FINGERPRINT RECOGNITION USING FEATURE EXTRACTION IZNI ...

FINGERPRINT RECOGNITION USING FEATURE EXTRACTION

IZNI SYARINAZ BINTI ISMAYUDDIN

A thesis submitted in fulfillment

of the requirements for the award of the degree of

Bachelor of Computer Sciene (Software Engineering)

Faculty of Computer System & Software Engineering

University College of Engineering & Technology Malaysia

PERPUTAKMN KOLEJ UWIVERS!Ti

KEJURUTERAAN & TEK!4OLOG !AAYSA No ProIehflT fl

O2I223 TA Tatft 15S

ri

3 1J ANoo( rs TL

4) -

NOVEMBER, 2006

ABSTRACT

Lately, fingerprint recognition usage among users is to make sure the safety

of security level or pin code user is very encourage. There are so many applications

that using fingerprint recognition such as fingerprint recognition for password and it

is also used in order to recognize individual identity card. However, the fingerprint

among individual is so unfamiliar and not for common known. Throughout this

project, 'Fingerprint Recognition using Feature Extraction' can recognize the types

of fingerprint for every individual and the unique is, with every characteristics of the

difference fingerprint among individual, it can be read in computer language by

using the method that was shown in image processing. The reason why this system

was developed is to analyze the method that always been used to recognize

fingerprint. Besides that, fingerprint recognition from analog image into digital

image can be carry out by using feature extraction method that is can recognize the

characteristic in fingerprint image.

vi

ABSTRAK

Kebelakangan mi, penggunaan cap jar adalah digalakkan bagi memastikan

tahap keselamatan atau kod rahsia pengguna terpelihara. Pelbagai aplikasi yang

menggunakan cap jar seperti penggunaan cap jar sebagai kata laluan dan selalu

digunakan untuk mengenal pasti kad pengenalan setiap individu. Namun, jenis cap

jar bagi setiap individu jarang di ketahui umum. Melalui project 'Fingerprint

Recognition using Feature Extraction' dapat mengenal pasti jenis cap jar yang di

miliki bagi setiap individu dan keunikannya adalah dengan ciri-ciri cap jar yang

berbeza-beza bagi setiap individu berupaya diterjemahkan dalam bahasa komputer

menggunakan kaedah-kaedah yg terdapat dalam pemprosesan imej. Tujuan system

mi di bangunkan ialah untuk menganalisa kaedah-kaedah yang sering digunakan bagi

pengecamancapjari. Selain itu, pengecaman bentuk cap jar dari gambar analog ke

gambar digital dilaksanakan mengunakan kaedah pengekstrakan ciri-ciri yang mana

dapat mengenal pasti karakter dalam bentuk gambar cap jar.

vi'

TABLE OF CONTENT

CHAPTER TITLE PAGE

STUDENT DECLARATION

SUPERVISOR DECLARATION

DEDICATION iv

ACKNOWLEDGEMENT v

ABSTRACT vi

ABSTRAK vii

TABLE OF CONTENT viii

LIST OF FIGURES xi

LIST OF TABLES xiii

LIST OF APPENDICES xiv

ABBREVIATIONS xv

INTRODUCTION 1

1.1 Introduction 1

1.2 Problem Statement 2

1.3 Objective of System 2

1.4 Scope of System 3

2 LITERATURE REVIEW 4

2.1 Introduction 4

2. 1.1 History of Fingerprint 4

2.1.2 Type of Fingerprint 6

2.2 Fingerprint recognition 9

2.2.1 Image Acquisition 9

2.2.2 Pre-processing Image 10

vu'

