GhostNet-YOLO Algorithm for Object Detection in UAV Image

The utilization of unmanned aerial vehicles (UAVs) in various fields has increased significantly in recent years, and object detection in UAV images is one of the critical applications. In this article, we suggest a new algorithm, GhostNet-YOLO, for object detection in UAV images. Our algorithm combines YOLOv5’s high accuracy and quick processing speed with GhostNet’s ability to reduce model complexity. TheGhostNet backbone network is first trained on a sizable dataset of object images, and it is then fine-tuned on UAV images to conform to their specific requirements. Then, using a UAV image dataset, we train the combined network from beginning to end by integrating the GhostNet backbone network with the YOLOv5 object detection head. Experimental findings indicate that our GhostNet-YOLO algorithm performs better than state-of-the-art object detection algorithms in terms of both accuracy and processing speed. It can accurately locate objects in UAV images and has real-time performance, making it suitable for numerous applications in fields such as agriculture, surveillance, and disaster response.