CHAPTER 1

OVERVIEW

1.1 INTRODUCTION:

Information is an essential element in decision making and the world is generating

increasing amounts of information in various forms with different degrees of complexity.

Hence, the need for improved information systems has become more conspicuous. One of

the major problems in the design of modern information systems is automatic pattern

recognition. A pattern is the description of an object. Pattern recognition can be defined

as the categorization of input data into identifiable classes via the extraction of significant

features or attributes of the data from a background of irrelevant detail. Recognition is

regarded as a basic attribute of human beings, as well as other living organisms. A human

being is a very sophisticated information system, partly because he possesses a superior

pattern recognition capability. There are so many pattern recognition systems and

biometrics is one of them, which gives the notion of life measure. Biometrics refers to a

broad range of technologies, systems, and applications that automate the identification or

verification of an individual based on his or her physiological or behavioral

characteristics. Physiological biometric is based on direct measurements of a part of the

human body at a point in time. The most common physiological biometrics involves

Fingerprint biometric system

Face biometric system

Iris biometric system

Retina biometric system

Palm biometric system

Wrist vein biometric system

Hand geometry biometric system

1

Behavioral biometrics is based on measurements and data derived from the

method by which a person carries out an action over an extended period of time. The

most common behavioral biometrics involves

Speech biometric system

Signature biometric system

Gait biometric system

Identifying an individual from his or her face is one of the most nonintrusive

modalities in biometrics. In fact, an automatic people identification system based

exclusively on fingerprints or on face recognition. The fingerprint verification is

trustworthy but it is intrusive and it can cause resistance to the users, depending on the

application. Face recognition is a compelling biometric because it is integral to everyday

human interaction and it has the following features.

Face recognition is the primary means for people recognizing one another so it is

natural.

Face recognition is nonintrusive.

Since the advent of photography, faces have been the primary method of

identification in passports and ID card systems.

Because optical imaging devices can easily capture faces, there are large legacy

databases that can provide template sources for facial recognition technology.

This includes police mug-shot, data-bases and television footage. Cameras can

acquire the biometric passively.

As such, facial recognition is more acceptable than other biometrics and is easy to

use. Face biometric system involves face detection and face recognition. Face detection is

the first step in automated face recognition. Face detection can be performed based on

several cues: skin color motion facial/head shape facial appearance or a combination of

these parameters. However this works considers only face recognition.

To make the process more authentic, we add palmprint identification in this project.

Palmprint identification is the measurement of palmprint features for recognizing the

identity of a user. Palmprint is universal, easy to capture and does not change much

2

across time. Palmprint biometric system does not requires specialized acquisition devices.

It is user-friendly and more acceptable by the public. Besides that, palmprint contains

different types of features, such as geometry features, line features, point features,

statistical features and texture features. The palmprint geometry features are insufficient

to identify individuals. This is because the palmprint geometry features such as palm size,

palm width and others for adults are relatively similar. The palmprint line features

include principal lines, wrinkles and ridges. Ridges are the fine lines of the palmprint. It

requires high-resolution image or inked palmprint image to obtain its features. Wrinkles

are the coarse line of the palmprint while the principle lines are major line that is

available on most of the palm (headline, lifeline & heart line). The separation of wrinkles

& principle line are difficult since some wrinkles might be as thick as principle lines. In

this work, peg-less right hand images different individuals were acquire. No special

lighting is used in this setup. The hand image is segmented and its key points are located.

The hand image is aligned and cropped according to the key points. The palmprint image

is enhanced and resized. Sequential modified Haar transform is applied to the resized

palmprint image to obtain Modified Haar Energy (MHE) feature. The sequential

modified Haar wavelet can maps the integer-valued signals onto integer-valued signals

without abandoning the property of perfect reconstruction. The MHE feature is compared

with the feature vectors stored in the database using Euclidean Distance. The accuracy of

the MHE feature and Haar energy feature under different decomposition levels and

combinations are compared. 94.3678 percent accuracy can be achieved using proposed

MHE feature.

1.2 AIM OF THE PROJECT:

The main goal of the project is to design a Matlab code for face recognition using

principal component analysis. It also includes palm identification using sequential

modified Haar wavelet energy to provide better authentication. The entire simulation is

done with the help of Matlab software.

1.3 METHODOLOGY:

3

BLOCK DIAGRAM:

Figure 1.1 Block Diagram.

In this project, we take different images of face and right hand palm of different

persons and load them in the database. Then an unknown face and right hand palm of a

person are captured and compared with the images in the database with their respective

algorithms. For face recognition, we use principal component analysis algorithm and for

palm, we use modified Haar transformation model. If the query images of face and palm

are same as the images in the database, then the person is authentic otherwise not.

1.4 SIGNIFICANCE OF THE WORK:

4

Input face image

Face detection

Face feature Extraction

Feature Matching Decision maker

Output result

Face and palm print image database

Input palm print image

Palm print detection

Palm print feature Extraction

Face or palm recognition plays an important role in today’s world. It has many real-

world applications like human/computer interface, surveillance, authentication and video

indexing. Humans often use faces to recognize individuals and advancements in

computing capability over the past few decades now enable similar recognitions

automatically. It is desirable to have a system that has the ability of learning to recognize

unknown faces. Face recognition has become an important issue in many applications

such as security systems, credit card verification and criminal identification. For example,

the ability to model a particular face and distinguish it from a large number of stored face

models would make it possible to vastly improve criminal identification. Face and palm

recognition depends heavily on the particular choice of features used by the classifier.

One usually starts with a given set of features and then attempts to derive an optimal

subset (under some criteria) of features leading to high classification performance with

the expectation that similar performance can also be displayed on future trials using novel

(unseen) test data.

1.5 ORGANIZATION OF THE REPORT:

The chapters are arranged in the following manner:

Chapter 1 deals with the introduction, aim of the project, methodology,

significance of the work and organization of the report.

Chapter 2 deals with biometric system vulnerabilities.

Chapter 3 deals with Face recognition using principal component analysis.

Chapter 4 deals with Palmprint Identification Using Sequential Modified Haar

Wavelet Energy.

Chapter 5 deals with Matlab software.

Chapter 6 deals with Results and Conclusion.

CHAPTER 2

5

BIOMETRIC SYSTEM VULNERABILITIES

2.1 INTRODUCTION:

All security measures, including mechanisms for authenticating identity have

ways of being circumvented. Certainly, the processes in working around these measures

vary difficulty based on effort and resources needed to carry out the deceptive

act. Authentication mechanisms based on secrets are particularly vulnerable to "guessing"

attacks. Token mechanisms that rely on the possession of an object, most notably a card

or badge technology are most vulnerable to theft or falsified reproduction. Biometric

technologies closely tie the authenticator to individual identity of the user through the use

of physiological or behavioral characteristics. While this property is an added advantage

over the previous two authentication mechanisms mentioned. It places a great emphasis

on validating the integrity of the biometric sample acquired and transferred in the

biometric system. Rather, N., et al. provided a model identifying vulnerabilities in

biometric systems. An example of the threat model is shown below in Figure 2.1 and

builds on the general biometric model outlined in Mansfield and Wayman.

Figure 2.1 Biometric Threat Model.

6

This model contains 11 individual areas at which vulnerabilities in biometric

systems exist. In addition to the five main internal modules that are characterized in the

General Biometric Model (data collection, signal processing, matching, storage and

decision), an additional component is added to represent the transfer of the authentication

decision to the greater application that relies on the decision from the biometric

system. Such applications could be identity management systems (IDMS) or access

control systems for logical and or physical access to resources. These systems can vary in

complexity and size ranging from a local computer log-in all the way to a wide scale

distributed architecture seen in the cases of the Transportation Worker Identification

Credential (TWIC) or Personal Identity Verification (PIV) of Federal Employees and

Contractors. The remaining points of vulnerability are communication channels between

these six modules. It is worth noting that not all 11 vulnerability points are unique to the

biometric system. Many of the same points such as storage and communication channels

are vulnerable in other authentication systems and similar methods can be used to limit

those particular vulnerabilities.

The most publicized vulnerability in biometric systems resides at the Data

Collection module in the form of spoofing or presentation artificial representations of

biometric samples (module #1 in Figure 2.1). If an artificial or fake biometric sample is

accepted by the biometric system at this initial stage, the entire biometric system is

corrupted and the system has been compromised. Attacks on the biometric system are not

new, it is popular culture to circumvent security systems and biometric systems are not

immune to this. Several online resources are available that describe such attacks on the

data collection module and many movies and television programs highlight attacks on

such systems. One such attack at this data collection module was outlined in the work of

Matsumoto, T., et al. in 2002 using "Gummy Fingers". The biometric research

community as well as industry has focused on research on preventing such attacks by

using the concept of “live-ness" detection techniques. Today, the newer sensors are

improving their resilience against a spoofing attack at this module. Previously, an acetate

spoofing attack where an image of a fingerprint placed on acetate was accepted as a

genuine live finger, were easy to do such attacks are providing increasingly difficult to do

and hence the more complicated approaches in vulnerability attacks being waged on the

7

sensor. And as such, techniques for live-ness detection within the fingerprint modality

focus on moisture content, temperature, electrical conductivity, and challenge response.

Figure 2.2 Various Biometrics.

A number of biometric identifiers are used in various applications. Each biometric

has its strengths and weaknesses and the choice typically depends on the application. No

single biometric is expected to effectively meet the requirements of all the applications.

The match between a biometric and an application is determined depending upon the

characteristics of the application and the properties of the biometric.

At its most simple level, biometric systems operate on a three-step process. First,

a sensor captures a physical or a behavioral sample. Second, the biometric system

develops a way to describe the observation mathematically that is obtaining a biometric

signature by separating the noise and other irrelevant features from the captured sample.

Third, the computer system inputs the biometric signature into a comparison algorithm

and compares it to one or more biometric signatures previously stored in its database.

8

Figure 2.3 Steps in Biometric Systems.

Figure 2.4 Major steps involved in any Biometric identification.

9

2.2 MAJOR STEPS OF ANY BIOMETRIC SYSTEM:

Data collection

Signal processing

Decision

Transmission

Store

10

CHAPTER 3

FACE RECOGNITION USING ONE DIMENSIONAL

PRINCIPALCOMPONENT ANALYSIS

3.1 INTRODUCTION:

Face recognition is one of the popular biometric systems. It is very high level task

for which developing a computational model is difficult because faces are complex,

multidimensional and meaningful visual stimuli. Face recognition can be treated as a

two-dimensional recognition problem taking the advantage of the fact that faces are

normally upright and thus maybe described by a small set of two-dimensional

characteristic views. Although face recognition is a high level visual problem, there is

quite a bit of structure imposed on the task. The advantage of this structure for

recognition scheme is based on an information theory approach. In the language of

information theory, relevant information is to be extracted in a face image, encode it as

efficiently as possible, and compare one face encoding with a database of models

encoded similarly. A simple approach to extract the information contained in an image of

a face is to somehow capture the variation in a collection of face images and use this

information to encode and compare individual face images. A common problem in

statistical pattern recognition is that of feature selection or feature extraction. Feature

selection refers to a process whereby a data space is transformed into a feature space that,

in theory, has exactly the same dimension as the original data space. However, the

transformation is done in such a way that the data set may be represented by reduced

number of effective features yet retain most of the intrinsic information content of the

data. In other words, the data set undergoes a dimensionality reduction. Principal

component analysis (PCA) or Karhunen-Loeve transformation is popularly used

technique for prediction, redundancy removal, feature extraction and dimensionality

reduction in which the main distinguishable features of the individual faces are to be

11

extract when applied to face recognition. In mathematical terms, the principal

components of the distribution of faces or the Eigen vectors of the covariance matrix of

the set of face images are to be found. Face images are projected onto a feature space that

best encodes the variation among known face images. Principal Component Analysis

(PCA) is discussed in section 3.2. Algorithm for face recognition using one dimensional

PCA is presented in section 3.3.

