JP6911312B2 - Object identification device - Google Patents

Description

The present invention relates to an object identification device suitable for use in an automated driving system.

In Patent Document 1, if the inclination of the AF point group constituting the cluster with respect to the traveling direction of the own vehicle is within a preset threshold value, the target represented by the cluster is installed along the road instead of a moving object. A method for determining that the wall-like object is formed is disclosed.

Japanese Unexamined Patent Publication No. 2013-36978

At curved entrances and exits where the curvature of the road is large, the inclination of the wall-shaped object installed along the road with respect to the traveling direction of the own vehicle becomes large. In this case, in the above method, since the inclination of the AF point group with respect to the traveling direction of the own vehicle exceeds the threshold value, there is a risk of erroneously determining that the object is not a wall object even though it is a wall object. be.

The present invention has been made in view of the above-mentioned problems, and an object of the present invention is to provide an object identification device capable of reducing misidentification of a wall-shaped object installed along a road as a moving object. And.

In order to achieve the above object, the object identification device according to the present invention is
Distance measurement data acquisition means for acquiring distance measurement data of an object existing in front of the own vehicle, and
A clustering means for clustering the distance measurement data acquired by the distance measurement data acquisition means, and
For each of the AF point groups targeted by the clustering means, the target size indicating the size is calculated, and the AF point group targeted by the clustering means whose target size is equal to or larger than a predetermined value. The first geometric feature calculation means for calculating the geometric features of
The road boundary in front of the own vehicle is recognized based on the map information of the road on which the own vehicle is traveling and the position information of the own vehicle, or a white line with respect to the photographed image taken in front of the own vehicle. A road boundary recognition means that recognizes the road boundary by recognition processing,
A second geometric feature calculation means for calculating the geometric feature of the road boundary recognized by the road boundary recognition means, and a second geometric feature calculation means.
A preset threshold value is the difference between the geometric feature calculated by the first geometric feature calculation means and the geometric feature calculated by the second geometric feature calculation means. If it is within the range, the distance measuring point group targeted by the clustering means whose target size is equal to or larger than a predetermined value is determined to be a wall-shaped object installed along the road.
Equipped with a,
The clustering means
When the height of the AF point cloud targeted by the clustering in the previous frame is equal to or less than a predetermined threshold value, the position and velocity of the AF point cloud targeted in the previous frame are used in the previous frame. Estimate the position of the target in the current frame related to the targeted AF point cloud, and
It is characterized in that the distance measurement data included in the estimated position range of the target is excluded from the target of the clustering in the current frame .

The determination means is the difference between the geometric feature amount calculated by the first geometric feature amount calculation means and the geometric feature amount calculated by the second geometric feature amount calculation means. Is larger than the threshold value, but when the AF point group targeted by the clustering means does not move along the shape of the road boundary for a predetermined time at a speed equal to or higher than a predetermined speed, the above. The AF point group may be configured to determine that it is a wall-like object installed along the road. According to such a configuration, it is possible to further reduce the misidentification of the wall-shaped object as a moving object.

According to the object identification device according to the present invention, the geometric feature amount of the targeted AF point group is compared with the geometric feature amount of the road boundary, so that the curvature of the road is the identification result of the object. It is possible to suppress the influence on. That is, according to the present invention, it is possible to correctly identify the wall-shaped object installed along the road as a wall-shaped object even at a curved entrance / exit having a large curvature of the road. It is possible to reduce misidentification as a moving object. In addition, when a long vehicle such as a truck or a bus is entering the road by excluding a group of AF points whose target size is less than a predetermined value from the calculation of geometric features. However, it is possible to reduce the misidentification of a long vehicle as a wall-like object. Further, in the clustering means, by excluding specific ranging data from the target of clustering in the frame this time, a stationary object that is not a wall-like object such as a rim stone is excluded from the identification target, and the amount of calculation required for object identification. Can be reduced.

