Vehicle-mounted radar target tracking method and device based on target prior information

Abstract

The invention discloses a vehicle-mounted radar target tracking method and device based on target prior information, and the method comprises the steps: S1, obtaining an echo signal of a vehicle-mounted radar in a target vehicle for target detection and tracking, and obtaining an initial target tracking result; S2, screening out a target group needing to be fused from the initial target detectiontracking result according to prior information of the tracked target, wherein the prior information comprises a speed characteristic and a position characteristic; S3, fusing the target groups screened in the step S2 to obtain a final target tracking result and outputting the final target tracking result. The device comprises a target detection tracking module, a fusion target screening module anda target group fusion module. According to the invention, the target priori information can be fully utilized to realize target tracking, the tracking effect of regularly arranged targets such as trees is improved, and the method has the advantages of simple implementation method, low cost, good tracking precision and tracking effect and the like.

Description

technical field [0001] The invention relates to the technical field of vehicle radar target tracking, in particular to a vehicle radar target tracking method and device based on target prior information. Background technique [0002] Vehicle-mounted radars such as automotive forward obstacle avoidance radars can be used to assist vehicles in judging obstacles ahead to avoid collisions and ensure driving safety. Therefore, this type of radar has become a research hotspot. Such as automotive forward obstacle avoidance radar and other common application scenarios in practice such as figure 1 As shown in the figure, when a vehicle is driving on the road, there are usually some regularly arranged targets on both sides of the road, such as neatly arranged trees, utility poles, etc., and there are other driving vehicles in front of the car. In this type of application scenario, a prerequisite for the car to avoid obstacles is stable and accurate tracking of the surrounding environ...

AI Technical Summary

Problems solved by technology

[0003] In the prior art, when the radar performs target tracking, it usually filters and tracks the target directly after detecting the position and motion information of the target. However, in the actual environment, there will be various noise interference, and it is necessary to accurately obtain The real-time noise of the target is almost impossible to achieve, that is, the filtering result cannot completely filter out the noise interference, resulting in the noise interference still exists in the output result, so the actual measurement value obtained by the radar usually has various errors

in such as figure 1 In the vehicle-mounted radar application scenario shown, when the above method is used to track regularly arranged targets such as trees and utility poles, due to various errors in the actual measurement values ​​obtained by the radar, the actual tracking effect is usually not as expected. It is expected that they are arranged in a neat and regular manner, and the desired tracking effect cannot be achieved

Taking the tracking effect of the car's forward obstacle avoidance radar on the trees beside the road as an example, the actual tracking effect obtained by using the above-mentioned traditional target tracking method is usually as follows figure 2 As shown in the figure, the boxes in the figure represent other vehicles in front of the own vehicle. It can be seen from the figure that the tracked row of trees is skewed and not straight and neatly arranged, that is, the actual tracking effect and the expected tracking effect There is a large error

Images