A blind spot detection and warning system based on millimeter wave radar for driver assistance

This paper proposed a blind spot detection and warning system named BSDW for driver assistance under daytime and nighttime conditions. The proposed BSDW system includes system architecture platform, radar hardware system structure, intermediate frequency signal processing, radar target detection algorithms, blind spot area calibration method, system control strategy and system integration. Line frequency modulated continuous wave millimeter-wave radar was used to detect moving targets which come into the rear blind spot alarm area, including left, right and behind area of the subject vehicle. Based on Rayleigh clutter distribution model, a cell greatest, smallest and averaging constant false-alarm rate target detection algorithm named CGSA-CFAR was proposed to maintain higher detection rate and lower false-alarm rate by adjusting power detection threshold in time based on the noise power level, which was estimated according to the proposed target detection algorithm. The proposed system was implemented on embedded hardware platform and verified on the Chery Arrizo7 vehicle. Under urban daytime and nighttime conditions, the early warning rates of the proposed system were up to 98.20% and 98.21% respectively. The results show that the proposed system can detect the targets which come into the rear warning area of the subject vehicle and give early alarm to driver effectively in various urban road environments.