AUTOMATED BIOMETRICS Technologies and Systems

Transcription

1 AUTOMATED BIOMETRICS Technologies and Systems

2 The Kluwer International Series on ASIAN STUDIES IN COMPUTER AND INFORMATION SCIENCE Series Editor Kai-Yuan Cai Beijing University of Aeronautics and Astronautics Beijing, CHINA Editorial Advisory Board Han-Fu Chen, Institute of System Science, Chinese Academy of Sciences Jun-Liang Chen, Beijing University of Post and Telecommunication Lin Huang, Peking University Wei Li, Beijing University of Aeronautics and Astronautics Hui-Min Lin, Institute of Software Technology, Chinese Academy of Sciences Zhi-Yong Liu, Institute of Computing Technology, Chinese Academy of Sciences Ru-Qian Lu, Institute of Mathematics, Chinese Academy of Sciences Shi-Tuan Shen, Beijing University of Aeronautics and Astronautics Qing-Yun Shi, Peking University You-Xian Sun, Zhejiang University Lian-Hua Xiao, National Natural Science Foundation of China Xiao-Hu You, Southeast University Bo Zhang, Tsinghua University Da-Zhong Zheng, Tsinghua University Bing-Kun Zhou, Tsinghua University Xing-Ming Zhou, Changsha University of Technology Also in the Series: FUZZY LOGIC AND SOFT COMPUTING, by Guoqing Chen, Mingsheng Ying, Kai-Yuan Cai; ISBN: INTELLIGENT BUILDING SYSTEMS, by Albert Ting-pat So, Wai Lok Chan; ISBN: PERFORMANCE EVALUATION, PREDICTION AND VISUALIZATION OF PARALLEL SYSTEMS by Xingfu Wu; ISBN: ROBUST MODEL-BASED FAULT DIAGNOSIS FOR DYNAMIC SYSTEMS by Jie Chen and Ron J. Patton; ISBN: FUZZY LOGIC IN DATA MODELING: Semantics, Constraints, and Database Design by Guoqing Chen; ISBN: SOFTWARE DEFECT AND OPERATIONAL PROFILE MODELING by Kai-Yuan Cai; ISBN:

3 AUTOMATED BIOMETRICS Technologies and Systems by David D. Zhang Hong Kong Polytechnic University SPRINGER SCIENCE+BUSINESS MEDIA, LLC

4 Library of Congress Cataloging-in-Publication Data Zhang, David, Automated biometrics : technologies and systems / by David D. Zhang. p. cm. -- (The Kluwer international series on Asian studies in computer and information science ; 4) Includes bibliographical references and index. ISBN ISBN (ebook) DOI / Pattern recognition systems. 2. Identification--Automation. 3. Biometry. 1. Title. Il. Series. TK7882.P3 Z dc Copyright by Springer Science+Business Media New York Originally published by Kluwer Academic Publishers, New York in 2000 Softcover reprint ofthe hardcover Ist edition 2000 AU rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, mechanical, photocopying, recording, or otherwise, without the prior written permission ofthe publisher, Springer Science+Business Media, LLC. Printed an acid-free paper.

5 CONTENTS PREFACE ACKNOWLEDGEMENTS ix xi 1 INTRODUCTION TO BIOMETRICS 1.1 Why Biometrics? 1.2 What is Biometrics? 1.3 How does a Biometrics System Work? 1.4 Where can Biometrics Systems be Applied? 1.5 Book Perspective Part I BIOMETRICS TECHNOLOGIES 2 HUMAN BODY AND BIOMETRICS 2.1 Our Body: Overview 2.2 Physical Structures 2.3 Behavioral Characteristics 2.4 Ways of Behaving SIGNAL AND IMAGE PROCESSING 3.1 Digital Image Processing 3.2 Transformation Technology 3.3 Image Enhancement 3.4 Image Restoration

