Development of Classifier for Wetland Feature Extraction

Abstract

Water is one of the most important and essential natural resource for the survival of life on the earth. Geographical spread of this resource is over a vast area and in different forms, though the usable freshwater availability is less than 1%. For optimizing the utilization of the water resources, it is important to monitor the change dynamics, plan the usage and conserve the resource. newlineWetlands are observed to act as hydrological filters, buffer and storage for the purpose of natural balance and regulation of ecosystem, pollutant management, and hydrological cycle etc., for the environmental equilibrium, which should be considered in wider perspectives for water resource management and wetland conservation. Thus, restoration and conservation of wetland should be prime motto for the sustainability of all the living beings. newlineConservation of wetland and management of water resources demands for continuous monitoring and large amount of effort, which are observed to be, reduced remarkably by various applications of Geo-informatics. Despite of many approaches and models developed in this domain, the question is to find out the most proper approach for feature extraction and apply the algorithm efficiently to fetch information of interest for the specific purpose. newlineHowever, the accuracy of wetland detection by means of pixel-based classification has been limited for the coverage of large area and especially the identification of mud (Soil with Water Presence) is always a greater challenge. For overcoming these difficulties, a new Hybrid Echelons Model has been proposed for extracting and classifying the wetland features. The Echelon Analysis is based on the areas of relative high and low values of response variables. By the application of Principal Component Analysis (PCA), based on the visual interpretation and statistical analysis, the redundant data is eliminated. A set of knowledge based rules is proposed for mapping and classifying wetland features with Hybrid Echelons Approach for the State of Gujarat, India. Resu