基于提示学习的一体化全色锐化模型

doi:10.3969/j.issn.1000-5641.2026.04.010

摘要/Abstract

摘要：

全色锐化技术通过融合高空间分辨率全色图像和低空间分辨率多光谱图像, 提高多光谱图像的空间分辨率. 随着深度学习的发展, 构建深度模型逐渐成为了全色锐化的主流方法. 然而, 目前的主流技术都是基于某个特定卫星的数据集进行模型训练, 训练出的模型也仅能用于该卫星数据集的全色锐化, 而对于数据集不足的卫星, 相关研究尚未充分开展. 针对这种现象, 提出了一种一体化的全色锐化方法, 通过提示学习技术, 同时针对不同卫星的数据集进行训练, 训练出的一体化模型可应用于不同卫星数据集的全色锐化. 实验表明, 该模型能够达到当前主流水平并且具有良好的泛化性.

关键词: 全色锐化, 遥感图像, 图像融合, 深度学习, 提示学习

Abstract:

Pansharpening is an image processing technique primarily used to enhance the spatial resolution of remote sensing images by fusing high spatial resolution panchromatic images with low spatial resolution multispectral images. With the development of deep learning, building deep models has gradually become the mainstream method for pansharpening. However, current mainstream technologies are based on training models using datasets from specific satellites, limiting the application of the trained models to fuse images from only those satellite datasets, without covering satellites with limited datasets. Addressing this issue, this paper proposes a unified pansharpening approach that leverages prompt learning technology to train models on datasets from different satellites. The trained unified model can be applied to fuse images from various satellite datasets. Experimental results demonstrate that the model achieves state-of-the-art performance and exhibits good generalization capabilities.

Key words: pansharpening, remote sensing images, image fusion, deep learning, prompt learning

中图分类号:

TP751.1

赵鸿飞, 汪婷婷, 方发明, 张桂戌. 基于提示学习的一体化全色锐化模型[J]. 华东师范大学学报（自然科学版）, 2026, 2026(4): 92-101.

Hongfei ZHAO, Tingting WANG, Faming FANG, Guixu ZHANG. A unified pansharpening model based on prompt learning[J]. J* E* C* N* U* N* S*, 2026, 2026(4): 92-101.

图/表 7

图1

表1

表2

表3

图2

表4

图3

参考文献 26

1	Zhang B, Wu D, Zhang L, et al.. Application of hyperspectral remote sensing for environment monitoring in mining areas. Environmental Earth Sciences, 2012, 65 (3): 649- 658.
2	Asokan A, Anitha J.. Change detection techniques for remote sensing applications: a survey. Earth Science Informatics, 2019, 12 (2): 143- 160.
3	Sharma R, Kamble S S, Gunasekaran A, et al.. A systematic literature review on machine learning applications for sustainable agriculture supply chain performance. Computers & Operations Research, 2020, 119, 104926.
4	Vivone G, Dalla Mura M, Garzelli A, et al.. A new benchmark based on recent advances in multispectral pansharpening: Revisiting pansharpening with classical and emerging pansharpening methods. IEEE Geoscience and Remote Sensing Magazine, 2021, 9 (1): 53- 81.
5	Meng X C, Shen H F, Li H F, et al.. Review of the pansharpening methods for remote sensing images based on the idea of meta-analysis: practical discussion and challenges. Information Fusion, 2019, 46, 102- 113.
6	Ghassemian H.. A review of remote sensing image fusion methods. Information Fusion, 2016, 32, 75- 89.
7	Gao H, Yang J, Zhang Y, et al.. Prompt-based ingredient-oriented all-in-one image restoration. IEEE Transactions on Circuits and Systems for Video Technology, 2024, 34 (10): 9458- 9471.
8	Zhang C, Zhu Y, Yan Q S, et al. All-in-one multi-degradation image restoration network via hierarchical degradation representation [C]//Proceedings of the 31st ACM International Conference on Multimedia. ACM, 2023: 2285-2293.
9	Masi G, Cozzolino D, Verdoliva L, et al.. Pansharpening by convolutional neural networks. Remote Sensing, 2016, 8 (7): 594.
10	Bandara W G C, Patel V M. HyperTransformer: a textural and spectral feature fusion transformer for pansharpening [C]//2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2022: 1757-1767.
11	Li B Y, Liu X, Hu P, et al. All-in-one image restoration for unknown corruption [C]//2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2022: 17431-17441.
12	Chen L Y, Lu X, Zhang J, et al. HINet: half instance normalization network for image restoration [C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). IEEE, 2021: 182-192.
13	Li R T, Tan R T, Cheong L F. All in one bad weather removal using architectural search [C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2020: 3172-3182.
14	Zhou K Y, Yang J K, Loy C C, et al. Conditional prompt learning for vision-language models [C]//2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2022: 16795-16804.
15	Chen J Y, Dong J X, Ju Y K, et al. PromptRestorer: a prompting image restoration method with degradation perception [C]//Advances in Neural Information Processing Systems 36. Neural Information Processing Systems Foundation, Inc. (NeurIPS), 2023: 8898-8912.
16	Khan S, Potlapalli V, Shahbaz Khan F, et al. PromptIR: prompting for all-in-one image restoration [C]//Advances in Neural Information Processing Systems 36. Neural Information Processing Systems Foundation, Inc. (NeurIPS), 2023: 71275-71293.
17	Zamir S W, Arora A, Khan S, et al. Restormer: efficient transformer for high-resolution image restoration [C]//2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2022: 5718-5729.
18	Meng X C, Xiong Y M, Shao F, et al.. A large-scale benchmark data set for evaluating pansharpening performance: Overview and implementation. IEEE Geoscience and Remote Sensing Magazine, 2021, 9 (1): 18- 52.
19	Peng S Y, Zhu D, Gao Q W, et al.. PSCF-net: deeply coupled feedback network for pansharpening. IEEE Transactions on Geoscience and Remote Sensing, 2023, 61, 5401812.
20	Wang J M, Lu T, Huang X, et al.. Pan-sharpening via conditional invertible neural network. Information Fusion, 2024, 101, 101980.
21	Wang H, Gong M Q, Mei X G, et al.. Deep unfolded network with intrinsic supervision for pan-sharpening. Proceedings of the AAAI Conference on Artificial Intelligence, 2024, 38 (6): 5419- 5426.
22	Yang J F, Fu X Y, Hu Y W, et al. PanNet: a deep network architecture for pan-sharpening [C]//2017 IEEE International Conference on Computer Vision (ICCV). IEEE, 2017: 1753-1761.
23	Jin Z R, Zhang T J, Jiang T X, et al.. LAGConv: local-context adaptive convolution kernels with global harmonic bias for pansharpening. Proceedings of the AAAI Conference on Artificial Intelligence, 2022, 36 (1): 1113- 1121.
24	Yuan Q Q, Wei Y C, Meng X C, et al.. A multiscale and multidepth convolutional neural network for remote sensing imagery pan-sharpening. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018, 11 (3): 978- 989.
25	Xu S, Zhang J S, Zhao Z X, et al. Deep gradient projection networks for pan-sharpening [C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, 2021: 1366-1375.
26	Cai J J, Huang B.. Super-resolution-guided progressive pansharpening based on a deep convolutional neural network. IEEE Transactions on Geoscience and Remote Sensing, 2021, 59 (6): 5206- 5220.

