An Illumination Invariant Face Recognition Method Based on Gradientfaces and 2DLDA
Yongfeng Qi, Yuanlian Huo · Advances in computer science research · 2015
Extracting illumination invariant features is an effective method to improve recognition rate of algorithm with varying illumination.In this paper, a novel illumination invariant method which combines Gradientface and 2DPDA is proposed.First, this approach transforms a face image into Gradientface, then the 2DPCA is used to reduce the dimension of Gradientface.The experimental results verify the effectiveness of our approach on CAS-PEAR-R1 and CMU PIE face databases.