An efficient method for judging false targets of vehicle-mounted radar

Abstract

The present invention belongs to the field of target tracking technology, and specifically relates to a method for efficiently judging false targets of vehicle-mounted radar, comprising the following steps: S1: the vehicle-mounted radar detects the target point track, and enters the target track tracking state after track initiation, track association, and track matching; S2: the current position, driving speed, and driving acceleration of the target are tracked using the Kalman filter algorithm; S3: the driving speed of the target is directly estimated using multiple frames of point tracks; S4: the filter value and the estimated value are compared, and if the preset conditions are met, the current target is identified as a false target. The present invention provides a method for efficiently judging false targets of vehicle-mounted radar, and the vehicle-mounted radar tracks the target continuously and stably, and the algorithm implementation process is concise and efficient; therefore, the method can not only effectively improve the efficiency of false target judgment, and has strong applicability and simple implementation, but also can effectively improve the accuracy of radar target tracking.

Description

Technical Field

[0001] The invention belongs to the technical field of target tracking, and in particular relates to a method for efficiently judging false targets of a vehicle-mounted radar. Background Art

[0002] The target detection and tracking process based on vehicle-mounted radar, especially diagonal radar, has problems such as severe jitter in the track initiation and tracking of close-range target detection areas, false alarms caused by splitting of false tracks due to large target tracking, inaccurate detection of stationary targets in the large-angle detection area of ​​the radar installation, or false moving tracks caused by associated tracking errors. These false tracks may lead to false alarms or missed alarms in blind spot detection.

[0003] Common false target judgments are mostly based on the following prior knowledge: 1) There is a maximum and minimum range for close-range target tracking speed; 2) Based on the rigid motion of the vehicle, its lateral acceleration should be very small; 3) The track association and tracking of the real target should be continuous and stable, with very few missed associations or missed detections; 4) The target's starting position, speed, and heading should match the vehicle's driving environment (altitude, low speed, parking, etc.). However, the above assumptions are very broad, so that some false target tracks cannot be truly identified.

[0004] Moreover, the common vehicle-mounted radar target tracking currently is basically that the radar on one side completes target detection and tracking separately. When the target moves from the radar detection area on one side to the radar detection area on the other side, it is easy for the radar target tracking on one side to be interrupted and the target tracking on the other side to start being unstable, resulting in discontinuous tracking. Summary of the invention

[0005] In order to solve the deficiencies of the prior art, the present invention provides an efficient false target judgment method for vehicle-mounted radar, which not only improves the efficiency of false target judgment, has strong applicability and simple implementation, but also can improve the continuity and stability of target tracking through the fusion of left and right vehicle-mounted radars, making the estimation of vx and vy more stable. At the same time, the continuous tracking of the track also ensures that the points involved in the optimal estimation are continuous, which helps to improve the accuracy of false target judgment. In addition, in the target detection and tracking process of the vehicle-mounted radar, the target tracking result of the radar on one side is integrated with the target tracking result of the radar on the other side at the initial stage of target tracking, so that the tracking of the target from the detection area of ​​the radar on one side to the detection area of ​​the radar on the other side is continuous and stable. Compared with the common fusion method of completing target detection and tracking by each single-side radar, the fusion effect of the vehicle-mounted radar target tracking result in the present invention is more stable, and the algorithm implementation process is concise and efficient.

[0006] The technical effects to be achieved by the present invention are achieved through the following technical solutions:

[0007] The vehicle-mounted radar false target efficient judgment method of the present invention comprises the following steps: S1: the vehicle-mounted radar detects the target point track, and enters the target track tracking state after track initiation, track association and track matching; S2: the current position x_filter, y_filter, driving speed vx_filter, vy_filter and driving acceleration ax_filter, ay_filter of the target are tracked by using the Kalman filter algorithm; S3: the driving speed Speed_optimized of the target is directly estimated by using multiple frame point tracks; S4: the filter values ​​vx_filter, vy_filter and the estimated value Speed_optimized are compared, and if the preset conditions are met, the current target is identified as a false target.

