Face And Palmprint Identification System Research

Biometrics, which refers to the automatic recognition of an individual by using certain physiological or behavioral traits associated with the person. It involves some knowledge about pattern recognition, artificial intelligence, image processing and other subjects.Face recognition is a widely researched and used biometric technology. In this paper, we propose a novel approach for face recognition using skin color model to locate face position, singular value decomposition (SVD) to extract features, and support vector machines (SVMs) to do the classification task. Compared with other traditional methods, SVD only needs a little computation to extract face feature effectively. SVMs are known to generalize well even in high dimensional spaces under small training sample conditions, which are typically encountered in the context of face recognition. The face images are captured with a USB CCD camera in our lab, using these images a recognition accuracy rate of up to 90% has been achieved by our face recognition system.Palmprint recognition is a relatively new branch of biometric technology, and Curvelet transform is a new mathematical tool that has been used for image processing, such as image denoising and contrast enhancement. When dealing with the edges, which are discontinuity curves in the image, Curvelet transform is more effective than other methods such as Fourier transform and wavelet transform. The digital Curvelet transform is revised in this paper and used to extract the palmprint features, with these features our palmprint recognition experiments achieve satisfactory results.Compared with single modal biometric system, multi-modal biometric system has several advantages. The information fusion in biometric systems can be achieved at three levels, i.e., feature extraction level, matching score level and decision level. In this paper, we combine the information of face and palmprint at these last two different levels respectively, the results of our experiments prove the advantages of multi-modal biometric systems.