3.2 PRINCIPAL COMPONENTS ANALYSIS:

Principal component analysis is an optimal feature extraction and universally used

dimensionality reduction technique based on extracting the desired number of principal

components of the multi-dimensional data. PCA finds an alternative set of parameters for

a set of raw data (or features) such that most of the variability in the data is compressed

down to the first few parameters. A face image is a two-dimensional N by N array of

intensity values or a vector of dimension N2. A typical image of size 256 by 256

describes a vector of dimension 65,536, or equivalently a point in 65,536-dimensional

space. An ensemble of images maps to a collection of points in this huge space. Images

of faces, being similar in overall configuration will not be randomly distributed in this

huge image space and thus can be described by a relatively low dimensional subspace.

The face images in the training set are normalized by subtracting the average face, m of

the set. In one dimensional PCA, two dimensional face image matrices must be

previously transformed into one dimensional image vectors. The main idea of the

principal component analysis is to find the vectors which best account for the distribution

of face images within the entire image space. These vectors define the subspace of face

images that call “face space”. Each vector is of length N2, describes an N by N image,

and is a linear combination of the original face images. Because these vectors are the

Eigen vectors of the covariance matrix corresponding to the original face images and

because they are face like in appearance, they are referred as eigenfaces. Each face image

in the training set can be represented exactly in terms of linear combination of the

eigenfaces. The number of possible eigenfaces is equal to the number of face images in

the training set. However, the faces can also be approximated using only the best

12

eigenfaces, those accounts for the most variance with in the set of face images. The

primary reason for using fewer eigenfaces is computational efficiency. The best k

Eigenvectors or eigenfaces span a k-dimensional subspace of all possible images.

The feature matrix of a particular image gives the information about the

composition of the eigenfaces that belongs to the Eigen space. PCA is to find out the

projection matrix through which the features of the given data set are obtained by

projecting them onto this projection matrix.

Let A denote N-dimensional random vector and it has zero mean

E [A] = 0 …………………………………………………………………………....... (3.1)

Where, E is the statistical expectation operator. If A has a nonzero mean, subtract

the mean from it. Let X denote a unit vector, also has dimension n, onto which the vector

A is to be projected. The projection is defined by the inner product of the vectors A and

X, as shown by

Y=ATX=XTA …………………………………………………………...……………. (3.2)

Subject to the constraint,

ǁXǁ = (XTX) 1/2 = 1 …………………………………………………...………………. (3.3)

The projection Y is a random variable with the mean and variance related to the

statistics of the vector A. Under the assumption that the random vector A has zero mean,

it follows that the mean value of the projection Y is zero too.

E[Y] = XT .E[A] = 0 ……………………………………………………….………… (3.4)

The variance of Y is therefore the same as its mean-square value, and so

σ2 = E [Y2]

= E [(XTA) (ATX)]

= XT E [[AAT].X

13

= XT CX ……………………………………………………………………………… (3.5)

The N-by-N matrix C is the correlation matrix of the random vector A, formally

defined as the expectation of the outer product of the vector A with itself, as shown by

C = E [AAT] …………………………………………………………………….…… (3.6)

The correlation matrix C is symmetric CT= C and it follows that if a and b are any

N-by-1 vectors then

aTCb = bTCa …………………………………………………………………………. (3.7)

The variance σ2 of the projection Y is a function of the unit vector and is given as

Ψ(X) = σ2.

= XT.C.X …………………………………………………………………….… (3.8)

The maximum variation between the features can be obtained only when the

projection matrix is having its vectors uncorrelated. The issue to be considered is that of

finding those unit vectors X along which the features obtained will have maximum

variation in between them. The solution to this problem lies in the Eigen structure of the

correlation matrix C. The Eigen vectors obtained for a correlation matrix will be

orthogonal to each other i.e. maximally uncorrelated. This is discussed as follows

CX = λX …………………………………………………………………….……….. (3.9)

This is the equation that governs the unit vectors X for which the equation has

maximum variance. The problem has nontrivial solutions (i.e., X! = 0) only for special

values of λ that are called the eigenvalues of the correlation matrix C. The associated

values of X are called eigenvectors. A correlation matrix is characterized by real,

nonnegative eigenvalues. The associated eigenvectors are unique, assuming that the

eigenvalues are distinct. Let the eigenvalues of the n-by-n matrix C be denoted by λ1 ,

λ2 ,L, λN and the associated eigenvectors be denoted by X1 , X2 ,L, XN respectively. Hence

CXj = λjXj where j = 1, 2... N …………………………………………………… (3.10)

14

Let the corresponding eigenvalues be arranged in decreasing order:

λ1> λ2, > L >λj>λN…………………………………………………………….…….. (3.11)

So that λ1 = λmax. Let the associated eigenvectors corresponding to Eigen values in

descending order are represented as X1, X2, L, XN be used to construct an N-by-N matrix

X = [X1, X2, L, XN] …………………………………………………………...……. (3.12)

The set of N equations are combined into a single equation to represent as

CX = XΛ ………………………………………………………………………........ (3.13)

Where, Λ is a diagonal matrix defined by the eigenvalues of the covariance matrix

and is given as

Λ = diag[λ1, λ2 ,L, λN] ………………………………………………………...…….(3.14)

The matrix X is an orthogonal (unitary) matrix in the sense that its column vectors

(i.e., the eigenvectors of C) satisfy the conditions of orthonormality.

XiTXj= 1, j = i

= 0, j ≠ i

= X, X = i

The inverse of matrix X is the same as its transpose, XiT = X-1 and the orthogonal

similarity transformation is XT CX= Λ, and is elaborated as follows

XiTCXk= λj, k = j

= 0, k! = j

The orthogonal similarity (unitary) transformation transforms the correlation

matrix C into a diagonal matrix of eigenvalues. The correlation matrix C may itself be

expressed in terms of its eigenvalues and eigenvectors and is referred as the spectral

theorem as follows The outer product Xi is of rank 1 for all i. Principal component

analysis and Eigen decomposition of matrix C are basically one and the same, just

15

viewing the problem in different ways. The eigenvectors corresponding to nonzero

eigenvalues of the covariance matrix produce an orthonormal basis for the subspace

within which most image data can be represented with a small amount of error. The

eigenvectors are sorted from high to low according to their corresponding eigenvalues.

The eigenvector associated with the largest Eigen value is one that reflects the greatest

variance in the image. The smallest Eigen value is associated with the eigenvector that

finds the least variance. The k significant Eigen vectors of the projection matrix are

chosen as those with the largest associated Eigen values. The number of Eigen faces to be

used is chosen heuristically based on the Eigen values. A face image in the training set is

projected by simple multiple operation with the projection matrix X which is of size ------

to obtain the feature vector y of size k, and it is applied to all the face images in the

training set to form Y = [Y1 ,Y2 ,L,YM]

Where, M is the number of face images in the training set. A new face image Q,

which is transformed into a vector is projected onto face space by a simple operation Y q=

XT (Q − m). The new face image is recognized as a known face image from the training

set by considering the minimum Euclidean distance between the unknown face feature

vector and trained face feature vector.

3.3 ALGORITHM:

The steps for face recognition using one dimensional principal component analysis

are as follows:

Step 1: Obtain face images I1, I2, L, IM and represent every image as one dimensional

vector. Every image of same size NxN must be resized to (N2x1) in the database.

16

Step 2: Suppose Г is an N2x1 vector, corresponding to an NxN face image I. The idea is

to represent Г (Φ = Г - mean face) into a low dimensional space.

Step 3: Computation of the average face vector Ψ:

Step 4: Subtract the mean face:

Step 5: Compute the covariance matrix C:

Step 6: Compute the Eigen vectors ui of AAT :

The matrix AAT is very large and computing it is not practical.

Step 6.1: Compute the matrix ATA (MxM matrix).

Step 6.2: Compute the Eigen vectors vi of ATA.

17

ATAvi = uivi

The relationship between ui and vi:

ATAvi = μivi => AATAvi = μiAvi =>

CATAvi = μiAvi => Cvi = μivi where ui = Avi

Thus, AAT and ATA have the same eigenvalues and eigenvectors are related as

follows:

ui = Avi

Note 1: AAT can have upto N2 eigenvalues and eigenvetors.

Note 2: ATA can have upto M eigenvalues and eigenvectors.

Note 3: The M eigenvalues of ATA (along with their corresponding

eigenvectors) correspond to the M largest eigenvalues of AAT:ui = Avi (along

with their corresponding eigenvectors).

Step 6.3: Compute the M best eigenvectors of AAT: ui = Avi.

(Important: normalize ui such that ǀǀuiǀǀ = 1)

Step 7: Keep only K eigenvectors (corresponding to the k largest eigenvalues). Each face

(minus the mean) Φi in the training set can be represented as a linear combination of the

K best eigenvectors:

(We call the uj’s eigenfaces)

18

Each normalized training face Φi is represented in this basis by a vector:

Unknown Image:

Given an unknown face image Г (centered and of the same size like the training

faces) follow these steps:

Step 1: Normalize Φ = Г – Ψ

Step 2: Projection on eigenspace:

Step 3: Represent Φ as:

Step 4: Find er = minlǀǀΩ- Ω lǀǀ

Step 5: If er < Tr, then Г is recognized as face l from the training set. The distance er is

called distance within the face space (difs).

Comment: we can use the common Euclidean distance to compute er, however, it has

been reported that the Mahalanobis distance performs better:

(Variations along all axes are treated as equally significant)

19

CHAPTER 4

PALMPRINT IDENTIFICATION USING

SEQUENTIAL MODIFIED HAAR WAVELET

ENERGY

4.1 INTRODUCTION:

Palmprint identification had been introduced a decade ago. It is defined as the

measurement of palmprint features to recognize the identity of a person. Palmprint is

universal because everyone has palmprint. It is easy to capture using digital cameras.

