Research On Removing Noise For Iris Image

With the development of information technology in contemporary society, the demands for security are becoming more and more important. So the technology of biometrics identification was improved and applied on many fields. As iris identification is one of potential biometrics technologies, it has been gradually applied to safety fields. Iris identification is based on iris texture to identify persons, and it is one of most accurate biometrics technologies. Because the iris recognition is in the ascendant abroad, and in domestic it is just on the beginning. So, it is important and valuable to analyze processing the iris imaged soundly.The paper discusses and summarizes the particular advantage of iris for recognition firstly. And then introduce the structure of iris recognition system. Owing to iris noise in iris region, the probability of iris recognition is very low. So technology removing iris image noise is one of keys in iris recognition system. After the preprocessing of iris image is analyzed, removing iris image noise is mainly studied. The removing methods which were proposed in some documents are applied to removing iris image noise for attempts. But the experimental results demonstrate that those methods aren't suitable for iris segmentation.Iris images which have been located and normalized still include iris noise, such as sclera, pupil, eyelash and eyelid. Then the iris noise is removed, the phase information is more important compared to amplitude information because the difference of amplitude information is not obvious and is easily influenced by outside factors such as contrast of iris image, intensity of lamp-house and amplification of optical system. The paper proposes a virtual method of removing based on phase congruency. By locating and normalized iris images with log Gabor filters, we can acquire information of image texture in a series frequencies and orientations. Iris noise regions are determined based on phase congruency by a group of log Gabor filters whose kernels are suitable for edge detection. These regions are filled according to characteristics of iris noise. The experimental results show that the proposed method can remove iris images effectively.The algorithm of encoding iris texture based on 2-D Gabor wavelet is applied to encode, and the algorithm of matching based on Hamming Distance is applied to matching two iris eigenvectors. When the method proposed in paper and another method proposed in document are performed respectively, the experimental results of 36 iris...