Proceedings of the
The Nineteenth International Conference on Computational Intelligence and Security (CIS 2023)
December 1 – 4, 2023, Haikou, China
Vision-based UAV Line Patrol System
1School of Information and Communications, Guilin University of Electronic Technology, China.
2Guilin University of Electronic Technology, Guangxi Educational Big Data and Cyberspace Security Collaborative Innovation Center, China.
ABSTRACT
With the ongoing advancement of Unmanned Aerial Vehicle (UAV) technology and computer vision, the utilization of UAV line patrol systems has expanded across various sectors, including industry, agriculture, urban construction, and environmental protection. Conventional patrol methods often require manual intervention, resulting in inefficiency and safety hazards. In contrast, vision-based UAV patrol systems leverage visual information to rapidly detect and identify ground trajectories, facilitating autonomous flight patrolling. This enhances patrol efficiency, accuracy, and mitigates safety risks associated with manual patrolling. This paper centers on exploring the application and significance of integrating UAVs and computer vision, specifically focusing on the design of an indoor vision-based UAV patrol system. The system operates in real-time, continuously computing UAV and related coordinates. It harnesses OpenCV's graph drawing function, performs affine transformations such as rotation and translation, and employs image compression and algorithms to consistently map the ground trajectory onto a canvas while saving the resultant trajectory map. This approach presents a novel method for the fusion of UAVs and computer vision, offering substantial practical value and promising prospects for broader implementation. The system finds effective application in indoor line patrolling, site surveys, security patrols, and other relevant fields.
Keywords: Unmanned aircraft, Digital image processing, PID control algorithm.
Download PDF
1School of Information and Communications, Guilin University of Electronic Technology, China.
2Guilin University of Electronic Technology, Guangxi Educational Big Data and Cyberspace Security Collaborative Innovation Center, China.
ABSTRACT
With the ongoing advancement of Unmanned Aerial Vehicle (UAV) technology and computer vision, the utilization of UAV line patrol systems has expanded across various sectors, including industry, agriculture, urban construction, and environmental protection. Conventional patrol methods often require manual intervention, resulting in inefficiency and safety hazards. In contrast, vision-based UAV patrol systems leverage visual information to rapidly detect and identify ground trajectories, facilitating autonomous flight patrolling. This enhances patrol efficiency, accuracy, and mitigates safety risks associated with manual patrolling. This paper centers on exploring the application and significance of integrating UAVs and computer vision, specifically focusing on the design of an indoor vision-based UAV patrol system. The system operates in real-time, continuously computing UAV and related coordinates. It harnesses OpenCV's graph drawing function, performs affine transformations such as rotation and translation, and employs image compression and algorithms to consistently map the ground trajectory onto a canvas while saving the resultant trajectory map. This approach presents a novel method for the fusion of UAVs and computer vision, offering substantial practical value and promising prospects for broader implementation. The system finds effective application in indoor line patrolling, site surveys, security patrols, and other relevant fields.
Keywords: Unmanned aircraft, Digital image processing, PID control algorithm.
Download PDF