The Research Of Fingerprint Image Preprocessing Algorithm

Biometrics, automatic identification an individual by using certain physiological or behavioral characteristics associated with the person, has gained wide acceptance recently because of the unbridled violence of terrorism and hacker. Fingerprint recognition is a hot spot and the most widely used branch of biometric with many advantages, such as easier to use, higher reliability and lower cost.The fingerprint image preprocessing system is mainly studied the parts below: (1) Fingerprint image segmentation;(2) Fingerprint image enhancement;(3) Fingerprint image thinning;(4) Post processing of thinned fingerprint image.The method of variances was used in fingerprint image segmentation, and the method of variances was improved. First of all, filter fingerprint images with the edge-keep filter, Dilution the strong noise in fingerprint image and then normalize the image, make the grey of the image tend to a unified scope. Finally, segment the image with the variance method.Fingerprint image enhancement mainly uses the method of direction filter. Propose an image-filtering algorithm that is able to choose image block size and corresponding filter intelligently. This algorithm overcomes the problems as follows: Image block size is not suitable;Direction information is not correct;can not connect the breakages;Image is not smooth after filter. The algorithm tested by a series of experiments, is proven to be effective.Fingerprint image thinning, erase the non-skeleton point according to the follow rules: (1) Internal points can not be deleted;(2) Single point can not be deleted;(3) End point of the line can not be deleted;(4) If P is a border point and when point P is deleted, the component is not increase, and then point P can be deleted. Design a table according to the rules above. From 0 to 255, there are 256 elements, each element is 0 or 1. Find the table according the 8-connected boundary of one point, if the element is 1, delete the point otherwise maintain the point.The fingerprint image preprocessing system, tested by a series of experiments, is proven to be effective.