Browse > Article
http://dx.doi.org/10.7780/kjrs.2022.38.6.2.8

The Performance Improvement of U-Net Model for Landcover Semantic Segmentation through Data Augmentation  

Baek, Won-Kyung (Department of Geoinformatics, University of Seoul)
Lee, Moung-Jin (Center for Environmental Data Strategy, Korea Environment Institute)
Jung, Hyung-Sup (Department of Geoinformatics, University of Seoul)
Publication Information
Korean Journal of Remote Sensing / v.38, no.6_2, 2022 , pp. 1663-1676 More about this Journal
Abstract
Recently, a number of deep-learning based land cover segmentation studies have been introduced. Some studies denoted that the performance of land cover segmentation deteriorated due to insufficient training data. In this study, we verified the improvement of land cover segmentation performance through data augmentation. U-Net was implemented for the segmentation model. And 2020 satellite-derived landcover dataset was utilized for the study data. The pixel accuracies were 0.905 and 0.923 for U-Net trained by original and augmented data respectively. And the mean F1 scores of those models were 0.720 and 0.775 respectively, indicating the better performance of data augmentation. In addition, F1 scores for building, road, paddy field, upland field, forest, and unclassified area class were 0.770, 0.568, 0.433, 0.455, 0.964, and 0.830 for the U-Net trained by original data. It is verified that data augmentation is effective in that the F1 scores of every class were improved to 0.838, 0.660, 0.791, 0.530, 0.969, and 0.860 respectively. Although, we applied data augmentation without considering class balances, we find that data augmentation can mitigate biased segmentation performance caused by data imbalance problems from the comparisons between the performances of two models. It is expected that this study would help to prove the importance and effectiveness of data augmentation in various image processing fields.
Keywords
Landcover; Semantic segmentation; U-Net; Data augmentation;
Citations & Related Records
Times Cited By KSCI : 5  (Citation Analysis)
연도 인용수 순위
1 Shi, P., M. Duan, L. Yang, W. Feng, L. Ding, and L. Jiang, 2022. An improved U-net image segmentation method and its application for metallic grain size statistics, Materials, 15(13): 4417. https://doi.org/10.3390/ma15134417   DOI
2 Stoian, A., V. Poulain, J. Inglada, V. Poughon, and D. Derksen, 2019. Land cover maps production with high resolution satellite image time series and convolutional neural networks: Adaptations and limits for operational systems, Remote Sensing, 11(17): 1986. https://doi.org/10.3390/rs11171986   DOI
3 Vali, A., S. Comai, and M. Matteucci, 2020. Deep learning for land use and land cover classification based on hyperspectral and multispectral earth observation data: A review, Remote Sensing, 12(15): 2495. https://doi.org/10.3390/rs12152495   DOI
4 Yu, J.-W., Y.-W. Yoon, W.-K. Baek, and H.S. Jung, 2021. Forest Vertical Structure Mapping Using Two-Seasonal Optic Images and LiDAR DSM Acquired from UAV Platform through Random Forest, XGBoost, and Support Vector Machine Approaches, Remote Sensing, 13(21): 4282. https://doi.org/10.3390/rs13214282   DOI
5 Halevy, A., P. Norvig, and F. Pereira, 2009. The unreasonable effectiveness of data, IEEE Intelligent Systems, 24: 8-12. https://doi.org/10.1109/MIS.2009.36   DOI
6 Ronneberger, O., P. Fischer, and T. Brox, 2015. U-net: Convolutional networks for biomedical image segmentation, Proc. of 2015 International Conference on Medical Image Computing and Computer-Assisted Intervention, Munich, Germany, Oct. 5-9, pp. 234-241.
7 Joo, G., C. Park, and H. Im, 2020. Performance evaluation of machine learning optimizers, Institute of Korean Electrical and Electronics Engineers, 24(3): 766-776. https://doi.org/10.7471/ikeee.2020.24.3.766   DOI
8 Baek, W.-K., H.-S. Jung, and D. Kim, 2020. Oil spill detection of Kerch strait in November 2007 from dual-polarized TerraSAR-X image using artificial and convolutional neural network regression models, Journal of Coastal Research, 102(SI): 137-144. https://doi.org/10.2112/SI102-017.1   DOI
9 Baek, W.-K., Y.-S. Lee, S.-H. Park, and H.-S. Jung, 2021b. Classification of Natural and Artificial Forests from KOMPSAT-3/3A/5 Images Using Deep Neural Network, Korean Journal of Remote Sensing, 37(6-3): 1965-1974 (in Korean with English abstract). https://doi.org/10.7780/kjrs.2021.37.6.3.5   DOI
10 Chen, S., S. Abhinav, S. Saurabh, and G. Abhinav, 2017. Revisting unreasonable effectiveness of data in deep learning era, Proc. of the IEEE International Conference on Computer Vision (ICCV), Venice, Italy, Oct. 22-29, pp. 843-852. https://doi.org/10.48550/arXiv.1707.02968   DOI
11 Jin, Y.W., S. Jia, A.B. Ashraf, and P. Hu, 2020. Integrative data augmentation with U-Net segmentation masks improves detection of lymph node metastases in breast cancer patients, Cancers, 12(10): 2934. https://doi.org/10.3390/cancers12102934   DOI
12 Shorten, C. and T.M. Khoshgoftaar, 2019. A survey on image data augmentation for deep learning, Journal of Big Data, 6(1): 1-48. https://doi.org/10.1186/s40537-019-0197-0   DOI
13 Kim, M.J., S.M. Lee, J.C. Park, H.W. Lee, C.M. Kwon, and I.Y. Won, 2018. A Poisonous Plants Classification System Using Data Augmentation And Transfer Learning, Proc. of the Korea Information Processing Society Conference, Busan, Korea, Nov. 2-3, pp. 660-663.
