Illumination invariant face recognition using SQI and weighted LBP histogram

The illumination variation is one of the main challenges in real-world face recognition systems. Face recognition under uneven illumination is still an open problem. In this paper, we proposed a novel illumination invariant face recognition approach based on Self Quotient Image and weighted Local Binary Pattern. We improved the performance of the system by using different sigma values of SQI for training and testing. Furthermore, using two multi-region uniforms LBP for feature extraction simultaneously, made the system more robust to illumination variation. This approach gathers information of the image in both local and global levels. The weighted Chi square statistic is used for histogram comparison. The used weighted approach emphasizes on the more important regions in faces. The proposed approach is compared with some methods like QI, SQI, QIR, MQI, DMQI, DSFQI, PCA and LDA on Yale face database B and CMU-PIE database. The experimental results show that our method outperforms other tested methods.