Implementation of Online Signature Verification by Slope Calculation Method 1

Size: px

Start display at page:

Download "Implementation of Online Signature Verification by Slope Calculation Method 1"

Barbra Bryan
5 years ago
Views:

1 Implementation of Online Signature Verification by Slope Calculation Method 1 A.P.Malode, Dr.P.T.Karule 1 Lecturer, Shri Datta Meghe Polytechnic,Nagpur Maharashtra,India, apmalode@gmail.com Professor, Yashwantrao chavan college of engineering, Wanadongri, Nagpur,Maharashtra,India, ptkarule@rediffmail.com ABSTRACT In this paper we have discussed on the implementation of the online signature verifier using the advanced RISC processor. The circuit is wired around the Atmel s AVR microcontroller.we also developed the algorithm based on slop calculation method. In that the signatures are generated & stored in the permanent memory in the form of look up table. During runtime, the signatures are again generated in real time and are compared to the stored, expected signatures. If disagreement occurs, the occurrence of an error is detected & reported.. Keywords: Signature, RISC,AVR, slop calculation, Online array 1. INTRODUCTION Signature is socially accepted authentication method & is widely used as proof of identity in our daily life. Depending on the format of input information, automatic signature verification can be classified into two categories a) Online signature verification b) Offline signature verification. FEATURES In this algorithm, we take advantage of the online registration data to recover writing trajectory of an offline input signature & make verification decision based on the 3. BLOCK DIAGRAM OF THE SYSTEM Touch screen A/D converter Processor Display Database in memory Figure1: Block diagram of online signature verifier

2 In this system touch screen is used as input device. It is used to sense the signatures which are in the form of analog signals. These signals are converted into digital signal by using analog to digital converter. After processing, Microcontroller unit stored the data (digital form of signature) in the memory & the corresponding message i.e. signature registered will be displayed on the LCD display. Digital forms of signatures are stored in the form of array. When same signature comes on the touch screen, the same process will be repeated. Microcontroller unit will verify the signature with the database signatures & if the signature matches with corresponding signature of database, the message signature correct will be displayed on the display. If the signature not matches with the database, the message signature incorrect will be displayed on the display 4. CIRCUIT DIAGRAM OF THE SIGNATURE VERIFIER +Vcc +Vcc 1 3 RS PA 0 RXD PA 1 PA TXD PA 3 Atmega16 PB Resistive Touch Screen (Input) LCD D7D SDA SCL 451 PD 4 PD PD 0 PD 1 PD RS RW EN Figure : Circuit diagram of signature verifier Above circuit diagram of signature verifier is wired around Atmel s AVR Atmega16. Resistive type touch screen is used as an input device. Resistive touch screen are so named because they are basically resistive voltage dividers. The four wires of the touch screen namely x+, x, y+, y are connected with the port A pins. LCD 16 display is used as output device,the data lines D 7 D 0 of LCD are connected with the port B, control signals RS,R/W,EN are connected with the PD 0,PD 1,PD pins of the AVR microcontroller respectively. For storing additional database the additional memory, serial EEPROM (IC 4C51) is connected. The IC protocol is used to perform the data transfer between the AVR microcontroller & serial EEPROM. The AVR microcontroller communicates with the other processor like PC through serial port using RS3 converter. Program can be burn in the microcontroller via serial port.

3 5. FLOWCHART The system will work in two modes namely registration mode & verification mode. In registration mode the algorithm will store the signature in EEPROM in the form of array. The array present in EEPROM is called as the registration array. For the n no. of signatures the n no. of arrays are required. In verification mode algorithm will store the signature in RAM in the form of array. The array present in RAM is called as online array. 5.1 Data Base Management The output of the A/D converter is a binary data of the analog signature. By using algorithm, Slopes of the consecutive coordinates of the signature is calculated, then each slope is converted into the sample.each sample consists of 8 bit data.for given signature, samples are generated. So the look up table of samples is created in the EEPROM memory for the one given signature.data base is required for comparing two signature samples. The number of memory locations required for storing the different encoded signatures. 5. Flowchart of Signature Registration Start Initialization Taking input from touch screen A/D conversion of signature data Wait for 5 sec Slope calculation Conversion of each slope in samples Storing all samples in array of EEPROM Stop Figure 3: Flowchart of signature registration

