(19)
(11)EP 0 232 081 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
30.03.1994 Bulletin 1994/13

(21)Application number: 87300555.7

(22)Date of filing:  22.01.1987
(51)International Patent Classification (IPC)5G06F 15/62

(54)

Image processing system with capability of enlarging and reducing operations

Bildverarbeitungssystem mit Fähigkeit zum Vergrössungs- und Reduzier-Betrieb

Système de traitement image capable d'opérations d'agrandissement et de réduction


(84)Designated Contracting States:
DE FR GB

(30)Priority: 22.01.1986 JP 9946/86
22.03.1986 JP 64503/86
22.03.1986 JP 64505/86
22.03.1986 JP 64506/86

(43)Date of publication of application:
12.08.1987 Bulletin 1987/33

(60)Divisional application:
92120468.1 / 0538908

(73)Proprietor: KONICA CORPORATION
Tokyo 163 (JP)

(72)Inventors:
  • Abe, Yoshinori
    Hachioji-shi Tokyo (JP)
  • Matsunawa, Masahiko
    Hachioji-shi Tokyo (JP)
  • Yamamoto, Hiroyuki
    Hachioji-shi Tokyo (JP)

(74)Representative: Senior, Alan Murray et al
J.A. KEMP & CO., 14 South Square, Gray's Inn
London WC1R 5LX
London WC1R 5LX (GB)


(56)References cited: : 
EP-A- 0 198 269
US-A- 4 288 821
  
      
    Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


    Description


    [0001] The present invention relates to an image processing device to enlarge and reduce images by an interpolation method.

    [0002] Image recorders to record and enlarge or reduce an image have been developed. When enlarging or reducing an image to be recorded, image recorders having a solid state image sensor, such as a charge coupled device (CCD) or the like, convert an original image into electrical image signals, and the spaces among pixels read out by the solid state image sensor on the original image are continuously compensated so as to produce an image signal having near-curvilinear waveform which is sampled at varied sampling intervals corresponding to the required enlarging or reducing ratio, thus providing an enlarged or reduced image.

    [0003] However, such an arrangement disadvantageously involves many and complicated arithmetic operations to fill data into the spaces among pixels from the solid state image sensor, resulting in complicated circuitry. There is also the known method of modifying the frequency of a reading signal, that is, a transfer clock signal, for an original image which has been read by a solid state image sensor (Japanese Patent Publication Open to Public Inspection, hereinafter referred to as Japanese Patent O.P.I. Publication, No. 146358/1981). More specifically, in the case of an enlargement recording operation, the interval of the transfer clock signal is set larger than that of the same scale ratio recording operation which records the image with a larger size than originally. On the contrary, in the case of reducing recording operation, the interval is set smaller than that of the same scale ratio recording operation. However, especially in the case of enlarging operation, such an arrangement fails to provide interpolation data which compensate the data between the arbitrary pixels. With the above-mentioned method, as the transfer clock signal to be used for enlarging or reducing recording operation is generated based on the reference clock to be used in the recording operation with the same scale ratio, a circuit for generating such a transfer clock signal is required. Additionally, modifying the interval of transfer clock signal for enlarging or reducing operation necessitates controlling the exposure (time), and a circuit for this purpose. Furthermore, as an image is recorded employing the image data sampled in compliance with the transfer clock signal having a specified frequency, the enlargement of, for example, an oblique line emphasizes ruggedness and deteriorates smoothness, resulting in disadvantageously poor image quality.

    [0004] In the case where a photoelectric conversion element such as a CCD or the like is utilized as an image reading means in an image recording device which is capable of enlarging and reducing an original image, some methods provide enlarged or reduced image signals by adding or reducing appropriate image data to or from the pixel data read on an original image in accordance with an enlarging or reducing ratio.

    [0005] Fig. 30 of the accompanying drawings is a block diagram illustrating the principle area of one processing system, employed to carry out enlarging and reducing operation, which is incorporated into such an image processing device.

    [0006] In this figure, numeral 40 denotes a memory for image data, and, image data read by an image reading means is fed into an input terminal 4l on the memory, above, after being processed for enlargement or reduction. Output image data outputted from an output terminal 42 is transferred to a recorder or the like, which reproduces an enlarged or reduced image.

    [0007] In enlarging and reducing operations, the recording width of a recorder restricts the amount of image data being fed into the memory 40. In this case, the timing signals, which is generated at the address signal generator 47 and is fed into the memory 40, is controlled in accordance with either an enlarging or reducing operation.

    [0008] To achieve this, preset-capable first and second counters, 43 and 44, are incorporated. When a clock signal CLK 2 (31-c in Fig. 31) having a predetermined frequency has been counted until the preset value P1 of the counter 43 and the preset value P2 of the counter 44 are reached, the first and second output pulses C1 and C2 (31-d and 31-e in Fig. 31) are generated. The first output pulse C1 sets a flipflop 45, and the second output pulse C2 resets the flipflop 45, generating a window pulse WP shown by F in the figure. Such a window pulse WP is supplied as a gate pulse into a gate circuit 46, and a clock signal CLK2 is supplied as long as the duration W1 of the window pulse WP into the address signal generator 47. Incidentally such a clock signal CLK2 is a clock signal synchronous to the enlarged or reduced image data.

    [0009] As a result, as the address data to be fed into the memory 40 is generated for the duration W1, the image data (31-g in Fig. 31) corresponding with the duration W1, among the image data (31-b in Fig. 31) restricted by a horizontal direction valid signal (H-VALID) shown by 31-a in Fig. 31, is written into the memory 40.

    [0010] Accordingly, by varying the preset values P1 and P2 in accordance with the enlarging or reducing scale ratio, the duration W1 of the window pulse WP varies in proportion to the variation in the scale ratio, this in turn defines the amount of the image data to be written into the memory 40.

    [0011] In the case of reducing operation, the processing is carried out with the duration of the window pulse WP being equal to that of the horizontal direction valid signal (H-VALID). Contrary, in the case of enlarging operation, as the amount of image data increase, the duration of the window pulse is set shorter beforehand than that of the horizontal direction valid signal (H-VALID) so as to decrease the amount of data.

    [0012] Incidentally, such a conventional image processing system, mentioned above, causes the following disadvantages.

    [0013] In other words, with a system constitution in Fig. 30, the first address (address 0) is always designated as the initial address from where the data is to be written into, regardless of the scale ratio, though the amount of the image data to be written into the memory 40 is limited in compliance with the enlarging or reducing scale ratio. For this reason, especially when such a constitution is incorporated into an image processing system whose reading or recording device reads or records an original draft by referring to the center of a recording paper, the image to be recorded may be, depending on the scale ratio, recorded outside the recording zone of the recording paper.

    [0014] For example, assuming that W in Fig. 32 is the maximum readable width of an image reading means, an image can be recorded as shown in Fig. 33 at the same scale ratio mode with the system wherein the image data of an original draft 52 is read out by using the center line ℓ of a draft table 51 and the image is recorded based on the center line ℓ. However, in the case of reducing operation, such a recorded image appears as shown by 33-a in Fig. 33.

    [0015] This is because the initial writing address on the memory 40, that is address 0, corresponds with a initial writing address in an output device (a recorder such as a laser printer). Consequently, if a recording paper 53 to record an image is small-sized, the reduced image may not be correctly recorded on the paper, because an image may overflow from the recording-capable zone on the paper.

    [0016] Even if a recording paper 53 has a larger size, the reduced image is disadvantageously recorded as being unexpectedly located to one corner of the recording paper 53.

    [0017] Additionally, the blank area of an original draft is also enlarged in the enlarging operation, and, the resultant enlarged image appears as shown by 33-c in Fig. 33. For this reason, a needed area of image may fail to be recorded on a specific recording paper 53.

    [0018] To solve such disadvantages, it is possible to conceive a method wherein the image data with enlarging or reducing data incorporated is temporarily stored in an output buffer circuitry, then, transferred to and stored in a memory at final stage or fed into an output device which records the image.

    [0019] As can be understood from the reasons described later, such a method can solve by controlling the timing for writing the image data into or reading out such data from the output buffer circuitry in correspondance with the enlarging or reducing scale ratio, the disadvantages such as the recording of a reduced image offset to one side of a recording paper and the recording of an enlarged image with its part missing.

    [0020] However, the incorporation of such an output buffer circuitry poses still another disadvantage.

    [0021] In other words, an output buffer circuitry sometimes employs a line memory or the like to store the image data. In this case, with such a line memory, the data in line memory may uncertainly become either "1" or "0" during a rising stage of an operation for example when power is turned on for an image processing device. That is, the probability is 50%.

    [0022] Consecutively, if the data is in all clear ("0") status, there is no specific disadvantage. However, if not so, the inability to identify the original image data may render the reading of correct image data impossible.

    [0023] Similar problems may happen even in modification of scale ratio. That is, in the case where the scale ratio is changed to smaller value, for example, from the enlarging mode to the reducing mode, the former image data remains being unchanged in the line memory. In such a case, again it is impossible to distinguish the newly written image data from the former image data, posing a new disadvantage similar to the above-mentioned one.

    [0024] EP-A-0,198,269 (which is prior art by virtue of Article 54(3) only) discloses a method and apparatus for interpolation signals representative of images to be enlarged or reduced. In the disclosed device the number of original image data input varies according to the ratio of magnification.

    [0025] US-4,288,821 discloses an image signal processing apparatus in which data is mapped between high and low resolution arrays for image enhancement and half tone processing.

    [0026] Therefore, it is the object of the present invention to provide, in order to eliminate the above-mentioned conventional disadvantages, an image processing having a capability of enlarging and reducing operations and being able to correctly write image data even if a line memory is in unstable status.

    [0027] With the invention, an reduced image is not recorded at one corner on a recording paper, or, a part of an image to be recorded is not missing.