[0008] Further, in step S1, the vehicle-mounted radar includes a left-side vehicle-mounted radar arranged on the left side of the vehicle, and a right-side vehicle-mounted radar arranged on the right side of the vehicle.

[0009] Further, in step S1, the target dot track monitored by the left vehicle-mounted radar enters the target track tracking state, including the following steps, S11: obtaining the left vehicle-mounted radar's own detection dot track Dots_Left and the target tracking track Track_Left at the previous moment; S12: simultaneously obtaining the target tracking track Track_Right transmitted by the right vehicle-mounted radar through the left and right communications, and both the track Track_Left and the track Track_Right are stored in the current track tracking list Tracks, and the left vehicle-mounted radar data processing module marks whether the track originates from Track_Left or Track_Right; S13: the dot track Dots_Left is associated with the track tracking list Tracks; S14: after the association is completed, the target track enters the tracking state.

[0010] Furthermore, in step S13, the point track that is not associated with the track will be used as a new target, and a new track will be started, and the track tracking list Tracks will be updated; the point track associated with the track will be matched to the corresponding track, and the track tracking list Tracks will be updated.

[0011] Furthermore, in step S13, for the track with successful point track association, the track mark from Track_Right is updated to the track mark from Track_Left, that is, it is integrated into the left vehicle-mounted radar target tracking track list and enters the tracking at the next moment; the status of the integrated track in the right vehicle-mounted radar target tracking list Track_Right is updated to invisible, and it is no longer displayed and no longer communicated to the left vehicle-mounted radar.

[0012] Further, in step S1, the target dot track monitored by the right vehicle-mounted radar enters the target track tracking state, including the following steps, S11': obtain the right vehicle-mounted radar's own detection dot track Dots_Right and the target tracking track Track_Right at the previous moment; S12': simultaneously obtain the target tracking track Track_Left transmitted by the left vehicle-mounted radar through the left and right communications, and both the track Track_Left and the track Track_Right are stored in the current track tracking list Tracks, and the right vehicle-mounted radar data processing module marks whether the track originates from Track_Left or Track_Right; S13': the dot track Dots_Right is associated with the track tracking list Tracks; S14': after the association is completed, the target track enters the tracking state.

[0013] Furthermore, in step S13', the point track that is not associated with the track will be used as a new target, and a new track will be started, and the track tracking list Tracks will be updated; the point track associated with the track will be matched to the corresponding track, and the track tracking list Tracks will be updated.

[0014] Further, in step S13', for the track with successful point track association, the track mark from Track_Left is updated to the track mark from Trackk_Right, that is, it is integrated into the right-side vehicle-mounted radar target tracking track list and enters the tracking at the next moment; the status of the integrated track in the left-side vehicle-mounted radar target tracking list Track_Left is updated to invisible, and it is no longer displayed and no longer communicated to the right-side vehicle-mounted radar.

[0015] Further, in step S3, using multi-frame point traces to directly estimate the target's driving speed Speed_optimized includes the following steps: S31: obtaining the associated detection points dots within a period of time deltT of the current target tracking track, and collecting the detection speed vr and detection angle δ of these points; S32: assuming that the motion state of the target remains unchanged within the deltT time, that is, the associated detection points dots within the deltT time are all detected in the same motion state vx, vy, ax, ay; S33: based on the measured vr and δ, as well as vx, vy, ax, ay under the same motion state, the least squares method is used to estimate the actual driving speed Speed_optimized of the target within the deltT time.

[0016] Further, in step S4, the filter values ​​vx_filter, vy_filter and the estimated value Speed_optimized are compared. If That is, the current target is identified as a false target.

