Please use this identifier to cite or link to this item:
DC FieldValueLanguage
dc.contributor.authorArumugam, P.en_US
dc.contributor.authorKer, P.J.en_US
dc.contributor.authorTan, J.Y.en_US
dc.contributor.authorMani, D.en_US
dc.description.abstractIn recent years, car theft cases are increasing at a drastic rate across the globe. Existing car security system can be hacked into and there is no alert system to notify the car owner. A face recognition-based car security system was proposed and developed, where only authentic user was allowed to have access to the car ignition system. Hardware and software designs were carried out to develop a car security system using microcontroller-based face recognition system. This system was developed using a Raspberry Pi 2 microcontroller. Incorporating with other hardware modules such as Pi NoIR camera, liquid crystal display (LCD) panel, modem router and Wi-Fi adapter, the face recognition system is capable of performing various tasks efficiently. The system uses Eigenfaces recognition algorithm, which was developed using OpenCV and Python scripts, to perform face detection and recognition. The performance of the Eigenfaces recognition system was improved with light emitting diode (LED) support lighting and increased input image database. A short message service (SMS) alert system was set up using a modem router to notify car owner in case of unauthorized user access. In addition, a web-based application was designed and developed to remotely control the car security access and to trigger the car alarm. Compared to other existing car security systems, this microcontroller-based face recognition system offers a secured environment for cars at lower power consumption and with remote monitoring and control features. © 2017 American Scientific Publishers All rights reserved.
dc.titleMicrocontroller-based face recognition system with web-based application for car securityen_US
item.fulltextNo Fulltext-
Appears in Collections:UNITEN Scholarly Publication
Show simple item record

Google ScholarTM



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.