Underwater Image Enhancement and Trash Detection Using Deep Learning

Underwater environments present formidable challenges for maintaining image quality and identifying Trash due to factors like water turbidity and low light conditions. In our research, we propose a unique method that combines the UGAN (Underwater Generative Adversarial Network) for enhancing images and YOLOv9 for detecting trash. Our goal is to not only improve the clarity and detail of underwater images but also to accurately locate and identify Trash. Our experimental results demonstrate the effectiveness of our approach in enhancing underwater images and detecting trash with precision. This work not only pushes the boundaries of underwater image processing but also holds promise for applications in marine conservation and underwater exploration, highlighting the importance of technology in understanding and protecting our oceans.