[0017] In summary, the present invention has at least the following benefits:

[0018] 1. The vehicle-mounted radar false target efficient judgment method of the present invention not only improves the efficiency of false target judgment, has strong applicability and simple implementation, but also can improve the continuity and stability of target tracking through the fusion of left and right vehicle-mounted radars, making the estimation of vx and vy more stable. At the same time, the continuous tracking of the track also ensures that the points participating in the optimal estimation are continuous, which helps to improve the accuracy of false target judgment.

[0019] 2. The method for efficiently judging false targets of vehicle-mounted radar of the present invention integrates the target tracking result of the radar on one side with the target tracking result of the radar on the other side at the initial stage of target tracking of the vehicle-mounted radar during the target detection and tracking process of the vehicle-mounted radar, so that the tracking of the target from the detection area of ​​the radar on one side to the detection area of ​​the radar on the other side is continuous and stable, and the algorithm implementation process is concise and efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a flow chart of a method for efficiently judging false targets of a vehicle-mounted radar in an embodiment of the present invention;

[0021] Figure 2 is a flow chart of a vehicle-mounted radar target tracking and fusion method in an embodiment of the present invention;

[0022] Figure 3 It is a structural schematic diagram of a vehicle-mounted radar tracking a target when the target crosses the rear end of a car in the prior art. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention, not all of them.

[0024] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] Embodiment 1:

[0026] The vehicle-mounted radar false target efficient determination method in this embodiment includes the following steps:

[0027] S1: The vehicle-mounted radar detects the target point track and enters the target track tracking state after track initiation, track association, and track matching;

[0028] S2: Use the Kalman filter algorithm to track the current position x_filter, y_filter, driving speed vx_filter, vy_filter and driving acceleration ax_filter, ay_filter of the target;

[0029] S3: Use multi-frame traces to directly estimate the target's driving speed Speed_optimized;

[0030] S4: Compare the filter values ​​vx_filter, vy_filter and the estimated value Speed_optimized. If the preset conditions are met, the current target is determined to be a false target.

[0031] Preferably, in step S1, the vehicle-mounted radar includes a left vehicle-mounted radar arranged on the left side of the vehicle, and a right vehicle-mounted radar arranged on the right side of the vehicle. The target point track monitored by the left vehicle-mounted radar enters the target track tracking state, including the following steps:

[0032] S11: Obtain the detection point trace Dots_Left of the left vehicle-mounted radar itself and the target tracking track Track_Left at the previous moment;

[0033] S12: simultaneously obtaining the target tracking track Track_Right transmitted by the right vehicle-mounted radar through the left-right communication, storing both the track Track_Left and the track Track_Right in the current track tracking list Tracks, and marking whether the track is from Track_Left or Track_Right by the left vehicle-mounted radar data processing module;

[0034] S13: Dots_Left is associated with the track list Tracks;

[0035] S14: After the association is completed, the target track enters the tracking state.

[0036] In step S13, the specific steps of associating the dot track Dots_Left with the track tracking list Tracks are as follows: the dot track that is not associated with the track will be used as a new target, and a new track will be started, and the track tracking list Tracks will be updated; and the dot track associated with the track will be matched to the corresponding track, and the track tracking list Tracks will be updated. For the track that is successfully associated with the dot track, and the track mark from Track_Right is updated to the track mark from Track_Left, that is, it is fused in the target tracking track list of the left vehicle-mounted radar, and enters the tracking at the next moment; and the status of the fused track in the target tracking list Track_Right of the right vehicle-mounted radar is updated to invisible, which is no longer displayed and no longer communicated to the left vehicle-mounted radar.

[0037] Embodiment 2:

[0038] The vehicle-mounted radar false target efficient determination method in this embodiment includes the following steps:

[0039] S1: The vehicle-mounted radar detects the target point track and enters the target track tracking state after track initiation, track association, and track matching;

[0040] S2: Use the Kalman filter algorithm to track the current position x_filter, y_filter, driving speed vx_filter, vy_filter and driving acceleration ax_filter, ay_filter of the target;

[0041] S3: Use multi-frame traces to directly estimate the target's driving speed Speed_optimized;

[0042] S4: Compare the filter values ​​vx_filter, vy_filter and the estimated value Speed_optimized. If the preset conditions are met, the current target is determined to be a false target.

