JP6638531B2 - Peripheral object detection device - Google Patents

Description

  The present invention relates to a peripheral object detection device.

  Conventionally, U.S. Patent Application Publication No. 2013/0332061 is known as a technical document relating to a peripheral object detection device. This specification is a technique for detecting an object around the host vehicle using a radar sensor.For a set of objects including at least two objects, a relative position between the objects, a speed difference, an azimuth difference, and the like, What describes whether these objects are separate objects or a single object is described.

US Patent Application Publication No. 2013/0332061

  By the way, in order to realize more appropriate driving support, it is desirable to determine whether an object around the own vehicle is a moving object or a stationary object. However, in the conventional determination using a grid map, when a long track or the like traveling in an adjacent lane is overtaking the own vehicle, the detection point (reflection point) of the radar sensor is set at a position corresponding to the middle of the overtaking truck. Point) is repeatedly detected, and there is a possibility that it may be erroneously determined that a stationary object exists at this position, which has been a problem.

  Therefore, in the present technical field, it is desired to provide a peripheral object detection device that can accurately determine a moving object and a stationary object.

  In order to solve the above-described problems, the present invention provides a method for detecting an object around a host vehicle based on a detection result of a detection point by a radar sensor mounted on the host vehicle, using a grid map preset for the host vehicle. A peripheral object detection device that detects the number of votes that is the number of detection points of the radar sensor included in each grid of the grid map based on the detection result of the detection points by the radar sensor. And a cluster setting unit that sets a cluster, which is an object to be detected, on a grid map by clustering the detection points based on a detection result of the detection points by the radar sensor, and a number of votes for each grid calculated multiple times. A grid discriminator for discriminating a stationary grid and a moving grid from the grid occupied by the cluster based on the total value; Based on the arrangement of the stationary grid and move the grid included in the grid comprises a determining moving object determination unit that the cluster is stationary object or a moving object, the.

  As described above, according to the present invention, a moving object and a stationary object can be accurately determined.

It is a block diagram showing a peripheral object detection device concerning this embodiment. (A) It is a top view showing a grid map at time t-1. (B) It is a top view showing a grid map at time t. 9 is a flowchart illustrating a process of determining a moving object and a stationary object by the peripheral object detection device.

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

  As shown in FIG. 1, a peripheral object detection device 100 according to the present embodiment is mounted on a vehicle such as a passenger car (hereinafter, referred to as “own vehicle”), and detects objects around the own vehicle. The objects include moving objects such as other vehicles (truck, bus, passenger car, motorcycle, etc.), pedestrians, bicycles, and stationary objects such as telephone poles, guardrails, buildings, and parked vehicles.

[Configuration of peripheral object detection device]
The peripheral object detection device 100 includes a radar sensor 1, a vehicle speed sensor 2, a yaw rate sensor 3, and an ECU 10. The radar sensor 1 is a sensor for detecting an object around the host vehicle using radio waves (for example, millimeter waves) or light. The radar sensor 1 detects a detection point (reflection point) by transmitting radio waves or light to the vicinity of the own vehicle and receiving radio waves or light reflected by an object. The radar sensor 1 detects a detection point at each preset detection timing. The radar sensor 1 transmits information on the detected detection point to the ECU 10.

  The vehicle speed sensor 2 is a detector that detects the speed of the host vehicle. As the vehicle speed sensor 2, a wheel speed sensor that is provided for a wheel of the host vehicle or a drive shaft that rotates integrally with the wheel, and detects a rotation speed of the wheel is used. The vehicle speed sensor 2 transmits information on the detected vehicle speed to the ECU 10.

  The yaw rate sensor 3 is a detector that detects a yaw rate (rotational angular velocity) around a vertical axis of the center of gravity of the vehicle. For example, a gyro sensor can be used as the yaw rate sensor 3. The yaw rate sensor 3 transmits information on the detected yaw rate of the host vehicle to the ECU 10.