It is a block diagram which shows the structure of the automatic operation system which concerns on Embodiment 1. FIG. It is a figure explaining the function of the distance measurement data acquisition part. It is a figure explaining the function of the clustering part. It is a figure explaining the function of the 1st geometric feature calculation part. It is a figure explaining the function of the road boundary recognition part. It is a figure explaining the function of the 2nd geometric feature calculation part. It is a figure explaining the function of the determination part. It is a flowchart which shows the flow of the object identification process which concerns on Embodiment 1. FIG. It is a figure explaining the function of the road boundary recognition part which concerns on Embodiment 2. FIG. It is a flowchart which shows the flow of the object identification process which concerns on Embodiment 2. It is a figure explaining the method of preventing the misidentification of a wall-shaped object and a long vehicle. It is a flowchart which shows the flow of the object identification process which concerns on Embodiment 3. It is a flowchart which shows the flow of the object identification process which concerns on Embodiment 4.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, when the number, quantity, quantity, range, etc. of each element is referred to in the embodiment shown below, the reference is made unless otherwise specified or clearly specified by the number in principle. The present invention is not limited to the number of the above. In addition, the structure described in the embodiments shown below is not necessarily essential to the present invention, except when explicitly stated or when it is clearly specified in principle.

Embodiment 1
FIG. 1 is a diagram showing a configuration of an automatic driving system according to the first embodiment. The autonomous driving system performs autonomous driving of the vehicle on which it is mounted. The automatic driving system 1 includes a control device 10 that functions as an object identification device, a lidar (LIDAR: Laser Imaging Detection and Ranging) 2 that is an autonomous recognition sensor, a radar 3, and a camera 4. These autonomous recognition sensors are directly connected to the control device 10 or via a communication network such as a CAN (Controller Area Network) constructed in the vehicle. The control device 10 can grasp the situation around substantially the entire vehicle from the object information obtained by the autonomous recognition sensor.

The automatic driving system 1 further includes a GPS (Global Positioning System) receiver 5, a map database 6, a navigation system 7, an HMI (Human Machine Interface) 8, and an actuator 9. The GPS receiver 5 is a means for acquiring position information indicating the position of the own vehicle based on a signal transmitted by a GPS satellite. The map database 6 is formed in, for example, a storage such as an HDD or SSD mounted on a vehicle. The map information included in the map database 6 includes road position information, road shape information, intersection and branch point position information, road lane information, and the like. The navigation system 7 calculates the route on which the own vehicle travels based on the position information of the own vehicle measured by the GPS receiver 5 and the map information of the map database 6, and the calculated route information is transmitted via the HMI 8. It is transmitted to the driver and output to the control device 10. The HMI 8 is an interface for outputting and inputting information between the occupant and the control device 10. The actuator 9 is a device that operates in response to an operation signal from the control device 10 and changes the traveling state of the vehicle according to the operation. The actuator 9 is provided in each of the drive system, the braking system, and the steering system, for example. In addition to these, the automatic driving system 1 is provided with internal sensors such as a vehicle speed sensor and an acceleration sensor for obtaining information on the running state of the own vehicle.

The control device 10 is an ECU (Electronic Control Unit) having at least one CPU, at least one ROM, and at least one RAM. Various data including various programs and maps for automatic operation are stored in the ROM. The program stored in the ROM is loaded into the RAM and executed by the CPU, so that the control device 10 realizes various functions. The control device 10 may be composed of a plurality of ECUs.

In FIG. 1, among the functions for automatic operation of the control device 10, the functions related to object identification are represented by blocks. Illustration of other functions of the control device 10 for automatic operation is omitted.

The control device 10 has a function of recognizing an object existing around the own vehicle and discriminating between a wall-shaped object installed along the road and a moving object. This function includes a distance measurement data acquisition unit 11, a clustering unit 12, a first geometric feature calculation unit 13, a road boundary recognition unit 14, and a second geometric feature calculation unit 15 included in the control device 10. And it is realized by the judgment unit 16. However, these parts 11, 12, 13, 14, 15, and 16 do not exist as hardware in the control device 10, but are realized by software when the program stored in the ROM is executed by the CPU. Will be done.

