Biometric Tech.

ABSTRACT Biometrics technique such as iris pattern, facial characteristic, fingerprints and other biometrics for identification is becoming more popular and common every day. Based on a characteristics and unique feature showed by individual, biometrics provides special and

automatics

identification

of

an

individual.

Recently,

iris

recognition has received increasing attention due to its high reliability. Iris recognition aims to identify persons using iris characteristics of human eyes. The iris recognition system consists of four processes which are image acquisition, pre-processing, feature extraction and identification. In order to identify both the uniqueness of the human iris and also its performance as a biometrics, this project developed an iris recognition system. The system two databases of digitized greyscale eye images were use to determine the recognition performance. Using MATLAB with its Acquisition Toolbox, we have developed an iris recognition system that allow user to read and identify persons by their iris. This system can easily capture biometrics and store them with information. Hough Transform is able to served an automatic segmentation system and able to localise the circular iris and pupil region, occluding eyelids and eyelashes, and reflections. For imaging inconsistencies, the extracted iris region was then normalised into rectangular block with constant dimensions to account. Finally the image code is calculated by using Gabor filter method.

CHAPTER 1

INTRODUCTION

1.1 Background 1.1.1

Biometric Technology Biometric system provided an automatic recognition based on some sort of

unique feature or individual characteristics. Biometric system developed based on finger print, handwriting, facial features, voice and iris. The system will then captured the biometrics sample such as digital sound signal for voice recognition or digital iris image for iris recognition and create a new template. This template is a transformation from sample by using some sort of mathematical function. This biometrics template will provide an efficient, normalised and highly discriminating representation of the feature. The template is compared with other templates in order to determine identity. This processes created during enrolment and if the templates match within a set tolerance, access to the desired resources is granted. The minimal chances of any two peoples having the same characteristic showed a good biometric. There are some requirements in order to provide convenience to the user and prevent misrepresentation of the feature which are uniqueness so that there would not be any identical appear in two person, universal which means it can occurs as many people as possible, measurable and user friendly. 1.1.2

The Human Iris

The iris lies between the cornea and the lens of the human eyes. The iris image is shown in Figure 1.1. The iris is perforated close to its center by a circular aperture. This is known as pupil. The iris function is to control the amount of light entering through pupil. This is done by the dilator muscles and sphincter, which adjust the size of pupil. The pupil size can vary from 10% to 80% of the iris diameter. The iris diameter is 12mm.

Figure 1.1 – A front-on view of the human eye.

The human iris consisted of three boundaries. The boundaries are shown in Figure 1.2. The first boundary is inner boundary which lies between the pupil and the iris. The pupil area has a low grey level and looks dark in the eyes image. The second boundary is an outer boundary. The outer boundary is between the iris and the sclera and the last boundary is the collarette boundary. A collarette boundary can be discriminated by adjacent pixels of variety. The papillary zone is vastly various and the ciliary area is seldom various. An inner boundary and outer boundary is divided by the collarette boundary.

Figure 1.2: Image of the boundaries and the localization iris area 1.1.3

Iris Recognition As a physiological biometric, iris recognition aims to identify persons using

iris characteristics of human eyes. Recently, iris recognition has received increasing attention due to its high reliability. Iris recognition combines pattern recognition, computer vision and statistics. It can serve as a living password that one need not to remember but always carries along. Iris recognition can be done by image processing. Image processing technique can be used to extract the iris pattern from a digitised image of the eyes. From the mathematical function, these biometrics templates can represent the unique information of an iris. This allows comparison to be made between templates. To use the iris recognition system, the subject eye is first photographed and then a template is created for their iris region. The template is then compares against all users in the system. If the template matching is found, the user is identified and access is granted. If no match is found, the subject remains unidentified.

1.2 Objectives The purposes of this project are: i.

Implement an iris recognition system using MATLAB

ii.

Increase the quality of security system by recognizing the iris patterns.

iii.

To verify the claimed performance of the technology.

1.3 Scope of Project The scope of project is to develop a system that can read an iris pattern and identify individual through the comparison with other template stored in database. MATLAB was a language choice due to its ease of use resulting in rapid project turnaround times. Additional scope that required is to learn the biological characteristics of the iris. Other scope is the process of developing the software that related on how to process the data and image techniques. 1.4 Expected Result By the end of this project, this system supposed can read iris pattern using MATLAB and compared with other template stored in a database until a template matching is found. 1.5 Statement of Problems In today’s world, system that used password and user login name is a requirement for each individual. The main issue is there is no guarantee that the system is secure and cannot easily crack. Problem occurred when user forgetting password, sharing password, using common password or stolen password. These techniques for avoiding password security obviously negate the purpose of the password. They also have hidden complications in terms of some accidental damage brought about by sharing level access and password. This system might be slightly complicated even though is practical because it is hardly to hack.

PSM REPORT 1 Prepared by Nurul Akmar Binti Mohamad Ramli On the week fourth, we decided to develop a system that can read iris pattern using C++ builder. The main output that we expect is the system can showed us type of diseased. The process of recognize the disease using iris pattern is known as irridiology. First, we are used any image such as face instead of iris to test the system. Problem occurred when our program cannot read any type of format image except bitmap. We try to manipulate the program and we successfully create a program that can read in any format image such as JPEG. We successfully invert the image, find the histogram and the threshold and also filter the image.

PSM REPORT 1 Prepared by Nurul Akmar Binti Mohamad Ramli On the week fifth and six, problem occurred when the program cannot compile the image more than 8 bit. We try to convert the 24 bit image to 8 bit but the quality of the image became poor. We also try to make a system that can read 8bit x 8bit x 8bit, but the final result is also same. We try to use the MATLAB software and we successfully read the image. We try to figure out how to transform the MATLAB source code to C++ builder. We make some research on internet and book references. We also try to find the journal and book related with iridiology and our project which is image processing using C++ builder.

PSM REPORT 1 Prepared by Nurul Akmar Binti Mohamad Ramli On the week seven, because the program using C++ builder cannot read the image as what we expected, we decide to change our software. We choose to use the MATLAB. We also change our expected result from the identify type of disease to identify the identity of individual. We also finished the chapter 1 of the project which is the introduction. For the program source code, it is still in process.