  The ECU [Electronic Control Unit] 10 is an electronic control unit having a CPU [Central Processing Unit], a ROM [Read Only Memory], a RAM [Random Access Memory], a CAN [Controller Area Network] communication circuit, and the like. The ECU 10 implements various functions by, for example, loading a program stored in the ROM into the RAM and executing the program loaded into the RAM by the CPU. The ECU 10 may include a plurality of electronic control units.

  Next, a functional configuration of the ECU 10 will be described. The ECU 10 includes a vehicle speed recognition unit 11, a yaw rate recognition unit 12, a grid map setting unit 13, a vote count calculation unit 14, a cluster setting unit 15, a grid determination unit 16, and a moving object determination unit 17.

  The vehicle speed recognition unit 11 recognizes the vehicle speed of the own vehicle. The vehicle speed recognition unit 11 recognizes the vehicle speed of the own vehicle based on the detection result of the vehicle speed sensor 2. The yaw rate recognition unit 12 recognizes the yaw rate of the own vehicle. The yaw rate recognition unit 12 recognizes the yaw rate of the own vehicle based on the detection result of the yaw rate sensor 3.

  The grid map setting unit 13 sets a grid map obtained by dividing the periphery of the vehicle into a plurality of grids. The grid map is a map in which the periphery of the vehicle is divided into a grid shape in a plan view and divided into a plurality of grids. The grid means a unit area of a fixed size used in object detection. The grid map setting unit 13 sets a grid map around the own vehicle in an absolute coordinate system by a known method.

  Here, FIG. 2A is a plan view showing a grid map at time t-1. FIG. 2A shows the vehicle M, the object N, the grid map GM, the grid G, and the voting grid Ga at time t-1. The object N is a long track, and is overtaking the host vehicle M in the lane adjacent to the host vehicle M. The voting grid Ga will be described later. FIG. 2A shows an XY orthogonal coordinate system including an X axis extending in the traveling direction of the host vehicle M and a Y axis extending in the vehicle width direction of the host vehicle M.

  If the grid map GM has been set at least once after the start of the engine of the own vehicle M, the grid map setting unit 13 calculates the movement amount of the own vehicle M from the elapsed time based on the vehicle speed of the own vehicle M and the yaw rate of the own vehicle M. And the direction of movement. The grid map setting unit 13 sets a new grid map GM by moving the grid map GM according to the estimation result of the moving amount and the moving direction of the vehicle M. The grid of the new grid map GM and the grid of the past grid map GM are associated with each other.

  The vote number calculation unit 14 calculates the number of votes, which is the number including the detection points of the radar sensor 1 for each grid of the grid map GM, based on the detection result of the detection points by the radar sensor 1. In other words, the vote count calculation unit 14 calculates the number of detection points included in each grid as the number of votes for each grid based on the detection points detected by the radar sensor 1. The voting grid Ga shown in FIG. 2A is a grid (a grid including a detection point) in which the number of votes is not zero at the time t-1. In FIG. 2A, the detection point is detected by the radar sensor 1 of the own vehicle M so as to surround the own vehicle M side of the object N, so that the voting grid Ga appears so as to surround the own vehicle M side of the object N. .

  FIG. 2B is a plan view showing the grid map GM at time t. FIG. 2B shows the voting grid Gb, the stationary grid GS, and the moving grid GV at the time t. The voting grid Gb at time t is a grid including the detection points of the radar sensor 1 at time t. The stationary grid GS and the moving grid GV will be described later.

  The cluster setting unit 15 sets a cluster as an object to be detected on the grid map GM by performing clustering of the detection points based on the detection result of the detection points by the radar sensor 1. The cluster setting unit 15 groups a plurality of detection points and sets a cluster based on the shape of the grouped detection points. A well-known method can be adopted as such clustering. In the situation illustrated in FIG. 2B, the cluster setting unit 15 sets the object N as a cluster on the grid map GM. Hereinafter, it is described as a cluster N.

