A radar -based vehicle tracking detection method and device

Abstract

The invention discloses a radar-based vehicle tracking detection method and device. The method steps include: S1. Selecting a target point group corresponding to the same vehicle target from the returned target detection point cloud every time the radar performs measurement; S2. In the target point group, the edge of the target shape is constructed according to the shape distribution of the vehicle; S3. Extract the measurement points within the specified range around the edge of the constructed target shape as the edge point, and fuse all the extracted edge points to obtain the final target position information; S4. Perform target tracking filtering according to the obtained final target position information. The invention has the advantages of simple implementation method, high target tracking accuracy and high efficiency.

Description

technical field [0001] The invention relates to the technical field of radar target tracking, in particular to a radar-based vehicle tracking detection method and device. Background technique [0002] For example, the 77GHZ radar hardware system has been widely used in various hot applications such as automobile obstacle avoidance and security. Because the 77GHz radar has a wider operating bandwidth, this type of radar has a higher range resolution. It is for this reason that this type of radar has the following problems when tracking targets: due to the improvement of radar resolution, A target often occupies multiple distance units in a single detection, that is, a single detection target will be returned with multiple measurement values. At this time, the detected target cannot be regarded as a point target like a traditional radar. Under this condition, the target Tracking is called extended target tracking. [0003] When the traditional tracking method deals with the ...

AI Technical Summary

Problems solved by technology

Since the 77GHz radar has a wider operating bandwidth, this type of radar has a higher distance resolution. It is for this reason that the following problems will exist when this type of radar performs target tracking: due to the improvement of the radar resolution, A target often occupies multiple distance units in a single detection, that is, a single detected target will be returned with multiple measurement values. At this time, the detected target cannot be regarded as a point target like traditional radar. Under such conditions, the target Extended Object Tracking

However, there are various noises in the real environment, and each returned measurement value will be mixed with noise, that is, there will be a deviation between the measurement value and the real value. This deviation depends on the influence of the noise. If the target is detected, the noise is relatively large, the deviation between the measured value and the real value is relatively large, and when the noise is relatively small, the deviation between the corresponding measured value and the real value is small, and the noise fluctuation has randomness, that is, The deviation between each measurement value and the true value in multiple measurement values ​​cannot be determined. The above-mentioned traditional tracking method directly averages all measurement values ​​without considering the different degrees of each measurement value being affected by noise, so There will be a large error between the final measured value obtained by averaging all the measured values ​​and the true value of the target, and the actual target tracking accuracy is not high

Images