Urban and rural road surface extraction network based on road topological correlation features

doi:10.11947/j.AGCS.2025.20240124

Abstract

Abstract:

Deep learning methods have become the mainstream technology for classifying and extracting urban and rural road networks based on remote sensing image data. However, the existing methods suffer from issues such as mixed occlusion of neighboring objects (such as vegetation and buildings), long model training time, and high computational complexity, and most of them only focus on independent targets such as road surface, edge line, and center line, resulting in low accuracy of road classification extraction results. In order to fully utilize the constrains of the spatial topological relationship features between road edges and road surfaces, this paper proposes a road surface extraction network, CAS-DeepNet, based on the road topological correlation feature information. Firstly, based on the DeepLabV3+network architecture, the lightweight MobileNetV2 feature extraction network is improved, and the edge enhancement module based on residual connection is embedded to capture road edge information. Secondly, a CS-ASPP structure based on dense connections is designed to improve the model performance. Then, the channel attention mechanism is introduced to effectively fuse multiple branch channels in the image to improve the feature representation ability. Finally, the road connectivity constraints are constructed through the topological association information of road edges to enhance the integrity of the road network extraction results. The experimental results on datasets, such as CHN6-CUG and DeepGlobe, show that the CAS-DeepNet designed in this paper has more advantages over popular methods, such as U-Net++, DeepLabV3+, D-LinkNet, RoadNet, ACNet, and SDUNet, in terms of accuracy rate, recall rate, F₁ score, and overall accuracy rate, significantly improving the accuracy and completeness of the extraction results of road network. This road surface fine extraction method based on road topology correlation features proposed in this study can provide basic support for natural resource survey and monitoring, as well as geospatial environment perception modeling.

Key words: road extraction, edge enhancement module, improved DeepLabV3+, CS-ASPP, attention mechanism

CLC Number:

P237

Yanjun WANG, Xuchao TANG, Cheng WANG, Hengfan CAI. Urban and rural road surface extraction network based on road topological correlation features[J]. Acta Geodaetica et Cartographica Sinica, 2025, 54(1): 75-89.

