The Study Of Fingerprint Identification Based On The Deltas And Cores

Biometrics is a technology which identifies a person based on his/herphysiological characteristics. AFIS(Auto Fingerprint Identification System)is a more practical,ideal bimetric identify technique relative to the technologyof face recognition or other biometrics recognition.The paper introduces all of the essential point.,discussing grayfingerprint image pretreatment in details and completing the extracting ofdeltas and cores.Basing on the cores and deltas we make the image classifiedand study the fingerprint identification in details.There are 5 chapters in the paper.Chapter 1 is Introduction. Chapter 2 isthe description of the essential point in the AFIS. Chapter 3 is the extractingof glocal earmark. Chapter 4 studyies the minutia matching algorithm. At lastis the summarization and prospects.The paper gets the images by the fingerprint gathering instrument, andstudies the theories of fingerprint deeply based on the the PC. Discussing therealization of the pretreatment in details ,the paper designs a kind of themethods of extracting of glocal earmark which has been improved. Basing onthe the classfication algorithm, we put forward the new ways of fingerprintidentification.The main jobs that we do in the paper as followed:(1) Establish the AFIS in hardware based on the PC and the fingerprintgathering instrument.(2) Study the rule for the auto biometrics recognition. Design the AFIS by theaspects of working pattern, the constituent parts of system and the algorithmflow.(3) Put forward and discuss several essential points which appear inAFIS.And make the introduce for the gathering images,enhancing,extractingorientation,extracting global extracting minutiae and theways for the basic algorithm in fingerprint identification.(4) Research the technique in pretreatment.For the extracting orientationrealize the adjustable direction algorithm and the rapid algorithm from thegrey picture directly of fingerprint . For the results using both of the twomethods we make the analysis and comparison. From the testing we can see that the different combinations of two-valuealgorithms and thin algorithm have different experimental effects. We shouldpay attention to that, despite of any two-value and thin algorithm, more falsetraits in which comprise lots of false bifurcation points or false end pointswill be produced after thinning for the bad quality of original fingerprintenhanced picture and have the relative centralized distributing. For the areasof good quality will have the desired effect after thinning. In the part of the thinning, the paper realizes the Hilditch thin algorithm,the rapid thinning and template matching thin method. Discuss the differentcombinations of the methods and the different effects of the experiment. Improve the algorithm of the extracting of minutia of fingerprint. First,based on the deleting of the burrs and short-lines etc on the picture afterthinning, get the initial eigenvector. Then analyze the traits of all kinds ofnoises existing in the fingerprint picture and consider the distributing law andthe direction of local lines to adopt the relative algorithms based on thedifferent noises. In this chapter, improve the original extracting of minutiaealgorithm and delete all the noises as the most possible for eliminating theeffects in extracting of minutiae from the noises. Experimental data show thatthis improvement definitely better the following speed but delete somecorrect information wrong. Though, the experiment indicates that if weprecisely extract 30 or so minutiae traits, the following matching algorithmwill work smoothly. So this improvement is worthy.(5) Locate precisely the core points and delta points of fingerprint so as toconfirm datum marks from the singularity points set in matching algorithm. ,the paper adopts the methods of Poincare function checked, and locatesprecisely the core points and delta points of fingerprint so as to confirmdatum marks from the singularity points set in matching algorithm. Bycomputing the distance between datum marks, and the number of the line ,wemake the finger images matched coarsely in order to accelerate the rapid ofthe matching.(6) Core points and delta points are considered in the matching algorithm.