新型语义分割D-UNet的建筑物提取

龙丽红; 朱宇霆; 闫敬文; 刘敬瑾; 王宗跃

doi:10.11834/jrs.20211029

模型与方法 | 浏览量 : 0 下载量: 331 CSCD: 0 更多指标

R-PDF
PDF
导出
分享
收藏
专辑

新型语义分割D-UNet的建筑物提取
New building extraction method based on semantic segmentation
2023年27卷第11期页码：2593-2602
纸质出版日期： 2023-11-07 ，
DOI： 10.11834/jrs.20211029

扫描看全文

龙丽红，朱宇霆，闫敬文，刘敬瑾，王宗跃.2023.新型语义分割D-UNet的建筑物提取.遥感学报，27（11）： 2593-2602

Long L H，Zhu Y T，Yan J W，Liu J J and Wang Z Y. 2023. New building extraction method based on semantic segmentation. National Remote Sensing Bulletin， 27（11）：2593-2602
龙丽红，朱宇霆，闫敬文，刘敬瑾，王宗跃.2023.新型语义分割D-UNet的建筑物提取.遥感学报，27（11）： 2593-2602 DOI： 10.11834/jrs.20211029.

Long L H，Zhu Y T，Yan J W，Liu J J and Wang Z Y. 2023. New building extraction method based on semantic segmentation. National Remote Sensing Bulletin， 27（11）：2593-2602 DOI： 10.11834/jrs.20211029.

摘要

高分辨率遥感图像语义分割在航空图像分析领域中具有重要的理论价值和应用价值。但由于高分辨率遥感图像中建筑物语义的丰富性和图像背景的复杂性，以往的分割方法往往容易产生边缘模糊、细节信息丢失和分辨率低等缺点。为了解决高分辨率卫星图像语义分割边界模糊和信息丢失的问题，本文提出一种端到端的卷积神经网络Dilated-UNet（D-UNet）。首先，通过改进U-Net网络结构，采用Dilation技术拓展四通道的多尺度空洞卷积模块，每个通道采用不同的卷积扩张率来识别多尺度语义信息，从而提取更丰富的细节信息。其次，设计了一种交叉熵和Dice系数的联合损失函数，更好的训练模型以达到预期分割效果。最后，在Inria航空图像数据集上进行综合评估与检验。实验结果表明，本文提出的遥感图像分割方法能够有效地从高分辨率遥感图像中进行像素级城市建筑物的分割，与其他方法相比，分割精度更高，具有较高的实际应用价值。

Abstract

Semantic segmentation of high-resolution remote sensing image has important theoretical and practical value in the field of aerial image analysis. However

the traditional segmentation methods are prone to edge blur

loss of detail information

and low resolution due to the richness of building semantics and the complexity of image background in high-resolution remote sensing images.

An end-to-end convolutional neural network called Dilated-UNet (D-UNet) is proposed to solve the problem of fuzzy boundary and information loss in high-resolution satellite image semantic segmentation. First

the U-Net network structure is improved and the multiscale dilated convolution module of four channels is expanded using the division technology. Each channel uses different convolution expansion rates to identify the multiscale semantic information for extracting richer detailed information. Second

a joint loss function of cross entropy and Dice coefficient is designed to achieve the desired segmentation effect.

The model is comprehensively evaluated and tested on the Inria aerial image dataset. Experimental results show that the proposed remote sensing image segmentation method can effectively segment urban buildings at pixel level from high-resolution remote sensing images

and the segmentation accuracy is higher and is therefore better than those of other methods.

Our proposed D-UNet can deliver automatic building segmentation from high-resolution remote sensing images with high accuracy. Thus

it is a useful tool for practical application scenarios.

关键词

遥感图像语义分割多尺度空洞卷积图像处理

Keywords

remote sensing imagessemantic segmentationmultiscaledilated convolutionimage processing

references

Bischke B, Helber P, Folz J, Borth D and Dengel A. 2019. Multi-task learning for segmentation of building footprints with deep neural networks//Proceedings of 2019 IEEE International Conference on Image Processing. Taipei, China: IEEE: 1480-1484 [DOI: 10.1109/ICIP.2019.8803050http://dx.doi.org/10.1109/ICIP.2019.8803050]

Chaurasia A and Culurciello E. 2017. LinkNet: exploiting encoder representations for efficient semantic segmentation//Proceedings of 2017 IEEE Visual Communications and Image Processing. St. Petersburg: IEEE [DOI: 10.1109/vcip.2017.8305148http://dx.doi.org/10.1109/vcip.2017.8305148]

Felzenszwalb P F and Huttenlocher D P. 2004. Efficient graph-based image segmentation. International Journal of Computer Vision, 59(2): 167-181 [DOI: 10.1023/B:VISI.0000022288.19776.77http://dx.doi.org/10.1023/B:VISI.0000022288.19776.77]

