Früher|Nächster

Road extraction of high-resolution satellite remote sensing images in U-Net network with consideration of connectivity

  • role： First author第一作者
  • Affiliation：

    School of Geography and Information Engineering, China University of Geosciences (Wuhan), Wuhan 430078, China

  • Email：bz21422@163.com
  • Introduction：王斌，1996年生，男，硕士研究生，研究方向为遥感图像处理和地质大数据。E-mail：bz21422@163.com
WANG Bin1,  
  • Affiliation：

    School of Geography and Information Engineering, China University of Geosciences (Wuhan), Wuhan 430078, China

CHEN Zhanlong1,  
  • Affiliation：

    School of Geography and Information Engineering, China University of Geosciences (Wuhan), Wuhan 430078, China

WU Liang1,  
  • Affiliation：

    Xi 'an Research Institute of Surveying and Mapping, Xi 'an 710054, China

    State Key Laboratory of Geo-Information Engineering, Xi 'an 710054, China

XIE Peng23,  
  • role： Corresponding author通信作者
  • Affiliation：

    College of Geomatics and Geoinformation,Guilin University of Technology, Guilin 541004, China

  • Email：bodhifan@163.com
  • Introduction：范冬林，1988年生，男，讲师，研究方向为图像智能解译。E-mail：bodhifan@163.com
FAN Donglin4*,  
  • Affiliation：

    College of Geomatics and Geoinformation,Guilin University of Technology, Guilin 541004, China

FU Bolin4

Resümee

Existing remote sensing image interpretation methods can obtain high-precision classification results but also have some defects. On the one hand, the problem of disconnection in the road extraction result of remote sensing image reduces the extraction precision. On the other hand, this problem affects the integrity of the road morphology. Thus, the extraction result cannot be directly applied to spatial decision making and analysis. Therefore, this study solves the problem of disconnection in the road extraction results of remote sensing images. A road extraction method for high-resolution remote sensing image of U-Net network with consideration of connectivity is also proposed.On the basis of the advantages of global feature representation of U-Net network in road extraction of high-resolution remote sensing images, this study proposes a broken road repair method that considers connectivity to improve the local features of U-Net network. First, the sample data after data enhancement and data volume expansion are used as an input of the U-Net network to train the model and perform road extraction of the optimal model. Then, the road breakpoint detection, the road breakpoint clustering, and the cubic polynomial curve fitting are organically combined to optimize the result because of the broken road appearing in the extraction result.The proposed method is practical and feasible according to the verification of experimental results. The method is also universal. The experiment shows that the accuracy and shape integrity of the road extraction in this method are improved significantly compared with similar networks. The precision is 86.25%, the recall rate is 85.50%, and F1-score reaches 85.87%.The roads extracted by the proposed method have good connectivity and morphological integrity. The resulting image can be directly applied to geographic decision analysis, especially for post-disaster path planning. The proposed method has certain reference significance for the occurrence of similar disconnection problems in the classification results of linear objects, such as roads, power grids, orbits, and rivers.

Schlüsselwort

remote sensing;the integrity of the road morphology;U-Net network;high-resolution remote sensing images;connectivity;broken road

