JP2007255978A - Object detection method and object detector - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an object detector for enhancing object detection accuracy and moreover detecting objects other than vehicles. <P>SOLUTION: This object detector is equipped with a control unit CU for detecting an object based on information obtained from a camera 1 and a radar 2. The control unit CU executes a change calculation process for finding a change in the size of a noted area which is an area where an object in an image is shown while finding a change in distance to an object on a measuring point obtained from detection information, an object detection process for collating the change in size with the change in distance to determine that an object has been detected if a correlation is obtained between the two, and a kind determination process for determining the kind of an object based on at least either of the image information in the noted area on which the correlation is obtained and the detection information on the measurement point. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an object detection method and an object detection apparatus for detecting an object based on image information obtained from image information acquisition means such as a camera and distance information obtained from detection information acquisition means such as a radar.

  2. Description of the Related Art Conventionally, driving support devices such as a collision prevention device, an inter-vehicle distance control device, and a following traveling device have been proposed for vehicles. In these driving support devices, an object detection device that detects an object (obstacle) such as a vehicle traveling in front of the host vehicle is used.

  Such an object detection device is required to detect an object with high accuracy.

As a conventional technique for improving such detection accuracy, for example, a camera and a radar are provided as detection sensors, and the vehicle is based on the object edge extracted from the camera image and the object detection information by the laser radar. There has been proposed a technique for accurately detecting and determining (see, for example, Patent Document 1).
JP 2005-157875 A

  However, the above conventional technology determines the region on the image that is assumed to be a vehicle from the edge histogram and the detection direction of the radar. May be included. In this case, when the other object included in the region facing the vehicle is an artificial object such as a building having a straight edge, for example, even if the reflection information from the radar is detected, There is a risk of erroneously detecting an object as a vehicle.

  The present invention has been made paying attention to the above-described problems, and an object thereof is to provide an object detection method and an object detection apparatus capable of improving the detection accuracy of an object.

  In order to achieve the above-described object, the present invention obtains image information related to an object in the object detection target area from the image information acquisition means and scans the object detection target area with a detection wave from the detection information acquisition means. Obtaining detection information relating to the detected object, obtaining a change in the size of the region indicating the object on the image based on the image information, and determining the distance between the detection information obtaining means and the object based on the detection information. The change is obtained, the obtained magnitude change and the distance change are collated, and when the correlation between the two is obtained, it is determined that the object is detected, and at least one of the image information and the detection information in the area where the correlation is obtained. This is an object detection method whose main feature is to determine the type of an object based on the above.

  In the object detection method of the present invention, the object detection is performed based on the change in size in the region indicating the object on the image and the change in the distance from the object measured by the detection information. It is possible to suppress the detection of an object based on information on an object that is different from a determination target such as an adjacent object, and improve the object detection accuracy.

  Furthermore, in the object detection method of the present invention, the determination of the type of the object is performed on at least one of the image information and the detection information of the region determined as the same object by collating the size change and the distance change as described above. Based on the determination of the object.

  For this reason, it is possible to prevent the determination of the type of the object based on the information of the object that is different from the determination target, such as the background and the adjacent object, and to improve the detection accuracy of the object. Can be determined.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  The object detection apparatus according to this embodiment is based on image information acquisition means (1) for acquiring image information related to an object in the object detection target region and reflection of the detection wave obtained by scanning the object detection target region with the detection wave. An object detection apparatus comprising: detection information acquisition means (2) for acquiring detection information related to an object; and object detection means (CU) for detecting an object based on the image information and detection information. Means (CU) obtains the image information and detection information, obtains a change in size within a region of interest, which is an area indicating an object on the image, and on the measurement point obtained from the detection information Change calculation processing for obtaining a change in distance between the object and detection information acquisition means, object detection processing for detecting an object based on the magnitude change and the distance change, and when an object is detected, And the region of the image information, and executes the and determining the type determination processing type of the object based on at least one of the detection information of the measuring point.

  An object detection apparatus according to Example 1 of the best mode of the present invention will be described with reference to FIGS.

  As shown in FIG. 1, the object detection device according to the first embodiment is mounted on a vehicle MB (hereinafter, the vehicle MB provided with the object detection device is referred to as a host vehicle in order to be distinguished from another vehicle AB described later). And a camera 1 as image acquisition means and a radar 2 as detection information acquisition means. Information on the object detected by the object detection device is output to a driving support device or the like.

  For example, the camera 1 is mounted in the vicinity of a room mirror (not shown) in the passenger compartment. The camera 1 includes at least one of a visible light camera that captures a luminance image using an image sensor such as a CCD (Charge Coupled Devices) and a CMOS (Complementary Metal Oxide Semiconductor), and an infrared camera that captures an infrared image. In the first embodiment, a visible light camera is used.

  The radar 2 is attached to the front part of the host vehicle MB, scans the vehicle front (arrow FR direction) in the horizontal direction, and detects the distance and the reflection intensity from the radar 2 to an object (detection point) existing in front of the vehicle. . The detection point is a position where an object is detected, and is detected as a position on the XZ-axis coordinate shown in FIG.

  As the radar 2, a millimeter wave radar, a laser radar, or an ultrasonic radar can be used. In the first embodiment, a laser radar is used. In the case of laser radar, the distance and the reflection intensity of the laser beam can be obtained as information on the detected object. In the case of millimeter wave radar, distance, radio wave reflection intensity, and relative speed between the subject vehicle and the detected object can be obtained as detected object information.

  Information obtained by the camera 1 and the radar 2 is input to a control unit CU as object detection means. The control unit CU inputs signals from in-vehicle sensors including the camera 1 and the radar 2, detects an object, and performs object detection control for discriminating the type, and includes a RAM (Random Access Memory), a ROM ( It is a known device including a read only memory (CPU), a central processing unit (CPU), and the like.

  The flow of processing in the object detection control in the control unit CU will be briefly described with reference to FIG. 2. First, in step S1, information is input from sensors including the camera 1 and the radar 2, and the memory 11 (see FIG. 3). Execute acquisition process to save.

  In the next step S2, information conversion processing for converting the stored input information into information used in the subsequent processing (this converted information is referred to as conversion information) is executed.

  In the next step S3, an object presence determination process is performed for determining the presence of an object based on image information and its conversion information, and determining the presence of an object based on detection information and its conversion information.

  In the next step S4, a change calculation process for calculating a change in distance from the object determined in step S3 and a change in size on the image is executed.

