Reduction Of Noise From Fingerprint Images Using Stationary Wavelet Trasnform
Nasar Iqbal · Zenodo (CERN European Organization for Nuclear Research) · 2017
In Automatic Fingerprint Identification Systems (AFIS) the quality of image is a very important factor as the minutiae extraction from fingerprint image heavily depends on image quality. To enhance the quality of fingerprint images a large number of denoising methods has been used. In this paper fingerprint image enhancement using stationary wavelet transform has been analyzed using different wavelets with different thresholds. Four different wavelets namely Haar DB4 (Daubechies), Coif2 (Coilflets) and Bior1.3 (Biorthogonal) were selected with four thresholds namely VisuShrink, NormalShrink, NeighShrink and BaysShrink. The methods were applied on three types of noises which were Speckle noise, Gaussian noise and Salt and Pepper noise. The effect of changing decomposition level on noise removal efficiency based on PSNR (Peak Signal to Noise Ratio).