Fingerprint Enhancement Using Coherence Enhance Diffusion

Automatic fingerprint recognition technology is now one of the most commonly used biometric identification technologies.Usually it includes three stages,namely,fingerprint image preprocessing(fingerprint image segmentation,enhancement,refinement),feature extraction and matching,However due to surface conditions of fingers,such as the fingers are too wet,with scars,dry,or the reason of collector itself(Such as stains on the instrument)makes the collected fingerprints with a variety of different noise,so the acquired fingerprint image needs to be enhanced to facilitate the subsequent steps of the treatment.In this paper,we mainly study several fingerprint enhancement methods,especially the use of coherence enhancing diffusion method to enhance the fingerprint image.To begin with,this paper presents the importance of fingerprint recognition technology and fingerprint image enhancement.The next,we summarize the algorithms of fingerprint image enhancement,such as Gabor filter and STFT filter.Once more,the fingerprint image is enhanced by the combination of stationary wavelet transform or shock filter with coherence enhancing diffusion.Finally,summarize the current work and discuss the need for further improvement of the future research direction.The main research work of this paper is as follows:First,to solve the problem of the enhanced fingerprint image will appear the edge blur and low contrast of fingerprint ridges,we propose a weighted model method using shock filter and coherence enhancing diffusion.This new method cannot only keep the advantage of coherence enhancement diffusion,but also sharpen the edge of the fingerprint ridge improving contrast of fingerprint ridges.Second,in order to solve the problem of the enhanced fingerprint image will appear edge blur and feature points fusion using coherence enhancing diffusion,we put forward a method of combining wavelet transform and coherence enhancing diffusion to maintain the advantage of coherence enhancement diffusion,and to reduce the difference between the original fingerprint image and the enhanced fingerprint image to eliminate the drift of fingerprint minutiaes.