    [0028] Additionally, a conventional image processing system is such that the reading resolution of an image reading means is designed to conform to the recording resolution of a recording means for the image data having been read.

    [0029] However, as various recording systems have become more popular recently, a recording means having a recording resolution different from a reading resolution is sometimes connected for use with such a system. In such an image processing system, the scale ratio designation conducted externally is exclusively carried out on an image reading means. Therefore, the difference between the reading resolution and the recording resolution causes the failure in correct recording of an image in the designated scale ratio.

    [0030] For example, when the reading resolution is 16 dots/mm and the recording resolution is 8 dots/mm, the recorded image is enlarged twice as large as the original image, even if the same scale ratio is designated externally. On the other hand, when the reading resolution is 8 dots/mm and the recording resolution is 16 dots/mm, the recorded image is reduced to half the original image.

    [0031] With such a conventional system, as no means is provided to solve such disadvantages, it is necessary to preset a designated scale ratio corresponding to the relevant resolution in order to allow the original image to be recorded correctly in the designated scale ratio, when the reading resolution and the recording resolution are different.

    [0032] However, such a modification is extremely cumbersome and sometimes causes errors.

    [0033] Therefore, it is the object of the present invention to propose an enlarging/reducing capable image processing device which can provide the recorded image in accordance with the specified scale ratio, even though a recording means having the recording resolution different from the reading resolution is connected to the device in order to process an image.

    [0034] Among image recording devices which can enlarge or reduce an original image, an image processing system is available, wherein the image data outputted from the image processing device is transferred to a host computer in order to display the data on a display device or the like provided in the host computer side.

    [0035] Incidentally, with such a conventional image processing system, various image processing operations are executed based on the instructions from the host computer. Additionally, with this type of image processing system, a plurality of recording means are provided in the output device side which is connected to the image processing side, wherein the most suitable recording means may be selected by the instructions from the host computer.

    [0036] In such an image processing system, the scale ratio instructions from the host computer are exclusively directed to the image reading means. Accordingly, when the reading resolution differs from the recording resolution, the image is not correctly recorded, even if the image is processed in accordance with the designated scale ratio.

    [0037] Therefore, it is the object of the invention to provide, in regard to the above-mentioned image processing systems, an enlarging/reducing capable image processing system which can produce a recorded image in accordance with the scale ratio designated from the host computer side even if the image processing is executed by connecting a recording means which has the recording resolution different from the reading resolution.

    [0038] Accordingly, the present invention provides an image processing apparatus for interpolating between image data of an original image to produce an enlarged image or a reduced image in accordance with a set scale ratio, the image data representing digital pixel density data being input in synchronisation with an input clock having a frequency f1, characterised in that the apparatus comprises
       means for generating a control clock having a frequency f2, where the ratio f2/f1 corresponds to the maximum enlargement ratio;
       means for gating the control clock in accordance with the set scale ratio to produce a writing clock to write the enlarged image or the reduced image, the number of pulses of the writing clock during a given period being changed in accordance with the set scale ratio;
       means for receiving the input image data and for holding two successive image data points of the input image data;
       means for producing selection data to determine an interpolation position between the image data points, in accordance with the set scale ratio under the control of the control clock; and
       interpolating means including an interpolation data memory to store a plurality of predetermined interpolation data representing interpolating density data, the interpolation means interpolating between the two image data points based on the selection data so as to address a corresponding interpolation data in the interpolation data memory and outputting the corresponding interpolation data in synchronisation with the writing clock.

    [0039] An output buffer circuit may be provided following an image processing circuit where the enlarging or reducing operation is carried out. The initial writing or reading address in line memories provided in the output buffer circuit may be modified in accordance with either the enlarging or reducing operation, and, at the same time, the line memory may be cleared at the timing when reading is carried out on the invalid zone during the reading operation.

    [0040] Furthermore, in the enlarging/reducing capable image processing device, according to the invention, which enlarges or reduces an image by using the image data readout obtained by photoelectrically converting image information, the actual reading resolution of the image reading means maybe modified in accordance with the recording resolution of the recorder.

    [0041] Figs. 1-a and 1-b are the schematic diagrams showing the outline of an enlargement/reduction capable image recorder. Fig. 2 is the waveform diagram used for the description of functions of such an recorder. Fig. 3 is a block diagram illustrating one example of an image processing circuit. Fig. 4-a and 4-b are diagrams showing an example of interpolation data set employed in image enlarging operation. Fig. 5 is a diagram showing another example of interpolation data set employed likewise. Fig. 6 exemplifies data selection used in image enlarging operation. Fig. 7 is a data table illustrating data selection signals and processing-timing signals corresponding to Fig. 6. Fig. 8 is a diagram showing one set of data selection signals employed in the image reducing operation. Fig. 9 is a data table illustrating contents of data selection signals and processing-timing signals corresponding to Fig. 8. Fig. 10 is a signal waveform diagram used for the description of the image enlarging functions. Fig. 11 is a timing chart for the similar processing. Fig. 12 is a signal waveform diagram used for the description of the image reducing functions. Fig. 13 is a timing chart for the similar processing. Fig. 14 is a diagram employed to describe line memories. Fig. 15 is a schematic diagram of recorded images. Fig. 16 illustrates initial writing and reading addresses. Fig. 17 is a block diagram illustrating one example of an output buffer circuit. Fig. 18 is a waveform diagram used to describe the operation of the similar circuit. Fig. 19 is a waveform diagram used to describe the clearing function. Fig. 20 is a block diagram illustrating one example of a main control circuit. Fig. 21 is a table showing the correlation between the designated scale ratio and the rectified scale ratio signal which is determined by the reading resolution. Fig. 22 is a schematic diagram showing one example of an image processing system involving a host computer. Figs. 23 ∼ 25 give waveform diagrams used to describe the operations of an image processing system. Figs. 26 and 28 show other examples of image reading and image recording. Figs. 27 and 29 illustrate initial read and write values for arbitrary scale ratios. Fig. 30 is a block diagram illustrating one example of the principle area of a conventional enlarging/reducing capable image processing device. Fig. 31 is a waveform diagram used to describe the operation of the similar device. Fig. 32 is a schematic diagram of an image reading system. Fig. 33 indicate diagrams showing the status of recorded images.

    [0042] The following is a detailed description of one example of an enlarging/reducing capable image processing system, according to the invention, wherein the system is applied in the device where the processing is carried out by referring the center line ℓ as a reference line.

    [0043] Fig. 1-a is the schematic diagram of an image recording system employing an image processing device according to the invention. Numeral 401 denotes an image reader which reads image information on an original draft or the like by employing a photoelectric conversion element such as a CCD or the like and converts it into electrical signals. The reader further converts the analog signals into digital signals, and after carrying out shading rectification or the like the reader outputs the digital signals as the image data having 16 tone levels (0 ∼ F as hexadecimal notation). Numeral 402 is an image processing device which enlarges or reduces the image data at a scale ratio which is, for example, designated externally. Numeral 403 denotes a recorder, such as a laser printer, LED printer or the like, which records an image by using the data having received image-processing.

    [0044] In Fig. 1-b, the image information of an original draft 52 is read with an image reading means 60 such as a CCD or the like and converted into analog image signals. Fig. 2 illustrates the interrelation between the image data and various timing signals. The horizontal direction valid signal (H-VALID) (2-c in the figure) corresponds to the maximum draft reading width W of the CCD 60. The image data represented by 2-f in the figure is read out, being synchronous to the synchronizing clock signal CLK (2-e in the figure).

    [0045] The image signal generated at the CCD 60 is converted by the A/D converter 61 into, for example, the image data having 16 tone levels (0 ∼ F). The image data is rectified for shading at the rectificating circuit for shading 62. This arrangement is for compensating the shading caused by uneven sensitivity on the CCD 60, disaligned optical system, or uneven illumination from an irradiation lamp. For this reason, the data (for one line) of an even density plate (a white plate or the like) provided in the non-image zone of the reading device are read by the CCD 60 prior to the data reading on the original draft information, and the data readout is stored as the uneven data in the memory 63. Such uneven data for rectifying shading is together with the image data to be processed fed into the rectificating circuit 62, so as to execute rectification for shading on each pixel.

    [0046] The image data, having received rectification for shading, is transferred to the image processing circuit 2, where the real-time enlarging or reducing processing is carried out in accordance with the designated scale ratio. The data defining the scale ratio is outputted from the main processing circuit 70.

    [0047] The image data having undergone the image processing is binary coded at a binary coding circuit 23 by referring to the threshold data (for example, the dither matrix data) stored in a threshold table 69. The binary coded image data is fed into the output buffer circuit 90. The output buffer circuit 90 is provided in order to control the timing for writing or reading the image data, and is controlled based on the initial writing or reading address data outputted from the main control circuit 70.

    [0048] The image data outputted from the output buffer circuit 90 is finally transferred to the image memory 64 which stores the image data, or directly supplied to an output device, where the image is recorded as required. As an output device 65, the recorder incorporating a laser printer, LED printer or the like is available.

    [0049] Numeral 71 represents an operation key for externally setting a scale ratio. Numeral 150 is a designating means for reading-resolution. When the reading-resolution of the CCD 60 differs from the recording-resolution of a recorder provided on the output device 65, the designating means for reading-resolution is incorporated in order to adjust the reading-resolution to the recording-resolution.

    [0050] The main control circuit 70 automatically rectifys the scale ratio based on the externally designated scale ratio signal as well as the reading-resolution data in order to record the image in accordance with the designated scale ratio. The rectified scale ratio signal is transmitted to the above-mentioned image processing circuit 2. The designating means for reading-resolution is described later.

    [0051] Additionally, numeral 66 denotes a reference clock generating circuit.

