 This study compared two types of satellite images, WorldView2, WV2, and QuickBird2 simulated QB2 to determine how they could best be used for object-based urban land cover classification. The authors first identified optimal segmentation parameters for both datasets, then compared the resulting segmentations and assessed the accuracy of the resulting classifications. They found that the WV2 dataset produced more accurate results due to its additional spectral bands, while the QB2 dataset was more suitable for simpler classifications. Additionally, the authors found that certain features from the WV2 dataset were not present in the QB2 dataset, suggesting that the former may provide more detailed information for urban land cover classification. This article was authored by Tessio Novak, Hermann Krux, Uwe Stille, and others.