Title :
Research on fingerprint enhancement and classification algorithm
Author :
Liu, Yuan-ning ; Li, Zhi-hui
Author_Institution :
Comput. Dept., JiLin Univ., China
Abstract :
In this paper, because Gabor filter has the optimum connection resolution ratio of the time and frequency field and its good band-pass, Gabor filter is used to filter the noise of fingerprint images with low quality. The ridge voting algorithm is presented to get ridge structure of fingerprint images, based on which the direction information of fingerprint images is extracted, and the image is enhanced by use of direction information. At the same time, based on K. Karu´s (1996) classification model of core and delta, its applicable condition and physical meaning is analyzed, and the topologic structure and characteristic is considered, moreover, the strict constraint condition and relevant concept, definition are created in our paper. Its correctness is enunciated in theory, and fingerprint experiment results verify the validity and correctness of the previous enhancement and classification algorithm.
Keywords :
feature extraction; filtering theory; fingerprint identification; image classification; image denoising; image enhancement; Gabor filter; constraint condition; direction information extraction; fingerprint classification; fingerprint enhancement; fingerprint images; image classification; image enhancement; noise filtering; optimum connection resolution ratio; ridge structure; ridge voting algorithm; topologic structure; Band pass filters; Classification algorithms; Data mining; Fingerprint recognition; Frequency; Gabor filters; Image matching; Image resolution; Signal to noise ratio; Voting;
Conference_Titel :
Computer and Information Technology, 2004. CIT '04. The Fourth International Conference on
Print_ISBN :
0-7695-2216-5
DOI :
10.1109/CIT.2004.1357197
