Ammad Fyp

FINGERPRINT RECOGNITION AND PASSWORD SECURITY SYSTEM PROJECT SUPERVISOR Dr. SYED WAQAR SHAH PROJECT LEADER AMMAD UDDIN MEMBERS AMEER ULLAH RUMMAN KHAN MUHAMMAD SHAKEEL

1

INTRODUCTION

INTRODUCTION

1.1 The era of biometrics In the 21st century the use of biometric based systems have seen an exponential growth. This is all because of tremendous progress in this field making it possible to bring down their prices, easiness of use and its diversified use in every day life. Biometrics is becoming new state of art method of security systems. Biometrics are used to prevent unauthorized access to ATM, cellular phones , laptops , offices, cars and many other security concerned things. Biometric have brought significant changes in security systems making them more secure then before, efficient and cheap. They have changed the security system from what you remember (such as password) or what you possess (such as car keys) to something you embody (retinal patterns, fingerprints, voice recognition).

2

1.2 What is biometrics? Biometrics is the science of verifying the identity of an individual through physiological measurements or behavioral traits. Since biometric identifiers are associated permanently with the user they are more reliable than token or knowledge based authentication methods. 1.3 Advantages of Biometrics Biometrics offers several advantages measures. Some of them are presented below.

over

traditional

security

1.3.1 Accuracy and Security Biometrics based security systems are far most secure and accurate than traditional password or token based security systems. For example a password based security system has always the threat of being stolen and accessed by the unauthorized user. Further more the traditional security systems are always prone to accuracy as compared to biometrics which is more accurate. 1.3.2 One individual, Multiple IDs Traditional security systems face the problem that they don’t give solution to the problem of individuals having multiple IDs. For examples a person having multiple passports to enter a foreign country. Thanks to biometrics!!! They give us a system in which an individual can’t possess multiple IDs and can’t change his ID through out his life time. Each individual is identified through a unique Biometric identity throughout the world. 1.3.3 One ID, multiple individuals In traditional security systems one ID can be used by multiple individuals. For example in case of a password based security system a single password can be shared among multiple individuals and they can share the resources allotted to a single individual. Biometric based security system doesn’t allow such a crime. Here each individual has a single unique ID and it can’t be shared with any other individual.

1.4 Biometrics categories Biometrics can be categorized in various categories as follow. 3

1.4.1 Physical biometrics This biometrics involves measurement of physical characteristics of individuals. The most prominent of these include •

Fingerprints

•

Face

•

Hand geometry

•

Iris scans

Fingerprints Fingerprints recognition has been present for a few hundred years. Due to tremendous research this field has reached such a point where the purchase of fingerprint security system is quite affordable. For this reason these systems are becoming more widespread in a variety of applications.

Fingerprint image. Face There has been significant achievement in face recognition system in past few years. Due to these advancements this problem appears to be eventually technologically feasible and economically realistic. In addition,

4

current research involves developing more robust approaches that accounts for changes in lighting, expression, and aging, where potential variations for a given person are illustrated in Figure. Also, other problem areas being investigated include dealing with glasses, facial hair, and makeup. Facial Expression Image

Hand geometry Hand geometry is one of the most basic biometrics in use today. A two-dimensional system can be implemented with a simple document scanner or digital camera, as these systems only measure the distances between various points on the hand. Meanwhile, a three dimensional system provides more information and greater reliability. These systems, however, require a more expensive collection device than the inexpensive scanners that can be used in a two-dimensional system. An example of a commercial three-dimensional scanner is shown in Figure.

Commercial three-dimensional scanner As seen in this image, the physical size of the scanner limits its application in portable devices. The primary advantage of hand geometry systems is that they are simple and inexpensive to use. Also, poor weather and individual anomalies such as dry skin or cuts along the hand do not appear to negatively affect the system. The geometry of the hand, however, is not a very distinctive quality. In addition, wearing jewelry or other items on the fingers may adversely affect the system’s performance.