    [0052] The reference clock signal outputted from the reference clock generating circuit 66 is transmitted to a timing control circuit 67, where various timing signals necessary for the image processing are generated. More specifically, in addition to timing signals (a transfer signal or the like) for driving CCD, the generated signals include a timing signal for driving an address signal generator 68 which controls the memory 63, a timing signal for the image processing circuit 2 and a timing signal to the threshold value table 69 which controls the binary coding.

    [0053] Fig. 3 is a block diagram illustrating one example of the image processing circuit 2.

    [0054] This example is so arranged that the enlarging or reducing operation can be executed every 1/64 in increments from the scale ratios 0.5 (32/64) to 2.0 (128/64).

    [0055] Principally, according to the invention, an interpolation processing is so applied that an image data is added in the enlarging operation and the image data is selectively reduced in the reducing operation. The enlargement or reduction in the main scanning direction shown in Fig. 32 is effected by the electrical signal processing. The enlargement or reduction in the sub-scanning direction is carried out by varying the scanning speed in relation to the photoelectric conversion element or image information while keeping the exposure of the similar element unchanged.

    [0056] The slowdown in scanning rate along the sub-scanning direction enlarges the original image, and, the speedup reduces the image.

    [0057] In Fig. 3, the timing signal generating circuit 10 is so arranged to generate the timing signals, one of which controls the processing timing of the image processing circuit 2 as a whole. The synchronizing clock signal (CLK), the horizontal direction valid signal (H-VAL), the vertical direction valid signal (V-VAL) and the horizontal direction synchronizing signal (H-SYNC) are fed into the timing signal generating circuit 10 in the same manner as for the CCD 60.

    [0058] In addition to the timing signal, mentioned above, the timing signal generating circuit 10 at the same time outputs the clock signal CLK2 having twice as large frequency value compared to the synchronizing clock signal CLK, in order to cope with the real-time enlarging operation with the enlarging ratio up to 2.0 as one example.

    [0059] A series of image data outputted from the CCD 60 and capable of having 16 tone levels incorporated within them are fed into two latch circuits 11 and 12 arranged in series, wherein the image data D1 and D0 corresponding with two adjacent pixels, among 4-bit-configured image data, are latched at the timing of the synchronizing clock signal. These latched data are used as the address data for the interpolation data memory 13.

    [0060] The interpolation data memory 13 is a data table which stores the new image data (which is, hereinafter, called the interpolation data) capable of being addressed based on two adjacent image data, and comprises a ROM or the like.

    [0061] As the address data for the interpolation data memory 13, the data selection signal SD, in addition to the above-mentioned single pair of latched data D0 and D1, is utilized.

    [0062] The data selection signal SD is employed as the address data to determine which data should be used as an interpolation data among the data table group selected by a pair of latched data D0 and D1.

    [0063] The data selection signal SD is designated, as described later, by the predetermined scale ratio for the enlarging or reducing operation.

    [0064] Fig. 4-a shows one example of the interpolation data S designated by the latched data D0 and D1 as well as the data selection signal SD.

    [0065] The figure illustrates the theoretical values (represented by five decimal positions) for each step as well as the actually stored interpolation data values S in the two cases, that is, the increasing inclination and the decreasing inclination, by assuming that D0 = 0 and D1 = F.

    [0066] As shown in Fig. 4, the interpolation data (4 bits) S is outputted in 16 tone levels. The image data D0 and D1 employed as the latched data respectively have 16 tone levels, which means the interpolation data table includes 16 x 16 = 256 data blocks as shown in Fig. 4-b.

    [0067] In fact, the interpolation data S is stored as exemplified in Fig. 5. However, the example data shows the case where D0 = 4 and

    .

    [0068] In Fig. 5, ADRS means the base address. This figure illustrates the correlation, between the data selection signal SD (the data, 0 through F, arranged horizontally) and the interpolation data S to be outputted, when D0 = 4 and D1 takes levels from 0 through F. The value of address data ADRS together with the value of data signal SD found in the corresponding horizontal line makes the actual address used in the interpolation memory 13.

    [0069] For example, in the case where D0 = 4 and D1 = F, when the data selection signal SD = 0, 4F0 is addressed and then 4 is outputted as the interpolation data S, and while SD = F, 4FF is addressed and then E is outputted.

    [0070] The interpolation data S outputted from the interpolation memory 13 is latched at the latch circuit 14.

    [0071] At the same time, numeral 16 denotes the interpolation data selection signal memory where the data selection signal SD is stored. In this memory, too, a data table is employed, and the data (hereinafter called the data selection signal SD) used as an address for selecting the interpolation daat is stored.

    [0072] Fig. 6 illustrates a portion of the data selection data signal SD employed in the image enlarging operation. The exemplified data is based on the enlarging ratio M set at 124/64. The enlarging scale ratio can be designated in 1/64 increments. In the figure, the symbol * denotes invalid data.

    [0073] In this way, such scale ratios are set with 1/64 increments, as shown in Fig. 6, it is preferred to set 64 as reading numbers of image data to be read with a predetermined reading interval during one cycle of an interpolating pattern, as one example, thereby processing 64 dots of image data during the one cycle. If the enlarging ratio is 124/64, since the interpolation data to be outputted during the one cycle are 124 dots, the sampling interval for addressing the interpolation data memory 13 is 64/124 (= 0.51613). Accordingly, the correlation between the sampling position (theoretical value) in relation to the phase of processing order in the one cycle and the data selection signal SD to be referred then is shown in the figure.

    [0074] In regard to the data selection signal SD at the phase "0" in the processing order, the data (0) in the left column is the leading data selection signal at the sampling position (0.00000), while the data (8) in the right column is the following data selection signal SD at the sampling position (0.51613). Such data selection signals SD in a pair differs with each other according the value of the phase in the processing order.

    [0075] Incidentally, in the phases 15, 32 and 48 of the processing order during the one cycle, the value of the data selection signal SD in the right column are null. This means there is only one data in each of these phases.

    [0076] These data are in fact stored in the status shown in Fig. 7 within the interpolation data selection signal memory 16. In this figure, within the data table where reference is made by the base address ADRS (vertical column) and the step number (horizontal line), the left figure represents the data selection signal SD, and the right figure denotes the data (referred to as the processing timing signal TD) for writing clock signal control, which is described later.

    [0077] If the processing timing signal TD is "1", the writable (write enable) status is made effective. If the signal is "0", the write inhibit status is made effective. Accordingly, the data "00" in the figure denotes an invalid data.

    [0078] Fig. 8 illustrates a portion of the data table of the interpolation data selection signal SD used in the reducing operation. The exemplified data is attained based on the reducing ratio 33/64. In the figure, the symbol * denotes thinned-out data. The data selection signal, in this case too, is stored in the memory as shown in Fig. 9.

    [0079] Now, the scale ratio signal designated with the operation key 71 is provided as an address data fed into the interpolation data selection signal memory 16 through the left 7-bit address terminals A7 ∼ A13. The scale ratio signal is, as mentioned above, supplied from the main processing circuit 70. At the same time, the counter output from the counter circuit 15 is as an address data (step number) fed into the right 7-bit address terminals A0 ∼ A6. Accordingly, the synchronizing clock signal CLK2 is supplied into the counter circuit 15.

    [0080] In this example according to the invention, since the reducing/enlarging operation can be carried out every 1/64 increment from the scale ratio 0.5 (32/64) to 2.0 (128/64), 97 kinds of scale ratio codes may be set as the scale ratio address signals.

    [0081] Further, since the expected maximum enlarging scale ratio is 2.0, the step number proceeding during the one cycle of the interpolating pattern are composed of 128 steps.

    [0082] Accordingly, 128 kinds of step number codes may be set as the step number address signals.

    [0083] Taking the scale ratio 128/64 as an enlarging example in Fig. 7, when step number is in it's initial step (0), the address 2D00 can be addressed, thereby the data 01 is outputted as the data selection signal SD and the processing-timing signal TD, and then the step number proceeds to next step while synchronizing with CLK2.

    [0084] When the step number reaches it's 128th step, the address data 2D7F can be addressed, thereby the data F1 is outputted as SD and TD.

    [0085] The interpolation data selection signal memory 16 outputs together with the interpolation data selection signal SD the processing timing signal TD.

    [0086] The processing timing signal SD is designated "1" when the interpolation data exists, as mentioned above, and, is designated "0" when the interpolation data does not exist or a part of the data is selectively removed.

    [0087] The data selection signal SD as well as the processing timing signal TD are latched at the latch circuit 17. The synchronizing clock signal CLK2 controls the latch timing.

    [0088] The processing timing signal TD is for controlling the timing of the interpolation data S which is to be latched up at the latch circuit 14. For this reason, the processing timing signal TD is supplied via the latch circuit, and is allowed to delay by an access time of the interpolation memory 13.

    [0089] The processing timing signal TD as delayed by a specific duration (one cycle of the synchronizing clock signal CLK2) is transferred as a gate signal to the gate circuit 19. The synchronizing clock signal CLK2 is supplied to the gate circuit 19 which is controlled so as to open when the processing timing signal TD is at "1" and to close when the signal is at "0". The clock signal is outputted only when the gate signal is at "1".

    [0090] The synchronizing clock signal CLK2 outputted from the gate circuit 19 is used as a latch pulse for the latch circuit 14, so as to latch up only valid data within the interpolation data S outputted from the interpolation memory 13. The synchronizing clock signal CLK2 is also employed as a writing clock signal for the output buffer circuit 90 in a following stage.

    [0091] The discussion, above, is for the principle constitution of the image processing circuit 2. Now, the output data generated at the image processing circuit 2 is, after being binary coded, transferred to the output device 65 or image memory 64 through the output buffer circuit 90 (its details are given later).

    [0092] One example of a circuit constitution for binary coding is described by referring to Fig. 3 again.

