Sea Fog Identification from GOCI Images Using CNN Transfer Learning Models
This study proposes an approaching method of identifying sea fog by using Geostationary Ocean Color Imager (GOCI) data through applying a Convolution Neural Network Transfer Learning (CNN-TL) model. In this study, VGG19 and ResNet50, pre-trained CNN models, are used for their high identification per...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-02-01
|
Series: | Electronics |
Subjects: | |
Online Access: | https://www.mdpi.com/2079-9292/9/2/311 |
id |
doaj-12024c5e609c434dbc06733a20253f03 |
---|---|
record_format |
Article |
spelling |
doaj-12024c5e609c434dbc06733a20253f032020-11-25T01:40:00ZengMDPI AGElectronics2079-92922020-02-019231110.3390/electronics9020311electronics9020311Sea Fog Identification from GOCI Images Using CNN Transfer Learning ModelsHo-Kun Jeon0Seungryong Kim1Jonathan Edwin2Chan-Su Yang3Marine Security and Safety Research Center, Korea Institute of Ocean Science & Technology, Busan 49111, KoreaMarine Security and Safety Research Center, Korea Institute of Ocean Science & Technology, Busan 49111, KoreaMarine Security and Safety Research Center, Korea Institute of Ocean Science & Technology, Busan 49111, KoreaMarine Security and Safety Research Center, Korea Institute of Ocean Science & Technology, Busan 49111, KoreaThis study proposes an approaching method of identifying sea fog by using Geostationary Ocean Color Imager (GOCI) data through applying a Convolution Neural Network Transfer Learning (CNN-TL) model. In this study, VGG19 and ResNet50, pre-trained CNN models, are used for their high identification performance. The training and testing datasets were extracted from GOCI images for the area of coastal regions of the Korean Peninsula for six days in March 2015. With varying band combinations and changing whether Transfer Learning (TL) is applied, identification experiments were executed. TL enhanced the performance of the two models. Training data of CNN-TL showed up to 96.3% accuracy in matching, both with VGG19 and ResNet50, identically. Thus, it is revealed that CNN-TL is effective for the detection of sea fog from GOCI imagery.https://www.mdpi.com/2079-9292/9/2/311sea fogremote sensinggociclassifciationcnntransfer learning |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Ho-Kun Jeon Seungryong Kim Jonathan Edwin Chan-Su Yang |
spellingShingle |
Ho-Kun Jeon Seungryong Kim Jonathan Edwin Chan-Su Yang Sea Fog Identification from GOCI Images Using CNN Transfer Learning Models Electronics sea fog remote sensing goci classifciation cnn transfer learning |
author_facet |
Ho-Kun Jeon Seungryong Kim Jonathan Edwin Chan-Su Yang |
author_sort |
Ho-Kun Jeon |
title |
Sea Fog Identification from GOCI Images Using CNN Transfer Learning Models |
title_short |
Sea Fog Identification from GOCI Images Using CNN Transfer Learning Models |
title_full |
Sea Fog Identification from GOCI Images Using CNN Transfer Learning Models |
title_fullStr |
Sea Fog Identification from GOCI Images Using CNN Transfer Learning Models |
title_full_unstemmed |
Sea Fog Identification from GOCI Images Using CNN Transfer Learning Models |
title_sort |
sea fog identification from goci images using cnn transfer learning models |
publisher |
MDPI AG |
series |
Electronics |
issn |
2079-9292 |
publishDate |
2020-02-01 |
description |
This study proposes an approaching method of identifying sea fog by using Geostationary Ocean Color Imager (GOCI) data through applying a Convolution Neural Network Transfer Learning (CNN-TL) model. In this study, VGG19 and ResNet50, pre-trained CNN models, are used for their high identification performance. The training and testing datasets were extracted from GOCI images for the area of coastal regions of the Korean Peninsula for six days in March 2015. With varying band combinations and changing whether Transfer Learning (TL) is applied, identification experiments were executed. TL enhanced the performance of the two models. Training data of CNN-TL showed up to 96.3% accuracy in matching, both with VGG19 and ResNet50, identically. Thus, it is revealed that CNN-TL is effective for the detection of sea fog from GOCI imagery. |
topic |
sea fog remote sensing goci classifciation cnn transfer learning |
url |
https://www.mdpi.com/2079-9292/9/2/311 |
work_keys_str_mv |
AT hokunjeon seafogidentificationfromgociimagesusingcnntransferlearningmodels AT seungryongkim seafogidentificationfromgociimagesusingcnntransferlearningmodels AT jonathanedwin seafogidentificationfromgociimagesusingcnntransferlearningmodels AT chansuyang seafogidentificationfromgociimagesusingcnntransferlearningmodels |
_version_ |
1725047712809746432 |