Infrared face recognition based on LBP co-occurrence matrix
Zhihua Xie · 2014
Infrared face recognition, being light- independent, and not vulnerable to facial skin, expressions and posture, can avoid or limit the drawbacks of face recognition in visible light. To extract spatial relations among the local binary pattern (LBP), a new infrared face recognition method based on LBP co-occurrence matrix is proposed in this paper. In traditional LBP-based features such as LBP histogram, space locations information, which is important feature for recognition, is discarded. To consider such spatial relations in infrared faces, co-occurrence matrix of LBP codes, instead of histogram, is used to represent the infrared face. LBP co-occurrence matrix not only preserves great invariance to environmental temperature, but also greatly enhances the discriminative power of the descriptor as co-occurrence matrix of LBP code well captures the correlation between locally adjacent regions. The experimental results show the infrared face recognition method based on LBP co-occurrence matrix outperforms the traditional methods based on LBP histogram or PCA+LDA.