Optimized discriminative LBP patterns for infrared face recognition

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. Local binary pattern (LBP), as a classic local feature descriptor, is appreciated for infrared face feature representation. To extract discriminative subset in LBP features, infrared face recognition based on optimized discriminative patterns (ODP) is proposed in this paper. Firstly, LBP operator is applied to infrared face for texture information. Secondly, based on two-class discriminative ability, we adaptively select a personalized subset of features from LBP for each subject. Then, dissimilarity metrics between the personalized features is computed base on chi-square distance. Finally, the final recognition algorithm is built on all two-classifiers using voting mechanism. The experimental results show the optimized discriminative patterns can extract compact and discriminative features for infrared face recognition, which outperform the LBP uniform and discriminative patterns.