    [0093] In this figure, a threshold value table 69 contains a main scanning counter 20 to count the writing clock signal, a sub scanning counter 21 to count the horizontal direction synchronizing signal and a dither matrix 22 to output the dither threshold value based on the counts of the counters 20 and 21.

    [0094] Then, a binary coding circuit 23 binary-codes pixel by pixel by comparing the image data outputted from the latch circuit 14 with the dither threshold value from the dither matrix 22.

    [0095] Next, the image processing operation by the above-mentioned image processing device 2 is described in detail, starting with the enlarging operation by referring to Fig. 10 and those following it. For the sake of convenience, the enlarging ratio M is designated 124/64 (= 1.94).

    [0096] Fig. 10 is an analog illustration of the correlation between the original data and the post-interpolation data, where D denotes the original data, and S the post-interpolation output data.

    [0097] The correlation between the image data level and the post-interpolation data is same as in Fig. 4. Additionally, the interrelation between the sampling pitch and the data selection signal SD, in this interpolation, is identical to the example in Fig. 6.

    [0098] The chart for signal timing at each zone, during this interpolation, appears as shown in Fig. 11.

    [0099] Accordinly, it is assumed that the original image data currently obtainable from the CCD 60 are D0(0), D1(F), D2(F), D3(0) and D4(0), where each numeral in parentheses () denotes the tone level in each image data. Being synchronous with synchronizing clock signal, the latch circuit 11 outputs D1(F), and the latch circuit 12 outputs D0(0).

    [0100] On the other hand, the data table shown in Fig. 7 is referred to based on the externally desiganted scale ratio signal as well as the output of counter circuit 15. As the data selection signal SD, the following are outputted: 0, 8; 0, 8; 1, 9; 1, 9; ... (11-e in Fig. 11) As the processing timing signal TD, the following are outputted: 1, 1, 1, .... (see 11-f in the same figure)
       The interpolation memory 13 outputs the necessary interpolation data S (11-g in the same figure), by referring to the interpolation data table based on the image data D0 and D1, and the data selection signal SD.

    [0101] It is appreciated from Fig. 4-b, as the data selection signal SD is set at 0, then 8, between the image data D0(0) and D1(F), 0 is outputted as the interpolation data S0, and 1 is outputted as the similar data S1.

    [0102] As the data selection signal SD is set at 0, then 8, between the image data D1(F) and D2(F), F is outputted as the interpolation data S2, and F is also outputted as the similar data S3. As the data selection signal SD is set at 1, then 9, between the image data D2(F) and D3(0), E is outputted as the interpolation data S4, and 7 is outputted as the similar data S5. As the selection data SD is set at 1, then 9, between the image data D3(0) and D4(0), 0 is outputted as the interpolation data S6, and 0 is outputted as the similar data S7.

    [0103] Likewise, the similar data reading of the interpolation data S is further carried out for the following image data D5, D6, .....

    [0104] Accordingly, the post interpolation data are designated with symbol x, producing the diagram in Fig. 10. This demonstrates that the image data having specific level have been inserted between the original image data and outputted.

    [0105] In this way, based on actual image data D0 ∼ D4, the interpolation data S0 ∼ S7 are sequentially read out, and these interpolation data are sequentially transferred to the latch circuit 14 (I in the figure).

    [0106] On the other hand, the processing timing signal TD outputted from the latch circuit 17 is delayed by the period t (see Fig. 11) in the latch circuit 18. Such delaying period t is, as mentioned above, the necessary time span for data access in the interpolation data memory 13, and also, is the necessary time span for reading out the interpolation data S with the latch circuit 14.

    [0107] As the processing timing signal TD from the latch circuit 18 controls the turning on or off of the gate circuit 19, the latch circuit 14 executes latching only when the gate circuit 19 is on, otherwise the circuit does not latch.

    [0108] Now, the reducing operation is described as follows.

    [0109] Fig. 12 is an analog illustration of the image signals in the reducing operation, where symbol o marks the image data D0, D1, D2, D3, ....., and symbol x marks the interpolation data S0, S1, ..... Fig. 13 illustrates the timing chart of signals in such an operation. In relation to such an operation, the interrelation between the original image data D employed and the interpolation data S is identical with that illustrated in Fig. 4. The interrelation between the similar image data D and the data selection signal SD is identical with that illustrated in Fig. 8.

    [0110] Additionally, the reducing ration M stated here is 33/64 (= 0.52). Also, the tone levels of the image data are the same as those in the above-mentioned enlarging operation.

    [0111] Two neighboring image data (for example, the image data D1 and D0) are fed as address data from the latch circuits 11 and 12 into the interpolation data memory 13, and the externally designated reducing ratio (33/64) is inputted into the interpolation data selection signal memory 16, and the synchronizing clock signal CLK2 is counted with the counter circuit 15. This operation is identical with the above-mentioned enlarging operation.

    [0112] As can be understood from Figs. 8 and 9, the selection memory 16 outputs the following data selection signal SD: 0, *; F, *;, *, *; E, 0; .... As the processing timing signal TD, the following are outputted: 1, 0, 1, 0, 0, 0, 1, ...... However, as the symbol * denotes invalid data, 0 data is instead stored in the interpolation data selection signal memory 16.

    [0113] Correspondingly, the interpolation data S shown in Fig. 13 is read out from the interpolation data memory 13.

    [0114] More specifically, as the data selection signal SD between the image data D0(0) and D1(F) gives a 0 and *, only a 0 is outputted as the interpolation data S (= S0).

    [0115] As the data selection signal SD between the image data D1(F) and D2(F) gives an F and *, an F is outputted as the interpolation data S1. As the data selection signal SD between the image data D2(F) and D3(0) gives an * and *, nothing is outputted as the interpolation data S. As the data selection signal SD between the image data D3(0) and D4(0) gives an E and *, only a 0 is outputted as the interpoaltion data S2.

    [0116] Afterwards, the reading of the interpolation is, similarly to the above-mentioned, executed on the image data D4, D5, ....

    [0117] As mentioned above, because the data are read out on the actual image data D0, D1, ... with the interpolation, the interpolation data S0, S1, .... are sequentially read out, and the interpolation data S comprising S0, S1, .... are sequentially transferred to the latch circuit 14.

    [0118] At the same time, the processing timing signal TD gives 0, 1, 0, 0, 0, 1 .... (13-b in Fig. 13), and, accordinly, the write clock signal outputted from the gate circuit 19 appears as shown by 13-h in Fig. 13, and specific data are thinned out and the interpolation data S0, S1, .... are outputted.
    (13-i in Fig. 13)

    [0119] Additionally, as mentioned above, in the course of the enlarging or reducing operation, the image data are outputted by adding new image data between pixels of the original image information, or by thinning out some of the image data of the original pixels, or by outputting the values unchanged. These output image data are generally called the interpolation data.

    [0120] In the case of the above-mentioned example, it is apparent that the modification of enlarging or reducing ratio varies the data selection signal SD outputted from the interpolation data selection signal memory 16, and the interpolation data memory is accordingly addressed to output the corresponding interpolation data S.

    [0121] Now, the image data with enlarging or reducing process incorporated and binary coded are fed into the output buffer circuit 90. The output buffer circuit 90 controls, according to the enlarging or reducing ratio, data write timing as well as write address on the line memory provided within the output buffer circuit 90 itself. The following examples illustrate the cases where write timing and write address are controlled.

    [0122] The reason to control the write timing and write address according to the scale ratio is described by referring to Figs. 14 and 15.

    [0123] When assuming that the maximum image reading size of the CCD 60 is a B4 size, and if the resolution of the CCD 60 is 16 dots/mm, the amount of image data per line is 4096 bits. In consideration of the scale ratio up to two, the line memory having 8192 bit capacity and shown in Fig. 14 is provided as a line memory for storing the image data.

    [0124] Then, the image data is written or read so that the recorded results may be based on the center (2048th bit).

    [0125] Accordingly, in the reducing operation, for example, the 1/2 reduction of the image, the initial writing address for the line memory is set at the address (1024th address) which is equal to 1/4 of 4096 bits. In this case, the reducing image data is written into the line memory in the manner shown by 14-a in Fig. 14.

    [0126] At the same time, the initial reading address is set at address 0. Correspondingly, the reduced image shown by 14-d in Fig. 15 is recorded.

    [0127] This is because that the image data are "0s" between addresses 0 through 1023 and the space between the addresses is recorded on a recording paper as blank, and that the recording based on reducing image data only begins at address 1024.

    [0128] For example, in the case of the reducing ratio 32/64, the writing of the reducing image data starts at address 1024. Similarly, in the case of the reducing ratio 33/64, the writing starts at address 992. If the reducing ratio is 34/64, the writing starts at address 960.

    [0129] In this way, an image is recorded based on the center line 1 of the recording paper 53, by writing the image data so that the writing result is adjusted to the center and by starting the reading at address 0.

    [0130] Correspondingly, the initial writing address of the reducing operation is determined as shown below.







       In the case of the enlarging operation, contrary to the reducing operation, the initial reading address is controlled because of the increased image data.

    [0131] If the maximum enlarging ratio is two, then the image data is twice as many as the equal ratio operation.

    [0132] This means the four-fold area of the image to be recorded. Natuarlly, it is impossible to enlarge, for example, a B4 sized draft two times and record the whole enlarged image on a recording paper if the maximum size of the recording paper is B4.

    [0133] In view of this, it is convenient in obtaining a proportionally enlarged image to limit in advance the image data to be written in accordance with the maximum size of a recording paper.

    [0134] Therefore, as shown by 14-b in Fig. 14, in the enlarging operation, the total 4096 bits, or 2048 bits before and 2048 bits after the intermediate point (corresponds with the location of the center line 1 of the image to be enlarged) of the total amount of enlarging image data are read out.

