Combination of Multiple Robots SLAM

This system is a multiple robots SLAM (Simultaneous Localization And Mapping) system. This system can be used in a variety of scenarios, such as collaborative military tasks, in agricultural planting, multi-robots work together to improve efficiency, etc. Our system consists of five modules: tracking, mapping, communication, optimization and place recognition. In the tracking module, each agent can extract GCNv2 feature points and combine the IMU (Inertial Measurement Unit) data to determine its rotation matrix and translation vector, and then the agent will run the local-mapping module to get local map points. After the server receives the agents’ messages through the communication module, it will run optimization and place recognition modules to get more accurate data. Our contribution is as follows. On the one hand, our algorithm to extract feature points is more robust than the ORB one. On the other hand, with the fuse of IMU, our system is more accurate than the systems where only feature points are used. Experiments are conducted to verify the proposed algorithms.