Adaptive filter in low quality fingerprint images

A method based on BEG algorithm and wavelet transformation was introduced to enhance the image fingerprint quality. First, a method to estimate the local image gradient, called the BEG (best estimation of gradient) algorithm was proposed to extract the texture orientation parameters. Then the frequency parameters in local fingerprint window were calculated. At last an adaptive Morlet wavelet filter was used to enhance the fingerprint image according to those parameters. The performance test shows that this method is valuable in practical uses.