摘要：This paper presents a multi-robot mapping and localization system. Learning maps and efficient exploration of unknown environment is a fundamental problem in mobile robotics usually called SLAM (simultaneous localization and mapping problem). Our approach uses a team of mobile robots which uses scan-matching and fast-slam techniques for mapping. The single maps generates for each robot are more accurate than map generated by odometry data. When a robot detects another, send its processed map and the master robot can generates a global map very accurate. This way, time necessary to built the global map is reduced.
