A review study on Vehicle Anti-Theft Immobilization System using Face Recognition

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 09 Issue: 10 | Oct 2022 www.irjet.net p-ISSN: 2395-0072

A review study on Vehicle Anti-Theft Immobilization System using Face Recognition

1-4 Student, Dept. of Electronics and Telecommunication Engineering, Vivekanand Education Society’s Institute of Technology, Maharashtra, India

5 Assistant Professor, Dept. of Electronics and Telecommunication Engineering, Vivekanand Education Society’s Institute of Technology, Maharashtra, India ***

Abstract - For preventing vehicle thefts, we propose an advanced security system for vehicles based on facial recognition using Raspberry Pi. It contains a facial detection system that authorizes the user, and if the userisdiscoveredto be unknown, the car is immobilized via a relay circuit in the car’s internal circuitry, and a notification is issued to the user via a mobile application. The authorized personnel can also add other users and get the vehicle’s location as the GPS (Global Positioning System) module, connected to the Raspberry Pi, will store the location informationinareal-time database Google Firebase, which will provide the vehicle’s location on the mobile application.

Key Words: FaceDetection,Anti-theft,Vehicletracking, Raspberry pi, Relay circuit, Vehicle Immobilization

1. INTRODUCTION

In today’s world, almost everyone owns a car. People are concernedaboutadvancedtechnologiesintheautomobile industry since buying a car is such a large expenditure. Today’sbiggestissueforautomobileownersisthecontinual concernoftheirvehiclesbeingstolenfromapublicparking lotorfromoutsidetheirhouses.Duplicatesofcarkeysare easy to create, and employing them raises the danger of thievery.Therehasbeenasignificantincreaseinautomobile theftsasthenumberofvehiclesontheroadhasincreased.

Car security system design and analysis are continually improving as a result of the expansion and use of various RaspberryPitechniques.Manymoderntechniques,suchas biometric recognition, image processing, communication, and so on, have been integrated into vehicle security systems. At the same time, the number of car accidents, particularlyfatalities,remainshigh.Asaresult,apractical car security system should be efficient, sturdy, and dependable.Traditionalcarsecuritysystemsareexpensive andrelyonalargenumberofsensors.Whenanautomobile istrulylost, nofurtherfeedback canbeuseful inassisting people in locating it. We propose that the automobile immobilizationapproachwithfacedetectionbeusedincar securitysystemssinceitisefficientfastandcheapinthatit ensures that the theft is permanently prevented before it evenbegins.Facedetectionalgorithmshavereceivedalotof attention in recent years, and it is a significant computer

visionproblemwithapplicationsinsurveillance,multimedia processing, and consumer products. Many modern face detectiontechniqueshavebeendevelopedinordertoattain a greater detection rate and a faster detection rate. By interpreting the received GPS signal, the GPS module determinesthespecificlocation

Fig-1:Casesinthelast10years(2011-2020),

Atotalof307000vehicleswerestolen

[Source: <https://www.hindustantimes.com/cities/delhi-news/cityof-cars-delhi-reports-maximum-vehicle-thefts-95-stoleneveryday-ncrb-report-101631730601005.html>]

The Raspberry Pi control center module controls all activities, including image acquisition, face detection, car immobilization through relay, GPS data acquisition, and notificationtotheuser.AccordingtoDelhiPolicepersonnel, cartheftaccountsforaround14percentofallcrimesinthe city.Evenduringthefirstharshlockdown,whichrestricted people’smovement,betweenMarch15and30,2020,over 83automobiles(morethanthreeperhour)werestolenin

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 09 Issue: 10 | Oct 2022 www.irjet.net p-ISSN: 2395-0072

Delhi.Tolimitthenumberofthefts,engineimmobilizersand sophisticated alarm systems have become standard equipmentinmanyautos.

Despitethetoughregulationsinplaceandsecuritymeasures implemented byauto manufacturersandthe government, criminalshavegrownincreasinglytech-savvy,discoveringa plethoraofwaystogetaccesstovehicles.Inthisdocument, weproposeanewmechanismthatwouldhelppreventthe carfrombeingstolen.

2. Literature Survey

S.Padmapriyaetal.[1]proposedanimprovedfacedetection andrecognitionmethodbasedoninformationofskincolor. Skin color detection is first performed on the input color image to reduce the computational complexity. Morphological operations were used which gives prior knowledgeforfacedetection.Here,theAdaboostalgorithm wasusedforfacedetection. Byusingprincipalcomponent analysis(PCA)algorithm,aspecificfacecanberecognizedby comparingtheprincipalcomponentsofthecurrentfaceto thoseoftheknownindividualsinafacialdatabasebuiltin advance.Thismethodofsecuritywascostefficient.

