Adaptive Wavelet Threshold Function for Fingerprint Image De-noising
Lei Li · Journal of Information and Computational Science · 2014
Based on the characteristics of fingerprint image as well as the traditional soft and hard threshold functions for the fingerprint image de-noising, this paper puts forward a new threshold function. The result of the experiment confirms that if the parameter of the new threshold function is appropriately defined, the de-noising performance is better than those methods [1-3]. In the simulation experiment, a fitting model about the noise intensity and the decomposition level is proposed on the basis of a large number of experiment data. The model can adaptively avoid the noise. Hence the new threshold function is supposed to help to improve the efficiency of the fingerprint image de-noising and to have certain application value.