lx

2.2.3 Image Threshold

10

2.2.4 Image Segmentation

10

2.2.5 Feature Extraction

11

2.2.6 Template Matching

11

2.3 Current System Using Feature Extraction for

11

Fingerprints Recognition

2.3.1 Feature Extraction Using Chain Code

11

Contours

2.3.2 Fingerprint Identification Based on The

13

Minutiae Located in Fingerprint

3 METHODOLOGY

18

3.1 Introduction

18

3.2 Waterfall Methodology

19

3.2.1 Analysis Phase

20

3.2.2 Design Phase

20

3.2.2.1 Database Process Phase

21

3.2.2.2 Image Acquisition Phase

21

3.2.2.3 Pre-Processing Phase

22

3.2.2.4 Matching Phase

28

3.2.2.5 Identification Phase

29

3.2.3 Implementation Phase

30

3.2.3.1 Development and Deployment

30

3.2.3.2 Hardware and Software

30

Requirements

3.2.4 Testing Phase

31

32

32

32

33

34

4 RESULT AND DISCUSSION

4.1 Introduction

4.2 Database Process

4.3 Image Acquisition

4.4 Pre-Processing

5

4.5 Result of Matching 39

4.6 Result of Identification 39

4.7 Result of Fingerprint Recognition System Using 42

Feature Extraction

4.8 Constraint of System 44

4.9 Discussion 44

4.10 Assumption and Further Research 45

4.10.1 Assumption 45

4.10.2 Further Research 45

CONCLUSION 46

REFERENCES 48

APPENDICES A 49

APPENDICES B 51

APPENDICES C 53

x

LIST OF FIGURES

NO TITLE PAGE

2.1 Type of Fingerprint 7

2.2 Global Levels 7

2.3 Local Levels 8

2.4 Very-fine Levels 9

2.5 Binarization Approach 12

2.6 Different Minutiae Types 13

2.7 Fingerprint Recognition Systems 13

2.8 Training Set 16

2.9 Core Points on Different Fingerprint Patterns 16

3.1 Waterfall Methodology 19

3.2 The Fingerprint Recognition Process 20

3.3 Function Get Fingerprint Image from File Saved 21

3.4 Function Get Fingerprint Image without Pick from File 22

Saved

3.5 Function Convert Color Image to Grayscale Image 23

3.6 Function of Filtering Using Predefined Filtering 24

3.7 Average Threshold Algorithms 25

3.8 Function of Threshold Using Average Technique 26

3.9 Feature Extraction Algorithms 27

3.10 Functions of Feature Extraction Using Smaller 27

Windows and By Row and Column

3.11 Euclidean Distance Algorithm 28

3.12 Functions for Template Matching Using Euclidean 28

Distance

3.13 Function for Identification the Type of Image 29

xi

4.1 Database Images 32

4.2 Interface for Image Capture 33

4.3 Image File Folders 33

4.4 Result for the Grayscale Image 34

4.5 Result for Filtering Images 36

4.6 Result for Threshold Images 37

4.7 Feature Extraction Result of the Input Image 38

4.8 Basic Concept of Euclidean Distance 40

4.9 Example Message Box Shows The Type of Fingerprint 40

4.10 Bar Chart That Show Percentages Recognize Images 42

and Unrecognized Images

4.11 The Fingerprint Recognition of Type 'Tended Arch' 43

4.12 Unrecognized the Fingerprint Recognition 43

xli

xl"

LIST OF TABLES

NO

TITLE PAGE

2.1 The History of Fingerprint

5

3.1 Minimum Hardware Requirement 30

3.2 Software Requirement

31

4.1 Example Result on Value Using Euclidean Distance 39

Technique between Database Image and Test Image

4.2 Result on Testing 20 Samples of Images 41

xlv

LIST OF APPENDICES

APPENDIX TITLE PAGE

A The Gantt Chart

WE

B Images of Fingerprint

52

C User Manual 53

ABBREVIATIONS

JPEG - Joint Photographic Experts Group

MATLAB - Matrix laboratory

SDLC - System Development Life Cycle

KUKTEM - Kolej Universiti Kejuruteraan & Teknologi Malaysia

TM - Template Matching

RGB - Red Green Blue

STFT - Short Time Fourier Transform

xv

CHAPTER 1

INTRODUCTION

1.1 Introduction

Every human has their own fingerprints and from the research, there are

several types of fingerprints, which are tended arch, arch, right loop, left loop and

whorl. That is why fingerprints are unique because the texture for every human are

different but it can be recognize or classified for each types.

Among all of the biometric techniques, fingerprint-based identification is the

oldest method, which has been successfully used in numerous applications.

