Target stable tracking method, device, radar and storage medium

Abstract

The present application provides a method, device, radar and storage medium for stable target tracking, the method comprising: judging whether the area where the target is located has been switched according to the target point cloud in multiple consecutive frames; if it has been switched, determining the tracking corner points of the target based on the corner point selection rule of the current area; determining the corner point jump value according to the tracking corner points of the target in the current area and the tracking corner points of the previous area; compensating the corner point position information of the tracking corner points of N frames before the area switching according to the corner point jump value, and calculating the speed information of the target based on the compensated corner point position information; estimating the state quantity of the target in the current frame according to the corner point position information, speed information and corner point jump value of the target in the current frame. The present application is based on the principle that the contours of the target detected by radars in different areas are different, and selects different corner points for the target point cloud in different detection areas of the radar to track the target, thereby ensuring the continuity of the target and improving the stability of target tracking.

Description

Technical Field

[0001] The present application relates to the field of radar technology, and in particular to a method, device, radar and storage medium for stable target tracking. Background Art

[0002] Compared with traditional sensors, high-resolution millimeter-wave radar has the advantages of high measurement accuracy, rich detection point cloud, and all-day and all-weather coverage. It has been widely used in L3 and above assisted driving environment perception. In the assisted driving process, stable tracking of the detected target and correct speed estimation are crucial for subsequent planning and decision-making.

[0003] Due to the characteristics of corner radar, existing algorithms usually select fixed corner points from the target point cloud for target tracking. However, when the target is in a blind spot or the target is not fully detected, if fixed corner points are used, for example, the points with the largest longitudinal and transverse coordinates in the target point cloud are selected as tracking corner points, it will cause large position measurement errors in target tracking, resulting in inaccurate velocity filtering or even tracking failure. Summary of the invention

[0004] The present application provides a method, device, radar and storage medium for stable target tracking to solve the problem of unstable target tracking by corner radar in the prior art.

[0005] In a first aspect, the present application provides a method for stable target tracking, comprising:

[0006] Determine whether the area where the target is located has been switched according to the target point cloud of the target detected by the radar in multiple consecutive frames; the area includes a reference area and a non-reference area;

[0007] If it is determined that the area where the target is located has been switched, determining the tracking corner point of the target based on the corner point selection rule of the current area;

[0008] Determining a corner point jump value according to a tracking corner point of the target in a current area and a tracking corner point of the target in a previous area;

[0009] Compensating corner point position information of tracking corner points N frames before region switching according to the corner point jump value, and calculating speed information of the target based on the compensated corner point position information and the corner point position information of tracking corner points in the current region;

[0010] The state quantity of the target in the current frame is estimated according to the corner point position information, speed information and corner point jump value of the target in the current frame.

[0011] In a second aspect, the present application provides a target stable tracking device, comprising:

[0012] An area switching judgment module is used to judge whether the area where the target is located has been switched according to the target point cloud of the target detected by the radar in multiple consecutive frames; the area includes a reference area and a non-reference area;

[0013] A tracking corner point determination module, configured to determine the tracking corner points of the target based on the corner point selection rule of the current area if it is determined that the area where the target is located has been switched;

[0014] A corner point jump value calculation module, used to determine the corner point jump value according to the tracking corner point of the target in the current area and the tracking corner point of the target in the previous area;

[0015] A speed information calculation module, used to compensate the corner point position information of the tracking corner points N frames before the area switching according to the corner point jump value, and calculate the speed information of the target based on the compensated corner point position information and the corner point position information of the tracking corner points in the current area;

[0016] The state estimation module is used to estimate the state quantity of the target in the current frame according to the corner point position information, speed information and corner point jump value of the target in the current frame.

[0017] In a third aspect, the present application provides a radar, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the method described in any possible implementation of the first aspect are implemented.

[0018] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in any possible implementation manner of the first aspect above are implemented.

[0019] The embodiment of the present application provides a method, device, radar and storage medium for stable target tracking. The method determines whether the area where the target is located has been switched based on the target point cloud of the target detected by the radar in multiple consecutive frames; if it is determined that the area where the target is located has been switched, the tracking corner points of the target are determined based on the corner point selection rule of the current area; the corner point jump value is determined based on the tracking corner points of the target in the current area and the tracking corner points of the target in the previous area; the corner point position information of the tracking corner points of N frames before the area switching is compensated according to the corner point jump value, and the speed information of the target is calculated based on the compensated corner point position information and the corner point position information of the tracking corner points in the current area; the state quantity of the target in the current frame is estimated according to the corner point position information, speed information and corner point jump value of the target in the current frame. Based on the principle that the contours of targets detected by radars in different areas are different, this embodiment selects different corner points for target point clouds in different areas of the radar to track the target, thereby ensuring the continuity of the target and improving the stability of target tracking. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0021] Figure 1 is a flowchart of a method for stable target tracking provided by an embodiment of the present application;

[0022] Figure 2 is a schematic diagram of a reference area of ​​a left rear corner radar of the vehicle provided in an embodiment of the present application;

[0023] Figure 3 It is a schematic diagram of target tracking corner points in different areas when the host vehicle and the target vehicle are parallel to each other provided in an embodiment of the present application;

[0024] Figure 4 It is a schematic diagram of target tracking corner points in different areas when the host vehicle is perpendicular to the target vehicle provided by an embodiment of the present application;

[0025] Figure 5 is a schematic diagram of the structure of a target stable tracking device provided in an embodiment of the present application;

[0026] Figure 6 Schematic diagram of a radar provided in an embodiment of the present application. DETAILED DESCRIPTION

[0027] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0028] In order to make the purpose, technical solutions and advantages of the present application clearer, specific embodiments will be described below in conjunction with the accompanying drawings.

[0029] See also Figure 1 , which shows a flowchart of the implementation of the target stable tracking method provided by the embodiment of the present application, which is described in detail as follows:

[0030] S101: judging whether the area where the target is located has been switched according to the target point cloud of the target detected by the radar in multiple consecutive frames; the area includes a reference area and a non-reference area.

