Research On Key Technologies Of Fingerprint Identification Processes

Among the numerous bio-metrics identification,fingerprint identification is most used because of its convenience and high accuracy.This technology is widely used in electronic commerce,criminal identification and information safety etc.The automatic fingerprint identification system(AFIS)generally consists of fingerprint image collection,fingerprint image preprocessing,minutiae extraction,minutiae matching and fingerprint classification.In this thesis,the development history,current situation and future development trend of fingerprint identification are researched.We give a detailed analysis of fingerprint preprocessing,including region segmentation,image normalization,computing orientation map,filtering enhancement,binaryzation,thinning and minutiae extraction.The output of each stage is given through MATLAB software.Especially we have carried on a comprehensive analysis of the fingerprint image preprocessing and minutiae extraction.On the basis of this,we propose a new fingerprint matching algorithm to improve its accuracy and robustness.Then we use a fingerprint database from the FVC2000 to test the experiment performance.The experiments show my program can surely improve the accuracy compared with two other algorithms mentioned in two recent literature.