Multi-Robot Point Cloud Map Fusion Algorithm Based on Visual SLAM

Abstract

To solve the problem of inaccurate judgment of overlapping areas in multi-robot system point cloud map fusion, an overlapping areas judgment method based on visual SLAM key frames relative motion size is mooted. On the basis of SLAM mapping, the relative motion is determined comprehensively through features matching and geometric constraint between key frames. Then overlapping areas of maps are determined by relative motion size. At last initial transformation matrix between maps can be calculated. We run our algorithm in both open datasets and real world environment. The results show that the accuracy of this algorithm is higher than that of traditional algorithms.