Segmentation and Separation of Overlapped Latent Fingerprints (eBook)

Algorithms, Techniques, and Datasets

Branka Stojanović, Oge Marques, Aleksandar Nešković (Autoren)

eBook Download: PDF

2019
X, 51 Seiten
Springer International Publishing (Verlag)
978-3-030-23364-8 (ISBN)

Infos

This Springerbrief presents an overview of problems and technologies behind segmentation and separation of overlapped latent fingerprints, which are two fundamental steps in the context of fingerprint matching systems. It addresses five main aspects: (1) the need for overlapped latent fingerprint segmentation and separation in the context of fingerprint verification systems; (2) the different datasets available for research on overlapped latent fingerprints; (3) selected algorithms and techniques for segmentation of overlapped latent fingerprints; (4) selected algorithms and techniques for separation of overlapped latent fingerprints; and (5) the use of deep learning techniques for segmentation and separation of overlapped latent fingerprints.

By offering a structured overview of the most important approaches currently available, putting them in perspective, and suggesting numerous resources for further exploration, this book gives its readers a clear path for learning new topics and engaging in related research. Written from a technical perspective, and yet using language and terminology accessible to non-experts, it describes the technologies, introduces relevant datasets, highlights the most important research results in each area, and outlines the most challenging open research questions.

This Springerbrief targets researchers, professionals and advanced-level students studying and working in computer science, who are interested in the field of fingerprint matching and biometrics. Readers who want to deepen their understanding of specific topics will find more than one hundred references to additional sources of related information.

Inhalt

Preface 7
Contents 9
1 Latent Fingerprints Matching Systems 11
1.1 Introduction 11
1.2 Latent Fingerprints 12
1.3 Manual Latent Fingerprints Matching: The ACE-V Method 12
1.4 Automated Latent Fingerprints Matching System 13
1.5 Overlapped Fingerprints 14
1.6 Concluding Remarks 16
References 17
2 Latent Fingerprint Datasets 19
2.1 Introduction 19
2.2 Latent Fingerprint Datasets 20
2.2.1 Publicly Available Latent Fingerprint Datasets 20
2.2.2 Specialized Overlapped Latent Fingerprint Datasets 21
2.3 The Vlatacom Dataset 22
2.3.1 Motivation 22
2.3.2 Requirements 23
2.3.3 Dataset Structure and Organization 23
2.3.4 Additional Aspects 26
2.3.5 Experiments and Results 27
2.4 Concluding Remarks 29
References 30
3 Overlapped Latent Fingerprints Segmentation: Problem Definition 31
3.1 Introduction 31
3.2 The Mask Segmentation Problem 32
3.3 Evaluation Methodology 35
3.4 Concluding Remarks 37
References 38
4 Machine Learning Based Segmentation of Overlapped Latent Fingerprints 39
4.1 Introduction 39
4.2 Description of the Algorithm 39
4.2.1 CNN Architecture 40
4.2.2 CNN Training 41
4.3 Experimental Evaluation 42
4.4 Concluding Remarks 44
References 44
5 Overlapped Latent Fingerprints Separation: Problem Definition 45
5.1 Introduction 45
5.2 Orientation Field Estimation 46
5.2.1 Orientation Field Estimation Algorithms for Single Fingerprint 47
5.2.2 Orientation Field Estimation Algorithms for Overlapped Fingerprints 48
5.3 Overlapped Latent Fingerprint Separation: Representative Approaches 49
5.4 Evaluation Methodology 51
5.5 Concluding Remarks 52
References 52
6 Machine Learning Based Separation of Overlapped Latent Fingerprints 55
6.1 Introduction 55
6.2 Segmentation 56
6.3 Initial Orientation Field Estimation and Enhancement 57
6.3.1 Initial Orientation Field Estimation 57
6.3.2 Single Region Orientation Field Enhancement 57
6.4 Orientation Separation 58
6.5 Neural Network: Architecture and Training 58
6.6 Experimental Evaluation 59
6.7 Concluding Remarks 60
References 61

Produkt-Details

• Erscheint lt. Verlag: 22.10.2019
• Reihe/Serie: SpringerBriefs in Computer Science
• Reihe/Serie: SpringerBriefs in Computer Science
• Zusatzinfo: X, 51 p. 32 illus., 21 illus. in color.
• Sprache: englisch
• Themenwelt: Informatik ► Theorie / Studium ► Künstliche Intelligenz / Robotik
• Schlagworte: biometrics • convolutional neural networks • Deep learning • Fingerprint datasets • fingerprint verification • Forensics • Latent fingerprints • machine learning • Neural networks • Overlapped fingerprint segmentation • Overlapped fingerprint separation
• ISBN-10: 3-030-23364-2 / 3030233642
• ISBN-13: 978-3-030-23364-8 / 9783030233648