Research On Fingerprint Enhancement Algorithm Using Gabor Filter

With the rapid development of science and technology,the fingerprint verification has become the common way of personal authentication.In the process of fingerprint acquisition,the images not only contain the background noise,but the problems of fracture and adhesion in the ridges may also easily occur on account of the external environment factors such as humidity,scars and surface stains.Therefore,running the pretreatment algorithms before fingerprint matching is necessary.Gabor filters have both frequency selective and orientation selective properties,it is one of the most widely used fingerprint enhancement algorithm.In order to solve time-consuming problem of traditional Gabor filtering method,this paper analyze and optimize the intermediate steps,including fingerprint segmentation,orientation field,filtering and minute extraction and purification.In terms of fingerprint segmentation,a method based on genetic algorithm combined with linear classifier is proposed;In terms of orientation field,a new neighbor orientation template and a new orientation reliability calculation method combined with maximum differences of gray means are proposed;In terms of filtering,a discrete Gabor filters method is proposed,and block frequency is divided into three levels,match the different size of the Gabor filters;In terms of minutia extraction,a minutia purification method based on path tracking and path recovery is proposed;In terms of parameter testing,this paper design a human-computer interaction interface to show the distribution of orientation field,frequency field,image detail and computational results using different colors and lines based on MATLAB GUI tools.In order to verify the performance of the proposed algorithm in this paper,the experimental results such as GI(goodness index)values of the minute extraction and the average execution time are compared and analyzed.The experiment results show that the proposed fingerprint segmentation algorithm based on genetic algorithm have lower error rates and more practicable,and the discrete Gabor filters method realizes a higher efficiency and better real-time character.