Everybody was known to have unique and immutable fingerprints. So that, it can

recognize and differentiate fingerprint one (1) to another. A fingerprint is made of a

series of ridges and furrows on the surface of the finger. The uniqueness of a

fingerprint can be determined by the pattern of ridges and furrows as well as the

minutiae points. Minutiae points are local ridge characteristics that occur at either a

ridge bifurcation or a ridge ending.

This project proposed for Undergraduate project that is "Fingerprint

recognition system using Feature Extraction process". The function of this system is

to detect or match the suitable image when the image of fingerprint is requested.

Therefore, this system can recognize the type of fingerprint of the users when the

image of the fingerprint is the same.

2

1.2 Problem Statement

Every human have the various types of fingerprint. However, each person

has a unique type of fingerprint. Even though the types of fingerprint are the same

but the minutiae of fingerprint is always different for each human being. What is the

importance or advantages of knowing the types of fingerprint?

The importance of knowing the types of fingerprint is to identify the identity

of individual. For example in finding and identify the fingerprint of criminal case, it

is easier if the types of fingerprint being grouped into its own group of fingerprint

which are basically in five (5) main group that consists of arch, tended arch, left

loop, right loop and whorl. Therefore, it can reduce time for searching and identify

the data of the criminal. Besides that, this is a new approach in image processing

application.

By then, after analyze several factors that related to the current problem, the

suitable solution is by using fingerprint recognition system using Feature Extraction.

Therefore, it is easier, safety, and reliability compare to the system that being used

for the time being.

1.3 Objectives of System

The objectives of this project are:

i) To analyze techniques used fingerprint recognition.

To develop prototype of the fingerprint recognition system using

feature extraction.

1.4 Scope of System

The scopes of this project consist of:

i) The format of image is using JPEG format.

Size of image is using 240 x 240 pixels.

The image output is in grayscale image (black and white color).

iv) The prototype system will be developing using MATLAB 7.0.

V) Twenty (20) pieces sample of image are used.

CHAPTER 2

LITERATURE REVIEW

2.1 Introduction

Fingerprint is one of the most widely used biometrics. The advent of several

inkless fingerprint scanning technologies coupled with the exponential increase in

processor performance. It has taken fingerprint recognition beyond criminal

identification applications to several civilian applications such as access control, time

and attendance, and computer user login.

Fingerprints are widely believed to be unique. Basically, fingerprints do not

change significantly with age. If a finger is damage, it normally heals in such a way

that the fingerprint restored. Because of these characteristics, a person's fingerprint

can be used as a method to identify human individuals.

2.1.1 History of Fingerprints

Fingerprints have long been involved in recognition and signature. Humans

have used fingerprints for personal identification for a very long time. Modern

fingerprint matching techniques initiated in the late 16th century. Henry Fauld, in

1880, first scientifically suggested the individuality and uniqueness of fingerprints.

At the same time, Herschel asserted that he had practiced fingerprint identification

for about 20 years. This discovery established the foundation of modern fingerprint

identification. In the late 19th century, Sir Francis Galton conducted an extensive

study of fingerprints. He introduced the minutiae features for single fingerprint

classification in 1888. Edward Henry made an important advance in fingerprint

identification in 1899 an elaborate method of indexing fingerprints very much tuned

to facilitating the human experts performing (manual) fingerprint identification. In

the early 20th century, fingerprint identification formally accepted as a valid personal

identification method by law enforcement agencies and became a standard procedure

in forensic. Fingerprint identification agencies were setup worldwide and criminal

fingerprint databases were established. With the advent of lives can fingerprinting

and availability of cheap fingerprint sensors, fingerprints are increasing used in

government and commercial applications for positive person identification [1]. The

conclusion of the history is show in Table 2.1 below.

Table 2.1: The History of Fingerprint [1]

Year Description

1686 Marcello Malpighi, a professor of anatomy at the University of

Bologna, noted in his; ridges, spirals and loops in fingerprints. He

made no mention of their value as a tool for individual

identification.

