Research on denoising optimization algorithm of point cloud in 3d reconstruction

Abstract

3 d reconstruction have become increasingly important in today’s information age, has become a bridge between real world and computer, usually in the access to the image due to factors such as equipment or environment are people hope there are a large number of high frequency signal, the pretreatment of the 3 d reconstruction is the most basic and crucial step in the technology. In this paper, an optimized adaptive filtering de-noising algorithm is proposed. The concept of minimum intra-class dispersion is introduced into the algorithm, and the genetic algorithm is used to find the optimal threshold, so as to eliminate the disadvantages of the algorithm in searching the scattered point cloud data and promote the overall performance of the algorithm. In this paper, the algorithm improves the speed, simplifies the initial point cloud model, and eliminates a large number of invalid points from the subsequent filtering algorithm.