With the pioneering studies of Galton and Henry on ?ngerprint-based human id- ti?cation over a hundred years ago, biometrics recognition is now permeating many sectors of our society. Even though the origins and the most frequent use of b- metrics, primarily ?ngerprints, have been in forensics and law enforcement app- cations, biometric recognition is now needed to address a number of societal pr- lems related to security risks, terror threats and ?nancial fraud. In some situations, for example, detecting duplicate enrolments (for issuing of?cial documents such as passports and driver licenses), biometrics is the only method to con?rm that an - plicant has not been previously enrolled in the database. In addition to ?ngerprints, new anatomical and behavioral traits, viz. , face, iris, voice and hand shape have also become popular. Biometrics recognition is an excellent proving ground for pattern recognition. While pattern recognition systems have been available for document imaging, speech recognition, medical image analysis, remote sensing and variety of insp- tion tasks, no other application domain offers as many challenges as the biometrics domain.