Fusion of infrared and range data: Multi-modal face images

Abstract

Infrared and range imagery are intriguing sensing modalities for face recognition systems. They may offer better performance than other modalities due to their robustness to environmental effects and deliberate attempts to obscure identity. Furthermore, a combination of these modalities may offer additional discrimination power. Toward this end, we present a semi-automatic system that captures range and infrared data of a human subject's face, registers and integrates multiple 3D views into one model, and applies the infrared measurements as a registered texture map. © Springer-Verlag Berlin Heidelberg 2005.