    [0135] Accordingly, in the case of the enlarging ratio 128/64, the data in the initial bit through the 2047th bit, among the enlarging image data, are ignored, and the reading from the line memory starts at the data in 2048th bit. This means the image data of total 4096 bits are read out.

    [0136] At the same time, the writing address signal is set at "0".

    [0137] Likewise, in the case of the enlarging ratio 127/64, the reading starts at the 2016th bit. In the case of the enlarging ratio 124/64, the reading starts at the 1984th bit, and the total of 4096 bit image data are to be read out.

    [0138] If the other enlarging ratio is designated, too, it is the matter of course that the image data is designated so that the reading may start at the initial reading address in accordance with the ratio.

    [0139] To sum up, the initial reading address for enlarging opeeration is determined by the following formula.





       Incidentally, if the data is read from the line memory, starting at the midway of the enlarging image data, as mentioned above, the probability of missing image due to the enlargement of the unnecessary zone is greatly reduced, because the image data based on the central zone of the original image is read from.

    [0140] Consequently, the initial address for reading or writing in enlarging or reducing operation is determined as shown in Fig. 16.

    [0141] Fig. 17 gives a circuit diagram for one example to enable the above-mentioned functions.

    [0142] Fig. 17 illustrates one example of an output buffer circuit 90. A pair of line memories 100 and 101 are provided in the output buffer circuit 90, and image data for one lnie is fed into each of the line memories. A pair of line memories are so provided in order to process the reading or writing of the image data on the real-time bases, by alternately feeding one line of image data. For the line memories 100 and 101, a memory having 8192 bit capacity is, as mentioned previously, utilized.

    [0143] The writing or reading for the line memories 100 and 101 is controlled in the following manner.

    [0144] During the data writing into the line memories, the write clock signal (the clock signal generated at the image processing circuit 2) available via the timing signal generating circuit 10 is employed. During the data reading from the line memories, as the clock signal for reading is employed for the output device 65, these clock signals are respectively transmitted via the first and second switches 102 and 103 for selecting clock signals to each of address counters 104 and 105.

    [0145] Both of the first and second switches 102 and 103 are complementarily controlled with each other so that one line memory is in the reading mode when the other line memory is in the writing mode. To enable this, the switches are controlled with the control signal for horizontal synchronization (18-c in Fig. 18) outputted from a control circuit 107.

    [0146] Additionally, each address data to determine either the initial writing address or initial reading address for the line memories 100 and 101 is fed into each of the address counters 104 and 105 via each of the third and fourth switches 108 and 109. Both of the third and fourth switches 108 and 109 are also complementarily controlled with each other so that one address counter is in the reading mode when the other address counter is in the writing mode. The control signal for horizontal synchronization as shown by 18-c in Fig. 18 is also fed into these switches 108 and 109.

    [0147] The initial writing address or initial reading address is preset, at the address counter 104 or 105, in synchronization with the horizontal synchronizing signal (1-a in Fig. 18).

    [0148] Either one of the outputs from the line memories 100 and 101 is selected at the fifth switch 110, then fed into the previously mentioned output device 65 or the image memory 64. As the fifth switch 110 is for selecting the image data during the reading mode, the signal having a phase reverse to the control signal shown by 18-c in Fig. 18 is employed.

    [0149] Data supply lines of a pair of the line memories 100 and 101 are provided with the sixth switch 140, where clearing data as well as the image data transferred from the image processing circuit 2 are selectively supplied.

    [0150] The clearing data is utilized in order to clear the data stored within a pair of line memories 100 and 101 before the image data is supplied there. Such an arrangement is provided because data (invalid image data) different from the appropriate image data may remain within the line memories 100 and 101 if these line memories are not cleared when the power is turned on or the scale ratio is modified.

    [0151] As a clearing data, data "0" or "1" is employed. Because a binary code is used as an image data, this example employs a data "0" as a clearing data. Such a clearing data may be generated at the main control circuit 70.

    [0152] The sixth switch 140 is controlled with the clearing signal PE generated at the control circuit 107.

    [0153] The clearing timing for the line memories 100 and 101 is set at the period when the zone other than the valid one is read in the image reading. In this example, the reading on the vertical invalid zone is utilized for this purpose. Correspondingly, a signal shown by 19-d in Fig. 19 is used as the clearing signal PE, by latching the vertical direction valid signal (V-VALID) with the horizontal synchronizing signal (H-SYNC) as shown by 19-a and 19-b in Fig. 19.

    [0154] By switching the sixth switch 140 to the clearing data side only for the period where the clearing signal PE is at "1", the control signal (19-c in Fig. 19) supplied during this period allows the clearing data "0" to be written into all the addresses in the line memories 100 and 101.

    [0155] Consequently, even if any data have been written in the line memories 100 and 101 by the moment when the power is turned on, all of such data are cleared utilizing the vertical direction valid signal and the horizontal synchronizing signal, thereby completely erasing the invalid image data.

    [0156] Additionally, if the clearing signal PE is made available for the period at least two horizontal scannings are conducted, the data stored within a pair of line memories 100 and 101 can be completely deleted.

    [0157] Incidentally, the above-mentioned initial writing/reading addresses (data) are generated at the main processing circuit 70 shown in Fig. 20.

    [0158] In Fig. 20, numeral 75 represents a CPU. Numeral 76 represents a ROM containing a control program. Numeral 77 denotes a ROM storing the writing/reading addresses data shown in Fig. 16.

    [0159] The scale ratio determined with the operation key 71 is fed into the CPU 75 via an I/O port 78. Correspondingly, the initial writing/reading addresses data in compliance with the scale ratio are fed into the above-mentioned third or fourth switch 108 or 109 via an I/O port 79.

    [0160] The main control circuit 70 is provided with the designating means 150 for reading-resolution. In this example, three stages of reading resolution can be designated with the means in order to cope with the various recording-resolutions of a recorder.

    [0161] For example, if the first reading-resolution is 16 dots/mm, 12 dots/mm and 8 dots/mm are respectively designated for the second and third reading resolutions.

    [0162] Such designation of the reading-resolution is effected by switching unshown dip switches provided on the image processing device.

    [0163] The reading-resolution (digital data) obtainable from the designating means 150 for reading-resolution is fed into the CPU 75 via the I/O port 151. The CPU 75 determines the scale ratio necessary to correctly record an image in the designated ratio through arithmetic operation or by referring to the data table, based on the reading-resolution, above, as well as the externally predetermined scale ratio signal. Fig. 20 shows the case where the new scale ratio, that is, a rectified scale ratio signal, is generated utilizing the data table. One example of such a data table is shown in Fig. 21.

    [0164] As mentioned above, if the reading-resolution of the CCD is 16 dots/mm, and if the recording-resolution of the recorder is the same as the reading-resolution, the rectified scale ratio signal equals the designated scale ratio.

    [0165] However, if the recording-resolution is 8 dots/mm, this recording-resolution is transmitted to the main control circuit 70 as the designated reading-resolution. In this case, the reading-resolution is set twice as large as the recording-resolution, accordingly, if the designated scale ratio is set, for example, at twofold, the scale ratio 1.0 (64/64) is designated as the rectified scale ratio.

    [0166] The image processing circuit 2 executes enlarging or reducing operation based on the rectified scale ratio signal. Correspondingly, in the above-mentioned example, the data processing operation is affected by regarding the designated scale ratio twofold as the equal ratio. However, as the recording-resolution of the recorder is 1/2, the image is recorded twice as large as the original image even with the same ratio image data.

    [0167] Accordingly, in the case where the reading resolution of CCD is 16 dots/mm and various recording resolutions such as 16, 12 and 8 dots/mm are adopted, the correlation between a designated scale ratio and a rectified scale ratio signal appears as shown in Fig. 21.

    [0168] Additionally, as the reducing ratio 0.5 corresponding to the recording-resolution 8 dots/mm becomes 1/4 of the scale ratio of the reading-resolution being 8 dots/mm, accordingly, the rectified scale ratio is 16/64.

    [0169] If the designated scale ratio is 1, or a reducing ratio, the rectified scale ratio signal in compliance with each scale ratio is selected.

    [0170] In this way, even if a recorder having a recording-resolution different from a reading-resolution is employed, the image can be recorded with the externally designated scale ratio by using a rectified scale ratio signal.

    [0171] Likewise, even if a reading resolution will be differently changed from the recording resolution, on the contrary to the above, the image cna be recorded with same manner by using a rectified scale ratio signal.

    [0172] Fig. 22 illustrates one example of an image processing system having a host computer incorporated.

    [0173] Image information of an original draft 52 is read out with a CCD 60 serving as a means for reading, and the image signal (analog) is fed into an image processing device 200, where various image processing operations are carried out. The processed image data is transferred via a switch 700 to a plurality of recorders 65A ∼ 65C provided on an output device 65, where a required image is recorded.

    [0174] A host computer 300 connects the image processing device 200, and sends various image processing instructions to the image processing device 200. To achieve this, a plurality of data bus connect between the image processing device 200 and the host computer 300.

    [0175] A bus 301 is a data bus to transfer the image data and control command. A bus 302 is a control bus to control the transfer timing of the image data. A bus 303 is a control bus to further control the control command, above. A bus 304 is a control bus (selection bus) to designate one of the recorders 65.

    [0176] The host computer 300 is controlled by the control signal entered with a keyboard 400.

    [0177] Numeral 600 denotes a display used to display the data inputted into the host computer or a job menu or the like. A mouse 500 is employed to enter a coordinate necessary for arranging a layout of an image presented on the display 600 or to designate a job on the job menu.

    [0178] Additionally, the image processing operation based on the instruction from the host computer is detailed later.

