Surface water waste hotspot detection

This dataset has been compiled to support research in detecting surface water waste hotspots using computer vision and machine learning approaches. It contains a diverse collection of images depicting various water bodies such as rivers, lakes, ponds, and coastal areas, under different environmental and lighting conditions, with and without visible waste accumulation. The images were collected from open-access sources including Google Images and publicly available datasets on Kaggle, focusing solely on content permissible for educational and research use. After collection, all images were manually reviewed, cleaned, and organized into categories representing clean and polluted water conditions. Duplicate and low-quality samples were removed to ensure dataset integrity and visual diversity. The dataset is suitable for tasks such as image classification, object detection, and deep learning model training aimed at identifying polluted water regions.

This work was carried out collaboratively by Kanav Dogra, Vishal Menia, Amit Kumar, and Nishchal Singh Bishnoi under the guidance of ,
Shri Nishchay Malhotra - Scientist C (CWPRS)
Dr. Sheetal S. Patil - Assistant Prof. BV(DU)COEP Computer Engineering Department