A Fuzzy Clustering Algorithm for Fingerprint Enhancement

Fingerprint Identification and Verification are important tools for both forensics (when dealing with crime evidence) and authentication when used as a biometricidentifier, say for access purposes. Although fingerprint authentication is considered an effective method, the quality of the verification is highly dependent upon sample quality. In the case of crimes, since most people would not leave clear footprints, it is important to have an effective enhancing method that would be able to assist narrow down the matching specimen while reducing the false-positives. We use a fuzzy clustering algorithm to enhance fingerprints, and report the encouraging results found.