Contactless 3D Fingerprint Identification.pdf-资料库

weixin_38290023-11274798-4744300845395866904.pdf-第1页.png

第1页 / 共129页

weixin_38290023-11274798-4744300845395866904.pdf-第2页.png

第2页 / 共129页

weixin_38290023-11274798-4744300845395866904.pdf-第3页.png

第3页 / 共129页

weixin_38290023-11274798-4744300845395866904.pdf-第4页.png

第4页 / 共129页

weixin_38290023-11274798-4744300845395866904.pdf-第5页.png

第5页 / 共129页

weixin_38290023-11274798-4744300845395866904.pdf-第6页.png

第6页 / 共129页

weixin_38290023-11274798-4744300845395866904.pdf-第7页.png

第7页 / 共129页

weixin_38290023-11274798-4744300845395866904.pdf-第8页.png

第8页 / 共129页

Advances in Computer Vision and Pattern Recognition Ajay Kumar Contactless 3D Fingerprint Identification

Advances in Computer Vision and Pattern Recognition Founding editor Sameer Singh, Rail Vision, Castle Donington, UK Series editor Sing Bing Kang, Microsoft Research, Redmond, WA, USA Advisory Board Horst Bischof, Graz University of Technology, Austria Richard Bowden, University of Surrey, Guildford, UK Sven Dickinson, University of Toronto, ON, Canada Jiaya Jia, The Chinese University of Hong Kong, Hong Kong Kyoung Mu Lee, Seoul National University, South Korea Yoichi Sato, The University of Tokyo, Japan Bernt Schiele, Max Planck Institute for Computer Science, Saarbrücken, Germany Stan Sclaroff, Boston University, MA, USA

More information about this series at http://www.springer.com/series/4205

Ajay Kumar Contactless 3D Fingerprint Identiﬁcation 123

Ajay Kumar The Hong Kong Polytechnic University Kowloon, Hong Kong ISSN 2191-6586 Advances in Computer Vision and Pattern Recognition ISBN 978-3-319-67680-7 https://doi.org/10.1007/978-3-319-67681-4 ISSN 2191-6594 (electronic) ISBN 978-3-319-67681-4 (eBook) Library of Congress Control Number: 2018956591 © Springer Nature Switzerland AG 2018 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, speciﬁcally the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microﬁlms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, in this publication does not imply, even in the absence of a speciﬁc statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional afﬁliations. trademarks, service marks, etc. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface Person identiﬁcation using epidermal ridge impressions from ﬁngers has been widely studied for over hundred years. It is widely employed in a range of forensic, e-business and e-governance applications around the world. Traditional acquisition of ﬁngerprint images by rolling or pressing of ﬁngers against hard surface like glass or polymer often results in degraded images due to skin deformations, slippages, smearing or residue of latent from previous impressions. As a result, full potential from the ﬁngerprint biometric cannot be realized. Contactless 2D ﬁngerprint sys- tems have emerged to provide improved hygiene and ideal solutions to above intrinsic problems. Contactless 3D ﬁngerprints can potentially provide signiﬁcantly more accurate personal identiﬁcation, as rich information is available from con- tactless 3D ﬁngerprint images. Contactless 3D ﬁngerprints offer exciting opportunities to improve the user convenience, hygiene and the matching accuracy over the ﬁngerprint biometric technologies available today. Introduction of videos, or addition of an additional temporal dimension, was a leap forward that revolutionized the usage of 2D images in the entertainment, e-governance and e-business. Similarly, the addition of one more dimension from 3D ﬁngerprints, has potential to signiﬁcantly alter the way this biometric is perceived and employed for the civilian and e-governance appli- cations. Such advancements will not be limited in e-security or e-business, but also enable dramatic advancements in forensics where the latent or lifted ﬁngerprint impressions are matched with suspects ﬁngerprint images. For example, the 3D ﬁngerprints from possible suspects can be employed to simulate latent ﬁngerprint impressions on a variety of hard or soft real-life materials (door, paper, glass, gun, etc.) and under variety of pressure, occlusions and deformations, which is expected to enable more accurate match with the corresponding latent ﬁngerprints that are lifted from the crime scene. The potential of contactless 3D ﬁngerprints offers exciting opportunities but requires signiﬁcant research and development efforts for its realization. Availability of a book that is exclusively devoted to the techniques, comparisons and promises from the contactless 3D ﬁngerprint identiﬁcation is expected to help in advancing much needed further research in this area. Some of the contents in this v

