Research On Fingerprint Recognition Applied To Attendance Check System

Biometric technology is the use of human biological characteristics for their own individual identification technology. Comparing with the facial recognition, retina recognition and other biometric identification technologies, the automated fingerprint identification technology is a better and more practical identification technology. Due to their uniqueness and stability, fingerprints are used as the principal basis of personal identity by people. Automated fingerprint identification system, a technology of fingerprint identification using computer, is convenience, high efficiency, security and reliability. It has been applied in many fields such as financial security, data encryption, electronical business and so on, and will play the more and more vital role in our production and the life. This paper discusses the fingerprint identification process and the algorithm involved in and introduced the problems of the Automated Fingerprint Identification System faced and need to solved. The pretreatment of gray fingerprint image is disscussed in details.The paper compared three thin methods included of the Hilditch thin algorithm, the rapid thinning and template matching thin method About the problem that fingerprint image's thoroughly detailed and complete repair technique are difficult and complex, and the time consumption is too more, discussed the algorithm of the extracting of minutia of fingerprint based on the image which not completed thin. Focus on the fingerprint matching algorithm. The overall feature points of fingerprint to the matching process is introduced and the problem is solved that the complicated time of matching method which based on the geometric characteristics of point pattern is too higher. This method is that find reference point for the calibration of template image and the to be recognized fingerprint image through the calculation of the overall features and curve fitting techniques, rotate and parallel transfer the template image and the to be recognized image, then process the feature points's pattern matching. Improve the algorithm based on the curve fitting. The experiments indicate that the improved algorithm of matching can correctly execute the matching function and have the fixity to the rotation and movement Though the precision falls a little, the speed of execution have an evidently increase. Experimental results show that this method can quickly process fingerprint matching. The fingerprint recosssssgnition system is implement and testing plan is designed. The experiment results show that system and classify and identify the better image.