[0031] The radar mentioned in this embodiment can be any corner radar of the vehicle. After acquiring each frame of target point cloud, the radar determines the area where the target is located based on the contour of the target point cloud in that frame. If the areas where the target is located in the previous and next frames are inconsistent, it is determined that the area where the target is located has been switched.

[0032] Specifically, the detection area of ​​the radar includes a reference area and a non-reference area. The user can pre-set the reference area range and the non-reference area range in the radar detection area, where the reference area is the area where the radar can fully detect the extended target outline of the unobstructed part of the target, and the non-reference area is the area where the radar cannot fully detect the extended target outline. Figure 2 A schematic diagram showing the reference area and non-reference area of ​​the left rear corner radar of the vehicle is shown in FIG. Figure 2 As shown, the installation position of the left rear corner radar 10 is a, and the detection area is the area below the EK line, wherein the shaded area is the reference area of ​​the left rear corner radar, and the area below the EK line except the reference area is the non-reference area.

[0033] S102: If it is determined that the area where the target is located has been switched, determining the tracking corner points of the target based on the corner point selection rule of the current area.

[0034] In this embodiment, since targets in reality are usually body targets, it would be cumbersome to use the position information of a point cloud of the body target to calculate the speed. Therefore, in the prior art, a point is usually selected from the target point cloud as a tracking corner point, and the target is tracked according to the position coordinates of the tracking corner point.

[0035] At present, when radar performs corner point tracking, it usually selects the tracking point cloud based on the position coordinates of all points in the target point cloud. For example, the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud is selected as the tracking corner point. Figure 2 The four-pointed star position is the tracking corner point of the radar, such as Figure 2 As shown in the figure, since the target is in a moving state, when the target moves toward the detection boundary M of the radar, the target contour that the radar can detect changes from the contour corresponding to the straight line box to the contour corresponding to the dotted line box, and the position information (four-pointed star position) of the point corresponding to the minimum value of the horizontal coordinate and the minimum value of the vertical coordinate in the target point cloud calculated by the radar remains unchanged at two moments. This will cause the problem of unstable target tracking and poor tracking effect.

[0036] Therefore, when the target enters the non-reference area from the reference area, or enters the reference area from the non-reference area, the radar needs to select corresponding tracking corner points for the target based on the new area the target has entered. The tracking corner points can enable the radar to be detected in the area, thereby ensuring that the target speed is relatively stable.

[0037] S103: Determine a corner point jump value according to the tracking corner points of the target in the current area and the tracking corner points of the target in the previous area.

[0038] S104: compensating corner point position information of the tracking corner points N frames before region switching according to the corner point jump value, and calculating speed information of the target based on the compensated corner point position information and the corner point position information of the tracking corner points in the current region.

[0039] In this embodiment, in order to calculate the speed information of the target, this embodiment needs to compensate for the corner point position information of the tracking corner points in the latest frames that are different from the tracking corner points in the current area, and based on the relative position relationship of the corresponding tracking corner points in the two areas, the tracking corner points in the previous frames that are different from the tracking corner points in the current area are transformed into the tracking corner points in the current area, and the corner point position information corresponding to the tracking corner points in each frame after the transformation is determined. The radar can only calculate accurate speed information based on the corner point position information of the same tracking corner point in the target point cloud in the latest frames.

[0040] S105: Estimate the state quantity of the target in the current frame according to the corner point position information, speed information and corner point jump value of the target in the current frame.

[0041] In this embodiment, the state quantity represents the real data of the target, including the distance and azimuth of the target relative to the radar.

[0042] This embodiment is based on the principle that the contours of targets detected by radars in different areas are different. Different corner points of the target point cloud are selected in different areas of the radar to track the target, thereby ensuring the continuity of the target and improving the stability of target tracking.

[0043] In a possible embodiment, the specific implementation process of S101 includes:

[0044] Determine the position information of the target in a plurality of consecutive frames based on the target point cloud;

[0045] If the target is located in the same area in the latest m frames, and the area where the target is located in the previous m+1 frames is inconsistent with the area where the target is located in the latest m frames, then it is determined that the area where the target is located in the current frame has been switched; otherwise, it is determined that the area where the target is located in the current frame has not been switched; m≥1.

[0046] In this embodiment, in order to avoid misjudgment of area switching caused by target jitter, this embodiment determines that the target has switched areas only when the target point clouds of the most recent m frames are all located in the same area and the area where the m frames are located is inconsistent with the area where the point clouds of the previous m+1 frames are located. Otherwise, it is determined that the target has not switched areas.

[0047] When determining m, if m is too small, erroneous switching may occur due to jitter; if m is too large, the entire outline of the target may not be effectively observed during switching, resulting in errors. Therefore, m=3 in this embodiment.

[0048] Specifically, when determining the area where the target is located based on the target point cloud, when the target point cloud is completely in the reference area, the target is determined to be located in the reference area; when the target point cloud enters the non-reference area from the reference area, and the points of the target point cloud are partially in the reference area and partially in the non-reference area, the target is determined to be located in the reference area; if the target point cloud is completely in the non-reference area, the target is determined to be located in the non-reference area; when the target point cloud enters the reference area from the non-reference area, and the points of the target point cloud are partially in the reference area and partially in the non-reference area, the target is determined to be located in the non-reference area.

[0049] Through the above method, this embodiment can not recognize that the target area has changed when the target point cloud is partially in the reference area and partially in the non-reference area. Only when the target point cloud completely enters another area will it be recognized that the target area has changed, thereby further ensuring the stability of target tracking.

[0050] In a possible embodiment, the specific implementation process of S102 includes:

[0051] S201: Determine a corner point selection rule for the current area according to the installation information of the radar and the type of the current area;

[0052] S202: Determine the tracking corner points of the target based on the corner point selection rule of the current area.

[0053] In this embodiment, the corner radars installed at different positions of the vehicle have different corner point selection rules due to the different relative positions of the detection areas and the corner radars.

