PROTOTYPE OF A FINGERPRINT BASED LICENSING SYSTEM FOR DRIVING

S.Ashwin Post Graduate Student Embedded & Real-Time Systems PSG College of Technology ashwin.s20@gmail.com S.Loganathan Post Graduate Student Embedded & Real-Time Systems PSG College of Technology logu22.89@gmail.com S.Santosh Kumar Post Graduate Student Embedded & Real-Time Systems PSG College of Technology sanuiq@gmail.com P.Sivakumar Assistant Professor Dept of EEE PSG College of Technology psk@eee.psgtech.ac.in

Abstract- To prevent non-licensees from driving and therefore causing accidents, a new system is proposed. An important and very reliable human identification method is fingerprint identification. Fingerprint identification is one of the most popular and reliable personal biometric identification methods. The proposed system consists of a smart card capable of storing the fingerprint of particular person. While issuing the license, the specific persons fingerprint is to be stored in the card. Vehicles such as cars, bikes etc should have a card reader capable of reading the particular license. The same automobile should have the facility of fingerprint reader device. A person, who wishes to drive the vehicle, should insert the card (license) in the vehicle and then swipe his/her finger. If the finger print stored in the card and fingerprint swiped in the device matches, he/she can proceed for ignition, otherwise ignition will not work. Moreover, the seat belt detector verifies and then prompts the user to wear the seat belt before driving. This increases the security of vehicles and also ensures safe driving by preventing accidents. Keywords- Safe Driving, Fingerprint, fingerprint reader, License, Ignition system, Smart Card

approximately half of all drunken driving takes place with drivers who do not have a valid driving license (Goldberg, 1997). Also in Sweden, unlicensed driving has been estimated as the cause of 100 deaths and 2500 injuries per year at a cost of more than one billion US dollars. In the USA, in 1995, more than 10,000 lives were lost in fatal accidents with unlicensed drunk drivers (approximately a quarter of all road deaths in that year). The equivalent figure in Great Britain would therefore be over 900 deaths if this rate prevailed. An in built system [2] in an automobile which prevents such cases has therefore become vital. This paper aims to introduce a hardware architecture which detects the fingerprint as well as the validity of the license of the driver and takes a robust decision to turn on or off the ignition system based on the validity. Section II describes the smart card and Section III describes the fingerprint matching algorithm. Section IV elucidates the architecture followed by results in Section V and conclusion in Section VI. II. SMART CARD The license issued by the Government is a smart card which stores different fields such as name, license no., date of expiry, fingerprints of 10 fingers, type of license and blocked status of the license as well as fingerprint templates. These fingerprint templates are derived from the fingerprint scan by the process as shown in Fig.1.

I. INTRODUCTION Unlicensed driving is a matter of concern for several reasons. It is possible that drivers who have not undergone appropriate training and testing may be deficient in some aspect of the knowledge and skills required to drive safely and efficiently. Also, drivers who are unauthorized may have less incentive to comply with road traffic laws in that they would not be influenced by the rewards and penalties set up under the licensing system. On this argument, drivers who do not hold a valid license may disregard the threat of license sanctions or the benefits of reduced insurance premium due to not having made a claim. It is noticeable in the literature [1] that the term unlicensed is used interchangeably to mean one of the below subcategories, as follows: A) Drivers who drive but who have never possessed any form of license; B) Drivers who have previously held a license but who have been disqualified; and C) Drivers possessing only a provisional license but whom, nevertheless, drive unaccompanied. For many unlicensed drivers, enforcement and penalties are not strong deterrents and in addition there are also administrative loopholes which some exploit. There appears to be a general laxity in the system of checking the validity of documents and their ownership for example it is claimed to be straightforward for an unlicensed driver to pass himself off as a friend (with a license) and later present the friends documents at a police station. According to a survey by the AA Foundation for Road Safety Research it has been estimated that in Sweden