    [0179] The general constitution of the image processing device appropriate for the incorporation into the above-mentioned image processing system is essentially the same as in Fig. 1-b which was previously described in detail. The difference with this system is that the scale ratio signal as well as reading resolution designation data are determined with the keyboard 400. These signal and data are transferred to the main control circuit 70 via the host computer 300. The reading resolution designation data is also used as the designation signal for designating the recorder 65. At the same time, an image memory 64 is contained in the host computer 300.

    [0180] With the main control circuit 70 in Fig. 20, the scale ratio data designated with the keyboard 400 instead of the operation key 71 is transmitted from the host computer 300 side to the CPU 75 via the I/O port 78. Accordingly, the initial writing address data in compliance with the ratio is fed into the third or fourth switch 108 or 109 shown in Fig. 17 via the I/O port 79.

    [0181] As mentioned above, the reading resolution designation data is supplied to the main control circuit 70. As shown in Fig. 22, with this example, an arrangement is provided so that one of the three recorders 65A ∼ 65C respectively having a different recording-resolution may be designated from the host computer 300 side.

    [0182] It is assumed that when the reading-resolution of the first recorder 65A is 16 dots/mm, the similar resolutions of the second and third recorders 65B and 65C are respectively 12 dots/mm and 8 dots/mm.

    [0183] One of these reading-resolutions is designated with a command signal from the host computer 300 side.

    [0184] The reading-resolution (digital data) determined with the keyboard instead of the designation means 150 for reading-resolution shown in Fig. 20 is inputted into the CPU 75 via the I/O port 151. The CPU 75 determines the new scale ratio required for correctly recording an image in the designated scale ratio, through arithmetic operation or by referring to the data table, based on the reading-resolution, mentioned above, as well as the scale ratio signal also designated with the keyboard 400. The method to determine the new scale ratio, that is, the rectified scale ratio signal, is same as the case where the previously mentioned designating means 150 for reading-resolution, and, the data table shown in Fig. 21 is likewise employed.

    [0185] As mentioned above, if a recorder having a recording-resolution different from a reading-resolution is designated from the host computer 300 side, an image can be recorded in the externally designated scale ratio by utilizing a rectified scale ratio signal.

    [0186] Next, the operation of the image processing system illustrated in Fig. 22 is described in detail with reference to Figs. 23 ∼ 25.

    (1) The case where a control command is transferred from the host computer 300 to the image processing device 200.



    [0187] The description is provided by referring to the timing chart in Fig. 23. When transferring a control command, the host computer 300 sets a command data valid signal OBF on a command control bus 302 at "L", and feeds 8 bit control command DB0 ∼ DB7 to a data bus 301. When detecting the command data invalid signal OBF (23-a in Fig. 23) being at "L", the CPU 75 outputs a negative logic acknowledging signal ACK (23-b in Fig. 23) into the command control bus 302, and simultaneously introduces within itself the control command data DB0 ∼ DB7 (23-c in Fig. 23) already established on the data bus 301. The fetched control command data is transferred to the CPU 75. The CPU 75 interprets the transferred control command data and operates in accordance with the command data.

    [0188] For example, if such a command is for designating a scale ratio, the CPU 75 outputs a scale ratio designation signal to the image processing circuit 2, and simultaneously displays it on an unshown display provided on the host computer 300 side. In this case, the displayed scale ratio is same as the ratio preset on the keyboard 400.

    (2) The case where a control command is transferred from the image processing device 200 to the host computer 300.



    [0189] The description is made by referring to the timing chart in Fig. 24. The CPU 75 outputs a command data strobe pulse STB (24-a in Fig. 24) onto the command control bus 302, and, at the same time, outputs in synchronization with the pulse an 8 bit image data or control command DB0 ∼ DB7 (C in Fig. 24) onto the data bus 301.

    [0190] Upon reception of the STB pulse, the host computer 300 latches the control command DB0 ∼ DB7 onto an unshown internal latch circuit, and set a command data introducing signal IBF (24-b in Fig. 24) on the command control bus 302 at "H". At this point, as the host computer 300 has not yet read out the introduced command data DB0 ∼ DB7, it cannot receive the next data. The host computer 300 sets the IBF signal at "L", in synchronization with the trailing edge of STB pulse, after completing reading of the control command.

    [0191] After confirming that the IBF signal is at "L", the CPU 75 transfers the next image data or control command.

    [0192] After completion of such transfer operations for the control command as described in (1) and (2), the reading operation for draft information is commenced.

    (3) Reading operation for draft information



    [0193] The description is made by referring to the timing chart in Fig. 25. The host computer 300 outputs an unshown read start command via the image signal control bus 302 when reading of image information has become possible. The CPU 75 on the image processing device 200 side starts reading operation for the original draft information after interpreting the command.

    [0194] When the image processing circuit 2 starts enlarging or reducing operation, a start pulse STP (25-a in Fig. 25) is outputted via the image signal control bus 302 to the host computer 300 side.

    [0195] The host computer 300, upon reception of the STP pulse, sets the internally provided image memory 64 at the writing state (25-b in Fig. 25), then allows the image data line-synchronizing signal VLS (25-c in Fig. 25) fall to "L", in order to have one line synchronized. In the VLS signal, a zone where the signal is at "L" state corresponds with the valid image zone, and anothter zone where the signal is at "H" state corresponds with an invalid image zone (blanking zone).

    [0196] Upon confirmation that the VLS signal is at "L", a binary coded image data D0 ∼ D7 (25-d in Fig. 25) is outputted onto the data bus 301 from the image processing device 200 side. In this case, the image data established on the data bus is written into the image memory 64, by means of athe image data strobe signal VSTB (25-e in Fig. 25) transmitted from the image processing device 200 side. Once the image processing device 200 has read out the image information corresponding with one page of original draft, it outputs an end signal WEND (25-f in Fig. 25) onto the image signal control bus 302, which in turn terminates the reading operation an original draft.

    [0197] Additionally, with the above-mentioned example, the present invention was applied to an image processing device which reads an image based on the center of an original draft and records an image based on the center of a recording paper, however, the invention may be applied to still another type of image processing device.

    [0198] Firstly, if an image processing device carries out image reading and image recording operation based on one side of an original draft and recording paper, the present invention may be applied without any disadvantages, because the initial reading location of the CCD 60 is identical with the initial recording location (in the case of a laser printer, the initial illuminating location of a laser beam for recording).

    [0199] Secondly, with an image processing device wherein the image reading is carried out based on the center line of an original draft and the image recording is carried out based on one side of a recording paper, the writing into the output buffer circuit 90 as well as the initial reading address are as follows.

    [0200] In this case, the initial writing address for a line memory either 100 or 101 is always address 0. Contrary, the initial reading address cannot be determined solely based on a scale ratio signal and varies in accordance with the size of an original draft.

    [0201] Accordingly, with such a type of image processing device, a signal designating the size of an original draft determines together with a scale ratio the initial reading address.

    [0202] A case, shown in Fig. 26, where the size of an original draft 52 to be read is A4, is described below.

    [0203] As mentioned previously, if the resolution is 16 dots/mm, the horizontal bit number of an A4 paper is;





    correspondingly, if the maximum readable draft size is B4, the value obtainable by multiplying width Y in Fig. 26 together with the scale ratio is the initial reading address to be provided for the line memory.

    [0204] Accordingly, the initial reading address in an same ratio operation is;





       Fig. 27 illustrates values of initial writing and initial reading addresses at an arbitrary scale ratio. In this case, the size of an original draft is A4.

    [0205] Thirdly, as shown in Fig. 28, with an image processing device wherein the image reading is carried out based on one side of an original draft and the image recording is carried out based on the center line 1 of a recording paper, the initial writing and reading addresses of the buffer circuit 90 are determined in the following manner.

    [0206] In this case, the initial writing address is determined based on the maximum bit number (3360 bits) of A4 size as well as the maximum bit number (4096 bits) on B4 size. Therefore;





    Accordingly, the initial reading address in this case is address 0.

    [0207] If the initial writing address is a negative number (an enlarging operation), the value is used as the initial reading address. Consequently, in this case, the initial writing address is address 0.

    [0208] Fig. 29 illustrates values of initial writing and initial reading addresses at an arbitrary scale ratio.

    [0209] In this way, the initial writing or reading address may be modified in compliance with the reference point for reading an original draft or the similar point for writing. Instead, the initial writing address for the line memory 100 or 101 may be modified in accordance with the size of a recording paper.

    [0210] Additionally, in regard to the above-mentioned embodiment, the frequency of a synchronizing clock signal CLK2 generated by the timing signal generating circuit 10 is set twice that of a synchronizing clock signal for reference, under the condition that the enlarging or reducing scale ratio is so arranged to enable designation of the ratio ranging from 128/64 to 33/64 in 1/64 increments, wherein such a frequency is determined based on the maximum enlarging ratio.

    [0211] If the maximum enlarging ratio is 3, for example, the frequency of cynchronizing clock signal CLK2 is set three times as large as the synchronizing clock signal for reference. Accordingly, the frequency of the synchronizing clock signal CLK2 is modified in accordance with the maximum enlarging ratio employed.

    [0212] Instead of ROMs, RAMs may be utilized as the memories 13 and 16. Additionally, an arithmetic circuit may be employed instead of the memory 13.

    [0213] As described above, the present invention provides an arrangement, wherein an interpolation data to compensate an image data between pixels derived from image information is easily determined by storing and reading such an interpolation data. Additionally, as an arrangement in the example can eliminate the conventional requirements of modifying the frequency of transfer clock signal in accordance with an enlarging or reducing ratio, neithre a complicated clock generating circuit nor control of exposure is required. More specifically, the timing function of the circuitry has been simplified, enabling the operation at a specified scale ratio, because the image processing is effected, whichever in an enlarging or reducing operation, with a clock frequency identical with that in the same ratio operation.

