Multispectral Face Recognition in Texture Space

Abstract

This work introduces the use of LBP like texture descriptors for efficient multispectral face recognition. LBP has been widely used in visible spectrum face recognition. This work extend its use to non visible spectrums (active and passive infrared spectrums). Local Binary Pattern (LBP) and Local Ternary Pattern (LTP) descriptors are used. Also a simple differential LTP descriptor (DLT) is introduced. The proposed texture space is less sensitive to noise, illumination change and facial expressions. These characteristics make it a good candidate for efficient multispectral face recognition. Linear and non linear dimensionality reduction techniques are introduced and used for performance evaluation of multispectral face recognition in the texture space. The obtained results show that the use of the proposed texture descriptors permit to achieve high recognition rates in multispectral face recognition.