Evaluation and optimization of intersection process between voxelized mesh models and homocentric spheres

Abstract

In the field of 3D model matching and retrieval, an effective method for feature extraction is spherical harmonic or its mutations, and is acted on the spherical images. But the obtainment of spherical images from 3D models is very time-consuming, which greatly restrains the responding speed of such systems. In this paper, we propose a quantitative evaluation of the whole process and give a detailed two-sided analysis based on the comparative size between pixels and voxels. The experiments show that the resultant optimized parameters are fit for the practical application and exhibit a satisfactory performance.