In the process of point cloud alignment,in order to improve the accuracy of point cloud alignment,the point cloud data is processed by Super 4PCS+ICP "coarse first and then fine" for the limitations of ICP algorithm for initial poses.The corresponding points are first searched using the KDTree,and the feature point set is determined using the feature degree of the local area,and then the coarse alignment is achieved using the Super 4PCS algorithm.KDTree is proposed for fine registration to speed up the process,SVD to solve for the corresponding point parameters,weighted average with constant 1,and solving for the error function to realize the improvement of ICP algorithm and to find out the rotation translation matrix after rigid body transformation to improve cloud registration accuracy.Experiments show that compared with the traditional ICP algorithm,the registration accuracy of the proposed method is significantly improved.The method studied in this paper can provide some reference for the further study on point cloud registration.