A Fast Reduction and Denoising Method of Scattered Point Cloud Data Based on Spherical Model

Abstract

Aiming at the problems of high complexity, long computing time and high hardware requirements of common point cloud reduction and denoising algorithms, a fast point cloud data reduction and denoising method based on ball model is designed. The method of building the sphere model and including the scattered point cloud model of 3D object in the sphere model is introduced. The coding and sorting of the vertebral body region in the sphere and the distribution strategy of each coordinate point is discussed. The different distribution of point clouds in each cone region and the algorithm of data reduction and denoising is studied. A method to transform the scattered point cloud model of 3D object from spherical coordinate to rectangular coordinate is presented. Experimental simulation and analysis show that the method proposed in this paper can simplify and denoise the scattered point cloud at the same time, and the algorithm is simple, with low requirements for the hardware system. This algorithm is more suitable for embedded system operation.