Video Face Recognition Based On External General Dictionary

Recently, video surveillance has been widely applied in various fields.Accordingly, the intelligent surveillance based on face recognition has been considered a promising technology. With the development of science and technology, face recognition technology has made considerable progress. However, because of the difficulty with the prediction on the facial variations of a query sample by the gallery samples in a certain scenario, face recognition with single sample per person is still a critical challenge. Due to the fact that the intra-class facial variations canbe shared across different subjects. Thus, based on the theory of compressed sensing.This paper propose an external general dictionary, which improves single sample face recognition rate. More specifically, this methodresorts to construct a generic variationdictionary D to predict the possible facial variations(e.g. illuminations, pose, expressions).by usingthe difference betweenthe reference subset and variation subset, we could get a dictionary. Then after the dictionary was introduced to face recognition algorithm, it provides a considerable accuracy even under the condition of the single sample per person. In this paper I employ there database like AR face database; SCface surveillance face database; ORLface database verify the proposed algorithm. Compared with the traditional algorithm, the accuracy of the proposed algorithm increases by nearly 15%.This algorithm is not only illumination-insensitive but alsoretained the advantages of the single sample face recognition, which is using theless training sample number to improve the system usability Furthermore, making one sample video takes step forward towards commercial face recognition system is more.