Font Size: a A A

The Study And Implementation Of Iris Recognition Algorithms Based On Multi-scale Grayscale Matching

Posted on:2012-10-22Degree:MasterType:Thesis
Country:ChinaCandidate:C Y WuFull Text:PDF
GTID:2218330338469596Subject:Basic mathematics
Abstract/Summary:PDF Full Text Request
As one of the most effective ways in biometrics, iris identification has been drawn more and more attention in recognition technology and widely applied to various fields and industries. However, iris identification is easily affected by noises such as eyelashes, eyelids, sampling flares, and its quality and speed in actual practice is not so effective. Based on feature extraction and matching code way used widely. In the paper, we systematically introduced in research fields of the iris recognition technology, and made analysis and research on key technologies such as iris orientation and segmentation, iris images normalization, iris feature point extraction and matching.And then designed an iris identification algorithm based on multi-scale gray matching. Then it proposes a standardized solution for the noise integration of using in login and sample iris images, and uses directly surface match way to judge recognition.Based on this method, we can extract more junior primitive iris data, analyses and process them, and increase area used to iris identification, reduce residual caused by the no unified available area. In the thesis it is put forward two methods to improve the matching speed; The first is to seek for steady recognition results through multi-resolution, the Second is extract iris information quickly through sure matrix Using the MATLAB7 .0 platform ,we designed the multi-scale gray iris identification system, did simulation experiments for the Version1.0 Database of iris images.The results show that the system identification has virtues of accuracy and speed, and has some applicated fields.
Keywords/Search Tags:Iris identification, Gray surface matching, Iris segmentation, Feature extraction, Noise unification
PDF Full Text Request
Related items