Analysis of Proposed Noise Detection & Removal Technique in Degraded Fingerprint Images

Abstract

The quality of fingerprint images is important to ensure good performance of fingerprint recognition since recognition process depends heavily on the quality of fingerprint images. Fingerprint images obtained from the acquisition phase are either contaminated with noise or degraded due to poor quality machines. Several factors such as scars, moist in scanner and many more noises affect the quality of the images during scanning process. This paper performed an analysis and compared noise removal techniques reported in the literature for fingerprint images. We also implemented histogram equalization, median filter, Fourier transform, unsharp mask and grayscale enhancement techniques. The quality of enhanced images is measured by peak signal to noise ratio (PSNR) calculation for analysis and comparisons.