Based On The Feature Descriptor Of The Fingerprint Algorithm

Fingerprint identification authentication is closely related to human life all the time,from the start of human civilization to the age of network and information today. It hasmany advantages, such as easier to use, higher reliability and lower cost. The targent ofauthor is based on the feature descriptor of the fingerprint algorithm for the detection ofsingular points used in embedded system.This dissertation put the emphases on the question described as follow:(1) After the detection of singular pionts, we get the number, position and type ofthese singular points. We can get the result of fingerprint classification by count thenumber of core points and calculate the relative position of core point and delta point.(2) The random local region descriptor(RLRD) of image-based features representthe orientation features in local region, and thanks to the random number, the feature iscancelable. The representation of RLRD is a fixed-length real or binary vector whichcan be used in singular point feature extraction conveinently.(3) The article introduces based on the characteristics of munutiae descriptorfingerprints to global feature matching. System makes a decision whether the twofingerprints are come from the same finger by count the number of matching minutiaepairs of the two fingerprints.