最全图像融合论文及代码整理
News
[2022-07-29] 我们的综述论文《基于深度学习的图像融合方法综述》被《中国图象图形学报》正式接收！[论文下载]
Github项目地址：https://github.com/Linfeng-Tang/Image-Fusion （欢迎大家Start、 Fork、 Fellow一键三连哦~）
图像融合(Image Fusion)
- 多模图像融合(Multi-Modal Image Fusion)
- 红外和可见光图像融合(Infrared and visible image fusion)
- 医学图像融合(Medical image fusion)
- 数字摄影图像融合(Digital Photography Image Fusion)
- 多曝光图像融合(Multi-exposure image fusion)
- 多聚焦图像融合(Multi-focus image fusion)
- 遥感影像融合(Remote Sensing Image Fusion)
- 全色图像锐化(Pansharpening)
- 通用图像融合框架(General Image Fusion Framerwork)
- 综述(Survey)
- 数据集(Dataset)
- 评估指标(Evaluation Metric)
- 通用评估指标(General evaluation metric)
- 遥感影像融合评估指标
- Citation
图像融合系列博客还有：
- 图像融合综述论文整理参见：图像融合综述论文整理
- 图像融合评估指标参见：红外和可见光图像融合评估指标
- 图像融合常用数据集整理参见：图像融合常用数据集整理
- 通用图像融合框架论文及代码整理参见：通用图像融合框架论文及代码整理
- 基于深度学习的红外和可见光图像融合论文及代码整理参见:基于深度学习的红外和可见光图像融合论文及代码整理
- 更加详细的红外和可见光图像融合代码参见：红外和可见光图像融合论文及代码整理
- 基于深度学习的多曝光图像融合论文及代码整理参见：基于深度学习的多曝光图像融合论文及代码整理
- 基于深度学习的多聚焦图像融合论文及代码整理参见：基于深度学习的多聚焦图像融合(Multi-focus Image Fusion)论文及代码整理
- 基于深度学习的全色图像锐化论文及代码整理参见：基于深度学习的全色图像锐化(Pansharpening)论文及代码整理
- 基于深度学习的医学图像融合论文及代码整理参见：基于深度学习的医学图像融合(Medical image fusion)论文及代码整理
- 彩色图像融合程序参见：彩色图像融合
- SeAFusion:首个结合高级视觉任务的图像融合框架参见：SeAFusion:首个结合高级视觉任务的图像融合框架
- DIVFusion:首个耦合互促低光增强&图像融合的框架参见：DIVFusion:首个耦合互促低光增强&图像融合的框架
多模图像融合(Multi-Modal Image Fusion)
红外和可见光图像融合(Infrared and visible image fusion)
| 方法 | 标题 | 论文 | 代码 | 发表期刊或会议 | 基础框架 | 监督范式 | 发表年份 |
|---|---|---|---|---|---|---|---|
| DenseFuse | DenseFuse: A Fusion Approach to Infrared and Visible Images | Paper | Code | TIP | AE | 自监督 | 2019 |
| FusionGAN | FusionGAN: A generative adversarial network for infrared andvisible image fusion | Paper | Code | InfFus | GAN | 无监督 | 2019 |
| DDcGAN | Learning a Generative Model for Fusing Infrared and VisibleImages via Conditional Generative Adversarial Network with DualDiscriminators | Paper | Code | IJCAI | GAN | 无监督 | 2019 |
| NestFuse | NestFuse: An Infrared and Visible Image Fusion ArchitectureBased on Nest Connection and Spatial/Channel Attention Models | Paper | Code | TIM | AE | 自监督 | 2020 |
| DDcGAN | DDcGAN: A dual-discriminator conditional generativeadversarial network for multi-resolution image fusion | Paper | Code | TIP | GAN | 无监督 | 2020 |
| RFN-Nest | RFN-Nest: An end-to-end residual fusion network for infraredand visible images | Paper | Code | InfFus | AE | 自监督 | 2021 |
| CSF | Classification Saliency-Based Rule for Visible and InfraredImage Fusion | Paper | Code | TCI | AE | 自监督 | 2021 |
| DRF | DRF: Disentangled Representation for Visible and InfraredImage Fusion | Paper | Code | TIM | AE | 自监督 | 2021 |
| SEDRFuse | SEDRFuse: A Symmetric Encoder–Decoder With Residual BlockNetwork for Infrared and Visible Image Fusion | Paper | Code | TIM | AE | 自监督 | 2021 |
| MFEIF | Learning a Deep Multi-Scale Feature Ensemble and anEdge-Attention Guidance for Image Fusion | Paper | Code | TCSVT | AE | 自监督 | 2021 |
| Meta-Learning | Different Input Resolutions and Arbitrary Output Resolution: AMeta Learning-Based Deep Framework for Infrared and Visible Image Fusion | Paper | TIP | CNN | 无监督 | 2021 | |
| RXDNFuse | RXDNFuse: A aggregated residual dense network for infrared andvisible image fusion | Paper | InfFus | CNN | 无监督 | 2021 | |
