Risk prediction method and device for millimeter wave radar plots

Abstract

The invention discloses a risk prediction method and device for millimeter wave radar plots. The method comprises the steps of obtaining the current feature information and the historical feature information corresponding to a plurality of plots; based on the current feature information, obtaining the combination feature information corresponding to each of the plurality of feature combinations; according to the combination feature information corresponding to the feature combinations and the historical feature information, determining the probability information corresponding to the feature combinations; determining the likelihood probability and weight of a target plot under the plurality of feature combinations based on the probability information; and determining the risk information of the target plot according to the likelihood probability and the weight. According to the technical scheme of the invention, the risk information of the target plot is determined through the likelihood probability and the weight of the target plot under the plurality of feature combinations, and the risk prediction is realized, so that the risk information can be used as the reference information for judging the authenticity of each plot before the target track processing, and the target automatic starting and tracking quality of a data processing system is improved.

Description

technical field [0001] The invention relates to the technical field of radar systems, in particular to a risk prediction method and device for millimeter-wave radar traces. Background technique [0002] The working environment of millimeter-wave radar is complex, and there are a large number of clutter / false target echoes in the radar echo, resulting in a sharp increase in the number of false point clouds, which brings great challenges to the automatic initiation and tracking of targets in the millimeter-wave radar data processing system. [0003] The millimeter-wave radar point cloud data will be clustered according to certain rules, but in the existing processing flow, the point traces formed after clustering are not processed, and false point traces will also participate in the data processing system. During the process of starting and tracking, it will affect the target track formation. Therefore, it is necessary to carry out risk prediction for the falsity of the dot t...

AI Technical Summary

Problems solved by technology

[0002] The working environment of millimeter-wave radar is complex, and there are a large number of clutter / false target echoes in the radar echo, resulting in a sharp increase in the number of false point clouds, which brings great challenges to the automatic start and tracking of targets in the millimeter-wave radar data processing system

[0003] The millimeter-wave radar point cloud data will be clustered according to certain rules, but in the existing processing flow, the point traces formed after clustering are not processed, and false point traces will also participate in the data processing system. In the process of starting and tracking, it will affect the target track formation

Images