Palmprint Recognition Using 2D Gabor Filter and Hamming Classifier
Fingerprint identification has drawn considerable attention over the past 25 years. With the rapid development of hardware and other devices, other biometric features of people come into use, for example face, voice, and palmprint etc. Palmprint has its own advantages considering some flaws in fingerprint collecting. This paper presents one new method on palmprint recognition using 2D Gabor filter and hamming classifier as Gabor filters have the advantages of accurate time-frequency location and providing robustness against varying brightness and contrast of images and Hamming classifier is very easy and simple for usage. The method proposed in this paper is as below: Firstly, the palmprint feature on a low-resolution palmprint image is extracted and encoded using 2D Gabor filters; Secondly the hamming distance in the real part and the imaginary part of the palmprint feature gained from the first step calculating are calculated, which support the matching; Lastly, the public and famous PolyU palmprint database is adopted to evaluate the performance of palmprint feature extraction. The experiment result not only illustrates the effectiveness of the method, but also shows that the recognition accuracy is related to the different sets of parameters of the Gabor filter.
Palmprint recognition 2D Gabor filter Hamming distance feature extraction
WANG Xi LU Jiwen
Department of Information Science, Xian University of Technology, Xian, 710048
国际会议
北京
英文
2007-08-05（万方平台首次上网日期，不代表论文的发表时间）