| STDFusionNet | STDFusionNet: An Infrared and Visible Image Fusion NetworkBased on Salient Target Detection | Paper | Code | TIM | CNN | 无监督 | 2021 |
| D2LE | A Bilevel Integrated Model With Data-Driven Layer Ensemble forMulti-Modality Image Fusion | Paper | TIP | CNN | 无监督 | 2021 | |
| HAF | Searching a Hierarchically Aggregated Fusion Architecture forFast Multi-Modality Image Fusion | Paper | Code | ACM MM | CNN | 无监督 | 2021 |
| SDDGAN | Semantic-supervised Infrared and Visible Image Fusion via aDual-discriminator Generative Adversarial Network | Paper | Code | TMM | GAN | 无监督 | 2021 |
| Detail-GAN | Infrared and visible image fusion via detail preservingadversarial learning | Paper | Code | InfFus | GAN | 无监督 | 2021 |
| Perception-GAN | Image fusion based ongenerative adversarial network consistent with perception | Paper | Code | InfFus | GAN | 无监督 | 2021 |
| GAN-FM | GAN-FM: Infrared and VisibleImage Fusion Using GAN With Full-Scale Skip Connection and Dual MarkovianDiscriminators | Paper | Code | TCI | GAN | 无监督 | 2021 |
| AttentionFGAN | AttentionFGAN: Infrared and Visible Image Fusion UsingAttention-Based Generative Adversarial Networks | Paper | TMM | GAN | 无监督 | 2021 | |
| GANMcC | GANMcC: A GenerativeAdversarial Network With Multiclassification Constraints for Infrared andVisible Image Fusion | Paper | Code | TIM | GAN | 无监督 | 2021 |
| MgAN-Fuse | Multigrained Attention Network for Infrared and Visible ImageFusion | Paper | TIM | GAN | 无监督 | 2021 | |
| TC-GAN | Infrared and Visible ImageFusion via Texture Conditional Generative Adversarial Network | Paper | TCSVT | GAN | 无监督 | 2021 | |
| TarDAL | Target-aware Dual Adversarial Learning and a Multi-scenario Multi-Modality Benchmark to Fuse Infrared and Visible for Object Detection | Paper | Code | CVPR | GAN | 无监督 | 2022 |
| RFNet | RFNet: Unsupervised Network for Mutually Reinforcing Multi-modal Image Registration and Fusion | Paper | Code | CVPR | CNN | 无监督 | 2022 |
| SeAFusion | Image fusion in the loop ofhigh-level vision tasks: A semantic-aware real-time infrared and visibleimage fusion network | Paper | Code | InfFus | CNN | 无监督 | 2022 |
| PIAFusion | PIAFusion: A progressive infrared and visible image fusionnetwork based on illumination aware | Paper | Code | InfFus | CNN | 无监督 | 2022 |
| UMF-CMGR | Unsupervised Misaligned Infrared and Visible Image Fusion via Cross-Modality Image Generation and Registration | Paper | Code | IJCAI | CNN | 无监督 | 2022 |
| DetFusion | DetFusion: A Detection-driven Infrared and Visible Image Fusion Network | Paper | Code | ACM MM | CNN | 无监督 | 2022 |
| DIVFusion | DIVFusion: Darkness-free infrared and visible image fusion | Paper | Code | InfFus | CNN | 无监督 | 2023 |
医学图像融合(Medical image fusion)
| 方法 | 标题 | 论文 | 代码 | 发表期刊或会议 | 基础框架 | 监督范式 | 年份 |
|---|---|---|---|---|---|---|---|
| CNN | A medical image fusion method based onconvolutional neural networks | Paper | ICIF | CNN | 无监督 | 2017 | |
| Zero-LMF | Zero-Learning Fast Medical Image Fusion | Paper | Code | ICIF | CNN | 无监督 | 2019 |
| DDcGAN | Learning a Generative Model for Fusing Infrared and VisibleImages via Conditional Generative Adversarial Network with DualDiscriminators | Paper | Code | IJCAI | GAN | 无监督 | 2019 |
