Prediction of Water Contamination Detection using Thermal Image Processing

Abstract

The increased rate of water-borne diseases is quite alarming and it is caused due to the poor quality of water. The toxins directly or indirectly find their way into water posing serious threats to the entire ecosystem. The work presented uses the technique in thermal image processing to monitor and detect the contamination in water. This method pertains to altering and analyzing the pictorial information of water with data sets representing various contaminants. The objective here is to visually monitor and statically evaluate some aspect of the images of water that is not readily apparent in its original form and enhance the image quality. The clean water templates and the infrared templates are compared using the GUI created in the MatLab platform. The non-contaminated areas and the contaminated areas are differentiated using color thresholding and effective matching is done with high accuracy using a pure water template.