Evaluation of RSI Classification Methods for Effective Land Use Mapping

An efficient remote sensing image (RSI) classification method is pre-requisite for an effective land use land cover (LULC) mapping. In order to provide the realistic LULC information of Ralegaon Siddhi watershed, for its future development, the present study has been carried out using the remotely sensed IRS-1D LISS III satellite image. The RSI classification is performed using various classification methods with the help of image processing capabilities of GIS software, namely Integrated Land and Water Information System (ILWIS). The results were assessed using a sample ground truth map through systematic random sampling. The five major LULC (e.g. Agriculture land, barren land, shrubs, waste land with scrub and water body) classes could be determined. It is found that maximum likelihood classifier performs better in terms of overall accuracy with 88.52% and 0.842 kappa coefficients.