Font Size: a A A

Research And Implementation Of The Preprocessing Algorithm Of Iris Recognition

Posted on:2014-05-29Degree:MasterType:Thesis
Country:ChinaCandidate:Y F HeFull Text:PDF
GTID:2268330401465657Subject:Communication and Information System
Abstract/Summary:PDF Full Text Request
Biometric identification technology is widely applied in many fields, it plays animportant role to improve information security. Iris recognition as a new technology inbiometric identification technology, because of the stability, uniqueness and advantagesof non invasive and easy collection already have quite a number of applications. Irishas higher accuracy compared with other biometric technologies methods such as face,fingerprint, voice. Iris recognition system main includes iris sample acquisition, irisimage preprocessing, iris feature extraction, identification authentication.This paper finish the following works:1. Based on the iris texture feature, the method is applied to judge the iris imageclarity. In this paper, based on three main factors: the Fourier frequency spectrumcomponent of the iris, the sharpness of pupil boundary and the gradient engine ofiris to design a optimal hyper plane to evaluate the clarity of iris.2. Iris segmentation of the inner and outer boundary. We abandoned the classic irislocalization algorithm based on edge extraction of Hough transform detection in irisimage positioning. We take the method that based on binary gray level and graylevel gradient to fit the inner boundary. Iris outer boundary positioning is toestablish circular template classifier to eliminate the noise points and segment irisboundary point class. Combining with circle fitting method, to fast accurate fittingiris outer boundary.3. We translation,rotation and scaling transformation the segmented iris image. Thenstretching the histogram after normalization to enhance the iris texture details.These algorithms has simulated on Matlab platform. Then debugging andoptimizing in C++language. The result shows a significant improvement comparedwith prvious version.
Keywords/Search Tags:Iris recognition, Quality assessment, Clarity, Iris localization, Circularclassifier
PDF Full Text Request
Related items