The distance measurement data acquisition unit 11 acquires distance measurement data of an object existing around the own vehicle from the autonomous recognition sensor. The distance measurement data includes information on the distance to the distance measurement point, information on the height of the distance measurement point, and information on the direction of the distance measurement point. The ranging data acquisition unit 11 uses the information obtained by the rider 2, the information obtained by the radar 3, the information obtained by the camera 4, and the information of a plurality of types of autonomous recognition sensors by sensor fusion. It is possible to acquire distance measurement data by at least one method of using them in combination. The distance measurement data acquisition unit 11 specifies the position of each distance measurement point in the traveling surface based on the acquired distance measurement data. FIG. 2 shows an example of the distribution of AF points on a plane (hereinafter referred to as XY plane) in which the distance in the forward direction with respect to the own vehicle is taken on the X-axis and the position in the lateral direction is taken on the Y-axis. ..

The clustering unit 12 clusters the distance measurement data acquired by the distance measurement data acquisition unit 11. Clustering is performed based on the position and height of each AF point on the X plane. In addition, it is performed so as to maintain continuity with the result of clustering in the previous frame. The specific method of clustering by the clustering unit 12 is not particularly limited. A known clustering method can be used. FIG. 3 shows an example of the result of targeting the AF points distributed on the XY plane by clustering. The rectangular frame indicates a target, and the range-finding point group surrounded by the rectangular frame is the target range-finding point group.

The first geometric feature calculation unit 13 calculates the geometric feature of the AF point cloud targeted by the clustering unit 12. The geometric feature amount includes the slope of the straight line with respect to the traveling direction of the own vehicle when the target range-finding point cloud is approximated by a straight line using a fitting method such as the least squares method. In addition, the geometric feature amount includes the curvature of the curve or the rate of change in curvature when the target range-finding point cloud is approximated by a curve using a fitting method such as the least squares method. The first geometric feature calculation unit 13 calculates at least one of the inclination, the curvature, and the rate of change of curvature. In FIG. 4, a straight line that approximates the targeted AF point cloud is drawn by a dotted line. In the example shown in FIG. 4, the inclination of the AF point cloud with respect to the X axis, specifically, the angle formed by the straight line that approximates the AF point cloud with respect to the X axis is calculated as the first geometric feature amount. Will be done.

The road boundary recognition unit 14 recognizes the road boundary by white line recognition processing for the captured image of the front of the own vehicle. The white line recognition process is performed in the order of detection of edge points from the image of the camera 4, detection of edge line segments, pairing of edge line segments, and selection of white line boundaries. FIG. 5 shows a method of detecting an edge point from the image of the camera 4. The image of the camera 4 is scanned horizontally from left to right to search for an edge point at which the pixel changes from black to white and an edge point at which the pixel changes from white to black. By performing this process vertically from top to bottom, an edge line segment in which edge points are aligned on a straight line is detected. If the edge line segment is that of a white line, a pair of edge line segments can be obtained corresponding to both edges of the white line. Edge line segment pairing is performed to exclude edge line segments that are not white lines and select edge line segments that may be white lines. The position of the edge line segment selected by pairing is compared with the past white line detection position. If there is continuity between the two, the selected edge segment is recognized as a white line, or road boundary.

The second geometric feature calculation unit 15 maps the road boundary recognized by the road boundary recognition unit 14 onto the XY plane, and calculates the geometric feature of the road boundary on the XY plane. The geometric feature calculated by the road boundary recognition unit 14 is the slope of the straight line with respect to the traveling direction of the own vehicle when the point sequence data of the road boundary is approximated by a straight line using a fitting method such as the least squares method. And, the curvature of the curve or the rate of change in curvature when the point sequence data of the road boundary is approximated by a curve using a fitting method such as the least squares method is included. The second geometric feature calculation unit 15 has the same type of geometric features as the geometric features calculated by the first geometric feature calculation unit 13 among the inclination, curvature, and curvature change rate. Calculate the quantity. Further, the calculation of the geometric feature amount by the second geometric feature amount calculation unit 15 is performed on the road boundary in the frame indicating the target set by the clustering unit 12. In FIG. 6, a straight line that approximates the road boundary in the target is drawn as a solid line. In the example shown in FIG. 6, the inclination of the road boundary with respect to the X-axis, specifically, the angle formed by the straight line that approximates the road boundary with respect to the X-axis is calculated as the second geometric feature amount.

