Integrated approach to assess the quality of river

Water pollution is a critical concern affecting ecosystems and water quality. This project presents an integrated approach to assess river quality using automated detection of pollutants and objects present in water. The system utilizes a dataset containing various pollution types, including acid pollution, dead animals, eutrophication, fish oil, and plastic pollution, along with objects like bottles, cardboard, glass, metal, leaves, paper, trash, and water containers. YOLOv10 and YOLOv8 and v9, an advanced object detection algorithm, is employed to identify these pollutants and objects in images of river water. The system is implemented using a Flask framework with a front-end developed in HTML, CSS, and JavaScript, and a Python back-end for processing. Users can register, log in, and access a detection module that analyzes uploaded images for pollutants. Detected pollution instances can trigger automated notifications to the administrator, enabling effective monitoring. This approach allows systematic evaluation of water quality, highlighting areas affected by different types of contamination. By combining object detection and user interaction modules, the project provides a framework for assessing and documenting water conditions. The methodology can be extended to other water sources and environmental studies.

Keywords: water quality, river pollution, YOLOv10, object detection, acid pollution, plastic pollution, eutrophication, fish oil, Flask, automated monitoring