Irvine, Flavius Ryaas Rasyid and Natalia, Friska and Sudirman, Sud and Ko, Chang Seong (2023) Automated Physical Distancing Monitoring Using Yolov3. ICIC Express Letters, 14 (8). pp. 869-876. ISSN 2185-2766
|
Text
Automated Physical Distancing Monitoring Using Yolov3.pdf Download (3MB) | Preview |
Abstract
Physical distancing has been practiced and proven to be part of a solution to reduce the spread of COVID-19 during this pandemic. The method, when implemented together with other COVID-19 protocols such as face mask wearing, maintaining personal hygiene, and mass mobility limitation is very effective in reducing the airborne virus infection rate. As more and more countries and communities are returning to normal life during this pandemic, the enforcement of COVID-19 rules will need to be more automated to make it as least intrusive as possible. In this paper, we designed an automated physical distancing monitoring system using the YOLOv3 object detection library to detect people in the video frames of the system’s camera and determine the physical distance between them if more than one person is detected. The system has been implemented on our campus and has been shown to be sufficiently accurate in achieving those tasks.
| Item Type: | Article |
|---|---|
| Keywords: | COVID-19 mitigation, Person-object detection, Physical distance monitoring, Rapid application development, YOLOv3 |
| Subjects: | 000 Computer Science, Information and General Works > 000 Computer Science, Knowledge and Systems > 003 Systems (Computer Modeling and Simulation) 000 Computer Science, Information and General Works > 000 Computer Science, Knowledge and Systems > 005 Computer Programming > 005.4 System Programming, Operating System, Computer Interface |
| Divisions: | Faculty of Engineering & Informatics > Information System |
| Depositing User: | Administrator UMN Library |
| Date Deposited: | 06 Nov 2023 05:03 |
| Last Modified: | 06 Nov 2023 05:03 |
| URI: | https://kc.umn.ac.id/id/eprint/27035 |
Actions (login required)
 |
View Item |
