A High-Level Radar and Vision Sensor Fusion for Autonomous Vehicle

摘要

Considering that a single sensor has low fault tolerance and low accuracy in recognizing and tracking the target, this paper has proposed a high-level, radar and camera information fusion algorithm based on ROS (Robot Operating System). Millimeter-wave radar and camera targets need to be filtered and time-synchronized before being fused. The Global Nearest Neighbor algorithm is used in the target matching process. We set the appropriate rules to fuse the targets of the radar and camera sequences. It is verified by experiments that more accurate information on the position of the preceding vehicle can be obtained through this fusion system.