Highway Traffic Detection Method Based on Ray-Vision Fusion
Aiming at the insufficiency of single-sensor detection and the complex road environment,this paper designs a fusion algorithm for vehicle target detection based on visual sensors and millimeter-wave radar.The spatial fusion model is constructed through the parameter calibration and coordinate transformation of the sensors,and the internal reference of the sensors and the position information of the actual installation are obtained firstly,and then the millimeter-wave radar's coordinates are transformed to the pixel coordinates of the image to realize the temporal fusion of the sensors.The experimental results show that the fusion algorithm of visual sensors and millimeter-wave radar can significantly improve the problems of video clutter,multipoint distribution of radar,and segmentation of detected targets compared with the detection results of a single sensor,and the accuracy of target detection is higher than that of a single sensor,especially in complex scenes,the fusion algorithm has a higher stability of detection.
NETL
NSTL
万方数据
