Research on road extraction considering road boundaries and connectivity

doi:10.11947/j.AGCS.2025.20240271

Abstract

Abstract:

Road extraction is a crucial task in remote sensing image interpretation. This study addresses the issue of occlusion in road extraction tasks within remote sensing imagery by proposing a novel feature fusion network structure, KDLinkNet. The network incorporates a graph-based inference module, the road connectivity module (RCM), designed to enhance road connectivity and rectify missing details in complex scenes. Additionally, the study introduces an edge optimization (EO) method based on multi-task learning, which incorporates prior knowledge of road boundaries to improve the network's ability to extract boundary information. Experimental results demonstrate that this method achieves F₁ scores of 94.0%, 79.8%, and 86.1% on the LRSNY, Massachusetts, and DeepGlobe datasets, respectively, outperforming current state-of-the-art methods. This research provides an effective solution for road extraction in complex remote sensing image scenarios.

Key words: road extraction, shelter, edge optimization, priori knowledge

CLC Number:

P237

Yongyang XU, Jian WANG, Liang WU, Zhong XIE. Research on road extraction considering road boundaries and connectivity[J]. Acta Geodaetica et Cartographica Sinica, 2025, 54(7): 1254-1264.

Figures/Tables 9

Fig. 1

Fig. 2

Fig. 3

Tab. 1

Tab. 2

Tab. 3

Fig. 4

Tab. 4

Fig. 5

References 32

[1]	FRIZZELLE B G, EVENSON K R, RODRIGUEZ D A, et al. The importance of accurate road data for spatial applications in public health: customizing a road network[J]. International Journal of Health Geographics, 2009, 8(1): 24.
[2]	YOUSSEF A M, SEFRY S A, PRADHAN B, et al. Analysis on causes of flash flood in Jeddah city (Kingdom of Saudi Arabia) of 2009 and 2011 using multi-sensor remote sensing data and GIS[J]. Geomatics, Natural Hazards and Risk, 2016, 7(3): 1018-1042.
[3]	XU Yongyang, SHI Zhaolun, XIE Xuejing, et al. Residual channel attention fusion network for road extraction based on remote sensing images and GPS trajectories[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2024, 17: 8358-8369.
[4]	连仁包, 张振敏, 廖一鹏, 等. 结合测地距离场与曲线平滑的遥感图像道路中心线快速提取[J]. 测绘学报, 2023, 52(8): 1317-1329. DOI: . doi: 10.11947/j.AGCS.2023.20220002
	LIAN Renbao, ZHANG Zhenmin, LIAO Yipeng, et al. A quick road centreline extraction method from remote sensing images combining with geodesic distance field and curve smoothing[J]. Acta Geodaetica et Cartographica Sinica, 2023, 52(8): 1317-1329. DOI: . doi: 10.11947/j.AGCS.2023.20220002
[5]	黄福蓉. 用于实时道路场景的语义分割算法CBR-ENet[J]. 中国电子科学研究院学报, 2021, 16(3): 277-284.
	HUANG Furong. Semantic segmentation network CBR-eNet for real-time road scenes[J]. Journal of China Academy of Electronics and Information Technology, 2021, 16(3): 277-284.
[6]	胡功明, 杨春成, 徐立, 等. 改进U-Net的遥感图像语义分割方法[J]. 测绘学报, 2023, 52(6): 980-989. DOI: . doi: 10.11947/j.AGCS.2023.20210684
	HU Gongming, YANG Chuncheng, XU Li, et al. Improved U-Net remote sensing image semantic segmentation method[J]. Acta Geodaetica et Cartographica Sinica, 2023, 52(6): 980-989. DOI: . doi: 10.11947/j.AGCS.2023.20210684
[7]	GAO Lin, SONG Weidong, DAI Jiguang, et al. Road extraction from high-resolution remote sensing imagery using refined deep residual convolutional neural network[J]. Remote Sensing, 2019, 11(5): 552.
[8]	LU Xiaoyan, ZHONG Yanfei, ZHENG Zhuo, et al. Multi-scale and multi-task deep learning framework for automatic road extraction[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(11): 9362-9377.
[9]	DING Lei, BRUZZONE L. DiResNet: direction-aware residual network for road extraction in VHR remote sensing images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 59(12): 10243-10254.
[10]	REN Yongfeng, YU Yongtao, GUAN Haiyan. DA-CapsUNet: a dual-attention capsule U-Net for road extraction from remote sensing imagery[J]. Remote Sensing, 2020, 12(18): 2866.
[11]	TAN Yongqiang, GAO Shanghua, LI Xuanyi, et al. VecRoad: point-based iterative graph exploration for road graphs extraction[C]//Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 8907-8915.
[12]	WEI Yao, JI Shunping. Scribble-based weakly supervised deep learning for road surface extraction from remote sensing images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 60: 1-12.
[13]	李卉. 集成LiDAR和遥感影像城市道路提取与三维建模[J]. 测绘学报, 2011, 40(1): 133.
	LI Hui. Road extraction and modeling with LiDAR and RS image in urban area[J]. Acta Geodaetica et Cartographica Sinica, 2011, 40(1): 133.
[14]	ZHOU Gaodian, CHEN Weitao, GUI Qianshan, et al. Split depth-wise separable graph-convolution network for road extraction in complex environments from high-resolution remote-sensing images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 3128033.
[15]	商建东, 刘艳青, 高需. 多尺度特征提取的道路场景语义分割[J]. 计算机应用与软件, 2021, 38(11): 174-178.
	SHANG Jiandong, LIU Yanqing, GAO Xu. Semantic segmentation of road scene based on multi-scale feature extraction[J]. Computer Applications and Software, 2021, 38(11): 174-178.
[16]	ZHOU Lichen, ZHANG Chuang, WU Ming. D-LinkNet: LinkNet with pretrained encoder and dilated convolution for high resolution satellite imagery road extraction[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Salt Lake City: IEEE, 2018: 192-1924.
[17]	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[J]. Advances in Neural Information Processing Systems, 2017, 30: 5998-6008.
[18]	CHEN L C, PAPANDREOU G, SCHROFF F, et al. Rethinking atrous convolution for semantic image segmentation[EB/OL]. [2017-12-05]. https://arxiv.org/abs/1706.05587.
[19]	ZHANG Yu, YANG Qiang. An overview of multi-task learning[J]. National Science Review, 2018, 5(1): 30-43.
[20]	LI X, SUN X, MENG Y, et al. Dice loss for data-imbalanced NLP tasks[EB/OL]. [2020-08-29]. https://arxiv.org/abs/1911.02855.
[21]	LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]//Proceedings of 2017 IEEE International Conference on Computer Vision. Venice: IEEE, 2017: 2999-3007.
[22]	FU J, LIU J, TIAN H, et al. Dual attention network for scene segmentation[C]//Proceedings of 2019 IEEE/CVF conference on computer vision and pattern recognition.[S.l.]: IEEE, 2019: 3146-3154.
[23]	PAN Huihui, HONG Yuanduo, SUN Weichao, et al. Deep dual-resolution networks for real-time and accurate semantic segmentation of traffic scenes[J]. IEEE Transactions on Intelligent Transportation Systems, 2023, 24(3): 3448-3460.
[24]	CHEN Ziyi, DENG Liai, LUO Yuhua, et al. Road extraction in remote sensing data: a survey[J]. International Journal of Applied Earth Observation and Geoinformation, 2022, 112: 102833.
[25]	NI Hongyin, JIANG Shan. Deep dual-resolution road scene segmentation networks based on decoupled dynamic filter and squeeze-excitation module[J]. Sensors, 2023, 23(16): 7140.
[26]	CHEN Ziyi, LUO Yuhua, WANG Jing, et al. DPENet: dual-path extraction network based on CNN and transformer for accurate building and road extraction[J]. International Journal of Applied Earth Observation and Geoinformation, 2023, 124: 103510.
[27]	ZHU Qiqi, ZHANG Yanan, WANG Lizeng, et al. A Global Context-aware and Batch-independent Network for road extraction from VHR satellite imagery[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2021, 175: 353-365.
[28]	LIU Pengfei, WANG Qing, YANG Gaochao, et al. Survey of road extraction methods in remote sensing images based on deep learning[J]. PFG-Journal of Photogrammetry, Remote Sensing and Geoinformation Science, 2022, 90(2): 135-159.
[29]	MADHUMITA D, BHARATH H A. Deep learning based approach for road distress mapping using VHR images[C]//Proceedings of 2023 IEEE International Geoscience and Remote Sensing Symposium. Pasadena: IEEE, 2023: 1549-1552.
[30]	CHEN Xin, SUN Qun, GUO Wenyue, et al. GA-Net: a geometry prior assisted neural network for road extraction[J]. International Journal of Applied Earth Observation and Geoinformation, 2022, 114: 103004.
[31]	李彦胜, 武康, 欧阳松, 等. 地学知识图谱引导的遥感影像语义分割[J]. 遥感学报, 2024, 28(2): 455-469.
	LI Yansheng, WU Kang, OUYANG Song, et al. Geographic knowledge graph-guided remote sensing image semantic segmentation[J]. National Remote Sensing Bulletin, 2024, 28(2): 455-469.
[32]	李彦胜, 张永军. 耦合知识图谱和深度学习的新一代遥感影像解译范式[J]. 武汉大学学报(信息科学版), 2022, 47(8): 1176-1190.
	LI Yansheng, ZHANG Yongjun. A new paradigm of remote sensing image interpretation by coupling knowledge graph and deep learning[J]. Geomatics and Information Science of Wuhan University, 2022, 47(8): 1176-1190.