Fig 1. Fingerprint Template Generation. The biometric fingerprint sensor takes a digital picture of a fingerprint. The fingerprint scan detects the ridges and valleys of a fingerprint and converts them into ones and zeroes. Complex algorithms analyze this raw biometric scan to identify characteristics of the fingerprint, known as the "minutiae". Minutiae are stored in a fingerprint template (a data file usually smaller than the initial scans). Up to 200 minutiae are stored in a template, but only a subset of these has to match for identification or verification. In most systems, if 10 to 20 minutiae match, the fingerprint is considered a match. In today's smart card

systems approximately 40 minutiae are stored, because of space restrictions. This template is stored in the smart card. III. FINGERPRINT MATCHING ALGORITHM Fingerprint identification is one of the most popular and reliable personal biometric identification methods. This paper describes an on-line fingerprint identification system consisting of image acquisition, edge detection, thinning, feature extractor and classifier. The preprocessing part includes steps to acquire binarized and skeletonized ridges, which are needed for feature point extraction. Feature points (minutia) such as endpoints, bifurcations, and core point are then extracted, followed by false minutia elimination. Human fingerprints are rich in details called minutiae, which can be used as identification marks for fingerprint verification. The algorithm that was implemented for finger print matching in this research work is discussed below. Anil Jain et al [3] proposed a hybrid matching algorithm for matching. Our algorithm is described in detail below. Step 1: Histogram Equalization: Histogram equalization is to expand the pixel value distribution of an image so as to increase the perceptional information. The original histogram of a fingerprint image has the bimodal type the histogram after the histogram equalization occupies all the range from 0 to 255 and the visualization effect is enhanced. Step 2: Fourier Transform: Because the image in the Fourier domain is decomposed into its sinusoidal components, it is easy to examine or process certain frequencies of the image, thus influencing the geometric structure in the spatial domain. Step 3: Binarization: A locally adaptive binarization method is performed to binarize the fingerprint image. Such a named method comes from the mechanism of transforming a pixel value to 1 if the value is larger than the mean intensity value of the current block (16x16) to which the pixel belong. Step 4: Direction: Field orientation and filtered field orientation map computation, which consists of the calculation of the dominant direction of ridges and valleys in each local region. Step 5: Region of Interest (ROI): 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. Step 6: Thinning:

The built-in Morphological thinning function in MATLAB is used for ridge thinning. The thinned ridge map is then filtered by other three Morphological operations to remove some H breaks, isolated points and spikes. Step 7: Matching: A bounding box is placed around each template minutia. If the minutia to be matched is within the rectangle box and the direction discrepancy between them is very small, then the two minutia pair is regarded as a matched minutia pair. Each minutia in the template image either has no matched minutia or has only one corresponding minutia. The number of matched minutia pair is calculated as percentage of matching. IV. SYSTEM ARCHITECTURE The system consists of smart card reader, controller module, seat belt sensing module, ignition system module and the smart card which is inserted into the system by the user. A fingerprint match causes the data pins to be in a high logic level and ideally output about 5volts while a fingerprint mismatch makes the data pins to be in a low logic level and ideally output 0volts. An interface control circuit was constructed to link the PC parallel port to the ignition system of a vehicle. This circuit provides a high degree of electrical isolation between the PC and the ignition system which operate at different voltage levels, through the use of components called optocouplers.

Fig 2. System Architecture. The circuit also provides capabilities for the controlling the ignition system via the interconnection of electronic components such as relays, bipolar junction transistors, resistors and diodes. Three wires from the ignition system of a vehicle are required to be connected to the interface circuit. When the parallel port data pins which form part of the connection to the interface circuit are in a HIGH logic level, the interface circuit is triggered to ignite the vehicle. On the other hand, the vehicle is not ignited when the circuit is in a LOW logic level. The principal

components of the prototype system are the fingerprint recognition software and the interface control circuits which are to form a continuous connection with a vehicle ignition system.The interaction between the various blocks are graphically represented in Fig 2. A) SMART CARD READER: The smart card is inserted into the slot of the reader. The card reader integrates with it a fingerprint scanner. This setup increases security by adding something you are along with something you have. When a fingerprint impression is received, it is analyzed for its minutiae information and is matched with the template stored in the card to find out if the fingerprints match. The process is depicted.