Gui Y F and Wu J P. 2011. An image edge detection method base on Laplacian operator and degree of gray relation. Journal of Shantou University (Natural Science), 26(2): 69-73

桂预风, 吴建平. 2011. 基于Laplacian算子和灰色关联度的图像边缘检测方法. 汕头大学学报(自然科学版), 26(2): 69-73 [DOI: 10.3969/j.issn.1001-4217.2011.02.011http://dx.doi.org/10.3969/j.issn.1001-4217.2011.02.011]

Guo M Q, Liu H, Xu Y Y and Huang Y. 2020. Building extraction based on U-net with an attention block and multiple losses. Remote Sensing, 12 (9): 1400 [DOI: 10.3390/rs12091400http://dx.doi.org/10.3390/rs12091400]

Han B B, Zhang Y T, Pan Z X, Tai X Q and Li F F. 2020. Building extraction in remote sensing image based on residual dense pyramid network. Foreign Electronic Measurement Technology, 39(5): 5-10

韩彬彬, 张月婷, 潘宗序, 台宪青, 李芳芳. 2020. 基于稠密金字塔网络的遥感影像建筑物提取. 国外电子测量技术, 39(5): 5-10 [DOI: 10.19652/j.cnki.femt.1901954http://dx.doi.org/10.19652/j.cnki.femt.1901954]

He K M, Zhang X Y, Ren S Q and Sun J. 2016. Deep residual learning for image recognition//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE [DOI: 10.1109/CVPR.2016.90http://dx.doi.org/10.1109/CVPR.2016.90]

Hinton G E, Osindero S and Teh Y W. 2006. A fast learning algorithm for deep belief nets. Neural Computation, 18(7): 1527-1554 [DOI: 10.1162/neco.2006.18.7.1527http://dx.doi.org/10.1162/neco.2006.18.7.1527]

Hu X Y, Shen J J, Shan J and Pan L. 2013. Local edge distributions for detection of salient structure textures and objects. IEEE Geoscience and Remote Sensing Letters, 10(3): 466-470 [DOI: 10.1109/LGRS.2012.2210188http://dx.doi.org/10.1109/LGRS.2012.2210188]

Huang B H, Lu K K, Audeberr N, Khalel A, Tarabalka Y, Malof J, Boulch A, Le Saux B, Collins L, Bradbury K, Lefèvre S and El-Saban M. 2018. Large-scale semantic classification: outcome of the first year of Inria aerial image labeling benchmark//Proceedings of 2018 IEEE International Geoscience and Remote Sensing Symposium. Valencia: IEEE [DOI: 10.1109/IGARSS.2018.8518525http://dx.doi.org/10.1109/IGARSS.2018.8518525]

Hui J, Qin Q M, Xu W and Sui J. 2019. Instance segmentation of buildings from high-resolution remote sensing images with multitask learning. Acta Scientiarum Naturalium Universitatis Pekinensis, 55(6): 1067-1077

惠健, 秦其明, 许伟, 隋娟. 2019. 基于多任务学习的高分辨率遥感影像建筑实例分割. 北京大学学报(自然科学版), 55(6): 1067-1077 [DOI: 10.13209/j.0479-8023.2019.106http://dx.doi.org/10.13209/j.0479-8023.2019.106]

Kang M, Xu Q G and Wang B S. 2008. A roberts’ adaptive edge detection method. Journal of Xi’an Jiaotong University, 42(10): 1240-1244

康牧, 许庆功, 王宝树. 2008. 一种Roberts自适应边缘检测方法. 西安交通大学学报, 42(10): 1240-1244 [DOI: 10.7652/xjtuxb200810010http://dx.doi.org/10.7652/xjtuxb200810010]

Li L, Chai W T and Mei S L. 2013. Segmentation of remote sensing images based on adaptive global threshold and fused markers. Transactions of the Chinese Society for Agricultural Machinery, 44(7): 222-228

李丽, 柴文婷, 梅树立. 2013. 基于自适应全局阈值融合标记的遥感图像建筑群分割. 农业机械学报, 44(7): 222-228 [DOI: 10.6041/j.issn.1000-1298.2013.07.039http://dx.doi.org/10.6041/j.issn.1000-1298.2013.07.039]

Li X, Jiang Y C, Peng H and Yin S. 2019. An aerial image segmentation approach based on enhanced multi-scale convolutional neural network//Proceedings of 2019 IEEE International Conference on Industrial Cyber Physical Systems. Taipei, China: IEEE: 47-52 [DOI: 10.1109/ICPHYS.2019.8780187http://dx.doi.org/10.1109/ICPHYS.2019.8780187]

Maggiori E, Tarabalka Y, Charpiat G and Alliez P. 2017. Can semantic labeling methods generalize to any city? The inria aerial image labeling benchmark//Proceedings of 2017 IEEE International Geoscience and Remote Sensing Symposium. Fort Worth: IEEE [DOI: 10.1109/IGARSS.2017.8127684http://dx.doi.org/10.1109/IGARSS.2017.8127684]