| GFPPC-GAN | Green Fluorescent Protein and Phase-ContrastImage Fusion via Generative Adversarial Networks | Paper | CMMM | GAN | 无监督 | 2019 | |
| CCN-CP | Multi-modality medical image fusion using convolutional neuralnetwork and contrast pyramid | Paper | Sensors | CNN | 无监督 | 2020 | |
| DDcGAN | DDcGAN: A Dual-Discriminator ConditionalGenerative Adversarial Network for Multi-Resolution Image Fusion | Paper | Code | TIP | GAN | 无监督 | 2020 |
| MGMDcGAN | Medical Image Fusion Using Multi-GeneratorMulti-Discriminator Conditional Generative Adversarial Network | Paper | Code | Access | GAN | 无监督 | 2020 |
| D2LE | A Bilevel Integrated Model With Data-Driven Layer Ensemble forMulti-Modality Image Fusion | Paper | TIP | CNN | 无监督 | 2021 | |
| HAF | Searching a Hierarchically Aggregated FusionArchitecture for Fast Multi-Modality Image Fusion | Paper | Code | ACM MM | CNN | 无监督 | 2021 |
| EMFusion | EMFusion: An unsupervised enhanced medicalimage fusion network | Paper | Code | InfFus | CNN | 无监督 | 2021 |
| DPCN-Fusion | Green Fluorescent Protein and Phase ContrastImage Fusion Via Detail Preserving Cross Network | Paper | Code | TCI | CNN | 无监督 | 2021 |
| MSPRN | A multiscale residual pyramid attentionnetwork for medical image fusion | Paper | Code | BSPC | CNN | 无监督 | 2021 |
| DCGAN | Medical image fusion method based on denseblock and deep convolutional generative adversarial network | Paper | NCA | GAN | 无监督 | 2021 |
数字摄影图像融合(Digital Photography Image Fusion)
多曝光图像融合(Multi-exposure image fusion)
| 方法 | 标题 | 论文 | 代码 | 发表期刊或会议 | 基础框架 | 监督范式 | 年份 |
|---|---|---|---|---|---|---|---|
| DeepFuse | DeepFuse: A Deep Unsupervised Approach for Exposure Fusionwith Extreme Exposure Image Pairs | Paper | Code | ICCV | CNN | 无监督 | 2017 |
| CNN | Multi-exposure fusion with CNN features | Paper | Code | ICIP | CNN | 无监督 | 2018 |
| MEF-Net | Deep guided learning for fast multi-exposure image fusion | Paper | Code | TIP | CNN | 无监督 | 2020 |
| ICEN | Multi-exposure high dynamic range imaging with informativecontent enhanced network | Paper | NC | CNN | 无监督 | 2020 | |
| MEF-GAN | MEF-GAN: Multi-Exposure Image Fusion via GenerativeAdversarial Networks | Paper | Code | TIP | GAN | 无监督 | 2020 |
| CF-Net | Deep coupled feedback network for joint exposure fusion andimage super-resolutions | Paper | Code | TIP | CNN | 无监督 | 2021 |
| UMEF | Deep unsupervised learning based on color un-referenced lossfunctions for multi-exposure image fusion | Paper | Code | InFus | CNN | 无监督 | 2021 |
| PA-AGN | Two exposure fusion using prior-aware generative adversarialnetwork | Paper | TMM | GAN | 无监督 | 2021 | |
| AGAL | Attention-guided Global-local Adversarial Learning forDetail-preserving Multi-exposure Image Fusion | Paper | Code | TCSVT | GAN | 无监督 | 2022 |
| GANFuse | GANFuse: a novel multi-exposure image fusion method based ongenerative adversarial networks | Paper | NCAA | GAN | 无监督 | 2021 | |
| DRLF | Automatic Intermediate Generation With Deep ReinforcementLearning for Robust Two-Exposure Image Fusion | Paper | TNNLS | CNN | 无监督 | 2021 | |
