Enhanced PCA reconstruction method for eyeglass frame auto-removal

Author

Guo Pei ; Su Fei

Author_Institution

Beijing Univ. of Posts & Telecommun., Beijing, China

fYear

2014

fDate

19-21 Sept. 2014

Firstpage

359

Lastpage

363

Abstract

A robust face recognition system needs to address the problem of partial occlusion like frame glasses. In this paper, a new glass auto-removal scheme is proposed, which includes the SVM-based glass detection, recursive PCA reconstruction, GVF snake model and the multi-image patch match. Experimental results show good performance in aspects of quantitative measure and face recognition. It implies that this method can be used to improve the face recognition accuracy in real applications.

Keywords

face recognition; image matching; image reconstruction; principal component analysis; support vector machines; GVF snake model; SVM-based glass detection; enhanced PCA reconstruction method; eyeglass frame auto-removal scheme; multiimage patch match; partial occlusion problem; quantitative measure; recursive PCA reconstruction; robust face recognition system; Face; Face recognition; Feature extraction; Glass; Image reconstruction; Principal component analysis; Reconstruction algorithms; eyeglass removal; gradient vector flow snake; multi-image patch match; recursive PCA reconstruction;

fLanguage

English

Publisher

ieee

Conference_Titel

Network Infrastructure and Digital Content (IC-NIDC), 2014 4th IEEE International Conference on

Conference_Location

Beijing

Print_ISBN

978-1-4799-4736-2

Type

conf

DOI

10.1109/ICNIDC.2014.7000325

Filename

7000325