  Next, in step S5, an object detection process for performing object detection is executed based on the collation with the distance change and the magnitude change obtained in step S4.

  In the next step S6, a type determination process for determining the type of the detected object is executed.

  FIG. 3 is a block diagram conceptually showing a portion for performing each of the processes performed in the control unit CU.

The memory 11 stores each piece of information obtained by the acquisition process in step S1.
The conversion process part 12 is a part which performs the information conversion process of step S2.
The object presence determination processing unit 13 is a part that executes the object presence determination process in step S3.
The distance change calculation processing unit 14 and the magnitude change calculation processing unit 15 are portions that execute the change calculation processing in step S4.
The object detection processing unit 16 is a part that executes the object detection process of step S5.
The type determination processing unit 17 is a part that executes the type determination processing in step S6.

Next, each process of said step S1-S6 is demonstrated in detail.
First, in the acquisition process of step S1, as shown in FIG. 4, the luminance image information captured by the camera 1 is stored in the memory 11, and the radar 2 scans in the horizontal direction as detection information detected by the radar 2. For each resolution (angle), the distance from the radar 2 to the object and the reflection intensity of the laser beam at each angle are stored.

Here, an example of the image information by the camera 1 is shown in FIGS.
FIG. 5 and FIG. 6 show an example of an image when another vehicle (preceding vehicle) AB, a person (pedestrian) PE, and a wall WO as a road structure exist in front of the host vehicle MB. Is projected on the imaging surface 1a of the camera 1 as shown in FIG.

  5A shows a state seen from the side, and FIG. 6A shows a state seen from above. Moreover, in both figures, (c) shows an infrared image when an infrared camera is used as the camera 1, and MP in the vehicle MB diagram shows a portion (muffler portion) captured as a high temperature. .

  5 and 6, the distance from the vertical projection point PA on the road surface of the camera 1 to the end point PF on the own vehicle MB side in the vertical projection image on the road surface of the person (pedestrian) PE is z, the point PA and the point PF. The interval in the x-axis direction is xs.

Therefore, when the reference coordinate is the origin of the center of the lens 1b of the camera 1, the position of the point PF in the reference coordinate system is represented as (xs, -H, z), and the point PF is located on the image on the imaging surface 1a. The coordinates (xc, yc) to be expressed are expressed by the relationship of the following expressions (1) and (2) using the focal length f.
xc = xs · f / z (1)
yc = −H · f / z (2)

Next, an example of detection information by the radar 2 is shown in FIG.
FIG. 7 shows an object detection example similar to the image information shown in FIGS. 5 and 6 shown in FIG. As shown in this figure, the presence of an object is detected by detecting laser light reflected by another vehicle AB, a person (pedestrian) PE, and a wall WO existing in front of the host vehicle (host vehicle MB). It can be detected. In the figure, qP, qA, and qW indicated by circles indicate detection points of the respective objects.

  Therefore, the radar memory stores the distance to the measurement point and the reflection intensity for each direction in which the reflected wave is observed. In addition, although reflected waves are observed at distances z1 and z3 (person PE and wall WO) at the measurement point of azimuth θ in FIG. 7, the reflection intensity is stored at the measurement points of the respective distances.

Next, the information conversion process of step S2 executed by the conversion processing unit 12 will be described.
In this information conversion process, a process for converting image information and a process for converting detection information are performed.

  As processing for converting this image information, as shown in FIG. 4, edge detection processing for extracting vertical edges, horizontal edges, and edge strengths for luminance image information obtained from the camera 1, and direction vectors are extracted. A direction vector calculation process and an optical flow process for calculating an optical flow are performed.

  Further, as processing for converting detection information, processing for obtaining a relative velocity based on measurement points obtained from detection information from the radar 2 is performed.

  Edge detection in the above-described edge detection processing can be calculated by a convolution (convolution (multiplication)) operation with a Sobel filter or the like.

  FIG. 8 shows an example of a simple Sobel filter, in which (a) and (b) are filters for vertical edges, and (c) and (d) are filters for horizontal edges. By multiplying such a filter and image information, a vertical edge and a horizontal edge can be obtained. Note that these edge strengths can be obtained as absolute values of these values, for example.

Further, the direction vector can be obtained by the following equation (3) when the vertical edge strength is Dx and the horizontal edge strength is Dy.
Direction vector = Dx / Dy (3)

  FIG. 9 shows the relationship between the direction vector angle and the edge strength.

Next, the optical flow will be described.
The optical flow is an arrow connecting a video displayed at a certain point (xc, yc) on the image at a certain time and a point on the image where the video is located after Δt seconds, In general, the movement of a certain point on a certain object captured on the image is shown. Such an optical flow can be obtained by applying any conventionally proposed method such as block matching or gradient method.

This optical flow will be specifically described with reference to FIGS.
7 and 10 show a case where the person PE is stopped and the other vehicle AB is moving forward in the same direction as the host vehicle MB. FIG. 10 shows Δt seconds after the time shown in FIG. Indicates the state. In both figures, (a) is an image of the camera 1 similar to that shown in FIGS. 5 and 6, and (b) shows a state in which the detection range by the radar 2 is viewed from above.

  Here, the values xc1, yc1, and hc1 indicating the person PE in FIG. 7 are smaller than the value of z as the denominator in the above-described equations (1) and (2) after Δt seconds shown in FIG. It becomes larger as the vehicle MB moves forward. Therefore, the arrow of the optical flow is the vanishing point VP (the point at which the forward infinity point on the image is imaged, and the optical axis LZ of the camera 1 is parallel to the road surface RS as in the settings of FIGS. In this case, the image center becomes longer in the direction away from the vanishing point VP.

  Similarly, since the points on the wall WO are also stopped, the optical flow becomes long. Also, these optical flows become arrows with the vanishing point VP as the center and from there toward the outside of the image.

  Therefore, the optical flow of the person PE shown in FIG. 10A is directed downward to the right at the feet, and directed to the right near the head near the center of the image.

  On the other hand, assuming that the other vehicle AB is moving at a constant speed with the host vehicle MB, the distance relationship is substantially constant and the value of z does not change in the above formulas (1) and (2). The length will be shorter.

