HRRNet: Hierarchical Refinement Residual Network for Semantic Segmentation of Remote Sensing Images
Semantic segmentation of high-resolution remote sensing images plays an important role in many practical applications, including precision agriculture and natural disaster assessment. With the emergence of a large number of studies on convolutional neural networks, the performance of the semantic se...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-02-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/15/5/1244 |
| _version_ | 1827752337879334912 |
|---|---|
| author | Shiwei Cheng Baozhu Li Le Sun Yuwen Chen |
| author_facet | Shiwei Cheng Baozhu Li Le Sun Yuwen Chen |
| author_sort | Shiwei Cheng |
| collection | DOAJ |
| description | Semantic segmentation of high-resolution remote sensing images plays an important role in many practical applications, including precision agriculture and natural disaster assessment. With the emergence of a large number of studies on convolutional neural networks, the performance of the semantic segmentation model of remote sensing images has been dramatically promoted. However, many deep convolutional network models do not fully refine the segmentation result maps, and, in addition, the contextual dependencies of the semantic feature map have not been adequately exploited. This article proposes a hierarchical refinement residual network (HRRNet) to address these issues. The HRRNet mainly consists of ResNet50 as the backbone, attention blocks, and decoders. The attention block consists of a channel attention module (CAM) and a pooling residual attention module (PRAM) and residual structures. Specifically, the feature map output by the four blocks of Resnet50 is passed through the attention block to fully explore the contextual dependencies of the position and channel of the semantic feature map, and, then, the feature maps of each branch are fused step by step to realize the refinement of the feature maps, thereby improving the segmentation performance of the proposed HRRNet. Experiments show that the proposed HRRNet improves segmentation result maps compared with various state-of-the-art networks on Vaihingen and Potsdam datasets. |
| first_indexed | 2024-03-11T07:12:34Z |
| format | Article |
| id | doaj.art-dc2142725c77495c9f32e55d5896e76a |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-11T07:12:34Z |
| publishDate | 2023-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-dc2142725c77495c9f32e55d5896e76a2023-11-17T08:30:24ZengMDPI AGRemote Sensing2072-42922023-02-01155124410.3390/rs15051244HRRNet: Hierarchical Refinement Residual Network for Semantic Segmentation of Remote Sensing ImagesShiwei Cheng0Baozhu Li1Le Sun2Yuwen Chen3School of Computer Science, Nanjing University of Information Science and Technology, Nanjing 210044, ChinaInternet of Things & Smart City Innovation Platform, Zhuhai Fudan Innovation Institute, Zhuhai 519031, ChinaSchool of Computer Science, Nanjing University of Information Science and Technology, Nanjing 210044, ChinaChongqing Institute of Green and Intelligent Technology, Chinese Academy of Sciences (CAS), Chongqing 400714, ChinaSemantic segmentation of high-resolution remote sensing images plays an important role in many practical applications, including precision agriculture and natural disaster assessment. With the emergence of a large number of studies on convolutional neural networks, the performance of the semantic segmentation model of remote sensing images has been dramatically promoted. However, many deep convolutional network models do not fully refine the segmentation result maps, and, in addition, the contextual dependencies of the semantic feature map have not been adequately exploited. This article proposes a hierarchical refinement residual network (HRRNet) to address these issues. The HRRNet mainly consists of ResNet50 as the backbone, attention blocks, and decoders. The attention block consists of a channel attention module (CAM) and a pooling residual attention module (PRAM) and residual structures. Specifically, the feature map output by the four blocks of Resnet50 is passed through the attention block to fully explore the contextual dependencies of the position and channel of the semantic feature map, and, then, the feature maps of each branch are fused step by step to realize the refinement of the feature maps, thereby improving the segmentation performance of the proposed HRRNet. Experiments show that the proposed HRRNet improves segmentation result maps compared with various state-of-the-art networks on Vaihingen and Potsdam datasets.https://www.mdpi.com/2072-4292/15/5/1244deep convolution convolutional neural networkattention mechanismsemantic segmentationremote sensing imagesresidual structure |
| spellingShingle | Shiwei Cheng Baozhu Li Le Sun Yuwen Chen HRRNet: Hierarchical Refinement Residual Network for Semantic Segmentation of Remote Sensing Images Remote Sensing deep convolution convolutional neural network attention mechanism semantic segmentation remote sensing images residual structure |
| title | HRRNet: Hierarchical Refinement Residual Network for Semantic Segmentation of Remote Sensing Images |
| title_full | HRRNet: Hierarchical Refinement Residual Network for Semantic Segmentation of Remote Sensing Images |
| title_fullStr | HRRNet: Hierarchical Refinement Residual Network for Semantic Segmentation of Remote Sensing Images |
| title_full_unstemmed | HRRNet: Hierarchical Refinement Residual Network for Semantic Segmentation of Remote Sensing Images |
| title_short | HRRNet: Hierarchical Refinement Residual Network for Semantic Segmentation of Remote Sensing Images |
| title_sort | hrrnet hierarchical refinement residual network for semantic segmentation of remote sensing images |
| topic | deep convolution convolutional neural network attention mechanism semantic segmentation remote sensing images residual structure |
| url | https://www.mdpi.com/2072-4292/15/5/1244 |
| work_keys_str_mv | AT shiweicheng hrrnethierarchicalrefinementresidualnetworkforsemanticsegmentationofremotesensingimages AT baozhuli hrrnethierarchicalrefinementresidualnetworkforsemanticsegmentationofremotesensingimages AT lesun hrrnethierarchicalrefinementresidualnetworkforsemanticsegmentationofremotesensingimages AT yuwenchen hrrnethierarchicalrefinementresidualnetworkforsemanticsegmentationofremotesensingimages |