Accuracy Assessment of Land Cover Classification in Jodhpur City Using Remote Sensing and GIS

Abstract: A This study examines the accuracy assessment of land use land cover (LU-LC) classification mapping using remote sensing and Geographic Information System (GIS) carried out in Jodhpur city. Landsat-8 OLI image of year 2015 was used and analyzed using image processing software. The land use and land cover classes area Built-up area, Vegetation, Mining area, Water bodies and Other area (crop land & shrub land) were studied. Supervised classification method was used to classify the image. The accuracy assessment model was used to measure of how many ground truth pixels are properly classified. The adequate accuracy assessment of the Land cover classification requires that a number of random location in the study area to be chosen that represent the classes in the classified map. These locations are visited and GPS data are collected so that they can be accurately classified. The classified locations are then compared with the land cover map to generate the overall accuracy. The result shows that overall accuracy of land use and land cover is 89.69% and Kappa (K) is 0.84 which is acceptable in both accuracy overall and Kappa accuracy.

Keywords: Accuracy Assessment, GIS, Kappa, Land Use Land Cover.