Palmprint does not change much across time. Palmprint has advantages compared to

other biometric systems. Iris scanning biometric system can provides a high accuracy

biometric system but the cost of iris scanning devices is high. Palmprint biometric system

can captures hand images using a conventional digital camera. Palmprint biometric

system is user-friendly because users can grant the access frequently by only presenting

their hand in front of the camera. Palmprint biometric system can achieve higher

accuracy than hand geometry biometric system, because the geometry or shape of the

hand for most of the adults is relatively similar. Palmprint contains geometry features,

line features, point features, statistical features & texture features. The palmprint

geometry features are insufficient to identify individuals. This is because the palmprint

geometry features such as palm size, palm width and others for adults are relatively

similar. The palmprint line features include principal lines, wrinkles and ridges. Ridges

are the fine lines of the palmprint. It requires high-resolution image or inked palmprint

image to obtain its features. Wrinkles are the coarse line of the palmprint while the

principle lines are major line that is available on most of the palm(headline, lifeline and

heartline).The separation of wrinkles & principle line are difficult since some wrinkles

might be as thick as principle lines. Palmprint point features use the minutiae points or

delta points to identify an individual. Point features require high resolution hand image

20

because low-resolution hand image does not have a clear point’s location. Palmprint

statistical features represent the palmprint image in a statistical form to identify and

individual. Some of statistical methods available are Principle Component Analysis

(PCA) and Independent Component Analysis (ICA). Palmprint texture features are

usually extracted using transform-based method such as Fourier Transform and Discrete

Cosine Transform. Besides that, Wavelet transform is also used to extract the texture

features of the palmprint. In this work, a sequential modified Haar wavelet is proposed to

find the Modified Haar Energy (MHE) feature. The sequential modified Haar wavelet can

maps the integer- valued signals onto integer-valued signals without abandoning the

property of perfect reconstruction shows the proposed palmprint identification using

sequential "S-transform" modified Haar transform. In this work, ten images from the

right hand of 100 individuals are acquired using a digital camera. The hand image is

segmented and the key points are located. By referring to the key points, the hand image

is aligned and the central of the palm is cropped. The palmprint image is enhanced and

resized. The energy features of the palmprint are extracted using sequential modified

Haar wavelet. The Modified Haar Energy (MHE) is represented using feature vector and

compared using Euclidean distance with the feature vectors stored in the database.

4.2 BLOCK DIAGRAM:

Figure 4.1 Block Diagram of Palmprint Identification Using Sequential Modified Haar

Wavelet.

21

4.3 IMAGE ACQUISITION:

Ten right hand images from 100 different individuals are captured using Canon

PowerShot A420 digital camera. During the image acquisition, no pegs alignment is

made. The hand image is taken without any special lighting condition. Dark intensity

backgrounds, such as black and dark blue are used in this work. The usage of the low

intensity background is to ease the hand image segmentation. Users are required to spread

their fingers apart and lean their hand against the background during the image

acquisition process. Figure 4.2 shows the hand image acquired. The hand image is saved

using JPEG format in 1024 x 768 pixels. The hand image has the resolution of 180 dpi

(dot per inches).

Figure 4.2 Hand Image.

4.4 IMAGE PREPROCESSING:

The hand images are preprocessed to obtain the region-of- interest (ROI) or also

known as palmprint image. In image preprocessing stage, hand image segmentation, key

point’s determination and palmprint extraction are done.

The skin has higher red intensity color while the background (black or dark blue

color) has lower red intensity color. Thus, by referring to the red component of the hand

image, the hand image is segmented using Otsu's Method.

22

Figure 4.3 Binary Hand Image

Otsu Segmentation:

In 1979, Japanese scholar Otsu presented a global thresholding algorithm.

Suppose that an image compose of target and background, which have different gray-

level, and the target jsat higher gray level. The gray-level based on the statistical

histogram ranges &om 0 to L. Between 0 and L , threshold K is chosen to segment the

image into two classes: the background whose gray-level is from 0 to K and the target

whose gray-level is komK+i to L. If a certain threshold K can make the value of

interclass variance the highest among all the possible 'values.

The threshold K is the one with which target and background can be accurately

divided. The K is the final are the variance values of target, background and image

threshold we are looking for. The equations involved are as follows:

23

where is the amount of pixel at gray-level I, N is the total pixel amount in the

image; p ( i ) is probability, , and , are the probabilities of target and background

respectively are the mean gray-level values of target ,background and image

respectively; are the variance values of target, background and image

respectively; , and , are the interclass and intraclass variance value respectively;

is the function for threshold selection.

24

After hand image segmentation, the boundary pixels for the binary hand image

are tracked using boundary-tracking algorithm. Figure 4.4 shows the boundary pixels of

the hand image.

The central of the wrist, Pw, is the middle pixels of the boundary pixels located at

the edge of the hand image. Figure 4.4 marks the location of Pw. The distance between

Pw and boundary pixels, Distp, in clockwise direction is calculated. Figure 4.5 shows the

graph of Distp plotted against the index of the boundary pixels.

Figure 4.4 Boundary Pixel of Palm.

From Figure 4.5, Key Point 1, KI, is the first local minima in the graph while Key

Point 2, K2, is the third local minima in the graph. The coordinate of the key points in the

hand image is determined using the index of the boundary pixels for KI and K2. KI is the

bottom pixels at the gap between little and ring fingers while K2 is the bottom pixels at

the gap between the middle finger and index fingers as in Figure 4.6.

Figure 4.5 Graph of Dist, Plotted against boundary pixel index.

25

The angle of rotation for the hand image is calculated according to the Figure 4.6.

Let LineK is the line between KI and K2. LineK is rotated 0 degrees clockwise at origin

KI so that LineK is parallel with the x-axis for all the hand images. The distance between

KI and K2 is calculated as (1).

Figure 4.6 Location of k1, K2 with rotation angle

From the experimental results, the ROI mask is selected 0.2 x DistK below the

LineK while the length of the square ROI is 1.4 x DistK. The selection of ROI mask with

0.2 x DistK below the LineKallow concentration of the ROI mask at the central of the

palm. The variable ROI mask size eases the extraction of the palmprint in different size.

Figure 7 shows the length of the lineK (DistK), length of the square ROI (1.4 x DistK)

and distance between lineK and square ROI (0.2 x DistK).

Figure 4.7 Distk, Length of the Square ROI and Distance between Linekand Square ROI

26

A smaller hand image is obtained by cropping the hand image according to the

minimum and maximum of the x and y coordinates of the square ROI. The cropped hand

image is rotated by 0 degrees clockwise. The diagonal pixels of the rotated image are

determined. By referring to the index of the first non-black pixel, m, the rotated hand

image is cropped m pixels from its sides (top, bottom, left and right). Figure 4.8 shows

the extracted palmprint image using the proposed method.

The palmprint image is in RGB format. It is converted into grayscale intensity

format, ImG before image enhancement. Figure 4.9 shows the palmprint image in

grayscale intensity format.

Figure.4.8.Palmprint Image Figure 4.9 Palmprint Image in Gray scale Format

The grayscale image is enhanced using two different methods. In the first method,

the grayscale hand image's histogram is adjusted to the full 256-color bins to form

Adjusted Palmprint Image, ImA. In Histogram Equalized Palmprint Image, ImH, the

histogram of the grayscale hand image is equalized into the full 256 bins. To test for the

effectiveness of the enhanced image, the grayscale image is also used in further

processing. Figure 4.10 shows the Adjusted Palmprint Image, ImA and Histogram

Equalized Palmprint Image, ImH from the same individual.

Fig 4.10 Adjusted Palmprint Image ImA and Histogram Equalized palmprintImage ImH

27

4.5 SEQUENTIAL HAAR TRANSFORM:

Wavelet transform is a multi-resolution tool that can be used to analyze the

palmprint image in different decomposition levels. Level one palmprint decomposition is

used to extract the fine lines of the palmprint. The higher the decomposition level, the

coarser the extracted palm lines, such as wrinkles and principal lines. Haar wavelet is

used to find the discontinuity between two pixels. It is not calculation expensive

compared to other types of wavelet such as Daubechies wavelets, Mexican hat wavelets

and Morlets wavelets.

Haar wavelet decomposes the palmprint image into approximation (A), horizontal

detail (H), vertical detail (V) and diagonal detail (D). Figure 4.11 shows one of the

methods used to obtain the 2-dimensional Haar coefficients.

Figure 4.11 Haar Wavelet Decomposition.

Where, X or Y axis represents the direction of addition and subtraction. Even is

the even pixels number (2, 4, 6 ... 256) and odd is the odd pixels number (1, 3, 5 ... 255).

For wavelet decomposition in multiple levels, the approximation for the previous level,

Ai-1 are used to further decompose into approximation, horizontal detail, vertical detail

and diagonal detail for the next decomposition level (Ai, Hi, Vi, and Di).

The Haar wavelet coefficients are represented using decimal numbers. It required

eight bytes for storing each of the Haar coefficients. The division of the subtraction

results in y-axis operation will also reduce the difference between two adjacent pixels.

28

Due to these reasons, sequential Haar wavelet that maps an integer-valued pixel onto

another integer-valued pixel is suggested. Sequential Haar coefficient requires only two

bytes to store each of the extracted coefficients. The cancellation of the division in

subtraction results avoids the usage of decimal numbers while preserving the difference

between two adjacent pixels. Figure 4.12 shows the decomposition using sequential Haar

wavelet.

Figure 4.12 Sequential Haar Transform

Where, X or Y-axis represents the direction of addition and subtraction. 1/2 is the

rounding of the "divide by 2" results towards the negative infinity.

From Figure 4.12 it is shown that the addition results are divided by 2 and

rounded to negative infinity. It is different than the original Haar wavelet decomposition

in Figure 4.11 that divides the results by 2 after the operation in Y-axis. No operation was

done to the subtraction results in modified Haar wavelet decomposition.

The grayscale image and enhanced images (ImA and ImH) is decomposed using

sequential modified Haar wavelets into six-decomposition levels. Figure 4.13 shows the

coefficients for the six levels of the decomposition using sequential modified Haar

wavelet in image representation. Figure 4.14 shows the location of horizontal details,

vertical details and diagonal details for every decomposition levels in Figure 4.13.

29

Figure 4.13 Image Representation for Coefficients in Six Level of Decomposition using

Sequential Modified Haar Wavelet.

In this work, only the details coefficients, C, of the sequential modified Haar

wavelet is used to find the energy feature. This is because the approximation coefficients

contain mean component of the palmprint image.

Figure 4.14 Locations of Details in Every Decomposition Levels.

For every image of the detail coefficients, the image is further divided into 4 x 4

blocks. The Modified Haar Energy (MHE) for each of the block is calculated using (2).

Where i is the level of decomposition, j is Horizontal, Vertical or diagonal details,

k is the block no of 1 to 16, P *Q is the size of the block.

30

The MHE energy feature for every detail coefficients are arranged as (3).

Let DetailHVD, i represents the combination of horizontal, vertical and diagonal

detail coefficients in i decomposition level, as in (4).

The DetailHVD in different decomposition levels are normalized using (5).

Several types of feature representation are tested in this work. They are:

FV1 - Level one normalized DetailHVD, D1.

FV2 - Level two normalized DetailHVD, D2.

FV3 - Level three normalized DetailHVD, D3.

FV4 - Level four normalized DetailHVD, D4.

FV5 - Level five normalized DetailHVD, D5.

FV6 - Level six normalized DetailHVD, D6.

FV7 - Combination of D1, and D2.

FV8 - Combination of Dl, D2 and D3.

FV9 - Combination of D1, D2, D3 and D4.

FV1O - Combination of DI, D2, D3, D4 and D5.

FV11 - Combination of D1, D2, D3, D4, D5 and D6.

FV1 to FV6 is the feature representation for different decomposition level while

FV7 to FV11 is the feature representation for different composition levels. Figure 4.15

31

shows intra-class feature vectors (Feature Vector obtained from the same individual in

different time interval) while Figure 4.16 shows the inter-class feature vectors (feature

vectors obtained from different individuals).

Figure 4.15 Intra-class Feature Vectors Figure 4.16 Inter-class Feature Vectors

From Figure 4.15 it is shown that the feature vector does not vary much in intra-

class. From Figure 4.16 the feature vector for Figure 4.15 is plotted in gray color while

the feature vector for another individual is plotted in black color. It is shown that the

feature vector for different individuals (inter-class) varied in amplitude and peaks.

