Analysis and monitoring of pollution by feature extraction in lichen using enhanced image processing and machine learning algorithm

Abstract

Air pollution has become most vulnerable for the existence of the living organisms. Pollution level in lichens reflects the same effect in living organisms. The advanced techniques in computational field reduce the time complexity and increase the accuracy of system. Pollution monitoring system gives reliable and accurate data about changes in the contamination intensity in urban areas. The extraction of the features like size, shape and color give the analysis of effect of pollution rate in the city and also explains how much pollutant is deposited in the human body. The feature parameters reveal the difference between characteristics of lichen when the environment is healthy and polluted. Monitoring system is helpful in finding the metals deposited in lichen is same as the human body while breathing in polluted newlinearea. Lichens are the bio indicator that are very sensitive to the air quality and absorbs the entire metal element present in the air. The measurement of the color and size assists in the monitoring of pollution according to the environmental factors present in the site. The main aim of the pollution monitoring system by the bio indicators is to categories the color into normal or polluted area and to find the metal present atmosphere. The lichen growth is continuously monitored by finding the circumference value of the lichens. The noise in the lichen image is removed by various denoising filter. The results show that the proposed wavelet transform reduces the redundant noise level and give the high pass frequency detail of the image. The wavelet transform consider both the time and frequency value of the image that is used to give the edge detail of the image accurately. newline newline