    [0214] Furthermore, the present invention does not involve a conventional method, wherein following interpolation of image data, the data sampling is carried out using a clock signal having a different frequency, consequently, a rapid ROM is not necessarily employed as an interpolation data memory. Additionally, according to the invention, the enlarging or reducing operation is carried out by employing a data into which a completely compensated image data has been incorporated, therefore, the method according to the invention can provide a better quality image as well as rapid processing when compared to a conventional method.

    [0215] With this invention, the writing data stored in a line memory provided within an output buffer circuit is cleared immediately before the image data is written into the line memory. Accordingly, the invalid image data written when power was applied to the device or the similar data remaining until the modification of scale ratio is completely cleared immediately before the writing of proper image data. Because of this arrangement, only a proper image data may be always written into the line memory.

    [0216] According to the invention, as the matter of course, the initial writing or reading address of the line memory is designated in correspondance with a scale ratio. Consequently, the enlarging or reducing operation is carried out as effective as the case where such an operation is based on the center of reading means, and an image may be also recorded based on the center of a recording paper.

    [0217] As a result, the system according to the invention has various features. For instance, there is no possibilities of the reduced image being recorded to one side of a recording paper or the image being recorded outside the printable zone of a recording paper. Additionally, as the blank zone of an original draft is not enlarged in an image enlarging operation, the proper image may be correctly recorded as required.

    [0218] According to the invention, the rectified scale ratio signal is generated based on the reading-resolution signal and the externally predetermined scale ratio signal, by designating the reading-resolution. The enlarging or reducing operation is, in turn, carried out based on the rectified scale ratio signal. For this reason, the image may be recorded in the externally predetermined scale ratio, even if the reading-resolution differs from the recording-resolution.

    [0219] As a result, the present invention provides an advantage that an image in compliance with a designated scale ratio may be easily recorded without fail.

    [0220] According to the invention, the enlarging or reducing operation is carried out based on a rectified scale ratio signal formed from the reading-resolution and the scale ratio signal being set from the host computer side, by designating a reading resolution-in compliance with a recording-resolution of a recorder selected on the host computer. This arrangement allows an image to be recorded in a scale ratio designated by the host computer, even if the reading-resolution differs from the recording-resolution.

    [0221] In this case, as the scale ratio is automatically rectified, it is not necessary any more to designate a scale ratio by taking the designated recording-resolution into account. This drastically simplifys the procedure for designating a sclae ratio.

    [0222] Additionally, a display provided on the host computer side indicates a scale ratio as designated with a keyboard, instead of a rectified scale ratio. This is because the indication of a rectified scale ratio on the display may produce a misunderstanding that a wrong scale ratio has been designated.


    Claims

    1. An image processing apparatus for interpolating between image data of an original image to produce an enlarged image or a reduced image in accordance with a set scale ratio, the image data representing digital pixel density data being input in synchronisation with an input clock (CLK) having a frequency f1, characterised in that the apparatus comprises
       means (10) for generating a control clock (CLK2) having a frequency f2, where the ratio f2/f1 corresponds to the maximum enlargement ratio;
       means (19) for gating the control clock (CLK2) in accordance with the set scale ratio to produce a writing clock to write the enlarged image or the reduced image, the number of pulses of the writing clock during a given period being changed in accordance with the set scale ratio;
       means (11, 12) for receiving the input image data and for holding two successive image data points (D0, D1) of the input image data;
       means (16) for producing selection data (SD) to determine an interpolation position between the image data points (D0, D1) in accordance with the set scale ratio under the control of the control clock (CLK2); and
       interpolating means (13) including an interpolation data memory to store a plurality of predetermined interpolation data (S) representing interpolating density data, the interpolation means interpolating between the two image data points (D0, D1) based on the selection data (SD) so as to address a corresponding interpolation data (S) in the interpolation data memory and outputting the corresponding interpolation data (S) in synchronisation with the writing clock.
     
    2. The apparatus of claim 1, wherein the interpolation data memory comprises a look-up table listing plural sets of pairs of image data points differing in density, each pair of image data points being provided with plural interpolation positions, and each of the interpolation positions being provided with the interpolation data (S).
     
    3. The apparatus of claim 2, wherein the interpolation means addresses the list of plural sets of pairs of image data points with the input image data (D0, D1) and further addresses the plural interpolation positions with the selection data (SD) so that the addressed interpolation data (S) is output from the interpolation data memory.
     
    4. The apparatus of claim 1, 2 or 3, wherein said means for producing selection data further produces timing signals (TD) to control the gating means to form the writing clock in accordance with the set scale ratio.
     
    5. The apparatus of claim 4, wherein the same number of the timing signals (TD) are provided for each available scale ratio during a given period in synchronisation with the control clock (CLK2).
     
    6. The apparatus of claim 5, wherein the timing signals (TD) include enable signals and inactive signals, and the gating means allows the control clock (CLK2) to form the writing clock in response to the enable signals and does not allow the control clock (CLK2) to form the writing clock in response to the inactive signals.
     
    7. The apparatus of claim 6, wherein the number of the enable signals of the timing signal (TD) during the given period of the control clock (CLK2) is changed for each of the available scale ratios so that the number of pulses of the writing clock during the given period is changed in accordance with the set scale ratio and that the number of the interpolation data output from the interpolation means is increased or decreased to form an enlarged image or a reduced image.
     
    8. The apparatus of any one of claims 4 to 7, wherein the means for producing selection data includes a selection data memory to store the selection data (SD) and the timing signal (TD) for each available scale ratio.
     
    9. The apparatus of claim 8, wherein the selection data (SD) and the timing signal (TD) are stored in a form of pairs of data in the selection data memory.
     
    10. The apparatus of claim 9, wherein the pairs of data are simultaneously output from the selection data memory in synchronization with the control clock (CLK).
     
    11. The apparatus of claim 9, wherein each available scale ratio is provided with a group of the pairs of data, and the same number of pairs of data are output from the selection data memory during a given period of the control clock (CLK2) for every available ratio.
     
    12. The apparatus of any one of the preceding claims, wherein the apparatus further comprises an image reader to read an original image with a reading pixel density, an image recorder to record the interpolation image data with a recording pixel density, and a control means to compensate for a difference between the reading pixel density and the recording pixel density.
     
    13. The apparatus of claim 12, wherein the control means adjusts the set scale ratio in accordance with the difference between the reading pixel density and the recording pixel density.
     
    14. The apparatus of any one of the preceding claims wherein the apparatus further comprises an output buffer to temporarily store the interpolation data, and an address control means to control the writing address to write the data in the output buffer and the reading address to read the data from the output buffer in accordance with the set scale ratio.
     


    Ansprüche

    1. Bildverarbeitungsvorrichtung für die Interpolation zwischen Bilddaten eines Vorlagenbildes zur Erzeugung eines vergrößerten oder verkleinerten Bildes in Übereinstimmung mit einem eingestellten Maßstabsverhältnis, wobei die Bilddaten digitale Pixeldichtedaten darstellen, welche synchron mit einem Eingangstakt (CLK) mit einer Frequenz f1 eingegeben werden, dadurch gekennzeichnet, daß die Vorrichtung folgendes umfaßt:

    - eine Einrichtung (10) zur Generierung eines Steuertakts (CLK2) mit einer Frequenz f2, wobei das Verhältnis f2/f1 dem maximalen Vergrößerungsverhältnis entspricht;

    - eine Einrichtung (19) zum Leiten des Steuertakts (CLK2) gemäß dem eingestellten Maßstabsverhältnisses, um einen Schreibtakt zum Schreiben des vergrößerten oder verkleinerten Bildes zu erzeugen, wobei sich die Anzahl der Impulse des Schreibtakts während einer gegebenen Periode entsprechend dem eingestellten Maßstabsverhältnis ändert;

    - eine Einrichtung (11, 12) zum Empfang der eingegebenen Bilddaten und zum Halten zweier aufeinanderfolgender Bilddatenpunkte (D0, D1) der eingegebenen Bilddaten;

    - eine Einrichtung (16) zur Erzeugung von Auswahldaten (SD) zur Bestimmung einer Interpolationsposition zwischen den Bilddatenpunkten (d0, D1) gemäß dem eingestellten Maßstabsverhältnis unter der Kontrolle des Steuertaktes (CLK2); und

    - eine Interpolationseinrichtung (13) einschließlich eines Interpolationsdatenspeichers zur Speicherung einer Vielzahl vorgegebener, die Interpolationsdichte repräsentierender Interpolationsdaten (S), wobei die Interpolationseinrichtung zwischen den beiden Bilddatenpunkten (D0, D1) auf Basis der Auswahldaten (SD) interpoliert, um so ein entsprechendes Interpolationsdatum (S) im Interpolationsdatenspeicher zu adressieren und das entsprechende Interpolationsdatum (S) synchron mit dem Schreibtakt auszugeben.


     
    2. Vorrichtung gemäß Anspruch 1, in der der Interpolationsdatenspeicher eine Nachschlagetabelle umfaßt, welche eine Vielzahl von Sätzen gepaarter Bilddatenpunkte enthält, welche sich in der Dichte unterscheiden, wobei jedem Paar von Bilddatenpunkten eine Vielzahl von Interpolationspositionen und jeder der Interpolationspositionen die Interpolationsdaten (S) zugeordnet sind.
     
    3. Vorrichtung gemäß Anspruch 1, in der die Interpolationseinrichtung die Liste der Vielzahl von Sätzen gepaarter Bilddatenpunkte mit den eingegebenen Bilddaten (D0, D1) adressiert, und des weiteren die Vielzahl von Interpolationspositionen mit den Auswahldaten (SD) adressiert, so daß die adressierten Interpolationsdaten (S) vom Interpolationsdatenspeicher ausgegeben werden.
     
    4. Vorrichtung gemäß Anspruch 1, 2 oder 3, in der die Einrichtung zur Erzeugung von Auswahldaten außerdem Taktsignale (TD) zur Steuerung der Gate-Einrichtung generiert, um den Schreibtakt in Übereinstimmung mit dem eingestellten Maßstabsverhältnis zu bilden.
     