4 5.3 Flowcharts of Signature Sensing & Verification Start A Initialization Adding differences D1+D+ +Dn = SUM Taking input from touch screen In ideal case SUM=0 Practically SUM=Deviation A/D conversion of signature data Deciding threshold as per the Deviation Wait for 5 sec Slope calculation Is SUM<Threshold No Conversion of each slope into samples Yes Storing all samples in the form array of EEPROM Subtracting samples Sample1 of EEPROMSample1 of RAM =difference (D) =00h Sample nsample n = Dn Message Sign correct Stop Message Sign incorrect A Figure 4: Flowchart of signature sensing & verification

5 5.4 Data Processing The processing unit will check the signature with the database signature (sample codes) by using algorithm. The algorithm will perform the percentage match with the reference (i.e. given coded samples of signature).depends on the percentage match or error, system will decide whether it is correct or not & the corresponding message will be display on the output device. 6. ALGORITHM Algorithm of the system based on the slope calculation method Y (0,) (,) (6,6) o o (5,5) (0,0) (,0) X Figure 5: Signature Model 6.1 Representation of Signature In the very first stage the algorithm will find the coordinates of the origin (starting point i.e. Xo,Yo).With reference to the origin algorithm will find the other coordinates of the signature i.e.(x 1,Y 1 ),(X,Y ), (X 3,Y 3 )... (X n, Y n ). In the second stage the algorithm will calculate the slopes of the consecutive coordinates by using the standard formula Slope = (Y 1 Y 0 ) (X 1 X 0 ) Let s assume Origin = (X 0, Y 0 ) = (5, 5) & (X 1, Y 1 ) = (6, 6) The slope will be Slope = (65) / (65) =1 Likewise the other slopes are calculated. One sample is generated for one slope.

6 6. Technique Used For Signature Registration Presuming signature time 5 sec for each signature. For each signature, there are samples generated. The size of each sample is 8 bit. Sample rate calculated by Time required for signature Total no. of samples 5 sec = = 50 m sec All the samples of signature are stored in EEPROM in the form of look up table (array). It is called as registration array. In this project 5 arrays are created for 5 different signatures Namely Sign1, Sign, Sign3, Sign4, and Sign5 Sample1 Sample1 Sample1 Sample1 Sample1 Sample Sample Sample Sample Sample Sample3 Sample3 Sample3 Sample3 Sample3 Sample Sample Sample Sample Sample Sign 1 Sign Sign 3 Sign 4 Sign 5 Table 1: s present in EEPROM database 6.3 Technique used for signature sensing & verification For online verification of signatures, again samples created for each signature. At this stage the temporary look up table is created for storing all the samples in RAM memory in the form of array, it is called as online array. In this technique, initially sample1 (8 bit) of online array is subtracted from sample 1 of registration array. Then we get difference between two samples. Like that all the samples of online array are subtracted from the corresponding samples of registration array. Ultimately we get all the differences

7 Sample no. 1 of sign1 array present in EEPROM X 1 Sample no. 1 Eg. of 00 array present in RAM Y 1 no.1 D 1 Eg. 00 Eg. 00 Eg Sample no. of sign1 array present in EEPROM X Sample no. of array present in RAM Y no. D Eg Eg Eg Figure 6: Subtracting samples from arrays & obtaining differences Finally added all the differences & summation is stored in the variable named SUM. If all differences are zero, the content of variable SUM is zero.in ideal case the differences should be zero. The content of variable SUM will be zero. We conclude that the sample matches with their corresponding samples. But practically the differences are non zero. We got some deviation. Hence the content of variable sum is non zero value. At this stage the threshold is fixed.if the sum is less than the threshold, we conclude that the current signature matches with the database signature & displayed the message correct sign. If the deviation more than threshold, signature not matches with the database signature. Then we displayed the message incorrect sign No.1 (D 1 ) No. (D ) No. (D ) SUM Figure 7: Summations of all differences