[0043] Preferably, in step S1, the vehicle-mounted radar includes a left vehicle-mounted radar arranged on the left side of the vehicle, and a right vehicle-mounted radar arranged on the right side of the vehicle. The target point track monitored by the right vehicle-mounted radar enters the target track tracking state, including the following steps:

[0044] S11': Get the detection point trace Dots_Right of the right vehicle-mounted radar itself and the target tracking track Track_Right at the previous moment;

[0045] S12': simultaneously obtain the target tracking track Track_Left transmitted by the left vehicle-mounted radar through the left and right communication, and store the track Track_Left and the track Track_Right in the current track tracking list Tracks. The right vehicle-mounted radar data processing module marks whether the track originates from Track_Left or Track_Right;

[0046] S13': Dots_Right is associated with the track list Tracks;

[0047] S14': After the association is completed, the target track enters the tracking state.

[0048] In step S13', the specific steps of associating the dot track Dots_Right with the track tracking list Tracks are as follows: the dot track that is not associated with the track will be used as a new target, and a new track will be started, and the track tracking list Tracks will be updated; the dot track associated with the track will be matched to the corresponding track, and the track tracking list Tracks will be updated. For the track that is successfully associated with the dot track, and the track mark from Track_Left is updated to the track mark from Trackk_Right, that is, it is fused in the right-hand vehicle radar target tracking track list and enters the tracking at the next moment; and the status of the fused track in the left-hand vehicle radar target tracking list Track_Left is updated to invisible, which is no longer displayed and no longer communicated to the right-hand vehicle radar.

[0049] The common target fusion tracking method in the prior art is to fuse the tracks after the radars on each side complete target detection and tracking. Figure 3 As shown; generally there will be the following situations:

[0050] ① The target is within the FOV (field of view) of the left radar but not within the FOV (field of view) of the right radar. At this time, the left radar detects and tracks the target to form a track, but the target is close to the FOV (field of view) boundary. The point detection quality is not high or the speed and angle measurement are inaccurate, which can easily lead to track interruption.

[0051] ② The target is in the intersection area of ​​the left and right radar FOV (field of view), that is, the target can be detected by both the left and right radars; at this time, the left radar can still track the target track, and the right radar detects the target. The track may not have been formed yet, or the target track may have been started, but the tracking state is quite different from the left tracking state, and the left and right tracks may not match;

[0052] ③ The target has left the left radar FOV (field of view) and is completely within the right radar FOV (field of view). At this time, the left radar target tracking is lost and the right radar has not yet formed a track. At this time, the target tracking is obviously interrupted, or a stable track has been formed, but the motion state is discontinuous.

[0053] Therefore, compared with the above-mentioned target tracking fusion method, the efficient method for judging false targets of the vehicle-mounted radar in the present embodiment 1-2, during the target detection and tracking process of the vehicle-mounted radar, integrates the target tracking result of the radar on one side at the initial stage of target tracking of the radar on the other side, so that the tracking of the target from the radar detection area on one side to the radar detection area on the other side is continuous and stable, and the algorithm implementation process is concise and efficient.

[0054] Embodiment 3:

[0055] The vehicle-mounted radar false target efficient determination method in this embodiment includes the following steps:

[0056] S1: The vehicle-mounted radar detects the target point track and enters the target track tracking state after track initiation, track association, and track matching;

[0057] S2: Use the Kalman filter algorithm to track the current position x_filter, y_filter, driving speed vx_filter, vy_filter and driving acceleration ax_filter, ay_filter of the target;

[0058] S3: Use multi-frame traces to directly estimate the target's driving speed Speed_optimized;

[0059] S4: Compare the filter values ​​vx_filter, vy_filter and the estimated value Speed_optimized. If the preset conditions are met, the current target is determined to be a false target.

