Improving Road Semantic Segmentation Using Generative Adversarial Network

被引:48
作者
Abdollahi, Abolfazl [1 ]
Pradhan, Biswajeet [1 ,2 ,3 ]
Sharma, Gaurav [4 ]
Maulud, Khairul Nizam Abdul [3 ,5 ]
Alamri, Abdullah [6 ]
机构
[1] Univ Technol Sydney UTS, Sch Informat Syst & Modelling, Fac Engn & IT, Ctr Adv Modelling & Geospatial Informat Syst CAMG, Sydney, NSW 2007, Australia
[2] Sejong Univ, Dept Energy & Mineral Resources Engn, Seoul 05006, South Korea
[3] Univ Kebangsaan Malaysia, Inst Climate Change, Earth Observat Ctr, Bangi 43600, Malaysia
[4] Univ Rochester, Dept Elect & Comp Engn, Rochester, NY 14627 USA
[5] Univ Kebangsaan Malaysia, Fac Engn & Built Environm, Dept Civil Engn, Bangi 43600, Malaysia
[6] King Saud Univ, Dept Geol & Geophys, Coll Sci, Riyadh 11451, Saudi Arabia
关键词
Roads; Image segmentation; Generative adversarial networks; Semantics; Remote sensing; Generators; Feature extraction; GAN; road segmentation; remote sensing; deep learning; U-Net; REMOTE-SENSING IMAGES; EXTRACTION; FRAMEWORK;
D O I
10.1109/ACCESS.2021.3075951
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Road network extraction from remotely sensed imagery has become a powerful tool for updating geospatial databases, owing to the success of convolutional neural network (CNN) based deep learning semantic segmentation techniques combined with the high-resolution imagery that modern remote sensing provides. However, most CNN approaches cannot obtain high precision segmentation maps with rich details when processing high-resolution remote sensing imagery. In this study, we propose a generative adversarial network (GAN)-based deep learning approach for road segmentation from high-resolution aerial imagery. In the generative part of the presented GAN approach, we use a modified UNet model (MUNet) to obtain a high-resolution segmentation map of the road network. In combination with simple pre-processing comprising edge-preserving filtering, the proposed approach offers a significant improvement in road network segmentation compared with prior approaches. In experiments conducted on the Massachusetts road image dataset, the proposed approach achieves 91.54% precision and 92.92% recall, which correspond to a Mathews correlation coefficient (MCC) of 91.13%, a Mean intersection over union (MIOU) of 87.43% and a F1-score of 92.20%. Comparisons demonstrate that the proposed GAN framework outperforms prior CNN-based approaches and is particularly effective in preserving edge information.
引用
收藏
页码:64381 / 64392
页数:12
相关论文
共 48 条
[1]   Road Extraction from High-Resolution Orthophoto Images Using Convolutional Neural Network [J].
Abdollahi, Abolfazl ;
Pradhan, Biswajeet ;
Shukla, Nagesh .
JOURNAL OF THE INDIAN SOCIETY OF REMOTE SENSING, 2021, 49 (03) :569-583
[2]   VNet: An End-to-End Fully Convolutional Neural Network for Road Extraction From High-Resolution Remote Sensing Data [J].
Abdollahi, Abolfazl ;
Pradhan, Biswajeet ;
Alamri, Abdullah .
IEEE ACCESS, 2020, 8 :179424-179436
[3]   Deep Learning Approaches Applied to Remote Sensing Datasets for Road Extraction: A State-Of-The-Art Review [J].
Abdollahi, Abolfazl ;
Pradhan, Biswajeet ;
Shukla, Nagesh ;
Chakraborty, Subrata ;
Alamri, Abdullah .
REMOTE SENSING, 2020, 12 (09)
[4]   Extraction of road features from UAV images using a novel level set segmentation approach [J].
Abdollahi, Abolfazl ;
Pradhan, Biswajeet ;
Shukla, Nagesh .
INTERNATIONAL JOURNAL OF URBAN SCIENCES, 2019, 23 (03) :391-405
[5]   Integrated technique of segmentation and classification methods with connected components analysis for road extraction from orthophoto images [J].
Abdollahi, Arnick ;
Pradhan, Biswajeet .
EXPERT SYSTEMS WITH APPLICATIONS, 2021, 176
[6]   Hierarchical graph-based segmentation for extracting road networks from high-resolution satellite images [J].
Alshehhi, Rasha ;
Marpu, Prashanth Reddy .
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING, 2017, 126 :245-260
[7]  
[Anonymous], U NET CONVOLUTIONAL, DOI DOI 10.1007/978-3-031-20364-0_45
[8]   Fast Local Laplacian Filters: Theory and Applications [J].
Aubry, Mathieu ;
Paris, Sylvain ;
Hasinoff, Samuel W. ;
Kautz, Jan ;
Durand, Fredo .
ACM TRANSACTIONS ON GRAPHICS, 2014, 33 (05)
[9]   Integrated Facility and Assets Management using GIS-Web Application [J].
Bahri, Mohd Aizat Saiful ;
Maulud, Khairul Nizam Abdul ;
Rahman, Muhammad Amartur ;
Oon, Aslinda Oon Ridzuan ;
Hashim, Che Hafify Che .
10TH IGRSM INTERNATIONAL CONFERENCE AND EXHIBITION ON GEOSPATIAL & REMOTE SENSING, 2020, 540
[10]   Semi automatic road extraction from digital images [J].
Bakhtiari H.R.R. ;
Abdollahi A. ;
Rezaeian H. .
Egyptian Journal of Remote Sensing and Space Science, 2017, 20 (01) :117-123