The invasive Crown-of-Thorns starfish (COTS) poses a significant threat to coral reef ecosystems. Early detection and removal are crucial to mitigating their impact. This research proposes a YOLOv5-based object detection system to accurately identify and locate COTS in underwater images. By fine-tuning the YOLOv5 model with a dataset of COTS images, the system can effectively detect and classify COTS with high precision and recall. The output of the system includes bounding boxes and confidence scores for each detected COTS, enabling timely intervention and removal efforts. This research contributes to the conservation of coral reefs by providing a valuable tool for monitoring and controlling COTS populations. The proposed system can be integrated into underwater drones or diveroperated devices to facilitate efficient and accurate COTS detection and removal.
Detection of the Crown of Thorns starfish using YOLOv5
2024-11-06
842112 byte
Conference paper
Electronic Resource
English
British Library Online Contents | 2015
|Military Aircraft Detection Using YOLOv5
Springer Verlag | 2023
|STD-Yolov5: a ship-type detection model based on improved Yolov5
Taylor & Francis Verlag | 2024
|Real-Time Military Aircraft Detection using YOLOv5
IEEE | 2024
|