C.Nandakumaretal.[2]presentedanembeddedautomotive security system involving face recognition. This system is used to reduce the increased vehicle theft and allows the owner to identify the intruder thereby having the vehicle undercontrol.Microprocessor(ARM7)isusedasthecontrol unitinthesystem. The processing ofimagesinvolves two parts, face detection which makes use of Viola-Jones Algorithmbycascadeobjectdetectionandfacerecognition which makes use of Linear Discriminant Analysis (LDA) algorithm which performs the feature extraction of the storedimagesinthedatabase.Whentheimageprocessing unitclassifiestheimageasunknown,theunauthenticated faceimageissenttotheownermobilethroughMultimedia MessagingServices(MMS)usingaGSMmodem.Thesystem updates the GPS information about the vehicle location to theowner.

AhmedA.Elngaretal.[3]proposedanewintelligentvehicle security system called VSS − IoT using a secure, efficient, low-cost,andLowPowerProcessingchipwiththeInternet as its important part. Moreover, this paper employed a hybridmechanism(HaarCascade+PCA)forfacedetection andrecognitionofthecapturedimagebyaUSBcamera.This system converts the images captured by the USB camera installed in the vehicle to Gray-scale images to detect and recognize faces in real-time. While PCA is a mathematical method that is expressed as a transformation of highdimensional facial images into few-dimensional principal components called Eigenface which is used to recognize facesinthedigitalimage.

The primary purpose of Prof. P.R. Shahane et al. [4] is to reducethenumberofvehicletheftusingfacedetectionbyan

image processing algorithm. Only a set of authorized members(feededinthedatabase)willbeabletoaccessthe ignitionsystem.Thefaceisdetectedandrecognizedusing EigenFaceinPCAalgorithm.Therecognizedimageisthen compared with the authorized image of the users in the database.Ifthefaceisunauthorized,thesystemwillnotlet theenginesturnon.Thesystemwillsendarequesttothe owner through a Local application using MQTT protocol demandingapasscodetostarttherelay.

The paper [5] by Sagnik Basu Choudhuri et al. aims to replacetheexistingRFID-basedengineimmobilizerwitha better and foolproof system which is achieved by implementing Face Recognition as the primary defense mechanismagainstvehicletheft,usingNILabVIEWandits toolkits.AThird-PartyAccessmodehasalsobeendeveloped tohelppeoplewhodonothavetheirtemplatesstoredinthe system run the vehicle for a pre-programmed amount of time.ThePassiveDefenceSystem(PDS),whichincludesthe OTPgenerationanditsmailing,isalsoimplementedusing the State Machine in NI LabVIEW. A Transdermal Alcohol Sensorinterfaceisproposedwhichaddstothesafetyofthe driverandthesurroundingsbyavoidingDrivinginImpaired conditions. Apart from these, an ultrasonic sensor-based guidancesystemisalsointegratedintothevehicletoprovide aguidancesystemtothedriverduringadversecasessuchas heavy fog or poor visibility. All these systems work as a packageandoffergreaterpassengersafetywhilereducing theriskofvehicletheft.

ThesystemproposedbyProf.M.MBulheetal.[6]consistsof Raspberrypiasthecentralprocessingunitthatreceivesthe datausingacamera.Thishiddencameracapturesthefacial dataandsendsittotheprocessingunitforcomparison.A databaseismaintainedthatcontainsthefacialdataasgiven by the user and cannot be manipulated by any other unauthorizeduser.ThissystemalsohasaGSMmodulethat helpstheCPUtogenerateSMStotheownerofthecarwhen unauthorized access is noticed. Then the CPU generates a command that switches off the motor. The algorithm is programmed in Raspberry pi which helps to compare cameracapturedimageswithdatabaseinformation,italso generatescommandstooperateotherdevices/components orotheradd-onsasdefinedbytheprogrammer.