    5. Vorrichtung gemäß Anspruch 4, in der dieselbe Anzahl von Taktsignalen (TD) für jedes verfügbare Maßstabsverhältnis während einer vorgegebenen Periode synchron mit dem Steuertakt (CLK2) bereitgestellt wird.
     
    6. Vorrichtung gemäß Anspruch 5, in der die Taktsignale (TD) Freigabesignale und Sperrsignale enthalten, und in der die Gate-Einrichtung dem Steuertakt (CLK2) ermöglicht, den Schreibtakt bei Vorliegen der Freigabesignale zu bilden, und bei Vorliegen der Sperrsignale dem Steuertakt (CLK2) nicht gestattet, den Schreibtakt zu bilden.
     
    7. Vorrichtung gemäß Anspruch 7, in der sich die Anzahl der Freigabesignale des Taktsignals (TD) während der vorgegebenen Periode des Steuerstaktes (CLK2) für jedes der vorfügbaren Maßstabsverhältnisse ändert, so daß sich die Anzahl der Schreibtaktimpulse während der vorgegebenen Periode in Übereinstimmung mit dem eingestellten Maßstabsverhältnis ändert und die Anzahl der von der Interpolationseinrichtung ausgegebenen Interpolationsdaten größer oder kleiner wird, um eine vergrößertes oder verkleinertes Bild zu erzeugen.
     
    8. Vorrichtung gemäß einem der Ansprüche 4 bis 7, in der die Einrichtung zur Erzeugung der Auswahldaten einen Auswahldatenspeicher zur Speicherung der Auswahldaten (SD) und des Taktsignals (TD) für jedes verfügbare Maßstabsverhältnis enthält.
     
    9. Vorrichtung gemäß Anspruch 8, in der die Auswahldaten (SD) und das Taktsignal (TD) in Form von Datenpaaren im Auswahldatenspeicher gespeichert sind.
     
    10. Vorrichtung gemäß Anspruch 9, in der die Datenpaare gleichzeitig aus dem Auswahldatenspeicher synchron mit dem Steuertakt (CLK) ausgegeben werden.
     
    11. Vorrichtung gemäß Anspruch 9, in der jedem verfügbaren Maßstabsverhältnis eine Gruppe der Datenpaare zugeordnet ist, und in der dieselbe Anzahl von Datenpaaren aus dem Auswahldatenspeicher während einer vorgegebenen Periode des Steuertaktes (CLK2) für jedes verfügbare Maßstabsverhältnis ausgegeben wird.
     
    12. Vorrichtung gemäß jedem der vorstehenden Ansprüche, die des weiteren ein Bildlesegerät zum Lesen eines Vorlagenbildes mit einer Lesepixeldichte, einen Bildrecorder zum Aufzeichnen der Interpolationsbilddaten mit einer Aufzeichnungspixeldichte und eine Steuereinrichtung zum Ausgleich der Differenz zwischen der Lesepixeldichte und der Aufzeichnungspixeldichte umfaßt.
     
    13. Vorrichtung gemäß Anspruch 12, in der die Steuereinrichtung das eingestellte Maßstabsverhältnis gemäß der Differenz zwischen der Lesepixeldichte und der Aufzeichnungspixeldichte korrigiert.
     
    14. Vorrichtung gemäß jedem der vorstehenden Ansprüche, welche des weiteren einen Ausgangspuffer zur vorübergehenden Speicherung der Interpolationsdaten und eine Adreßsteuereinrichtung zur Steuerung der Schreibadresse zum Schreiben der Daten in den Ausgangspuffer und der Leseadresse zum Lesen der Daten aus dem Ausgangspuffer gemäß dem eingestellten Maßstabsverhältnis umfaßt.
     


    Revendications

    1. Dispositif de traitement d'image pour réaliser une interpolation entre les données d'image d'une image originale afin de produire une image agrandie ou une image réduite selon un rapport d'échelle préétabli, les données d'image représentant des données numériques de densité de pixels qui sont introduites en synchronisme avec un signal d'horloge (CLK) d'entrée ayant une fréquence f1, caractérisé en ce que le dispositif comprend
       un moyen (10) pour produire un signal d'horloge de commande (CLK2) ayant une fréquence f2, où le rapport f2/f1 correspond au rapport d'agrandissement maximal;
       un moyen (19) pour déclencher le signal d'horloge de commande (CLK2) en fonction du rapport d'échelle préétabli afin de produire un signal d'horloge d'écriture pour enregistrer l'image agrandie ou l'image réduite, le nombre d'impulsions du signal pendant un laps de temps donné étant modifié en fonction du rapport d'échelle préétabli;
       des moyens (11, 12) pour recevoir les données d'image introduites et pour stocker deux points successifs (D0, D1) des données d'image introduites;
       un moyen (16) pour produire des données de sélection (SD) afin de déterminer une position d'interpolation entre les points (D0, D1) de données d'image en fonction du rapport d'échelle préétabli sous le contrôle du signal d'horloge de commande (CLK2); et
       un moyen d'interpolation (13) comportant une mémoire de données d'interpolation pour stocker une pluralité de données d'interpolation prédéterminées (S) représentant des données de densité d'interpolation, le moyen d'interpolation effectuant une interpolation entre les deux points (D0, D1) de données d'image à partir des données de sélection (SD) de façon à accéder à des données d'interpolation correspondantes (S) dans la mémoire de données < d'interpolation et à transmettre les données d'interpolation correspondantes (s) en synchronisme avec le signal d'horloge d'écriture.
     
    2. Dispositif selon la revendication 1, dans lequel la mémoire de données d'interpolation comporte une table de consultation ayant une liste d'une pluralité d'ensembles de paires de points de données d'images différant par la densité, chaque paire de points de données d'images possédant plusieurs positions d'interpolation et chacune des positions d'interpolation possédant les données d'interpolation (S).
     
    3. Dispositif selon la revendication 2, dans lequel le moyen d'interpolation accède aux données d'images introduites (D0, D1) dans la liste d'une pluralité d'ensembles de paires de points de données et accède en outre aux données de sélection (SD) dans la pluralité de positions d'interpolation de façon que les données d'interpolation consultées (S) soient sorties de la mémoire de données d'interpolation.
     
    4. Dispositif selon la revendication 1, 2 ou 3, dans lequel ledit moyen pour produire des données de sélection produit en outre des signaux de synchronisation (TD) pour commander le moyen de déclenchement afin de former le signal d'horloge d'écriture en fonction du rapport d'échelle préétabli.
     
    5. Dispositif selon la revendication 4, dans lequel le même nombre de signaux de synchronisation (TD) est réalisé pendant un laps de temps donné pour chaque rapport d'échelle existant, en synchronisme avec le signal d'horloge de commande (CLK2).
     
    6. Dispositif selon la revendication 5, dans lequel les signaux de synchronisation (TD) comprennent des signaux de validation et des signaux inactifs, et le moyen de déclenchement permet au signal d'horloge de commande (CLK2) de former le signal d'horloge d'écriture en réponse aux signaux de validation et ne permet pas au signal d'horloge de commande (CLK2) de former le signal d'horloge d'écriture en réponse aux signaux inactifs.
     
    7. Dispositif selon la revendication 6, dans lequel le nombre des signaux de validation du signal de synchronisation (TD) pendant la durée donnée du signal d'horloge de commande (CLK2) est changé pour chacun des rapports d'échelle existants de façon que le nombre d'impulsions du signal d'horloge d'écriture pendant le laps de temps donné soit changé en fonction du rapport d'échelle préétabli et que le nombre des données d'interpolation transmises par le moyen d'interpolation soit accru ou réduit pour former une image agrandie ou une image réduite.
     
    8. Dispositif selon l'une quelconque des revendications 4 à 7, dans lequel le moyen pour produire des données de sélection comporte une mémoire de données de sélection pour stocker les données de sélection (SD) et le signal de synchronisation (TD) pour chaque rapport d'échelle existant.
     
    9. Dispositif selon la revendication 8, dans lequel les données de sélection (SD) et le signal de synchronisation (TD) sont stockés sous la forme de paires de données dans la mémoire de données de sélection.
     
    10. Dispositif selon la revendication 9, dans lequel les paires de données sont transmises simultanément par la mémoire de données de sélection en synchronisme avec le signal d'horloge de commande (CLK).
     
    11. Dispositif selon la revendication 9, dans lequel à chaque rapport d'échelle existant est affecté un groupe des paires de données, et le même nombre de paires de données est transmis par la mémoire de données de sélection pendant une durée donnée du signal d'horloge de commande (CLK2) pour chaque rapport existant.
     
    12. Dispositif selon l'une quelconque des revendications précédentes, dans lequel le système comporte en outre un lecteur d'image pour lire une image originale avec une densité de pixels de lecture, un enregistreur d'image pour enregistrer les données d'interpolation d'image avec une densité de pixels d'enregistrement, et un moyen de commande pour compenser une différence entre la densité des pixels de lecture et la densité des pixels d'enregistrement.
     
    13. Dispositif selon la revendication 12, dans lequel, dans lequel le moyen de commande règle le rapport d'échelle préétabli en fonction de la différence entre la densité de pixels de lecture et la densité de pixels d'enregistrement.
     
    14. Dispositif selon l'une quelconque des revendications précédentes, dans lequel le système comporte en outre une mémoire tampon de sortie pour stocker temporairement les données d'interpolation, et un moyen de commande d'adresse pour commander l'adresse d'écriture pour enregistrer les données dans la mémoire tampon de sortie et l'adresse de lecture pour lire les données de la mémoire tampon de sortie en fonction du rapport d'échelle préétabli.
     




    Drawing