Integration of 2D and 3D images for enhanced face authentication

Abstract

This work presents a complete face authentication system integrating 2D intensity and 3D range data, based on a low-cost, real-time structured light sensor. Novel algorithms are proposed that exploit depth data to achieve robust face detection, localization and authentication under conditions of background clutter, occlusion, face pose alteration and harsh illumination. The well known embedded hidden Markov model technique for face authentication is applied to depth maps. A method for the enrichment of face databases with synthetically generated views depicting various head poses and illumination conditions is proposed. The performance of the proposed system is tested on an extensive face database of 3,000 images. Experimental results demonstrate significant gains resulting from the combined use of depth and intensity.