Fingerprint Enhancement Using Wavelet Transform Combined With Gabor Filter
Weipeng Zhang,Yuan Yan Tang,Xinge You
DOI: https://doi.org/10.1142/S0218001404003861
IF: 1.261
2004-01-01
International Journal of Pattern Recognition and Artificial Intelligence
Abstract:The performance of automatic fingerprint identification system (AFIS) is heavily determined by the quality of the input image, thus an effective method to enhance the fingerprint image is essential in such a system. In this paper, we combine the filter-base method, which is mostly used nowadays with wavelet transform to achieve a more reliable and effective approach to fingerprint enhancement. This novel approach consists of five main steps, namely: (1) normalization, (2) decomposition, (3) wavelet coefficient adjustment, (4) Gabor filtering, and (5) reconstruction. Using this new method, a more clear fingerprint image can be obtained, which can distinctly improve the accuracy of the minutiae extraction module and finally achieve a better performance of the entire system. Experiments have been conducted in our study and positive experimental results have been received, which show that the proposed combined method is more effective and robust than other existing methods such as the filter-based and direct gray-level approaches.