5

Iris Iris recognition has taken on greater interest in recent years. As this technology advances, purchasing these systems has become more affordable. These systems are attractive because the pattern variability of the iris among different persons is extremely large. Thus, these systems can be used on a larger scale with a small possibility of incorrectly matching an imposter. Also, the iris is well protected from the environment and remains stable over time. In terms of localizing the iris from a face, its distinct shape allows for precise and reliable isolation. Figure shows the unique iris pattern data extracted from a sample input.

Iris pattern 1.4.2 Behavioral biometrics This category of biometrics is temporal in nature. They are evolved during the life time of an individual. It involves measuring the way in which an individual performs certain tasks. Behavioral biometrics include •

Gait 6

•

Handwriting

•

Speech

•

Signature

Now let discuss some the behavioral biometrics in a little detail. Gait Gait-based recognition involves identifying a person’s walking style. Although these systems are currently very limited, there is a significant amount of research being conducted in this area. Furthermore, studies have shown that gait changes over time and is also affected by clothes, footwear, walking surfaces, and other conditions. Figure below outlines the various stages of a gait cycle. Samples recorded from a gait cycle

1.4.3 Chemical biometrics: This is a new emerging field. It involves measuring of chemical or biological composition of an individual different body parts such as •

DNA

•

Blood glucose

1.5 Multi-biometric systems Similar to multimodal systems, there are several other techniques aimed at improving the performance of a biometric system, as outlined in Figure below. 1.5.1 Multimodal 7

Multimodal systems employ more than one biometric recognition technique to arrive at a final decision. These systems may be necessary to ensure accurate performance. Combining several biometrics in one system allows for improved performance as each individual biometric has its own strengths and weaknesses. Using more than one biometric also provides more diversity in cases where it is not possible to obtain a particular characteristic for a person at a given time. Although acquiring more measurements increases the cost and computational requirements, the extra data allows for much greater performance.

1.5.2 Multialgorithmic These techniques acquire a single sample from one sensor and process this signal with two or more different algorithms. 1.5.3 Multi-instance These systems use a sensor to obtain data for different instances of the same biometric, such as capturing fingerprints from different fingers of the same person.

1.5.4 Multi-sensorial

8

These systems sample the same biometric trait with two or more different sensors, such as scanning a fingerprint using both optical and capacitance scanners.

Multi-biometric categories

FINGERPRINT IENTIFICATION 9

SYSTEM

Fingerprint Identification system 2.1 A brief history Fingerprints have been scientifically studied for many years in our society. The characteristics of fingerprints were studied as early as 1600s. Meanwhile, using fingerprints as a means of identification first occurred in the mid-1800s. Sir William Herschel, in 1859, discovered that fingerprints do not change over time and that each pattern is unique to an individual. With these findings, he was the first to implement a system using fingerprints and handprints to identify an individual in 1877. By 1896, police forces in India realized the benefit of using fingerprints to identify criminals, and they began collecting the fingerprints of prisoners along with their other measurements. With a growing database of fingerprint images, it soon became desirable to have an efficient manner of classifying the various images. Between 1896 and 1897, Sir Edward Henry developed the Henry Classification System, which quickly found worldwide acceptance within a few years. This system allows for logical categorization of a complete set of the ten fingerprint images for a person. By establishing groupings based on fingerprint pattern types, the Henry System greatly reduces the effort of searching a large database. Until the mid-1990s, many organizations continued to use the Henry Classification System to store their physical files of fingerprint images. As fingerprints began to be utilized in more fields, the number of requests for fingerprint matching began to increase on a daily basis. At the same time, the size of the databases continued to expand with each passing day. Therefore, it soon became difficult for teams of fingerprint experts to provide accurate results in a timely manner. In the early 1960s, the FBI, Home Office in the United Kingdom, and Paris Police Department began to devote a large amount of resources in developing automatic fingerprint identification systems. These systems allowed for an improvement in 10

operational productivity among law enforcement agencies. At the same time, the automated systems reduced funding requirements to hire and train human fingerprint experts. Today, automatic fingerprint recognition technology can be found in a wide range of civilian applications.

2.2 Fingerprint details In this section structure and detail about fingerprint image will be explained. We will emphasize and give greater detail of only those terminologies which are related to our project. 2.2.1 What is a fingerprint? A fingerprint is the feature pattern of one’s finger as shown in figure below. It is believed with strong evidence that these feature patterns are unique for each individual. So each individual has its own fingerprint with permanent uniqueness. That’s why fingerprints have been used for identification and forensic investigation for a long time.