In the combination of different decomposition levels (FV7 to FVII), the Euclidean

distance for all decomposition level is added up. Since some of the decomposition levels

might give a lower minimum total error rate (MTER) than others, a suitable weight can

be used to enhance the better discriminate decomposition levels. The Euclidean distance

and MTER will be explained in the next section (Section Feature Matching). The weight

of the decomposition level i is defined as

32

Where, MTERi is the Minimum Total Error Rate at i-th level decomposition, K is

the total number of decomposition levels. The feature representations for weighted

combination of different decomposition levels are:

FV12 - Weighted combination of D1 and D2.

FV13 - Weighted combination of Dl, D2 and D3.

FV14 - Weighted combination of D1, D2, D3 and D4.

FV15 - Weighted combination of D1, D2, D3, D4 and D5.

FV16 - Weighted combination of D1, D2, D3, D4, D5 and D6.

4.6 FEATURE MATCHING:

Euclidean distance measures the similarity between two different feature vectors

using (7).

Where J is the length of the feature vector, FVi is the feature vector for individual i.

Each of the feature vectors is matched using Euclidean Distance with the

remaining 999 feature vectors in the database. The genuine ED distribution graph and

imposter ED distribution graph are normalized because for every feature vector, there

will be nine genuine matching and 990 imposter matching. If both of the graphs are

plotted directly, the imposter ED distribution graph with 990000 values will cover the

genuine ED distribution graph with only 9000 value. Thus, the FAR and FRR is

normalized to 0.5 each, so that the total is equal to one. Theoretically, the system will get

a zero when there are no false acceptance and no false rejection. Since there is no perfect

system, Figure 4.17 shows the graph of false acceptance rate (FAR) and false rejection

rate (FRR) versus the threshold index.

33

Figure 4.17 Graph of FAR and FRR versus the Threshold Index

The interceptions between these two graphs yield a minimum total error rate

(MTER). MTER is defined as the minimum error a system can achieve regardless the

system is tends for civil application or forensic application. By referring to the MTER, a

suitable threshold is selected to differentiate the genuine user and the imposter. Besides

that, the MTER for the FV1 to FV6 are also used to find the weight in FV12 to FV16.

34

CHAPTER 5

SOFTWARE DESCRIPTION

5.1 INTRODUCTION:

5.1.1 MATLAB:

MATLAB® is a high-performance language for technical computing. It integrates

computation, visualization, and programming in an easy-to-use environment where

problems and solutions are expressed in familiar mathematical notation. Typical uses

include:

Math and computation

Algorithm development

Modeling, simulation, and prototyping

Data analysis, exploration, and visualization

Scientific and engineering graphics

Application development, including graphical user interface building.

MATLAB is an interactive system whose basic data element is an array that does not

require dimensioning. This allows you to solve many technical computing problems,

especially those with matrix and vector formulations, in a fraction of the time it would

take to write a program in a scalar non interactive language such as C or FORTRAN.

The name MATLAB stands for matrix laboratory. MATLAB was originally written

to provide easy access to matrix software developed by the LINPACK and EISPACK

projects. Today, MATLAB uses software developed by the LAPACK and ARPACK

projects, which together represent the state-of-the-art in software for matrix computation.

35

MATLAB has evolved over a period of years with input from many users. In

university environments, it is the standard instructional tool for introductory and

advanced courses in mathematics, engineering, and science. In industry, MATLAB is the

tool of choice for high-productivity research, development, and analysis.

MATLAB features a family of application-specific solutions called toolboxes. Very

important to most users of MATLAB, toolboxes allow you to learn and apply specialized

technology. Toolboxes are comprehensive collections of MATLAB functions (M-files)

that extend the MATLAB environment to solve particular classes of problems. Areas in

which toolboxes are available include signal processing, control systems, neural

networks, fuzzy logic, wavelets, simulation, and many others.

5.1.2 The MATLAB System:

The MATLAB system consists of five main parts:

1. Development Environment:

This is the set of tools and facilities that help you use MATLAB functions

and files. Many of these tools are graphical user interfaces. It includes the

MATLAB desktop and Command Window, a command history, and browsers for

viewing help, the workspace, files, and the search path.

2. The MATLAB Mathematical Function Library:

This is a vast collection of computational algorithms ranging from

elementary functions like sum, sine, cosine, and complex arithmetic, to more

sophisticated functions like matrix inverse, matrix eigenvalues, Bessel functions,

and fast Fourier transforms.

3. The MATLAB Language:

This is a high-level matrix/array language with control flow statements,

functions, data structures, input/output, and object-oriented programming

features. It allows both "programming in the small" to rapidly create quick and

36

dirty throw-away programs, and "programming in the large" to create complete

large and complex application programs.

4. Handle Graphics®:

This is the MATLAB graphics system. It includes high-level commands

for two-dimensional and three-dimensional data visualization, image processing,

animation and presentation graphics. It also includes low-level commands that

allow you to fully customize the appearance of graphics as well as to build

complete graphical user interfaces on your MATLAB applications.

5. The MATLAB Application Program Interface (API):

This is a library that allows you to write C and FORTRAN programs that

interact with MATLAB. It include facilities for calling routines from MATLAB

(dynamic linking), calling MATLAB as a computational engine, and for reading

and writing MAT-files.

5.2 DEVELOPMENT ENVIRONMENT:

Introduction:

This chapter provides a brief introduction to starting and quitting

MATLAB, and the tools and functions that help you to work with MATLAB

variables and files. For more information about the topics covered here, see the

corresponding topics under Development Environment in the MATLAB

documentation, which is available online as well as in print.

Starting and Quitting MATLAB:

Starting MATLAB:

On a Microsoft Windows platform, to start MATLAB,

double-click the MATLAB shortcut icon on your Windows

desktop.

37

On a UNIX platform, to start MATLAB, type Matlab at the

operating system prompt.

After starting MATLAB, the MATLAB desktop opens -

see MATLAB Desktop.

You can change the directory in which MATLAB starts,

define startup options including running a script upon startup, and

reduce startup time in some situations.

Quitting MATLAB:

To end your MATLAB session, select Exit MATLAB from

the File menu in the desktop, or type quit in the Command

Window. To execute specified functions each time MATLAB

quits, such as saving the workspace, you can create and run a

finish.m script.

MATLAB Desktop:

When you start MATLAB, the MATLAB desktop appears, containing

tools (graphical user interfaces) for managing files, variables, and applications

associated with MATLAB.

The first time MATLAB starts, the desktop appears as shown in the

following illustration, although your Launch Pad may contain different entries.

You can change the way your desktop looks by opening, closing, moving

and resizing the tools in it. You can also move tools outside of the desktop or

return them back inside the desktop (docking). All the desktop tools provide

common features such as context menus and keyboard shortcuts.

You can specify certain characteristics for the desktop tools by selecting

Preferences from the File menu. For example, you can specify the font

characteristics for Command Window text. For more information, click the Help

button in the Preferences dialog box.

38

Desktop Tools:

This section provides an introduction to MATLAB's desktop tools. You can also

use MATLAB functions to perform most of the features found in the desktop tools.

The tools are:

1. Current Directory Browser.

2. Workspace Browser.

3. Array Editor.

4. Editor/Debugger.

5. Command Window.

6. Command History.

7. Launch Pad.

8. Help Browser

Command Window:

Use the Command Window to enter variables and run functions and M-

files.

Command History:

Lines you enter in the Command Window are logged in the Command

History window. In the Command History, you can view previously used

functions, and copy and execute selected lines. To save the input and output from

a MATLAB session to a file, use the diary function.

Running External Programs:

You can run external programs from the MATLAB Command Window.

The exclamation point character! is a shell escape and indicates that the rest of the

input line is a command to the operating system. This is useful for invoking

utilities or running other programs without quitting MATLAB. On Linux, for

example, emacsmagik.m invokes an editor called emacs for a file named

39

magik.m. When you quit the external program, the operating system returns

control to MATLAB.

Launch Pad:

MATLAB's Launch Pad provides easy access to tools, demos, and

documentation.

Help Browser:

Use the Help browser to search and view documentation for all your Math

Works products. The Help browser is a Web browser integrated into the

MATLAB desktop that displays HTML documents.

To open the Help browser, click the help button in the toolbar, or type help

browser in the Command Window. The Help browser consists of two panes, the

Help Navigator, which you use to find information, and the display pane, where

you view the information.

Help Navigator:

Use to Help Navigator to find information. It includes:

Product filter:

Set the filter to show documentation only for the products you specify.

Contents tab:

View the titles and tables of contents of documentation for your products.

Index tab:

Find specific index entries (selected keywords) in the MathWorks

documentation for your products.

40

Search tab:

Look for a specific phrase in the documentation. To get help for a specific

function, set the Search type to Function Name.

Favorites tab:

View a list of documents you previously designated as favorites.

Display Pane:

After finding documentation using the Help Navigator, view it in the

display pane. While viewing the documentation, you can:

Browse to other pages: Use the arrows at the tops and bottoms of the

pages, or use the back and forward buttons in the toolbar.

Bookmark pages: Click the Add to Favorites button in the toolbar.

Print pages: Click the print button in the toolbar.

Find a term in the page: Type a term in the Find in page field in the

toolbar and click Go.

Other features available in the display pane are: copying information,

evaluating a selection, and viewing Web pages.

Current Directory Browser:

MATLAB file operations use the current directory and the search path as

reference points. Any file you want to run must either be in the current directory

or on the search path.

Search Path:

To determine how to execute functions you call, MATLAB uses a search

path to find M-files and other MATLAB-related files, which are organized in

directories on your file system. Any file you want to run in MATLAB must reside

in the current directory or in a directory that is on the search path. By default, the

files supplied with MATLAB and MathWorks toolboxes are included in the

search path.

41

Workspace Browser:

The MATLAB workspace consists of the set of variables (named arrays)

built up during a MATLAB session and stored in memory. You add variables to

the workspace by using functions, running M-files, and loading saved

workspaces.

To view the workspace and information about each variable, use the

Workspace browser, or use the functions who and who’s.

To delete variables from the workspace, select the variable and select

Delete from the Edit menu. Alternatively, use the clear function.

The workspace is not maintained after you end the MATLAB session. To

save the workspace to a file that can be read during a later MATLAB session,

select Save Workspace As from the File menu, or use the save function. This

saves the workspace to a binary file called a MAT-file, which has a .mat

extension. There are options for saving to different formats. To read in a MAT-

file, select Import Data from the File menu, or use the load function.

Array Editor:

Double-click on a variable in the Workspace browser to see it in the Array

Editor. Use the Array Editor to view and edit a visual representation of one- or

two-dimensional numeric arrays, strings, and cell arrays of strings that are in the

workspace.

Editor/Debugger:

Use the Editor/Debugger to create and debug M-files, which are programs

you write to run MATLAB functions. The Editor/Debugger provides a graphical

user interface for basic text editing, as well as for M-file debugging.

You can use any text editor to create M-files, such as Emacs, and can use

preferences (accessible from the desktop File menu) to specify that editor as the

default. If you use another editor, you can still use the MATLAB Editor/Debugger

42

for debugging, or you can use debugging functions, such as dbstop, which sets a

breakpoint.

