The present invention relates to a method of generating data by which the optimal number and placement of a plurality of energy beams, used to detect the presence of an object within a space under surveillance, can be determined to define a sensor system for detecting the presence of objects in that space.
Several technologies are known for use in detecting the presence of an object entering a volume of space under surveillance. These energy beams may comprise, for example, radio frequency energy (microwaves), sound waves (ultrasonics), or optical energy (e.g., infrared waves). One application of such technology involves collision avoidance systems used in on-the-road vehicles to detect the presence of an adjacent vehicle or object which poses a risk of collision (see for example US-A-4 447 800). Each of the technologies mentioned above possesses certain advantages and limitations, but all of them present certain common problems in their implementation for a specific application. For example, such collision avoidance systems are typically employed to monitor a specific volume of space immediately adjacent the "host" vehicle to be protected. Energy having unique characteristics, e.g. wavelength, modulation frequency, etc., is projected from the host vehicle into the space, and reflections of this energy toward the host vehicle are monitored to detect the presence of a vehicle or other object presenting a risk of danger. However, cost, size and other constraints impose practical limits on the amount of energy that can be projected into the monitored space; therefore, it is necessary to project the energy in a limited number of energy beams which normally cover only a portion of the volume being monitored.
Since only a limited number of energy beams may be employed in most collision avoidance systems, it is therefore important to judiciously choose the number, size and placement of the beams so as to maximize the likelihood of detection. This task is substantially complicated by several factors. For example, the size, geometry and features of the various vehicles which might enter the monitored zone vary widely, and therefore each present unique reflective characteristics. Also, adjacent curbing, road features and reflective road striping have the potential to reflect energy and therefore may result in false detections. On the other hand, the monitored volume of space must be sufficiently "covered" with beam energy so that smaller objects and vehicles, such as motorcycles and pedestrians, may be reliably detected.
Accordingly, there is a need in the art for a method of determining the optimal number and positions of a plurality of energy beams employed to detect the presence of an object within a volume of space being monitored and particularly detection beams used in collision avoidance systems for vehicles.
It is an object of the present invention to provide a method of generating data for determining the optimal number and placement of a plurality of energy detection beams employed to detect the presence of an object, such as a vehicle within a volume of space under surveillance.
According to the invention, this is achieved by the features in the claim 1. Advantageous further embodiments are described in the subclaims 2 through 9.
The method is preferably carried out using a plurality of sample beams of energy, such as infrared optical energy. The source of the sample beams or the object under surveillance is moved in each of a plurality of linear, parallel paths so that reflections are sensed at a plurality of locations over the area under surveillance. For each relative position between the beam source and the object, the beam source is rotated about a substantially vertical axis to determine the reflections based on a plurality of angles of incidence.
Similarly, the beam source is rotated about a substantially horizontal axis to each of a plurality of positions to determine the reflectivity of features on the object for each of the relative positions between the beam source and the object. The recorded data is employed to produce a composite map of the locations in space where reflections have been noted. The map of reflective features may then be used to determine the optimal number and placement of a limited number of detection beams.
In the drawings, wherein like reference numerals are employed to designate identical components and which are to be used in conjunction with the following description:
The present invention relates to a method of generating data for determining the optimal number and placement of energy beams which are used to detect the presence of an object within a zone to be monitored. One such application of object detection requiring strategic placement of the energy beams is in the context of collision avoidance systems for vehicles, where it is desired to determine the presence of another vehicle, pedestrian or other object which is within a zone of danger relative to a protected vehicle and poses a risk of collision therewith.
A typical collision avoidance system using multiple energy beams to detect adjacent objects is depicted in Figures 1-3. A tractor-trailer type vehicle is depicted which comprises a tractor or truck 24 that tows a long trailer 22. Large trucks of this type are typically referred to as "Class 8" trucks, and because of their size and configuration, the operator of such a vehicle normally has a relatively large blind spot within which an adjacent vehicle may be concealed from his view, thus presenting a severe hazard during lane-changing maneuvers, turns and the like. The collision avoidance system includes a collision avoidance detector 20 mounted on each side of the truck 54 and more particularly on the door 30 of the truck cab. The precise mounting location of the detector 20 will vary depending upon the nature, size and configuration of the truck or other "host" vehicle with which the detector 20 is used. In many applications, such as the truck 24, it may be unnecessary to employ a detector 20 on both sides thereof, and therefore only those energy beams 28 associated with the detector 20 on the right side of the truck 24 will be discussed herein.
