Image Enhancement Research Based On Automtic Fingerprint Identification System

In today's complex, geographically mobile and increasingly electronically inter-connected information society, accurate personal identification is becoming more and more important and difficult. The aim and object of this paper is to improve the equality of poor fingerprint from optical sensor, especially the over or under inking or incomplete images.This paper mainly discusses follows:1. Discuss and summarize the method, the principle, the main realizing method, of Automatic Fingerprint Identification System. A comprehensive preview is presented on fingerprint image preprocessing on the state of the core algorithm related to AFIS.2. The paper design an automatic fingerprint identification system, included test and identification module. Binarization, thinning, minutia detection module are designed using exiting method. The enhancement module is the core content. The test result has proved identification rate improved highly after image enhancement.3. The segmentation of fingerprint area from image background is one of the important module in image preprocessing. Based the position difference between finger and sensor, the huge average of four corner is used as segmentation threshold. The test result has proved that this method can segment the fingerprint image from background quickly.4. Gabor-based fingerprint image enhanced algorithm in wavelet domain was proposed based on the idea that the dry and wet can be looked as high-frequency disturbance. The using above estimated ridge orientation, each direction coefficient images were enhanced based on Gabor function and further reconstruction.