Comparison of two Classification methods (MLC and SVM) to extract land use and land cover in Johor Malaysia

7th IGRSM International Remote Sensing & GIS Conference and Exhibition

地球科学;计算机科学

Deilmai, B. Rokni^1 ; Ahmad, B. Bin^1 ; Zabihi, H.^2

Department of Remote Sensing, Faculty of Geoinformation and Real Estate, Universiti Teknologi Malaysia, 81310 Johor, Malaysia^1

Department of Geoinformation, Faculty of Geoinformation and Real Estate, Universiti Teknologi Malaysia, 81310 Johor, Malaysia^2

关键词: Classification methods;    Classification results;    Environmental process;    Land cover classification;    Land use and land cover;    Landsat Thematic Mapper;    Maximum likelihood classifiers;    Remotely sensed data;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/20/1/012052/pdf
DOI  :  10.1088/1755-1315/20/1/012052

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

Mapping is essential for the analysis of the land use and land cover, which influence many environmental processes and properties. For the purpose of the creation of land cover maps, it is important to minimize error. These errors will propagate into later analyses based on these land cover maps. The reliability of land cover maps derived from remotely sensed data depends on an accurate classification. In this study, we have analyzed multispectral data using two different classifiers including Maximum Likelihood Classifier (MLC) and Support Vector Machine (SVM). To pursue this aim, Landsat Thematic Mapper data and identical field-based training sample datasets in Johor Malaysia used for each classification method, which results indicate in five land cover classes forest, oil palm, urban area, water, rubber. Classification results indicate that SVM was more accurate than MLC. With demonstrated capability to produce reliable cover results, the SVM methods should be especially useful for land cover classification.