Fig 3. Fingerprint Reader Block Diagram. B) CONTROLLER: The role of the controller is to enable the logic flow (shown aside). The microcontroller is fed with the required input signals from card reader and seat belt controller. The card reader sends signals, each to individual pins of the microcontroller, and the signals include fingerprint matching information, license expiry status, license suitability status and license blocked status, all in the form of bits. The microcontroller then branches out to any one of the logical paths and delivers the output at one of its pins, which is used by the ignition control unit. The flow logic also includes checking of expiry of the license. If the license expires in 10 days, it prompts the user to renew the license, once the license expires, the ignition does not happen. When the user holds a learners license, it accepts the license after the fingerprint matches, then it prompts the user to insert a valid license and once again the checking process continues. The valid license is to be present in the vehicle until the vehicle is switched off. If the license is taken out before ignition is OFF, the vehicle automatically comes to OFF, and this ensures that the license is not used in another vehicle to switch it ON. We have implemented the proposed prototype using PIC 16F877 microcontroller. PIC microcontroller can also be used if the system is going to be more sophisticated and makes use of interrupts to control the switching ON and OFF of the ignition system. PIC 16F877 from Microchip [4] is a powerful yet easy-to-program (only 35 single word instructions) CMOS

FLASH-based 8-bit microcontroller packs PIC16F877 features 256 bytes of EEPROM data memory, self programming, an ICD, 8 channels of 10-bit Analog-toDigital (A/D) converter, 2 additional timers, 2 capture/compare/PWM functions, the synchronous serial port can be configured as either 3-wire Serial Peripheral Interface (SPI) or the 2-wire Inter-Integrated Circuit (IC) bus and a Universal Asynchronous Receiver Transmitter (USART). All of these features make it ideal for more advanced level A/D applications in automotive, industrial, appliances and consumer applications. All port connectors are brought out to standard headers for easy connect and disconnect. All the necessary support components are included, together with a Power and Programming LED for easy status indication. Plus a reset switch for program execution and a RS232 connection for data transfer to and from a standard RS232 port, available on most computers. The PIC16F877 Controller is the ideal solution for use as a standard controller in many applications. The small compact size combined with easy program updates and modifications, make it ideal for use in machinery and control systems, such as alarms, card readers, real-time monitoring applications and much more. The input to the PIC microcontroller is the result of the fingerprint matching algorithm through port pins RC6 and RC7. The port B is configured as output port is connected to the LCD to display the result of the fingerprint matching. This port can be directly connected to the ignition control or as LCD display to the user to indicate the current status of the fingerprint based ignition system so that the user can take the necessary action. This makes the proposed system user friendly. The table 1 below enumerates the various situations that might occur and their corresponding actions and Fig 4 shows the flowchart of this algorithm. Table 1. Input and Output for the System Possible inputs to system When valid license inserted Valid license and correct finger print input Valid license, correct fingerprint, seat belt is ON Valid license and incorrect fingerprint License with validity ending in 10 days Invalid license to the vehicle used Blocked license License with no validity Learners license Expected output from system Wait for fingerprint input Wait for seat belt YES for ignition Rejects and requests for input again Alerts the user License is rejected License is rejected License is rejected Asks to insert another valid

license

Fig 4. Flowchart for the algorithm. C) IGNITION CONTROL: The ignition system [5] of an internal-combustion engine is an important part of the overall engine system that provides for the timely burning of the fuel mixture within the engine. The ignition system is usually switched on/off through a lock switch, operated with a key or code patch. The part of the ignition system that first initiates the process of moving a vehicle is the key system in conjunction with the kick starter. A wire from the battery in the vehicle connects to the kick starter and other wires connect the kick starter to the key system. When the car key in the ignition system is turned once, two wires coming from the kick starter to the key system are bridged. This causes the engine and some other parts of the vehicle to be put in a READY or ON state. Turning the key again makes a third wire to temporarily join the already bridged wires, causing voltage to flow from the battery to the necessary parts vehicle so as to enable the vehicle move. The ignition control is brought about by placing a relay between the battery and the ignition key unit of the vehicle, as shown in Fig-5. The control of the relay is by a signal from the microcontroller. This signal is activated when the logic flow presented earlier satisfies. The relay on successful turn on, gives a feedback to the microcontroller. A solenoid valve attached to the fuel pipe of the engine opens when this feedback is obtained. If feedback was not received, then it is understood that the relay was manipulated and so the solenoid does not open thus preventing the engine from starting. The ignition control is thus made tamper-proof. The outline circuit diagram of the ignition control system is shown in Fig 5.