The determination unit 16 has a first geometric feature calculated by the first geometric feature calculation unit 13 and a second geometry calculated by the second geometric feature calculation unit 15. Calculate the difference from the target feature amount for each target. For example, when the slope and the curvature are calculated as the geometric features, the determination unit 16 calculates the difference in the slope and the difference in the curvature, respectively. In FIG. 7, a straight line that approximates the targeted AF point cloud is drawn by a dotted line, and a straight line that approximates the road boundary in the target is drawn by a solid line. In the example shown in FIG. 7, the difference between the inclination of the AF point cloud and the inclination of the road boundary with respect to the X-axis, specifically, the angle formed by the straight line that approximates the AF point cloud and the straight line that approximates the road boundary. It is calculated.

If there is no difference between the first geometric feature and the second geometric feature, or if the difference is small, the AF point cloud targeted by the clustering unit 12 is It is likely to be a wall-like object installed along the road. When the absolute value of the difference is within a preset threshold value, the determination unit 16 determines that the targeted AF point group is a wall-shaped object installed along the road. The determination unit 16 determines whether to include the targeted AF point cloud in the tracking process or exclude it from the tracking process according to the determination result.

The control device 10 functions as an object identification device by the functions of the above-mentioned parts 11, 12, 13, 14, 15, and 16. FIG. 8 is a flowchart showing the flow of the object identification process according to the present embodiment. The control device 10 as the object identification device repeatedly executes the process shown in this flowchart at a predetermined cycle.

First, in step S1, the road boundary in front of the own vehicle is recognized by the white line recognition process for the image of the camera 4. Further, the processes of steps S2 and S3 are performed in parallel with the process of step S1. In step S2, the distance measurement data of the objects existing around the own vehicle are acquired by the rider 2 or the radar 3. In step S3, the distance measurement data acquired in step S2 is clustered, and the distance measurement point cloud is targeted.

In step S4, the geometric feature amount of the AF point group targeted in step S3 is calculated for each target. Here, it is assumed that the slope of the AF point group is calculated as the geometric feature of the AF point group. Further, in step S5, the geometric feature amount of the road boundary recognized in step S1 is calculated for each target. The geometric feature calculated here is the same slope as the geometric feature calculated in step S4.

In step S6, the slope of the AF point cloud calculated in step S4 is compared with the slope of the road boundary calculated in step S5. Then, the absolute value of the difference is calculated. In step S7, it is determined whether or not the difference (absolute value) between the slope of the AF point cloud calculated in step S6 and the slope of the road boundary is within the threshold value. If the difference in slope is within the threshold value, step S8 is selected as the next process, and if the difference in slope is larger than the threshold value, step S9 is selected as the next process.

In step S8, the AF point cloud targeted in step S3 is excluded from the tracking process. This is because if there is no clear difference between the inclination of the AF point group and the inclination of the road boundary, the AF point group is likely to be a wall-like object installed along the road. On the contrary, in step S9, the AF point cloud targeted in step S3 is targeted for the tracking process. Then, the target information related to the AF point group is output. The target information includes information such as the position, speed, and acceleration of the target. The target information is used for vehicle operation control such as inter-vehicle distance control and lane change control.

According to the object identification process performed in the above procedure, the geometric feature of the targeted AF point group is compared with the geometric feature of the road boundary, so that the curvature of the road is the object. It is possible to suppress the influence on the identification result. Therefore, according to the automatic driving system according to the present embodiment, it is possible to correctly identify the wall-shaped object installed along the road as a wall-shaped object even at a curved entrance / exit having a large curvature of the road. Therefore, it is possible to reduce the misidentification of a wall-shaped object as a moving object.

Embodiment 2
The automatic driving system according to the present embodiment has the configuration shown in FIG. 1 like the automatic driving system according to the first embodiment. The difference between the automatic driving system according to the present embodiment and the automatic driving system according to the first embodiment lies in the method of recognizing the road boundary by the road boundary recognition unit 14.

