Research on robotic mapping based on fusion of LiDAR and depth camera

In order to solve the problems of indoor mobile robot mapping, such as not detecting small obstacles or unsatisfactory mapping results, this paper proposes a depth camera and LiDAR fusion method for indoor mobile robot mapping. First, the LIDAR odometer was estimated using LIDAR distance scanning. Then the LIDAR odometer, wheeled odometer, and IMU are fused using Kalman filtering to improve the odometer accuracy. Next, the depth image is converted to pseudo-laser information, and the pseudo-laser information and laser information are simply spliced, weighted average and other operations are performed to fuse them into laser information with more comprehensive information. Finally, the mapping algorithm is used to receive the fused laser information and the fused odometer to complete the 2D construction of the surrounding environment. The Gazebo simulation environment and the robot model are built through the ROS2 system for verification. The research results show that the optimized fusion algorithm makes the indoor mobile robot clearer in the mapping of hollow-out objects, more accurate in obstacle, and more accurate in robot. This provides a certain reference for mobile robot map.