Font Size: a A A

A Study On The Blur Iris Images Texture Feature Extraction And Recognition

Posted on:2012-06-14Degree:MasterType:Thesis
Country:ChinaCandidate:K Y ZhangFull Text:PDF
GTID:2178330332492730Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
Iris recognition system is one of research directions of biometric technology. Compared with other biometric technology, iris has the advantage of unique, stability, easy-collected and non-invasive, iris recognition system has been used in customs and some high-security fields.This paper analyzes and compares iris research status at home and broad, at the same time, it finds out the gap which iris images are researched between domestic and overseas and the shortcomings of the existing system. In favor to improve existing iris recognition systems, we work hard to struggle so as to achieve a kind of fast speed and high recognition rate's product to make things convenient for the customers.In order to further expand the depth scope of collection system, the purpose is the iris images more suitable for iris recognition systems in different position to make things convenient to use for the customers. This paper mainly researches which the optic system is invariant to defocus the blur iris images in iris recognition application. There are two methods to extract iris features in order to find out stable feature points. A method is the iris recognition algorithms based on texture stable feature points and anther method is the iris recognition algorithm based on extracting the sequence images of stable feature points. The two algorithms'core thoughts are that iris feature codes are extracting from changing images, the stable feature points can be extracted from the sequence feature templates of the same iris. At the same time, the stable points are extracting from many iris images to implement and-operation. The stable feature is that invariant feature and the unstable feature is that changing feature. Only if that unstable feature points are eliminating, correct recognition rate is improved and the proposed algorithm is effective for improving iris recognition performance. A large number of experimental results show that the two algorithms can be applied in the actual iris recognition system. At the same time, the proposed method is effect guiding value in iris recognition system's research development fields. Two algorithms are applied to iris recognition system. This paper designs and builds based on asynchronous electrical machine control linear unit to gather iris images experimental platform and iris recognition application software. The experimental platform is used to gather iris sequence images, in order to extract stable feature points from iris sequence images. Iris recognition application software proves that this paper's design algorithm is effective and feasible in practice.
Keywords/Search Tags:Biometric Feature Identification, Defocus and Blur, Iris Recognition, Stable Points
PDF Full Text Request
Related items