Studies On An Automatic Fingerprint Verification System Based On Gabor Filter

Fingerprint is the most reliable research object for its uniqueness and permanence.Automatic fingerprint verification(identification) system is applied in a wide range of application domains such as national ID card,electronic commerce,and automated banking.In this paper,we have analysed traditonal fingerprint identification alogthium and provide a new fingerprint orientation-based alogthium.The alogthium is mainly composed of two sections,which are fingerprint preprocessing and fingerprint matching. Fingerprint preprocess includes image normalizing,orientation estimation,orientation smooth,ridge extraction ,ridge removed noise,thinning and removed noise. Fingerprint matching includes feature extraction ,feature code and minutiae matching. The main purpose of image normalizing is to reduce the variations in gray level values along ridges and furrows,but not to change the clarity of the ridge and furrow structures.The orientation image estimated by Sobel gradient operator is compared for the filtered fingerprint image and minutiae matching .The Gabor filter is applied to enhance the quality of fingerprint image.Thus the system is robust to low quality of fingerprint image.Then we extract ridges from filtered fingerprint image by 7X7 mask and fill holes between the ridges.The purpose of thinning image is to reduce a lot of inessential information and to get very integrated minutiae information.At last we align the minutia of collected fingerprint image which are affected by some unfavorable factors such as tranlation ,rotation and match the fingerprints to find the similar degree,then identify whether they belong to a person's fingerprint .We have done a great deal of improvement on previouswork and put forward our own fingerprint algothium which is to some degree different from others' fingerprint verification algothiums. By comparision to traditional fingerprint algothium,our systems have been extensively on a large number of captured fingerprint images .Experimental results show that our system has a relatively high rate of fingerprint recognition and a strong ability to resist image noises.In addition, it is proved to be relatively high effective in recognizing damaged or blurry fingerprint images.The system is relatively high valuable in practice.