[0054] Figure 3 A schematic diagram of corner points of the left rear corner radar in different areas when the host vehicle and the target vehicle are traveling in opposite directions provided by this embodiment is shown; Figure 4 FIG. 2 shows a schematic diagram of corner points of the left rear corner radar in different areas when the driving direction of the vehicle provided by the present embodiment is perpendicular to that of the target vehicle; Figure 3 and Figure 4 As shown, the vehicle coordinate system takes O as the origin, the left side of the vehicle as the positive direction of the horizontal coordinate, and the front of the vehicle as the positive direction of the vertical coordinate.

[0055] In a possible embodiment, the specific implementation process of S201 includes:

[0056] In the case where the radar is installed at the left rear corner of the vehicle and the current area is the reference area of ​​the radar, if the target meets any one of the first condition group, the corner point selection rule is to select the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud in the vehicle coordinate system as the tracking corner point; if the target meets any one of the second condition group, the corner point selection rule is to select the point with the largest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point; if the target meets the third condition, the corner point selection rule is to select the point with the smallest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point;

[0057] If the radar is installed at the left rear corner of the vehicle and the current area is the non-reference area of ​​the radar, the corner point selection rule is to select the point with the smallest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point;

[0058] The first condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose ordinate is greater than the radar ordinate; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are no points in the target point cloud whose abscissa is less than the abscissa of the radar; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are points in the target point cloud whose ordinate is greater than the ordinate of the radar;

[0059] The second condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose horizontal coordinates are smaller than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are points in the target point cloud whose horizontal coordinates are smaller than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are no points in the target point cloud whose vertical coordinates are larger than the radar vertical coordinates;

[0060] The third condition includes: the target point cloud is the first frame point cloud of the target detected by the radar and there is no point whose ordinate is greater than the radar ordinate and no point whose ordinate is less than the radar ordinate in the target point cloud;

[0061] In the case where the radar is installed at the right rear corner of the vehicle and the current area is the reference area of ​​the radar, if the target satisfies any one of the fourth condition group, the corner point selection rule is to select the point with the largest horizontal coordinate and the smallest vertical coordinate in the target point cloud in the vehicle coordinate system as the tracking corner point; if the target satisfies any one of the fifth condition group, the corner point selection rule is to select the point with the smallest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point; if the target satisfies the sixth condition, the corner point selection rule is to select the point with the largest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point;

[0062] If the radar is installed at the right rear corner of the vehicle and the current area is the non-reference area of ​​the radar, the corner point selection rule is to select the point with the largest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point;

[0063] The fourth condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose ordinate is greater than the radar ordinate; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are no points in the target point cloud whose abscissa is greater than the radar abscissa; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are points in the target point cloud whose ordinate is greater than the radar ordinate;

[0064] The fifth condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose horizontal coordinates are greater than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are points in the target point cloud whose horizontal coordinates are greater than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are no points in the target point cloud whose vertical coordinates are greater than the radar vertical coordinates;

[0065] The sixth condition includes: the target point cloud is the first frame point cloud of the target detected by the radar and there is no point whose ordinate is greater than the radar ordinate and no point whose ordinate is greater than the radar ordinate in the target point cloud;

[0066] In the case where the radar is installed at the left front corner of the vehicle and the current area is the reference area of ​​the radar, if the target satisfies any one of the seventh condition group, the corner point selection rule is to select the point with the smallest horizontal coordinate and the largest vertical coordinate in the target point cloud in the vehicle coordinate system as the tracking corner point; if the target satisfies any one of the eighth condition group, the corner point selection rule is to select the point with the largest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point; if the target satisfies the ninth condition, the corner point selection rule is to select the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point;

[0067] If the radar is installed at the left front corner of the vehicle and the current area is the non-reference area of ​​the radar, the corner point selection rule is to select the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point;

[0068] The seventh condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose ordinate is less than the radar ordinate; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are no points in the target point cloud whose abscissa is less than the abscissa of the radar; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are points in the target point cloud whose ordinate is less than the ordinate of the radar;

[0069] The eighth condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose horizontal coordinates are smaller than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are points in the target point cloud whose horizontal coordinates are smaller than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are no points in the target point cloud whose vertical coordinates are smaller than the radar vertical coordinates;

[0070] The ninth condition includes: the target point cloud is the first frame point cloud of the target detected by the radar and there is no point whose ordinate is smaller than the ordinate of the radar and no point whose ordinate is smaller than the ordinate of the radar and whose ordinate is smaller than the ordinate of the radar in the target point cloud;

[0071] In the case where the radar is installed at the right front corner of the vehicle and the current area is the reference area of ​​the radar, if the target satisfies any one of the tenth condition group, the corner point selection rule is to select the point with the largest horizontal coordinate and the largest vertical coordinate in the target point cloud in the vehicle coordinate system as the tracking corner point; if the target satisfies any one of the eleventh condition group, the corner point selection rule is to select the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point; if the target satisfies the eleventh condition, the corner point selection rule is to select the point with the largest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point;

[0072] If the radar is installed at the right rear corner of the vehicle and the current area is the non-reference area of ​​the radar, the corner point selection rule is to select the point with the largest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point;

[0073] The tenth condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose ordinate is less than the radar ordinate; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are no points in the target point cloud whose abscissa is greater than the abscissa of the radar; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are points in the target point cloud whose ordinate is less than the ordinate of the radar;

[0074] The eleventh condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose horizontal coordinates are greater than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are points in the target point cloud whose horizontal coordinates are greater than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are no points in the target point cloud whose vertical coordinates are less than the radar vertical coordinates;

[0075] The twelfth condition includes: the target point cloud is the first frame point cloud of the target detected by the radar and there is no point whose ordinate is less than the radar ordinate and no point whose abscissa is greater than the radar abscissa in the target point cloud;

[0076] The coordinates are all located in the vehicle coordinate system, and the vehicle coordinate system is a two-dimensional coordinate system established with the left side of the vehicle as the positive direction of the horizontal axis and the front of the vehicle as the positive direction of the vertical axis.

[0077] In this embodiment, if the current frame target point cloud is not the first frame target point cloud of the target detected by the radar, the direction relationship between the target and the vehicle can be determined based on the heading angle of the target and the heading angle of the vehicle. The direction relationship includes vertical and parallel.