方法	GF1			QB
方法	PSNR↑/dB	SSIM↑	ERGAS↓	PSNR↑/dB	SSIM↑	ERGAS↓
PSCF	39.6091	0.9861	6.2063	43.6585	0.9895	3.8693
PSCINN	40.0811	0.9867	5.7596	42.1693	0.9871	4.7166
DISPNet	46.6074	0.9949	2.9815	48.1770	0.9940	2.4140
PNN	45.1481	0.9931	3.4656	47.0071	0.9912	2.7640
PanNet	46.1890	0.9947	3.1255	47.7010	0.9940	2.3748
LAGConv	47.6586	0.9952	2.7133	47.8898	0.9937	2.3554
MSDCNN	46.2428	0.9941	3.1067	47.6549	0.9927	2.5328
GPPNN	43.7842	0.9916	4.1566	46.2982	0.9912	3.0878
SRPPNN	46.9454	0.9953	2.8718	48.5581	0.9945	2.1984
本文方法	48.5660	0.9957	0.8423	48.5080	0.9947	0.6159

方法	WV2			WV4
方法	PSNR↑/dB	SSIM↑	ERGAS↓	PSNR↑/dB	SSIM↑	ERGAS↓
PSCF	40.3956	0.9821	17.1725	42.1347	0.9918	3.5732
PSCINN	38.0735	0.9761	21.3113	32.4285	0.9776	10.922
DISPNet	36.7243	0.9663	25.6244	38.1584	0.9781	5.5745
PNN	40.4835	0.9813	16.8310	40.7926	0.9895	4.1806
PanNet	39.5244	0.9785	18.7344	40.3312	0.9876	4.4710
LAGConv	39.9678	0.9804	17.8407	40.1784	0.9876	4.4442
MSDCNN	40.4649	0.9820	16.8003	41.5396	0.9905	3.8784
GPPNN	40.7654	0.9826	16.4375	42.2707	0.9913	3.5499
SRPPNN	39.8992	0.9797	18.0026	40.7506	0.9891	4.2770
本文方法	39.8846	0.9797	18.2235	41.4290	0.9898	3.9733

