Advances in the aerospace industry have driven growing research into the use of artificialintelligence for analyzing objects of interest in satellite imagery. Unlike typical 8-bit RGB camera images,however, satellite imagery often contains 16-bit pixel values, which can result in outliers that darken theimage. This issue leads to difficulty in object identification and negatively impacts analysis performance.

To address this issue, various image enhancement techniques have been proposed, but the effectiveness ofeach technique depends on the specifics of each satellite and the task to be performed. To address thisissue, various image enhancement techniques have been proposed. However, since each satellite has uniquecharacteristics and the effectiveness of each technique varies depending on the specific task, it is necessaryto carefully evaluate which technique is most suitable. This research analyzed which of the five imageenhancement techniques is most suitable for semantic segmentation tasks using the dataset from theKOMPSAT-3A satellite, which is widely used in South Korea. Experimental results using five semanticsegmentation models indicated that percentile stretching performed well in three models, suggesting it asthe most universally applicable method. In addition, for buildings and roads, which are important objectsin urban analysis, recursive separated and weighted histogram equalization (RSWHE) and percentilestretching were found to be effective.



Keywords KOMPSAT, Image enhancement method, Semantic segmentation