A method of 3D face recognition based on principal component analysis algorithm

Author

Yuan, Xue ; Lu, Jianming ; Yahagi, Takashi

fYear

2005

fDate

23-26 May 2005

Firstpage

3211

Abstract

We present a method of face recognition using 3D images. We first compensate the poses of 3D original facial images using geometrical measurement and extract 2D texture data and the 3D shape data from 3D facial images for recognition. Based on a principal component analysis (PCA) algorithm, all the 2D texture images and the 3D shape images are normalized to 32×32 pixels. In the second step, we propose a method for face recognition based on fuzzy clustering and parallel neural networks. Experimental results for 70 persons with different poses demonstrate the efficiency of our algorithm.

Keywords

face recognition; feature extraction; fuzzy logic; image texture; neural nets; principal component analysis; 1024 pixel; 2D texture data extraction; 2D texture images; 32 pixel; 3D face recognition; 3D images; 3D shape data extraction; 3D shape images; facial images; fuzzy clustering; parallel neural networks; principal component analysis; Clustering algorithms; Data mining; Face recognition; Feature extraction; Fuzzy neural networks; Image recognition; Lighting; Neural networks; Principal component analysis; Shape measurement;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems, 2005. ISCAS 2005. IEEE International Symposium on

Print_ISBN

0-7803-8834-8

Type

conf

DOI

10.1109/ISCAS.2005.1465311

Filename

1465311