A novel face recognition approach based on strings of minimum values and several distance metrics

Abstract

This paper proposes a novel approach to face recognition using string of minimum values (SMV) as a new face feature extractor for face representation. Unlike most of the face representative methods, which focus only on micro-structures information in image analysis, by surrounding the treated pixel with a mask. The proposed descriptor uses the chains of unit vectors in four directions, by moving from the current pixel to the next one, from which to a new next pixel, and so on, in order to encode also the global appearance of the face image. Furthermore, seven distance metrics from the nearest neighbour classifier are evaluated in the classification stage. The experimental results show which metrics perform well and demonstrate the efficiency of the proposed approach in terms of recognition rate compared to the existing face recognition methods.