Figures/Tables 14

Fig. 1

Fig. 2

Tab. 1

Fig. 3

Fig. 4

Fig. 5

Tab. 2

Tab. 3

Fig. 6

Fig. 7

Tab. 4

Tab. 5

Tab. 6

Fig. 8

References 51

[1]	WANG Weixing, YANG Nan, ZHANG Yi, et al. A review of road extraction from remote sensing images[J]. Journal of Traffic and Transportation Engineering (English Edition), 2016, 3(3): 271-282.
[2]	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.
[3]	罗巍, 王东亮. 利用角度纹理特征提取高分辨率遥感影像中城市主干道路[J]. 中国图象图形学报, 2017, 22(11): 1584-1591.
	LUO Wei, WANG Dongliang. Method using the angle texture feature to extract urban trunk road information from high-resolution remote sensing images[J]. Journal of Image and Graphics, 2017, 22(11): 1584-1591.
[4]	戴激光, 杜阳, 方鑫鑫, 等. 多特征约束的高分辨率光学遥感影像道路提取[J]. 遥感学报, 2018, 22(5): 777-791.
	DAI Jiguang, DU Yang, FANG Xinxin, et al. Road extraction method for high resolution optical remote sensing images with multiple feature constraints[J]. Journal of Remote Sensing, 2018, 22(5): 777-791.
[5]	林鹏, 阮仁宗, 王玉强, 等. 一种基于面向对象的城镇道路自动提取方法研究[J]. 地理与地理信息科学, 2016, 32(1): 49-54.
	LIN Peng, RUAN Renzong, WANG Yuqiang, et al. Research on extraction of road based on object oriented in an urban context[J]. Geography and Geo-Information Science, 2016, 32(1): 49-54.
[6]	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
[7]	曹云刚, 王志盼, 慎利, 等. 像元与对象特征融合的高分辨率遥感影像道路中心线提取[J]. 测绘学报, 2016, 45(10): 1231-1240. DOI:. doi: 10.11947/j.AGCS.2016.20160158
	CAO Yungang, WANG Zhipan, SHEN Li, et al. Fusion of pixel-based and object-based features for road centerline extraction from high-resolution satellite imagery[J]. Acta Geodaetica et Cartographica Sinica, 2016, 45(10): 1231-1240. DOI:. doi: 10.11947/j.AGCS.2016.20160158
[8]	ZANG Yu, WANG Cheng, YU Yao, et al. Joint enhancing filtering for road network extraction[J]. IEEE Transactions on Geoscience and Remote Sensing, 2017, 55(3): 1511-1525.
[9]	HE Hao, WANG Shuyang, WANG Shicheng, et al. A road extraction method for remote sensing image based on encoder-decoder network[J]. Journal of Geodesy and Geoinformation Science, 2020, 3(2): 16-25.
[10]	ABDOLLAHI A, PRADHAN B, ALAMRI A. SC-RoadDeepNet: a new shape and connectivity-preserving road extraction deep learning-based network from remote sensing data[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 3143855.
[11]	ALSHEHHI R, MARPU P R, WOON W L, et al. Simultaneous extraction of roads and buildings in remote sensing imagery with convolutional neural networks[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2017, 130: 139-149.
[12]	WANG Qi, GAO Junyu, YUAN Yuan. Embedding structured contour and location prior in siamesed fully convolutional networks for road detection[J]. IEEE Transactions on Intelligent Transportation Systems, 2018, 19(1): 230-241.
[13]	林云浩, 王艳军, 李少春, 等. 一种耦合DeepLab与Transformer的农作物种植类型遥感精细分类方法[J]. 测绘学报, 2024, 53(2): 353-366.DOI:. doi: 10.11947/j.AGCS.2024.20220692
	LIN Yunhao, WANG Yanjun, LI Shaochun, et al. A coupled DeepLab and Transformer approach for fine classification of crop cultivation types in remote sensing[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(2): 353-366.DOI:. doi: 10.11947/j.AGCS.2024.20220692
[14]	SHELHAMER E, LONG J, DARRELL T. Fully convolutional networks for semantic segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(4): 640-651.
[15]	RONNEBERGER O, FISCHER P, BROX T. U-Net: convolutional networks for biomedical image segmentation[EB/OL]. [2024-02-02]. https://arxiv.org/pdf/1505.04597.
[16]	BADRINARAYANAN V, KENDALL A, CIPOLLA R. SegNet: a deep convolutional encoder-decoder architecture for image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(12): 2481-2495.
[17]	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.
[18]	ZHANG Zhengxin, LIU Qingjie, WANG Yunhong. Road extraction by deep residual U-Net[J]. IEEE Geoscience and Remote Sensing Letters, 2018, 15(5): 749-753.
[19]	ZHONG Z L, LI J, CUI W H, et al. Fully convolutional networks for building and road extraction: preliminary results[C]//Proceedings of 2016 IEEE International Geoscience and Remote Sensing Symposium. Beijing: IEEE, 2016: 1591-1694.
[20]	LU Xiaoyan, ZHONG Yanfei, ZHENG Zhuo, et al. GAMSNet: globally aware road detection network with multi-scale residual learning[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2021, 175: 340-352.
[21]	MEI Jie, LI Roujing, GAO Wang, et al. CoANet: connectivity attention network for road extraction from satellite imagery[J]. IEEE Transactions on Image Processing, 2021, 30: 8540-8552.
[22]	CHENG Guangliang, WANG Ying, XU Shibiao, et al. Automatic road detection and centerline extraction via cascaded end-to-end convolutional neural network[J]. IEEE Transactions on Geoscience and Remote Sensing, 2017, 55(6): 3322-3337.
