Presentation 2014-03-07
Block Matching Algorithm Weighting Eye Area for Face Recognition Across Age Progression
Taro OKAKURA, Daiki TAKEDA, Kaoru ARAKAWA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Personal authentication from face images is now utilized in various applications, however, the correct recognition rate gets lowered, if age progression exists between the captured face image and the registered one. Authors proposed before a method for personal authentication with such age progression using the block matching algorithm. This method is based on the principle that the face components shift according to age progression, while the features of the face components are kept unchanged. The experimental results showed advantage to the conventional methods without block matching, such as a method with KL transform. In this paper, we newly propose a method to improve the performance of the previous block-matching method by weighting the eye area. The eye area usually expresses each person's characteristics clearly. Since the eye area of a person is similar to that of the same person at least in either right area or left area, the degree of similarity in right-eye area or left-eye area, the value of which is more dominant, is additionally involved in the evaluation function. Computer simulations show the advantage of the new method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) face image / block matching / personal authentication / age-related change
Paper # SIS2013-67
Date of Issue

Conference Information
Committee SIS
Conference Date 2014/2/27(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Smart Info-Media Systems (SIS)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Block Matching Algorithm Weighting Eye Area for Face Recognition Across Age Progression
Sub Title (in English)
Keyword(1) face image
Keyword(2) block matching
Keyword(3) personal authentication
Keyword(4) age-related change
1st Author's Name Taro OKAKURA
1st Author's Affiliation Graduate School of Science and Technology, Meiji University()
2nd Author's Name Daiki TAKEDA
2nd Author's Affiliation Graduate School of Science and Technology, Meiji University
3rd Author's Name Kaoru ARAKAWA
3rd Author's Affiliation School of Interdisciplinary Mathematical Sciences, Meiji University
Date 2014-03-07
Paper # SIS2013-67
Volume (vol) vol.113
Number (no) 467
Page pp.pp.-
#Pages 5
Date of Issue