Target classification method of millimeter wave radar

Abstract

The invention relates to the technical field of radars, in particular to a target classification method for a millimeter-wave radar, and the method comprises the following specific steps: S1, obtaining a plurality of feature parameters outputted by the millimeter-wave radar; s2, analyzing the plurality of characteristic parameters to obtain a plurality of candidate targets, the plurality of candidate targets including pedestrians, cars and non-motor vehicle targets; s3, establishing an effective feature vector based on the plurality of candidate targets, and extracting representative target features; s4, a classifier based on multi-feature fusion is constructed, a classification recognition model based on various road targets is constructed and designed according to the KNN algorithm principle, and the adopted KNN algorithm is simple, visual and easy to implement; targets are effectively classified by using a target classification and recognition technology, target types and behavior intentions are judged according to category information, and then target threat degrees are described and judged, so that information is fed back to vehicle owners in time, and the accident occurrence probability is reduced.

Description

technical field [0001] The invention relates to the technical field of radar, in particular to a target classification method of millimeter wave radar. Background technique [0002] Safety issues have always been accompanied by the development of the automobile industry. Traffic accidents that occur around the world every year have caused huge losses of human life and property. It is understood that the number of road traffic deaths is still on the rise. Therefore, how to improve the safety of driving has become a A problem faced by both governments and researchers. [0003] In recent years, the advanced driver assistance system (ADAS) based on active safety control has become the mainstream development trend to ensure safe driving. ADAS uses on-board sensors to perceive changes in the surrounding environment at any time and collect data, thereby realizing functions such as target detection, tracking, identification and classification, so as to effectively eliminate potenti...

AI Technical Summary

Problems solved by technology

However, in order to realize the target classification of vehicle-mounted millimeter-wave radar, there are still problems such as the selection of target features, the establishment of classification mechanism, and the research of fusion classification algorithm.

Images