Research On Pressure Sensed Fingerprint Identification System And Algorithem

The requirement for accuracy, security, and practicability in identification become higher, along with the continuing development of society and economy. Traditional identification methods based on what you have or what you know have some hidden trouble, as they are prone to be lost, forgotten, to be modified, and not convenient to take with. Some kind of biometric identification methods based on biology characteristics provide more reliable and convenient schemes. Biometric identification method based on what you are, overcome those problems radically by conbining biology and information technology together. Among all biometric identification metods, fingerprint identification is recognized as its convenience, accuracy, small volume, and low cost, which is now an efficient identification metod, and is widely used electronic commerce, criminal identification, access control, and information security.Fingerprint lies on the tip of fingers, whose line information is recognizable by our eyes. It's formed in the early time as embryo, and while this processs is affected by the environment at the beginning, it's invariable during the rest of human life. The automatic fingerprint identification system (AFIS) integrates photoelectric technique, image processs and pattern recognition technique, and it mainly consists of some processs as fingerprint image collection, fingerprint image processs, minutiae extraction, and pattern template matcing. In this paper some key algorithms used in AFIS are studied based on system principle and summarizing exsited research results. Also a pratical and effective scheme is presented, which is prepare for the alogorithm implementation of an on going pressure sensor base embedded fingerprint identifyication system. The main content and research results include:1,The compare of various biometric identification technique is discussed. AFIS architecture, principle, and processs flow is summarized, and current research state is presented through analyse of various algorithm. 2,A agile system architecture is designed based on MFC application frame, in which algorithm and parameters can be implementd and changed easily. And a flow of processs algorithms is presented as programs.3,Noise decrease, line enhancement, dynamic binarization, and thinning algorithm of fingerprint image is studied. And a binarization method utilizing orientation information is proposed, aimed for increase accuracy of fingerprint line extraction.4,A method of fingerprint ridge repairment is proposed, for decreasing pseudo-minutiae extraction, and hence increase matching accuracy.5,The result system is test using FVC fingprint test library, as to chose and try different algorithms and parameters.As is show by the testing results analyse, the implemented system is applicable in accuracy as is required.