Algorithm of 3D reconstruction based on point cloud segmentation denoising

The triangular mesh has am undesirable noise. By the point cloud segmentation approach we can denoise the data just only for the characteristic point, thereby reducing the number of point cloud processing and precision. This paper proposes an algorithm which can smooth nosing data quickly and use this method to reconstruction 3D model. Experimental result proves the validity of the algorithm.