Intraurban land cover classification using IKONOS II images and data mining techniques: a comparative analysis

Abstract

High spatial resolution image analysis acquired over urban areas has been performed with success using Geographic Object Based Analysis (GEOBIA). However, it was observed that the use of data mining techniques in the image analysis procedures can speed up the processing time by selecting the most appropriate parameters for classification process without decreasing the classification accuracy. Therefore, this work aims at comparing some algorithms for classifying intra-urban land cover using IKONOS II images and data mining techniques. Three classification algorithms, KNN, MLP and C4.5 were analyzed.