A cloud and snow detection method of TH-1 image based on combined ResNet and DeepLabV3+
Cloud and snow detection is an important part of satellite remote sensing image processing, and also a key step for its following analysis and interpretation. In this paper, a simultaneous cloud and snow detection method for satellite remote sensing images based on ResNet and DeepLabV3+ fully convol...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | zho |
Published: |
Surveying and Mapping Press
2020-10-01
|
Series: | Acta Geodaetica et Cartographica Sinica |
Subjects: | |
Online Access: | http://xb.sinomaps.com/article/2020/1001-1595/2020-10-1343.htm |
id |
doaj-a2714b3d69a245fd8095f595ad0b6491 |
---|---|
record_format |
Article |
spelling |
doaj-a2714b3d69a245fd8095f595ad0b64912021-08-18T02:32:11ZzhoSurveying and Mapping PressActa Geodaetica et Cartographica Sinica1001-15951001-15952020-10-0149101343135310.11947/j.AGCS.2020.2019042020201011A cloud and snow detection method of TH-1 image based on combined ResNet and DeepLabV3+ZHENG Kai0LI Jiansheng1YANG Jianfeng2OUYANG Wen3WANG Gaojie4ZHANG Xun5Information Engineering University, Zhengzhou 450001, ChinaInformation Engineering University, Zhengzhou 450001, ChinaInformation Engineering University, Zhengzhou 450001, ChinaInformation Engineering University, Zhengzhou 450001, ChinaInformation Engineering University, Zhengzhou 450001, ChinaInformation Engineering University, Zhengzhou 450001, ChinaCloud and snow detection is an important part of satellite remote sensing image processing, and also a key step for its following analysis and interpretation. In this paper, a simultaneous cloud and snow detection method for satellite remote sensing images based on ResNet and DeepLabV3+ fully convolutional neural network is proposed. The ResNet50 backbone is used, and the DeepLabV3+ network structure is optimized and improved according to the characteristics of cloud and snow on TH-1 remote sensing image. The ELU activation function, Adam gradient descent method and Focal Loss function are used to speed up convergence and improve segmentation accuracy. The network is trained and tested with the cloud and snow image dataset of TH-1 satellite. The experimental results show that it has strong robustness compared with Otsu method, and the detection accuracy of proposed method exceeds FCN-8s and original DeepLabV3+ network, meanwhile the detection speed of proposed method is faster than original DeepLabV3+, which can be applied to a variety of different remote sensing images through some adjustment and has favorable application prospects.http://xb.sinomaps.com/article/2020/1001-1595/2020-10-1343.htmsatellite imagecloud snow detectionth-1resnetdeeplabv3+ |
collection |
DOAJ |
language |
zho |
format |
Article |
sources |
DOAJ |
author |
ZHENG Kai LI Jiansheng YANG Jianfeng OUYANG Wen WANG Gaojie ZHANG Xun |
spellingShingle |
ZHENG Kai LI Jiansheng YANG Jianfeng OUYANG Wen WANG Gaojie ZHANG Xun A cloud and snow detection method of TH-1 image based on combined ResNet and DeepLabV3+ Acta Geodaetica et Cartographica Sinica satellite image cloud snow detection th-1 resnet deeplabv3+ |
author_facet |
ZHENG Kai LI Jiansheng YANG Jianfeng OUYANG Wen WANG Gaojie ZHANG Xun |
author_sort |
ZHENG Kai |
title |
A cloud and snow detection method of TH-1 image based on combined ResNet and DeepLabV3+ |
title_short |
A cloud and snow detection method of TH-1 image based on combined ResNet and DeepLabV3+ |
title_full |
A cloud and snow detection method of TH-1 image based on combined ResNet and DeepLabV3+ |
title_fullStr |
A cloud and snow detection method of TH-1 image based on combined ResNet and DeepLabV3+ |
title_full_unstemmed |
A cloud and snow detection method of TH-1 image based on combined ResNet and DeepLabV3+ |
title_sort |
cloud and snow detection method of th-1 image based on combined resnet and deeplabv3+ |
publisher |
Surveying and Mapping Press |
series |
Acta Geodaetica et Cartographica Sinica |
issn |
1001-1595 1001-1595 |
publishDate |
2020-10-01 |
description |
Cloud and snow detection is an important part of satellite remote sensing image processing, and also a key step for its following analysis and interpretation. In this paper, a simultaneous cloud and snow detection method for satellite remote sensing images based on ResNet and DeepLabV3+ fully convolutional neural network is proposed. The ResNet50 backbone is used, and the DeepLabV3+ network structure is optimized and improved according to the characteristics of cloud and snow on TH-1 remote sensing image. The ELU activation function, Adam gradient descent method and Focal Loss function are used to speed up convergence and improve segmentation accuracy. The network is trained and tested with the cloud and snow image dataset of TH-1 satellite. The experimental results show that it has strong robustness compared with Otsu method, and the detection accuracy of proposed method exceeds FCN-8s and original DeepLabV3+ network, meanwhile the detection speed of proposed method is faster than original DeepLabV3+, which can be applied to a variety of different remote sensing images through some adjustment and has favorable application prospects. |
topic |
satellite image cloud snow detection th-1 resnet deeplabv3+ |
url |
http://xb.sinomaps.com/article/2020/1001-1595/2020-10-1343.htm |
work_keys_str_mv |
AT zhengkai acloudandsnowdetectionmethodofth1imagebasedoncombinedresnetanddeeplabv3 AT lijiansheng acloudandsnowdetectionmethodofth1imagebasedoncombinedresnetanddeeplabv3 AT yangjianfeng acloudandsnowdetectionmethodofth1imagebasedoncombinedresnetanddeeplabv3 AT ouyangwen acloudandsnowdetectionmethodofth1imagebasedoncombinedresnetanddeeplabv3 AT wanggaojie acloudandsnowdetectionmethodofth1imagebasedoncombinedresnetanddeeplabv3 AT zhangxun acloudandsnowdetectionmethodofth1imagebasedoncombinedresnetanddeeplabv3 AT zhengkai cloudandsnowdetectionmethodofth1imagebasedoncombinedresnetanddeeplabv3 AT lijiansheng cloudandsnowdetectionmethodofth1imagebasedoncombinedresnetanddeeplabv3 AT yangjianfeng cloudandsnowdetectionmethodofth1imagebasedoncombinedresnetanddeeplabv3 AT ouyangwen cloudandsnowdetectionmethodofth1imagebasedoncombinedresnetanddeeplabv3 AT wanggaojie cloudandsnowdetectionmethodofth1imagebasedoncombinedresnetanddeeplabv3 AT zhangxun cloudandsnowdetectionmethodofth1imagebasedoncombinedresnetanddeeplabv3 |
_version_ |
1721204113998348288 |