Clustering Algorithm in ILWIS GIS for Classification of Landsat TM Scenes: a Case Study of Mecsek Hills Region, Hungary

The emphasis of this research is application of spatial analysis using clustering algorithm, and remote sensing data (Landsat TM imagery) for agricultural mapping of land cover types. The study area covers Mecsek Hills region, located in south-western Hungary. This region is characterized by high land heterogeneity and complex landscape structure, caused by intense agricultural land use in the region with mixed vegetation type and high environmental value. The research data consists in Landsat TM scenes taken for years for years 1992, 1999 and 2006. The methodology is based on cluster classification algorithm available in ILWIS GIS. Based on clustering technique, the agricultural land cover classes were identified by association of pixels on the Landsat TM scenes to thematic clusters. Different land use types were classified, which include natural vegetation coverage, anthropogenic areas and agricultural fields, sub-divided to various crop types. Once classification was complete, agricultural thematic maps have been created. The final research output consists in three independent agricultural thematic maps of land cover types for years 1992, 1999 and 2006.