Research of the reconstruction method for the image feature of non-rigid 3D point cloud

Abstract

In the reconstruction of non-rigid three dimensional (3D) point cloud image features, the image data is large, which leads to the difficulty for analysis, and the achieving process is complex. In this paper, the reconstruction method for non-rigid 3D point cloud image features based on NURBS curve is proposed to collect non-rigid 3D point cloud image data, and achieve triangulation. The operation of point cloud data neighborhood search is limited in the local area, and the data of non-rigid 3D point cloud images are divided with the idea of space division. The least square plane is obtained by using preprocessed non-rigid 3D point cloud data, and the non-rigid 3D point cloud data is mapped and converted into a two dimensional (2D) point cloud model. The cumulative chord length parameterization method is adopted to introduce weighting factor for NURBS curve description using the rational polynomial function, based on the definition and properties of NURBS curve, point, line, surface reconstruction model is utilized to reconstruct the non-rigid 3D point cloud image feature. The simulation results show that the proposed method can reconstruct the 3D point cloud images feature with less time and the reconstruction effect is better than the Crust method.