Efficient feature extraction using hybrid face recognition method

Recognition of faces for human is always a challenging task due to the absence of sufficientfeatures. To solve this problem we propose a novel and efficient face recognition method which is invariant to slight variation in pose, illumination and background. Proposed method uses discrete cosine transform (DCT) and gray-level co-occurrence matrix (GLCM) for the extraction of both visual features and texture features respectively. The Proposed method helps in removing some high frequency details and thus reduce the size of images. These visual features and texture features are fused to form the hybrid feature leading to improve the feature space and nearest neighbor classifier is used for the classification purpose. The proposed method shows better result on variation of background and pose and it also reduces recognition time greatly. This method improves the efficiency and performance of existing method.