Next, returning to FIG. 4, the detection information conversion processing from the radar 2 will be described.
As described above, the conversion process of the detection information of the radar 2 is a process for obtaining the relative velocity based on the distance data of the measurement point. This relative speed is obtained when the distance information from the radar 2 is acquired in time series (for example, every 0.1 second), that is, the distance change with respect to the same measurement point for a certain fixed time, that is, the distance change / the fixed time It can be calculated by length.

  Alternatively, a least square error line between the same measurement points may be obtained, and the inclination thereof may be obtained as a relative speed. Because there is variation in distance accuracy, for example, by taking the least square error straight line obtained for distance results every 0.1 seconds, the error of about one measurement point or non-measurement point is complemented. Thus, the relative speed can be obtained stably.

  Next, the object presence determination process in step S3 executed by the object presence determination processing unit 13 will be described.

  This object presence determination is performed in parallel with determination based on image information and determination based on detection information.

  In the first embodiment, the determination based on the image information is performed by determining that an object exists in a portion where the same movement is concentrated by the optical flow.

  And in the area | region where it determined with the object existing on the image, the rectangular area | region surrounding the position where the optical flow was observed is set. For example, in the case of the other vehicle AB shown in FIG. 7A, a rectangle having a width wc0 and a height hc0 is set based on the coordinates (xc0, yc0), and in the case of the person PE, the coordinates (xc1, yc1) are used as a reference. Rectangular areas SG1 and SG2 having a width wc1 and a height hc1 are set. Note that the areas SG1 and SG2 determined to include the object are hereinafter referred to as object determination areas.

  On the other hand, the determination based on the detection information is basically performed by determining that an object exists at that position when some measurement points are observed continuously in time at the same distance. For example, in the first embodiment, when the scanning period of the radar 2 is set to 0.1 msec and measurement points are observed in the same direction and the same distance for 0.3 msec continuously, it is determined that an object exists in the direction. The

  In the first embodiment, when the measurement points qA and qP are azimuthally continuous (continuous in the left-right direction in the figure), like the other vehicle AB and the person PE in FIG. It is determined that the object exists without the above-described time-series observation.

  Furthermore, in the first embodiment, when the reflection intensity is higher than a preset threshold value among the measurement points qA, qP, and qW, the time series continuity and the azimuth continuity are small. Even if it exists, it determines with an object existing in the measurement point.

  That is, in general, noise has a low reflection intensity and often has a single measurement point. From this, when a measurement point with high reflection intensity is observed as in the other vehicle AB or the two-wheeled vehicle MS, it is immediately determined that an object exists.

  Since the reflection intensity depends on the intensity of the laser beam emitted by the radar 2, the threshold value is obtained based on the experimental value obtained by obtaining a general reflection intensity when the detection target object is experimentally measured. Is set. Further, the reflection intensity decreases as the distance from the radar 2 increases. For this reason, the threshold value is set to a smaller value as the distance from the object increases. In this way, when the threshold value is correlated with the distance, the threshold value correlated with the distance may be stored in advance based on the experimental value, or the threshold value may be multiplied by a coefficient corresponding to the distance. Thus, a correlation with the distance may be provided.

Next, the distance change calculated by the distance change calculation processing unit 14 in the change calculation process of step S4 will be described.
FIG. 11 shows a measurement state by the radar 2 in (a), and an image captured by the camera 1 in (b).

  In the radar 2, measurement point groups P0, P1, P2, and P3 are obtained for each of the person PE, the other vehicle AB, the wall WO, and the two-wheeled vehicle MS on the image, and the distances z0, z1, z2, and these are measured. z3 (here, the average value of the distances of the measurement points included in each measurement point group for each measurement point group) is measured.

  The distance change calculation processing unit 14 obtains a change in distance observed continuously in time for an object measured by the radar 2, that is, the measurement point groups P0, P1, P2, and P3.

  Further, the distance change calculation processing unit 14 also obtains a change in reflection intensity (here, an average value of the reflection intensity at each measurement point included in each measurement point group) along with the distance change. Then, when a threshold value is set in advance for the distance change and the reflection intensity change, and the distance change and the reflection intensity change are continuously observed in a range not exceeding the threshold value, It is determined that the same object is continuously measured, and a flag or the like is given to this object.

  On the other hand, when at least one of the change in distance and the change in reflection intensity exceeds a preset threshold value, it is determined that the object being observed has changed.

  That is, the distance change needs to continuously observe the same object. Therefore, in order to check whether the same object is being observed continuously, the threshold of the distance change and the reflection intensity is set as described above, and whether or not the other object is replaced during the observation. Confirmation is made. Here, the distance change threshold is set to a value indicating that the distance has changed at a speed that is not possible as an object to be detected, and a different threshold may be used for each type of object. it can. Similarly, the reflection intensity is also set to a value indicating that an impossible reflection intensity change has occurred on the object to be detected.

  Next, the magnitude change calculated by the magnitude change calculation processing unit 15 will be described. In the first embodiment, the change in size is not the size of the detection target object itself but the change of the size of a part of the object.

  The size of the object imaged by the camera 1 is inversely proportional to the distance. In addition, the change in size on the image is in an inversely proportional relationship with the distance in any part of the object and the entire object.

  For example, when the other vehicle AB is detected and the distance z1 to the other vehicle AB is halved, the size of the other vehicle AB on the image is doubled. Further, even in the number plate portion of the other vehicle AB, the size of the number plate on the image is doubled when the distance is halved as described above.

  In this way, the change in the size of an image of a certain object being observed is the same for the entire object or any part of the object.

  Therefore, in the first embodiment, the calculation target of the size change is not the rectangular object determination areas SG1 and SG2 surrounding the entire object (other vehicle AB, person PE) as shown in FIG. This is performed for a part of the object determination areas SG1 and SG2. An area that is a target of calculation of the change in size is referred to as an attention area in this specification.

  In the example shown in FIG. 11 (b), rectangular attention areas R0, R1, R2, and R3 surrounded by dotted lines in the figure are shown in part of the person PE, other vehicle AB, wall WO, and motorcycle MS on the image. Has been. These attention areas R0, R1, R2, and R3 reliably capture an image of an object based on edge strength and the like in an object determination area (see SG1 and SG2 in FIG. 7) set based on image information. It is a part of the area determined to be.

  That is, the attention area R0 is an area only in the center of the person PE, the attention area R1 is an area only including the rear bumper 11c and the number plate 11d of the other vehicle AB, and the attention area R2 is from the other vehicle AB of the wall WO. The only part of the area that is surely separated, the attention area R3, is an area of only the central part of the motorcycle MS.

