Biometrics: Proving Ground for Image and Pattern Recognition

Abstract

The emerging requirements of reliable and highly accurate personal identification in a number of government and commercial applications (e.g., international border crossings, access to buildings, laptops and mobile phones) have served as an impetus for a tremendous growth in biometric recognition technology. Biometrics refers to the automatic recognition of an individual by using anatomical or behavioral traits associated with that person. By using biometrics, it is possible to recognize a person based on who you are, rather than by what you possess (e.g., an ID card) or what you remember (e.g., a password). Most of the authentication systems based on passwords (knowledge-based security) and ID cards (token-based security) can be easily breached when a password is divulged to an unauthorized user or a card is stolen by an impostor. Biometric systems have addressed the security problems that plague traditional verification systems because they make use of a person's fingerprint, hand shape, iris, face and voice that are supposed to be unique to that person. Besides bolstering security, biometric systems also enhance user convenience by alleviating the need to design and remember multiple complex passwords. No wonder large scale systems have been deployed in such diverse applications as US-VISIT and entry to Disney Park, Orlando. The revenues for the global biometric recognition market are projected to grow from about US $2.1 billion in 2006 to US $5.7 billion in 2010. In spite of the fact that automatic biometric recognition systems based on fingerprints (called AFIS) have been used by law enforcement agencies worldwide for over 40 years, biometric recognition continues to remain a very difficult pattern recognition problem. A biometric system has to contend with problems related to noisy images (failure to enroll), lack of distinctiveness (finite error rate), large intra-class variations (false reject), and spoof attacks (system security). This talk will present an overview of biometrics, its advantages and limitations, state-of-the-art error rates and current research in sensor design, representation, fusion and security issues.