[0078] Specifically, the radar can calculate the target's heading angle based on the state quantity of the target point cloud in the previous N frames.

[0079] In one embodiment of the present invention, the radar determines the tracking corner points of the first frame target point cloud of each detection area according to the above method, and uses the tracking corner points corresponding to the first frame target point cloud as the tracking corner points of other frame target point clouds in the area without switching the area. In this way, the amount of calculation of the radar when tracking the target can be reduced, and the target tracking efficiency of the radar can be improved.

[0080] In a possible embodiment, the corner point jump value includes a horizontal corner point jump value and a vertical corner point jump value; the specific implementation process of S103 includes:

[0081] For each target point cloud in the first m frames of the target point cloud, the difference between the horizontal coordinate of the corresponding tracking corner point in the current region of the target point cloud and the horizontal coordinate of the corresponding tracking corner point in the previous region is calculated as the horizontal coordinate difference of the corner point corresponding to the target point cloud of the current frame; the difference between the vertical coordinate of the corresponding tracking corner point in the current region of the target point cloud and the vertical coordinate of the corresponding tracking corner point in the previous region is calculated as the vertical coordinate difference of the corner point corresponding to the target point cloud of the current frame;

[0082] The horizontal coordinate difference of the corner points corresponding to each frame of the target point cloud is averaged to obtain the horizontal corner point jump value;

[0083] The vertical coordinate differences of the corner points corresponding to the target point cloud in each frame are averaged to obtain the vertical corner point jump value.

[0084] In this embodiment, when each frame of target point cloud is obtained, the corner point position coordinates of the tracking corner points corresponding to the frame point cloud in the reference area and the corner point position coordinates of the tracking corner points corresponding to the frame point cloud in the non-reference area are recorded at the same time. If it is determined that the target enters the non-reference area from the reference area or enters the reference area from the non-reference area, the horizontal coordinate difference and the vertical coordinate difference of the tracking corner points in the reference area and the non-reference area of ​​each frame point cloud in the m frames of point cloud before the determination area switching moment are calculated, and the horizontal coordinate difference and the vertical coordinate difference of the corner points of the m frames of point cloud are averaged respectively to obtain the horizontal corner point jump value and the vertical corner point jump value.

[0085] After obtaining the corner point jump value, determine whether the tracking corner point of the previous m frames of point cloud is the same as the tracking corner point of the current frame of point cloud. If they are the same, the corner point position coordinates are not compensated. If they are different, the corner point jump value is added / subtracted to the corner point position coordinates of the different tracking corner points to obtain the compensated corner point position coordinates of the tracking corner points.

[0086] Specifically, the plus and minus signs when performing compensation are determined based on the driving direction of the vehicle.

[0087] In a possible embodiment, the corner point position information includes a horizontal coordinate and a vertical coordinate of the corner point; the speed information includes a horizontal speed and a vertical speed;

[0088] The specific implementation process of S104 includes:

[0089] Based on the formula calculating a lateral velocity of the target;

[0090] Based on the formula calculating the longitudinal velocity of the target;

[0091] Among them, m represents the number of point cloud frames, timeVector[i] represents the time corresponding to the i-th frame point cloud, and Y i represents the ordinate of the corner point corresponding to the i-th frame point cloud, vy represents the longitudinal velocity, vx represents the lateral velocity, X i Indicates the horizontal coordinate of the corner point corresponding to the i-th frame point cloud.

[0092] In this embodiment, taking the longitudinal speed as an example, after obtaining the corner point position coordinates of each frame of point cloud, the radar saves the corner point position coordinates and time of each frame of point cloud, and records them in the array timeVector{(timeVector[i],Yi)} corresponding to the first m frames of point cloud, i=1,2,…m.

[0093] The linear least squares method is used, that is, the sum of squares of the errors between the measured point (timeVector[i], Yi) and the straight line y=kx+b is minimized, and the derived formula is as follows:

[0094] make Y i is the measured value, Y j is the calculated value, that is, Y j = k*timeVector[i]+b. We can get

[0095] Let φ take partial derivative with respect to k and b, and let the partial derivative equal to 0, and we can get

[0096] Similarly, based on the formula Calculate the lateral velocity of the target.

[0097] In a possible embodiment, the corner point position information includes the corner point abscissa and the corner point ordinate; the speed information includes the lateral speed and the longitudinal speed; the corner point jump value includes the lateral corner point jump value and the longitudinal corner point jump value; the specific implementation process of S105 includes:

[0098] Determine the measurement value of the target in the current frame

[0099] Determine the predicted value of the target in the current frame

[0100] The measured value and predicted value of the target in the current frame are input into the Kalman filter model to estimate the state quantity of the target in the current frame.

[0101] Among them, rx represents the horizontal coordinate of the corner point, ry represents the vertical coordinate of the corner point; vx represents the horizontal speed, vy represents the vertical speed, rxJump represents the horizontal corner point jump value, ryJump represents the vertical corner point jump value, represents the predicted value of the target in the kth frame, x k|k-1 Indicates the predicted value of the horizontal coordinate of the corner point of the target in the kth frame; y k|k-1 Represents the predicted value of the corner point ordinate of the target in the kth frame; Represents the predicted lateral velocity of the target in the kth frame; represents the predicted longitudinal velocity of the target in the kth frame; Z Krepresents the measurement value of the target in the kth frame; G k represents the Kalman gain, H k represents the measurement transfer matrix.

[0102] In this embodiment,

[0103] It can be seen from the above embodiments that this embodiment can avoid the problem of corner point tracking failure caused by the tracking corner points in a certain area changing as the target moves by switching the tracking corner points between the reference area and the non-reference area, thereby ensuring stable target tracking and improving the accuracy of target tracking.

[0104] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0105] The following is an embodiment of the device of the present application. For details not described in detail, please refer to the corresponding method embodiment described above.