vi Preface book have appeared in some of our research publications and US patents. However, many of the important details, explanation and results that have been missed in the publications are included in this book. The contents in this book attempt to provide a systematic introduction to the 3D ﬁngerprint including most updated advancements in contactless 2D and 3D sensing technologies, and expla- nation of every important aspect towards the development of an effective 3D ﬁn- gerprint identiﬁcation system. identiﬁcation, This book is organized into eight different chapters. Chapter 1 introduces current trends in the acquisition and identiﬁcation of ﬁngerprint images. This introductory chapter discusses the nature of ﬁngerprint impressions and the sensing techniques, which includes completely contactless 2D ﬁngerprint sensors. This chapter bridges the journey from rolled and inked ﬁngerprint impressions, to the more advanced smartphone-based ﬁngerprint sensors, in terms of their resolution and sensing area. It also provides details on publicly accessible implementations on ﬁngerprint matchers and most updated list/details on publicly available ﬁngerprint databases along with respective weblinks to enable easy accessibility. Chapter 2 in this book presents a range of 3D ﬁngerprint imaging techniques along with their comparative technical details. Image acquisition methods presented in this chapter have been grouped into four categories: optical, non-optical, geo- metric and photometric methods. Details on ﬁve different methods to acquire 3D ﬁngerprint images using stereo vision, pattern lighting, optical coherence tomog- raphy, ultrasound imaging and photometric stereo, along with potential from other methods, appear in this chapter. Chapter 3 in this book is devoted to in-depth details on a low-cost and effective method for the online 3D ﬁngerprint image acquisition. Systematic details on such photometric stereo-based setup are detailed in this chapter, i.e. from hardware, calibration, preprocessing and specular reﬂection removal, to the choice of recon- struction methods. This chapter also shares our insights and results on the attempts made to consider non-Lambertian nature of ﬁnger surface. Resulting computational complexity for such online 3D ﬁngerprint imaging system also appears in this chapter. Chapter 4 provides details on more efﬁcient 3D ﬁngerprint imaging approach using coloured photometric stereo. This approach is introduced to address two key problems associated with practical 3D ﬁngerprint involuntary ﬁnger motion and complexity for online applications. This approach revisits the method detailed in Chap. 3 and contactless 3D ﬁngerprint images acquired using the setup introduced in this chapter are also publicly made available. imaging: Contactless 3D ﬁngerprint data often requires preprocessing operations to sup- press the accompanying noise and to enhance or accentuate the ridge–valley fea- tures. Chapter 5 in this book details on such preprocessing operations on the cloud point 3D ﬁngerprint data. This chapter also provides detailed explanation on spe- cialized enhanced operations required for the contactless 2D ﬁngerprint images that are employed for the reconstruction of 3D ﬁngerprints.

Preface vii Chapter 6 systematically introduces representation of minutiae in 3D space and provides details on recovering these features from cloud point data. Therefore, the techniques discussed in this chapter are generalized and quite independent of method used for the 3D imaging. With the help of many illustrations, most from real 3D ﬁngerprint data, this chapter systematically details alignment and relative representation of 3D minutiae in order to generate numerical match score between two arbitrary 3D ﬁngerprint minutiae templates. This chapter also details a minutiae selection algorithm and in-depth study on the variation of ﬁve-tuple relative 3D minutiae components with distance, which resulted in the introduction of a uniﬁed matching distance. Detailed experimental results presented in this chapter underline the effectiveness of our 3D minutiae template-based approach. Contactless 3D ﬁngerprints can be matched using a range of methods than those detailed in Chap. 6. Therefore, Chap. 7 details on such efﬁcient 3D ﬁngerprint matching methods, binary surface code-based approach and its variants, along with tetrahedron-based matching approach. Methods detailed in this chapter offer computationally efﬁcient alternatives that can justify their usage in a range of e-business or civilian applications. Chapter 8 in this book is devoted for the study on the uniqueness of 3D ﬁn- gerprints. This chapter scientiﬁcally deﬁnes the individuality of 3D ﬁngerprints and comparatively evaluates its improvement, over the 2D ﬁngerprints, using practical 3D minutiae template matching criterion and imaging resolutions. The numbers illustrated in this chapter provide upper bound on the expected performance from the contactless 3D ﬁngerprint systems. I wish to thank many student and staff members in The Hong Kong Polytechnic University, who have directly or indirectly supported in the completion of this book. Cyril Kwong and Chenhao Lin deserve special thanks here as they have been instrumental in advancing many of the research outcome reported in this book. Cyril has worked with me for several months as research assistant, while Chenhao has been working towards his doctoral degree research. Kowloon, Hong Kong June 2018 Ajay Kumar

资料库

Contactless 3D Fingerprint Identification.pdf

相关推荐

音视频

热门标签

最新资料