A fingerprint image acquired by a sensor. 2.2.3 Fingerprint features

11

A fingerprint pattern is composed of a sequence of ridges and valleys. In a fingerprint image, the ridges appear as dark lines while the valleys are the light areas between the ridges. A cut or burn to a finger does not affect the underlying ridge structure, and the original pattern will be reproduced when new skin grows. Ridges and valleys generally run parallel to each other, and their patterns can be analyzed on a global and local level. Ridges and valleys generally run parallel to each other, and their patterns can be analyzed on a global and local level. 2.2.4 Global level At the global level, the fingerprint image will have one or more regions where the ridge lines have a distinctive shape. 2.2.5 Local level While the global level allows for a general classification of fingerprints, analyzing the image at the local level provides a significant amount of detail. These details are obtained by observing the locations where a ridge becomes discontinuous, known as minutiae points. Our project is also based on the minutiae based recognition. The most common types of minutiae are shown in Figure below.

12

Ridge Types

In general, a ridge can either come to an end, which is called a termination, or it can split into two ridges, which is called a bifurcation. The other types of minutiae are slightly more complicated combinations of terminations and bifurcations. For example, a lake is simply a sequence of two bifurcations in opposing directions, while an independent ridge features two separate terminations within a close distance. In our project we will use only the ridge ending and ridge bifurcation for identification.

13

Minutiae Types (Terminations and Bifurcations)

2.3 FINGERPRINT MATCHING TECHNIQUES Two main approaches are used for fingerprint matching which are described below. 2.3.1 Minutiae based The first approach, which is minutia-based, represents the fingerprint by its local features, like terminations and bifurcations. This approach has been intensively studied, also is the backbone of the current available fingerprint recognition products. I am also using this approach in my project.

2.3.2 Image based The second approach, which uses image-based methods, tries to do matching based on the global features of a whole fingerprint image. It is an advanced and newly emerging method for fingerprint recognition. It is useful to solve some problems of the first approach. But my project does not aim at this method, so further study in this direction is not expanded in my thesis.

14

SYSTEM DESIGN 15

3.1 System Level Design A fingerprint recognition system consists of image acquisition, minutiae extractor and minutiae matcher. In our project there is also a password security added so a password acquisition and password matcher is also needed. For fingerprint acquisition, optical or semi-conduct sensors are widely used. But for my project I have used the available fingerprints on the net for testing. So no acquisition stage is implemented. The minutia extractor and minutia matcher modules are explained in detail in the next part for algorithm design and other subsequent sections.

3.2 Algorithmic Level Design To implement our project a four stage approach has been used. These four stages are given in the figure below.

Preprocessing

• Image enhancement • Image Binarization • Image segmentation

Minutiae Extraction • Thinning • Minutiae marking 16

Post-processing • False minutiae removal

Minutiae Matching

Figure. Project

System Process flow Chart

Each of these stages will now be explained in detail in the next chapter.

17

FINGERPRINT IMAGE PROCESSINGS

4.1 Fingerprint Image Enhancement The performance a fingerprint recognition system basically depends upon the quality of the input image. Since the images acquired with different kinds of sensors are not of the perfect quality and so they can’t be used directly for the matching. Therefore to ensure the accurate working of the system the image is first enhanced using different algorithms and image processing techniques. I have used histogram equalization in my project for the fingerprint image enhancement. It is explained below.

4.1.1 Histogram Equalization

18

Histogram equalization is used to expand the pixels to all the intensity values from 0 to 255. Some of the original images are very dark i.e. their histogram is such that their intensity values are concentrated towards the origin i.e. near zero intensity value. On the other hand some of the images are very bright i.e. their intensity values are concentrated towards 255 and nearby intensity values. After applying the histogram equalization the pixels are distributed uniformly all over the intensity values from 0 to 255. Due to this process the contrast of the image is increased and so the visual effect of the image is increased. 2000 1800 1600 1400 1200 1000 800 600 400 200 0 0

50

100

150

200

250

200

250

Histogram of image before enhancement.

