고해상도 위성 데이터 처리를 위한 GCP chip 구축 및 비교

Recently, with a growing global interest in satellite imagery, the significance of high-accuracy preprocessing technology has become increasingly apparent. Employing an automated image matching method is crucial to enhance the geometric accuracy of the data, particularly through the utilization of GCP (Ground Control Point) chips. The NGII (National Geographic Information Institute) of Korea has generated GCP chips from aerial orthoimages, centered on points derived from traditional topographic surveying. However, these points prove ineffective for the image matching process with satellite images. In response to this limitation, we propose an innovative approach that involves extracting GCP chips from geographical features found in aerial images. This method aims to achieve superior matching results compared to existing chips. To execute this approach, matching was conducted using the generated GCP chips and KOMPSAT-3A images captured in the Taebaeksan region, employing an area-based matching method. The accuracy of the results was assessed by evaluating the matching rate and deviation correction after eliminating mismatches through data snooping. The outcomes revealed a higher matching rate for the new GCP chips compared to the existing ones. Furthermore, there were instances where image matching succeeded with the new chip, whereas it failed when using the existing GCP chip.