Fast 3D recognition for forensics and counter-terrorism applications

Abstract

The development of advanced techniques for fast 3D reconstruction and recognition of human faces in unconstrained scenarios can significantly help the fight against crime and terrorism. We describe a 3D solution developed within Sheffield Hallam University that satisfies a number of important requirements such as operating close to real-time, high accuracy in recognition rates and robust to local illumination. Experimental results in 3D face recognition are reported and two scenarios are provided that can be used to exploit the outcomes of this research for forensic analysis and for flagging potential threats in counter-terrorism.