GISA & IAG'i 2023

講演情報

口頭発表(国際セッション)

Disaster

2023年10月28日(土) 15:20 〜 16:40 D会場 (C-201 C棟2階)

座長: 畑山 満則 (京都大学 防災研究所)

16:20 〜 16:40

[D4-05] Comparison of the shorelines separately detected from Sentinel-1 and 2 satellite imagery acquired on the east coast of Korea

*Yun-Jae Choung1 (1. Geo C&I Co., Ltd.)

キーワード:Shoreline, Sentinel satellite imagery, Land cover classification, SAR sensor, Image classification technique

In this research, the shorelines separately detected from the Sentinel-2 satellite imagery with the multispectral bands and the Sentinel-1 satellite imagery with the SAR C-band were compared as follows. First, the Sentinel-1 GRD product was preprocessed, including multiple steps such as applying the orbit files, removing the thermal and border noises, the radiometric calibration, and the terrain correction, and the first land cover classification map was generated from the Sentinel-1 GRD product by using the image classification technique. Next, the second land cover classification map was generated from the multispectral bands of the Sentinel-2 L2A product using the same image classification technique. Then, the boundaries between the water and the other features in both land cover classification maps were separately selected as the first and second shorelines. Finally, the accuracy of both shorelines were measured for mapping the accurate shoreline positions in various types of coasts in the east coast of Korea.