Landsat8-9 Satellite Image Classification with ERDAS for Mapping the Urban Spaces of East Sikkim

This research initiative examines the categorization of Landsat 8-9 satellite imagery from the Operational Land Imager Sensor(OLI) using remote sensing to quantify land cover and delineate urban areas in East Sikkim for the year 2023. It employs a system for satellite image processing utilizing supervised classification methods. The processing was transitioned to spectrum preparation and enhancement of spatial resolution utilizing the blue, green, red, near-infrared, and panchromatic channels. LandSat 8-9 OLI and the Thermal Infrared Sensor (TIRS) satellites image has been used for this study. ERDAS Imagine software has been used for performing image classification. The validation of the findings from the categorization has been performed using multiple reference points, including a previously validated national processing result and a Google Earth image from the same year. Upon multiple iterations of image classification, the supervised method achieved overall accuracy of 91% and Kappa indices of 0.84 for the Maximum Likelihood classification technique where as the unsupervised method achieved overall accuracy of 76% and Kappa indices of 0.63 for ISODATA classification technique.