An Appropriate Interpretation Approach for Land Use/Land Cover Map in Chomthong, Phitsanulok, Thailand

Abstract

The study of land use/land cover (LU / LC) has belonged to an important aspect of resource management, planning and mapping, etc. The acquisition of these results is due to the interpretation process, consists of two methods which are visual interpretation and digital interpretation. The tools used in the interpretation are classifiers in image processing software. Currently, there are many classifiers, but there is one method that provides better results than the other, maximum likelihood classifier (MLC). MLC is one of supervised classifiers, this method classifies the calculation of the probability for each given pixel in LU/LC class. The comparative study of the interpretation results with MLC through three image processing software (ERDAS, ENVI, QGIS) is the purpose of this study. Interpretation results show that ERDAS (OA = 83%, Kappa = 76%) provides better overall accuracy (OA) and kappa coefficient (Kappa) than ENVI (OA = 81%, Kappa = 74%) and QGIS (OA = 79%, Kappa = 71%).