14 Lee, S.H. and M.J. Lee, 2020. A study on deep learning optimization by land cover classification item using satellite imagery, Korean Journal of Remote Sensing, 36(6-2): 1591-1604 (in Korean with English abstract). https://dx.doi.org/10.7780/kjrs.2020.36.6.2.9   DOI
15 Lee, S.H. and M.J. Lee, 2021. A Study of Establishment and application Algorithm of Artificial Intelligence Training Data on Land use/cover Using Aerial Photograph and Satellite Images, Korean Journal of Remote Sensing, 37(5-1): 871-884 (in Korean with English abstract). https://doi.org/10.7780/kjrs.2021.37.5.1.4   DOI
16 Baek, W.-K., 2022, Phase Unwrapping Using Modified U-Net Regression Model: Focusing on Network Structure and Training Data Optimization, University of Seoul, Seoul, Korea (in Korean with English abstract).
17 Lee, S., W.-K. Baek, H.-S. Jung, and S. Lee, 2020. Susceptibility Mapping on Urban Landslides Using Deep Learning Approaches in Mt. Umyeon, Applied Sciences, 10(22): 8189. https://doi.org/10.3390/app10228189   DOI
18 Lee, S.H. and M.J. Lee, 2022. Comparisons of Multi Resolution Based AI Training Data and Algorithms Using Remote Sensing Focus on Landcover, Frontiers in Remote Sensing, 8: 832753. https://doi.org/10.3389/frsen.2022.832753   DOI
19 Oliveira, G.L., 2019. Encoder-decoder Methods for Semantic Segmentation: Efficiency and Robustness Aspects, Albert-Ludwigs-Universitat Freiburg, Freiburg, Germany.
20 Wu, R., S. Yan, Y. Shan, Q. Dang, and G. Sun, 2015. Deep image: Scaling up image recognition, arXiv preprint arXiv:1501.02876. https://doi.org/10.48550/arXiv.1501.02876   DOI
21 Yuan, K., X. Zhuang, G. Schaefer, J. Feng, L. Guan, and H. Fang, 2021. Deep-Learning-Based Multispectral Satellite Image Segmentation for Water Body Detection, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 14: 7422-7434. https://doi.org/10.1109/JSTARS.2021.3098678   DOI
22 Nazi, Z.A. and T. A. Abir, 2020. Automatic skin lesion segmentation and melanoma detection: Transfer learning approach with u-net and dcnn-svm, Proc. of International Joint Conference on Computational Intelligence. Budapest, Hungary, Nov. 2-4, pp. 371-381.
23 Zhang, P., Y. Ke, Z. Zhang, M. Wang, P. Li, and S. Zhang, 2018. Urban land use and land cover classification using novel deep learning models based on high spatial resolution satellite imagery, Sensors, 18(11): 3717. https://doi.org/10.3390/s18113717   DOI
24 AI Hub, 2020. 2020 Satellite-derived landcover dataset, https://www.aihub.or.kr/aihubdata/data/list.do?pageIndex=1&currMenu=115&topMenu=100&dataSetSn=&srchdataClCode=DATACL001&srchOrder=&SrchdataClCode=DATACL002&searchKeyword=%ED%86%A0%EC%A7%80%ED%94%BC%EB%B3%B5, Accessed on Nov. 30, 2022.
25 Baek, W.-K. and H.-S. Jung, 2021a. Performance comparison of oil spill and ship classification from x-band dual-and single-polarized SAR image using support vector machine, random forest, and deep neural network, Remote Sensing, 13(16): 3203. https://doi.org/10.3390/rs13163203   DOI
26 Baek, W.-K., S.-H. Park, N.-K. Jeong, S. Kwon, W.-J. Jin, and H.-S. Jung, 2017. A study for the techniques and applications of NIR remote sensing based on statical analyses of NIR-related papers, Korean Journal of Remote Sensing, 33(5-3): 889-900 (in Korean with English abstract). https://doi.org/10.7780/kjrs.2017.33.5.3.11   DOI
27 Choi, D., C.J. Shallue, Z. Nado, J. Lee, C.J. Maddison, and G.E. Dahl, 2019. On empirical comparisons of optimizers for deep learning, arXiv preprint arXiv:1910.05446. https://doi.org/10.48550/arXiv.1910.05446   DOI
28 Johnson, J.M. and T.M. Khoshgoftaar, 2019. Survey on deep learning with class imbalance, Journal of Big Data, 6(1): 1-54. https://doi.org/10.1186/s40537-019-0192-5   DOI
29 Liu, M., B. Fu, S. Xie, H. He, F. Lan, Y. Li, P. Lou, and D. Fan, 2021. Comparison of multi-source satellite images for classifying marsh vegetation using DeepLabV3 Plus deep learning algorithm, Ecological Indicators, 125: 107562. https://doi.org/10.1016/j.ecolind.2021.107562   DOI