3D Model Retrieval Based on Multi-Fourier Spectra Representation

Atsushi TATSUMA  Yohei SEKI  Masaki AONO  Ryutarou OHBUCHI 

Publication
D - Abstracts of IEICE TRANSACTIONS on Information and Systems (Japanese Edition)  Vol.J91-D  No.1  pp.23-36
Publication Date: 2008/01/01
Online ISSN: 1881-0225
Print ISSN: 1880-4535
Type of Manuscript: PAPER
Category: 
Keyword: 
3D shape retrieval,  3D model,  Fourier spectrum,  feature vector, 

Full Text(in Japanese): PDF(1.2MB)

Summary: 
We propose a novel method for retrieving 3D shape models. Major characteristics of our method include "Periphery Emphasized Image (PEI)" transformed intensities as the features for representing 3D shape models, and multiple Fourier spectra representations as a combined feature vector for each 3D model, which we call Multiple Fourier Spectra Descriptor (MFSD). As experiments, we use three benchmark datasets, and demonstrate that our method outperforms all the previously known methods to our knowledge for 3D shape retrieval in terms of the average recall and precision.