Generation of CAS500-4 Orthoimages Using Automatic GCP Extraction Method Based on Sentinel-2 Images; [Sentinel-2 영상 기반 자동 지상기준점 추출 방법을 이용한 농림위성영상 정사영상 생성]

초록

The Korean government plans to launch the CAS500-4 specialized for efficient management of the land. The CAS500-4 enables periodic monitoring of the Korean Peninsula with a spatial resolution of 5m along a 120 km swath width and a short revisit cycle of 3 days. To increase the usability of the CAS500-4, a process of removing geometric distortion is necessary. For this process, ground control points (GCPs) are used as key data. Acquiring GCPs through a field survey takes time and cost and cannot be achieved for inaccessible and overseas areas. In this study, we propose a method to automatically extract GCP based on Sentinel-2 Level-1C (L1C) images and generate CAS500-4 ortho-images using Rapideye 1B images with specifications like the CAS500-4 images. The proposed method is performed by generating GCP and ortho-images generation through geometric correction. After verifying the accuracy of the generated ortho-images, the sensor model accuracy before and after the geometric correction was analyzed. As a result, the sensor model accuracy with an error of less than 2 pixels was confirmed in all experimental images. Therefore, the proposed method is possible to automatically acquire GCP, it is expected to be utilized in the generation of ortho-images of CAS500-4 images in the future. © 2024 Korean Society of Remote Sensing.

키워드