PENGEMBANGAN SISTEM PENGAWASAN UNTUK DETEKSI DRONE DALAM LINGKUNGAN TERSAMAR

• Main Author: Exfrensive Nyaw, Joesef
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/77970
• Institution: Institut Teknologi Bandung

The widespread use of drones or UAVs has reached various sectors; however, drones are often operated by irresponsible individuals in restricted areas, such as around airports. This phenomenon poses significant risks to airport operations. Moreover, drones can also infiltrate hidden areas, such as among trees or buildings, making detection challenging. In this study, Writer proposes a drone detection approach using the YOLO method along with supporting algorithms. The writer developed a surveillance system that combines the YOLO model with the DeepSORT, Region of Interest and Background Subtraction algorithms. The training dataset includes drone images and video clips. The selection of the YOLO model is based on its efficient real-time performance. The experimental results show that the proposed surveillance system successfully detects drones in ambiguous areas, including among trees and buildings. However, one limitation of the system is that it sometimes misidentifies birds as drones. In conclusion, the proposed surveillance system has the potential to effectively detect drones in challenging detection conditions. While there are still some aspects to be improved, this research can contribute to the development of surveillance systems for drone detection in restricted areas and advancements in security technology.