Font Size: a A A

Study And Application On Key Technologies Of Iris Pattern

Posted on:2011-01-25Degree:MasterType:Thesis
Country:ChinaCandidate:A X WuFull Text:PDF
GTID:2178330332962707Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
Biometric recognition technology is one of the hottest research fields, which is the multidiscipline technology,including biology, electronics, computer, anatomy, etc. Iris identification is potential because of its stability, uniqueness, non-invasive, can be collected advantages gradually attention has been paid. And the face image, voice, fingerprints, authentication methods, the iris is more accurate. In recent years, iris recognition technique and applications in research have been great progress, and has shown promising and the market.This paper studied the correlation of iris recognition technique, and developed a set of authentication based on iris recognition system. This system is composed of iris image preprocessing, iris image positioning, the normalized iris image, iris feature extraction and matching of several parts.Iris image preprocessing is based on iris recognition system and the key preprocessing directly affect the following feature extraction process and the matching process. In the pretreatment process, in view of some large traditional pretreatment method, the memory overhead of a large, long search time, the paper suggests the use of median filtering and improved salt and pepper filter two methods to filter the iris image to achieve a good denoising effect, this method also has simple, memory space overhead, time-consuming and less features.Positioning the iris image is divided into the outer edge of inner edge location and orientation. When the edge in positioning, the system first on the iris image segmentation, and then point Hough transform used to determine the pupil's center and radius; in positioning the outer edge, it will Daugman improve circulation integral operator in a certain region area integral operator, has been calculated using the inner circle's center and radius to search for the center of the circle and the radius of cylindrical.In normalized terms, this article details the first choice rich texture of the iris region, and then use the method of extracting line segments has selected the iris region normalization, which will ring normalized iris region into a rectangle. In the iris image to start the process, inevitably some point a new image with the original image can not match the point system for this situation to deal with bilinear interpolation. In order to improve recognition results, the system of the iris image after the commencement of the local histogram equalization.At the same time the system uses Gabor transform to encode the iris texture, and encoded using Hamming distance of the iris image pattern matching. The experiments show that the realization of personal identification based on iris recognition system to achieve a high recognition rate.
Keywords/Search Tags:Iris recognition, image preprocessing, iris positioning, Feature extraction, Pattern matching
PDF Full Text Request
Related items