If you just need to view the contents of an M-file, you can display it in the

Command Window by using the type function.

5.3 MANIPULATING MATRICES:

5.3.1 Entering Matrices:

The best way for you to get started with MATLAB is to learn how to handle

matrices. Start MATLAB and follow along with each example.

You can enter matrices into MATLAB in several different ways:

Enter an explicit list of elements.

Load matrices from external data files.

Generate matrices using built-in functions.

Create matrices with your own functions in M-files.

Start by entering Dürer's matrix as a list of its elements. You have only to follow a

few basic conventions:

Separate the elements of a row with blanks or commas.

Use a semicolon (;) to indicate the end of each row.

Surround the entire list of elements with square brackets [ ].

To enter Durer's matrix, simply type in the Command Window

A = [16 3 2 13; 5 10 11 8; 9 6 7 12; 4 15 14 1]

MATLAB displays the matrix you just entered.

43

A =

16 3 2 13

5 10 11 8

9 6 7 12

4 15 14 1

This exactly matches the numbers in the engraving. Once you have entered the

matrix, it is automatically remembered in the MATLAB workspace. You can refer to it

simply as A.

5.3.2 Expressions:

Like most other programming languages, MATLAB provides mathematical

expressions, but unlike most programming languages, these expressions involve entire

matrices. The building blocks of expressions are:

Variables

Numbers

Operators

Function

5.3.3 Variables:

MATLAB does not require any type declarations or dimension statements. When

MATLAB encounters a new variable name, it automatically creates the variable and

44

allocates the appropriate amount of storage. If the variable already exists, MATLAB

changes its contents and, if necessary, allocates new storage. For example,

num_students = 25

Creates a 1-by-1 matrix named num_students and stores the value 25 in its single

element.

Variable names consist of a letter, followed by any number of letters, digits, or

underscores. MATLAB uses only the first 31 characters of a variable name. MATLAB is

case sensitive; it distinguishes between uppercase and lowercase letters. A and a are not

the same variable. To view the matrix assigned to any variable, simply enter the variable

name.

5.3.4 Numbers:

MATLAB uses conventional decimal notation, with an optional decimal point and

leading plus or minus sign, for numbers. Scientific notation uses the letter e to specify a

power-of-ten scale factor. Imaginary numbers use either i or j as a suffix. Some examples

of legal numbers are

3 -99 0.0001

9.6397238 1.60210e-20 6.02252e23

1i -3.14159j 3e5i

All numbers are stored internally using the long format specified by the IEEE

floating-point standard. Floating-point numbers have a finite precision of roughly 16

significant decimal digits and a finite range of roughly 10-308 to 10+308.

45

5.3.5 Operators:

Expressions use familiar arithmetic operators and precedence rules.

+ Addition

- Subtraction

* Multiplication

/ Division

\ Left division (described in "Matrices and Linear Algebra" in

Using MATLAB)

^ Power

' Complex conjugate transpose

( ) Specify evaluation order

Table 5.1 Arithmetic Operators and Precedence Rules.

5.3.6 Functions:

MATLAB provides a large number of standard elementary mathematical

functions, including abs, sqrt, exp, and sin. Taking the square root or logarithm of a

negative number is not an error; the appropriate complex result is produced

automatically. MATLAB also provides many more advanced mathematical functions,

including Bessel and gamma functions. Most of these functions accept complex

arguments. For a list of the elementary mathematical functions, type helpelfun. For a list

of more advanced mathematical and matrix functions, type helpspecfun or helpelmat.

Some of the functions, like sqrt and sin, are built-in. They are part of the

MATLAB core so they are very efficient, but the computational details are not readily

accessible. Other functions, like gamma and sinh, are implemented in M-files. You can

46

see the code and even modify it if you want. Several special functions provide values of

useful constants.

Pi 3.14159265...

I Imaginary unit, √-1

I Same as i

Eps Floating-point relative precision, 2-52

Realmin Smallest floating-point number, 2-1022

Realmax Largest floating-point number, (2-ε)21023

Inf Infinity

NaN Not-a-number

Table 5.2 Special Functions.

5.4 GRAPHICAL USER INTERFACE (GUI):

A graphical user interface (GUI) is a user interface built with graphical objects, such

as buttons, text fields, sliders, and menus. In general, these objects already have meanings

to most computer users. For example, when you move a slider, a value change; when you

press an OK button, your settings are applied and the dialog box is dismissed. Of course,

to leverage this built-in familiarity, you must be consistent in how you use the various

GUI-building components.

Applications that provide GUIs are generally easier to learn and use since the

person using the application does not need to know what commands are available or how

they work. The action that results from a particular user action can be made clear by the

design of the interface.

47

The sections that follow describe how to create GUIs with MATLAB. This

includes laying out the components, programming them to do specific things in response

to user actions, and saving and launching the GUI; in other words, the mechanics of

creating GUIs. This documentation does not attempt to cover the "art" of good user

interface design, which is an entire field unto itself. Topics covered in this section

include:

5.4.1 Creating GUIs with GUIDE:

MATLAB implements GUIs as Figure windows containing various styles of

uicontrol objects. You must program each object to perform the intended action when

activated by the user of the GUI. In addition, you must be able to save and launch your

GUI. All of these tasks are simplified by GUIDE, MATLAB's graphical user interface

development environment.

5.4.2 GUI Development Environment:

The process of implementing a GUI involves two basic tasks:

Laying out the GUI components

Programming the GUI components

GUIDE primarily is a set of layout tools. However, GUIDE also generates an M-file

that contains code to handle the initialization and launching of the GUI. This M-file

provides a framework for the implementation of the callbacks - the functions that execute

when users activate components in the GUI.

48

5.4.3 The Implementation of a GUI:

While it is possible to write an M-file that contains all the commands to lay out a

GUI, it is easier to use GUIDE to lay out the components interactively and to generate

two files that save and launch the GUI:

A FIGURE-file - contains a complete description of the GUI Figure and all of its

Children (uicontrols and axes), as well as the values of all object properties.

An M-file - contains the functions that launch and control the GUI and the

Callbacks, which are defined as subfunctions. This M-file is referred to as the

Application M-file in this documentation.

Note that the application M-file does not contain the code that lays out the

uicontrols; this information is saved in the FIGURE-file.

Figure 5.1 Diagram illustrating parts of a GUI implementation.

49

5.4.4 Features of the GUIDE-Generated Application M-File:

GUIDE simplifies the creation of GUI applications by automatically generating

an M-file framework directly from your layout. You can then use this framework to code

your application M-file. This approach provides a number of advantages:

The M-file contains code to implement a number of useful features (see

ConFigureuring Application Options for information on these features). The M-file

adopts an effective approach to managing object handles and executing callback routines

(see Creating and Storing the Object Handle Structure for more information). The M-files

provides a way to manage global data (see Managing GUI Data for more information).

The automatically inserted subfunction prototypes for callbacks ensure

compatibility with future releases. For more information, see Generating Callback

Function Prototypes for information on syntax and arguments.

You can elect to have GUIDE generate only the FIGURE-file and write the

application M-file yourself. Keep in mind that there are no uicontrol creation commands

in the application M-file; the layout information is contained in the FIGURE-file

generated by the Layout Editor.

5.4.5 Beginning the Implementation Process:

To begin implementing your GUI, proceed to the following sections:

Getting Started with GUIDE - the basics of using GUIDE:

Selecting GUIDE Application Options - set both FIGURE-file and M-file

options.

Using the Layout Editor - begin laying out the GUI.

Understanding the Application M-File - discussion of programming techniques

used in the application M-file.

Application Examples - a collection of examples that illustrate techniques which

are useful for implementing GUIs.

50

5.4.6 Command-Line Accessibility:

When MATLAB creates a graph, the Figure and axes are included in the list of

children of their respective parents and their handles are available through commands

such as findobj, set, and get. If you issue another plotting command, the output is directed

to the current Figure and axes.

GUIs are also created in Figure windows. Generally, you do not want GUI

Figures to be available as targets for graphics output, since issuing a plotting command

could direct the output to the GUI Figure, resulting in the graph appearing in the middle

of the GUI.

In contrast, if you create a GUI that contains an axes and you want commands

entered in the command window to display in this axes, you should enable command-line

access.

5.4.7 User Interface Controls:

The Layout Editor Component palette contains the user interface controls that you

can use in your GUI. These components are MATLAB uicontrol objects and are

programmable via their Callback properties. This section provides information on these

components.

Push Buttons

Sliders

Toggle Buttons

Frames

Radio Buttons

Listboxes

Checkboxes

51

Popup Menus

Edit Text

Axes

Static Text

Figures

5.5 PROGRAMMING WITH MATLAB:

Push Buttons:

Push buttons generate an action when pressed (e.g., an OK button may

close a dialog box and apply settings). When you click down on a push button, it

appears depressed; when you release the mouse, the button's appearance returns to

its nondepressed state; and its callback executes on the button up event.

Properties to Set:

String - set this property to the character string you want displayed on the push

button.

Tag - GUIDE uses the Tag property to name the callback subfunction in the

application M-file. Set Tag to a descriptive name (e.g., close_button) before

activating the GUI.

Programming the Callback:

When the user clicks on the push button, its callback executes. Push

buttons do not return a value or maintain a state.

52

The callback routine needs to query the toggle button to determine what

state it is in. MATLAB sets the Value property equal to the Max property when

the toggle button is depressed (Max is 1 by default) and equal to the Min property

when the toggle button is not depressed (Min is 0 by default).

Toggle Buttons:

Toggle buttons generate an action and indicate a binary state (e.g., on or

off). When you click on a toggle button, it appears depressed and remains

depressed when you release the mouse button, at which point the callback

executes. A subsequent mouse click returns the toggle button to the nondepressed

state and again executes its callback.

From the GUIDE Application M-File:

The following code illustrates how to program the callback in the GUIDE

application M-file.

functionvarargout = togglebutton1_Callback(h,eventdata,handles,varargin)

button_state = get (h,'Value');

ifbutton_state == get (h,'Max') % toggle button is pressed

elseifbutton_state == get (h,'Min') % toggle button is not pressed

end

Adding an Image to a Push Button or Toggle Button:

Assign the CData property an m-by-n-by-3 array of RGB values that

define a truecolor image. For example, the array a defines 16-by-128 truecolor

image using random values between 0 and 1 (generated by rand).

a (:,:,1) = rand(16,128);

a (:,:,2) = rand(16,128);

a (:,:,3) = rand(16,128);

set (h,'CData',a)

53

Radio Buttons:

Radio buttons are similar to checkboxes, but are intended to be mutually

exclusive within a group of related radio buttons (i.e., only one button is in a

selected state at any given time). To activate a radio button, click the mouse

button on the object. The display indicates the state of the button.

Implementing Mutually Exclusive Behavior:

Radio buttons have two states - selected and not selected. You can query

and set the state of a radio button through its Value property:

Value = Max, button is selected.

Value = Min, button is not selected.

To make radio buttons mutually exclusive within a group, the callback for

each radio button must set the Value property to 0 on all other radio buttons in the

group. MATLAB sets the Value property to 1 on the radio button clicked by the

user.

The following subfunction, when added to the application M-file, can be

called by each radio button callback. The argument is an array containing the

handles of all other radio buttons in the group that must be deselected.

functionmutual_exclude(off)

set (off,'Value',0)

Obtaining the Radio Button Handles:

The handles of the radio buttons are available from the handles structure,

which contains the handles of all components in the GUI. This structure is an

input argument to all radio button callbacks.