3000 2500 2000 1500 1000 500

19

0 0

50

100

150

Histogram of image after enhancement.

Fingerprint Equalization

Before Histogram

After

Fingerprint Equalization

Histogram

image

image

4.2 Fingerprint Image Binarization Fingerprint Image Binarization is to transform the 8-bit Gray fingerprint image to a 1-bit image with 0-value for ridges and 1-value for furrows. After 20

the operation, ridges in the fingerprint are highlighted with black color while furrows are white. First a threshold value from the image is selected using matlab function ‘graythresh’. A threshold intensity value between 0 and 1 is obtained using the above function. This value is then used to convert the grey image to a black and white image. The value of pixel which is less then the above threshold value calculated is taken as 0 representing the ridge in black color. A value of the pixel in the image which is greater than the threshold value calculated is then converted to 1 representing the white color valley or furrow. The following image shows the image before and after Binarization

Fingerprint before Binarization.

image

Figure. Fingerprint after Binarization

image

21

Fingerprint image segmentation

In general, only a Region of Interest (ROI) is useful to be recognized for each fingerprint image. The image area without effective ridges and furrows is first discarded since it only holds background information. Then the bound of the remaining effective area is sketched out since the minutia in the bound region is confusing with that spurious minutia that is generated when the ridges are out of the sensor.

4.3 ROI by morphological operations Two Morphological operations called ‘OPEN’ and ‘CLOSE’ are adopted. The ‘OPEN’ operation can expand images and remove peaks introduced by background noise. The ‘CLOSE’ operation can shrink images and eliminate small cavities.

4.4 Fingerprint Ridge Thinning Ridge Thinning is to eliminate the redundant pixels of ridges till the ridges are just one pixel wide. For this purpose I have used matlab built in function ‘bwmorph’. This function repeats operation on the ridges until they are one pixel wide and are suitable for minutiae extraction phase.

22

MINUTIAE MATCHING 5.1 Minutia Marking

After the fingerprint ridge thinning, marking minutia points is relatively easy. In general, for each 3x3 window, if the central pixel is 1 and has exactly 23

3 one-value neighbors, then the central pixel is a ridge branch as shown in figure. If the central pixel is 1 and has only 1 one-value neighbor, then the central pixel is a ridge ending shown in figure below.

0 1 0 0 1 0 1 0 1

0 0 0 0 1 0 0 0 1

Bifurcation

Termination

Also the average inter-ridge width D is estimated at this stage. The average inter-ridge width refers to the average distance between two neighboring ridges. The way to approximate the D value is simple. Scan a row of the thinned ridge image and sum up all pixels in the row whose value is one. Then divide the row length with the above summation to get an interridge width. For more accuracy, such kind of row scan is performed upon several other rows and column scans are also conducted, finally all the interridge widths are averaged to get the D.

5.2 Post processing Post processing mainly involves removal of the false minutiae from the fingerprint image. As described earlier, the crossing number algorithm is used again to locate the terminations and bifurcations within the final thinned image. In this process, the locations where the ridges end at the outer boundaries of the image are classified as terminations. In the true sense, however, these locations are not unique termination minutiae. Instead, they only appear as terminations because the dimensions of the image force each ridge to come to an end. Knowing this, these locations should not be recorded as minutiae within the fingerprint. One way to eliminate such locations involves creating an ellipse to only select minutiae points inside the fingerprint image. 24

The center of the ellipse is established by locating the minimum and maximum rows and columns that contain a ridge pixel, then calculating the row and column that lie halfway between these extremes. 5.3 Minutiae Matching The matching process involves comparing one set of minutiae data to another set. In most cases, this process compares an input data set to a previously stored data set with a known identity, referred to as a template. The template is created during the enrollment process, when a user presents a finger for the system to collect the data from. This information is then stored as the defining characteristics for that particular user. In our project since the whole process is done in matlab so there is no need for database creation. We need to compare only two images. If both the images are from the same fingerprint they are matched otherwise they are unmatched. The following are the steps involved in the matching process. 1. First of all the two fingerprints which are to be matched are load into the matching function. 2. Minutiae points i.e ridge ending and ridge bifurcation extracted from both fingerprint images are laoded into the function. 3. A built in matlab function ‘isequal’ is used to compare the two minutiae points i.e ridge ending and ridge bifurcation of both the images with each other. 4. If both the ridge endings and ridge bifurcations of the two fingerprint images matches with each other, only then the fingerprint images are matched otherwise they are unmatched.