[0106] Figure 5 The schematic diagram of the structure of the target stable tracking device provided by the embodiment of the present application is shown. For the convenience of explanation, only the part related to the embodiment of the present application is shown, which is described in detail as follows:

[0107] like Figure 5 As shown, the target stabilization tracking device 5 includes:

[0108] The area switching judgment module 110 is used to judge whether the area where the target is located has been switched according to the target point cloud of the target detected by the radar in multiple consecutive frames; the area includes a reference area and a non-reference area;

[0109] A tracking corner point determination module 120, configured to determine the tracking corner points of the target based on a corner point selection rule of the current area if it is determined that the area where the target is located has been switched;

[0110] A corner point jump value calculation module 130, configured to determine a corner point jump value according to a tracking corner point of the target in a current area and a tracking corner point of the target in a previous area;

[0111] The speed information calculation module 140 is used to compensate the corner point position information of the tracking corner points N frames before the area switching according to the corner point jump value, and calculate the speed information of the target based on the compensated corner point position information and the corner point position information of the tracking corner points in the current area;

[0112] The state estimation module 150 is used to estimate the state quantity of the target in the current frame according to the corner point position information, speed information and corner point jump value of the target in the current frame.

[0113] In a possible embodiment, the area switching determination module 110 includes:

[0114] Determine the position information of the target in a plurality of consecutive frames based on the target point cloud;

[0115] If the target is located in the same area in the latest m frames, and the area where the target is located in the previous m+1 frames is inconsistent with the area where the target is located in the latest m frames, then it is determined that the area where the target is located in the current frame has been switched; otherwise, it is determined that the area where the target is located in the current frame has not been switched; m≥1.

[0116] In a possible embodiment, the tracking corner point determination module 120 includes:

[0117] A corner point selection rule determination unit, used to determine a corner point selection rule for a current area according to the installation information of the radar and the type of the current area;

[0118] The tracking corner point selection unit is used to determine the tracking corner points of the target based on the corner point selection rules of the current area.

[0119] In a possible embodiment, the corner point selection rule determination unit includes:

[0120] In the case where the radar is installed at the left rear corner of the vehicle and the current area is the reference area of ​​the radar, if the target meets any one of the first condition group, the corner point selection rule is to select the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud in the vehicle coordinate system as the tracking corner point; if the target meets any one of the second condition group, the corner point selection rule is to select the point with the largest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point; if the target meets the third condition, the corner point selection rule is to select the point with the smallest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point;

[0121] If the radar is installed at the left rear corner of the vehicle and the current area is the non-reference area of ​​the radar, the corner point selection rule is to select the point with the smallest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point;

[0122] The first condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose ordinate is greater than the radar ordinate; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are no points in the target point cloud whose abscissa is less than the abscissa of the radar; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are points in the target point cloud whose ordinate is greater than the ordinate of the radar;

[0123] The second condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose horizontal coordinates are smaller than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are points in the target point cloud whose horizontal coordinates are smaller than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are no points in the target point cloud whose vertical coordinates are larger than the radar vertical coordinates;

[0124] The third condition includes: the target point cloud is the first frame point cloud of the target detected by the radar and there is no point whose ordinate is greater than the radar ordinate and no point whose ordinate is less than the radar ordinate in the target point cloud;

[0125] In the case where the radar is installed at the right rear corner of the vehicle and the current area is the reference area of ​​the radar, if the target satisfies any one of the fourth condition group, the corner point selection rule is to select the point with the largest horizontal coordinate and the smallest vertical coordinate in the target point cloud in the vehicle coordinate system as the tracking corner point; if the target satisfies any one of the fifth condition group, the corner point selection rule is to select the point with the smallest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point; if the target satisfies the sixth condition, the corner point selection rule is to select the point with the largest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point;

[0126] If the radar is installed at the right rear corner of the vehicle and the current area is the non-reference area of ​​the radar, the corner point selection rule is to select the point with the largest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point;

[0127] The fourth condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose ordinate is greater than the radar ordinate; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are no points in the target point cloud whose abscissa is greater than the radar abscissa; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are points in the target point cloud whose ordinate is greater than the radar ordinate;

[0128] The fifth condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose horizontal coordinates are greater than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are points in the target point cloud whose horizontal coordinates are greater than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are no points in the target point cloud whose vertical coordinates are greater than the radar vertical coordinates;

[0129] The sixth condition includes: the target point cloud is the first frame point cloud of the target detected by the radar and there is no point whose ordinate is greater than the radar ordinate and no point whose ordinate is greater than the radar ordinate in the target point cloud;

[0130] In the case where the radar is installed at the left front corner of the vehicle and the current area is the reference area of ​​the radar, if the target satisfies any one of the seventh condition group, the corner point selection rule is to select the point with the smallest horizontal coordinate and the largest vertical coordinate in the target point cloud in the vehicle coordinate system as the tracking corner point; if the target satisfies any one of the eighth condition group, the corner point selection rule is to select the point with the largest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point; if the target satisfies the ninth condition, the corner point selection rule is to select the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point;

[0131] If the radar is installed at the left front corner of the vehicle and the current area is the non-reference area of ​​the radar, the corner point selection rule is to select the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point;

[0132] The seventh condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose ordinate is less than the radar ordinate; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are no points in the target point cloud whose abscissa is less than the abscissa of the radar; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are points in the target point cloud whose ordinate is less than the ordinate of the radar;

[0133] The eighth condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose horizontal coordinates are smaller than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are points in the target point cloud whose horizontal coordinates are smaller than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are no points in the target point cloud whose vertical coordinates are smaller than the radar vertical coordinates;

[0134] The ninth condition includes: the target point cloud is the first frame point cloud of the target detected by the radar and there is no point whose ordinate is smaller than the ordinate of the radar and no point whose ordinate is smaller than the ordinate of the radar and whose ordinate is smaller than the ordinate of the radar in the target point cloud;

[0135] In the case where the radar is installed at the right front corner of the vehicle and the current area is the reference area of ​​the radar, if the target satisfies any one of the tenth condition group, the corner point selection rule is to select the point with the largest horizontal coordinate and the largest vertical coordinate in the target point cloud in the vehicle coordinate system as the tracking corner point; if the target satisfies any one of the eleventh condition group, the corner point selection rule is to select the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point; if the target satisfies the eleventh condition, the corner point selection rule is to select the point with the largest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point;

