The requirement of 3D scanning of workpiece topography is increasing in workshop automation equipment,where point cloud registration is an important step in 3D data processing.The existing 3D point cloud registration has the problems of feature point mismatch,long registration time and poor registration accuracy and etc.In this paper,an NDT 3D point cloud registration algorithm based on intrinsic shape signatures 3D shape Context feature(ISS 3DSC)is proposed.Firstly,the intrinsic shape signatures(ISS)algorithm is used to extract 3D point cloud key points to improve the registration efficiency.Then,the 3D shape context feature(3DSC)is used to describe the features of key points,and the wrong point pairs are removed according to the median distance method of feature point pairs,and the initial transformation matrix is calculated by SVD decomposition.Finally,the NDT algorithm is used to complete the fine registration.The experimental results show that the accuracy of the algorithm in vamp and sole point cloud data matching can reach 0.025cm,which significantly improves the registration efficiency compared with the traditional SAC IA+NDT algorithm and has certain engineering application value.