The use of biometric in recognizing individuals is now a phenomenon in today?s world. As there is security need within wide array of applications including transactions security across computer networks, PIN (personal identification number) replacement in retail business and banking, high-security wireless access, televoting as well as restricted areas? admission, biometric is more and more a subject of interest. The traditional mode of personal identification verification is grounded on possession (ID card) or knowledge (secret code) but ID cards could be stolen or lost while codes could be forgotten. These problems could allow opportunity to criminals to steal someone?s identity. As such, using features that are uniquely belonging to a person would decrease fraud possibility. Also, in many situations, the use of biometry is convenience. This is because biometry can take over the use of codes, keys and cards. Also, fingerprint is regarded as biometric that is most practical. With regard to performance, applications of pattern recognition mostly rely on extraction of feature and classification or learning scheme, the extraction and selection of meaningful features. This generates major problem in identification of fingerprint; how to attain features in numerous fingerprint styles so that the right individual could be reflected. Geometric Moments is proposed by this paper for feature extraction with identification of fingerprints and analyze its performance.
展开▼
