Current work reports a case study of using ILWIS GIS for satellite image classification aimed at agricultural monitoring. Two Landsat TM scenes for 1992 and 2006 years have been compared and changes analysed. It has been illustrated that land cover types changes in the past decades which is caused by various environmental factors. The distributed habitats of agricultural landscapes and different crops cultivated in the foothills of the Mecsek hills, in 1992 and 2006 were interpreted based on the methods of raster images segmentation. This includes association of pixels into different thematic classes: vegetation, different categories of agricultural land agriculture and anthropogenic territories. This work contributes to the development of optimal algorithms of the agricultural monitoring. It furthermore details technical aspects of the GIS mapping applied for environmental monitoring: mapping soil and vegetation land cover types. Finally, the methods applied in this work enabled to perform recognition of different crop types on the satellite images based on pixel classification: wheat, barley, maize, sunflower, sugar beet, potatoes.