ACKNOWLEDGMENTS
First of all I thanks ALLAH for guiding me and taking care of me all of the time, my life is so blessed because of ALLAH.
I would also like to take this opportunity to thank my thesis supervisor, Assoc. Prof. Dr.
Rahib Abiyev for giving me the opportunity to work with him and to provide my with guidance and help throughout my thesis and other courses.
I would wish to thank all the faculty members of the Computer Engineering Department at Near East University for their generous help and tremendous support through the course of my master program
Lastly, it is my pleasure to acknowledge and thank people who helped me accomplish my goal to pursue graduate studies. I would like to thank my lovely parents for their constant support and encouragement. They have made lots of sacrifices to help me with my education, for which I will always be grateful.
i
ABSTRACT
Fingerprint recognition plays an important role in the biometric identification of humans. Fingerprint recognition has many applications. It has been used for decades in civilian applications, in criminal investigation, and other applications.
There are different techniques that can be used to recognize fingerprints. Artificial Neural Networks (ANNs) is an efficient method for Fingerprint recognition. Neural Networks has such characteristics as: vitality, parallel computation, learning and generalization abilities, and analytic description of linear and non-linear problems. Due to these characteristics neural networks become of great importance for applications in such areas like artificial intelligence, pattern recognition, theory of control and decision making, identification, optimal control, and robotics.
In this thesis, recognition of fingerprints using neural network (NN) is considered.
Using Discrete Fourier Transform DFT the directional images of fingerprints are obtained. The DFT achieves effective low frequency filtering, reducing the noise effects in fingerprint images. Then the feed-forward neural network is applied for fingerprint recognition. The fingerprint database is constructed and used to train a neural network.
Simulation of the fingerprint recognition system is carried out using MATLAB. The neural network is used to train and identify fingerprints. The training is carried out using adaptive learning Algorithm.
ii
CONTENTS
AKNOWLEDEMENT i
ABSTRACT ii
CONTENTS iii
INTRODUCTION 1
1. OVERVIEW OF BIOMETRIC SYSTEMS 4
1.1 Overview 4
1.2 Biometric Systems 4
1.3 A Comparison of Various Biometrics 8
1.4. Advantages of Biometric Systems 14
1.5. Applications of Biometric Systems 16
1.6 Summary 17
2. FINGERPRINT RECOGNITION 18
2.1 Overview 18
2.2 Fingerprints as a Biometric
18 2.3 Architecture of Fingerprint identification System 20
2.4 Fingerprint Sensing 21
2.5 Fingerprint Representation and Feature Extraction 23
2.6 Fingerprint Matching 25
2.7 Fingerprint Classification and Indexing 27
2.8 Synthetic Fingerprints 29
2.9 Applications of Fingerprint Recognition Systems 30
2.10 Summary 32
3. THE USE OF NEURAL NETWORK FOR FINGERPRINT RECOGNITION 33
3.1 Overview 33
3.2 Neural Network Definition 33
3.3 Model of a Neuron 35
3.4 Activation Functions 35
3.5 Back-Propagation 37
3.5.1 Back-Propagation Learning 37
3.6 Learning Processes 39
3.6.1 Neural Network Learning 40
3.7 Learning Tasks 41
3.7.1 Compression Networks 42
3.7.2 Architecture of Proposed Finger Verification System 42
3.8 Summary 43
4. DEVELOPMENT OF NEURAL NETWORK BASED FINGER PRINT RECOGNITION SYSTEM
44
4.1 Overview 44
4.2 Introduction 44
4.3 Construction of Directional Images 44
4.3.1 Finger Segmentation 47
4.3.2 Computation of Directional Vectors 48
4.3.3 Construction of Directional Image 49
4.4 Fingerprint Recognition Process 51
iii
4.4.1 Grayscale Analysis 53
4.5 Clip Images 53
4.6 Neural Network Based Fingerprint Recognition 54
4.7 Architecture 55
4.8 Initialization 56
4.8.1 Training 56
4.9 System Performance 56
4.10 Summary 59
CONCLUSION 60
REFERENCES 62
APPENDIX A 64
iv