Coral Detection using Image Enhancement and Discrete Wavelet Transform Techniques

Coral reefs provide vital environmental, economic, and medical benefits. They serve as breeding grounds for fish species, attract tourists, act as natural barriers against coastal erosion, and offer potential medical advancements. Coral reefs are under threat from pollution, overfishing, and climate change, thus effective protection and monitoring are essential. Accurate classification of underwater coral photos is challenging due to color distortion and poor contrast. The research in this article was supported by the Earth Sea Foundation (Thailand) to study a method for detecting and classifying coral species using Image Enhancement and Discrete Wavelet Transform (DWT) techniques with YOLO deep learning models. (YOLOv8, YOLOv9, and YOLOv10). Techniques such as color correction and red channel adjustment significantly improve image quality. Our comparative analysis demonstrates that YOLOv9 achieves the highest classification accuracy with a mean average precision (mAP) of 98.98%, followed by YOLOv10 with 97.57%, and YOLOv8 with 97.12%. This indicates that the performance of the models is ranked as YOLOv9, YOLOv10, and YOLOv8, respectively, in terms of their effectiveness in coral classification.