Research on Vehicle Target Detection Algorithm Based on Binocular Camera and LiDAR Fusion
Chenglin Dong, Siqi Chen, Ji‐Xiang Wang · 2023
Environment awareness technology is one of the key technologies of driverless cars, and the use of vision sensors or lidar alone has limitations in object detection. Based on the depth map generated by binocular camera, a vehicle target detection algorithm based on the fusion of binocular camera and liDAR is proposed in this paper. Firstly, the binocular image is generated by the adaptive stereoscopic matching network, and then the parallax map is converted into pseudo-point cloud, and the pseudo-point cloud is fused with the real Lidar point cloud. Then the Pillar rasterized target detection network is used to detect the true and false fusion point cloud, and the final vehicle target detection result is obtained. The proposed algorithm is verified on KITTI data set, and the experimental results show that the proposed method has significantly improved the accuracy of vehicle target detection compared with the advanced algorithms.