The Research Of Algorithm About Drawing The Details Feature Point In Fingerprint In The Automated Fingerprint Identification System

The fingerprint is the pattern which forms by the nipple bulge friction crestline on the fi- nger terminal positive skin,everyone has the different fingerprint which is not changed all the life from eachother. Just it Contains a large amount of personal individual information so it has high rights of the person discern value. Pepole have clear understanding to the characteristic system of fingerprint progressively from untiring studying and exploration in thelasthundred years and develop the more complete fingerprint automatic recognition system reling on alarge amount of research of the a lgorithm drawning the fingerprint characteristic.The technology drawing the characteristic in the fingerprint is a very important part in the recognition technology of fingerprint. The quality matching the fingerprint characteristic will be effected directly if the fingerprint characteristic can not be abstracted accurately and effectivly. A lot of scholars have carried on the study abute the methods dawning the chara- cteristic of the fingerprint, which includes two mainly classes: the first one is drawning characteristic in fingerprint from the grey vision about fingerprint; the other one is from the fingerprint vision valuing 1 or 0. The second one is better from a large of experiments.But all the methods are strenthening the picture by the straining waves from estimateing the direction picture of fingerprint by analyzeing some part of the fingerprint. The resolution ratio will be reduced greatly if the quality of source fingerprint picture is relatively low, these methods are difficulte to get the accurate estimated values of the fingerprint picture direction and strained and strengthend the picture effctively. In fact, The existing algorithms drawning the characterristic in the fingerprint are difficulte to play a great role because most of the fingerprint picture are polluted and lowing resolution.Under this situation, in order to improve the function of the recognition system of fingerprint to deal with fingerprint of low quality, This article proposes a new algorithm of drawning the detail characteristic of the fingerprint by summarizing relevant research results. Thise article is divided into three parts altogether. The firs part describe emphatically the rudimentary knowledge of fingerprint and categorised overview; The second part describe the overview of the automatic recognition system of fingerprint,Basic principle that the fingerprint matches automatically and the content and characteristic of the automatic recognition system of fingerprint; The third part describe one improved algorithm drawning the detail characteristicsof fingerprint. In this part what describe emphatically is the improved algorithm drawning the detail characteristics of fingerprint from the two value fingerprint visions and how to get rid of some typical false characteristics from the results by compareint the algorithms from the grey level vision and the two valued visions of the fingerprint. The finishing part descibe the value and prospect of this article. This improved algorithm contain seven parts: The first one is the algorithm of the preconditionde vision of the fingerprint; This algorithm includes the rule melting,cuting apart and the strengthening about the fingerprint picture; The scond part is the two value of the fingerprint vision. The purpose of two valuing fingerprint picture is conversing the picture into 0 or 1. The third one is the subdivision of the fingerprint vision. The fourth part is about the improved algorithm rely on the point of extremeing and branching. The fiveth part is the algorithm of getting rid of the false characteristics which contains burr and short spine ; The sixth part is about the influence factors which describes maintly threshold value and the two vaule; The seventh part is about the efficiency of the algorithm.The algorithm in this article can drwan high quality detail characteristc of the fingerprint at the selected part in the fingerprint picture, but can not drwan the detail characteristic from all the fingerprint picture,especially to the low quality fingerprint pictuer. So, about the more larger area of the low quality fingerprint picture or the whole fingerprint picture,the algorithm can not drwan the detail characteristic effectively, In order to drwan high quality detail characteristic about the fingerprint from the macroscopic,we need to expand the deal with area in the fingerprint picture of this algorithm with the characteristic of fingerprint. And what will be the direction is that the author continue to study.