融合密集连接和自注意力机制的高分辨率遥感影像变化检测方法

doi:10.11947/j.AGCS.2024.20230454

测绘学报 ›› 2024, Vol. 53 ›› Issue (12): 2244-2253.doi: 10.11947/j.AGCS.2024.20230454

融合密集连接和自注意力机制的高分辨率遥感影像变化检测方法

庞世燕¹(), 郝京京¹, 左志奇², 兰晶晶¹, 胡翔云³^,⁴()

^1.华中师范大学人工智能教育学部，湖北　武汉　430079
^2.华中农业大学信息学院，湖北　武汉　430070
^3.武汉大学遥感信息工程学院，湖北　武汉　430079
^4.湖北珞珈实验室，湖北　武汉　430079

收稿日期:2023-10-10 出版日期:2025-01-06 发布日期:2025-01-06
通讯作者: 胡翔云 E-mail:pangsy@ccnu.edu.cn;huxy@whu.edu.cn
作者简介:庞世燕（1987—），女，博士，副教授，研究方向为遥感影像解译、深度学习应用。E-mail：pangsy@ccnu.edu.cn
基金资助:
教育部人文社会科学项目(22YJC880058);中央高校基本科研业务费项目(CCNU22QN019)

A high-resolution remote sensing images change detection method via the integration of dense connections and self-attention mechanisms

Shiyan PANG¹(), Jingjing HAO¹, Zhiqi ZUO², Jingjing LAN¹, Xiangyun HU³^,⁴()

^1.Faculty of Artificial Intelligence in Education, Central China Normal University, Wuhan 430079, China
^2.College of Informatics, Huazhong Agricultural University, Wuhan 430070, China
^3.School of Remote Sensing and Information Engineering, Wuhan University, Wuhan 430079, China
^4.Hubei Luojia Laboratory, Wuhan 430079, China

Received:2023-10-10 Online:2025-01-06 Published:2025-01-06
Contact: Xiangyun HU E-mail:pangsy@ccnu.edu.cn;huxy@whu.edu.cn
About author:PANG Shiyan (1987—), female, PhD, associate professor, majors in remote sensing image interpretation and deep learning applications. E-mail: pangsy@ccnu.edu.cn
Supported by:
Humanities and Social Science Research Project Funded by the Ministry of Education(22YJC880058);The Fundamental Research Funds for the Central Universities(CCNU22QN019)

摘要/Abstract

摘要：

遥感影像变化检测是遥感影像分析中的一项重要任务，在城市动态监测、地理信息更新、自然灾害监测、违章建筑物查处、军事目标打击效果分析及国土资源调查等方面应用广泛。作为像素级别的预测任务，当前的变化检测有两个比较突出的问题：一是受限于自注意力计算的效率问题，遥感影像中的远程上下文信息利用不足；二是侧重于深度变化语义特征的提取，而忽略了包含高分辨率和细粒度特征的浅层信息的应用。针对第1个问题，本文采用Transformer对提取的两期深层影像特征进行上下文建模，提升深层变化特征的质量。为了兼顾Transformer的效率问题，本文将图像特征图转化为稀疏标记，以此来显著减少Transformer的令牌数量。针对第2个问题，本文采用密集跳跃连接的方式，以保留浅层变化特征中的高分辨率和细粒度信息。采用了3个公开的数据集进行试验验证，结果表明，与其他先进的变化检测方法相比，本文方法的IoU指标分别达到了85.44%、84.15%和94.61%，均优于其他对比方法。

关键词: 遥感影像, 变化检测, Transformer, 密集连接, 多尺度特征融合

Abstract:

Remote sensing image change detection is an important task in remote sensing image analysis, which is widely used in urban dynamic monitoring, geographic information update, natural disaster monitoring, illegal building investigation, military target strike effect analysis, and land and resources survey. As a pixel-level prediction task, the current methods of change detection have two prominent problems: one is the computational efficiency of self-attention between arbitrary pixel pairs is low, and the long context information in remote sensing images is insufficiently utilized; the other is that the current methods focus on the extraction of deep change image features while shallow information containing high-resolution and fine-grained features are ignored. To address the first problem, Transformer is used to perform context modeling on the extracted bitemporal image features to improve the quality of the deepest change features. To take into account the efficiency of the Transformer, the proposed method converts the images into sparse tokens, thereby significantly reducing the number of tokens of the Transformer. For the second problem, the proposed method uses dense skip connections to retain high resolution in shallow change features. Three publicly available datasets were used for experiments. Extensive experiments show that compared with the state-of-the-art change detection methods, the IoU metric of the proposed method reached 85.44%, 84.15% and 94.61%, respectively, which is better than other comparison methods.