In the present embodiment, the road boundary recognition unit 14 recognizes the road boundary by self-position estimation using the image of the camera 4 and the map information. FIG. 9 shows a method of self-position estimation. First, the image of the camera 4 is projected on a plane so that the position of the own vehicle is at the center. Further, a map image around the own vehicle is searched from the map database 6 based on the position of the own vehicle specified by the GPS receiver 5. Then, the camera image projected on the plane and the map image are collated. Since the approximate position of the road boundary can be known from the map image, the point sequence data of the road boundary can be extracted from the camera image by collating the two images.

FIG. 10 is a flowchart showing the flow of the object identification process according to the present embodiment. In the object identification process according to the present embodiment, the process of step S1'is performed instead of the process of step S1 of the first embodiment. In step S1', the road boundary is recognized by self-position estimation using the image of the camera 4 and the map information. The remaining processes from step S2 to step S10 are not changed from the first embodiment. As a road boundary recognition method, it is also possible to use both the recognition of the road boundary by self-position estimation using the image of the camera 4 and the map information and the recognition of the road boundary by the white line recognition process.

Embodiment 3
The automatic driving system according to the present embodiment has the configuration shown in FIG. 1 like the automatic driving system according to the first embodiment. In the automatic driving system according to the present embodiment, new functions for not misidentifying a long vehicle such as a truck or a bus and a wall-shaped object are the first geometric feature calculation unit 13 and the determination unit 16. Is given to.

In the present embodiment, the first geometric feature calculation unit 13 calculates the target size of the AF point cloud targeted by the clustering unit 12. The target size is at least one of the width, length, and ratio of the rectangular frame indicating the target. For the target size, a threshold value for distinguishing between a wall-shaped object and a long vehicle is set. When the target size is less than the threshold value, the first geometric feature calculation unit 13 excludes the AF point cloud indicated by the target from the calculation of the geometric feature. The excluded AF point cloud is likely to be a moving object that is not a wall-like object. Therefore, the determination unit 16 includes the AF point group excluded by the first geometric feature calculation unit 13 as the target of the tracking process, and outputs the target information related to the AF point group.

Further, in the present embodiment, the determination unit 16 determines whether or not the targeted AF point group is moving along the shape of the road boundary for a predetermined time at a speed equal to or higher than a predetermined speed. And a wall-like object. Specifically, in the determination unit 16, the speed of the predetermined threshold value or higher continues for a predetermined time or longer (condition A), and the lateral position error of the target position at each time with respect to the lane center position continues for the predetermined time or longer. It is determined whether or not both of the above (condition B) are satisfied. In the example shown in FIG. 11, the velocity of the targeted AF point cloud is calculated from the change in the position of the target at time T-2, T-1, and T and compared with the threshold velocity, and the lane is used. The lateral position error of the target position with respect to the center position is calculated. The lane center position is calculated from the positions of the white lines on the left and right. If either condition A or condition B is not satisfied, the determination unit 16 determines that the AF point group is likely to be a long vehicle, and the target information related to the AF point group. Stops the output of.

FIG. 12 is a flowchart showing the flow of the object identification process according to the present embodiment. In the object identification process according to the present embodiment, the process of step S101 is performed in addition to the object identification process according to the first embodiment or the second embodiment. The process of step S101 is performed after the process of step S3 and before the process of step S4. In step S101, it is determined whether or not the target size of the AF point cloud targeted in step S3 is equal to or larger than the threshold value. When the target size is equal to or larger than the threshold value, the process proceeds to step S4 to calculate the geometric feature amount (for example, inclination) of the AF point group. However, if the target size is less than the threshold value, steps S4 to S7 are skipped and the process proceeds to step S9, and the AF point cloud targeted in step S3 is subject to tracking processing. By adding the process of step S101, it is possible to prevent a long vehicle such as a truck or a bus from being mistaken for a wall-like object.

Further, in the object identification process according to the present embodiment, the process of step S102 is performed in addition to the object identification process according to the first embodiment or the second embodiment. The process of step S102 is performed after the process of step S9 and before the process of step S10. In step S102, the past movement history of the target targeted for tracking processing is analyzed using the white line information obtained in step S1. Then, if both the above-mentioned condition A and condition B are satisfied, the process proceeds to step S10 and the target information is output. However, if either of the above-mentioned conditions A and B is not satisfied, the output of the target information is stopped. By performing the process of step S102, it is possible to prevent the wall-shaped object from being mistaken for a long vehicle such as a truck or a bus.