MaheshR.PawarandImdadRizvi[7]proposedanIoTBased Embedded System for Vehicle Security. The system is designed and developed using raspberry pi, a highresolution camera, a vibration sensor and an open-source software. For face recognition, ‘Local Binary Patterns Histograms’(LBPH)isusedasitissimplerandgivesbetter resultsindifferentlightconditions.Theinternalpartofthe vehicle i.e., the key components of the vehicle (ignition system) will be handled by the controlling unit. The controlling unit and the owner’s device are connected wirelessly and the other devices like sensors will be interfacedusingwiredmedia.Iftheimagecaptureddoesnot match with the existing database (includes only the

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 09 Issue: 10 | Oct 2022 www.irjet.net p-ISSN: 2395-0072

authorizedmembers),anemailofthepictureissenttothe owner.

3. Proposed Solution

The idea behind this project is to implement a security systemforvehiclesusingRaspberrypiwhichwillhelpthe ownertoauthorizetheaccessibilityandkeepatrackofhis vehiclethroughamobileapplication.

The first step of this proposed system includes the face recognition technique using Raspberry pi which provides user authentication by verifying the real-time database of authorized users. The Raspberry pi is programmed using Pythonandconnectedtothereal-timedatabase.

Whenever a user tries to access the vehicle, Raspberry pi capturesanimageandverifiesitwiththereal-timedatabase toprovideauthentication.Iftheuserdetailsarenotpresent in the database a notification is sent to the user on the mobile application. Whenever the owner detects a suspicious activity, he can get the current location of his vehicleonthemobileapplicationthroughtheGPSmodule installedintheprojecthardwarecircuitry.

Whenanunauthorizeduserisdetected,Raspberrypisendsa signaltotherelaycircuitinstalledinthevehiclenetworking system.Thisdetectedsignal cutsoffthefuelsupplytothe engine hence, the mobility of the vehicle is restricted. Another important feature of this proposed system is the user-friendlymobileapplication.Thisapplicationallowsthe vehicle owner to authorize the new user by adding user information into the real-time database and also allows userstoviewthedatabaseofauthorizedusers.

Appropriate implementation of all the functions and methodsproposedinthissolutioncanbuildasecuresystem andprovidefoolproofprotectionagainstvehicletheft

Fig-3:Flowchartofthesystem

4. Methodology

A. Hardware:

Asauserenterstheautomobile,animageiscapturedusing thepicamera.TheRaspberryPiincludespre-programmed code that compares the taken image to the existing user databasecreatedinGoogle’sreal-timedatabasefirebase.If theuser’sfacematchesoneoftheusersinthedatabase,the user is authorized and has control of the automobile; otherwise,thefuelsupplytothecariscutoff.Thesystemwill also send a notification to the authorized employees if an unknown user attempts to enter the car. In addition, the owner can use the mobile application to track the whereabouts of the vehicle. A GPS module will save the vehicle’scurrentlocationinGoogle’sdatabase.

Fig-2:Proposedblockdiagram

1) RaspberryPi:TheproposedsystemincludesaRaspberry Pitoexecutefacialrecognition,databaseconnection,andto send notifications to the owner’s phone via a mobile application. The user database is kept in the Raspberry Pi memory unit. The Raspberry Pi does image detection and recognition.Itisalsoharnessedasacontrolplatformforthis project.Inthissystem,weusetheRaspberrypi3B+model thatworksona64-bitquad-coreprocessorrunningat 1.4GHz,hasadual-bandof2.4GHzand5GHzwirelessLAN, Bluetooth4.2/BLE,fasterEthernet,andPoEcapabilityviaa separatePoEHAT.TheRaspberryPi3ModelB+maintains

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 09 Issue: 10 | Oct 2022 www.irjet.net p-ISSN: 2395-0072

the same mechanical footprint as both the Raspberry Pi 2 ModelBandtheRaspberryPi3ModelB.

2) PiCamera:ThePicameramoduleisaportablelightweight camerathatworkswiththeRaspberryPi.Itcommunicates withtheRaspberryPiusingtheMIPIcameraserialinterface protocol.Itismostcommonlyutilizedinimageprocessing, machinelearning,andsurveillanceprojects.ThePiCamera module is a camera that can be used to take photos and record high-definition videos. Raspberry Pi Board has CSI (CameraSerialInterface)interfacetowhichwecanattachthe PiCameramoduledirectly.ThisPiCameramodulecanattach to theRaspberry Pi’sCSIport usinga 15-pin ribboncable. Here,wehaveusedPicamerav1.3.Ithasaresolutionof5MP andcancapturewide,still(motionless)imagesofresolution 2592x1944pixels.