1823 John Evangelist Purkinji, a professor of anatomy at the University

of Breslau, published his thesis discussing 9 fingerprint patters, but

he too made no mention of the value of fingerprint for personal

identification

1856 Sir William Herschel, first used fingerprints on native contracts in

Jungipoor, India. Sir Herschel began to note that the inked

impressions could, indeed, prove or disprove identity. While his

experience with fingerprinting was admittedly limited, Sir

Herschel's private conviction that all fingerprints were unique.

1880 Dr. Henry Faulds forwarded an explanation of his classification

system and sample forms for recording inked impressions to Sir

Charles Darwin. The same year Faulds also published an article in

the scientific journal "Nature" where he discussed fingerprints as

means of personal identification and the use of ink as a method for

obtaining these.

RI

1891 Juan Vucetich, an Argentine Police Official, began the first

fingerprint files based on Galton pattern types. At first, Vucetich

included the Bertillon System with the files.

1897 On 12 June 1987, the Council of the Governor General of India

approved a committee report that fingerprints should be used for

classification of criminal records. Haque and Bose are the two

Indian fingerprint experts credited with primary development of

the Henry System of fingerprint classification and it is still used in

all English-speaking countries.

1946 The F.B.I. had processed 100 million fingerprint cards in manually

maintained files; and by 1971, 200 million cards. With the

introduction of AFIS technology, the files were split into

computerized criminal files and manually maintained civil files.

Many of the manual files were duplicates though, the records

actually represented somewhere in the neighborhood of 25 to 30

million criminals, and an unknown number of individuals in the

civil files.

2.1.2 Type of Fingerprints

Fingerprints can be dividing into the three (3) major pattern type arches,

loops, and whorls. These major pattern types can be dividing further into different

subgroup such as right, left or twin loops, plain or tended arches and spiral or

concentric circles as whorls [2]. Figure 2.1 shows the type of fingerprint with their

images.

In addition, a fingerprint pattern, when analyzed at different scales, exhibits

different types of features. Actually, it can be looked in three levels, which are the

global level, the local level, and the very-fine level. The descriptions of each level are:

7

a) Arch b) Tended Arch c) Left Loop

d) Right Loop 0 Twin Loop e) Whorl

Figure 2.1 Type of Fingerprint [2]

i) Global Level

At global level, have singularity points called core and delta points. These

singularity points are very important for fingerprint classification but not sufficient

for accurate matching.

Figure 2.2 Global Levels [2]

8

According to Figure 2.2, core and delta marked on sketches of the two (2)

fingerprint patterns loop and whorl. Loop has one delta, whorls have two deltas and

the minutiae detail is not shown.

ii) Local Level

At local level have minutiae details or minutiae points. Minutiae details also

known as ridge characteristics, ridge details or Galton's details. Based on figure 2.3,

one way to classify the minutiae details are in terms of ridge termination, bifurcation,

independent ridge, dot or island, lake, spur and crossover. The two most prominent

minutiae details are ridge termination (ending) and ridge bifurcation.

RIDGE TERMINATION

BIFURCATION

INDEPENDENT RIDGE

DOT OR ISLAND

LAKE

SPUR

CROSSOVER

Figure 2.3 Local Levels [2]

iii) Very-fine Level

At very-fine level, have essentially the finger sweat pores. The position and

shape of the pores can be use to help identify a person.

Figure 2.4 Very-fine Levels [2]

Based on Figure 2.4, part of a fingerprint image with sweat pores and

minutiae details visible. The black lines in the image correspond to the ridges in the

fingerprint, and the white lines in the image correspond to the valleys in the

fingerprint. The white dots on the ridges correspond to the sweat pores in the

fingerprint and are marked with empty circles on a single ridge line. Minutiae details

are marked with black-filled circles.

2.2 Fingerprints Recognition

2.2.1 Image Acquisition

Image acquisition is the first process for capture image using a camera digital.

CCD sensor or satellite that provide analog image to convert into digital image.

Digital image can be considered a matrix whose row and column indices identify a

point in the image and the corresponding matrix element value identifies the gray

level at that point. Example image acquisition application is using fingerprint reader

to capture or read an image for fingerprint or using satellite to capture image in

location detecting of flooding area.