  Next, how to obtain the change in size will be described by taking the attention area R1 in the other vehicle AB as an example.

  This change in size is obtained from the distance between the edges in the attention area R1. That is, the attention area R1 includes a rear window 11a, a trunk lid 11b, a rear bumper 11c, and a number plate 11d where a horizontal edge is strongly detected in the other vehicle AB.

  FIG. 12 shows a region of interest R1, where (a) shows a state where the distance from the host vehicle MB is z1 at the time t0, and (b) shows that the host vehicle MB The distance is smaller than z1 (this distance is afz1).

  As shown in FIG. 12, the distance ww1 between the trunk lid 11b and each lower horizontal edge of the rear bumper 11c and the distance ww2 between the rear window 11a and each lower horizontal edge of the rear bumper 11c correspond to changes in the relative distance. Change.

  That is, the distance between horizontal edges afww1 and afww2 when distance = afz1 shown in (b) is closer to the distance between horizontal edges ww1 and ww2 when distance = z1 shown in (a). growing.

In this case, the distances z1 and afz1 and the horizontal edge distances ww1, ww2, afww1, and afww2 satisfy the following relationship (3).
z1 / afz1 = afww1 / ww1 = afww2 / ww2 (3)

  In addition, when the wall WO has a clear edge, the size change can be obtained by the change in the distance between the edges in the same manner as described above. However, when a repetitive pattern or texture is observed, as described above. The change in size is determined not by the distance between edges but by the frequency components of the observed repeated patterns and textures.

  As shown in FIG. 11B, FIG. 13 shows an image change corresponding to a change in the distance of the attention area R2 in the case where repeated oblique edges are observed on the wall WO. At the distance = z1 shown in FIG. 13 (a), the interval between the repetitive patterns in the attention area R2 is narrow. On the other hand, when the distance = afz1 shown in FIG. The interval of increases.

  Such a change in the interval between the repeated patterns is obtained as a change in the frequency of the edge in the attention area R2. Therefore, the frequency component of the repetitive edge pattern is obtained by wavelet transform, image FFT transform (fast Fourier transform), or the like.

  In general, when there are many edges like a repetitive pattern, there is a possibility that the combination of edges may be mistaken if there are many edges of the same type, and the object size change based on the change in the edge interval described above is reliable. It is difficult to calculate high. Therefore, in the first embodiment, for a repetitive pattern with a short distance between edges, a change in the frequency f is obtained and this is set as a magnitude change.

In this case, the relationship of the following formula (4) is established between the distances z1 and afz1 and the frequencies f1 and aff1.
z1 / afz1 = f1 / aff1 (4)

  Here, since the frequency only needs to be applied to a type of image in which a strong repetitive pattern is observed when the frequency is decomposed, it may be applied to the frequency f1 at which a strong peak is observed by FFT transformation or the like.

  Next, the object detection process in step S5 in the object detection processing unit 16 will be described. That is, in the object presence determination process in step S3, it is determined that an object exists on the image and on the detection information. However, the object detection process in step S5 includes a “size change” of a certain area where the object is detected on the image and a “distance change” at a measurement point in a certain direction where the object is detected by the radar. Are detected, and “object detection” is determined when there is an object that can be obtained as an inversely proportional relationship between the two transformations.

  Note that the regions of interest R0, R1, R2, and R3 on the image shown in FIG. 11 and the measurement point groups P0, P1, P2, and P3 where an object is detected based on detection information are used for this collation.

  Next, the object type determination process in step S6 executed by the type determination processing unit 17 will be described.

  This object type determination is performed based on the information on the attention areas R0, R1, R2, and R3 and the measurement point groups P0, P1, P2, and P3 used in the object detection process in step S5 described above.

  In the first embodiment, vertical / horizontal edges, edge intensity, direction vector dispersion, reflection intensity, and relative velocity are used as information used for determining the type of object, and the relationship between the information for determination and the type of object Is stored in advance in the database.

  FIG. 14 shows an example of the determination characteristic information. In the figure, “small”, “many”, “weak”, “strong”, “high”, and “low” indicate the characteristics of each information in the detected object, and “0” indicates that it is not used for the type of object determination. ing.

  Therefore, based on whether or not the characteristics of the information of each region of interest R0, R1, R2, R3 and the measurement point group P0, P1, P2, P3 match the type of each object in the determination characteristic information of FIG. The type of object is determined.

  That is, there are many vertical and horizontal edges that are information of the attention area, the edge intensity is strong, the direction vector dispersion is small, and the reflection intensity that is the information of the measurement point group is strong corresponding to the reflection from the reflector. When the value indicates a value and the relative speed is a slow speed corresponding to the preceding vehicle, it is determined that the vehicle is traveling other vehicle AB.

  Also, when the attention area has few vertical and horizontal edges, the edge strength is weak, the direction vector variance is small, the reflection intensity, which is the information held by the measurement point group, is weak, and the relative velocity is high Is determined to be a person (pedestrian) PE.

  Also, when there are few vertical and horizontal edges, which are the information of the attention area, the edge strength is weak, the direction vector variance is large, the reflection intensity, which is the information of the measurement point group, is strong, and the relative speed is slow Is determined to be a traveling motorcycle MS.

  In addition, when the attention area has many vertical edges, few horizontal edges, strong edge strength, small directional vector variance, and high relative speed, which is information held by the measurement point cloud, and low reflection intensity Is determined to be a road structure.

  Furthermore, when it is detected as an object but does not belong to any of these criteria, the type is determined as “other object”.

  As described above, the object detection apparatus according to the first embodiment is based on the comparison between the change in the size of the area on the image captured by the camera 1 and the change in the distance of the measurement point captured by the radar 2. The object is detected, and further, the type of the object is determined based on at least one of the image information and the detection information of the region where the comparison has been made.

  As described above, the object detection apparatus according to the first embodiment can detect various types of objects without being limited only to the “vehicle”. In addition, in detecting such various types of objects, It is possible to determine the type of object with high reliability and no information.

  That is, when extracting an object from image information, the means for extracting the outline of the object or extracting a rectangle including the object as shown in FIG. Or other adjacent objects.

  As described above, when the extracted portion includes information on the background and other adjacent objects in addition to the object to be detected, a portion inconsistent with the detection information by the radar 2 is included. There is a risk that accuracy such as distance may decrease. In addition, when the type of an object is determined based on the information extracted in this way, information regarding an object other than the object is included, so that the type determination accuracy may be lowered.