Key words: remote sensing images, change detection, Transformer, dense connections, multi-scale feature fusion

中图分类号:

P237

庞世燕, 郝京京, 左志奇, 兰晶晶, 胡翔云. 融合密集连接和自注意力机制的高分辨率遥感影像变化检测方法[J]. 测绘学报, 2024, 53(12): 2244-2253.

Shiyan PANG, Jingjing HAO, Zhiqi ZUO, Jingjing LAN, Xiangyun HU. A high-resolution remote sensing images change detection method via the integration of dense connections and self-attention mechanisms[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(12): 2244-2253.

图/表 12

图1

图2

图3

图4

图5

图6

图7

表1

图8

表2

表3

图9

参考文献 29

[1]	FANG S, LI K, SHAO J, et al. SNUNet-CD: a densely connected Siamese network for change detection of VHR images[J]. IEEE Geoscience and Remote Sensing Letters, 2021, 19: 1-5.
[2]	LIU W, JI X, LIU J, et al. A novel unsupervised change detection method with structure consistency and GFLICM based on UAV images[J]. Journal of Geodesy and Geoinformation Science, 2022, 5(1): 91-102.
[3]	张祖勋, 姜慧伟, 庞世燕, 等. 多时相遥感影像的变化检测研究现状与展望[J]. 测绘学报, 2022, 51(7): 1091-1107. DOI:. doi: 10.11947/j.AGCS.2022.20220070
	ZHANG Zuxun, JIANG Huiwei, PANG Shiyan, et al. Review and prospect in change detection of multi-temporal remote sensing images[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(7): 1091-1107. DOI:. doi: 10.11947/j.AGCS.2022.20220070
[4]	ZAGORUYKO S, KOMODAKIS N. Learning to compare image patches via convolutional neural networks[C]//Proceedings of 2015 IEEE Conference on Computer Vision and Pattern Recognition. Boston: IEEE, 2015: 4353-4361.
[5]	ZHAN Yang, FU Kun, YAN Menglong, et al. Change detection based on deep siamese convolutional network for optical aerial images[J]. IEEE Geoscience and Remote Sensing Letters, 2017, 14(10): 1845-1849.
[6]	DAUDT R C, LE SAUX B, BOULCH A. Fully convolutional siamese networks for change detection[C]//Proceedings of 2018 IEEE International Conference on Image Processing. Athens: IEEE, 2018: 4063-4067.
[7]	ZHANG Wuxia, LU Xiaoqiang. The spectral-spatial joint learning for change detection in multispectral imagery[J]. Remote Sensing, 2019, 11(3): 240.
[8]	CHEN Hao, LI Wenyuan, SHI Zhenwei. Adversarial instance augmentation for building change detection in remote sensing images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 60: 1-16.
[9]	PENG Daifeng, ZHANG Yongjun, GUAN Haiyan. End-to-end change detection for high resolution satellite images using improved UNet++[J]. Remote Sensing, 2019, 11(11): 1382.
[10]	HUANG Gao, LIU Zhuang, VAN DER M L, et al. Densely connected convolutional networks[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017: 4700-4708.
[11]	ZHOU Zongwei, SIDDIQUEE M M R, TAJBAKHSH N, et al. UNet++: redesigning skip connections to exploit multiscale features in image segmentation[J]. IEEE Transactions on Medical Imaging, 2020, 39(6): 1856-1867.
[12]	王超, 王帅, 陈晓, 等. 联合UNet++和多级差分模块的多源光学遥感影像对象级变化检测[J]. 测绘学报, 2023, 52(2): 283-296. DOI:. doi: 10.11947/j.AGCS.2023.20220202
	WANG Chao, WANG Shuai, CHEN Xiao, et al. Object-level change detection of multi-sensor optical remote sensing images combined with UNet++ and multi-level difference module[J]. Acta Geodaetica et Cartographica Sinica, 2023, 52(2): 283-296. DOI:. doi: 10.11947/j.AGCS.2023.20220202
[13]	梁哲恒, 黎宵, 邓鹏, 等. 融合多尺度特征注意力的遥感影像变化检测方法[J]. 测绘学报, 2022, 51(5): 668-676. DOI:. doi: 10.11947/j.AGCS.2022.20200540
	LIANG Zheheng, LI Xiao, DENG Peng, et al. Remote sensing image change detection fusion method integrating multi-scale feature attention[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(5): 668-676. DOI:. doi: 10.11947/j.AGCS.2022.20200540
[14]	ZHANG Chenxiao, YUE Peng, TAPETE D, et al. A deeply supervised image fusion network for change detection in high resolution bitemporal remote sensing images[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2020, 166: 183-200.
[15]	PENG Xueli, ZHONG Ruofei, LI Zhen, et al. Optical remote sensing image change detection based on attention mechanism and image difference[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 59(9): 7296-7307.
[16]	CHEN Hao, SHI Zhenwei. A spatial-temporal attention-based method and a new dataset for remote sensing image change detection[J]. Remote Sensing, 2020, 12(10): 1662-1684.
[17]	JIANG Huiwei, HU Xiangyun, LI Kun, et al. PGA-SiamNet: pyramid feature-based attention-guided Siamese network for remote sensing orthoimagery building change detection[J]. Remote Sensing, 2020, 12(3): 484.
[18]	SHI Qian, LIU Mengxi, LI Shengchen, et al. A deeply supervised attention metric-based network and an open aerial image dataset for remote sensing change detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 60: 1-16.
[19]	LIU Yi, PANG Chao, ZHAN Zongqian, et al. Building change detection for remote sensing images using a dual-task constrained deep siamese convolutional network model[J]. IEEE Geoscience and Remote Sensing Letters, 2020, 18(5): 811-815.
[20]	ZHANG Lin, HU Xiangyun, ZHANG Mi, et al. Object-level change detection with a dual correlation attention-guided detector[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2021, 177, 147-160.
[21]	CHEN Pan, ZHANG Bing, HONG Danfeng, et al. FCCDN: feature constraint network for VHR image change detection[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2022, 187: 101-119.
[22]	FANG Sheng, LI Kaiyu, LI Zhe. Changer: feature interaction is what you need for change detection[EB/OL]. [2023-09-01]. https://arxiv.org/abs/2209.08290v1.
[23]	CHEN Hao, QI Zipeng, SHI Zhenwei. Remote sensing image change detection with transformers[J]. IEEE Transactions on Geos-cience and Remote Sensing, 2021, 60: 1-14.
[24]	BANDARA W G C, PATEL V M. A Transformer-based Siamese network for change detection[C]//Proceedings of 2022 IEEE International Geoscience and Remote Sensing Symposium. Kuala Lumper: IEEE, 2022: 207-210.
[25]	ZHENG Zhuo, ZHONG Yanfei, TIAN Shiqi, et al. ChangeMask: deep multi-task Encoder-Transformer-Decoder architecture for semantic change detection[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2022, 183: 228-239.
[26]	CHEN Zhanlong, ZHOU Yuan, WANG Bin, et al. EGDE-Net: a building change detection method for high-resolution remote sensing imagery based on edge guidance and differential enhancement[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2022, 191, 203-222.
[27]	HE Kaiming, ZHANG Xiangyu, REN Shaoqing, et al. Deep residual learning for image recognition[C]//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 770-778.
[28]	DENG Jia, DONG Wei, SOCHER R, et al. Imagenet: a large-scale hierarchical image database[C]//Proceedings of 2009 IEEE Conference on Computer Vision and Pattern Recognition. Miami: IEEE, 2009: 248-255.
[29]	SELVARAJU R R, COGSWELL M, DAS A, et al. Gradcam: visual explanations from deep networks via gradient-based localization[C]//Proceedings of 2017 IEEE International Conference on Computer Vision. Venice: IEEE, 2017: 618-626.