References

  1. 1.
    Biswas S, Ghoshal D and Hazra R. 2016. A new algorithm of image segmentation using curve fitting based higher order polynomial smoothing. Optik, 127(20): 8916-8925
  2. 2.
    Cardim G P, Da Silva E A, Dias M A, Bravo I and Gardel A. 2018. Statistical evaluation and analysis of road extraction methodologies using a unique dataset from remote sensing. Remote Sensing, 10(4): 620
  3. 3.
    Chen Y S, Lin Z H, Zhao X, Wang G and Gu Y F. 2014. Deep learning-based classification of hyperspectral data. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 7(6): 2094-2107
  4. 4.
    Chowdhury K, Chaudhuri D, Pal A K and Samal A. 2019. Seed selection algorithm through K-means on optimal number of clusters. Multimedia Tools and Applications, 78(13): 18617-18651
  5. 5.
    He H, Wang S C, Yang D F, Wang S Y and Liu X. 2019. An road extraction method for remote sensing image based on Encoder-Decoder network. Acta Geodaetica et Cartographica Sinica, 48(3): 330-338
  6. 6.
    Hoberg T and Rottensteiner F. 2010. Classification of settlementareas in remote sensing imagery using conditional random fields. International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 38(7): 53-58
  7. 7.
    Hu J S, Kang J R, Zhang Q, Liu P C and Zhu M D. 2018. An improving image boundary tracking algorithm based on eight neighborhood. Bulletin of Surveying and Mapping, (12): 21-25
  8. 8.
    Kim J, Kim J, Jang G J and Lee M. 2017. Fast learning method for convolutional neural networks using extreme learning machine and its application to lane detection. Neural Networks, 87: 109-121
  9. 9.
    Kim J H, Lee H, Hong S J, Kim S, Park J, Hwang J Y and Choi J P. 2019. Objects segmentation from high-resolution aerial images using U-Net with pyramid pooling layers. IEEE Geoscience and Remote Sensing Letters, 16(1): 115-116
  10. 10.
    Lam L, Lee S W and Suen C Y. 1992. Thinning methodologies-a comprehensive survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 14(9): 869-885
  11. 11.
    Li H C, Chu H, Huo Y H. 2018. Multi-feature multiple kernels SVM-based urban road extraction. Bulletin of Surveying and Mapping, (2): 72-77
  12. 12.
    Li L W, Liu J P and Yin Z W. 2005. Road extraction from high resolution remote sensing image based on mathematic morphology. Remote Sensing Information, (5): 9-11
  13. 13.
    Long Y, Gong Y P, Xiao Z F and Liu Q. 2017. Accurate object localization in remote sensing images based on convolutional neural networks. IEEE Transactions on Geoscience and Remote Sensing, 55(5): 2486-2498
  14. 14.
    Martin D R, Fowlkes C C and Malik J. 2004. Learning to detect natural image boundaries using local brightness, color, and texture cues. IEEE Transactions on Pattern Analysis and Machine Intelligence, 26(5): 530-549
  15. 15.
    Meng X L, Currit N, Wang L and Yang X J. 2012. Detect residential buildings from Lidar and aerial photographs through object-oriented land-use classification. Photogrammetric Engineering and Remote Sensing, 78(1): 35-44
  16. 16.
    Mittal K, Aggarwal G and Mahajan P. 2019. Performance study of K-nearest neighbor classifier and K-means clustering for predicting the diagnostic accuracy. International Journal of Information Technology, 11(3): 535-540
  17. 17.
    Mnih V. 2013. Machine Learning for Aerial Image Labeling. Toronto: University of Toronto
  18. 18.
    Oon A, Shafri H Z M, Lechner A M and Azhar B. 2019. Discriminating between large-scale oil palm plantations and smallholdings on tropical peatlands using vegetation indices and supervised classification of LANDSAT-8. International Journal of Remote Sensing, 40(19): 7312-7328
  19. 19.
    Pan J P and Li Z. 2009. Road extraction based on remote sensing classification and mathematic morphology. Computer Engineering and Applications, 46(34): 213-214, 227.
  20. 20.
    Panboonyuen T, Jitkajornwanich K, Lawawirojwong S, Srestasathiern P and Vateekul P. 2017. Road segmentation of remotely-sensed images using deep convolutional neural networks with landscape metrics and conditional random fields. Remote Sensing, 9(7): 680
  21. 21.
    Peng D F, Zhang Y J and Guan H Y. 2019. End-to-End change detection for high resolution satellite images using improved UNet++. Remote Sensing, 11(11): 1382
  22. 22.
    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: 234-241
  23. 23.
    Saeed K, Tabędzki M, Rybnik M and Adamski M. 2010. K3M: a universal algorithm for image skeletonization and a review of thinning techniques. International Journal of Applied Mathematics and Computer Science, 20(2): 317-335
  24. 24.
    Salari E and Siy P. 1984. The ridge-seeking method for obtaining the skeleton of digital images. IEEE Transactions on Systems, Man, and Cybernetics, SMC-14(3): 524-528
  25. 25.
    Salehnia N, Salehnia N, Ansari H, Kolsoumi S and Bannayan M. 2019. Climate data clustering effects on arid and semi-arid rainfed wheat yield: a comparison of artificial intelligence and K-means approaches. International Journal of Biometeorology, 63(7): 861-872
  26. 26.
    Sobel L. 1978. Neighborhood coding of binary images for fast contour following and general binary array processing. Computer Graphics and Image Processing, 8(1): 127-135
  27. 27.
    Tang Z M, Luo L K, Peng H and Li S H. 2018. A joint residual network with paired ReLUs activation for image super-resolution. Neurocomputing, 273: 37-46
  28. 28.
    Wiedemann C. 2003. External evaluation of road networks[J]. ISPRS Archives, 34(3/W8): 93-98
  29. 29.
    Xu R and Wunsch D. 2005. Survey of clustering algorithms. IEEE Transactions on Neural Networks, 16(3): 645-678
  30. 30.
    Xu Y Y, Xie Z, Feng Y X and Chen Z L. 2018. Road extraction from high-resolution remote sensing imagery using deep learning. Remote Sensing, 10(9): 1461
  31. 31.
    Yuan M, Liu Z and Wang F. 2019. Using the wide-range attention U-Net for road segmentation. Remote Sensing Letters, 10(5): 506-515
  32. 32.
    Zhang M M, Li W and Du Q. 2018. Diverse region-based CNN for hyperspectral image classification. IEEE Transactions on Image Processing, 27(6): 2623-2634
  33. 33.
    Zhang Y D, Pan C C, Sun J D and Tang C S. 2018. Multiple sclerosis identification by convolutional neural network with dropout and parametric ReLU. Journal of Computational Science, 28: 1-10
  34. 34.
    Zhang Y H, He J, Kan X, Xia G H, Zhu L L and Ge T T. 2018. Summary of road extraction methods for remote sensing images. Computer Engineering and Applications, 54(13): 1-10, 51
  35. 35.
    Zhang Z X, Liu Q J and Wang Y H. 2018. Road extraction by deep residual U-Net. IEEE Geoscience and Remote Sensing Letters, 15(5): 749-753

Lesen Sie die ganze Passage

The above content is generated by Large Model Translation. The translated content is for reference only. We do not assume any commercial or legal responsibilty for any consequences arising from the use of our website