Using Post-Classification Enhancement in Improving the Classification of Land Use/Cover of Arid Region (A Case Study in Pishkouh Watershed, Center of Iran)

Classifying remote sensing imageries to obtain reliable and accurate LandUse/Cover (LUC) information still remains a challenge that depends on many factors suchas complexity of landscape especially in arid region. The aim of this paper is to extractreliable LUC information from Land sat imageries of the Pishkouh watershed of centralarid region, Iran. The classical Maximum Likelihood Classifier (MLC) was first applied toclassify Land sat image of 15 July 2007. The major LUC identified were shrubland(rangeland), agricultural land, orchard, river, settlement. Applying Post-ClassificationCorrection (PCC) using ancillary data and knowledge-based logic rules the overallclassification accuracy was improved from about 72% to 91% for LUC map. The improvedoverall Kappa statistics due to PCC were 0.88. The PCC maps, assessed by accuracymatrix, were found to have much higher accuracy in comparison to their counterpart MLCmaps. The overall improvement in classification accuracy of the LUC maps is significantin terms of their potential use for land change modeling of the region.