[23]	YANG Xiaofei, LI Xutao, YE Yunming, et al. Road detection and centerline extraction via deep recurrent convolutional neural network U-Net[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(9): 7209-7220.
[24]	JING Ran, GONG Zhaoning, ZHU Wending, et al. Island road centerline extraction based on a multiscale united feature[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2018, 11: 3940-3953.
[25]	XU Zhenhua, SUN Yuxiang, LIU Ming. iCurb: imitation learning-based detection of road curbs using aerial images for autonomous driving[J]. IEEE Robotics and Automation Letters, 2021, 6(2): 1097-1104.
[26]	ZHOU Linying, ZHOU Zhou, NING Hang. Road detection based on edge feature with GAC model in aerial image[J]. International Journal of Pattern Recognition and Artificial Intelligence, 2021, 35(14): 81-93.
[27]	LIANG J, HOMAYOUNFAR N, MA W C, et al. Convolutional recurrent network for road boundary extraction[C]//Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach: IEEE, 2019.
[28]	SHAO Zhenfeng, ZHOU Zifan, HUANG Xiao, et al. MRENet: simultaneous extraction of road surface and road centerline in complex urban scenes from very high-resolution images[J]. Remote Sensing, 2021, 13(2): 239.
[29]	WEI Yao, ZHANG Kai, JI Shunping. Simultaneous road surface and centerline extraction from large-scale remote sensing images using CNN-based segmentation and tracing[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(12): 8919-8931.
[30]	LIU Ruyi, MIAO Qiguang, SONG Jianfeng, et al. Multiscale road centerlines extraction from high-resolution aerial imagery[J]. Neurocomputing, 2019, 329: 384-396.
[31]	CHEN L C, ZHU Y K, PAPANDREOU G, et al. Encoder-decoder with atrous separable convolution for semantic image segmentation[C]//Proceedings of 2018 Computer Vision European Conference. Cham: Springer, 2018: 833-851.
[32]	CHOLLET F. Xception: deep learning with depthwise separable convolutions[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017: 1800-1807.
[33]	CANNY J. Acomputational approach to edge detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1986, PAMI-8(6): 679-698.
[34]	SANDLER M, HOWARD A, ZHU Menglong, et al. MobileNetV2: inverted residuals and linear bottlenecks[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 4510-4520.
[35]	CHEN L C, PAPANDREOU G, KOKKINOS I, et al. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(4): 834-848.
[36]	HUANG Gao, LIU Zhuang, VAN DER MAATEN L, et al. Densely connected convolutional networks[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017: 2261-2269.
[37]	HOU Qibin, ZHOU Daquan, FENG Jiashi. Coordinate attention for efficient mobile network design[C]//Proceedings of 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Nashville: IEEE, 2021: 13708-13717.
[38]	HU Jie, SHEN Li, SUN Gang. Squeeze-and-excitation networks[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 7132-7141.
[39]	WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]//Proceedings of 2018 Computer Vision European Conference. Cham: Springer, 2018: 3-19.
[40]	ZHANG Qinglong, YANG Yubin. SA-Net: shuffle attention for deep convolutional neural networks[C]//Proceedings of 2021 IEEE International Conference on Acoustics, Speech and Signal Processing. Toronto: IEEE, 2021: 2235-2239.
[41]	WANG Qilong, WU Banggu, ZHU Pengfei, et al. ECA-Net: efficient channel attention for deep convolutional neural networks[C]//Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 11531-11539.
[42]	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.
[43]	DEMIR I, KOPERSKI K, LINDENBAUM D, et al. DeepGlobe 2018: a challenge to parse the earth through satellite images[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Salt Lake City: IEEE, 2018.
[44]	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.
[45]	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.
[46]	LIU Yahui, YAO Jian, LU Xiaohu, et al. RoadNet: learning to comprehensively analyze road networks in complex urban scenes from high-resolution remotely sensed images[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(4): 2043-2056.
[47]	DING Xiaohan, GUO Yuchen, DING Guiguang, et al. ACNet: strengthening the kernel skeletons for powerful CNN via asymmetric convolution blocks[C]//Proceedings of 2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019: 1911-1920.
[48]	YANG Mengxing, YUAN Yuan, LIU Ganchao. SDUNet: road extraction via spatial enhanced and densely connected UNet[J]. Pattern Recognition, 2022, 126: 108549.
[49]	PREWITT J M S. Object enhancement and extraction[J]. Picture Processing and Psychopictorics, 1970, 10: 75-149.
[50]	MARR D, HILDRETH E. Theory of edge detection[J]. Proceedings of the Royal Society of London Series B, Biological Sciences, 1980, 207(1167): 187-217.
[51]	ROBINSON G S. Edge detection by compass gradient masks[J]. Computer Graphics and Image Processing, 1977, 6(5): 492-501.