10

2.2.2 Pre-processing Image

Pre-processing image consists of filtering image where the first image is not

clear and difficult to read. The purpose for filtering process is to improve the quality

of image after capture image from analog to digital image. Actually, there are

several technique in filtering technique such as order filter, median filter, average

filter, maximum or minimum filter. The technique that use is depends on the input of

image.

2.2.3 Image Threshold

Threshold is the part of image recognition. Fingerprint pixel in a grayscale

image are recognize as 'object' pixels if their value is greater than some threshold

value which assume an object to be brighter than the background and as

'background' pixels for otherwise. Basically, an object pixel is given a value of' 1'

while a background pixel is given a value of V. Threshold process is important

because able to differentiate between image of fingerprint and the background of

image. Besides, there are several techniques in threshold process such as median

threshold, minimum threshold, maximum threshold and fixed value threshold. The

technique that use is depends on the contrast between object pixels and background

pixels or for the foreground and background image.

2.2.4 Image Segmentation

The objective of image segmentation is to find regions that represent

fingerprint or meaningful parts fingerprint. They are several techniques using in

image segmentation like region growing and shrinking technique where an operating

by row and column based on image space. This method can be local that operate on

small neighborhoods or can be global that operating on the entire image or a

combination of local and global. Another technique is clustering technique where

I

segments the image by placing similar elements into groups or cluster based on the

similarity measure for example by color space or histogram space.

2.2.5 Feature Extraction

Feature extraction is the operation to extract image features for identifying or.

interpreting meaningful physical objects from images. That means from the feature

extraction, system recognized the feature or characteristic of the image and generates

the data by the binary image. Two (2) techniques of clustering images in feature

extraction are by row and column and divide the image into smaller window. The

example of application for feature extraction is fingerprint recognition. Whereby it

extract the fingerprint image features to recognize the type of images.

2.2.6 Template Matching

Template matching is the operation to match or compare the image based on

the feature image from the feature extraction process. This process was compared

the image from database and the input image. Using template matching, it can

recognize the image appropriately because using the data after feature extraction.

2.3 Current System Using Feature Extraction for Fingerprints Recognition

2.3.1 Feature Extraction Using Chain Code Contours 131

According to this thesis, the binarization approach is to convert the gray scale

image into a binary image prior to minutiae detection. While algorithms differ in

several implementation aspects, they have the following common stages.

-

12

According to the Figure 2.5, it shows the different stages in a typical feature

extractor following the binarization approach.

Enhancennent

Bniarieation

Thinning

Figure 2.5 Binarization Approach [3]

The first step starts in segmentation or binarization stage. In this step, the

gray scale image was converting to a binary image through the process of simple

threshold or some form of adaptive binarization. The quality of the binarization

output improved if the gray scale image was enhancing prior to this process. This

step is also referring to as segmentation in literature. However, this should not be

confused with segmentation of the fingerprint foreground from the background

during region mask generation.

The second step is the thinning stage. The resulting binary image is thinning

by an iterative morphological process resulting in a single pixel wide ride map.

Some algorithms and our proposed approach do not require this stage.

The third step is the Minutiae Detection where the problem of minutiae

detection is trivial. The resulting pixel wide map is scanned sequentially and

minutiae points are identified based on its neighborhood. Ridge endings are

characterized by single neighbor and bifurcations are identified by locating pixels

with three or more neighbors.

termination

bifurcation

lake

independent ridge

0 dot or island

spur

13

The last step is the post-processing stage. The minutiae extraction process

results in two forms of errors. The detection may introduce spurious minutiae where

they do not exist in the original or may miss genuine minutiae. While nothing can be

done about the missing minutiae, spurious minutiae can be eliminated by considering

their spatial relationships. Several heuristic rules may then be applied to filter out

these false positives.

2.3.2 Fingerprint Identification Based on The Minutiae Located in Fingerprint [4]

According to this thesis, the minutiae matching are the suitable method to use

for fingerprint comparison. Basically, minutiae are local discontinuities in the

fingerprint pattern. Only the ridge ending and ridge bifurcation minutiae type are

used in fingerprint recognition. Example of minutiae is shows in Figure 2.6.

(a)

(b)

Figure 2.6 (a) Different Minutiae Types, (b) Ridge Ending & Bifurcation