模型	WHU-CD						LEVIR-CD						CDD
模型	IoU	OA	Precision	Recall	F₁值	Kappa	IoU	OA	Precision	Recall	F₁值	Kappa	IoU	OA	Precision	Recall	F₁值	Kappa
DTCDSCN	82.19	99.31	92.70	87.87	90.22	89.86	79.90	98.86	89.04	88.61	88.82	88.23	86.15	98.22	94.08	91.08	92.56	91.55
SNUNet	79.09	99.17	90.68	86.08	88.32	87.89	82.16	98.97	94.83	86.01	90.21	85.76	88.13	98.47	93.99	93.39	93.69	92.82
BIT	81.29	99.26	90.66	88.71	89.68	89.29	82.63	98.99	93.88	87.33	90.49	89.95	89.19	98.62	95.12	93.47	94.29	93.50
ChangeFormer	76.28	99.07	92.11	81.62	86.55	86.07	78.21	98.78	89.74	85.89	87.77	87.13	84.53	98.00	94.99	88.47	91.62	90.48
ChangerEx	72.09	98.77	80.43	87.42	83.78	83.14	81.62	98.99	91.76	88.08	89.88	89.35	89.52	98.67	96.82	92.24	94.47	93.71
TNUNet-CD	85.44	99.44	93.61	90.74	92.15	91.86	84.15	99.13	92.10	90.70	91.40	90.94	94.61	99.33	97.97	96.50	97.23	96.85