The following code shows the call to mutual_exclude being made from the

first radio button's callback in a group of four radio buttons.

functionvarargout = radiobutton1_Callback(h,eventdata,handles,varargin)

off = [handles.radiobutton2,handles.radiobutton3,handles.radiobutton4];

mutual_exclude(off) % Continue with callback.

54

After setting the radio buttons to the appropriate state, the callback can

continue with its implementation-specific tasks.

Checkboxes:

Check boxes generate an action when clicked and indicate their state as

checked or not checked. Check boxes are useful when providing the user with a

number of independent choices that set a mode (e.g., display a toolbar or generate

callback function prototypes).

The Value property indicates the state of the check box by taking on the

value of the Max or Min property (1 and 0 respectively by default):

Value = Max, box is checked.

Value = Min, box is not checked.

You can determine the current state of a check box from within its

callback by querying the state of its Value property, as illustrated in the following

example.

function checkbox1_Callback(h,eventdata,handles,varargin)

if (get(h,'Value') == get(h,'Max'))

% then checkbox is checked-take approriate action

Else

% checkbox is not checked-take approriate action

end

Edit Text:

Edit text controls are fields that enable users to enter or modify text

strings. Use edit text when you want text as input. The String property contains

the text entered by the user.

To obtain the string typed by the user, get the String property in the

callback.

function edittext1_Callback(h,eventdata, handles,varargin)

user_string = get(h,'string'); % proceed with callback...

55

Obtaining Numeric Data from an Edit Test Component:

MATLAB returns the value of the edit text String property as a character

string. If you want users to enter numeric values, you must convert the characters

to numbers. You can do this using the str2double command, which converts

strings to doubles. If the user enters non-numeric characters, str2double returns

NaN.

You can use the following code in the edit text callback. It gets the value

of the String property and converts it to a double. It then checks if the converted

value is NaN, indicating the user entered a non-numeric character (isnan) and

displays an error dialog (errordlg).

function edittext1_Callback(h,eventdata,handles,varargin)

user_entry = str2double(get(h,'string'));

ifisnan(user_entry)

errordlg('You must enter a numeric value','BadInput','modal')

end % proceed with callback...

Triggering Callback Execution:

On UNIX systems, clicking on the menubar of the Figure window causes

the edit text callback to execute. However, on Microsoft Windows systems, if an

editable text box has focus, clicking on the menubar does not cause the editable

text callback routine to execute. This behavior is consistent with the respective

platform conventions. Clicking on other components in the GUI execute the

callback.

Static Text:

Static text controls displays lines of text. Static text is typically used to

label other controls, provide directions to the user, or indicate values associated

with a slider. Users cannot change static text interactively and there is no way to

invoke the callback routine associated with it.

56

Frames:

Frames are boxes that enclose regions of a Figure window. Frames can

make a user interface easier to understand by visually grouping related controls.

Frames have no callback routines associated with them and only uicontrols can

appear within frames (axes cannot).

Placing Components on Top of Frames:

Frames are opaque. If you add a frame after adding components that you

want to be positioned within the frame, you need to bring forward those

components. Use the Bring to Front and Send to Back operations in the Layout

menu for this purpose.

List Boxes:

List boxes display a list of items and enable users to select one or more

items.

The String property contains the list of strings displayed in the list box.

The first item in the list has an index of 1.

The Value property contains the index into the list of strings that

correspond to the selected item. If the user selects multiple items, then Value is a

vector of indices.

By default, the first item in the list is highlighted when the list box is first

displayed. If you don’t want any item highlight then set the Value property to

empty, [].

The ListboxTop property defines which string in the list displays as the

top most items when the list box is not large enough to display all list entries.

ListboxTop is an index into the array of strings defined by the String property and

must have a value between 1 and the number of strings. Noninteger values are

fixed to the next lowest integer.

57

Single or Multiple Selections:

The values of the Min and Max properties determine whether users can

make single or multiple selections:

If Max - Min > 1, then list boxes allow multiple item selection.

If Max - Min <= 1, then list boxes do not allow multiple item selection.

Selection Type:

Listboxes differentiate between single and double clicks on an item and

set the Figure SelectionType property to normal or open accordingly. See

Triggering Callback Execution for information on how to program multiple

selections.

Triggering Callback Execution:

MATLAB evaluates the list box's callback after the mouse button is

released or a keypress event (including arrow keys) that changes the Value

property (i.e., any time the user clicks on an item, but not when clicking on the list

box scrollbar). This means the callback is executed after the first click of a

double-click on a single item or when the user is making multiple selections.

In these situations, you need to add another component, such as a Done

button (push button) and program its callback routine to query the list box Value

property (and possibly the Figure SelectionType property) instead of creating a

callback for the list box. If you are using the automatically generated application

M-file option, you need to either Set the list box Callback property to the empty

string ('') and remove the callback subfunction from the application M-file. Leave

the callback subfunction stub in the application M-file so that no code executes

when users click on list box items.

The first choice is best if you are sure you will not use the list box callback

and you want to minimize the size and efficiency of the application M-file.

However, if you think you may want to define a callback for the list box at some

time, it is simpler to leave the callback stub in the M-file.

58

Popup Menus:

Popup menus open to display a list of choices when users press the arrow.

The String property contains the list of string displayed in the popup menu. The

Value property contains the index into the list of strings that correspond to the

selected item. When not open, a popup menu displays the current choice, which is

determined by the index contained in the Value property. The first item in the list

has an index of 1.

Popup menus are useful when you want to provide users with a number of

mutually exclusive choices, but do not want to take up the amount of space that a

series of radio buttons requires.

Programming the Popup Menu:

You can program the popup menu callback to work by checking only the

index of the item selected (contained in the Value property) or you can obtain the

actual string contained in the selected item.

This callback checks the index of the selected item and uses a switch

statement to take action based on the value. If the contents of popup menu are

fixed, then you can use this approach.

functionvarargout = popupmenu1_Callback(h,eventdata,handles,varargin)

val = get(h,'Value');

switchval

case 1 % The user selected the first item

case 2 % The user selected the second item, etc.

This callback obtains the actual string selected in the popup menu. It uses

the value to index into the list of strings. This approach may be useful if your

program dynamically loads the contents of the popup menu based on user action

and you need to obtain the selected string. Note that it is necessary to convert the

value returned by the String property from a cell array to a string.

functionvarargout = popupmenu1_Callback(h,eventdata,handles,varargin)

val = get(h,'Value');

string_list = get(h,'String');

59

selected_string = string_list{val}; % convert from cell array to string, etc.

Enabling or Disabling Controls:

You can control whether a control responds to mouse button clicks by

setting the Enable property. Controls have three states:

On - The control is operational

Off - The control is disabled and its label (set by the string property) is

grayed out.

Inactive - The control is disabled, but its label is not grayed out.

When a control is disabled, clicking on it with the left mouse button does

not execute its callback routine. However, the left-click causes two other callback

routines to execute:

First the Figure WindowButtonDownFcn callback executes. Then the

control's ButtonDownFcn callback executes.

A right mouse button click on a disabled control posts a context menu, if

one is defined for that control. See the Enable property description for more

details.

Axes:

Axes enable your GUI to display graphics (e.g., graphs and images). Like

all graphics objects, axes have properties that you can set to control many aspects

of its behavior and appearance. See Axes Properties for general information on

axes objects.

Axes Callbacks:

Axes are not uicontrol objects, but can be programmed to execute a

callback when users click a mouse button in the axes. Use the axes

ButtonDownFcn property to define the callback.

60

Plotting to Axes in GUIs:

GUIs that contain axes should ensure the Command-line accessibility

option in the Application Options dialog is set to Callback (the default). This

enables you to issue plotting commands from callbacks without explicitly

specifying the target axes.

GUIs with Multiple Axes:

If a GUI has multiple axes, you should explicitly specify which axes you

want to target when you issue plotting commands. You can do this using the axes

command and the handles structure. For example, ‘axes(handles.axes1)’ makes

the axes whose Tag property is axes1 the current axes, and therefore the target for

plotting commands. You can switch the current axes whenever you want to target

different axes. See GUI with Multiple Axes for an example that uses two axes.

Figure:

Figures are the windows that contain the GUI you design with the Layout

Editor. See the description of Figure properties for information on what Figure

characteristics you can control.

61

CHAPTER 7

RESULTS AND CONCLUSION

7.1 RESULTS:

The results of the project are shown in the following figures:

1. Initialize graphic user interface.

Figure 7.1 Project graphic user interface.

62

2. Click on query button and select an unknown face image.

Figure 7.2 Project graphic user interface.

3. Unknown face image loaded into axes1.

Figure 7.3 Project graphic user interface.

63

4. Click on training button and Eigen faces of database images will be calculated.

Figure 7.4 Project graphic user interface.

5. Click on Feature extraction button and unknown face image features gets

extracted.

Figure 7.5 Project graphic user interface.

64

6. Click on query button and select an unknown palm image.

Figure 7.6 Project graphic user interface.

7. Click on feature extraction button and Palm features get extracted.

Figure 7.7 Project graphic user interface.

65

8. Click on feature matching button and Euclidian distance between unknown

images and database images get calculated. If Euclidian distance is less than

threshold value then the person is genuine and Authentic is displayed on GUI as

shown below.

Figure 7.8 Project graphic user interface.

9. If the Euclidian distance is greater than the threshold value then the person is not

genuine and Not Authentic is displayed on GUI as shown below.

Figure 7.9 Project graphic user interface.

66

7.2 CONCLUSION:

The project “FACE RECOGNITION AND PALMPRINT IDENTIFICATION

FOR AUTHENTICATION PROCESS” has been successfully designed and simulated

using Matlab software.

7.3 FUTURE SCOPE OF THE PROJECT:

This project can be applied in many areas. Some of them are given as follows:

Law enforcement application: Face and palm recognition technology is primarily

used in law enforcement applications, especially mug shot albums (static

matching) and video surveillance (real-time matching by video image sequences).

In transactional authentication: static matching of photographs on credit cards,

ATM cards and photo ID to real-time.

In document control: digital chip in passports and driver’s licenses.

In computer security: user access verification.

In physical access control: smart doors.

In voter registration: election accuracy.

In time and attendance: entry and exit information.

In computer games: a virtual “you” plays against virtual opponents.

Also this project can be interfaced with other biometric system to provide more

authentications.

67

References

[1] M. Turk, A. Pentland, Eigenfaces for Recognition, Journal of Cognitive

Neurosicence, Vol. 3, No. 1, 1991, pp. 71-86.

[2] M.A. Turk, A.P. Pentland, Face Recognition Using Eigenfaces, Proceedings of the IEEE

Conference on Computer Vision and Pattern Recognition, 3-6 June 1991, Maui, Hawaii, USA,

pp. 586-591.

[3] A. Pentland, B. Moghaddam, T. Starner, View-Based and Modular Eigenspaces for Face

Recognition, Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,

21-23 June 1994, Seattle, Washington, USA, pp. 84-91.

[4] H. Moon, P.J. Phillips, Computational and Performance aspects of PCA-based Face

Recognition Algorithms, Perception, Vol. 30, 2001, pp. 303-321.

