| The social development and the safety about social life in the country are characterized with using information technology. Digital and network bring forward new requirements. Under this environment, the traditional security technology emerged insurmountable shortcomings. Society urgently needs a more convenient, safe and effective identity authentication technology. While, biometric technology is such the solution to this important issue on social security problem.Human biological authentication technology, known as biometric technology, is one kind of technology that use the physical characteristics which the human body itself inherents (such as fingerprint, iris, appearance, etc.) and behavioral characteristics (such as writing, sound, etc.), through the image process and pattern recognition approach to verify an individual's identity automatically. The core technology of biometric lies in how to get these biological characteristics, and converts them into digital information and stores them in the computer. Then it makes the use of reliable matching algorithm to complete the identification and verification of personal identity.In the different biological characteristics, the iris is the most robust and the most enduring features. As identification features, compared with other biological characteristics, iris has more good natures: invariability, stability, acquisition, unintrusion, and anti-forgery. In the recognition process of biometric, unintrusion study is the inevitable trend for the identity identification and application. Compared with human'face, voice and other non-contact identity identification method, iris recognition has a higher accuracy. According to the statistics, the error rate of Iris Recognition is the lowest among all those biometric technologies. Based on iris recognition system which has a higher rate of accuracy in the long run, the iris identity recognition technology gradually drawn the attention of the academic and the business community. However, as in the process of imaging, Iris is affected by collecting equipment, light changes, defocus, and other factors, which make iris image restoration one of the important research topics in iris recognition.This paper firstly gives a detailed explanation on the role of human biological certification in real life, the iris image restoration methods, the technology development, as well as some classical methods of the restoration. And it pointed out the study focus and difficulty in current days in the area of iris image restoration research.This paper also summarizes some new methods of the iris image restoration and expatiates the process of the iris restoration realization detailly. A complete iris recognition system consists of iris image acquisition, iris image preprocessing (including the iris image denoising, image restoration, image enhancement) and other taches.On the basis of predecessors'fruit on iris image De-noising and restoration work and combining the author's work in the iris recognition research, the paper does a deep study on iris image acquisition, pretreatment, and the algorithm that iris image restoration to the image recovery technology. And for the practical needs of the iris image restoration technology, I raised several new improved algorithm methods on the important and difficult points in the field of the iris image restoration, and I also did a number of useful attempts on iris image process. As a result, I achieved somel results. The main contributions and the study results are as follows:1. the relative study on iris image restoration both at home and abroadThe research of the iris image restoration at home and abroadIn Iris recognition, the more important step is iris collection. The iris recognition system proposed by Daugman has been very mature, but the precondition is on obtaining high-quality images to identification. Since the majority of Iris acquisition devices do not have automatic focusing function, the images that we have got are often defocus. The blurred images will bring great difficulties for Iris Recognition. How to obtain a high-quality iris image is becoming an important research topic.2, This paper focuses on the study to the defocus iris image, and restores the blurred images that the focus collection equipment got. And it put forward the following improved algorithm methods: (1) First, it analysises the defocus reasons of iris image, and proposes the auto-collimation iris image algorithm method based on the power spectrum.(2)Finally, to use the regular method for image restoration, thus we gained a clear iris imageIn the end of my paper, using the iris image database provided by CASIA, we gave a systematic test on the algorithm method in the PC machine, in which the CPU is 2.4 GHz, and EMS is 256. We got a lot experimental datas and the given test results finally.The theory and experiment prove that, the several improvements algorithm methods in the paper are superior to the traditional methods of iris image restoration, which can effectively overcome the impact of blurred Iris image caused by iris image defocus, light. And it has fast and accurate features.Iris image processing technology is rich in content. And it also has a very broad application prospects. The work on this paper is small. There are some things that the author does not have time to do. And the work done by the author also has many imperfect points. In the future study, great efforts should be made to realize the algorithm improvements or propose a new algorithm method, and do a great deal of experimental samples to verify the algorithm method.Although iris recognition technology has appeared in the 1990s, because the hardware under that condition was not developed enough, the algorithm research is not thorough enough. So it had not reached the stage of large-scale applications. The computer hardware is not what it was before. The stander of product performance is increasing. The price is decline relatively. Mathematics and Computer Science has also developed continuously. All those give the support to the development and application on biological character recognition technology. I believe that biometric technology will develop fast from theory into practical products. The performance of products will be more reliable, and the cost will be reduced. Among these, as the numerous advantages on iris, as well as the continuing development and mature of iris image restoration methods, iris recognition technology will become a model in biometric identification technology.
|
PDF Full Text Request