6 vi Automated Biometrics: Technologies and Systems 3.5 Data Compression PATTERN RECOGNITION 4.1 Pattern, Pattern Recognition and System 4.2 Image Segmentation 4.3 Feature Selection 4.4 Pattern Classification 4.5 Neural Networks Part II PHYSICAL BIOMETRICS 5 FINGERPRINT SYSTEM 5.1 Review of Fingerprint Systems 5.2 Definitions and Notations 5.3 Fingerprint Image Processing 5.4 Minutiae Determination 5.5 Fingerprint Matching PALMPRINT VERIFICATION 6.1 A New Type of Biometrics: Palmprint 6.2 Datum Point Determination 6.3 Two Typical Features in Palmprint 6.4 Palmprint Classification 6.5 Experimental Results 111 III FACE RECOGNITION 7.1 Introduction to Face Recognition System 7.2 Detection and Location of Faces 7.3 Features Extraction and Face Recognition

7 Contents vii 7.4 An Dual Eigenspaces Method for Face Recognition What Should We Do Next? IRIS BIOMETRICS Introduction Iris Recognition Coordinate System Texture Energy Feature Experimental Results Part III BEBA VIORAL BIOMETRICS 9 SPEAKER RECOGNITION Introduction Principles of Speaker Recognition GSMSV Method Experimental Results SIGNATURE SYSTEM Signature Verification Off-line Signature On-line Signature A Signature Verification System InternetlIntranet Signature Verification Application OTHER BEHAVIORAL BIOMETRICS Behavioral Biometrics System Keystroke Biometrics Gesture Recognition Gait Biometrics 234

8 viii Automated Biometrics: Technologies and Systems 239 Part IV BIOMETRICS APPLICATIONS 12 BIOMETRICS USER INTERFACES 12.1 Biometrics Interfaces 12.2 Human-Side Interface 12.3 Machine-Side Interface 12.4 Case Study: PalmScanner Interface PERSONAL AUTHENTICATION 13.1 Examples of Current Applications 13.2 Potential Application Areas 13.3 How to Select a Biometrics System 13.4 Application Programming Interface Standards 13.5 Information Resources BIOMETRICS CHINESE MEDICINE 14.1 Traditional Chinese Medicine (TCM) 14.2 Four Chinese Diagnostic Methods 14.3 Various Observation Types in TCM 14.4 Tongue Diagnosis FUTURE WORK IN BIOMETRICS 15.1 Biometrics in Internet-Ecommerce 15.2 Key Technologies in Biometrics 15.3 Integrated Biometrics 15.4 VLSI Biometrics INDEX 327

9 PREFACE In the modern, automated world, there is an ever-growing need to authenticate and identify individuals. The current technology of using a personal identification numbers (PIN) or password for these purposes hardly meets the requirements of an identification system because one has to remember too many passwords and the password or PIN is very insecure. As a result, in today's complex, geographically mobile, increasingly electronically wired information society, the problem of identifying a person continues to pose a great challenge. Biometrics-based authentication and identification are emerging as the most reliable method. Biometrics requires that the person to be identified be physically present at the point-of-identification and relies on "something which you are or you do" to provide better security, increased efficiency, and improved accuracy. It overcomes some of the limitations of the traditional automatic personal identification technologies such as ID cards and PIN: ID cards may be lost, stolen, forgotten, or misplaced whereas PIN may be forgotten or guessed by the impostors. In addition, the traditional identification methods are unable to differentiate between an authorized person and an imposter who fraudulently acquires the "knowledge" or "token" of the authorized person. Automated biometrics deals with physiological or behavioral characteristics such as fingerprints, signature, palmprint, iris, hand, voice and face that can be used to authenticate a person's identity or establish an identity from a database. With rapid progress in electronic and Internet commerce, there is also a growing need to authenticate identity of a person for secure transaction processing. Designing an automated biometrics system to handle large population identification, accuracy and reliability of authentication are challenging tasks. Currently, there are over ten different biometrics systems that are either widely used or under development. Some automated biometrics, such as fingerprint identification and speaker verification, have received considerable attention over the past 25 years, and some issues like face recognition and iris-based authentication have been studied extensively resulting in successful development of biometrics systems in commercial applications. However, according to the author's best knowledge, so far, very few books have been found exclusively devoted to such issues of automated biometrics. As such a book about automated biometrics, it will systematically introduce the relative technologies and systems, and explore how. to design the corresponding systems with in-depth discussion. But, this is not meant to suggest a low-relevance of the book to biometrics in general. Rather, the issues of the book addresses are highly relevant to many fundamental concerns of both researchers and practitioners of automated biometrics in security/detective. The materials in the book are the outgrowth of research the author has conducted for many years, and present the

