You are viewing a javascript disabled version of the site. Please enable Javascript for this site to function properly.
Go to headerGo to navigationGo to searchGo to contentsGo to footer
In content section. Select this link to jump to navigation

Blood perfusion construction for infrared face recognition based on bio-heat transfer

Issue title: Frontiers in Biomedical Engineering and Biotechnology – Proceedings of the 3rd International Conference on Biomedical Engineering and Biotechnology, 25–28 September 2014, Beijing, China

Article type: Research Article

Authors: Xie, Zhihua; | Liu, Guodong

Affiliations: Key Lab of Optic-Electronic and Communication, Jiangxi Sciences and Technology Normal University, Nanchang 330013, China

Note: [] Corresponding author: Zhihua Xie, Key Lab of Optic-Electronic and Communication, Jiangxi Sciences and Technology Normal University, Nanchang 330013, China. Tel.: 13970018047; Fax: 086-791-83831977; E-mail: [email protected].

Keywords: Feature extraction, infrared image, face recognition, blood perfusion, bio-heat transfer model

DOI: 10.3233/BME-141091

Journal: Bio-Medical Materials and Engineering, vol. 24, no. 6, pp. 2733-2742, 2014

Published: 2014
Get PDF open access

Abstract

To improve the performance of infrared face recognition for time-lapse data, a new construction of blood perfusion is proposed based on bio-heat transfer. Firstly, by quantifying the blood perfusion based on Pennes equation, the thermal information is converted into blood perfusion rate, which is stable facial biological feature of face image. Then, the separability discriminant criterion in Discrete Cosine Transform (DCT) domain is applied to extract the discriminative features of blood perfusion information. Experimental results demonstrate that the features of blood perfusion are more concentrative and discriminative for recognition than those of thermal information. The infrared face recognition based on the proposed blood perfusion is robust and can achieve better recognition performance compared with other state-of-the-art approaches.

Show more