Combining Speeded-up Robust Features with Principal Component Analysis in Face Recognition System

Recently, the techniques of face recognition have been widely used in security application such as security monitoring, and access control. However, there are still some problems in face recognition system in which the light changes, expression changes, head movements and accessory occlusion are the main issues. In this article, a robust face recognition scheme is proposed. Speeded-Up Robust Features algorithm is used for extracting the feature vectors with scale invariance and pose invariance from face images. Then PCA is introduced for projecting the SURF feature vectors to the new feature space as PCA-SURF local descriptors. Finally, the K-means algorithm is applied to clustering feature points, and the local similarity and global similarity are then combined to classify the face images. Experimental results show that the performance of the proposed scheme is better than other methods, and PCA-SURF feature is more robust than original SURF and SIFT local descriptors against the accessory, expression, and pose variations.