A Novel Approach to Face Recognition and Expression Analysis using Local Directional Number Pattern

Image analysis and understanding has recently received significant attention, especially during the past several years. At least two reasons can be accounted for this trend: the first is the wide range of commercial and law enforcement applications, and the second is the availability of feasible technologies after nearly 30 years of research. In this paper we propose a novel local feature descriptor, called Local Directional Number Pattern (LDN), for face analysis, i.e., face and expression recognition. LDN characterizes both the texture and contrast information of facial components in a compact way, producing a more discriminative code than other available methods. An LDN code is obtained by computing the edge response values in 8 directions at each pixel with the aid of a compass mask. This directional information is then encoded them into a 6 bit binary number using the relative strength of these edge responses. To analyze a face, it is divided into several regions, and the distribution of the LDN features are extracted from them. Then, these features are concatenated into a feature vector, and it is used as a face descriptor. We perform several experiments in which our descriptor showed consistent results under age, illumination, expression, and noise variations.