The evolution of Unmanned Aerial Vehicles (UAV s) drone technology has seen significant advancements, with drones moving from basic remote-control aircraft to systems capable of complex mission execution. It is the fastest-growing technology with a wide range of usage. UAVs have become more accessible to the masses over the recent years. As drone technology continues to evolve, safety concerns have become more significant. In today's era it is crucial to ensure security in airspace from unauthorized drones. To address these challenges while simultaneously maximizing performance, we have compared 3 versions of the You Only Look Once (YOLO) architecture namely YOLOv7, YOLOv8 and YOLOv9. YOLO model architecture is used for object detection in images or video streams. This paper explores the effectiveness of YOLOv9 over YOLOv7 and YOLOv8, a state-of-the-art deep learning model, for real-time drone detection. The YOLOv9 model achieved better results as compared to YOLOv7 as compared to YOLOv8, indicating its high accuracy in detecting U AV s with a balance between precision and recall, and its robustness across various Intersection over Union (IoU) thresholds.
Performance Comparison of YOLO Algorithms in Drone Detection
2024-07-19
2226267 byte
Conference paper
Electronic Resource
English
Drone-YOLO: An Efficient Neural Network Method for Target Detection in Drone Images
DOAJ | 2023
|Drone Detection using YOLO and SSD A Comparative Study
IEEE | 2022
|Drone system for solar panel defect detection using YOLO
European Patent Office | 2022
|