Abstract—In this paper, illumination uniformity on face images has been accessed. A database of 1056 visible light face images and 1056 near-infrared face images has been used in our assessment process. This database uses 22 volunteers, irrespective to race and gender. Each volunteer had to attend two photography sessions. Our investigation on intensity distributions, and intensity differences, supports the theory that near-infrared face images are more robust towards illumination variations.

Index Terms—Digital image processing, face images, face detection, illumination inhomogeneity, illumination uniformity, near-infrared face images, visible light face images.

The authors are with the School of Electical & Electronic Engineering, Engineering Campus, Universiti Sains Malaysia, 14300 Nibong Tebal, Penang, Malaysia (e-mail: haidi_ibrahim@ ieee.org, nkc99131@student.usm.my).

Cite: Haidi Ibrahim and Kuo Chue Neo, "Illumination Uniformity Evaluation on Visible Light andNear-Infrared Face Images," International Journal of Computer and Electrical Engineering vol. 4, no. 6, pp. 843-845, 2012.