서해 갈조탐지를 위해 원격탐사에 아노말리 탐지 기법을 적용한 연구

Abstract

Massive blooms of the Sargassum horneri (S.horneri) occurred in the Yellow Sea in 2015. It caused fish kills and beach fouling. Detecting the S.horneri is important to reduce the damages. This study presents the results of the S.horneri detection using Geostationary Ocean Color Imager (GOCI) data during the period between January and June in 2015. Since GOCI data are acquired every hour during the daytime, allowing near real-time ocean monitoring, it can be effectively used to observe the temporal variations of the S.horneri patches that float on the oceanic water. We first used normalized difference vegetation index (NDVI) with 660nm and 745nm of GOCI to detect floating vegetation algae. With NDVI only, however, S.horneri detection is challenging in turbid waters because NDVI is elevated by the high NIR signals due to the mineral particles in turbid waters. So, in the next step, we employed an anomaly detection technique to increase the detection rate of S.horneri both in clear and turbid waters. Finally, we tried to differentiate between S.horneri and other floating algae and then estimated its coverage within a pixel.