Improved face recognition algorithm using extended vector quantization histogram features

In this paper, we propose an improved face recognition approach based on the combination of Vector Quantization (VQ) and Markov Stationary Feature (MSF) which obtain the extended MSF-VQ features from facial sub-regions for face recognition. It can not only utilize the MSF framework to extend the VQ histogram based features with the spatial structure information but can also incorporate more location information extracted from different facial sub-regions so as to improve the accuracy of face recognition system. We demonstrate our proposed algorithm utilizing FB category of FERET face database and the maximum top1 recognition rate of 97.6% is obtained.