  The grid discriminating unit 16 discriminates the stationary grid GS and the moving grid GV from the grid occupied by the cluster N based on the total number of votes for each grid calculated a plurality of times. The total value of the number of votes for each grid is, for example, the sum of the number of votes for each grid at time t-1 shown in FIG. 2A and the number of votes for each grid at time t shown in FIG. Value. The grid occupied by the cluster N is a grid that at least partially overlaps the range in which the cluster N is set in the grid map GM, and is a grid having one or more votes. That is, the grid determination unit 16 does not determine a grid having a vote count of zero as a determination target. The grid occupied by the cluster N corresponds to the voting grids Ga and Gb overlapping the cluster N on the grid map GM in the situation shown in FIG.

  The stationary grid GS is a grid in which the total number of votes is equal to or greater than the vote count threshold among the grids occupied by the cluster N. Since the detection point of the radar sensor 1 is detected from a grid on which a stationary object such as a wall is located regardless of the time, the total value of the votes of the grid is a large value. For this reason, the grid discrimination unit 16 discriminates a grid in which the total number of votes is equal to or more than the threshold of the number of votes as the stationary grid GS.

  Specifically, in the situation shown in FIG. 2B, the grid discriminating unit 16 determines a grid in which the voting grid Ga at time t-1 and the voting grid Gb at time t overlap each other, and determines that the total value of the number of votes is the number of votes. It is determined as a static grid GS equal to or larger than the threshold value. Note that the vote count threshold value may be a fixed value, or may be a value that changes according to the vehicle speed of the host vehicle M or the like. A modified example of the vote count threshold value will be described later.

  The moving grid GV is a grid in which the number of votes is smaller than the number-of-votes threshold among the voting grids Ga and Gb occupied by the cluster N. If the moving object moves over time from the grid on which the moving object such as another vehicle is located, the detection point of the radar sensor 1 is no longer detected, and the total value of the vote count of the grid is small. For this reason, the grid discrimination unit 16 discriminates a grid in which the total number of votes is less than the threshold of the number of votes as the moving grid GV.

  Specifically, in the situation shown in FIG. 2B, the grid discrimination unit 16 determines whether the voting grid Gb at time t-1 does not overlap the voting grid Gb at time t-1 and the voting grid Gb at time t. Among them, a grid that does not overlap with the voting grid Ga at the time t-1 is determined as a moving grid GV whose total number of votes is less than the threshold of the number of votes.

  The moving object determination unit 17 determines whether the cluster N is a moving object or a stationary object based on the determination result of the grid determination unit 16. First, the moving object determination unit 17 determines that the cluster N is a moving object when the ratio of the moving grid GV included in the grid occupied by the cluster N is equal to or greater than the determination threshold. The determination threshold may be a fixed value or a value that changes according to the vehicle speed of the host vehicle M or the like.

  When the ratio of the moving grid GV is less than the determination threshold, the moving object determination unit 17 determines whether the cluster N is a moving object based on the arrangement of the stationary grid GS and the moving grid GV included in the grid occupied by the cluster N. It is determined whether the object is a stationary object.

  The moving object determination unit 17 determines that the cluster N is a moving object when the arrangement of the stationary grid GS and the moving grid GV included in the grid occupied by the cluster N satisfies a preset condition. The preset condition is satisfied, for example, in the arrangement of the stationary grid GS and the moving grid GV as shown in FIG.

  Specifically, the moving object determination unit 17 determines that the arrangement of the stationary grid GS and the moving grid GV occupied by the cluster N is in the order of the stationary grid GS and the moving grid GV along the traveling direction (X-axis direction) of the host vehicle M. When the moving grid GV is continuous along the vehicle width direction (Y-axis direction) of the host vehicle M, the cluster N is determined to be a moving object. It is not always necessary that the grids are connected to maintain continuity. The ECU 10 stores a plurality of arrangements (preset conditions) of the stationary grid GS and the moving grid GV when determining the cluster N as a moving object.

  The moving object determination unit 17 determines that the cluster N is a stationary object when the arrangement of the stationary grid GS and the moving grid GV included in the grid occupied by the cluster N does not satisfy a preset condition.