  Furthermore, in the first embodiment, “size change” is calculated at the time of detecting an object. If information such as the background is included in this way, the edge of the background has an adverse effect, and “size change” is detected. May be erroneously calculated.

  On the other hand, in the first embodiment, as described above, a part of the small area in the object on the image is extracted as the attention area among the objects on the image. There is a low possibility that information on other objects adjacent to each other is included, and the object detection accuracy is improved as compared to the case where the object detection is performed based on the outline of the object or the rectangle including the object as in the conventional case.

  Moreover, in the detection of an object, the “size change” obtained based on the information of a specific small area on the image is collated with the “distance change” of the object existing in the same direction obtained from the detection information. Since the same object is detected by the camera 1 and the radar 2 are collated, the object detected by the camera 1 and the radar 2 can be matched with high accuracy, and the object is further increased. The detection accuracy can be improved.

  Further, in the first embodiment, since the type of the object is determined based on the information of the attention area and the measurement point group that coincides with the measurement point by the radar 2, an object other than the vehicle can also be detected. In addition, this determination is unlikely to include information such as backgrounds and adjacent objects other than the detection target object, and the type of object based on the outline of the object and rectangular information including the object as in the past. Compared with the one that performs the determination, it is possible to improve the determination accuracy of the object type.

  In the first embodiment, the determination of the object type is performed based on the determination characteristic information stored in the database, so that the determination process can be simplified. In addition, since vertical / horizontal edges, edge strength, direction vector dispersion, reflection strength, and relative speed are used for this determination, the person PE, the other vehicle AB, the two-wheeled vehicle MS, and the road structure (wall WO) are identified. Can be determined.

  Next, an object detection apparatus according to Example 2 of the embodiment of the present invention will be described with reference to FIG. In the description of the second embodiment, the same or equivalent parts as those in the first embodiment will be denoted by the same reference numerals and different parts will be mainly described.

  In the object detection apparatus according to the second embodiment, a technique for extracting a region of interest from an object on an image in the size change calculation process in step S4 is different from that in the first embodiment.

  In other words, in the first embodiment, a partial area in the object on the image is extracted. In the second embodiment, first, a rectangular object determination area including an object as shown in FIG. After forming SG1 and SG2, they are subdivided to set a region of interest.

  FIG. 15 shows an example, and the object determination area SG3 including the other vehicle AB is further subdivided into equally divided dividing lines BL in each of the vertical direction and the horizontal direction, and each of the subdivided subdivisions. The activated region is defined as a region of interest TR.

  Then, in the object detection process in step S5, collation based on “size change” and “distance change” is performed, and only the attention region TR having an inversely proportional correlation is set as the type determination region RSG.

  Therefore, in the subsequent type determination process, the image information of the type determination region RSG is used for determination of the other vehicle AB.

  In the second embodiment, the other configuration is the same as that of the first embodiment, and thus the description thereof is omitted.

  In the second embodiment, the setting of the attention area TR simply divides the rectangular object determination area SG including the object, so that the processing is simple and the processing speed is improved or the configuration required for the processing is simplified. Can be planned.

  The object type is determined based on only information of the type determination region RSG in which “size change” and “distance change” correlate among the plurality of attention regions TR roughly set as described above. Since the object type is determined, the object type determination is less likely to include information other than the object to be detected, and the object detection accuracy can be improved, as in the first embodiment, and The accuracy of determining the type of object can be improved.

  Next, an object detection apparatus according to Example 3 of the embodiment of the present invention will be described. In the description of the third embodiment, the same or equivalent parts as those in the first embodiment will be denoted by the same reference numerals and different parts will be mainly described.

  In the object detection apparatus of the third embodiment, the contents of the object detection process are different from those of the first embodiment. That is, in the third embodiment, the following processing is executed in parallel with the object detection processing of the first embodiment, and processing for determining object detection before performing collation based on “size change” and “distance change” is added. did.

  Specifically, in the object presence determination process in step S3, the edge intensity obtained from the image information and the reflection intensity obtained from the detection information are higher than a preset threshold value, and any one of the detection results Even if the reliability is high, it is immediately determined that the object is detected.

  Thereby, it is possible to detect an object even in a situation where detection reliability of either one of the camera 1 and the radar 2 is low due to fog, rain, snow, backlight, or the like.

  Moreover, in the third embodiment, when an object is detected by only one of the camera 1 and the radar 2, a threshold value that can ensure reliability is set, so that only one of the two is detected. Even so, high reliability can be ensured.

  Next, an object detection apparatus according to Example 4 of the embodiment of the present invention will be described. In the description of the fourth embodiment, the same or equivalent parts as those in the first embodiment will be denoted by the same reference numerals and different parts will be mainly described.

  In the fourth embodiment, a part of the contents of the information conversion process of the conversion processing unit 12 in FIG. 3 is different from the first embodiment. That is, in the fourth embodiment, the conversion processing unit 12 performs a process of multiplying image information and detection information by direction differentiation, time series differentiation, and time series integration.

  In addition, the differentiation of an image calculates | requires convolution with the filter shown in FIG. Further, when differentiating the measurement value of the radar 2, for example, a simple process such as obtaining a difference in reflection intensity between adjacent measurement points may be applied. In the time-series differentiation and integration, attention is paid to a certain target pixel or measurement point, and a difference or addition value of the pixel or measurement point of the target point that is continuously input is calculated as a differential value and an integration value. .

  As described above, according to the fourth embodiment, even if a defect occurs in the edge by chance, the integral value is used, so that it is possible to detect such a stable object while preventing such a defect. .

  Further, since the differential value is calculated in advance, a differential image is used instead of the luminance value as information of the object determination areas SG1 and SG2 shown in FIG. 7A, or the luminance value and the differential image are changed. By using both, the detection of the linear component can be performed more accurately.

  Next, an object detection apparatus according to Example 5 of the embodiment of the present invention will be described with reference to FIGS. In the description of the fifth embodiment, the same or equivalent parts as those in the first embodiment will be denoted by the same reference numerals and different parts will be mainly described.

  The fifth embodiment is different from the first embodiment in the type determination processing in the type determination processing unit 17 shown in FIG.

  That is, in the fifth embodiment, for the object type determination, edge orientation dispersion, horizontal edge strength, luminance dispersion, luminance average, and vertical edge strength are used as image information from the camera 1. As detection information, reflection intensity dispersion, reflection intensity average, and luminance dispersion are used, and these are collectively referred to as feature amounts.

