Font Size: a A A

The Research Of Palmprint Recognition Algorithm Based On Moment Invariants

Posted on:2011-11-24Degree:MasterType:Thesis
Country:ChinaCandidate:H Q LiangFull Text:PDF
GTID:2178330332488268Subject:Computer system architecture
Abstract/Summary:PDF Full Text Request
With the development of information technology, it is the security of information have become increasingly demanding. In this case, traditional identification methods have been unable to meet the needs of social development, so palmprint recognition technology would come about.This paper in-depth research of the palmprint image positioning and the palmprint feature extraction. In the palmprint image pre-processing stage, combined with palmprint image capture device, this paper applies a method used in locating the ROI which based on hand edge points. In the palmprint feature extraction section, based on palm natural multi-resolution features, presents a palmprint feature extraction method which based on wavelet transform and moment invariants. The proposed scheme first decomposes a ROI image at different scales, so we get a series of similar images. After divide each similar image into some sub-images, extract features vector by computing moment invariants of each sub-image.In order to assess the performance of the algorithm, simulation experiments of the Hu moments and Zernike moments feature extraction and matching carry out in a large database, and compared the Hu moments and Zernike moments of recognition rate, verification of the Zernike moments feature extraction method superiority.
Keywords/Search Tags:Palmprint Recognition, Wavelet Analysis, Invariant Moments, Zernike Moments, Hu Moments
PDF Full Text Request
Related items