| Trans-MEF | TransMEF: A Transformer-Based Multi-Exposure Image FusionFramework using Self-Supervised Multi-Task Learning | Paper | Code | AAAI | AE | 自监督 | 2022 |
| DPE-MEF | Multi-exposure image fusion via deep perceptual enhancement | Paper | Code | InFus | CNN | 无监督 | 2022 |
多聚焦图像融合(Multi-focus image fusion)
| 方法 | 标题 | 论文 | 代码 | 发表期刊或会议 | 基础框架 | 监督范式 | 年份 |
|---|---|---|---|---|---|---|---|
| CNN | Multi-focus image fusion with a deep convolutional neuralnetwork | Paper | Code | InFus | CNN | 有监督 | 2017 |
| ECNN | Ensemble of CNN for multi-focus image fusion | Paper | Code | InFus | CNN | 有监督 | 2019 |
| MLFCNN | Multilevel features convolutional neural network formultifocus image fusion | Paper | TCI | CNN | 有监督 | 2019 | |
| DRPL | DRPL: Deep Regression Pair Learning for Multi-Focus ImageFusion | Paper | Code | TIP | CNN | 有监督 | 2020 |
| MMF-Net | An α-Matte Boundary Defocus Model-Based Cascaded Network forMulti-Focus Image Fusion | Paper | Code | TCI | CNN | 有监督 | 2020 |
| MFF-SSIM | Towards Reducing Severe Defocus Spread Effects for Multi-FocusImage Fusion via an Optimization Based Strategy | Paper | Code | Sensors | CNN | 无监督 | 2020 |
| MFNet | Structural Similarity Loss for Learning to Fuse Multi-FocusImages | Paper | TIP | CNN | 有监督 | 2021 | |
| GEU-Net | Global-Feature Encoding U-Net (GEU-Net) for Multi-Focus ImageFusion [GEU-Net | Paper | Code | TCI | CNN | 自监督 | 2021 |
| DTMNet | DTMNet: A Discrete Tchebichef Moments-Based Deep NeuralNetwork for Multi-Focus Image Fusion | Paper | TMM | CNN | 无监督 | 2021 | |
| SMFuse | SMFuse: Multi-FocusImage Fusion Via Self-Supervised Mask-Optimization | Paper | Code | NCA | CNN | 无监督 | 2021 |
| ACGAN | A generative adversarial network with adaptive constraints formulti-focus image fusion | Paper | Code | ICCV | GAN | 有监督 | 2021 |
| FuseGAN | Learning to fuse multi-focus image via conditional generativeadversarial network | Paper | TIP | GAN | 有监督 | 2020 | |
| D2FMIF | Depth-Distilled Multi-focus Image Fusion | Paper | TMM | CNN | 有监督 | 2019 | |
| SESF-Fuse | SESF-Fuse: an unsupervised deep model for multi-focus imagefusion | Paper | Code | NCAA | CNN | 有监督 | 2020 |
| MFF-GAN | MFF-GAN: An unsupervised generative adversarial network withadaptive and gradient joint constraints for multi-focus image fusion | Paper | Code | InFus | GAN | 无监督 | 2021 |
| MFIF-GAN | MFIF-GAN: A new generative adversarial network for multi-focusimage fusion | Paper | Code | SPIC | GAN | 有监督 | 2021 |
遥感影像融合(Remote Sensing Image Fusion)
全色图像锐化(Pansharpening)
| 方法 | 标题 | 论文 | 代码 | 发表期刊或会议 | 基础框架 | 监督范式 | 年份 |
|---|---|---|---|---|---|---|---|
| PNN | Pansharpening by Convolutional NeuralNetworks | Paper | Code | RS | CNN | 有监督 | 2016 |
| PanNet | PanNet: A deep network architecture forpan-sharpening | Paper | Code | PanNet | CNN | 有监督 | 2017 |
| TFNet | Remote sensing image fusion based ontwo-stream fusion network | Paper | Code | TFNet | CNN | 有监督 | 2020 |
| BKL | Unsupervised Blur Kernel Learning forPansharpening | Paper | IGARSS | CNN | 无监督 | 2020 | |
| Pan-GAN | Pan-GAN: An unsupervised pan-sharpeningmethod for remote sensing image fusion | Paper | Code | InFus | GAN | 无监督 | 2020 |
