Learning Modified CS-LBP for Face Recognition

The face of a human being conveys extensive information about the identity and emotional state of the person. However, the most important phase in face recognition is extraction. In which the most useful and unique features of the face image are extracted. The Center-Symmetric Local binary Pattern (CS-LBP) feature can be viewed a combination of the texture-based feature and the gradient-based feature. However, due to the low spatial support, the bit-wise comparison made between two single pixel values is significantly affected by noise and is sensitive to image translation and rotation. To address this problem, a modified feature called Multi-scale block Center-Symmetric Local Binary Pattern called as (MCS-LBP) is presented. Instead of pixels, in MCS-LBP the comparison is based on the sum of values of each block’s sub-regions.