25

PASSWORD SECURITY

26

Password Security Password? A password is a secret word or string of characters that is used to prove identity or gain access to a resource. The password must be kept secret from those not allowed access. Fingerprint is the state of art security measure as compared to password and other conventional security methods. But in our project we have summed up both the fingerprint and password security in order to design a more secure and efficient security system. Since our project design and simulation is in matlab so we have used matlab coding to implement this task. The steps used in method are as follow. 1. First of all an array of passwords is saved in m file of matlab. 2. When a user wants to enter the system he must enter the password to get access. 3. When the password is entered it is matched with the already stored passwords in the m file. 4. If the entered password matches with any of the passwords in the array stored in the m file then the password matches dialogue box appears. 5. If the password does not matches with any of the passwords stored in the array then an unmatched messages is displayed and access is not granted to the user.

27

In our project the Password security works and is presented in the following way as shown in different figures. 6.2 Password GUI a). when the user enters the Enter Password button a dialogue box appears asking the user to enter password.

Password GUI 1

28

b). when the user enters the password and presses OK on the dialogue box another dialogue box appears showing whether the password is matched or not.

Password GUI 2

29

SYSTEM GUI AND ITS USER MANUAL

30

7.1 GUI User Manual We have presented our project in matlab GUI. Following are different steps showing its operation. 1. Open the matlab 6.0 or above. 2. In the matlab command window type fig1 and press enter.

Matlab Command Window

31

The following window will appear when we press enter in the above figure.

Main GUI Window

There are three main panels in the main window as shown in the above figure. There is a control panel, view panel and exit panel. Control panel have two buttons of which match images is of main importance. The view panel shows different buttons used for performing different operations on the two images. 32

3. From the file menu click open and load two images as show in the figure below.

GUI Loading Images

4. After loading the two images we can view the images by clicking on the view images button in the view panel as show below.

33

GUI View Images

5. By clicking on the View Transform button in the View Panel we can view the transformed image after different image preprocessing and post processing algorithms applied as show below.

34

Transformed Images

6. When we click the histograms button in the view panel, the histograms of the two images are shown as in the figure below.

35

Histogram of Original Images

36

7. Click on the Hist Equalization to view the equalized histogram of the two images as shown below.

Histogram Of Images After Equalization

8. when we click the match images, it compares the two images and displays the result. If the two fingerprint images match then a message box is shown showing “Fingerprint matched”, otherwise 37

the message box displays “Fingerprint don’t match” as shown below.

Fingerprint Matching Result

9. In the Control Panel click on Enter Password to enter password for authentication. When we click on this button user is asked to enter password in the matlab command window. If the password is matched a message box appears showing that the password is matched otherwise the message box shows that the password did not match. 38

Password Matching Result

39

CONCLUSION

Conclusion In our project we have combined biometric i.e. fingerprint and conventional security i.e. password security system. The conventional password security has several advantages as well as disadvantages. Those disadvantages have been overcome by using it with fingerprint security system. So overall system has both the advantages of the state of art biometric and conventional security system which makes it more powerful than either of the two security measures. The efficiency of the fingerprint security system mainly depends upon the accuracy and the response time. Keeping these two things in mind we have used robust algorithms giving us fast response time and accuracy. Future work Following are the some of the recommendations and future work that can be done in order to enhance the system further. 1. There is a need in fingerprint security system to detect whether the fingerprint is from a living user or not. Experiments have shown that fingerprint security systems can be fooled by using copy of the fingerprint from a user. 2. Several biometric may be combined i.e. multi biometric security system may be used in order further increase the security and efficiency. 3. It is obvious that biometric systems will govern the security domain in future electronic world. So due increase in usage the identification speed and accuracy will be the crucial factors. Therefore the algorithms may be optimized in such a way in order to meet these demands.

40