Deep Learning Methods for Robotic Arm Workspace Scene Reconstruction

抄録

This research is part of the Yaskawa Motoman Robot Autonomous Control Project, which aims to map the real workspace in a virtual environment using a depth camera mounted on the robot, and to plan the robot's autonomous obstacle avoidance path based on the 3D octomap. The main tool used in this study is RTAB-Map, which is based on the built-in handheld mapping scheme to improve it to meet our actual needs. After the actual test, our solution shows finer mapping accuracy, can update the map data in real time, and the perception of obstacles within the field of view is more comprehensive, but there is still a lot of room for optimizing the mapping speed.