[0136] If the radar is installed at the right rear corner of the vehicle and the current area is the non-reference area of ​​the radar, the corner point selection rule is to select the point with the largest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point;

[0137] The tenth condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose ordinate is less than the radar ordinate; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are no points in the target point cloud whose abscissa is greater than the abscissa of the radar; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are points in the target point cloud whose ordinate is less than the ordinate of the radar;

[0138] The eleventh condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose horizontal coordinates are greater than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are points in the target point cloud whose horizontal coordinates are greater than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are no points in the target point cloud whose vertical coordinates are less than the radar vertical coordinates;

[0139] The twelfth condition includes: the target point cloud is the first frame point cloud of the target detected by the radar and there is no point whose ordinate is less than the radar ordinate and no point whose abscissa is greater than the radar abscissa in the target point cloud;

[0140] The coordinates are all located in the vehicle coordinate system, and the vehicle coordinate system is a two-dimensional coordinate system established with the left side of the vehicle as the positive direction of the horizontal axis and the front of the vehicle as the positive direction of the vertical axis.

[0141] In a possible embodiment, the corner point jump value includes a horizontal corner point jump value and a vertical corner point jump value; the corner point jump value calculation module 130 includes:

[0142] For each target point cloud in the first m frames of the target point cloud, the difference between the horizontal coordinate of the corresponding tracking corner point in the current region of the target point cloud and the horizontal coordinate of the corresponding tracking corner point in the previous region is calculated as the horizontal coordinate difference of the corner point corresponding to the target point cloud of the current frame; the difference between the vertical coordinate of the corresponding tracking corner point in the current region of the target point cloud and the vertical coordinate of the corresponding tracking corner point in the previous region is calculated as the vertical coordinate difference of the corner point corresponding to the target point cloud of the current frame;

[0143] The horizontal coordinate difference of the corner points corresponding to each frame of the target point cloud is averaged to obtain the horizontal corner point jump value;

[0144] The vertical coordinate differences of the corner points corresponding to the target point cloud in each frame are averaged to obtain the vertical corner point jump value.

[0145] In a possible embodiment, the corner point position information includes the corner point abscissa and the corner point ordinate; the speed information includes the lateral speed and the longitudinal speed; the speed information calculation module 140 includes:

[0146] Based on the formula calculating a lateral velocity of the target;

[0147] Based on the formula calculating the longitudinal velocity of the target;

[0148] Among them, m represents the number of point cloud frames, timeVector[i] represents the time corresponding to the i-th frame point cloud, and Y i represents the ordinate of the corner point corresponding to the i-th frame point cloud, vy represents the longitudinal velocity, vx represents the lateral velocity, X i Indicates the horizontal coordinate of the corner point corresponding to the i-th frame point cloud.

[0149] In a possible embodiment, the corner point position information includes the corner point abscissa and the corner point ordinate; the speed information includes the lateral speed and the longitudinal speed; the corner point jump value includes the lateral corner point jump value and the longitudinal corner point jump value;

[0150] The state estimation module 150 includes:

[0151] Determine the measurement value of the target in the current frame

[0152] Determine the predicted value of the target in the current frame

[0153] The measured value and predicted value of the target in the current frame are input into the Kalman filter model to estimate the state quantity of the target in the current frame.

[0154] Among them, rx represents the horizontal coordinate of the corner point, ry represents the vertical coordinate of the corner point; vx represents the horizontal speed, vy represents the vertical speed, rxJump represents the horizontal corner point jump value, ryJump represents the vertical corner point jump value, represents the predicted value of the target in the kth frame, x k|k-1 Indicates the predicted value of the horizontal coordinate of the corner point of the target in the kth frame; y k|k-1 Represents the predicted value of the corner point ordinate of the target in the kth frame; Represents the predicted lateral velocity of the target in the kth frame; represents the predicted longitudinal velocity of the target in the kth frame; Z K represents the measurement value of the target in the kth frame; G k represents the Kalman gain, H k represents the measurement transfer matrix.

[0155] Through the above scheme, this embodiment is based on the principle that the contours of targets detected by radars in different areas are different, and different corner points of the target point cloud are selected in different areas of the radar to track the target, thereby ensuring the continuity of the target and improving the stability of target tracking.

[0156] Figure 6 Schematic diagram of the radar provided in the embodiment of the present application. Figure 6 As shown, the radar 6 of this embodiment includes: a processor 60, a memory 61, and a computer program 62 stored in the memory 61 and executable on the processor 60. When the processor 60 executes the computer program 62, the steps in the above-mentioned various target stable tracking method embodiments are implemented, such as Figure 1 Alternatively, when the processor 60 executes the computer program 62, the functions of each module / unit in the above-mentioned device embodiments are realized, for example Figure 5 The functions of the modules 110 to 150 are shown.

[0157] Exemplarily, the computer program 62 may be divided into one or more modules / units, which are stored in the memory 61 and executed by the processor 60 to complete / implement the solution provided by the present application. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program 62 in the radar 6. For example, the computer program 62 may be divided into Figure 5 Modules 110 to 150 are shown.

[0158] The radar 6 may include, but is not limited to, a processor 60 and a memory 61. Those skilled in the art will appreciate that Figure 6 It is only an example of radar 6 and does not constitute a limitation of radar 6. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the radar may also include input and output devices, network access devices, buses, etc.

[0159] The processor 60 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0160] The memory 61 may be an internal storage unit of the radar 6, such as a hard disk or memory of the radar 6. The memory 61 may also be an external storage device of the radar 6, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the radar 6. Further, the memory 61 may also include both an internal storage unit and an external storage device of the radar 6. The memory 61 is used to store the computer program and other programs and data required by the radar. The memory 61 may also be used to temporarily store data that has been output or is to be output.

