Research on Point Cloud Registration Based on Topological Homeomorphism

In 3D point cloud processing technology, point cloud registration is one of the most important technologies. Current methods often struggle with these issues, leading to suboptimal registration accuracy. This paper introduces micro-distribution to judge the distribution of point clouds and improve registration speed; then, through topological homeomorphism, it is considered from the structure. Specifically, for the input point cloud data, firstly, based on the micro-distribution, a representative point cloud is selected. Then, the point cloud is topologically homeomorphic based on the theory of topological homeomorphism, combining the characteristics of the point cloud itself and the topological structure, and making a homeomorphic judgment from the structure. Experiments have proved the effectiveness of this algorithm.