  FIG. 2B shows a case where the cluster N, which is a truck, overtakes the own vehicle M. Even when the cluster N is overtaken by the own vehicle, the stationary grid GS and the moving grid occupied by the cluster N are also shown. The arrangement of the GVs is similar. In this case, the position of the cluster N is closer to the host vehicle M than in the situation shown in FIG. Further, even when the host vehicle M is located in front of the cluster N, the arrangement of the stationary grid GS and the moving grid GV occupied by the cluster N is the same as in FIG.

[Determination processing of moving object and stationary object by peripheral object detection device]
Hereinafter, determination processing of a moving object and a stationary object by the peripheral object detection device 100 according to the present embodiment will be described with reference to FIG. FIG. 3 is a flowchart illustrating a process of determining a moving object and a stationary object by the peripheral object detection device 100.

  As shown in FIG. 3, the ECU 10 sets the grid map GM by the grid map setting unit 13 in S10. If the grid map GM has been set at least once after the start of the engine of the own vehicle M, the grid map setting unit 13 moves the previous grid map GM based on the vehicle speed of the own vehicle M and the yaw rate of the own vehicle M. A new grid map GM is set. When the ECU 10 sets the grid map GM, the ECU 10 proceeds to S12.

  In S12, the ECU 10 calculates the number of votes for each grid by the vote number calculation unit 14. The vote number calculation unit 14 calculates the number of votes for each grid based on the detection result of the detection point by the radar sensor 1 at each detection timing of the radar sensor 1. The ECU 10 proceeds to S14 when the number of votes for each grid is calculated at least twice.

  In S14, the ECU 10 sets the cluster N, which is the object to be detected, on the grid map GM by the cluster setting unit 15. The cluster setting unit 15 sets a cluster N by grouping a plurality of detection points by well-known clustering. When the cluster N is set, the ECU 10 proceeds to S16.

  In S16, the ECU 10 causes the grid discriminating unit 16 to discriminate the stationary grid GS and the moving grid GV from the grid occupied by the cluster N. The grid discriminating unit 16 discriminates, among the grids occupied by the cluster N, a grid having a total vote count equal to or greater than the vote count threshold value as the stationary grid GS. On the other hand, the grid discrimination unit 16 discriminates, among the grids occupied by the cluster N, a grid whose total number of votes is less than the threshold of the number of votes as the moving grid GV. It should be noted that the grid determination unit 16 does not determine a grid for which the number of votes is zero as a determination target. When all the grids to be determined have been determined, the ECU 10 proceeds to S18.

  In S18, the ECU 10 causes the moving object determination unit 17 to determine whether the ratio of the moving grid GV included in the grid occupied by the cluster N is equal to or greater than a determination threshold. When it is determined that the ratio of the moving grid GV included in the grid occupied by the cluster N is less than the determination threshold (S18: NO), the ECU 10 proceeds to S20. When it is determined that the ratio of the moving grid GV included in the grid occupied by the cluster N is equal to or greater than the determination threshold (S18: YES), the ECU 10 proceeds to S22.

  In S20, the ECU 10 causes the moving object determination unit 17 to determine whether the arrangement of the stationary grid GS and the moving grid GV included in the grid occupied by the cluster N satisfies a preset condition. For example, in the case of the arrangement shown in FIG. 2B, the moving object determination unit 17 determines that the arrangement of the stationary grid GS and the moving grid GV included in the grid occupied by the cluster N satisfies a preset condition. I do.

  When it is determined that the arrangement of the stationary grid GS and the moving grid GV included in the grid occupied by the cluster N satisfies a preset condition (S20: NO), the ECU 10 proceeds to S22. When it is determined that the arrangement of the stationary grid GS and the moving grid GV included in the grid occupied by the cluster N does not satisfy the preset condition (S20: NO), the ECU 10 proceeds to S24.