[0161] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0162] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0163] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0164] In the embodiments provided in the present application, it should be understood that the disclosed devices / radars and methods can be implemented in other ways. For example, the device / radar embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0165] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0166] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0167] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various target stable tracking method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0168] In addition, the embodiments shown in the drawings of the present application or the features of the various embodiments mentioned in this specification are not necessarily understood as independent embodiments. Instead, each feature described in one example of an embodiment can be combined with one or more other desired features from other embodiments to produce other embodiments not described in words or with reference to the drawings.

[0169] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for stable target tracking, characterized in that: include: Determine whether the area where the target is located has been switched according to the target point cloud of the target detected by the radar in multiple consecutive frames; the area includes a reference area and a non-reference area; If it is determined that the area where the target is located has been switched, determining the tracking corner point of the target based on the corner point selection rule of the current area; Determining a corner point jump value according to a tracking corner point of the target in a current area and a tracking corner point of the target in a previous area; Compensating corner point position information of tracking corner points N frames before region switching according to the corner point jump value, and calculating speed information of the target based on the compensated corner point position information and the corner point position information of tracking corner points in the current region; Estimate the state quantity of the target in the current frame according to the corner point position information, speed information and corner point jump value of the target in the current frame; The step of judging whether the area where the target is located has been switched according to the target point cloud of the target detected by the radar in multiple consecutive frames includes: Determine the position information of the target in a plurality of consecutive frames based on the target point cloud; If the target is located in the same area in the latest m frames, and the area where the target is located in the previous m+1 frames is inconsistent with the area where the target is located in the latest m frames, it is determined that the area where the target is located in the current frame has been switched; otherwise, it is determined that the area where the target is located in the current frame has not been switched; m≥1; The determining the tracking corner points of the target based on the corner point selection rule of the current area includes: Determine a corner point selection rule for the current area according to the installation information of the radar and the type of the current area; The tracking corner points of the target are determined based on the corner point selection rule of the current area.

2. The target stable tracking method according to claim 1, characterized in that: The determining of the corner point selection rule of the current area according to the installation information of the radar and the type of the current area includes: In the case where the radar is installed at the left rear corner of the vehicle and the current area is the reference area of ​​the radar, if the target meets any one of the first condition group, the corner point selection rule is to select the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud in the vehicle coordinate system as the tracking corner point; if the target meets any one of the second condition group, the corner point selection rule is to select the point with the largest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point; if the target meets the third condition, the corner point selection rule is to select the point with the smallest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point; If the radar is installed at the left rear corner of the vehicle and the current area is the non-reference area of ​​the radar, the corner point selection rule is to select the point with the smallest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point; The first condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose ordinate is greater than the radar ordinate; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are no points in the target point cloud whose abscissa is less than the abscissa of the radar; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are points in the target point cloud whose ordinate is greater than the ordinate of the radar; The second condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose horizontal coordinates are smaller than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are points in the target point cloud whose horizontal coordinates are smaller than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are no points in the target point cloud whose vertical coordinates are larger than the radar vertical coordinates; The third condition includes: the target point cloud is the first frame point cloud of the target detected by the radar and there is no point whose ordinate is greater than the radar ordinate and no point whose ordinate is less than the radar ordinate in the target point cloud; In the case where the radar is installed at the right rear corner of the vehicle and the current area is the reference area of ​​the radar, if the target satisfies any one of the fourth condition group, the corner point selection rule is to select the point with the largest horizontal coordinate and the smallest vertical coordinate in the target point cloud in the vehicle coordinate system as the tracking corner point; if the target satisfies any one of the fifth condition group, the corner point selection rule is to select the point with the smallest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point; if the target satisfies the sixth condition, the corner point selection rule is to select the point with the largest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point; If the radar is installed at the right rear corner of the vehicle and the current area is the non-reference area of ​​the radar, the corner point selection rule is to select the point with the largest horizontal coordinate and the largest vertical coordinate in the target point cloud as the tracking corner point; The fourth condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose ordinate is greater than the radar ordinate; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are no points in the target point cloud whose abscissa is greater than the radar abscissa; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are points in the target point cloud whose ordinate is greater than the radar ordinate; The fifth condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose horizontal coordinates are greater than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are points in the target point cloud whose horizontal coordinates are greater than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are no points in the target point cloud whose vertical coordinates are greater than the radar vertical coordinates; The sixth condition includes: the target point cloud is the first frame point cloud of the target detected by the radar and there is no point whose ordinate is greater than the radar ordinate and no point whose ordinate is greater than the radar ordinate in the target point cloud; In the case where the radar is installed at the left front corner of the vehicle and the current area is the reference area of ​​the radar, if the target satisfies any one of the seventh condition group, the corner point selection rule is to select the point with the smallest horizontal coordinate and the largest vertical coordinate in the target point cloud in the vehicle coordinate system as the tracking corner point; if the target satisfies any one of the eighth condition group, the corner point selection rule is to select the point with the largest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point; if the target satisfies the ninth condition, the corner point selection rule is to select the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point; If the radar is installed at the left front corner of the vehicle and the current area is the non-reference area of ​​the radar, the corner point selection rule is to select the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point; The seventh condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose ordinate is less than the radar ordinate; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are no points in the target point cloud whose abscissa is less than the abscissa of the radar; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are points in the target point cloud whose ordinate is less than the ordinate of the radar; The eighth condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose horizontal coordinates are smaller than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are points in the target point cloud whose horizontal coordinates are smaller than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are no points in the target point cloud whose vertical coordinates are smaller than the radar vertical coordinates; The ninth condition includes: the target point cloud is the first frame point cloud of the target detected by the radar and there is no point whose ordinate is smaller than the ordinate of the radar and no point whose ordinate is smaller than the ordinate of the radar and whose ordinate is smaller than the ordinate of the radar in the target point cloud; In the case where the radar is installed at the right front corner of the vehicle and the current area is the reference area of ​​the radar, if the target satisfies any one of the tenth condition group, the corner point selection rule is to select the point with the largest horizontal coordinate and the largest vertical coordinate in the target point cloud in the vehicle coordinate system as the tracking corner point; if the target satisfies any one of the eleventh condition group, the corner point selection rule is to select the point with the smallest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point; if the target satisfies the twelfth condition, the corner point selection rule is to select the point with the largest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point; If the radar is installed at the right rear corner of the vehicle and the current area is the non-reference area of ​​the radar, the corner point selection rule is to select the point with the largest horizontal coordinate and the smallest vertical coordinate in the target point cloud as the tracking corner point; The tenth condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose ordinate is less than the radar ordinate; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are no points in the target point cloud whose abscissa is greater than the abscissa of the radar; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are points in the target point cloud whose ordinate is less than the ordinate of the radar; The eleventh condition group includes: the target point cloud is the first frame point cloud of the target detected by the radar and there are points in the target point cloud whose horizontal coordinates are greater than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is parallel to the driving direction of the vehicle and there are points in the target point cloud whose horizontal coordinates are greater than the radar horizontal coordinates; and, the target point cloud is not the first frame point cloud of the target detected by the radar and the target is perpendicular to the driving direction of the vehicle and there are no points in the target point cloud whose vertical coordinates are less than the radar vertical coordinates; The twelfth condition includes: the target point cloud is the first frame point cloud of the target detected by the radar and there is no point whose ordinate is less than the radar ordinate and no point whose abscissa is greater than the radar abscissa in the target point cloud; The coordinates are all located in the vehicle coordinate system, and the vehicle coordinate system is a two-dimensional coordinate system established with the left side of the vehicle as the positive direction of the horizontal axis and the front of the vehicle as the positive direction of the vertical axis.

