Implementation of Object Detection Method for Intelligent Surveillance Systems at the Faculty of Engineering, Universitas Sebelas Maret (UNS) Surakarta

Aris Maulana Fauzan (Universitas Sebelas Maret)
Sutrisno Ibrahim (Universitas Sebelas Maret)
Meiyanto Eko Sulistyo (Universitas Sebelas Maret)

Journal of Electrical, Electronic, Information, and Communication Technology (JEEICT)
Vol 4, No 1 (2022): JOURNAL OF ELECTRICAL, ELECTRONIC, INFORMATION, AND COMMUNICATION TECHNOLOGY

The number of positive Covid-19 cases in Indonesia continue to increase. This increase influenced by the behavior of Indonesian citizens in dealing with the pandemic, one of which is rarely wearing masks. In this study, we implemented an object detection method for intelligent surveillance systems (ISS) at the Faculty of Engineering, Universitas Sebelas Maret (UNS), Surakarta. By implementing face detection and mask detection, the surveillance system can recognize whether a person in a CCTV video frame is wearing a mask or not. In addition, deep metric learning and histogram of gradient (HOG) are applied to recognize faces of unmasked people in images. The test results show that the surveillance system can recognize the use of masks with 75%-87% accuracy rate. Furthermore, the accuracy rate for facial recognition on images ranges from 69% -100% for each person