Multi-robot SLAM and map merging

Abstract

Although it is a very useful technique for a robot to detect an unknown environment and plot the map, mapping is a classic problem in robotics. The mapping task can be sped up by multi-robot cooperation. Each robot in a team detects a part of the environment and finally merges each individual robot’s map into a single whole one. Because different robots may use different coordinate systems, the maps built by robots cannot merge together easily. The key problem of map merging is to coordinate unification. This thesis presents a way to solve this problem by comparing robot visions. The robots are at the same place when they have the same visions. The robots’ coordinates can be unified when their visions are matched and then their individual maps can be merged together.