Improved Security For Credit And Debit Cards

IMPROVED SECURITY FOR CREDIT AND DEBIT CARDS

SYNOPSIS Today, the World has become a place with very high security but there are hackers (code breakers) and thieves who are trying to bring down this security system. So we need keep updating our security system. Here one of the system which is presently in function could be more secured using Biometrics and Embedded System. In this paper we have used the biometrics (finger print recognition) for security purpose and embedded system to automate the complete process. There many methods in providing security using biometrics. We have used the finger print recognition. In embedded system we have used PIC micro controller (PIC16F877). Here our project is helping the banking sector to safe keep the credit cards so called the plastic money, ATM cards and the bank lockers. We have taken into consideration all factors that affect the security of the present system in those sectors and have given an updated version of security for them .The biometrics have been used as the lock in providing security and the embedded system has been interface that connect the user and the automated system. We hope to give an better security for the society which needs to be a secured place live for us. We believe that our idea provides better security than the existing system. PIC IN EMBEDDED We are going for micro controller called the PIC micro controller. The PIC micro controller has the following advantages of cost effectiveness, low power, high speed and high user interface and system integration. The PIC16F877 Micro controller belongs to the midrange PIC. ThePIC16F87X Micro controller (MCU) Family provides a migration path from OTP to FLASH in 28 to 44-pin packages, with a wide range of peripheral integration options. This family features a 14-bit instruction set, 5 to 8 channels of 10-bit Analog-to-Digital Converters, interrupt handling capability, various serial interface capabilities, Capture/Compare/PWM, Brown-out Detection and an 8-level deep stack.

BIOMETRICS Biometrics is defined as an automated method of identifying or authenticating the identity of a living person based on the physiological characteristics. It offers inherently stronger security than any other methods like iris, face and voice recognition. Basic steps involved in recognition systems are as follows:

•

A mechanism to scan and capture

•

Compression, Processing and Comparison

•

Interfacing with the application system.

FINGER PRINT VERIFICATION SYSTEM In order to have safety over the property, the people living in all parts of a country go for “Fully Secured Banking system”. But there are hackers and thieves who steel out of our own credit cards and debit cards People mainly look for the protective measures provided by the bankers. So to strengthen this system, we had chosen fingerprint of the user as the credit and debit cards along with an password for better security. Since, the fingerprints of all the human beings are unique in nature, we had prescribed it as key factor. But there are other factors, which are unique like that of fingerprint. They are face, iris, signature and voice. Iris recognition system has the highest accuracy in finding the fraudulent persons and to have the lowest error rate. But it has a disadvantage of high cost. Hence we selected the next better factor (fingerprint), as it has moderate cost and high accuracy. The statistics of various recognition techniques used are as follows:

Cost Iris --------------------------------------------------Signature--------------Fingerprint ---------------------------------Face ---------------------Voice --------------------Accuracy Various Recognition Systems- Comparison

Here according to the statistic we can see from the graph that iris recognition has the maximum security but we have chosen the fingerprint technique. Since finger print recognition has the next best security system to the iris recognition. The cost of the finger print recognition system is comparably less and most reliable among the various recognition system The above reasons that have made us go for finger print instead of the other biometrics technique for security system. This security techniques have made the credit and debit cards more secured. BASIC BLOCK DIAGRAM

LCD DISPLAY MODULE

CCD SCANNER MODULE

P I C

KEYPAD MODULE

M C U COMPUTER

MEMORY CARD READER MODULE

LCD DISPLAY

INTERFACING LCD AND PIC  Two-line liquid crystal display-each line can hold 16 characters

 Displays the characters with a bright background (yellow).  A potentiometer is varied to adjust the brightness of characters.  LCD consists of 16 pins to provide interfacing with PIC The LCD is interfaced through the PORT B of the PIC MCU. The port pins are connected to the data lines of a two row LCD. The three control pins ENABLE, READ/ WRITE & READ SELECT are connected to the port pins RE0, RE1, RE2 of the PIC MCU. CCD SCANNER MODULE This module is used to scan the user fingerprint. Then the user is asked for the certain password that has been provided by the bank. Here the password is used as filename which holds the information of the card holder and their fingerprint. The fingerprint obtain from the user is compared with fingerprint for that corresponding password code which is stored in the computer. The data comparison is done by certain techniques using MATLAB for simulation.

B KEYPAD MODULE The hex keypad consists of 16 keys to input the alphabets and numerals. This 4 X 4 matrix keypad is formed by using the D port of

the PIC MCU 16F877.

Here

the port pins RD0, RD1, RD2, RD3 act as input pins and RD4, RD5, RD6, RD7 act as output pins. The pins are connected to the power supply and when both the pins become high i.e., when the switch corresponding to both the pins is pressed then the corresponding letters (that are pre assigned in the program) are read from the program and input to two row LCD through the data lines. At the same time these letters are also got for processing. Hex-Keypad RD0

RD4

RD5

RD1

1

2

G H 3

K L 5

M N 6

O P 7

C

I 4

RD3

F

A B 0 J

RD2

D

E

RD6

Q R 8

S

T 9

U V W 0

XYZ . \0

RD7

NEXT

PREV

CLOSE

ENTER

The diagrammatic view of the hex keypad is shown below followed by the functions of each special keys. OPERATION  In our system each key (switch) is programmed to input multiple characters.  This can be explained by taking an example of the first key (i.e.,) the switch provided at the intersection of the pins RD0 & RD4.