3) GPS Module: The Global Positioning System (GPS) is a satellite-basednavigationsystemthatdeliverspositionand timing information. Anyone with a GPS receiver and an unobstructedlineofsighttoatleastfourGPSsatellitescan usethesystemforfree.AGPSreceiverestimatesitsposition byaccuratelytimingthesignalssuppliedbyGPSsatellites. TheGPSmoduleusedintheproposedsystemisNEO-6Mthat hasapositionupdaterateof5Hz,arechargeablebatteryfor Backup, EEPROM to save configuration settings, and an operating temperature in the range -40 to 85°CUART TTL socket.GPSisnowadayswidelyusedandalsohasbecomean integralpartofsmartphones.

B. Software:

1) FaceRecognition:Afacerecognitionsystemisbuiltusing pythonlibraries.OpenCVlibraryisusedforimportingimages and pandas library is used for data analysis. The face recognition library and OpenCV are used to process the imagescapturedthroughthepi-camera.Thepythonprogram uses Deep-Learning based facial embeddings which are highlyaccurateupto99.38percentandarecapableofbeing executedinreal-time.Thefacerecognitionlibraryisapython package that wrapsdlib’s face recognitionfunctionsinto a simple, easy-to-use API. It helps in training the model by obtainingthepatternandoutlinesofaperson’seyes,nose, mouth,andchin.

2) MobileApplication:Amobileapplicationintegratedwith thehardwarehasbeenbuiltusingAndroidStudioandFlutter whichwillperformthefollowingthreefunctionalities:

i)Accessibilitynotification:

Afterfacedetection,anotificationissenttotheownerwhen an unauthorized user tries to access the vehicle. This notificationisreceivedonthemobileapplicationintegrated withtheraspberrypi.

ii)Authorizingnewusers:

The mobile application has a user-friendly interface and providesthefeatureofauthorizinganewuser.Thevehicle ownercanauthorizenewusersbyaddingtheirdetailstothe real-time database. The owner has to click on the upload buttonpresentonthemobileapplicationtosendthephototo the database which contains the data of other authorized users.

iii)VehicleTracking:

Themobileapplicationalsoprovidesthefeatureoflocation tracking.Theownercanaccessthelocationofthevehicleat any time whenever a suspicious activity is detected. This trackingtakesplacethroughtheGPSmoduleconnectedto theraspberrypi.TheGPSmodulesendsthelocationinthe form of latitude and longitude values to the real-time databaseandthesevaluesreflectthecurrentlocationofthe vehicleonthemobileapplication.

C.CarImmobilizationCircuitry

In a vehicle circuitry, the fuse box is connected to several componentsunderthehoodofanautomobile,oneofwhichis the fuel injector. The ECU (Electronic Control Unit) is attached to one terminal ofthe injector, while the otheris connected to the fuse box, which is connected to a 12V battery.ECU,theenginemanagementsystem’sheart,controls fuel delivery, air management, fuel injection, and ignition. Relaysareelectronicswitchesthatopenorcloseacircuitto controlitsflow.Soanautomotiverelayisinstalledbetween the fuse box and the injector (figure 4). When the relay receivesasignalaboutunauthorizedaccessfromraspberry piitcutsoffthefuelsupplytotheengineandrestrictsthe mobilityofthevehicle.

Relaysarealsousedtocontrolhighvoltagecircuitsusinglow voltage signals hence, to send signals to the vehicle’s networkingsystemthroughalowvoltagesignaldevicelike raspberry pi a relay is used. The relay alone cannot be connectedastheraspberrypicangiveoutamaximumofonly 3.3V.Hence,aMOSFETisaddedtoenergizeandde-energize therelaywithoutdamagingtheraspberrypi.

After the face detection process, one of the two outcomes occurs:facerecognized,notrecognized.Whentheuser’sface doesnotmatchwiththedatabase,theraspberrypigivesa commandtotherelaywhichbreaksthefuelinjectorcircuit therebycuttingoffthefuelsupplytotheengine.

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 09 Issue: 10 | Oct 2022 www.irjet.net p-ISSN: 2395-0072

authorization of new users by adding a new user to the database and tracking the location of the car. The Facial Recognitionalgorithmusesanefficientpythonprogramthat gives excellent accuracy. In case of theft, when an unauthorizedpersontriestoaccessthevehicle,oursystem providesaneffectivesolutiontopreventtheunauthorized user from accessing the vehicle by adding a relay to the circuittherebycuttingoffthefuelsupplytotheengine.By usingthissystem,wecanensurethesafetyofthevehicles andcanpreventthecarfrombeingstolen.