| UCNN | Pansharpening via Unsupervised ConvolutionalNeural Networks | Paper | JSTARS | CNN | 无监督 | 2020 | |
| UPSNet | UPSNet: Unsupervised Pan-Sharpening NetworkWith Registration Learning Between Panchromatic and Multi-Spectral Images | Paper | ACCESS | CNN | 无监督 | 2020 | |
| GPPNN | Deep Gradient Projection Networks forPan-sharpening | Paper | Code | CVPR | CNN | 有监督 | 2021 |
| GTP-PNet | GTP-PNet: A residual learning network basedon gradient transformation prior for pansharpening | Paper | Code | ISPRS | CNN | 有监督 | 2021 |
| HMCNN | Pan-Sharpening Via High-Pass ModificationConvolutional Neural Network | Paper | Code | ICIP | CNN | 有监督 | 2021 |
| SDPNet | SDPNet: A Deep Network for Pan-SharpeningWith Enhanced Information Representation | Paper | Code | TGRS | CNN | 有监督 | 2021 |
| SIPSA-Net | SIPSA-Net: Shift-Invariant Pan Sharpeningwith Moving Object Alignment for Satellite Imagery | Paper | Code | CVPR | CNN | 有监督 | 2021 |
| SRPPNN | Super-resolution-guided progressivepansharpening based on a deep convolutional neural network | Paper | Code | TGRS | CNN | 有监督 | 2021 |
| PSGAN | PSGAN: A generative adversarial network forremote sensing image pan-sharpening | Paper | Code | TGRS | GAN | 有监督 | 2021 |
| MDCNN | MDCNN: multispectral pansharpening based on amultiscale dilated convolutional neural network | Paper | JRS | CNN | 有监督 | 2021 | |
| LDP-Net | LDP-Net: An Unsupervised PansharpeningNetwork Based on Learnable Degradation Processes | Paper | Code | Arxiv | CNN | 无监督 | 2021 |
| DIGAN | Pansharpening approach via two-stream detailinjection based on relativistic generative adversarial networks | Paper | ESA | GAN | 有监督 | 2022 | |
| DPFN | A Dual-Path Fusion Network for Pan-Sharpening | Paper | Code | TGRS | CNN | 有监督 | 2022 |
| MSGAN | An Unsupervised Multi-scale GenerativeAdversarial Network for Remote Sensing Image Pan-Sharpening | Paper | ICMM | GAN | 无监督 | 2022 | |
| UCGAN | Unsupervised Cycle-Consistent GenerativeAdversarial Networks for Pan Sharpening | Paper | Code | TGRS | GAN | 无监督 | 2022 |
| P2Sharpen | P2Sharpen: A progressive pansharpening network with deep spectral transformation | Paper | Code | INFFus | CNN | 有监督 | 2023 |
通用图像融合框架(General Image Fusion Framerwork)
| 方法 | 标题 | 论文 | 代码 | 发表期刊或会议 | 基础框架 | 监督范式 | 年份 |
|---|---|---|---|---|---|---|---|
| IFCNN | IFCNN: A general image fusion framework based on convolutionalneural network | Paper | Code | InFus | CNN | 有监督 | 2020 |
| FusionDN | FusionDN: A Unified Densely Connected Network for ImageFusion | Paper | Code | AAAI | CNN | 无监督 | 2020 |
| PMGI | Rethinking the Image Fusion: A Fast Unified Image FusionNetwork based on Proportional Maintenance of Gradient and Intensity | Paper | Code | AAAI | CNN | 无监督 | 2020 |
| CU-Net | Deep Convolutional Neural Network for Multi-Modal ImageRestoration and Fusion | Paper | Code | TPAMI | CNN | 有监督 | 2021 |
| SDNet | SDNet: A Versatile Squeeze-and-Decomposition Network forReal-Time Image Fusion | Paper | Code | IJCV | CNN | 无监督 | 2021 |
| DIF-Net | Unsupervised Deep Image Fusion With Structure TensorRepresentations | Paper | Code | TIP | CNN | 无监督 | 2021 |
| IFSepR | IFSepR: A general framework for image fusion based on separaterepresentation learning | Paper | TMM | AE | 自监督 | 2021 | |