[5] D. Swets, J. Weng, "Using Discriminant Eigenfeatures for Image Retrieval ", IEEE

Transactions on Pattern Analysis and Machine Intelligence, 18(8), pp. 831-836, 1996.

[6] Ilker Atalay, “Face recognition using eigenfaces,” M.Sc. Thesis, ISTANBUL Technical

University, Institute of science and technology, January, 1996.

[7] David Zhang, Palmprint Authentication, Kluwer Academic Publishers: Boston, 2004,

pp,195.

[8] Edward Wong Kie Yih, G. Sainarayanan, Ali Chekima, Narendra G, Palmprint

Identification Using Sequential Modified Haar Wavelet Energy, IEE-International

Conference and Signal processing, Communications and Networking, Madras Institute of

technology, Anna University Chennai India, Jan 4-6, 2008. Pp411-416.

[9] Website: www.mathworks.com.

68

APPENDIX

CODE:

function varargout = gaitrec(varargin)

% GAITREC M-file for gaitrec.Figure

% GAITREC, by itself, creates a new GAITREC or raises the existing

% singleton*.

%

% H = GAITREC returns the handle to a new GAITREC or the handle to

% the existing singleton*.

%

% GAITREC('CALLBACK',hObject,eventData,handles,...) calls the local

% function named CALLBACK in GAITREC.M with the given input arguments.

%

% GAITREC('Property','Value',...) creates a new GAITREC or raises the

% existing singleton*. Starting from the left, property value pairs are

% applied to the GUI before gaitrec_OpeningFunction gets called. An

% unrecognized property name or invalid value makes property application

% stop. All inputs are passed to gaitrec_OpeningFcn via varargin.

%

% *See GUI Options on GUIDE's Tools menu. Choose "GUI allows only one

% instance to run (singleton)".

%

% See also: GUIDE, GUIDATA, GUIHANDLES

% Edit the above text to modify the response to help gaitrec

% Last Modified by GUIDE v2.5 13-Apr-2010 08:20:57

% Begin initialization code - DO NOT EDIT

69

gui_Singleton = 1;

gui_State = struct('gui_Name', mfilename, ...

'gui_Singleton', gui_Singleton, ...

'gui_OpeningFcn', @gaitrec_OpeningFcn, ...

'gui_OutputFcn', @gaitrec_OutputFcn, ...

'gui_LayoutFcn', [] , ...

'gui_Callback', []);

if nargin && ischar(varargin{1})

gui_State.gui_Callback = str2func(varargin{1});

end

if nargout

[varargout{1:nargout}] = gui_mainfcn(gui_State, varargin{:});

else

gui_mainfcn(gui_State, varargin{:});

end

% End initialization code - DO NOT EDIT

% --- Executes just before gaitrec is made visible.

function gaitrec_OpeningFcn(hObject, eventdata, handles, varargin)

% This function has no output args, see OutputFcn.

% hObject handle to Figure

% eventdata reserved - to be defined in a future version of MATLAB

% handles structure with handles and user data (see GUIDATA)

% varargin command line arguments to gaitrec (see VARARGIN)

% Choose default command line output for gaitrec

handles.output = hObject;

a = ones(256,256);

axes(handles.axes1);

70

imshow(a);

axes(handles.axes2);

imshow(a);

% % % b = ones(64,150);

axes(handles.axes3);

imshow(a);

% Update handles structure

guidata(hObject, handles);

% UIWAIT makes gaitrec wait for user response (see UIRESUME)

% uiwait(handles.Figure1);

% --- Outputs from this function are returned to the command line.

function varargout = gaitrec_OutputFcn(hObject, eventdata, handles)

% varargout cell array for returning output args (see VARARGOUT);

% hObject handle to Figure

% eventdata reserved - to be defined in a future version of MATLAB

% handles structure with handles and user data (see GUIDATA)

% Get default command line output from handles structure

varargout{1} = handles.output;

% --- Executes on button press in query.

function query_Callback(hObject, eventdata, handles)

% hObject handle to query (see GCBO)

% eventdata reserved - to be defined in a future version of MATLAB

% handles structure with handles

[filename, pathname] = uigetfile('*.tif', 'Pick an M-file');

if isequal(filename,0) || isequal(pathname,0)

disp('User pressed cancel')

else

disp(['User selected ', fullfile(pathname, filename)])

end

71

inp=imread(fullfile(pathname,filename));

[r c p]=size(inp);

if p==3

inp=rgb2gray(inp);

end

axes(handles.axes1);

imshow(inp);

handles.inp = inp;

handles.filename = filename;

% Update handles structure

guidata(hObject, handles);

% --- Executes on button press in training.

function trainin_Callback(hObject, eventdata, handles)

% hObject handle to trainin (see GCBO)

% eventdata reserved - to be defined in a future version of MATLAB

% handles structure with handles and user data (see GUIDATA)

T = [];

for i = 1 : 20

% I have chosen the name of each image in databases as a corresponding

% number. However, it is not mandatory!

str = int2str(i);

str = strcat(str,'.jpg');

img = imread(str);

img = rgb2gray(img);

[irow icol] = size(img);

temp = reshape(img',irow*icol,1); % Reshaping 2D images into 1D image vectors

T = [T temp]; % 'T' grows after each turn

end

[m, A, Eigenfaces] = EigenfaceCore(T);

handles.m = m;

72

handles.A = A;

handles.Eigenfaces = Eigenfaces;

warndlg('Process Completed');

% Update handles structure

guidata(hObject, handles);

% --- Executes on button press in extratfeature.

function extratfeature_Callback(hObject, eventdata, handles)

% hObject handle to extratfeature (see GCBO)

% eventdata reserved - to be defined in a future version of MATLAB

% handles structure with handles and user data (see GUIDATA)

m = handles.m;

A = handles.A;

inp = handles.inp;

filename = handles.filename;

Eigenfaces = handles.Eigenfaces;

[OutputName h Recognized_index MLresult] = Recognition(filename, m, A, Eigenfaces,

inp);

handles.Recognized_index = Recognized_index;

handles.OutputName = OutputName;

handles.MLresult = MLresult;

handles.h = h;

warndlg('Process Completed');

% Update handles structure

guidata(hObject, handles);

% --- Executes on button press in palminput.

function palminput_Callback(hObject, eventdata, handles)

% hObject handle to palminput (see GCBO)

% eventdata reserved - to be defined in a future version of MATLAB

% handles structure with handles and user data (see GUIDATA)

73

[filename path2] = uigetfile('*.bmp','Pick an Image File');

if filename==0

warndlg('User Pressed Cancel');

else

a = imread(filename);

axes(handles.axes2);

imshow(a);

handles.pamin = a;

handles.filename = filename;

end

% Update handles structure

guidata(hObject, handles);

% --- Executes on button press in featureextract.

function featureextract_Callback(hObject, eventdata, handles)

% hObject handle to featureextract (see GCBO)

% eventdata reserved - to be defined in a future version of MATLAB

% handles structure with handles and user data (see GUIDATA)

% % % sharpened = handles.pamin;

filename = handles.filename;

len = length(filename);

newfile = filename(1:len-4);

[stat,mess]=fileattrib([newfile '.mat']);

if stat == 1

load (strcat(newfile,'.mat'))

handles.queryFV1 = queryFV1;

handles.queryFV_comb1 = queryFV_comb1;

else

input = imread(filename);

[queryFV, queryFV_comb] = Stransform(input);

74

queryFV1 = queryFV;

queryFV_comb1 = queryFV_comb;

handles.queryFV1 = queryFV1;

handles.queryFV_comb1 = queryFV_comb1;

save(newfile, 'queryFV1','queryFV_comb1');

end

guidata(hObject, handles);

warndlg('Feature Extraction is Completed');

% --- Executes on button press in featurematch.

function featurematch_Callback(hObject, eventdata, handles)

% hObject handle to featurematch (see GCBO)

% eventdata reserved - to be defined in a future version of MATLAB

% handles structure with handles and user data (see GUIDATA)

OutputName = handles.OutputName;

h = handles.h;

Eu=(h/1e+15);

MLresult = handles.MLresult;

inp = handles.inp;

Recognized_index = handles.Recognized_index;

if Recognized_index<11

indresult = 1;

else

indresult = 0;

end

filename = handles.filename ;

queryFV = handles.queryFV1 ;

if(filename=='rt1.bmp')

ind1=1;

else

ind1=0;

75

end

queryFV_comb = handles.queryFV_comb1;

file = dir(fullfile(cd,'*.bmp'));

len_file = length(file);

for i=1:len_file

imagename = file(i).name;

dist(i).file = imagename;

len = length(imagename);

newfile = imagename(1:len-4);

[stat,mess]=fileattrib([newfile '.mat']);

if stat == 1

load (strcat(newfile,'.mat'))

databaseFV = queryFV1;

databaseFV_comb = queryFV_comb1;

else

database_image = imread(imagename);

[databaseFV1,databaseFV_comb1] = Stransform(database_image);

databaseFV = databaseFV1;

databaseFV_comb = databaseFV_comb1;

queryFV1 = databaseFV;

queryFV_comb1 = databaseFV_comb;

save(newfile, 'queryFV1','queryFV_comb1');

end

[distanceFv, distanceFv_comb] = Euclidean(queryFV, databaseFV,

queryFV_comb, databaseFV_comb);

Euclidean_Dist_Fv(i) = distanceFv; %#ok<AGROW>

d = Euclidean_Dist_Fv(i);

Euclidean_Dist_Fv_comb(i) = distanceFv_comb;

end

[value ind] = sort(Euclidean_Dist_Fv/(10^3));

rec_imag = dist(ind(1)).file;

76

if MLresult>1

if value(1)<=0.565;

if (ind1==1)&&(MLresult>1)

set(handles.text2,'string','Authenticated');

axes(handles.axes3);

imshow(inp);

else

set(handles.text2,'string','Not Authenticated');

end

else

set(handles.text2,'string','Not Authenticated');

end

else

set(handles.text2,'string','Not Authenticated');

end

guidata(hObject, handles);

% --- Executes on button press in clear.

function clear_Callback(hObject, eventdata, handles)

% hObject handle to clear (see GCBO)

% eventdata reserved - to be defined in a future version of MATLAB

% handles structure with handles and user data (see GUIDATA)

a = ones(256,256);

axes(handles.axes1);

imshow(a);

axes(handles.axes2);

imshow(a);

axes(handles.axes3);

imshow(a);

set(handles.text2,'string',' ');

77

EigenfaceCore.m:

function [m, A, Eigenfaces] = EigenfaceCore(T)

% Use Principle Component Analysis (PCA) to determine the most

% discriminating features between images of faces.

%

% Description: This function gets a 2D matrix, containing all training image vectors

% and returns 3 outputs which are extracted from training database.

%

% Argument: T - A 2D matrix, containing all 1D image vectors.

% Suppose all P images in the training database

% have the same size of MxN. So the length of 1D

% column vectors is M*N and 'T' will be a MNxP 2D matrix.

