Graph Structure and Correlation Coefficient for Face Detection

Abstract

Human face detection has been the subject of extensive research for several decades. Face detection has become increasingly important with the advent of video surveillance, security access control, and content-based information retrieval applications. Due to the non-rigid nature of faces and their high degree of variability, face detection remains challenging. There have been many studies that examined various face detection methods under different conditions, including illumination, facial expressions, head rotations, occlusions, and aging. A novel approach to face detection is described in this paper using Local Graph Structure and Correlation Coefficient (LGS-CC). Through LGS-CC, texture information is enhanced by considering both texture and local shape, rather than relying solely on grayscale information. The proposed method has been shown to achieve encouraging results in extensive experiments.