Adaptive Reduction Algorithm of Scattered Point Clouds Based on Wavelet Technology

Abstract:

An adaptive reduction algorithm of scattered point clouds based on wavelet is proposed, in which the 3D point clouds are converted into point sets on the 2D plane firstly by using the slicing technology in rapid prototyping theory, and then the wavelet coefficients of sorted point clouds data after the wavelet transform can be obtained whose peaks represent the points to be reserved. According to the experiments, the rapid and high quality reduction of scattered point can be performed while the slice thickness is chosen as 2 or 3 times of the sampling interval. The result indicates that this algorithm has obvious advantages in terms of the feature preserving. It can preserve the feature information ultimately, thus the reduction results are more ideal. Due to peaks of wavelet coefficient can adaptively identify the objects’ details and features, this algorithm needlessly set a threshold, which explains the adaptability of the algorithm and also contributes to realizing the automatic reduction.