[0060] In step S1, the vehicle-mounted radars on both sides perform target tracking fusion through the steps in embodiments 1 and 2; in step S3, the target's driving speed Speed_optimized is directly estimated using multi-frame point traces, including the following steps:

[0061] S31: Obtain the associated detection points dots within a period of time deltT of the current target tracking track, and collect the detection speed vr and detection angle δ of these points;

[0062] S32: Assume that the motion state of the target remains unchanged within deltT time, that is, the associated detection points dots within deltT time are all detected in the same motion state vx, vy, ax, ay;

[0063] S33: Based on the measured vr and δ, as well as vx, vy, ax, and ay in the same motion state, the actual driving speed Speed_optimized of the target within the deltT time is estimated using the least squares method.

[0064] Embodiment 4:

[0065] The vehicle-mounted radar false target efficient determination method in this embodiment includes the following steps:

[0066] S1: The vehicle-mounted radar detects the target point track and enters the target track tracking state after track initiation, track association, and track matching;

[0067] S2: Use the Kalman filter algorithm to track the current position x_filter, y_filter, driving speed vx_filter, vy_filter and driving acceleration ax_filter, ay_filter of the target;

[0068] S3: Use multi-frame traces to directly estimate the target's driving speed Speed_optimized;

[0069] S4: Compare the filter values ​​vx_filter, vy_filter and the estimated value Speed_optimized. If the preset conditions are met, the current target is determined to be a false target.

[0070] In step S1, the vehicle-mounted radars on both sides perform target tracking fusion through the steps in embodiments 1 and 2; in step S3, the target's driving speed Speed_optimized is directly estimated through the steps in embodiment 3; the filter values ​​vx_filter, vy_filter and the estimated value Speed_optimized are compared, and if the result is satisfied, That is, the current target is identified as a false target.

[0071] It can be seen from the technical solutions of the above embodiments that the present invention provides an efficient method for judging false targets of vehicle-mounted radar, which can not only effectively improve the efficiency of false target judgment, but also has strong applicability and simple implementation, and can improve the continuity and stability of target tracking through the fusion of left and right vehicle-mounted radars, making the estimation of vx and vy more stable. At the same time, continuous tracking of the track also ensures that the points participating in the optimal estimation are continuous, which helps to improve the accuracy of false target judgment.

[0072] Although the present invention is described in conjunction with the above specific embodiments, it is obvious that those skilled in the art can make many substitutions, modifications and changes according to the above content. Therefore, all such substitutions, improvements and changes are included in the spirit and scope of the appended claims.

Claims

1. An efficient method for judging false targets of vehicle-mounted radar, characterized in that: The following steps are included: S1: The vehicle-mounted radar detects the target point track and enters the target track tracking state after track initiation, track association, and track matching; S2: Use the Kalman filter algorithm to track the current position x_filter, y_filter, driving speed vx_filter, vy_filter and driving acceleration ax_filter, ay_filter of the target; S3: Use multi-frame traces to directly estimate the target's driving speed Speed_optimized; S4: Compare the filter values ​​vx_filter, vy_filter and the estimated value Speed_optimized. If the preset conditions are met, the current target is considered a false target. In step S1, the target point track monitored by the vehicle-mounted radar enters the target track tracking state, including the following steps: S11: Obtain the detection point trace of the left / right vehicle-mounted radar itself and the target tracking track at the last moment; S12: simultaneously obtain the target tracking track transmitted by the on-board radar on the opposite side through the left and right communication, store the tracks on both sides in the current track tracking list Tracks, and mark which side the track comes from by the left / right on-board radar data processing module; S13: The point track is associated with the track tracking list Tracks; S14: After the association is completed, the target track enters the tracking state.

2. The method for efficiently judging false targets of vehicle-mounted radar according to claim 1 is characterized in that: In step S1 , the vehicle-mounted radar includes a left vehicle-mounted radar disposed on the left side of the vehicle, and a right vehicle-mounted radar disposed on the right side of the vehicle.

