains or pattern
701, Sc2 2102 related to the properties of stains or pattern
702 and
Sc3 2103 related to the properties of the overlap region
703.
[0220] Figure 22 shows an alternative contour extraction method by color threshold in the
CIE L*a*b* color space. In this case the obtained contours represent the initial stains
of INK1
2201 and INK2
2202, including the overlap region.
[0221] To illustrate the color blending effect that occurs for overlapping stains, two commercial
color dyes were employed (Lumogen® F Orange 240 (BASF) (DYE1) and Lumogen® F Yellow
083 (BASF) (DYE2). Although these dyes also produce fluorescence which was used for
demonstrating the cascade effect in the cascade embodiment of the invention, only
their color properties were used in the present example of the color blending.
[0222] An orange ink (INK1) and a yellow ink (INK2) where formulated the same way as for
the cascade effect example described above, except that the yellow INK2 was 3 times
less diluted in order to be visible.
[0223] Individual ink patterns samples and overlay samples were prepared the same way as
for the cascade effect example and imaged in reflectivity using white light illumination.
After color balance correction, the 3 regions were analyzed by colorimetric algorithms
to extract their respective color parameters in two different color spaces as shown
in Figure 20. From these graphs one can see that each of the 3 patterns has a distinct
color vector (irrespective of the color space used) which can be used to discriminate
the 3 patterns, and extract their contour from the image.
Different approaches can be used to extract the 3 pattern contours from their color
properties, among these e.g.
- 1. Color de-convolution
- 2. Color thresholding
Color de-convolution:
[0224] Color de-convolution is a known image processing technique in bio-medical imaging
for separation and quantification of immuno-histochemical staining (see reference
Quantification of histochemical staining by color de-convolution, by
Ruifrok AC, Johnston DA., in Anal Quant Cytol Histol. 2001 Aug 23(4), 291-9). This operation, also termed color unmixing, is a particular case of blind source
separation. Two random processes with two inks create random marks. The marks are
observed on a set of three spectra (RGB) and the objective is to extract the stains
created by each ink.
[0225] This task is better accomplished on optical density (OD) values obtained by the conversion
of RGB values using the Beer-Lambert law. Up to three OD ink vectors in the OD RGB
values can be defined. These vectors, once normalized, can be used to construct an
OD convolution matrix. The inversion of this matrix (de-convolution matrix) provides
a means to characterize the random marks in terms of the concentration of each ink.
Color thresholding:
[0227] Color thresholding is a common image processing technique for extracting colored
areas in images from their color parameters. It can be applied in various color spaces
with different discrimination capabilities.
Figure 22 illustrates the result of 3 suitable color thresholding operations in the
CIE L*a*b* color space (corresponding histograms are shown in Figure 23) to extract
the 3 contours corresponding to INK1, INK2, and the overlap. The result of the color
thresholding in the L*a*b* color space allows to recover the original contours of
INK1 and INK2 including the overlap
2201: signature
Sc1' (including the overlap)
2202: signature
Sc2' (including the overlap)
2203: signature
Sc3 (overlap only)
[0228] From the extracted color regions in the image, contours can be defined by suitable
image thresholding and signatures can be generated, the same way as for the fluorescence
cascade effect embodiment described previously.
Formulation Examples
[0229] In the following formulation examples, two different formulations of INK1 were tested
based on the following blank properties in order to determine the effect of ink compositions
on the observed cascade effect:
- 1) INK1 with solvent-based blank formulation (FORMULATION blank A) where the dry inks
layer is thin and hence pigments are concentrated and much of them are close to the
top surface
- 2) INK1 with UV-curable resin blank formulation (FORMULATION blank B) where dry ink
layer is significantly thicker, the pigments are homogeneously distributed along the
thickness (z axis) of the print layer, and consequently, the pigment concentration
at the surface of the print is reduced as compared to the solvent based ink
[0230] Both of these inks blanks are used for silk-screen inks and their respective formulations
are described in details below. For the formulation example described here, two examples
of INK1 using the two different blanks were formulated by adding 15%-weight of Lumilux®
green SN-F2Y pigment (Honeywell) as donor. A test patch of each formulation is first
silk-screen printed on a suitable white substrate (e.g. the white part of LENETA N2C-2
substrates) with a 90T silkscreen frame, followed by solvent evaporation drying for
the FORMULATION A and UV curing for the FORMULATION B.
[0231] A typical digital printing ink blank was formulated for INK2 (detailed FORMULATION
C below). INK2 is obtained by adding to the blank 0.3 %-weight of Lumogen® F Orange
240 fluorescent dye as acceptor.
[0232] For the purpose of the example described here, INK2 was applied with a spray device
(Nordson Microspray EFD Series 787MS-SS) and using a rectangular mask to produce a
distinguishable pattern on top of, and partially covering each of the two silk-screen
printed patterns. The spray parameters were adjusted so as to produce a dry film equivalent
to the one obtained from a wet film deposit of 12 µm thickness prepared using e.g.
a K Control Coater from RK Print Coat Instruments using, e.g., the HC2 coating bar,
followed by drying at room temperature.
[0233] The efficiency of the cascade effect for the 2 samples was measured with a camera
equipped with a lens and a long pass optical filter to transmit mainly the fluorescence
from INK2 in a wavelength range between 600 nm and 950 nm, while using a deep blue
LED emitting at a peak wavelength of 410 nm for the excitation of INK1 only (INK2
is only very weakly excited at the wavelength range where the deep UV LED is emitting).
The average intensity emitted by the patch of INK2 was obtained from stored bitmap
images and can be represented, with reference of FORMULATION B at 100%, as follows:
| INK1 blank formulation |
Cascade efficiency |
| Formulation A (solvent-based) |
152 % |
| Formulation B (UV-curable) |
100 % |
[0234] This example demonstrates that the cascade efficiency can be increased by more than
50% depending on the blank ink formulation of INK1, which comes from a larger amount
of the phosphorescent pigment being available at the interface to the layer formed
from INK2.
FORMULATION blank A (solvent based):
21.3% NeoCryl B-728, 51.6% Butylglycol acetate, 21.7% Ethyl-3 ethoxypropionate, 0.3%
Aerosil 200, 1.3% Byk-053 (anti-foam agent), 3.5% Dowanol DPM, 0.3% BYK-D410. (surfactant)
FORMULATION blank B (UV-curable):
31.5 wt.-% Tripropyleneglycol diacrylate monomer, 17.9 wt.-% trimethylolpropane triacrylate,
19.0 wt.-% EBECRYL™ 2959, 11.6 wt.-% EBECRYL™ 80, 2.1 wt.-% TEGO® Airex 900, 1.0 wt.-%
GENORAD™ 20, 9.5 wt.-% Calcium carbonate, 2.1 wt.-% Benzil dimethyl ketal and 5.3
wt.-% IRGACURE® 1173.
FORMULATION C (digital ink):
87 %-weight Methylethylketone, 10.3 %-weight of a hydroxyl-containing copolymer made
from 84 %-weight vinyl chloride and 16 %-weight of acrylic acid ester (commercially
available from Wacker Chemie under the tradename VINNOL E15/40 A) and 2 %-weight of
a terpolymer made from 84 %-weight vinyl chloride, 15 %-weight vinyl acetate and.
1 %-weight dicarboxylic acid (commercially available from Wacker Chemie under the
trade name VINNOL E15/45 M).