Input	Operator	t	c	n	s
512×512×3	Conv2d	—	32	1	2
256×256×32	Bottleneck	1	16	1	1
256×256×16	Bottleneck	6	24	2	2
128×128×24	Bottleneck	6	32	3	2
64×64×32	Bottleneck	6	64	4	2
32×32×64	Bottleneck	6	96	3	1
32×32×96	Bottleneck	6	160	3	1
32×32×160	Bottleneck	6	320	1	1

数据集	模型	Pre	Rec	F₁	OA	IoU
	U-Net++	0.706 7	0.746 2	0.725 9	0.952 7	0.569 8
	D-LinkNet	0.747 5	0.837 2	0.789 8	0.969 1	0.652 6
	RoadNet	0.760 5	0.795 1	0.777 4	0.967 8	0.635 9
CHN6-CUG	DeepLabV3+	0.738 8	0.810 0	0.772 8	0.960 5	0.629 7
	ACNet	0.786 7	0.843 4	0.814 1	0.963 6	0.684 1
	SDUNet	0.812 6	0.757 2	0.783 9	0.970 4	0.692 7
	CAS-DeepNet	0.781 2	0.860 5	0.818 9	0.974 4	0.693 4
	U-Net++	0.657 5	0.723 6	0.688 9	0.972 3	0.525 5
	D-LinkNet	0.669 1	0.759 3	0.711 3	0.971 8	0.552 0
	RoadNet	0.661 3	0.745 6	0.700 9	0.972 0	0.539 5
DeepGlobe	DeepLabV3+	0.726 9	0.738 3	0.732 5	0.967 0	0.578 0
	ACNet	0.743 3	0.772 7	0.757 7	0.970 3	0.589 7
	SDUNet	0.774 4	0.752 8	0.763 4	0.969 8	0.605 8
	CAS-DeepNet	0.743 5	0.789 2	0.765 6	0.976 4	0.578 0

模型	Params	MACs	Time/s
U-Net++	37.65×10⁶	186.43×10⁹	1.23
D-LinkNet	72.63×10⁶	133.23×10⁹	1.48
RoadNet	65.35×10⁶	167.46×10⁹	1.57
DeepLabV3+	71.42×10⁶	233.45×10⁹	1.82
ACNet	47.39×10⁶	192.36×10⁹	1.69
SDUNet	76.63×10⁶	265.67×10⁹	1.73
CAS-DeepNet	68.34×10⁶	243.82×10⁹	1.71

试验	CEEM	CA	ECA	CS-ASPP	Mob-DeepLabV3+	DeepLabV3+	Pre	Rec	F₁	OA	IoU
1	√				√		0.748 2	0.773 1	0.760 4	0.964 5	0.613 3
2		√	√		√		0.754 9	0.794 6	0.794 5	0.955 4	0.619 7
3				√	√		0.761 1	0.769 3	0.765 2	0.957 1	0.619 8
4	√	√	√	√	√		0.781 2	0.860 5	0.818 9	0.974 4	0.693 4
5					√		0.711 0	0.769 4	0.739 0	0.961 1	0.586 1
6						√	0.738 8	0.810 0	0.772 8	0.960 5	0.629 7

模型	Pre	Rec	F₁	OA	IoU
Prewitt+Mob-DeepLabV3+	0.739 8	0.732 3	0.736 0	0.957 6	0.582 4
Laplacian+Mob-DeepLabV3+	0.744 2	0.740 8	0.742 5	0.961 0	0.590 5
Robert+Mob-DeepLabV3+	0.745 8	0.760 9	0.753 2	0.958 0	0.604 2
Sobel+Mob-DeepLabV3+	0.747 1	0.768 7	0.757 7	0.957 5	0.610 0
Canny+Mob-DeepLabV3+	0.748 2	0.773 1	0.760 4	0.964 5	0.613 3