The detector 20 emits a plurality of beams 28 of energy forwardly, laterally and rearwardly into a volume of space 26 which comprises the zone to be monitored for the presence of vehicles, pedestrians or other objects posing a risk of collision with the truck 24. The energy beams 28 may comprise acoustical energy (e.g., ultrasonics), radio frequency energy (e.g., microwaves), or optical energy (e.g., infrared light waves). In the disclosed embodiment, nineteen of the beams 28 are employed to detect objects within the monitored zone 26. The zone 26 under surveillance is generally rectangular in plan view and is of a preselected height. The dimensions and geometry of the monitored zone 26 are such that the volume of space monitored is above the surface 26 of striping 32 on the road and extends laterally to a point which falls short of curbs 34, signs (not shown) or other objects which are not intended to be detected.
The beams 28 emanate from the detector 20 and impinge upon objects, such as an adjacent vehicle 38 within the zone 26. A portion of the energy in the beam 28 is reflected from surface features such as the trim piece 40 on the adjacent vehicle 38, a portion of which energy is reflected in a beam 36 back to the detector 20. The spacing between, location, direction, size and number of the beams 28 must be carefully selected in accordance with a particular application to optimize system operation and to assure that all objects of interest within the zone 26 may be reliably detected. In some cases, it may be determined that certain of the beams 28, such as the lateral beam 28a and rearwardly projecting beams 28b, should extend between the normal monitoring zone 26 in order to detect certain unique surface features of adjacent vehicles such as trim components, license plates or the like. In any event, it is important to use a relatively minimum number of the beams 28 to provide the required degree of detection reliability, since the use of additional beams is unnecessarily costly and may not significantly add to detection reliability. The present invention is therefore concerned with generating data for determining the number and optimal placement, positioning and length of the beams 28 so as to assure high detection reliability at minimum cost.
Attention is now directed to Figures 4-8 which depict an arrangement for carrying out a novel method of generating data for determining the optimal position, placement, length and number of the beams 28. First, an envelope shown in solid line and designated by the numeral 56 in Figure 4 is determined, which defines the area in plan view where it is desired to detect objects which might pose a risk of collision with the host vehicle 24. The envelope 56 essentially possesses a configuration in plan view similar or identical to the zone 26 shown in Figure 1. A source of sampling beam energy 42 is positioned at a point adjacent the detection envelope 56, essentially corresponding to the point on the host vehicle 24 where the detection unit 20 would ordinarily be mounted (see Figure 1). In one suitable embodiment, the sampling beam source 42 produces six horizontally aligned beams 66 of energy, for example, optical energy of the infrared wavelength. Four of such beams 66a-66d are depicted in Figure 7 as impinging on the door of an adjacent vehicle 38. In the present embodiment, the beams 66 are of circular cross-section and, as shown in Figure 8, impinge upon the side of the vehicle 38 as contiguous or closely adjacent circular spots 110a-110f.
For purposes of carrying out the method of the present invention, the detection envelope 56 is divided by sector lines 58-62 into four sections respectively subtended by essentially equal angles A-D. The sampling beam source 42 is shown in more detail in Figure 6. Two sets of three horizontally aligned light-emitting diodes (LED) are respectively mounted behind lenses 74, 76 on optical supports 70, 72. Two sets of three photodetectors 84 are likewise mounted behind lens 80, 82 which are disposed on the optical mounts 70, 72. Thus, each set of LEDs 78 are respectively associated with the photosensors 84 carried by the same optical mount 70, 72. The precise aiming of the LEDs 78 and photosensors 84 may be controlled by any suitable means such as micrometer adjustments 86. The optical mounts 70, 72 are respectively secured to bases 88, which in turn are independently pivotally mounted for rotation on pivot pins 90 about substantially vertical axes on a common support assembly 94. The support assembly 94, in turn, is mounted on a base 96 which may be rotated about a substantially vertical axis 106. The base 96 is coupled by a joint 98 to a bearing assembly 100 which is rotatable about a substantially horizontal axis 104. The bearing assembly 100 is in turn mounted on a suitable support 102. It may thus be appreciated that the optical mounts 70, 72 may be rotated about their vertical axes to achieve alignment between the two sets of respectively associated horizontal beams, and rotation about the horizontal and vertical axes 104, 106 allows precise aiming of the horizontal bank 66 of six beams to specific areas within the detection envelope 56. More particularly, it can be appreciated that the bank of six horizontal beams may be pivoted about a vertical axis in the direction of arrow 64 (Figure 4) to any of a plurality of indexing positions through an angle which will be referred to herein as the "equatorial" angle. Likewise, it may be appreciated that the bank of six beams 66 may be rotated about a horizontal axis 104 in the direction of arrow 65 (Figure 5) to any of a plurality of second index positions through angles which will be referred to herein as "azimuthal" angles.