  In the fifth embodiment, the ratio of these feature amounts is used for determining the type of object. That is, in the fifth embodiment, determination characteristic information indicating the average feature amount ratio based on the experiment result is stored in a database (not shown) for each object type.

  FIG. 16 shows an example of the determination characteristic information stored in the database. (A) shows the determination characteristic information of the other vehicle AB, and (b) shows the determination characteristic information of the person PE.

  Then, by calculating the ratio of the characteristic amounts of the attention areas R0 to R3 and the measurement point groups P0 to P3 that are actually detected, and comparing the results with the stored determination characteristic information, Make a decision.

  FIG. 17 shows a detection example of the feature amount ratio. In FIG. 17, (a) shows the ratios by type of feature amounts at the measurement point P1 where the other vehicle AB in FIG. 11 is detected and the attention area R1. FIG. 7B shows the ratios of the feature points such as vertical and horizontal edges and reflection intensity obtained in the measurement point group P0 and the attention area R0 of the person PE in FIG.

  As can be seen with reference to FIGS. 16 and 17, the other vehicle AB has many straight edges such as a horizontal line, strong reflection intensity, and small edge direction vector dispersion. Further, the person PE has few edges, but has a large direction vector dispersion and a high reflection intensity. Thus, the ratio of the feature amount varies depending on the type of object.

  Therefore, as in the fifth embodiment, the type of the object can be determined with high accuracy by comparing the ratio of the feature amount between the stored data and the detected object.

  The processing described below is performed when processing the above-described feature amount ratio as a value on the bar graph.

  First, the number of vertical and horizontal edges ECT, the reflection intensity LCT, and the edge direction vector dispersion value EV are calculated. The reference values ECT0, LCT0, and EV0 are set in advance for the number of vertical and horizontal edges, reflection intensity, and edge direction vector dispersion value, respectively.

Next, for each value, the ratio between the calculated value and the reference value, ECT / ECT0 (= RE), LCT / LCT0 (= RL), EV / EV0 (= RV) is calculated.
Next, the sum of the calculation ratios W = RE + RL + RV is calculated.

Further, the ratio of each ratio to the total sum W, that is, RE / W, RL / W, and RV / W are calculated.
Finally, these values RE / W, RL / W, and RV / W are processed as a bar graph as shown in FIG.

  As described above, in the fifth embodiment, it is possible to determine with high accuracy by determining the type of the object based on the ratio of the feature amount.

  Further, even when the type of the object to be discriminated is increased or changed, it can be dealt with by simply changing or increasing the data indicating the feature quantity stored in the database, and the type of the object to be discriminated can be easily changed.

  Next, an object detection apparatus according to Example 6 of the embodiment of the present invention will be described with reference to FIG. In the description of the sixth embodiment, the same or equivalent parts as those in the first embodiment will be denoted by the same reference numerals and different parts will be mainly described.

  The sixth embodiment is different from the first embodiment in setting the region for determining the type of object in the type determination processing unit 17 of FIG. In other words, in the sixth embodiment, there are two types of objects such as a horizontally long object such as the other vehicle AB and an object in which the region indicating the object is largely divided in the image, such as the wall WO, in two places, the edge and the inside of the object. A determination area is set.

  An example of this is shown in FIG. 18, for the other vehicle AB and the wall WO, respectively, end type determination regions TR1 and TR2 arranged at the edge of the object, and internal type determination arranged inside the object. The use areas NR1 and NR2 are set.

  These areas TR1, TR2, NR1, and NR2 are set using the processing described in the second embodiment. That is, first, as described in the second embodiment, a wide range of object determination areas including objects (in this example, the other vehicle AB and the wall WO) are set, and this is subdivided to set the attention area. Then, for each region of interest, “size change” and “distance change” are collated, and the region where the correlation is obtained is set as the type determination region. Up to this point, the processing is the same as in the second embodiment.

  Further, in the sixth embodiment, from the obtained type determination area, an area including a portion having a strong edge indicated by the edge of the object, and an area arranged inside the edge indicating the edge are Extracted. The former areas extracted in this way are the end type determination areas TR1 and TR2, and the latter areas are the internal type determination areas NR1 and NR2.

  It should be noted that an elongated object such as the person PE or the two-wheeled vehicle MS or an originally small object is likely to include a lot of background if an area including the object end is provided. Therefore, in the case of long objects and small objects other than the above-mentioned horizontally long and largely separated objects in the image, the feature amount is not present at the object end, and the type determination area set inside the object as in the second embodiment. An object is detected and its type is determined using only this information.

  The object type determination is performed based on the determination characteristic information indicating the ratio of the feature amount described in the fifth embodiment. In this case, the ratio of the feature amount is set differently between the edge portion and the inside depending on the type of the object.

  That is, in the other vehicle AB, in the image information, the object end has a strong edge, while the inside has a portion having an edge such as a bumper and a portion having no edge such as a window glass surface or a body surface. To do.

  In addition, as with other vehicles AB, the artifacts such as the wall WO and the building outside the figure have linearity at the object end, but the interior does not have an edge or texture, or the edge or texture is not. Even if it is, it may be weak.

  Therefore, even for the same type of object, the determination characteristic information is set differently between the edge portion and the inside.

  As described above, in the object detection apparatus according to the sixth embodiment, the edge type determination regions TR1 and TR2 set in the edge and inside of the object and the internal type determination regions NR1 and NR2 are used. The type of the object is determined based on the obtained image information. For this reason, the feature of the object can be grasped more clearly, and the object detection accuracy and the type determination accuracy can be improved.

  Next, an object detection apparatus according to Example 7 of the embodiment of the present invention will be described with reference to FIGS. In the description of the seventh embodiment, the same or equivalent parts as those of the first embodiment will be denoted by the same reference numerals and different parts will be mainly described.

  In the seventh embodiment, the object detection processing unit 16 shown in FIG. 3 executes the same direction processing when there are measurement points indicating a plurality of objects at the same direction and different distances. That is, when there are a plurality of objects in the same azimuth, and the entire object behind is not hidden behind the object in front, these objects existing behind are also measured by the radar 2.

  FIG. 19A shows a measurement example of such a radar 2, in the direction of the azimuth α, in order from the front side, the measurement point group P00 of the person PE, the measurement point group P01 of the other vehicle AB, the wall WO Measurement point group P02 exists at distances z0, z1, and z2 in the direction α. This shows a measurement example similar to that shown in FIG.

