Advances in 3D-based face recognition

Biometrics Methods use unique patterns of a bodily feature for identification. Unlike passwords or pin numbers, biometrics cannot be lost or forgotten. Biometrics include fingerprints, iris, height, voice, face. Automatic face recognition has attracted substantial research attention in the last fifteen years. Face recongnition is well accepted by the public and is non-intrusive in comparison to fingerprinting. The bulk of the research has concentrated on 2D-based face recongnition systems. Problems associated with 2D techniques are well known and include make-up and hair, expression, pose, illumination, time lapse between gallery and probes, gallery size, computing time. Research in 3D is becoming more popular as some of the difficulties brought by changes in illumination and pose may be eliminated. This article is a review of 3D-based face recognition. © 2006 IEEE.