Fig 5. Ignition Control. V. RESULTS The output of each stage of the fingerprint detection algorithm which is implemented is shown from Fig 6 to Fig 9. The simulation was done using MATLAB 2010a. The output from MATLAB is sent as an input to PIC microcontroller PIC16F877 through serial communication. PIC is programmed using MikroC PRO for PIC. The simulation is also performed using Proteus Professional 7 on Windows 7. The hex file generated from the MikroC PRO compiler is dumped into the PIC microcontroller in Proteus software and it is then executed. The result of the fingerprint matching is displayed on the LCD screen. The output of the PIC microcontroller can be connected to the ignition of the vehicle. Fig 10 and Fig 11 show two possible outcomes of the fingerprint matching algorithm implemented using the PIC microcontroller.

Fig 7. Fast Fourier Transform of histogram equalized image

Fig 8. Binarization and Direction estimation.

Fig 6. Histogram Equalization of fingerprint image

Fig 9. Thinning and Minutiae detection performed before minutiae matching.

The present module can be interfaced with GPS / GSM module which would be of great use in future. The combined module can be used to monitor from remote location about the vehicle. The data can be used to monitor about the person who is driving the vehicle, by this way, theft can be minimized since it would help to find the person driving along with location details. Furthermore to augment the safety, the seat belt detector can be interfaced to the controller as an additional input for ignition. REFERENCES
[1] David Silcock, Anna Sunter & Chris van Lottum, Ross Silcock Limited, Kris Beuret, Social Research Associates, Unlicenced Driving: A Scoping Study to Identify Potential Areas for Further Research Foundation for Road Safety Research. [2] Anil Jain, Arun Ross and Salil Prabhakar, Fingerprint Matching Using Minutiae And Texture Features, Fingerprint Matching Using Minutiae And Texture Features, in Proc. of Intl Conference on Image Processing (ICIP), pp.282-285, Thessaloniki, Greece, Oct 7 - 10, 2001 [3] Omidiora E. O., Fakolujo O. A., Arulogun O. T., Aborisade D. O., (2011), A Prototype of a Fingerprint Based Ignition Systems in Vehicles, European Journal of Scientific Research, ISSN 1450-216X Vol.62 No.2 (2011), pp. 164-171. [4]http://www.microchip.com/wwwproducts/Devices.aspx?dDocName= en010241 [5]http://auto.howstuffworks.com/ignitionsystem.htm,How Automobile Ignition Systems Work

Fig 10. Simulation of PIC microcontroller showing matched fingerprint

Authors Profile

Ashwin S is currently an intern in Intel Technologies Pvt Ltd, Bangalore. He is pursuing his Masters in Embedded and Real-Time Systems from PSG College of Technology, Coimbatore, India. His research areas include Biomedical Image processing, Embedded Systems and Networking. Loganathan S is currently pursuing his Masters degree in Embedded and Real-Time Systems from PSG College of Technology, Coimbatore. His research areas include Embedded Systems, Embedded Networks and Real-Time Operating Systems. Santosh Kumar S is currently pursuing his Masters degree in Embedded and Real-Time Systems from PSG College of Technology, Coimbatore. His research areas include Embedded Systems, Fault Testing and Support Vector machines. Sivakumar P is currently working as Assistant Professor in department of Electrical and Electronics Engineering, PSG College of Technology, Coimbatore. His research areas include Controller Area Networks, Microcontrollers and Automotive Embedded Systems.

Fig 11. Fingerprint mismatch

VI. CONCLUSION AND FUTURE PROSPECTS The above input and output analysis of the proposed system proves that the vehicle can be ensured that it is been driven only by the authorized persons. The system also provides facility for the learners licensees to drive by keeping a licensed person near them. It also gives time to get the system repaired if any malfunction exists. In cars, it also ensures that the seat belt is worn by the driver, so that it adds the safety feature to cars. Though implementation of the proposed system may take time, it would be of great use for the safety of drivers and irregularities can be kept at check without any loopholes. The developed prototype serves as an impetus to drive future research, geared towards developing a more robust and embedded real-time fingerprint based ignition systems in vehicles.