  FIG. 19B shows an image captured by the camera 1, and Hα in the figure indicates an area on the image corresponding to the azimuth α of the radar 2. As shown in FIG. 19B, in the area Hα on the image, the other vehicle AB and a part of the wall WO are hidden behind the person PE in the foreground. In the figure, a vertical dotted line on the image indicates a range in which the unit of the scanning resolution of the radar 2 is associated with the image.

  In the seventh embodiment, when there are a plurality of objects at different distances in the same azimuth α as shown in the figure, the same azimuth processing for distinguishing these objects in the same azimuth α is executed.

  The flowchart of FIG. 20 shows the flow of the object detection process of the seventh embodiment. In step S71, it is determined whether or not there are a plurality of measurement points at different distances in the same direction. In step S72, the same direction process is executed. On the other hand, if there are not a plurality of measurement points at different distances in the same direction, the process proceeds to the normal process of step S73. This normal process is the object detection process described in the first embodiment.

  Next, the same direction processing in step S72 will be described. This azimuth processing is performed by checking “distance change” and “size change” for each measurement point having different distances (for each measurement point group P00, P01, P02 in FIG. 19) on the image in ascending order of distance. This is a process of setting an object region indicating the object.

  That is, first, the “distance change” of the measurement point group P00 of the closest object (person PE) is collated with the size change of each region of interest, and the region of interest has a correlation (inverse proportional relationship). A region including the direction α is defined as an object region R00 indicating a front object (person PE). In the seventh embodiment, the object area is set as a rectangular area.

  Next, in the azimuth α, the “distance change” of the measurement point group P01 of the second closest object is collated with the “size change” of the remaining attention area. In this case, the object area R00 includes the object area R00. The attention area is excluded.

  Furthermore, when a region of interest correlating with the “distance change” in the measurement point group P01 of this second closest object (in this example, the other vehicle AB) is determined, an object region R01 including this region of interest is set. This object region R01 is set so as not to overlap with the object region R00 of the person PE set previously.

  Next, for the third closest measurement point group P02, the same distance change and size change as above are collated for the attention area outside the first and second object areas R00 and R01. , The object region R02 of the object is set.

  As described above, when the object areas R00, R01, R02 are set for each object, the object is detected based on the information contained in the object areas R00, R01, R02 and the measurement point groups P00, P01, P02. Also, the type of object is determined. Since this object detection and type determination are the same processing as in the first embodiment, the description thereof is omitted.

  As described above, in the seventh embodiment, when a plurality of objects are measured in the same direction and at different distances, the object regions R00, R01, and R02 set on the image are reliably images of the object in the foreground. It is distinguished from Therefore, in the subsequent object detection process and type determination process, information held by the object in front is erroneously extracted, and object detection and object type determination are not performed. Therefore, the object detection accuracy is further improved.

  The embodiment of the present invention and Examples 1 to 7 have been described in detail above with reference to the drawings, but the specific configuration is not limited to this embodiment and Examples 1 to 7. Design changes that do not depart from the gist of the present invention are included in the present invention.

  For example, in the first to seventh embodiments, the object detection method and the object detection device of the present invention are implemented by being mounted on a vehicle. However, the present invention is not limited to this, and the present invention is applicable to other than industrial vehicles such as industrial robots. Can do.

Further, in the object presence determination process, the determination based on the image information has shown an example in which the portion where the same movement is concentrated by the optical flow is an object in the first embodiment, but is not limited thereto. For example,
a. A model of an object is defined, a template for object detection is prepared, and detection is performed based on similarity to the template.
b. Using a means for detecting a straight line such as Hough transform, an area where there are many straight lines is defined as an object.
c. Detect by color area.
Or the technique described in Unexamined-Japanese-Patent No. 2004-235711 and Unexamined-Japanese-Patent No. 11-345336 can be used.

  Further, in the second embodiment, the example in which the rectangular object determination region SG3 is subdivided in the figure is shown in the drawing, but the present invention is not limited to this. For example, you may subdivide at random based on the edge etc. which are observed in the rectangular object determination area | region SG3.

It is the schematic which shows the own vehicle MB carrying the object detection apparatus of Example 1 of embodiment of this invention, (a) is the figure seen from the side, (b) is the figure seen from upper direction. It is a flowchart which shows the flow of the object detection control in the object detection apparatus of the said Example 1. FIG. It is a block diagram which shows notionally the part which performs each process performed by control unit CU in the object detection apparatus of the said Example 1. FIG. It is a conceptual diagram explaining the outline of the information preserve | saved in the memory 11 in the object detection control of the object detection apparatus of the said Example 1, and the conversion process in the conversion process part 12. FIG. It is a figure explaining the image information of the camera 1 in the object detection apparatus of the said Example 1, Comprising: (a) shows the state which looked at the object detection apparatus of Example 1 from the side, (b) is the imaging of the camera 1 The brightness | luminance image projected on the surface 1a is shown, (c) has shown the infrared image at the time of using an infrared camera as the camera 1. FIG. It is a figure explaining the image information of the camera 1 in the object detection apparatus of the said Example 1, Comprising: (a) shows the state which looked at the object detection apparatus of Example 1 from upper direction, (b) is the imaging of the camera 1 The brightness | luminance image projected on the surface 1a is shown, (c) has shown the infrared image at the time of using an infrared camera as the camera 1. FIG. It is explanatory drawing which shows the state of the distance detection of the image information and radar of the camera 1 in the object detection apparatus of the said Example 1. FIG. It is the schematic which shows the Sobel filter used by the information conversion process of the image information of the camera 1 in the object detection apparatus of the said Example 1. FIG. In the information conversion process of the image information of the camera 1 in the object detection apparatus of the first embodiment, it is an explanatory diagram of how to obtain an edge direction vector, (a) shows a filter for calculating a vertical edge component, (b) Represents a filter for calculating a horizontal edge component, and (c) represents a relationship between the edge strength and the edge direction vector. It is explanatory drawing which shows the state of the optical flow and the distance detection of a radar in the information conversion process of the image information of the camera 1 in the object detection apparatus of the said Example 1. FIG. FIG. 3 is an explanatory diagram showing image information of the camera 1 and a radar distance detection state in the object detection apparatus of the first embodiment, where (a) shows a measurement state by the radar 2 and (b) shows an image captured by the camera 1. ing. FIG. 4 is an explanatory diagram of a calculation example of “magnitude change” by the attention area R1 in the object detection device of the first embodiment, where (a) shows a state at a time point t0, and (b) shows a time elapsed from t0. Showing the body. FIG. 7 is an explanatory diagram of an example of calculating “size change” in a region of interest R2 in the object detection device of the first embodiment, where (a) shows a state of distance = z1, and (b) shows a distance longer than (a). The approaching distance = afz1 is shown. It is a characteristic view which shows an example of the determination characteristic information used in the object detection apparatus of the said Example 1. It is explanatory drawing explaining the example of a setting of the attention area in the object detection apparatus of Example 2 of embodiment of this invention, (a) shows the state which subdivided the object determination area and set the attention area, (b ) Shows a state in which the type determination area is set. In the object detection apparatus of Example 5 of an embodiment of the invention, it is a characteristic figure showing judgment characteristic information stored in a database, (a) is judgment characteristic information on other vehicles AB, and (b) is person PE. The judgment characteristic information is shown. It is a figure which shows the example of a detection of the ratio of the feature-value in the object detection apparatus of the said Example 5, a) shows the example of detection of other vehicles AB, (b) has shown the example of detection of person PE. It is explanatory drawing explaining the example of a setting of the area | region for type determination of the object detection apparatus of Example 6 of embodiment of this invention. It is a figure which shows the example of a measurement in the object detection apparatus of Example 7 of embodiment of this invention, (a) shows the example of a measurement of the radar 2, (b) has shown the image imaged with the camera 1. FIG. It is a flowchart which shows the flow of the object detection process in the object detection apparatus of Example 7 of embodiment of this invention.