3. The method for efficiently judging false targets of vehicle-mounted radar according to claim 2 is characterized in that: In step S1, the target point track monitored by the left vehicle-mounted radar enters the target track tracking state, including the following steps: S11: Obtain the detection point trace Dots_Left of the left vehicle-mounted radar itself and the target tracking track Track_Left at the previous moment; S12: simultaneously obtaining the target tracking track Track_Right transmitted by the right vehicle-mounted radar through the left-right communication, storing both the track Track_Left and the track Track_Right in the current track tracking list Tracks, and marking whether the track is from Track_Left or Track_Right by the left vehicle-mounted radar data processing module; S13: Dots_Left is associated with the track list Tracks; S14: After the association is completed, the target track enters the tracking state.

4. The method for efficiently judging false targets of vehicle-mounted radar according to claim 3 is characterized in that: In step S13, the point track that is not associated with the track will be used as a new target, and a new track will be started, and the track tracking list Tracks will be updated; the point track associated with the track will be matched to the corresponding track, and the track tracking list Tracks will be updated.

5. The method for efficiently judging false targets of vehicle-mounted radar according to claim 3 is characterized in that: In step S13, for the track that is successfully associated with the point track, the track flag from Track_Right is updated to the track flag from Track_Left, that is, the left vehicle radar target tracking track list is integrated and the tracking at the next moment is entered; The status of the fused track in the right vehicle-mounted radar target tracking list Track_Right is updated to invisible, and it is no longer displayed and no longer communicated to the left vehicle-mounted radar.

6. The method for efficiently judging false targets of vehicle-mounted radar according to claim 2 is characterized in that: In step S1, the target point track monitored by the right vehicle-mounted radar enters the target track tracking state, including the following steps: S11': Get the detection point trace Dots_Right of the right vehicle-mounted radar itself and the target tracking track Track_Right at the previous moment; S12': simultaneously obtain the target tracking track Track_Left transmitted by the left vehicle-mounted radar through the left and right communication, and store the track Track_Left and the track Track_Right in the current track tracking list Tracks. The right vehicle-mounted radar data processing module marks whether the track originates from Track_Left or Track_Right; S13': Dots_Right is associated with the track list Tracks; S14': After the association is completed, the target track enters the tracking state.

7. The method for efficiently judging false targets of vehicle-mounted radar according to claim 6 is characterized in that: In step S13', the point track that is not associated with the track will be used as a new target, and a new track will be started, and the track tracking list Tracks will be updated; the point track associated with the track will be matched to the corresponding track, and the track tracking list Tracks will be updated.

8. The method for efficiently judging false targets of vehicle-mounted radar according to claim 6 is characterized in that: In step S13', for the track that is successfully associated with the point track, the track mark from Track_Left is updated to the track mark from Track_Right, that is, the track list of the right vehicle-mounted radar target tracking is integrated to enter the tracking at the next moment; The status of the fused track in the left vehicle-mounted radar target tracking list Track_Left is updated to invisible, and it is no longer displayed and no longer communicated to the right vehicle-mounted radar.

9. The method for efficiently judging false targets of vehicle-mounted radar according to claim 1, characterized in that: In step S3, the target's driving speed Speed_optimized is directly estimated using multiple frame traces, including the following steps: S31: Obtain the associated detection points dots within a period of time deltT of the current target tracking track, and collect the detection speed vr and detection angle δ of these points; S32: Assume that the motion state of the target remains unchanged within deltT time, that is, the associated detection points dots within deltT time are all detected in the same motion state vx, vy, ax, ay; S33: Based on the measured vr and δ, as well as vx, vy, ax, and ay in the same motion state, the actual speed of the target within deltT is estimated using the least squares method. Speed_optimized.

10. The method for efficiently judging false targets of vehicle-mounted radar according to claim 1, characterized in that: In step S4, the filter values ​​vx_filter and vy_filter are compared with the estimated values Speed_optimized, if satisfied That is, the current target is identified as a false target.

Images