10 x Preface author's recent academic achievements made in the field, although, for the sake of completeness, related work of other authors will also be addressed. This book is organized into four parts. As an introduction, Chapter 1 first describes the basic concepts necessary for a good understanding of biometrics and answers some questions about biometrics like why, how, what and where. Then, Part I focuses on fundamental biometrics technologies. In Chapter 2, some basic definitions and notations of biometrics from human body are given. Image processing and pattern recognition technologies related to biometrics are discussed in Chapters 3 and 4, respectively. Part II explores physical biometrics recognition systems. Chapter 5 describes how fingerprint system can be built. In Chapter 6, we develop a biometrics system by palmprint. Chapter 7 defines face system with facial feature extraction, face database and matching techniques and reviews various face recognition approaches. Chapter 8 introduces iris recognition based on coordination system to solve head tilting problem and texture energy. Part III describes how some behavioral biometrics systems are designed. Chapter 9 presents voice system that deals with speech acquisition, feature extraction, and speaker verification. As another main behavioral biometrics, signature system is discussed in Chapter 10. In addition, Chapter II summarizes the other three behavioral biometrics systems, including keystroke biometrics, gesture recognition, and gait biometrics. Part IV deals with biometrics applications. As an important stage, biometrics user interface is first discussed in Chapter 12. Chapter 13 describes personal authentication application. In Chapter 14, we apply biometrics to Traditional Chinese Medicine (TCM) and such a biometrics Chinese medicine system is then developed. Finally, some main research directions in biometrics are discussed in Chapter 15. This book is not a primer in biometrics, in that a certain amount of prior knowledge, such as pattern recognition and image processing, is assumed. It is my hope that this book will contribute to our understanding of this new and exciting discipline: Automated Biometrics: Technologies and Systems. David D. Zhang Hong Kong Polytechnic University

11 ACKNOWLEDGEMENTS My sincere thank goes to Professor Zhaoqi Bian at Tsinghua University, Beijing, China, for his advices throughout this research. I would like to thank Professors Kuanquan Wang, Jane You and Jeffery Liu for their useful comments and supports on special biometrics applications. I would also like to express my gratitude to my research associates and assistants, Dr. Wei Shu, Dr. Changshui Zhang, Dr. Yiying Zhang, Dr. Jie Zhou, Dr. Y ong Fang and Mr. Lei Huang for their useful helping and supporting. My graduate students are excellent companions on a difficult road towards under&tanding Biometrics. I am very grateful to Wenxin Li, Adams Kong, Pang Bo, Xiangqian Wu, Yanlai Li, Shen Xu, George Bao and Cheng Cao, who have contributed to this book. Also thank is due to Mr. Carlos Chan for his help in the preparation of this book. The financial support from UGC/CRC Grants in Hong Kong is appreciated.

12 SERIES EDITOR'S ACKNOWLEDGMENTS I am pleased to acknowledge the assistance to the editorial work by Beijing University of Aeronautics and Astronautics and the National Natural Science Foundation of China Kai- Yuan Cai Series Editor Department of Automatic Control Beijing University of Aeronautics and Astronautics Beijing China