%

% Returns: m - (M*Nx1) Mean of the training database

% Eigenfaces - (M*Nx(P-1)) Eigen vectors of the covariance matrix of

the training database

% A - (M*NxP) Matrix of centered image vectors

%

% See also: EIG

%%%%%%%%%%%%%%%%%%%%%%%% Calculating the mean image

m = mean(T,2); % Computing the average face image m = (1/P)*sum(Tj's) (j = 1 : P)

Train_Number = size(T,2);

%%%%%%%%%%%%%%%%%%%%%%%% Calculating the deviation of each

image from mean image

A = [];

for i = 1 : Train_Number

temp = double(T(:,i)) - m; % Computing the difference image for each image in the

training set Ai = Ti - m

78

A = [A temp]; % Merging all centered images

end

%%%%%%%%%%%%%%%%%%%%%%%% Snapshot method of Eigenface methos

% We know from linear algebra theory that for a PxQ matrix, the maximum

% number of non-zero eigenvalues that the matrix can have is min(P-1,Q-1).

% Since the number of training images (P) is usually less than the number

% of pixels (M*N), the most non-zero eigenvalues that can be found are equal

% to P-1. So we can calculate eigenvalues of A'*A (a PxP matrix) instead of

% A*A' (a M*NxM*N matrix). It is clear that the dimensions of A*A' is much

% larger that A'*A. So the dimensionality will decrease.

L = A'*A; % L is the surrogate of covariance matrix C=A*A'.

[V D] = eig(L); % Diagonal elements of D are the eigenvalues for both L=A'*A and

C=A*A'.

%%%%%%%%%%%%%%%%%%%%%%%% Sorting and eliminating eigenvalues

% All eigenvalues of matrix L are sorted and those who are less than a

% specified threshold, are eliminated. So the number of non-zero

% eigenvectors may be less than (P-1).

L_eig_vec = [];

for i = 1 : size(V,2)

if( D(i,i)>1 )

L_eig_vec = [L_eig_vec V(:,i)];

end

end

%%%%%%%%%%%%%%%%%%%%%%%% Calculating the eigenvectors of

covariance matrix 'C'

% Eigenvectors of covariance matrix C (or so-called "Eigenfaces")

79

% can be recovered from L's eiegnvectors.

Eigenfaces = A * L_eig_vec; % A: centered image vectors

Euclidean.m:function [distanceFv, distanceFv_comb] = Euclidean(queryFV, databaseFV,

queryFV_comb, databaseFV_comb)

len2 = size(queryFV_comb);

k=1;

for j=1:6

for i =1:len1(2)

dist(k) = (queryFV(j,i)- databaseFV(j,i)).^2;

k =k+1;

end

end

ii=1;

len3 = length(dist);

for I =1:len3

if dist(I)>=0

Dist(ii) = dist(I);

ii=ii+1;

end

end

distanceFv = sqrt(sum(Dist));

%-----------------

l=1;

for J=1:5

for i =1:len2(2)

dist1(l) = (queryFV_comb(J,i)- databaseFV_comb(J,i)).^2;

l =l+1;

end

80

end

ii=1;

len4 = length(dist1);

for JJ =1:len4

if dist1(JJ)>=0

Dist1(ii) = dist1(JJ);

ii=ii+1;

end

end

if ii==1

Dist1(1)=0;

end

distanceFv_comb = sqrt(sum(Dist1));

Recognition.m:

function [OutputName Euc_dist_min Recognized_index MLresult] =

Recognition(TestImage, m, A, Eigenfaces,a)

% Recognizing step....

%

% Description: This function compares two faces by projecting the images into facespace

and

% measuring the Euclidean distance between them.

%

% Argument: TestImage - Path of the input test image

%

% m - (M*Nx1) Mean of the training

% database, which is output of 'EigenfaceCore' function.

%

% Eigenfaces - (M*Nx(P-1)) Eigen vectors of the

81

% covariance matrix of the training

% database, which is output of 'EigenfaceCore' function.

%

% A - (M*NxP) Matrix of centered image

% vectors, which is output of 'EigenfaceCore' function.

%

% Returns: OutputName - Name of the recognized image in the training

database.

%

% See also: RESHAPE, STRCAT

%%%%%%%%%%%%%%%%%%%%%%%% Projecting centered image vectors into

facespace

% All centered images are projected into facespace by multiplying in

% Eigenface basis's. Projected vector of each face will be its corresponding

% feature vector.

ProjectedImages = [];

Train_Number = size(Eigenfaces,2);

for i = 1 : Train_Number

temp = Eigenfaces'*A(:,i); % Projection of centered images into facespace

ProjectedImages = [ProjectedImages temp];

end

%%%%%%%%%%%%%%%%%%%%%%%% Extracting the PCA features from test

image

InputImage = imread(TestImage);

InputImage=imresize(InputImage,[200 180]);

temp = InputImage(:,:,1);

[irow icol] = size(temp);

82

InImage = reshape(temp',irow*icol,1);

Difference = double(InImage)-m; % Centered test image

ProjectedTestImage = Eigenfaces'*Difference; % Test image feature vector

%%%%%%%%%%%%%%%%%%%%%%%% Calculating Euclidean distances

% Euclidean distances between the projected test image and the projection

% of all centered training images are calculated. Test image is

% supposed to have minimum distance with its corresponding image in the

% training database.

[MLresult] = finding(a);

Euc_dist = [];

for i = 1 : Train_Number

q = ProjectedImages(:,i);

temp = ( norm( ProjectedTestImage - q ) )^2;

Euc_dist = [Euc_dist temp];

end

[Euc_dist_min , Recognized_index] = min(Euc_dist);

OutputName = strcat(int2str(Recognized_index),'.jpg');

Stransform.m:function [FV, FV_comb] = Stransform(rt)

%rt = imread('img_0286.jpg');

bm = imresize(rt,[256 256]);

bm = double(bm);

% bm = rgb2gray(bm1);

% bm=rt;

for i = 1:6

[A H V D] = sequential_Haar(bm);

% [A H V D] = dwt2(bm,'haar');

k=1;

83

% temp(i) = sum(sum(A));

[row col] = size(H);

for ii = 1:row/4:row

for jj = 1:col/4:col

blockH(:,:,k) = double(H((ii:row/4-1+ii),(jj:col/4-1+jj)))

blockV(:,:,k) = double(V((ii:row/4-1+ii),(jj:col/4-1+jj)));

blockD(:,:,k) = double(D((ii:row/4-1+ii),(jj:col/4-1+jj)));

k = k+1;

end

end

% The MHE energy feature for every detail coefficients are

for I = 1:16 % MHEi,j = [MHEi,j,1,...MHEi,j,16];---------(3)

mheH(i,I) =double( sum(sum(blockH(:,:,I).^2)));

mheV(i,I) = double(sum(sum(blockV(:,:,I).^2)));

mheD(i,I) =double( sum(sum(blockD(:,:,I).^2)));

end

% DetailHVD -> combination of H V D co-effs. in level-------(4)

HDetailHVD(i,:) = double(mheH(i,:));

VDetailHVD(i,:) = double(mheV(i,:));

DDetailHVD(i,:) = double(mheD(i,:));

DetailHVD_sum = double((sum(HDetailHVD(i,:))+sum(VDetailHVD(i,:))

+sum(DDetailHVD(i,:))));

% the detail HVD in diffrnt decomposition levels are normalized(5)

% tempH = (HDetailHVD(i,:)./DetailHVD_sum);

if DetailHVD_sum

HD(i,(1:16)) = HDetailHVD(i,:)./DetailHVD_sum;

VD(i,(1:16)) = VDetailHVD(i,:)./DetailHVD_sum;

84

DD(i,(1:16)) = DDetailHVD(i,:)./DetailHVD_sum;

else

HD(i,(1:16)) = HDetailHVD(i,:);

VD(i,(1:16)) = VDetailHVD(i,:);

DD(i,(1:16)) = DDetailHVD(i,:);

end

FV(i,:)=[double(HD(i,:)) double(VD(i,:)) double(DD(i,:))];

FV(i,:) = double(FV(i,:));

switch i

case 1

A1 = A;

H1 = H;

V1 = V;

D1 = D;

case 2

A2 = A;

H2 = H;

V2 = V;

D2 = D;

case 3

A3 = A;

H3 = H;

V3 = V;

D3 = D;

case 4

A4 = A;

H4 = H;

V4 = V;

D4 = D;

85

case 5

A5 = A;

H5 = H;

V5 = V;

D5 = D;

case 6

A6 = A;

H6 = H;

V6 = V;

D6 = D;

end

bm = A;

blockH =[];

blockV =[];

blockD =[];

end

%

% level6 = [A6,H6;V6,D6];

% level5 = [level6,H5;V5,D5];

level4 = [A4,H4;V4,D4];

level3 = [level4,H3;V3,D3];

level2 = [level3,H2;V2,D2];

level1 = [level2,H1;V1,D1];

FV_comb(1,:)=(FV(1,:)+FV(2,:));

FV_comb(2,:)=(FV(1,:)+FV(2,:)+FV(3,:));

FV_comb(3,:)=(FV(1,:)+FV(2,:)+FV(3,:)+FV(4,:));

FV_comb(4,:)=(FV(1,:)+FV(2,:)+FV(3,:)+FV(4,:)+FV(5,:));

86

FV_comb(5,:)=(FV(1,:)+FV(2,:)+FV(3,:)+FV(4,:)+FV(5,:)+FV(6,:));

% figure;imshow(level1,[]);

% tem = [A1,H1;V1,D1];

% figure;imshow(tem,[]);

function[A H V D] = sequential_Haar(fim1)

% if level==1

% temp=fim1;

% return;

% end

[ r c ] = size(fim1);

% even = zeros(r,(c/2));

%first level decomposition

%one even dimension

for j = 1:1:r

a = 2;

for k = 1:1:(c/2)

Yeven(j,k) = fim1(j,a);

a = a+2;

end

end

%one odd dim

odd = zeros(r,(c/2));

for j = 1:1:r

a = 1;

for k = 1:1:(c/2)

Yodd(j,k) = fim1(j,a);

a = a+2;

end

end

[ lenr lenc ] = size(Yodd) ;

87

%one dim haar

for j = 1:1:lenr

for k = 1:1:lenc

Ylow(j,k) = Yodd(j,k)-Yeven(j,k);

Yhigh(j,k) = round((Yeven(j,k)+Yodd(j,k))/2);

end

end

%2nd dimension

[len2r len2c ] = size(Ylow);

for j = 1:1:(len2c)

a=2;

for k = 1:1:(len2r/2)

%even separation of one dim

Yloweven(k,j) = Ylow(a,j);

Yhgheven(k,j) = Yhigh(a,j);

a=a+2;

end

end

%odd separtion of one dim

for j = 1:1:(len2c)

a=1;

for k =1:1:(len2r/2)

Ylowodd(k,j) = Ylow(a,j);

Yhghodd(k,j) = Yhigh(a,j);

a = a+2;

end

end

%2d haar

[ len12r len12c ] = size(Ylowodd) ;

for j = 1:1:len12r

for k = 1:1:len12c

88

%2nd level hh

A(j,k) = round((Yhgheven(j,k)+Yhghodd(j,k))/2);

%2nd level hl

H(j,k) = Yhgheven(j,k)-Yhghodd(j,k);

%2nd level lh

V(j,k) = round((Ylowodd(j,k)+Yloweven(j,k))/2);

%2nd level ll

D(j,k) = Yloweven(j,k) - Ylowodd(j,k);

end

end

% level=level-1;

% figure;imshow(A,[]);

% figure;imshow(H,[]);

% figure;imshow(V,[]);

% figure;imshow(D,[]);

temp = [A,H];

temp1=[V,D];

out = [temp;temp1];

return;

89