Abstract: Since most of existing fingerprint image enhancement methods using Gabor wavelet have the disadvantage that their filtering regions are fixed, an improved method is proposed. The performance of Gabor wavelet and inherent characteristics of fingerprint image are investigated in depth to determine the orientation of Gabor wavelet and relevant frequency parameters. And an area is set up as the filtering region to enhance a fingerprint image, which can be modulated automatically according to the change in local ridge orientation and frequency. Experimental results showed that the improved method can denoise the fingerprint images efficiently and that the information on ridge orientation is able to be well restored so as to improve the precision of subsequent operation for fingerprint matching.