An Illumination Robust Algorithm for Face Recognition Via SRC and Gradientfaces

Authors:
Zipeng Yu;Zhendong Wu;Jianwu Zhang
Affiliations:
College of Telecommunication Engineering, Hangzhou Dianzi University, Hangzhou 310018, China;College of Telecommunication Engineering, Hangzhou Dianzi University, Hangzhou 310018, China;College of Telecommunication Engineering, Hangzhou Dianzi University, Hangzhou 310018, China
Venue:
Proceedings of the Second International Conference on Innovative Computing and Cloud Computing
Year:
2013

Citing 6
Cited 0

Total Variation Models for Variable Lighting Face Recognition

IEEE Transactions on Pattern Analysis and Machine Intelligence
Robust Face Recognition via Sparse Representation

IEEE Transactions on Pattern Analysis and Machine Intelligence
Face recognition under varying illumination using gradientfaces

IEEE Transactions on Image Processing
Face recognition under varying lighting conditions using self quotient image

FGR' 04 Proceedings of the Sixth IEEE international conference on Automatic face and gesture recognition
Compressed sensing

IEEE Transactions on Information Theory
Near-Optimal Signal Recovery From Random Projections: Universal Encoding Strategies?

IEEE Transactions on Information Theory

Quantified Score

Hi-index	0.00

Visualization

Abstract

As the theory of Compressive Sensing being put forward, people began to study how to apply CS for face recognition. Allen Y. Yang and John Wright proposed a new method based on CS, named Sparse Representation-based Classification (SRC). SRC cast the recognition problem as one of finding a sparse representation of the test image in terms of the training set as a whole. SRC addressed the recognition problem with illumination invariant images successfully. However, SRC didn't consider the variable illumination and the complexity of calculation. In this paper, we propose an illumination robust algorithm based on SRC, named G_SRC. Moreover, our algorithm is faster than SRC. G_SRC applies the Gradientfaces to preprocess face images and extract illumination invariant. Then, G_SRC applies PCA to extract face feature and reduce the image dimensions. At last, we use SRC to face recognition. G_SRC can weaken the influence of light on the human face and improve the recognition. Experimental results on the ORL and the actual face databases show that G_SRC has better generalization ability than SRC for face recognition.