Multispectral visible and infrared imaging for face recognition

Multispectral imaging in the visible and near infrared spectra helps reduce color variations in the face due to changes in illumination source types and directions. Thermal infrared imaging provides useful signatures of the face that is insensitive to ambient lighting through the measurement of heat energy radiated from the object. This paper introduces the use of multispectral imaging and thermal infrared imaging as alternative means to conventional broadband monochrome or color imaging sensors in order to enhance the performance of face recognition in uncontrolled illumination conditions.

1. Introduction

Many automatic face recognition systems have demonstrated excellent recognition performances, often in controlled lighting conditions, and typically with frontal, centered and expressionless views of faces [1]. However, face recognition in uncontrolled illumination experiences significant degradation in performance due to changes in illumination directions and skin colors [2][3]. The color information acquired by conventional color CCD cameras with broadband sensor characteristics varies depending on the types of lighting sources. A main obstacle for universal color use in machine vision applications is that the cameras are not able to distinguish changes of surface color from color shifts caused by varying illumination. Multiple wavelengths of visible and infrared (IR) regions of electromagnetic spectrum provide useful information for face recognition [4]. Such information can facilitate recognition in the situations where an ordinary imaging system might not be able to separate the illumination effects from the changes in the object.