3. The target stable tracking method according to claim 1, characterized in that: The corner point jump value includes a horizontal corner point jump value and a vertical corner point jump value; and determining the corner point jump value according to the tracking corner point of the target in the current area and the tracking corner point of the target in the previous area includes: For each target point cloud in the first m frames of the target point cloud, the difference between the horizontal coordinate of the corresponding tracking corner point in the current region of the target point cloud and the horizontal coordinate of the corresponding tracking corner point in the previous region is calculated as the horizontal coordinate difference of the corner point corresponding to the target point cloud of the current frame; the difference between the vertical coordinate of the corresponding tracking corner point in the current region of the target point cloud and the vertical coordinate of the corresponding tracking corner point in the previous region is calculated as the vertical coordinate difference of the corner point corresponding to the target point cloud of the current frame; The horizontal coordinate difference of the corner points corresponding to each frame of the target point cloud is averaged to obtain the horizontal corner point jump value; The vertical coordinate differences of the corner points corresponding to the target point cloud in each frame are averaged to obtain the vertical corner point jump value.

4. The target stable tracking method according to claim 1, characterized in that: The corner point position information includes the corner point abscissa and the corner point ordinate; the speed information includes the lateral speed and the longitudinal speed; The calculating the speed information of the target based on the compensated corner point position information and the corner point position information of the tracking corner point in the current area includes: Based on the formula calculating a lateral velocity of the target; Based on the formula calculating the longitudinal velocity of the target; Among them, m represents the number of point cloud frames, timeVector[i] represents the time corresponding to the i-th frame point cloud, and Y i represents the ordinate of the corner point corresponding to the i-th frame point cloud, vy represents the longitudinal velocity, vx represents the lateral velocity, X i Indicates the horizontal coordinate of the corner point corresponding to the i-th frame point cloud.

5. The target stable tracking method according to claim 1, characterized in that: The corner point position information includes the corner point abscissa and the corner point ordinate; the speed information includes the lateral speed and the longitudinal speed; the corner point jump value includes the lateral corner point jump value and the longitudinal corner point jump value; The estimating the state quantity of the target in the current frame according to the corner point position information, speed information and corner point jump value of the target in the current frame includes: Determine the measurement value of the target in the current frame Determine the predicted value of the target in the current frame The measured value and predicted value of the target in the current frame are input into the Kalman filter model to estimate the state quantity of the target in the current frame. Among them, rx represents the horizontal coordinate of the corner point, ry represents the vertical coordinate of the corner point; vx represents the horizontal speed, vy represents the vertical speed, rxJump represents the horizontal corner point jump value, ryJump represents the vertical corner point jump value, represents the predicted value of the target in the kth frame, x k|k-1 Indicates the predicted value of the horizontal coordinate of the corner point of the target in the kth frame; y k|k-1 Represents the predicted value of the corner point ordinate of the target in the kth frame; Represents the predicted lateral velocity of the target in the kth frame; represents the predicted longitudinal velocity of the target in the kth frame; Z K represents the measurement value of the target in the kth frame; G k represents the Kalman gain, H k represents the measurement transfer matrix.

6. A target stable tracking device, characterized in that: include: An area switching judgment module is used to judge whether the area where the target is located has been switched according to the target point cloud of the target detected by the radar in multiple consecutive frames; the area includes a reference area and a non-reference area; A tracking corner point determination module, configured to determine the tracking corner points of the target based on the corner point selection rule of the current area if it is determined that the area where the target is located has been switched; A corner point jump value calculation module, used to determine the corner point jump value according to the tracking corner point of the target in the current area and the tracking corner point of the target in the previous area; A speed information calculation module, used to compensate the corner point position information of the tracking corner points N frames before the area switching according to the corner point jump value, and calculate the speed information of the target based on the compensated corner point position information and the corner point position information of the tracking corner points in the current area; A state estimation module, used to estimate the state quantity of the target in the current frame according to the corner point position information, speed information and corner point jump value of the target in the current frame; The area switching judgment module includes: Determine the position information of the target in a plurality of consecutive frames based on the target point cloud; If the target is located in the same area in the latest m frames, and the area where the target is located in the previous m+1 frames is inconsistent with the area where the target is located in the latest m frames, it is determined that the area where the target is located in the current frame has been switched; otherwise, it is determined that the area where the target is located in the current frame has not been switched; m≥1; The tracking corner point determination module comprises: Determine a corner point selection rule for the current area according to the installation information of the radar and the type of the current area; The tracking corner points of the target are determined based on the corner point selection rule of the current area.

7. A radar comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the target stable tracking method as described in any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the target stable tracking method as described in any one of claims 1 to 5 are implemented.

Images