Multi-resolution Surface Reconstruction (TP-P1)

Author(s) : Mingyi He (Northwestern Polytechnical University, China) Bangshu Xiong (Northwestern Polytechnical University, China) Huajing Yu (Northwestern Polytechnical University, China)

Abstract : Based on the relationship between local normal change of sample points and the surface Gaussian curvature, a novel multi-resolution surface reconstruction algorithm (MRSR) is proposed. First a min-max box containing all sample points is split into octree cells with different size according to the user-specified normal change threshold. If the local surface is flat or smoothly curved, a portion of the isosurface will be extracted in a larger cell, otherwise it will be extracted in a smaller one. Then the cracks at interface cells with different resolution are patched by perpendicular projection. Thus a multi-resolution triangular mesh is directly generated from the unorganized cloud points. The datasets of Terracotta warrior and horses of the Qin Dynasty and the Kettle were used to verify the efficiency of our algorithm. The experimental results show that the proposed algorithm can balance the tradeoff between surface reconstruction accuracy and speed very well.

Menu