An Image Fusion Method Based on Special Residual Network and Efficient Channel Attention
This paper presents an image fusion network based on a special residual network and attention mechanism. Compared with the traditional fusion network, the image fusion network has the advantages of an end-to-end network and integrates the feature extraction advantages of the attention mechanism resi...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2022-09-01
|
Series: | Electronics |
Subjects: | |
Online Access: | https://www.mdpi.com/2079-9292/11/19/3140 |
_version_ | 1797479791470837760 |
---|---|
author | Yang Li Haitao Yang Jinyu Wang Changgong Zhang Zhengjun Liu Hang Chen |
author_facet | Yang Li Haitao Yang Jinyu Wang Changgong Zhang Zhengjun Liu Hang Chen |
author_sort | Yang Li |
collection | DOAJ |
description | This paper presents an image fusion network based on a special residual network and attention mechanism. Compared with the traditional fusion network, the image fusion network has the advantages of an end-to-end network and integrates the feature extraction advantages of the attention mechanism residual network. It overcomes the shortcomings of the traditional network that need complex design rules and manual operation. In this method, hierarchical feature fusion is used to achieve effective fusion. A combined loss function is designed to optimize training results and improve image fusion quality. This paper uses many qualitative and quantitative experimental analyses on different data sets. The results show that, compared with the comparison algorithm, the method in this paper has a stronger retention ability of infrared and visible light information and better indexes. 72% of eleven indexes compared with some images in the public TNO data set are optimal or sub-optimal, and 80% are optimal or suboptimal in the RoadScene data set, which is much higher than other algorithms. The overall fusion effect is more in line with human visual perception. |
first_indexed | 2024-03-09T21:50:55Z |
format | Article |
id | doaj.art-68cced07c5e94feba890320785e6ba0a |
institution | Directory Open Access Journal |
issn | 2079-9292 |
language | English |
last_indexed | 2024-03-09T21:50:55Z |
publishDate | 2022-09-01 |
publisher | MDPI AG |
record_format | Article |
series | Electronics |
spelling | doaj.art-68cced07c5e94feba890320785e6ba0a2023-11-23T20:06:59ZengMDPI AGElectronics2079-92922022-09-011119314010.3390/electronics11193140An Image Fusion Method Based on Special Residual Network and Efficient Channel AttentionYang Li0Haitao Yang1Jinyu Wang2Changgong Zhang3Zhengjun Liu4Hang Chen5School of Space Information, Space Engineering University, Beijing 101416, ChinaSpace Security Research Center, Space Engineering University, Beijing 101416, ChinaSchool of Space Information, Space Engineering University, Beijing 101416, ChinaSchool of Space Information, Space Engineering University, Beijing 101416, ChinaSchool of Physics, Harbin Institute of Technology, Harbin 150001, ChinaSchool of Space Information, Space Engineering University, Beijing 101416, ChinaThis paper presents an image fusion network based on a special residual network and attention mechanism. Compared with the traditional fusion network, the image fusion network has the advantages of an end-to-end network and integrates the feature extraction advantages of the attention mechanism residual network. It overcomes the shortcomings of the traditional network that need complex design rules and manual operation. In this method, hierarchical feature fusion is used to achieve effective fusion. A combined loss function is designed to optimize training results and improve image fusion quality. This paper uses many qualitative and quantitative experimental analyses on different data sets. The results show that, compared with the comparison algorithm, the method in this paper has a stronger retention ability of infrared and visible light information and better indexes. 72% of eleven indexes compared with some images in the public TNO data set are optimal or sub-optimal, and 80% are optimal or suboptimal in the RoadScene data set, which is much higher than other algorithms. The overall fusion effect is more in line with human visual perception.https://www.mdpi.com/2079-9292/11/19/3140codec networkdeep learningimage fusionattention mechanism |
spellingShingle | Yang Li Haitao Yang Jinyu Wang Changgong Zhang Zhengjun Liu Hang Chen An Image Fusion Method Based on Special Residual Network and Efficient Channel Attention Electronics codec network deep learning image fusion attention mechanism |
title | An Image Fusion Method Based on Special Residual Network and Efficient Channel Attention |
title_full | An Image Fusion Method Based on Special Residual Network and Efficient Channel Attention |
title_fullStr | An Image Fusion Method Based on Special Residual Network and Efficient Channel Attention |
title_full_unstemmed | An Image Fusion Method Based on Special Residual Network and Efficient Channel Attention |
title_short | An Image Fusion Method Based on Special Residual Network and Efficient Channel Attention |
title_sort | image fusion method based on special residual network and efficient channel attention |
topic | codec network deep learning image fusion attention mechanism |
url | https://www.mdpi.com/2079-9292/11/19/3140 |
work_keys_str_mv | AT yangli animagefusionmethodbasedonspecialresidualnetworkandefficientchannelattention AT haitaoyang animagefusionmethodbasedonspecialresidualnetworkandefficientchannelattention AT jinyuwang animagefusionmethodbasedonspecialresidualnetworkandefficientchannelattention AT changgongzhang animagefusionmethodbasedonspecialresidualnetworkandefficientchannelattention AT zhengjunliu animagefusionmethodbasedonspecialresidualnetworkandefficientchannelattention AT hangchen animagefusionmethodbasedonspecialresidualnetworkandefficientchannelattention AT yangli imagefusionmethodbasedonspecialresidualnetworkandefficientchannelattention AT haitaoyang imagefusionmethodbasedonspecialresidualnetworkandefficientchannelattention AT jinyuwang imagefusionmethodbasedonspecialresidualnetworkandefficientchannelattention AT changgongzhang imagefusionmethodbasedonspecialresidualnetworkandefficientchannelattention AT zhengjunliu imagefusionmethodbasedonspecialresidualnetworkandefficientchannelattention AT hangchen imagefusionmethodbasedonspecialresidualnetworkandefficientchannelattention |