Explanation of symbols

1 Camera (image acquisition means)
1b Lens 1a Imaging surface 2 Radar (detection information acquisition means)
11 Memory 11a Rear window 11c Rear bumper 11b Trunk lid 11d Number plate 12 Conversion processing unit 13 Object presence determination processing unit 14 Distance change calculation processing unit 15 Size change calculation processing unit 16 Object detection processing unit 17 Type determination processing unit CU Control unit ( Object detection means)

Claims (13)

Acquiring image information related to the object in the object detection target area from the image information acquisition means, and acquiring detection information related to the object obtained by scanning the object detection target area with the detection wave from the detection information acquisition means,
Based on the image information, a change in the size of an area indicating the object on the image is obtained, and on the basis of the detection information, a change in the distance between the detection information acquisition unit and the object is obtained,
Check the change in size and the change in distance, and if the correlation between the two is obtained, it is determined that the object is detected,
An object detection method comprising: determining a type of an object based on at least one of image information and detection information of a region where a correlation is obtained. Image information acquisition means for acquiring image information relating to an object in the object detection target area;
Detection information acquisition means for acquiring detection information about an object based on reflection of a detection wave obtained by scanning an object detection target region with a detection wave;
Object detection means for detecting an object based on the image information and detection information;
An object detection device comprising:
The object detection means is
An acquisition process for acquiring the image information and the detection information;
A change calculation process for calculating a change in the size of a region of interest that is an area indicating an object on the image, and calculating a change in a distance between the object on the measurement point obtained from the detection information and the detection information acquisition unit; ,
An object detection process for performing object detection based on the size change and the distance change;
When an object is detected, a type determination process for determining the type of the object based on at least one of the image information in the attention area and the detection information of the measurement point;
The object detection apparatus characterized by performing.   The object detection apparatus according to claim 2, wherein the attention area is a small area obtained by extracting a part of an area indicating an object on an image.   The object detection apparatus according to claim 2, wherein the region of interest is each subdivided region obtained by subdividing a region indicating an object on an image.   5. The object detection apparatus according to claim 2, wherein the size change in the change calculation process is obtained from a change in an interval between a plurality of edges existing in the region of interest.   In obtaining the change in distance in the change calculation process, a threshold value for determining that the measurement object has changed is set to at least one of the distance change and the reflection intensity change of the measurement point to be calculated. The object detection apparatus according to claim 2, wherein it is confirmed that a change in distance is calculated for the same measurement object based on the comparison of the two.   In the object detection process, when at least one of the edge intensity obtained from the image information and the reflection intensity of the detection wave obtained from the detection information is equal to or greater than a predetermined threshold, the magnitude change and the distance change The object detection apparatus according to claim 2, wherein the object detection is determined without performing the correlation determination. In the object detection process, when there are a plurality of measurement points indicating the object at different positions in the same direction, the same direction process is executed.
This azimuth processing first collates the distance change of the nearest measurement point with the size change of the attention area of the object in the same direction and its vicinity, and shows a magnitude change correlated with the distance change. The region of the object on the image that includes the region of interest is set as the foremost object region. After that, the distance change at the next closest measurement point, the same direction excluding the set object region, and the object in the vicinity The process of collating with the size change of the region of interest and setting the object region on the image containing the correlated region of interest as the next closest object region is repeated sequentially for the number of remaining measurement points. The object detection apparatus according to claim 2, wherein the object detection apparatus is a process. The object detection means includes a conversion means for performing differential conversion or integral conversion of at least one of the image information and detection information in a time-series direction or detection information in an arbitrary direction on a scan plane,
9. The object detection apparatus according to claim 2, wherein at least one of differential information and integral information obtained by the conversion unit is used in the type determination process.   10. The type determination area as an area on the image used for determining the type of the object is set at each of an edge portion and an inside of the object on the image. The object detection apparatus according to claim 1.   For the determination of the type of the object, the direction vector and time-series conversion value obtained from the image information of the object edge and the internal type determination area, the intensity dispersion and the intensity average obtained from the detection information, The object detection device according to claim 10, wherein: is used. In the object detection means, determination characteristic information regarding the correlation between the object and the information is stored for each type of object to be determined,
3. The object detection unit according to claim 2, wherein in the type determination process, the type of the object is determined based on a comparison between the image information and the detection information and the determination characteristic information stored in advance. The object detection device according to any one of 10. The determination characteristic information is information indicating a ratio of the image information and the detection information with respect to a predetermined reference value set for each type of object,
In the type determination process, the object detection unit calculates a ratio of acquired information with respect to a reference value, and determines the type of object based on a comparison between the calculated ratio and a stored ratio of information. The object detection apparatus according to claim 12, wherein:

Images