Fingerprint pre-alignment for hybrid match-on-card system

Smartcard has limited memory and processing resources. If we can perform some generic and time consuming algorithm in the host PC that captures the query fingerprint image, the time taken for matching in the smartcard could be reduced considerably. For this reason, we propose a new fingerprint pre-alignment algorithm in the host PC that obtains small amount of information from the smartcard, find the alignment point and angle and send the aligned query fingerprint template to the smartcard for matching. Our pre-alignment algorithm achieved more than 90% genuine acceptance rate on the FVC 2000 database and within small distance and angle tolerance.