Abstract We propose a point cloud registration method based on the fusion of sample consensus initial alignment(SAC-IA) and normal distribution transformation(NDT) registration. First, we calculate the fast point feature histogram(FPFH) feature of the point cloud to be registered and the target point cloud; second, according to the feature of two pieces of point cloud and using the SAC-IA algorithm to find the initial transformation matrix, thereby completing the initial registration; finally, on the basis of the initial registration, we use the NDT algorithm to precisely register the two point clouds. Experimental results show that the registration accuracy of this method is significantly superior to the ICP algorithm, and the registration efficiency is also relatively improved.