A Comparative Study of Green Area Classification by Remote Sensing Techniques in Phayao Municipality, Thailand

Abstract

This research has two primary goals: first, to compare different image classification methods for identifying green areas within the Phayao Municipality, and second, to analyze and compare the resulting proportions of green areas. Using Sentinel 2B satellite imagery, the study classified green areas by applying two methods: Support Vector Machine (SVM) and Maximum Likelihood (ML). A Confusion matrix is used to verify the accuracy of the classification. Finally, the analysis focused on the proportion of green area in relation to the total urban area and the ratio of both total green area and recreational green area to the population count. The results show that both methods were able to classify green areas including community economic green areas, unused green areas, utility green areas, natural green areas, and public green areas. Although differences were found between the areas classified by the SVM and ML methods, the accuracy assessment indicated that the SVM method achieved higher classification accuracy than the ML method. Furthermore, the analysis reveals differences in the estimated proportion of green areas and the ratio of green space per capita between the two classification approaches. The findings suggest that the ratio obtained from the ML method is closer to the value reported by the government reported. Finally, the results can provide useful information for urban green space planning and management.

Keywords: satellite imagery, Support Vector Machine (SVM), Maximum Likelihood (ML), urban planning

© 2026 Serbian Geographical Society, Belgrade, Serbia.

This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Serbia.