  In S22, the ECU 10 determines the cluster N as a moving object by the moving object determining unit 17. After that, the processing for the cluster N ends. Similarly, in S24, the ECU 10 causes the moving object determination unit 17 to determine the cluster N as a stationary object. After that, the processing for the cluster N ends.

  According to the peripheral object detection device 100 according to the present embodiment described above, based on the detection result of the detection point by the radar sensor 1, when detecting the object around the own vehicle M using the grid map GM, the detection point is detected. Since it is determined whether the object is a moving object or a stationary object using the cluster N set by the clustering, the accuracy of the moving object and the stationary object can be compared with the case where the moving object and the stationary object are determined only from the grid map GM. A good judgment can be made.

  As a result, according to the peripheral object detection device 100, the improvement of the recognition accuracy of the surrounding environment of the own vehicle contributes not only to the reduction of the risk but also to the generation of a route plan for efficient and robust driving support. Efficient and robust route planning not only contributes to improving passenger comfort by reducing unnecessary accelerator and brake operations, but also improves fuel efficiency and leads to smooth traffic with other vehicles in the vicinity. Have. In addition, by stopping the tracking processing of the cluster determined to be a stationary object, the processing load on the ECU 10 can be reduced.

  The preferred embodiment of the present invention has been described above, but the present invention is not limited to the above-described embodiment. The present invention can be implemented in various forms including various modifications and improvements based on the knowledge of those skilled in the art, including the above-described embodiment.

  For example, the moving object determination unit 17 does not necessarily need to perform the determination using the ratio of the moving grid GV. That is, S18 in the flowchart shown in FIG. 3 is not essential. In this case, the determination in S20 is performed instead of S18 in the flowchart shown in FIG.

  Further, the vote count threshold value may change according to the vehicle speed of the vehicle M. Specifically, the grid determination unit 16 sets the vote count threshold to a larger value when the vehicle speed of the host vehicle M is less than the predetermined vehicle speed threshold value than when the vehicle speed of the host vehicle M is equal to or higher than the predetermined vehicle speed threshold value. May be changed. The larger the determination threshold is, the easier it is to determine the moving grid GV, and the smaller the vote count threshold, the more likely it is to determine the static grid GS.

  In addition, when there is no grid having the number of votes equal to or more than the predetermined first threshold in the grid occupied by the cluster N, the grid determination unit 16 may change the vote count threshold to a larger value. The predetermined first threshold is a value larger than the changed vote count threshold. On the other hand, when the grid occupied by the cluster N includes a grid having a vote count equal to or greater than the predetermined first threshold, the grid determination unit 16 may change the vote count threshold to a smaller value.

  DESCRIPTION OF SYMBOLS 1 ... radar sensor, 2 ... vehicle speed sensor, 3 ... yaw rate sensor, 11 ... vehicle speed recognition part, 12 ... yaw rate recognition part, 13 ... grid map setting part, 14 ... vote number calculation part, 15 ... cluster setting part, 16 ... grid Judgment unit, 17: moving object determination unit, 100: peripheral object detection device, G: grid, Ga, Gb: voting grid, GM: grid map, GV: moving grid, GS: stationary grid, M: own vehicle, N ... Cluster (object).

Claims (1)

A peripheral object detection device for detecting an object around the host vehicle based on a detection result of a detection point by a radar sensor mounted on the host vehicle using a grid map set in advance with respect to the host vehicle. hand,
Based on the detection results of the detection points by the radar sensor, for each grid of the grid map, a vote count calculation unit that calculates the number of votes that is the number that includes the detection points of the radar sensor,
Based on the detection result of the detection point by the radar sensor, by the clustering of the detection point, a cluster setting unit that sets a cluster that is an object to be detected on the grid map,
A grid discriminator that discriminates a stationary grid and a moving grid from the grid occupied by the cluster based on a total value of the vote count for each grid calculated a plurality of times;
A moving object determination unit that determines whether the cluster is a moving object or a stationary object based on the arrangement of the stationary grid and the moving grid included in the grid occupied by the cluster,
A peripheral object detection device comprising:

Images