인공신경망을 이용한 적조 탐지

Title
인공신경망을 이용한 적조 탐지
Alternative Title
Detecting Cochlodinium polykrikoides blooms using artificial neural network algorithms in Korean waters
Author(s)
유신재; 김예슬; 손영백
KIOST Author(s)
Kim, Yeseul(김예슬)Son, Young Baek(손영백)
Publication Year
2018-04-12
Abstract
This study is to develop detection algorithms for the blooms of Cochlodinium polykrikoides, which has caused extensive HABs for more than two decades in Korean waters. Various algorithms have been proposed to detect HABs. However, most of those algorithms are empirical and therefore limited in their applicability in the coastal areas where bio-optical conditions are highly variable spatially and temporally. In a previous study, we have shown that the remote sensing reflectance of Cochlodinium polykrikoides exhibits a distinctive depression in the blue-green wavelength band (Kim et al., Optics Express, 2016) based on a large data set of remote sensing reflectance simulated using HydroLight and IOCCG data. Based on this, we are developing neural network algorithms for in-water and satellite applications. We tested two kinds of neural networks: probabilistic neural network and resilient back-propagation neural network. Resilient back-propagation neural network performs better in terms of accuracy and calculation time, although training is more difficult than the former. The success rate for prediction was 0.95 for the simulated in-water remote sensing reflectance and 0.89 for the MODIS versus HAB sighting match-up data. We compare the HAB distribution from MODIS images processed by neural network algorithm and in-situ sighting information.hose algorithms are empirical and therefore limited in their applicability in the coastal areas where bio-optical conditions are highly variable spatially and temporally. In a previous study, we have shown that the remote sensing reflectance of Cochlodinium polykrikoides exhibits a distinctive depression in the blue-green wavelength band (Kim et al., Optics Express, 2016) based on a large data set of remote sensing reflectance simulated using HydroLight and IOCCG data. Based on this, we are developing neural network algorithms for in-water and satellite applications. We tested two kinds of neural networks: probabilistic neural network and resilient back-propagation neural network. Resilient back-propagation neural network performs better in terms of accuracy and calculation time, although training is more difficult than the former. The success rate for prediction was 0.95 for the simulated in-water remote sensing reflectance and 0.89 for the MODIS versus HAB sighting match-up data. We compare the HAB distribution from MODIS images processed by neural network algorithm and in-situ sighting information.
URI
https://sciwatch.kiost.ac.kr/handle/2020.kiost/23434
Bibliographic Citation
EGU 2018, pp.1, 2018
Publisher
EGU
Type
Conference
Language
English
Publisher
EGU
Related Researcher
Research Interests

Ocean Color Remote Sensing,Climate Change,UAV,해양원격탐사,기후변화,무인체계

Files in This Item:
There are no files associated with this item.

qrcode

Items in ScienceWatch@KIOST are protected by copyright, with all rights reserved, unless otherwise indicated.

Browse