模型	Params/MB	FLOPs/G
DTCDSCN	31.26	52.89
SNUNet	12.03	219.30
BIT	24.74	232.14
ChangeFormer	41.02	810.50
ChangerEx	11.39	23.82
TNUNet-CD	40.48	225.91

变化分析模块（共5层）							信息解码模块		评估指标（IoU）
第1~4层			第5层				信息解码模块		评估指标（IoU）
CCA	Sub	RBIT（32）	CCA	Sub	BIT（4）	RBIT（32）	UNet	UNet++	WHU-CD
	√			√			√		81.13
√			√				√		82.38
√					√		√		76.15
√						√	√		84.47
		√				√		√	83.55
√						√		√	85.44

融合密集连接和自注意力机制的高分辨率遥感影像变化检测方法

A high-resolution remote sensing images change detection method via the integration of dense connections and self-attention mechanisms

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 12

参考文献 29

相关文章 15

编辑推荐

Metrics

本文评价

[1]	季顺平, 刘瑾, 高建, 龚健雅. 多视影像深度学习密集匹配三维重建智能框架[J]. 测绘学报, 2025, 54(9): 1633-1646.
[2]	张继贤, 顾海燕, 倪欢, 李海涛, 杨懿, 丁少鹏, 隋淞蔓. 遥感智能变化检测的深度学习方法：演变与发展趋势[J]. 测绘学报, 2025, 54(8): 1347-1370.
[3]	程青, 汪博轩, 张洪艳. DRformer：一种渐进式耦合多尺度CNN与浓缩注意力Transformer的高光谱图像超分辨率方法[J]. 测绘学报, 2025, 54(7): 1230-1242.
[4]	万杰, 谢忠, 徐永洋, 陶留锋. 面向城市道路场景的车载LiDAR点云语义分割U形图卷积网络方法[J]. 测绘学报, 2025, 54(7): 1280-1293.
[5]	谢亚坤, 赵耀纪, 涂佳星, 夏瑞丰, 冯德俊, 刘苏凝, 陈虹宇, 朱军. 融合边缘与全局特征的遥感影像显著性目标检测方法[J]. 测绘学报, 2025, 54(7): 1265-1279.
[6]	董子博, 王竞雪, 卜丽静, 房琳, 许峥辉. MAFNet：基于多尺度空洞融合网络的遥感影像建筑物提取方法[J]. 测绘学报, 2025, 54(6): 1094-1106.
[7]	李海峰, 郭旺, 吴梦伟, 彭程里, 朱庆, 刘瑜, 陶超. 视觉-语言联合的遥感地物概念表达与智能解译：原理、挑战与机遇[J]. 测绘学报, 2025, 54(5): 853-872.
[8]	赵一鸣, 胡克林, 涂可龙, 卿雅娴, 杨超, 祁昆仑, 吴华意. 基于SAR与光学遥感影像融合的多标签场景分类方法[J]. 测绘学报, 2025, 54(5): 911-923.
[9]	王超, 陈天宇, 张同, AhmedTanvir, 纪立强, 谢涛, 杨佳俊, 王帅. 基于全局差分增强模块和平衡惩罚损失的多源光学遥感影像变化检测[J]. 测绘学报, 2025, 54(5): 873-887.
[10]	张新长, 齐霁, 陶超, 傅思扬, 郭明宁, 阮永检. 光学遥感影像去云研究进展、挑战与趋势[J]. 测绘学报, 2025, 54(4): 603-620.
[11]	龚良雄, 李星华, 程远明, 赵兴友, 谢仁平, 王红根. 时空差异增强与自适应特征融合的轻量级遥感影像变化检测网络[J]. 测绘学报, 2025, 54(1): 136-153.
[12]	李佳铃, 齐霁, 鲁伟鹏, 陶超. 面向城市功能区分类的光学遥感影像-OSM数据联合自监督学习方法[J]. 测绘学报, 2025, 54(1): 154-164.
[13]	徐涛, 杨元维, 高贤君, 王志威, 潘越, 李少华, 许磊, 王艳军, 刘波, 余静, 吴凤敏, 孙浩宇. 融合图卷积与多尺度特征的接触网点云语义分割[J]. 测绘学报, 2024, 53(8): 1624-1633.
[14]	鄢薪, 慎利, 潘俊杰, 戴延帅, 王继成, 郑晓莉, 李志林. 多尺度特征融合与空间优化的弱监督高分遥感建筑变化检测[J]. 测绘学报, 2024, 53(8): 1586-1597.
[15]	谢志伟, 翟帅智, 张丰源, 陈旻, 孙立双. 面向对象高分影像归纳式图神经网络分类法[J]. 测绘学报, 2024, 53(8): 1610-1623.