A Dense Encoder–Decoder Network with Feedback Connections for Pan-Sharpening
To meet the need for multispectral images having high spatial resolution in practical applications, we propose a dense encoder–decoder network with feedback connections for pan-sharpening. Our network consists of four parts. The first part consists of two identical subnetworks, one each to extract f...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-11-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/13/22/4505 |
| _version_ | 1797508604514795520 |
|---|---|
| author | Weisheng Li Minghao Xiang Xuesong Liang |
| author_facet | Weisheng Li Minghao Xiang Xuesong Liang |
| author_sort | Weisheng Li |
| collection | DOAJ |
| description | To meet the need for multispectral images having high spatial resolution in practical applications, we propose a dense encoder–decoder network with feedback connections for pan-sharpening. Our network consists of four parts. The first part consists of two identical subnetworks, one each to extract features from PAN and MS images, respectively. The second part is an efficient feature-extraction block. We hope that the network can focus on features at different scales, so we propose innovative multiscale feature-extraction blocks that fully extract effective features from networks of various depths and widths by using three multiscale feature-extraction blocks and two long-jump connections. The third part is the feature fusion and recovery network. We are inspired by the work on U-Net network improvements to propose a brand new encoder network structure with dense connections that improves network performance through effective connections to encoders and decoders at different scales. The fourth part is a continuous feedback connection operation with overfeedback to refine shallow features, which enables the network to obtain better reconstruction capabilities earlier. To demonstrate the effectiveness of our method, we performed several experiments. Experiments on various satellite datasets show that the proposed method outperforms existing methods. Our results show significant improvements over those from other models in terms of the multiple-target index values used to measure the spectral quality and spatial details of the generated images. |
| first_indexed | 2024-03-10T05:06:18Z |
| format | Article |
| id | doaj.art-a185dde662d44c2b930d3d38098e3ac6 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T05:06:18Z |
| publishDate | 2021-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-a185dde662d44c2b930d3d38098e3ac62023-11-23T01:18:23ZengMDPI AGRemote Sensing2072-42922021-11-011322450510.3390/rs13224505A Dense Encoder–Decoder Network with Feedback Connections for Pan-SharpeningWeisheng Li0Minghao Xiang1Xuesong Liang2College of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaCollege of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaCollege of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaTo meet the need for multispectral images having high spatial resolution in practical applications, we propose a dense encoder–decoder network with feedback connections for pan-sharpening. Our network consists of four parts. The first part consists of two identical subnetworks, one each to extract features from PAN and MS images, respectively. The second part is an efficient feature-extraction block. We hope that the network can focus on features at different scales, so we propose innovative multiscale feature-extraction blocks that fully extract effective features from networks of various depths and widths by using three multiscale feature-extraction blocks and two long-jump connections. The third part is the feature fusion and recovery network. We are inspired by the work on U-Net network improvements to propose a brand new encoder network structure with dense connections that improves network performance through effective connections to encoders and decoders at different scales. The fourth part is a continuous feedback connection operation with overfeedback to refine shallow features, which enables the network to obtain better reconstruction capabilities earlier. To demonstrate the effectiveness of our method, we performed several experiments. Experiments on various satellite datasets show that the proposed method outperforms existing methods. Our results show significant improvements over those from other models in terms of the multiple-target index values used to measure the spectral quality and spatial details of the generated images.https://www.mdpi.com/2072-4292/13/22/4505convolutional neural networkdouble-stream structurefeedbackencoder–decoder networkdense connections |
| spellingShingle | Weisheng Li Minghao Xiang Xuesong Liang A Dense Encoder–Decoder Network with Feedback Connections for Pan-Sharpening Remote Sensing convolutional neural network double-stream structure feedback encoder–decoder network dense connections |
| title | A Dense Encoder–Decoder Network with Feedback Connections for Pan-Sharpening |
| title_full | A Dense Encoder–Decoder Network with Feedback Connections for Pan-Sharpening |
| title_fullStr | A Dense Encoder–Decoder Network with Feedback Connections for Pan-Sharpening |
| title_full_unstemmed | A Dense Encoder–Decoder Network with Feedback Connections for Pan-Sharpening |
| title_short | A Dense Encoder–Decoder Network with Feedback Connections for Pan-Sharpening |
| title_sort | dense encoder decoder network with feedback connections for pan sharpening |
| topic | convolutional neural network double-stream structure feedback encoder–decoder network dense connections |
| url | https://www.mdpi.com/2072-4292/13/22/4505 |
| work_keys_str_mv | AT weishengli adenseencoderdecodernetworkwithfeedbackconnectionsforpansharpening AT minghaoxiang adenseencoderdecodernetworkwithfeedbackconnectionsforpansharpening AT xuesongliang adenseencoderdecodernetworkwithfeedbackconnectionsforpansharpening AT weishengli denseencoderdecodernetworkwithfeedbackconnectionsforpansharpening AT minghaoxiang denseencoderdecodernetworkwithfeedbackconnectionsforpansharpening AT xuesongliang denseencoderdecodernetworkwithfeedbackconnectionsforpansharpening |