A Novel Target Classification And Identification Algorithm For 77G LFMCW Automotive Collision Avoidance Radar

Study on Electromagnetic scattering and micro-Doppler signatures of pedestrians are fundamentally important in pedestrian recognition for developing robust pedestrian collision avoidance systems. To develop algorithms for person, obstacles and cars detection, classification, identification and tracking, it is necessary to acquire the knowledge of the radar characteristics of these targets. Firstly, this paper describes the principle of proposed 77G LFMCW radar and analytical formulas were derived to measure the distance, velocity and angle information of targets. Secondly, based on actual data, we focused on investigation of the echo characteristics of pedestrians, obstacles and cars using a radar system operating at 77 GHz. Further research based on their difference of spectral scattering characteristics presented in this paper will contributes to the classification and identification algorithm, ultimately forming to their respective track points.