3D mapping and visual navigation system for automatic driving scene

In view of the current problem that the navigation effect of the target point of the indoor small environment is not ideal in the V_SLAM of the current mobile robot through the depth camera, based on the RTAB-MAP algorithm, it is proposed to configure the multi-sensor configuration map navigation system based on the use of visual map navigation and the lidar odometer and robot visual mileage rgbd_odometry. Experimental results show that compared with single-V_SLAM vision or lidar mapping, RTAB-MAP after extended sensor configuration can obtain a three-dimensional depth map with richer information and more accurate map edges, which improves the accuracy and practicality of visual navigation to a certain extent.