A Study on Red Tide Detection Technique by Comparison of Spectral Similarity
-
Title
- A Study on Red Tide Detection Technique by Comparison of Spectral Similarity
-
Author(s)
- 백수호; 황도현; 김흥민; 강돈혁; 윤홍주
- KIOST Author(s)
- Kang, Don Hyug(강돈혁)
-
Alternative Author(s)
- 강돈혁
-
Publication Year
- 2017-09-08
-
Abstract
- This study proposes a method to detect red tide( Cochlodinium polykrikoides ) using Water-leaving radiance data and Absorption Coefficients data obtained from Geostationary Ocean Color Imager(GOCI) in the South Sea of Korea. The proposed algorithm detected the red tide pixels from the satellite image through the two-step filtering process. In the first step, clear water pixels were removed using the spectral slope change of Water-leaving radiance spectrum from 490nm to 555nm. In the second step, turbid water pixels were removed using the dynamic time warping algorithm for detects red tide pixels. The calculation process that proposed algorithm in this study was shorter than the previous study. In addition, proposed algorithm could be detected red tide pixel from turbid water as well as clear water.gorithm detected the red tide pixels from the satellite image through the two-step filtering process. In the first step, clear water pixels were removed using the spectral slope change of Water-leaving radiance spectrum from 490nm to 555nm. In the second step, turbid water pixels were removed using the dynamic time warping algorithm for detects red tide pixels. The calculation process that proposed algorithm in this study was shorter than the previous study. In addition, proposed algorithm could be detected red tide pixel from turbid water as well as clear water.
-
URI
- https://sciwatch.kiost.ac.kr/handle/2020.kiost/23847
-
Bibliographic Citation
- The 6th International Conference on Next Generation Computer and Information Technology (NGCIT 2017), pp.137 - 141, 2017
-
Publisher
- NGCIT 2017
-
Type
- Conference
-
Language
- English
- Files in This Item:
-
There are no files associated with this item.
Items in ScienceWatch@KIOST are protected by copyright, with all rights reserved, unless otherwise indicated.