| MTOE | Multiple Task-Oriented Encoders for Unified Image Fusion | Paper | ICME | CNN | 无监督 | 2021 | |
| U2Fusion | U2Fusion: A Unified Unsupervised Image Fusion Network | Paper | Code | TPAMI | CNN | 无监督 | 2022 |
| SwinFusion | SwinFusion: Cross-domain Long-range Learning for General Image Fusion via Swin Transformer | Paper | Code | JAS | Transformer | 无监督 | 2022 |
| DeFusion | Fusion from Decomposition: A Self-Supervised Decomposition Approach for Image Fusion | Paper | Code | ECCV | CNN | 无监督 | 2022 |
| UIFGAN | UIFGAN: An unsupervised continual-learning generative adversarial network for unified image fusion | Paper | Code | InFus | GAN | 无监督 | 2023 |
综述(Survey)
| 标题 | 论文 | 代码 | 发表期刊或会议 | 年份 | |
|---|---|---|---|---|---|
| A review ofremote sensing image fusion methods | Paper | InFus | 2016 | ||
| Pixel-levelimage fusion: A survey of the state of the art | Paper | InFus | 2017 | ||
| Deep learning for pixel-level image fusion: Recent advances and future prospects | Paper | InFus | 2018 | ||
| Infrared and visible image fusion methods and applications: A survey | Paper | InFus | 2019 | ||
| Multi-focu image fusion: A Survey of the state of the art | Paper | InFus | 2020 | ||
| Image fusion meets deep learning: A survey and perspective | Paper | InFus | 2021 | ||
| Deep Learning-based Multi-focus Image Fusion: A Survey and A ComparativeStudy | Paper | Code | TPAMI | 2021 | |
| Benchmarking and comparing multi-exposure image fusion algorithms | Paper | Code | InFus | 2021 | |
| Current advances and future perspectives of image fusion: A comprehensive review | Paper | InFus | Code | InFus | 2023 |
数据集(Dataset)
| 融合场景 | 数据集 | 下载链接 |
|---|---|---|
| 红外和可见光图像融合 | TNO | https://figshare.com/articles/dataset/TNO_Image_Fusion_Dataset/1008029 |
| INO | https://www.ino.ca/en/technologies/video-analytics-dataset/videos/ | |
| RoadScene | https://github.com/hanna-xu/RoadScene | |
| MSRS | https://github.com/Linfeng-Tang/MSRS | |
| LLVIP | https://bupt-ai-cz.github.io/LLVIP/ | |
| M3FD | https://github.com/JinyuanLiu-CV/TarDAL | |
| 医学图像融合 | Harvard | http://www.med.harvard.edu/AANLIB/home.html |
| 多曝光图像融合 | MEF | https://github.com/csjcai/SICE |
| MEFB | https://github.com/xingchenzhang/MEFB | |
| 多聚焦图像融合 | Lytro | https://mansournejati.ece.iut.ac.ir/content/lytro-multi-focus-dataset |
| MFI-WHU | https://github.com/HaoZhang1018/MFI-WHU | |
| MFFW | https://www.semanticscholar.org/paper/MFFW%3A-A-new-dataset-for-multi-focus-image-fusion-Xu-Wei/4c0658f338849284ee4251a69b3c323908e62b45 | |
| 全色图像锐化 | GaoFen | https://directory.eoportal.org/web/eoportal/satellite-missions/g |
| WorldView | https://worldview.earthdata.nasa.gov/ | |
| GeoEye | https://earth.esa.int/eogateway/missions/geoeye-1 | |
| QuickBird | https://www.satimagingcorp.com/satellite-sensors/quickbird/ |
评估指标(Evaluation Metric)
通用评估指标(General evaluation metric)
通用评估指标位于：https://github.com/Linfeng-Tang/Image-Fusion/tree/main/General%20Evaluation%20Metric or https://github.com/Linfeng-Tang/Evaluation-for-Image-Fusion
遥感影像融合评估指标(Evaluation metric for pansharpening)
如有疑问可联系：2458707789@qq.com; 备注 姓名+学校
更新维护不易，关注，收藏，点赞一键三连是我持续更新的动力哦~