After the beam source 42 is disposed in the proper position adjacent the envelope 56, it is rotated about a vertical axis to one of the sectors, such as sector D, at a given azimuthal angle. For example, as shown in Figure 5, the first azimuthal angle may correspond to the level indicated by the numeral 1, so that the beams effectively illuminate upper portions of the vehicle 38.
With the beam source 42 positioned at the first combination of equatorial and azimuthal angles, the vehicle 38 is moved along a first track 40 designated as track 1 to a first position where a portion thereof, for example, is within sector D. With the beam source 42 aimed at sector D, the LEDs 78 are pulsed to produce horizontal bank of beams 66. Those of the beams 66 which are incident on the surface of the vehicle 38 result in illumination spots, such as those designated by the numerals 110a-110f in Figure 8. Depending upon the geometry and nature of the surface features thusly illuminated, a portion of the incident energy will be reflected back toward the beam source 42 and is picked up by the photosensors 84. In other words, those particular illuminated surface features on the vehicle 38 which are disposed at certain angles relative to the beam source 42 reflect the incident beam at the proper angle back toward the beam source 42 so that the beam source 42 effectively "sees" or detects such reflective feature or surface area. The precise location in space of each such reflection produced by a surface feature is recorded for each of the tracks 40 (1-7). The precise points in space where such reflections occur can be determined using any of various known techniques, including triangulation. Normally, for future use, the reflection locations are recorded in the form of an x, y, z coordinate system, along with the equatorial and azimuthal angles of the beam source 42.
After the first set of data is recorded in the manner just described, the beam source 42 is rotated about its vertical axis to view the next sector, which in this case is sector C, while the vehicle 38 remains at its first incremental position in track 1. At this point, a second set of reflective data is taken, following which the beam source 42 is successively moved to view sectors B and A where similar sets of reflection data are recorded. After the beam source has been rotated through each of the sectors A-D for a given azimuthal angle, the beam source 42 is rotated to a second azimuthal angle so as to view, for example, that portion of the vehicle 38 at a lower elevation designated by the numeral 2. The beam source 42 is then successively indexed to differing equatorial angles to view sectors A-D for the second azimuthal angle, and the reflection data is recorded for each combination of these equatorial and azimuthal angles. In a similar manner, this same recording process is carried out for each of the additional azimuthal angles, which correspond to elevations 3 and 4 in Figure 5, where the beam source 42 is swung about its vertical axis to permit recording of reflection data in each of the four sectors A-D. Following the recording of such data, the vehicle 38 is then moved forwardly in track 1 to a second incremental position, where a second set of reflection data is recorded in the manner just described. It may be readily appreciated that this second set of reflection data may very well be different than the first set thereof since, due to the incremental movement of the vehicle 38, the angle of incidence on the various reflective features of the vehicle 38 is slightly changed. For example, as shown in Figure 8, the vehicle 38 is disposed at a particular position of incremental movement where a particular surface feature designated by the numeral 114 is coincidentally disposed at an angle substantially perpendicular to rays 116 of the beam 66d. As a result, the rays 116 incident on surface feature 114 are reflected back toward the beam source 42 and are detected by the corresponding photosensor 84 as a reflection. In contrast, it may be seen that the rays, e.g. 112, of the other beams 66a, 66b, 66c, 66e, 66f are incident on surface features which are inclined relative to the beam source 42 at an angle such that the incident light energy is reflected in a direction which prevents detection thereof by the photosensors 84.
Figure 7 illustrates the alignment of the bank of beams 66 along a substantially horizontal axis 108, four of such beams 66a-66d being shown incident on the door of the vehicle 38. From this figure, it may be seen that beam 66d is incident on the door handle 106 of the vehicle 38 which may possess surface areas of various angular dispositions that may increase the likelihood that a portion of the energy in beam 66d is reflected back toward and is detected by the corresponding photosensor 84. The successively lower elevational positions of the beam bank are designated by the numerals 2, 3 and 4 in Figure 7, these elevational positions being determined by the azimuthal angle of the beam source 42.
It may be readily appreciated that the reflection data determined in the manner described above can likewise be recorded by other techniques and sequences of relative movement between the beam source and the vehicle 38. For example, the vehicle 38 may remain stationary, and the beam source 42 may instead be moved to successive positions in a plurality of linear paths at successively greater lateral distances from the vehicle 38, and in this manner, a set of data would be produced which is identical to that previously described. Indeed, it may be preferably to linearly move the beam source 42 to successive incremental positions, rather than the vehicle 38, since the beam source 42 could easily be mounted on apparatus which can be more precisely indexed and displaced in comparison to the vehicle 38. Likewise, the sequence of indexing the beam source 42 to various azimuthal and equatorial angles could be altered. For example, for a given incremental position of the vehicle 38, the beam source 42 might be rotated to successive azimuthal positions for each selected equatorial position. It is only necessary that there be the required relative movement between the beam source and the vehicle 38 so as to determine the precise points in the envelope of space 56 at which reflections are obtained from various surface features on the vehicle 38.