8 VCC VSS VDD VEE RS RW E D0 D1 D D3 D4 D5 D6 D7 VCC Ideally SUM=0, But practically SUM 0 If SUM Threshold, then result= correct sign If SUM> Threshold, then result= incorrect sign Experimental study The experimental study of the signature verifier has done on the simulator named Proteus. In which 5 switches are connected for registration of the 5 different signatures. Also we have designed & fabricate the prototype of the signature verifier. U1(PA1/ADC1) (A) A B RV1 RV C D LCD1 LM016L 1k GND 1k U1 RESET XTAL1 XTAL PA0/ADC0 PA1/ADC1 PA/ADC PA3/ADC3 PA4/ADC4 PA5/ADC5 PA6/ADC6 PA7/ADC7 PB0/XCK/T0 PB1/T1 PB/INT/AIN0 PB3/OC0/AIN1 PB4/SS PB5/MOSI PB6/MISO PB7/SCK PC0/SCL PC1/SDA PC/TCK PC3/TMS PC4/TDO PC5/TDI PC6/TOSC1 PC7/TOSC PD0/RXD PD1/TXD PD/INT0 PD3/INT1 PD4/OC1B PD5/OC1A PD6/ICP PD7/OC AVCC AREF GND ATMEGA16 Figure 8: Experimental study on Proteus software References [1] Yasser Sedaghat, Seyed Ghassem Miremadi, Mahdi Fazeli Dependable Systems Laboratory (DSL) Sharif University of Technology, Tehran, Iran Offline Signature Verification Using Online Handwriting Registration, 006 IEEE [] K. Wilken and J.P. Shen, Continuous signature monitoring: lowcost concurrent detection of processor control errors, IEEE Trans. on ComputerAided Design of Integrated Circuits and Systems, Vol. 9, Issue 6, June 1990, pp [3] Yu Qiao, Jianzhuang Liu,Department of Information Engineering The Chinese University of Hong Kong A SoftwareBased Error Detection Technique Using Encoded Signatures, 007 IEEE [4] H. Madeira and J. G. Silvia, Online signature learning and checking, in Dependable Computing for Critical Applications, J. F. Schlichting and R. D. Schlichting, Eds: SpringerVerlag, 199, pp [5] Henrique Madeira, Jose Camoes, and Joao Gabriel Silva Department of Electrical Engineerin University of Coimbr 3000 Coimbra Portugal Signature Verification: A New Concept For Building Simple And Effective Watchdog IEEE AUTHOR

A.P. Malode received the B.E. and M.Tech. Degrees in Electronics Engineering from Rashtrasant Tukdoji Maharaj Nagpur University in 001 and 010, respectively.

9 A.P. Malode received the B.E. and M.Tech. Degrees in Electronics Engineering from Rashtrasant Tukdoji Maharaj Nagpur University in 001 and 010, respectively. During , he worked as a capacity design engineer for development of home & industrial automation products in intellect control systems, Nagpur. Also he worked as lecturer in Electronics engineering. His area of interest is embedded system, Digital & Analog system design. Now he is working as a lecturer in Electronics & Communication Department, Shri Datta Meghe Polytechnic, Nagpur.

PART 1 : MR-16. PART 2 : CPU Board. PART 3 : Software Tools. PART 4 : Compile and Download. 1. Introduction 2. Features

PART 1 : MR-16. PART 2 : CPU Board. PART 3 : Software Tools. PART 4 : Compile and Download. 1. Introduction 2. Features MR-6 User Manual C O N T E N T S PART : MR-6. Introduction. Features PART : CPU Board. Placement Diagram (Silkscreen). Circuit Diagram 3. Parts List PART 3 : Software Tools. AVR Development Program Installation.