An Improved RANSAC Surface Reconstruction Study. (August 2019)

• Record Type: Journal Article
• Authors: Li, Ma; Jinming, Zhang; Dongchao, Ma; Yingxun, Fu
• Is Part Of: Journal of physics. Volume 1284(2019)
• Publication Date: 2019-08
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1284/1/012020
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: Surface reconstruction technology has always been an important research area in graphic image. However, efficiently processing massive amounts of high-precision point cloud data is still a problem worth studying. Based on the characteristics of point cloud data and the existing research results, this paper proposes a RANSAC algorithm based on voxel segmentation and applies it to 3D surface reconstruction. By analyzing the characteristics of the original point cloud data, the method improves the reconstruction speed of the 3D model while retaining a large amount of effective feature information and ensuring high precision. The experimental results show that compared with the reconstruction algorithm based on the original RANSAC algorithm, the proposed algorithm can effectively reduce the surface reconstruction time of 47.51%, and the model error after processing is only one thousandth compared with the original model.