MFCSNet: Multi-Scale Deep Features Fusion and Cost-Sensitive Loss Function Based Segmentation Network for Remote Sensing Images
Semantic segmentation of remote sensing images is an important technique for spatial analysis and geocomputation. It has important applications in the fields of military reconnaissance, urban planning, resource utilization and environmental monitoring. In order to accurately perform semantic segment...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2019-09-01
|
Series: | Applied Sciences |
Subjects: | |
Online Access: | https://www.mdpi.com/2076-3417/9/19/4043 |
id |
doaj-b6235da1e4eb4000953122989ac33a5d |
---|---|
record_format |
Article |
spelling |
doaj-b6235da1e4eb4000953122989ac33a5d2020-11-25T02:32:55ZengMDPI AGApplied Sciences2076-34172019-09-01919404310.3390/app9194043app9194043MFCSNet: Multi-Scale Deep Features Fusion and Cost-Sensitive Loss Function Based Segmentation Network for Remote Sensing ImagesEnde Wang0Yanmei Jiang1Yong Li2Jingchao Yang3Mengcheng Ren4Qingchun Zhang5Key Laboratory of Optical Electrical Image Processing, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110016, ChinaDepartment of Electrical and Information Engineering, Hebei Jiaotong Vocational and Technical College, Shijiazhuang 050000, ChinaKey Laboratory of Optical Electrical Image Processing, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang 110016, ChinaDepartment of Electrical and Information Engineering, Hebei Jiaotong Vocational and Technical College, Shijiazhuang 050000, ChinaCollege of Information Science and Engineering, Northeastern University, Shenyang 110819, ChinaCollege of Information Science and Engineering, Northeastern University, Shenyang 110819, ChinaSemantic segmentation of remote sensing images is an important technique for spatial analysis and geocomputation. It has important applications in the fields of military reconnaissance, urban planning, resource utilization and environmental monitoring. In order to accurately perform semantic segmentation of remote sensing images, we proposed a novel multi-scale deep features fusion and cost-sensitive loss function based segmentation network, named MFCSNet. To acquire the information of different levels in remote sensing images, we design a multi-scale feature encoding and decoding structure, which can fuse the low-level and high-level semantic information. Then a max-pooling indices up-sampling structure is designed to improve the recognition rate of the object edge and location information in the remote sensing image. In addition, the cost-sensitive loss function is designed to improve the classification accuracy of objects with fewer samples. The penalty coefficient of misclassification is designed to improve the robustness of the network model, and the batch normalization layer is also added to make the network converge faster. The experimental results show that the classification performance of MFCSNet outperforms U-Net and SegNet in classification accuracy, object details and prediction consistency.https://www.mdpi.com/2076-3417/9/19/4043semantic segmentationremote sensing imagesfeature fusioncost-sensitive |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Ende Wang Yanmei Jiang Yong Li Jingchao Yang Mengcheng Ren Qingchun Zhang |
spellingShingle |
Ende Wang Yanmei Jiang Yong Li Jingchao Yang Mengcheng Ren Qingchun Zhang MFCSNet: Multi-Scale Deep Features Fusion and Cost-Sensitive Loss Function Based Segmentation Network for Remote Sensing Images Applied Sciences semantic segmentation remote sensing images feature fusion cost-sensitive |
author_facet |
Ende Wang Yanmei Jiang Yong Li Jingchao Yang Mengcheng Ren Qingchun Zhang |
author_sort |
Ende Wang |
title |
MFCSNet: Multi-Scale Deep Features Fusion and Cost-Sensitive Loss Function Based Segmentation Network for Remote Sensing Images |
title_short |
MFCSNet: Multi-Scale Deep Features Fusion and Cost-Sensitive Loss Function Based Segmentation Network for Remote Sensing Images |
title_full |
MFCSNet: Multi-Scale Deep Features Fusion and Cost-Sensitive Loss Function Based Segmentation Network for Remote Sensing Images |
title_fullStr |
MFCSNet: Multi-Scale Deep Features Fusion and Cost-Sensitive Loss Function Based Segmentation Network for Remote Sensing Images |
title_full_unstemmed |
MFCSNet: Multi-Scale Deep Features Fusion and Cost-Sensitive Loss Function Based Segmentation Network for Remote Sensing Images |
title_sort |
mfcsnet: multi-scale deep features fusion and cost-sensitive loss function based segmentation network for remote sensing images |
publisher |
MDPI AG |
series |
Applied Sciences |
issn |
2076-3417 |
publishDate |
2019-09-01 |
description |
Semantic segmentation of remote sensing images is an important technique for spatial analysis and geocomputation. It has important applications in the fields of military reconnaissance, urban planning, resource utilization and environmental monitoring. In order to accurately perform semantic segmentation of remote sensing images, we proposed a novel multi-scale deep features fusion and cost-sensitive loss function based segmentation network, named MFCSNet. To acquire the information of different levels in remote sensing images, we design a multi-scale feature encoding and decoding structure, which can fuse the low-level and high-level semantic information. Then a max-pooling indices up-sampling structure is designed to improve the recognition rate of the object edge and location information in the remote sensing image. In addition, the cost-sensitive loss function is designed to improve the classification accuracy of objects with fewer samples. The penalty coefficient of misclassification is designed to improve the robustness of the network model, and the batch normalization layer is also added to make the network converge faster. The experimental results show that the classification performance of MFCSNet outperforms U-Net and SegNet in classification accuracy, object details and prediction consistency. |
topic |
semantic segmentation remote sensing images feature fusion cost-sensitive |
url |
https://www.mdpi.com/2076-3417/9/19/4043 |
work_keys_str_mv |
AT endewang mfcsnetmultiscaledeepfeaturesfusionandcostsensitivelossfunctionbasedsegmentationnetworkforremotesensingimages AT yanmeijiang mfcsnetmultiscaledeepfeaturesfusionandcostsensitivelossfunctionbasedsegmentationnetworkforremotesensingimages AT yongli mfcsnetmultiscaledeepfeaturesfusionandcostsensitivelossfunctionbasedsegmentationnetworkforremotesensingimages AT jingchaoyang mfcsnetmultiscaledeepfeaturesfusionandcostsensitivelossfunctionbasedsegmentationnetworkforremotesensingimages AT mengchengren mfcsnetmultiscaledeepfeaturesfusionandcostsensitivelossfunctionbasedsegmentationnetworkforremotesensingimages AT qingchunzhang mfcsnetmultiscaledeepfeaturesfusionandcostsensitivelossfunctionbasedsegmentationnetworkforremotesensingimages |
_version_ |
1724816856585338880 |