Multi-resolution analysis of grid point clouds based on wavelet transform

An application of classical wavelet in the realization of the multi-resolution analysis of point cloud is presented in this paper. Firstly, a proper wavelet function and the expending method are determined by analyzing the character of grid point cloud. With the theory of discrete wavelet transform (DWT), the multi-level wavelet decomposition is realized. Based on the band-pass filter property of the wavelet function, the high frequency information in the point cloud is removed, and the low frequency part is reserved. And then the multi-level of detail and multi-resolution models are generated, which are the simplifications of the original point cloud. Finally, the feasibility and the efficiency are proved by a study case.