 So on the first press of this key ‘A’ would be input to the system. And on pressing it again a ‘B’ would be input and the next press would input a ‘0’.  The next press would again insert an ‘A’ and so on. MEMORY CARD MODULE This module has memory card reader and data editor. Every bank are provided with separate memory cards where the amount that have been credited or debited using banks credit or debit card is stored. This memory card holds the database as the account number and the amount debited or credited for the card. At the end of the day these memory cards are given to the corresponding banks the amount for the cards can be collected by the shopkeepers. ALGORITHMS An algorithm is defined as the step-by step procedure of instructions involved in the processing. We had used MATLAB to compare two fingerprint images. We had employed this finger print recognition system by simulation. In our actual system, we have planned to use a CCD camera to get the fingerprint from the user. But we have not implemented that part, since high precision cameras are too expensive. As already mentioned, we have developed two algorithms for comparing fingerprint impressions. They are as follows:

 Direct Comparison Of Pixels:  Rotational Algorithm In the former method, the pixels of two images are compared directly based on the position of pixels. In the second method, the image is rotated either to clockwise or anti-clockwise direction depending on the tilt of input image. The advantage and disadvantage of both the algorithms are discussed below. DIRECT COMPARISON OF PIXELS In this algorithm, we are going to compare the pixel(gray values) of the two images, directly. The result of this algorithm is based on number of pixels that are matched. The algorithm is given below.

Algorithm 1. Get the input from the user. 2. Retrieve the already stored fingerprint image from the database, corresponding to the account number. 3. Pre-process the two images. 4. Modify the two images by cutting the borders. 5. For the first 10 pixels in Y-axis of both the images, count the number of pixels matched and unmatched. Let they be ‘M’ (matched) and ‘UM’ (unmatched). 6. Find the difference between M and UM. 7. Repeat the process for next 10 pixels in y-axis and so on for entire image. 8. Display the ‘differences’ obtained from comparing of each block. 9. If the number of negative difference is greater than or equal to TWO (>=2), then display that the two images are not matching. (At the same time give the alarm signal to the alarm circuit). 10. Else display that the images are matching and provide the control signal to debit or credit the amount from the account. For the following two fingerprints (one from the database and other from the user) the result is “ Finger Prints Are Matched”, since they are exactly same. Image from the Database

Direct Comparison of Pixels - 1

Input Image

Advantage The main advantage of this algorithm is its simplicity in comparing the images. The program is very simple to implement and to understand. Disadvantage Since we are comparing the images directly, this algorithm has a disadvantage that if the input image of the user is tilted, then the result displayed may be faulty. For example, consider the following two images (one of which is tilted) to be compared. The result of the comparison is “Finger prints are not matching”. Image from the Database

Input Image from user

Direct Comparison of Pixels - 2 In order to overcome this difficulty, we have developed another algorithm called ROTATIONAL algorithm, which is as follows ROTATIONAL ALGORITHM In this algorithm, the images are modified in such a way that the borders are also tilted (rotated) in the direction of input image’s tilt. The algorithm is as follows. Assume hat the fingerprint of the user has been tilted in clockwise direction. Algorithm 1. Get the input from the user. 2. Get the already stored fingerprint image from the database corresponding to the account number. 3. Pre-process the two images.

4. Take the first image (already stored) and modify it by cutting the borders. 5. Take the second image (input image) and cut the borders in such a way that all the four border are tilted towards the clockwise direction. Now the entire image is rotated in CW direction. 6. Next count the number of black and white pixels in both the images. 7. Calculate the difference in the total black and white pixels. Let it be ‘B_diff’ and ‘W_diff’. 8. If the W_diff is greater than B_diff, then display that the images are not matching. Provide an alarm signal to the alarm circuit. Since we are finding the differences and we are mainly concerned with the black pixels, the above condition is developed. 9. Else display the images are matching and then provide the control signal to open the locker. Advantage: As already mentioned this algorithm gives additional security in avoiding the entry of fraudulent persons and to allow the correct user to access his/her locker, even when the user’s fingerprint image is tilted. The Efficiency of this algorithm is “NEARLY 80%”. If we combine the two above-mentioned algorithms, we can have the efficiency of “NEARLY 90% “ Disadvantage: In this algorithm, there is a disadvantage that care should be taken in finding whether the image is tilted in clockwise or in anti clock direction

CONCLUSION: Since we have provided an automated system using embedded system, it can serve the people up to its maximum ability. Nowadays, we can’t see a system, providing the credit and debit card like what we have provided here. Hence, we decided to provide this system similar to that of ATM’s. We have chosen fingerprint recognition system to provide security for the user to access his/her locker. Hardware parts of our project are implemented using PIC MCU, whereas the Software parts are implemented using MATLAB and HITEC-C (C language for PIC MCU). We tried to produce 24 hours working system that is very compatible in its structure. Since we had used computer for fingerprint authentication, we cannot provide a compatible system. But, we had reduced man power required as in conventional banking systems. If we use scanners directly to match the fingerprint of the user then we can have a very compatible system. In the future, it may come into exist with additional features like automatic ON-OFF switch, real-time clock, etc. We can control the additional features by using the same PIC micro controller. Hence, in the future we can hope, a fully automated locker system will develop.