Uav detection with convolutional neural networks: The role of artificial intelligence technologies for civil uav safety

In this research, the detection of UAV, which have recently become an important source of concern in terms of defense and security due to the increasing use of civilian unmanned aerial vehicles, has been carried out with deep learning. A dataset consisting of various images of fixed-wing and rotary-wing civilian UAVs was created. These images were then labeled and converted into a format suitable for deep learning models. Trainings were conducted using YOLOv5, YOLOv7, YOLOv8 and the most recent YOLOv10 deep learning models that are frequently used in the literature. As a result of the trainings, it was revealed that the YOLOv8 model performed better than the other models with a 97% success rate. A Windows application was developed using PyQt5 library and Qt Designer. With this application, a system that can detect civilian UAVs was designed by combining the model with the best success rate. This study shows that deep learning models can be used effectively to detect civilian drones. The results obtained provide an effective method that can be used in the detection of civilian aerial vehicles that are important for security.