Testing Different Values of Segmentation Scale Parameter in the Function of Detection of Illegal Waste Landfills

Abstract

Illegal waste landfills are a current problem facing a large number of countries. In order to control and better manage illegal landfills, it is necessary to know the updated locations and contents of illegal landfills. Remote sensing methods have proven to be extremely effective in the automated detection of potential sites of illegal landfills. This paper focuses on the experimental testing of different values of the segmentation scale parameter of the Pleiades 1B satellite image in the function of detecting illegal landfills in the area of Novo Sarajevo municipality. In this case, the optimal value of the scale parameter is considered to be the value that gives the highest accuracy of classification and whose segments best correspond to the real situation on the ground.