Hits:

Indexed by:期刊论文

Date of Publication:2007-03-01

Journal:Journal of Information and Computational Science

Included Journals:EI

Volume:4

Issue:1

Page Number:257-264

ISSN No.:15487741

Abstract:Singular values (SVs) have been used for face recognition by many researchers in the last two decades, However, SVs neglect the character of human faces and contain inadequate information for recognition. To address these problems, more useful information must be utilized. In this paper, we show that facial features represent the individual diversity adequately and these local regions are affected slightly by environment. Accordingly, a new representation of face image based on the above findings is proposed. The Combined SVs (CSVs) utilize the high discriminatory local regions to represent face, and their properties are similar to SVs'. The experimental results on the ORL database and our database indicate CSVs are effective and proper for face recognition. In order to detect the needed local regions conveniently, a new robust detection method based on grayscale mathematical morphology and gray-level projection is also proposed. The experimental results illustrate the detection algorithm is robust the variations of facial details, pose and expression, and appropriate for extraction of CSVs.