REFERENCES

Fig-4:FuelInjectorcircuitdiagram[Source: <www.semanticscholar.org>]

5. Future Scope

The fuel injection circuitry of all commercial vehicles controlledbytheECU(ElectronicControlUnit)aregenerally equippedwithlocksforprotectiontopreventtheusersfrom tamperingwiththecircuit. Inracecars,thecaseisslightly different;instructionscanalsobegiventotheECUtotune andenhancetheperformanceofthevehicle.

The On-Board Diagnostics (OBD) is a computer-based systemincludedinallautomobilesandtrucksmanufactured after1996.ECUisthecentralpartoftheOBDsystemasit collectsvariousdatafromthesensorsandusesittocontrol differentpartsofthevehicle(eg.thefuelinjector).Currently, theOBDisusedonlytoreadthestatusofavehicle.Innear future,thiscouldbetweakedtoreceiveinstructionsfromthe user to improve the performance and security of the vehicles.

Inacommercialvehicle,differentmodesareavailablesports mode,powermode,etc.forefficientusage.Similarly,anew virtual mode, the “Theft Mode” could be added so that, wheneveranunauthorizedusertriestoaccessthevehicle, the raspberry pi will send out specific codes given by the manufacturertotheECUtoturnontheTheftMode.These codes will instruct the ECU to shut down the vehicle networkingsystemorturnonthesleepmode.

For this to occur, the ECU should be able to function bidirectionally so that it can receive and execute the instructionsaccordingly.IntroducingthisfeaturetotheECU duringthemanufacturingstageitselfwillresultinareliable andinfalliblesecuresystem.

6. Conclusion

This paper proposes an efficient Vehicle anti-theft system that includes Raspberry Pi as its core processing unit and GPSmodule.Withthehelpofthissystem,wecanreducethe vehicle theft issue significantly. It includes a Facial Recognitionsystemandamobileapplicationwhichprovides user-friendly features which will not only be useful to eradicatethetheftbutalsoprovideadditionalfeatureslike

[1] S.Padmapriya Esther Annlin KalaJames “Real-time smart car lock security system using face detection and recognition” Published in: International Conference on Computer Communication and Informatics,2012.

[2] C. Nandakumar, G. Muralidaran and N. Tharani “Real Time Vehicle Security System through Face recognition”DivisionofMechatronics,Departmentof ProductionTechnology,MadrasInstituteofTechnology, AnnaUniversity,Chennai,INDIA.

[3] Ahmed A. Elngar and Mohammed Kayed, “Vehicle Security Systems using Face Recognition based on InternetofThings”Publishedin:InternationalJournal ofCreativeResearchThoughts(IJCRT),2019.

[4] Prof. P.R. Shahane, Subhashi Gupta, Rajat Shrivastav, Vignesh.SandSushantSingh,“VehicleTheftDetection Using Face Recognition” Published in: International Research Journal of Engineering and Technology (IRJET).

[5] Prof. M.M Bulhe, Bhavika Mahale, Rashmi Borika PrashantNakade,ShreyasNimje“RaspberryPiVehicle Anti-Theft Face Recognition System” Published in: 2020InternationalResearchJournalofEngineeringand Technology(IRJET)

[6] MaheshR.Pawar,ImdadRizvi“IoT Based Embedded SystemforVehicleSecurityandDriverSurveillance” Published in: 2018 2nd International Conference on Inventive Communication and Computational Technologies.

[7] ViskaMutiawani,SarahRahmany,TaufikFuadiAbidin “Anti-theftVehicleMonitoringandTrackingAndroid Application Using Firebase as Web Service” Published in: 2018 International Conference on ElectricalEngineeringandInformatics(ICELTICs).

[8] V.Akila,K.Sriharshini,P.Sravani,D.Sravanthi,Rishita Gopi “Intelligent Car Anti-Theft Face Recognition System” Published in: International Journal of Online andBiomedicalEngineering

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056

Volume: 09 Issue: 10 | Oct 2022 www.irjet.net p-ISSN: 2395-0072

[9] Syedfasiuddin,SyedOmer,Dr.KhanSohelrana,Amena Tamkeen, Mohammed Abdul Rasheed “Raspberry Pi Vehicle Anti-Theft Face Recognition System” Published in: 2020 International Research Journal of EngineeringandTechnology(IRJET).

[10] KingshukMukherjee“Anti-TheftVehicleTrackingand Immobilization System“ Published in: International ConferenceonPower,ControlandEmbeddedSystems (ICPCES)

2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal |