Assessment of Lead Contaminant in Ground Water using Image Processing and Machine Learning Algorithms

Water is a precious gift from the nature which is very essential for every living being on the earth. However this valuable source is polluted by various activities of human which adversely affects the eco system. Hence monitoring the quality of the water is necessary in global scenario. The assessment of water quality is done through laboratory procedure which is time consuming and requires expertise. Hence this study attempts to develop simplistic methodology for the assessment of water quality. As an initial phase, the research focuses on assessing the toxicity level of the ground water particularly Lead. To estimate the metal contamination, Araceae species such as Lemna Minor which has phytoremediation property is grown in the synthetically prepared water sample of various concentrations of Lead. The growth of the plant is analyzed using image processing techniques and the toxicity level is classified using k-NN, SVM and Logistic regression algorithms. The contamination is classified in to two - low and high level. The classification accuracy achieved with these methodologies is 90 % for all the classifiers.