The Technology Of Iris Recognition And Its Application In Identity Certification

With the development of science and technology and the popularity of Internet, identification is becoming more and more important. Because of the unique physiological characteristics, iris receives much concern in the field of biometrics recognition. Iris recognition is considered as one of the safest, most accurate identification method. It uses the features (rings, tiara, spots, wrinkles) in the Iris region of human eye image to form feature template, through which we compare and complete the identification.A complete iris recognition system should include the iris image acquisition, iris image preprocessing, feature extraction and feature matching. As a technology moving into society and application, iris recognition still has lots to be improved. This thesis aims to establish a more stable iris detection and identification systems, based of which we do a preliminary study at the image acquisition and image pre-processing issues. The thesis includes:1. The thesis introduced the theory of biometric feature identification. Besides the introduction of the classic fingerprint recognition, face recognition and speech recognition, iris recognition was treated importantly. The thesis also compared all kinds of biometric technologies. Then, through the introduction of iris physical features, iris recognition development and iris recognition system, the paper talks about the importance of iris identification in biometric identification.2. The thesis introduced the iris recognition system. The thesis elaborated the basic principles about iris identification in two aspects of Image processing and pattern recognition. On this basis, the classic iris recognition algorithms were compared, including Daugman’s phase-encoding method, Wildes and others’Laplacian pyramid-based image matching method and the Li Ma and others’space-based multi-channel filtering identification method.3. As for the bad identification result because of the unclear image and low resolution in the iris recognition system, the thesis proposed a super-resolution iris image reconstruction method based on compressed sensing. With the help of compressed sensing theory, the iris image was super-resolution reconstructed and ultimately clear high-resolution iris images could be abtained. Experiments show that the method has significantly improved the recognition rate of iris recognition systems and effectively improved system performance.4. Based on the existing conditions, a set of MATLAB-based iris attendance system was built here. The system uses an iris acquisition device as iris image acquisition device. Based on the classic algorithm as technical support, the system adds the iris image reconstruction module. After debugging, the system is stable and robust.