Case study: security system for solar panel theft based on system integration of GPS tracking and face recognition using deep learning

Suprapto, Bhakti Yudho and Dwijayanti, Suci (2020) Case study: security system for solar panel theft based on system integration of GPS tracking and face recognition using deep learning. In: The Nine Pillars of Technologies for Industry 4.0. The Institution of Engineering and Technology, London, England, pp. 431-453. ISBN 978-1-83953-006-7

Text (Book Chapter Case Study) Case Study antitheft.pdf - Published Version Download (3MB)

Abstract

Security system is important to protect the objects, including solar panel modules. In this study, an integrated system that combines image processing and object tracking is proposed as a security system of solar panel. Face recognition using deep learning is used to detect unknown face. Then, the stolen object can be tracked using Global Positioning System (GPS) that works using General Packet Radio Service and Global System for Mobile communication system. The results show that the inte�grated security system is able to find the suspect and track the stolen object. Using the combination of FaceNet and deep belief network, unknown face can be recognized with an accuracy of 94.4% and 87.5% for offline and online testing, respectively. Meanwhile, the GPS tracking system is able to track the coordinate data of the stolen object with an error of 2.5 m and the average sending time is 4.64 s. The duration of sending and receiving data is affected by the signal strength. The proposed method works well in real-time manner and they can be monitored through a website for both recorded unknown face and coordinate data location

Item Type:	Book Section
Subjects:	#1 Repository of Teaching and Learning Process (PBM) > Buku Ajar (Book)
Divisions:	03-Faculty of Engineering > 20201-Electrical Engineering (S1)
Depositing User:	Mr. Bhakti Suprapto
Date Deposited:	28 Apr 2023 23:16
Last Modified:	28 Apr 2023 23:16
URI:	http://repository.unsri.ac.id/id/eprint/98027

Actions (login required)

View Item