Embodiment 4
The automatic driving system according to the present embodiment has the configuration shown in FIG. 1 like the automatic driving system according to the first embodiment. The difference between the automatic driving system according to the present embodiment and the automatic driving system according to the first embodiment lies in the function of reducing the amount of calculation required for object identification given to the clustering unit 12.

In the present embodiment, as a method of reducing the amount of calculation required for object identification, clustering is performed by thinning out the AF points. There are the following two methods for thinning out the AF points. First, as a procedure common to the first method and the second method, the position of the target in the frame this time is estimated by using the position and velocity of the AF point cloud targeted in the previous frame. .. Then, assuming that the object is a rigid body, a target frame having the same height, width, and length as the previous frame is used as the target frame in the current frame.

In the first method, the focus points that fall into the frame of the target in the frame this time are thinned out at regular intervals, or the thinning amount is changed according to the size of the distance (for example, the distance is thinned out). The closer it is, the larger the amount of thinning out) is performed. In the second method, when the size of the target frame of the previous frame, particularly the height, is equal to or less than the threshold value, all the AF points that fall within the target frame of the current frame are excluded from the clustering target. By implementing such a method, it is possible to reduce the amount of calculation required for object identification. In particular, by implementing the second method, it is possible to exclude a stationary object that is not a wall-shaped object such as a curb from the identification target and reduce the amount of calculation required for object identification.

FIG. 13 is a flowchart showing the flow of the object identification process according to the present embodiment. In the object identification process according to the present embodiment, the process of step S201 is performed in addition to the object identification process according to the first embodiment or the second embodiment. The process of step S201 is performed after the process of step S2 and before the process of step S3. In step S201, the AF points are thinned out from the AF points acquired in step S2 by any one of the first method and the second method described above. Then, in step S3, clustering is performed on the thinned AF points. Since the contents after step S4 are not changed from the first embodiment or the second embodiment, the description in the flowchart is omitted.

1 Autonomous driving system 2 Rider 3 Radar 4 Camera 5 GPS receiver 6 Map database 7 Navigation system 8 HMI
9 Actuator 10 Control device (object identification device)
11 Distance measurement data acquisition unit (distance measurement data acquisition means)
12 Clustering unit (clustering means)
13 First geometric feature calculation unit (first geometric feature calculation means)
14 Road boundary recognition unit (road boundary recognition means)
15 Second geometric feature calculation unit (second geometric feature calculation means)
16 Judgment unit (judgment means)

Claims (1)

Distance measurement data acquisition means for acquiring distance measurement data of an object existing in front of the own vehicle, and
A clustering means for clustering the distance measurement data acquired by the distance measurement data acquisition means, and
For each of the AF point groups targeted by the clustering means, the target size indicating the size is calculated, and the AF point group targeted by the clustering means whose target size is equal to or larger than a predetermined value. The first geometric feature calculation means for calculating the geometric features of
The road boundary in front of the own vehicle is recognized based on the map information of the road on which the own vehicle is traveling and the position information of the own vehicle, or a white line with respect to the photographed image taken in front of the own vehicle. A road boundary recognition means that recognizes the road boundary by recognition processing,
A second geometric feature calculation means for calculating the geometric feature of the road boundary recognized by the road boundary recognition means, and a second geometric feature calculation means.
A preset threshold value is the difference between the geometric feature calculated by the first geometric feature calculation means and the geometric feature calculated by the second geometric feature calculation means. If it is within the range, the distance measuring point group targeted by the clustering means whose target size is equal to or larger than a predetermined value is determined to be a wall-shaped object installed along the road.
With
The clustering means
When the height of the AF point cloud targeted by the clustering in the previous frame is equal to or less than a predetermined threshold value, the position and velocity of the AF point cloud targeted in the previous frame are used in the previous frame. Estimate the position of the target in the current frame related to the targeted AF point cloud, and
An object identification device characterized in that the distance measurement data included in the estimated position range of the target is excluded from the target of the clustering in the current frame.

Images