A Novel Approach for Face Recognition Based on Weightiness Image Partition

In this paper we propose a novel method to improve the accuracy of face recognition . In the proposed method we partition the facial data into fifteen regions according to ´three courtyards and five eyes´ theory in pencil sketch on facial image in Chinese traditional art. Then in each partition we use LDA getting eigenvalues of feature and structure character to represent the facial data. We assign different weightiness to each sub-image according to the result of sub-image variety. Experiments on AR and ORL dataset show that the proposed method is robust and effective to face with expression, lighting and expression variance.