Land Cover Classification Assessment Using Decision Trees and Maximum Likelihood Classification Algorithms on Landsat 8 Data

<p><em>Classification technique on remote sensing images is an effort taken to identify the class of each pixel based spectral characteristics various channels. Traditional classifications such as Maximum Likelihood are statistical parameters standard deviation and mean, which have a probability model in class. While object-based classification method, one Decision Trees, rules for with mathematical functions. This study compares Trees algorithms land cover Surabaya Bangkalan areas using Landsat 8 data. research begins creating Regions Interest (ROIs) Rules greater than less functions Trees. The ROIs test was carried out Separability Index matching Confusion Matrix. experimental results show that accuracy value resulting from Matrix calculation 90.48%, Kappa Coefficient Value 0.87. method produces nigher actual condition method. difference distribution two ways not significant. limited because validation uses manual interpretation results. Future expected use large-scale relevant agencies verify field data, larger samples ROIs, high-resolution imagery order improve results.</em></p>