Published Paper Details:

Ocean pollution has become a serious environmental issue that threatens marine ecosystems, biodiversity, and human health. Traditional monitoring approaches, such as ship-based surveys and manual interpretation of satellite imagery, are costly, time-consuming, and limited in spatial coverage. To overcome these limitations, this paper presents an automated ocean pollution detection system using satellite imagery analysis and deep learning-based computer vision techniques. The proposed framework utilizes the You Only Look Once (YOLO) object detection model to identify and localize major surface-level pollutants, including oil spills, harmful algal blooms, and floating plastic waste. Prior to detection, satellite images are pre-processed using OpenCV to enhance image quality and ensure compatibility with the detection model. The trained YOLO model is integrated into a Flask-based web application that enables users to upload satellite images and visualize detection results in real time. The system highlights polluted regions using bounding boxes and provides confidence scores for each detection. Experimental evaluation demonstrates that the proposed approach achieves reliable detection performance with fast inference speed, making it suitable for practical environmental monitoring applications. The modular architecture of the system supports scalability and cloud deployment for large-scale ocean surveillance. The results indicate that combining satellite remote sensing with deep learning offers an efficient and cost-effective solution for automated marine pollution monitoring.