"Preface : the uniqueness of the physiological and behavioral characteristics of human beings is used to facilitate the identification or verification process, and it always results in correct classification. However, the distinctive evidences obtained from an individual do not guarantee a 100% matching to be performed to the biometric characteristics corresponding to another subject, even after considering all aspects of accurate recognition process. Tampered sensors, distorted evidences, recognition at a distance, and sometimes, motion of the target subject cause identification process to be a weak one. Moreover, picking up the right algorithms for image enhancement, feature extraction, matching, classification, and decision in a biometric system are critical. Most of the commercial, off-the-shelf biometrics systems together cannot deal with all the intrinsic perspectives that could easily make the system unstable. Therefore, to cope with this problem, we should consider standard biometrics properties and algorithmic paradigms for the identification or verification of human beings. Hence, experiments of biometric systems at large with various biometrics traits could be a good approach to identify the right and secured human recognition system. This book is a collection of biometrics solutions, which include both unimodal and multimodal biometrics. Also, this book has included up-to-date biometrics algorithms with novel feature"-- Provided by publisher.
Includes bibliographical references and index.