Evaluate the Accuracy of Supervisor Classification for Al-Shatrah Image, Using Random Points by Remote Sensing and GIS

Remote sensing and GIS techniques are one of the very important tools for producing land-use maps and land cover through a process called image classification. This study examines the evaluation of the accuracy of the supervisor classification of the land cover for land use using Google Earth case of Shatrah city in Iraq for the year 2017, where a Landsat-8 OLI_TIRS image was used and analyzed using Arc GIS 10.6. After classifying the landcover/ landuse types, 100 Random Points were created in Arc GIS and convert random points to KML to open in Google Earth. The value of each random point in Google Earth has been validated to assess accuracy. This research includes two parts (1) Landuse / Landcover (LULC) classification and (2) accuracy evaluation. the supervised classification was performed. The major classified LULC were uncovered agricultural (66.6%), water (1.6%), urban areas (8.5%), Dense agricultural (9.9%), and barren lands (13.4%). The results indicate that the overall accuracy of the rating was 78% and the Kappa Coefficient (K) of 0.73. The kappa coefficient is classified as acceptable and therefore the categorized image was found to be suitable for further research. This study provides an essential source of information where planners and decision makers can use it for sustainable environmental planning.