Font Size: a A A

Research On Palm Vein Recognition Algorithm

Posted on:2010-03-29Degree:MasterType:Thesis
Country:ChinaCandidate:J WangFull Text:PDF
GTID:2178330338979010Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
In recent years, biometric technology has been rapidly developed, such as: fingerprint recognition, iris recognition. Biometrics technology is widely in various applications. Currently palm vein recognition technology occupied an important position in the field of biometric identification. The palm vein matching algorithm is directly related to recognition speed and recognition rate of the palm vein recognition system. So,this article focuses on the palm vein matching algorithm.Currently, the research literature on palm vein recognition technology is relatively less. This article has draw primarily back of palm, fingerprint matching algorithm and so on. In order to improve the recognition rate of the algorithm, this paper summarize, analyze and sum up existing algorithms, study on palm vein matching algorithms in depth. The main works are reflected in the following areas:Firstly, the image correction algorithm is researched, and a fast correction algorithm is proposed. When collecting images, because non-contact collection devices, the image is very prone to translation and rotation. Therefore, the image is difficult to be identified. How to determine the calibration parameters is difficult in the correction algorithm, calibration parameters need to the two sets of corresponding matching points, this paper uses the centers of the endpoints and the centers of the intersection points as the two sets of matching feature points, the calibration algorithm greatly reduced. time complexity.Secondly, because the matching algorithm based on the minutiae is difficalut to process the faint image, based on Gabor wavelet energy and direction characteristics of palm vein recognition algorithm was proposed. The 2D Gabor transform is widely used in palmprint and fingerprint's image processing and recognition. Algorithm used 2D Gabor filter to filter palm vein image in different directions at different scales, and then extract the feature of the energy and the direction. The difference of Feature vector was calculated. Finally, the palm vein image was identify by two characteristic values. The experimental results show that the algorithm has higher recognition rate. Thirdly, if a matching algorithm only used one feature, the recognition rate of the algorithm is relatively low, the range of identification is relatively narrow. So a multi-feature fusion palm vein matching algorithm was proposed. In order to improve the recognition rate, reduce false acceptance rate and the rejection rate, algorithm has proposed a first calibration image and then matching image. A fusion of correlation coefficient, invariant moments and Hausdorff distance matching algorithm made up for the weakness of the only one feature matching. Algorithm greatly enhanced the recognition rate of the algorithm, expanded the scope of application of the algorithm, solved the problem of the narrow scope of a single algorithm application and the problem of the low recognition rate of special circumstances.
Keywords/Search Tags:Image Matching, Image Correction, 2D Gabor Transform, Hausdorff Distance
PDF Full Text Request
Related items