方法	WV3			IKONOS
方法	PSNR↑/dB	SSIM↑	ERGAS↓	PSNR↑/dB	SSIM↑	ERGAS↓
PSCF	29.6898	0.9147	47.1054	33.6036	0.9207	9.7883
PSCINN	25.7267	0.8128	70.7151	32.6893	0.8960	10.7471
DISPNet	30.5656	0.9203	43.9822	37.3609	0.9462	6.4953
PNN	30.1288	0.9184	44.5506	36.4540	0.9343	7.3178
PanNet	29.9108	0.8591	50.0833	36.7974	0.9397	6.9012
LAGConv	31.1914	0.9084	42.2248	37.7622	0.9333	7.6344
MSDCNN	30.0524	0.9250	44.8807	35.7193	0.9341	7.8613
GPPNN	26.0173	0.8344	69.8300	33.1218	0.9192	10.3322
SRPPNN	31.0001	0.9119	42.3682	37.5427	0.9494	6.4896
本文方法	34.0009	0.9852	2.8504	42.0446	0.9914	2.7870

方法	WV3			IKONOS
方法	D_λ↓	D_S↓	Q↑	D_λ↓	D_S↓	Q↑
PSCF	0.1073	0.0683	0.8317	0.1867	0.1236	0.7137
PSCINN	0.1214	0.0925	0.7967	0.4231	0.1696	0.4816
DISPNet	0.0423	0.0500	0.9099	0.0869	0.0708	0.8493
PNN	0.0519	0.0688	0.8829	0.1199	0.1081	0.7858
PanNet	0.0421	0.0841	0.8777	0.0827	0.1109	0.8161
LAGConv	0.0605	0.0722	0.8717	0.0916	0.0749	0.8410
MSDCNN	0.0329	0.0669	0.9030	0.1213	0.1023	0.7898
GPPNN	0.0732	0.1249	0.8107	0.3066	0.1491	0.5913
SRPPNN	0.0455	0.0661	0.8919	0.0773	0.0626	0.8656
本文方法	0.0336	0.0426	0.9252	0.0611	0.0430	0.8988

[1]	李旭阳, 高岩, 全红艳. 深度语义指导的遥感图像道路提取算法[J]. 华东师范大学学报（自然科学版）, 2026, 2026(4): 73-82.
[2]	何泽锋, 沈富可, 魏同权. 基于双重决策自适应冻结实现快速准确的迁移学习[J]. 华东师范大学学报（自然科学版）, 2025, 2025(6): 29-38.
[3]	赵思嘉, 韩凡宇, 王伟. 基于多维特征融合的GitHub开发者地理位置预测[J]. 华东师范大学学报（自然科学版）, 2025, 2025(5): 1-13.
[4]	曹鹭萍, 夏勇. 大气湍流下高分辨率高带宽分数涡旋光束的探测[J]. 华东师范大学学报（自然科学版）, 2025, 2025(3): 51-60.
[5]	梁怡萍, 肖路巍, 王琳琳. 基于标签感知增强的社交媒体心理亚健康归因方法[J]. 华东师范大学学报（自然科学版）, 2025, 2025(1): 124-137.
[6]	黄彩蝶, 王昕萍, 陈良育, 刘勇. 基于堆叠门控循环单元残差网络的知识追踪模型研究[J]. 华东师范大学学报（自然科学版）, 2022, 2022(6): 68-78.
[7]	马依琳, 陶慧玲, 董启文, 王晔. 基于Transformer的多特征融合的航空发动机剩余使用寿命预测[J]. 华东师范大学学报（自然科学版）, 2022, 2022(5): 219-232.
[8]	王泽杰, 沈超敏, 赵春, 刘新妹, 陈杰. 融合人体姿态估计和目标检测的学生课堂行为识别[J]. 华东师范大学学报（自然科学版）, 2022, 2022(2): 55-66.
[9]	刘波, 白晓东, 张更新, 沈俊, 谢继东, 赵来定, 洪涛. 深度学习在认知无线电中的应用研究综述[J]. 华东师范大学学报（自然科学版）, 2021, 2021(1): 36-52.
[10]	张旭, 黄定江. 基于深度学习的铝材表面缺陷检测[J]. 华东师范大学学报（自然科学版）, 2020, 2020(6): 105-114.
[11]	韩程程, 李磊, 刘婷婷, 高明. 语义文本相似度计算方法[J]. 华东师范大学学报（自然科学版）, 2020, 2020(5): 95-112.
[12]	杨康, 黄定江, 高明. 面向自动问答的机器阅读理解综述[J]. 华东师范大学学报(自然科学版), 2019, 2019(5): 36-52.
[13]	陈远哲, 匡俊, 刘婷婷, 高明, 周傲英. 共指消解技术综述[J]. 华东师范大学学报(自然科学版), 2019, 2019(5): 16-35.
[14]	刘恒宇, 张天成, 武培文, 于戈. 知识追踪综述[J]. 华东师范大学学报(自然科学版), 2019, 2019(5): 1-15.
[15]	叶健, 赵慧. 基于大规模弹幕数据监听和情感分类的舆情分析模型[J]. 华东师范大学学报(自然科学版), 2019, 2019(3): 86-100.