Fingerprint Registration Using Zernike Moments : An Approach for a Supervised Contactless Biometric System

International Journal of Image Processing

Antoine VIANOU¹  Marc Kokou ASSOGBA¹  Tahirou DJARA¹  Amine NAIT-ALI¹ 

[1] $$

关键词: Contactless Biometry;    Fingerprint;    Zernike Moments;    Image Registration.;   

DOI  :  

来源: Computer Science Journals

【 摘 要 】

In this work, we deal with contactless fingerprint biometrics. More specifically, we are interested in solving the problem of registration by taking into consideration some constraints such as finger rotation and translation. In the proposed method, the registration requires: (1) a segmentation technique to extract streaks, (2) a skeletonization technique to extract the center line streaks and (3) and landmarks extraction technique. The correspondence between the sets of control points, is obtained by calculating the descriptor vector of Zernike moments on a window of size RxR centered at each point. Comparison of correlation coefficients between the descriptor vectors of Zernike moments helps define the corresponding points. The estimation of parameters of the existing deformation between images is performed using RANSAC algorithm (Random SAmple Consensus) that suppresses wrong matches. Finally, performance evaluation is achieved on a set of fingerprint images where promising results are reported.

【 授权许可】

Unknown