Pan X R, Yang F, Gao L R, Chen Z C, Zhang B, Fan H R and Ren J C. 2019. Building extraction from high-resolution aerial imagery using a generative adversarial network with spatial and channel attention mechanisms. Remote Sensing, 11(8): 917 [DOI: 10.3390/rs11080917http://dx.doi.org/10.3390/rs11080917]

Ronneberger O, Fischer P and Brox T. 2015. U-Net: convolutional networks for biomedical image segmentation//Proceedings of the 18th International Conference on Medical Image Computing and Computer-Assisted Intervention. Munich: Springer [DOI: 10.1007/978-3-319-24574-4_28http://dx.doi.org/10.1007/978-3-319-24574-4_28]

Sebastian C, Imbriaco R, Bondarev E and de With P H N. 2020. Contextual pyramid attention network for building segmentation in aerial imagery. arXiv: 2004.07018

Shelhamer E, Long J and Darrell T. 2017. Fully convolutional networks for semantic segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(4): 640-651 [DOI: 10.1109/TPAMI.2016.2572683http://dx.doi.org/10.1109/TPAMI.2016.2572683]

Simonyan K and Zisserman A. 2014. Very deep convolutional networks for large-scale image recognition. arXiv: 1409.1556

Sun W W, Yang G, Chen C, Chang M H, Huang K, Meng X Z and Liu L Y. 2020. Development status and literature analysis of China’s earth observation remote sensing satellites. Journal of Remote Sensing, 24(5): 479-510

孙伟伟, 杨刚, 陈超, 常明会, 黄可, 孟祥珍, 刘良云. 2020. 中国地球观测遥感卫星发展现状及文献分析. 遥感学报, 24(5): 479-510 [DOI: 10.11834/jrs.20209464http://dx.doi.org/10.11834/jrs.20209464]

Wang J, Yang X C, Qin X B, Ye X and Qin Q M. 2015. An efficient approach for automatic rectangular building extraction from very high resolution optical satellite imagery. IEEE Geoscience and Remote Sensing Letters, 12(3): 487-491 [DOI: 10.1109/LGRS.2014.2347332http://dx.doi.org/10.1109/LGRS.2014.2347332]

Wang P Q, Chen P F, Yuan Y, Liu D, Huang Z H, Hou X D and Cottrell G. 2018. Understanding convolution for semantic segmentation//Proceedings of 2018 IEEE Winter Conference on Applications of Computer Vision. Lake Tahoe: IEEE [DOI: 10.1109/WACV.2018.00163http://dx.doi.org/10.1109/WACV.2018.00163]

Wei Y and Long H Y. 2019. A building segmentation method based on deep convolution networks for remote sensing imagery. Computer Technology and Development, 29(6): 57-61

余威, 龙慧云. 2019. 基于深度卷积网络的遥感影像建筑物分割方法. 计算机技术与发展, 29(6): 57-61 [DOI: 10.3969/j.issn.1673-629X.2019.06.012http://dx.doi.org/10.3969/j.issn.1673-629X.2019.06.012]

Wu S H, Wu Y Q and Zhou J J. 2018. Multi-level thresholding for remote sensing image of urban area based on line intercept histogram. CAAI Transactions on Intelligent Systems, 13(2): 227-235

吴诗婳, 吴一全, 周建江. 2018. 直线截距直方图城区遥感图像多阈值分割. 智能系统学报, 13(2): 227-235 [DOI: 10.11992/tis.201609012http://dx.doi.org/10.11992/tis.201609012]

Yu F and Koltun V. 2016. Multi-scale context aggregation by dilated convolutions. arXiv: 1511.07122

Zhao H S, Shi J P, Qi X J, Wang X G and Jia J Y. 2017. Pyramid scene parsing network//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE: 2881-2890 [DOI: 10.1109/CVPR.2017.660http://dx.doi.org/10.1109/CVPR.2017.660]

Zou B X, Zhang R and Miao J. 2013. Improved Prewitt method for image edge detection. Microelectronics and Computer, 30(5): 23-26

邹柏贤, 张然, 苗军. 2013. Prewitt图像边缘检测方法的改进. 微电子学与计算机, 30(5): 23-26 [DOI: 10.19304/j.cnki.issn1000-7180.2013.05.005http://dx.doi.org/10.19304/j.cnki.issn1000-7180.2013.05.005]

文章被引用时，请邮件提醒。

提交

结合协方差池化与跨尺度特征提取的高光谱分类

边缘增强的BECU-Net模型高分辨率遥感影像耕地提取

MAR20：遥感图像军用飞机目标识别数据集

多元信息监督的遥感图像有向目标检测

融合余弦退火与空洞卷积的遥感影像语义分割