网络	F₁值	Precision	Recall	IoU
D-LinkNet（ResNet 34）^[16]	92.8	94.2	91.7	86.8
DeepLabV3（ResNet 34）^[18]	91.8	93.4	90.4	85.0
DANet（ResNet 34）^[22]	92.8	92.9	92.7	86.6
DDRNet^[23]	92.4	92.5	92.4	85.8
AEML U-Nets^[29]	93.7	93.4	94.1	88.2
KDLinkNet（ResNet34）	94.0	94.4	93.6	88.7

网络	F₁值	Precision	Recall	IoU
D-LinkNet（ResNet 34）^[16]	73.7	79.3	68.9	58.8
DeepLabV3（ResNet 34）^[18]	76.1	78.3	74.0	61.4
DANet（ResNet 34）^[22]	77.6	81.2	74.2	63.3
DDRNet^[23]	73.3	72.5	74.1	58.0
KDLinkNet（ResNet34）	77.8	79.9	75.8	63.8

网络	F₁值	Precision	Recall	IoU
D-LinkNet（ResNet 34）^[16]	80.1	81.6	78.7	66.9
DeepLabV3（ResNet 34）^[18]	77.2	78.2	76.2	62.3
DiResNet^[9]	80.2	78.8	81.6	66.8
GA-Net^[30]	81.7	83.2	80.2	69.1
KDLinkNet（ResNet34）	86.1	86.7	85.8	75.8

网络	F₁值	Precision	Recall	IoU
基准网络	92.84	94.23	91.63	86.75
基准+道路连通性优化	93.71	93.78	93.72	88.26
基准+道路边界优化	92.81	93.71	90.93	86.66
基准+道路连通性优化+道路边界优化	93.96	94.43	93.61	88.68