Method for recognizing iris with matched characteristic and graph based on partial bianry mode

Abstract

The invention discloses an iris recognition method based on local two-value mode feature and graph matching. Firstly local two-value mode code is picked up according to the ordinal relation of each two pixel grey-scale values in an iris image neighborhood for describing the iris oriented texture feature having a constant characteristic under light irradiation; secondly the iris image is divided into a plurality of image pieces, and a local two-value mode histogram in each piece is calculated to describe the iris oriented texture statistical characterization having robustness on translation and deformation; each image piece is viewed as a node, the local two-value mode histogram in each piece is viewed as the attribute of the node, and the feature of each iris image is presented as a graph mode; during iris recognition, the graph matching method is used for searching matching node pairs in two graph modes; image recognition and registration of the number of the matching node pairs in the graph modes are carried out to show the similarity between the two graph mode, thereby judging the ID of a user. The invention is used for automatic identification in the application fields such as entrance guard, attendance check and clearance.

Description

technical field [0001] The invention relates to digital image processing, pattern recognition, computer vision and information coding, especially an iris recognition method based on Local Binary Pattern and Graph Matching. Background technique [0002] An effective identity authentication method can bring safety, convenience and benefit to our life, and can bring stability, order and efficiency to our society. In a modern society with highly developed networks and frequent personnel exchanges and flows, people have higher expectations for identification technology: automatic, fast, accurate, anti-counterfeiting, convenient, and non-repudiation. Driven by demand, identification methods based on biometric features such as fingerprints, faces, irises, hand shapes, and handwriting have emerged as the times require, trying to replace traditional methods based on knowledge or markers. [0003] The iris is a ring-shaped area located between the black pupil and the white sclera on ...

AI Technical Summary

Problems solved by technology

[0007] Like feature extraction, feature matching is also an important link in the iris recognition system, but it has been neglected for a long time in the field of iris recognition

The existing iris feature matching methods are all statistical classifiers, such as Hamming distance, Euclidean distance, correlation coefficient, neural network, etc. There is no matching algorithm based on structural information

Images