Camouflage Target Detection Using Deep Learning with Drone Images

In the task of camouflaged target detection, the target is highly integrated with the background textures, which makes it difficult to detect and leads to false detection or missed detection. A detection algorithm YOLOv5s has been used for the detection of camouflaged targets to explore the detection effect of using Artificial Intelligence on camouflaged targets. On the other hand, the use of drones for object detection as well as camouflage detection increases in many fields like military, civil, agriculture, etc. detection of hidden objects using drones and UAVs need to explore with efficient Deep Learning (DL) models. Drone data was used in this study to identify bounding boxes for military-like camouflage targets. This paper helps to fine-tune YOLO v5 and evaluate performance results. It is validated on the dataset obtained from the drone, and the experimental results show that the YOLOv5s can be well applied for detection of camouflaged target detection.