After a set of reflection data is obtained for a vehicle 38 of a given configuration, the process is repeated to obtain a second set of data for a vehicle of a differing configuration. This is necessary since the reflective features of vehicles differ depending upon the physical geometry, size and choice of materials used in the vehicle. It is known, however, that certain types of vehicles possess surface geometries which are more difficult to detect than others. Accordingly, it is preferable to select those vehicles having the surface features which are most difficult to detect for use in generating reflection data which is subsequently used to determine the number and displacement of sampling beams.
Having generated the necessary sets of data, such data can be plotted to effectively provide a map of the points in the envelope space 56 at which "detections" may be obtained for a given vehicle model. For example, Figures 9-12 are plots or "maps" of detections for a 1988 Pontiac Fiero which is black in color. The ordinate depicts the lateral distance from the beam source 42, which of course corresponds to the lateral distance from the detector 20. Also, it can be appreciated that the vertical plots are divided into seven tracks which correspond with the lateral, parallel tracks 40 shown in Figure 4. The abscissa is a plot of the distance from the beam source 42 (or detector 20) in a direction forward up to the leading edge of the envelope 56 (or zone 26) which is indicated as a vertical, dotted line. These plots are based on a four-level gray scale, with the totally white spaces indicating no-detects and the totally dark portions indicating four or more detects at any given x-y position. Figures 9-11 respectively correspond to the data recorded for three different elevational (azimuthal) positions, with Figure 9 being the lowest position, Figure 10 being at a middle position, and Figure 11 being at an uppermost elevational position. It can be seen in Figure 9 that the low beams fail to detect the vehicle at the upper left region of the zone due to the fact that they are aimed under the rear bumper at long ranges. It can also be seen that there is a strong area of response beyond the leading edge of the zone (vertical dash line) due to the very high reflectivity of the rear license plate of the vehicle.
Figure 10 shows that the middle height beams fail to detect the "test" vehicle in some parts of the middle zone because they sometimes all fall on highly reflective body panels at unfavorable angles of incidence. Again, there is a substantial response forward of the leading edge of the zone due to the high reflectivity of the license plate of the test vehicle.
Figure 11 demonstrates that the highest position of the beams are useful mainly for detecting at the outer lateral boundaries, as well as at relatively close distances, particularly near the beam source (or detection unit 20).
Figure 12 is a detection map which forms a composite of those shown in Figures 9-11, showing the detection responses for all of the beams at all of the chosen azimuthal elevations for the forwardmost sector of the envelope of volume 56. From Figure 12, it was determined that a subset of only ten of the total number of thirty-six beams that were used to generate the reflection data were needed to effectively cover the forward quadrant and could do so with less undesired response resulting from the highly reflective areas of the test vehicle forward of the leading edge of the zone being monitored.
Figures 13-15 are similar to Figures 9-11 but depict the detection data for a 1988 Ford EXP which is black in color. Figure 13 shows that, compared with the Fiero, there is less response at long ranges, due apparently to the relatively high position of the license plate on the Ford EXP. All of the beams fall under the plate at long ranges. Figure 14 shows that there is less response at the longer lateral distances compared to the Fiero, most probably due to the absence of reflective trim or side reflectors on the Ford EXP. Figure 15 shows that the higher beams used to generate the detection data in this particular map are sufficiently high in elevation to detect the comparatively high license plate of the Ford EXP forward of the leading edge of the detection zone. Figure 16 is a map of the detection responses which is a composite of the maps of Figures 13-15. From Figure 16, it can be concluded that the set of ten beams which were effective for detecting Fiero are likewise very effective for detecting the Ford EXP.
Using the recorded test data and detection maps such as those described above, the final number, positioning and length of the sampling beams may be determined. A typical set of such beams indicated by the numeral 28 are shown in Figure 17. Figure 17 depicts the relative angles of the beams as well as their heights and lengths.
From the foregoing, it may be appreciated that the method of the present invention not only provides for the reliable accomplishment of the object of the invention, but does so in a particularly effective and economical manner. It is recognized, of course, that those skilled in the art may make various modifications or additions to the preferred embodiment chosen to illustrate the invention without departing from the scope of the present contribution to the art. Accordingly, it is to be understood that the protection sought and to be afforded hereby should be deemed to extend to the subject matter claimed and all equivalents thereof fairly within the scope of the invention.