What is wrong with mesh PCA in coordinate direction normalization

Authors:
Heng Liu;Jingqi Yan;David Zhang
Affiliations:
Institute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, Shanghai 200030, PR China and Southwest University of Science and Technology, Mianyang 621000, PR China;Institute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, Shanghai 200030, PR China;Department of Computing, The Hong Kong Polytechnic University, Hong Kong, PR China
Venue:
Pattern Recognition
Year:
2006

Citing 1
Cited 1

Shape-Based 3D Model Retrieval

ICTAI '03 Proceedings of the 15th IEEE International Conference on Tools with Artificial Intelligence

Robust shape normalization of 3D articulated volumetric models

Computer-Aided Design

Quantified Score

Hi-index	0.01

Visualization

Abstract

This work makes a detailed analysis on why using mesh PCA for coordinate direction normalization is always uncertainty in 3D surface registration. Our analysis takes the view of discrete signal statistical analyzing and is based on the specific process research of mesh PCA. Then we present a corrected method to improve mesh PCA effects. Such corrected method comes from the fact that the principal axes directions of 3D surface should be those in which the vertex distances are the longest among all 3D vertex distances. Corresponding experimental results on range scan data and synthetic models are provided.