The enhancement of fingerprint plays an important role in automatic fingerprint identification system. It is the foundation that extracts fingerprint minutiae reliably. In order to make up for the defects of fingerprint image and enhance low quality fingerprint image effectively

this paper proposes a novel algorithm by using spectrum diffusion to enhancing the fingerprint image. The basic idea of the algorithm is to enhance the fingerprint by using block quality assessment and block spectrum matched filter in the frequency domain. First

the quality assessment result of block image is obtained according to the point direction coherence parameter. Second

this paper studies the diffusion characteristics of composite window spectrum

and then design out a block spectrum matched filter based on linear fuzzy k-means clustering and 2D Butterworth band-pass filter of one order adaptive bandwidth and two-dimensional Gaussian window. At last

the composite window spectrum is enhanced by block spectrum matched filter according to the block quality grade. Experiments on the database of FVC 2004 show that

compared with the previously proposed enhancement methods

the proposed method is more accurate and more robust against noise

and has a prominent effect on both low quality and high quality fingerprint image. This paper presents a fingerprint enhancement using filter method of weak parameters dependencies in frequency domain. This method is of good fault tolerant. The experimental results show that both high quality and low quality fingerprint image can be enhanced effectively.