Korean Journal of Remote Sensing (대한원격탐사학회지)

Korean Society of Remote Sensing (대한원격탐사학회)
권호 표지
DOI QR코드

DOI QR Code

Crop Classification for Inaccessible Areas using Semi-Supervised Learning and Spatial Similarity - A Case Study in the Daehongdan Region, North Korea -

준감독 학습과 공간 유사성을 이용한 비접근 지역의 작물 분류 - 북한 대홍단 지역 사례 연구 -

  • Kwak, Geun-Ho (Department of Geoinformatic Engineering, Inha University) ;
  • Park, No-Wook (Department of Geoinformatic Engineering, Inha University) ;
  • Lee, Kyung-Do (Climate Change and Agroecology Division, National Institute of Agricultural Sciences) ;
  • Choi, Ki-Young (Agricultural and Fisheries Statistics Division, Population & Social Statistics Bureau, Statistics Korea)
  • 곽근호 (인하대학교 공간정보공학과) ;
  • 박노욱 (인하대학교 공간정보공학과) ;
  • 이경도 (국립농업과학원 기후변화생태과) ;
  • 최기영 (통계청 사회통계국 농어업통계과)
  • Received : 2017.08.16
  • Accepted : 2017.10.12
  • Published : 2017.10.30
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper, a new classification method based on the combination of semi-supervised learning with spatial similarity of adjacent pixels is presented for crop classification in inaccessible areas. Iterative classification based on semi-supervised learning is applied to extract reliable training data from both the initial classification result with a small number of training data, and classification results of adjacent pixels are also considered to extract new training pixels with less uncertainty. To evaluate the applicability of the proposed method, a case study of the classification of field crops was carried out using multi-temporal Landsat-8 OLI acquired in the Daehongdan region, North Korea. From a case study, the misclassification of crops and forests, and isolated pixels in the initial classification result were greatly reduced by applying the proposed semi-supervised learning method. In addition, the combination of classification results of adjacent pixels for the extraction of new training data led to the great reduction of both misclassification results and isolated pixels, compared to the initial classification and traditional semi-supervised learning results. Therefore, it is expected that the proposed method would be effectively applied to classify areas in which it is difficult to collect sufficient training data.

이 논문에서는 비접근 지역의 작물 분류를 목적으로 준감독 학습에 인접 화소의 공간 유사성 정보를 결합하는 분류 방법론을 제안하였다. 적은 수의 훈련 자료를 이용한 초기 분류 결과로부터 신뢰성 높은 훈련 자료의 추출을 위해 준감독 학습 기반의 반복 분류를 적용하였으며, 새롭게 훈련 자료 추출시 인접한 화소의 분류 항목을 고려함으로써 불확실성이 낮은 훈련 자료를 추출하고자 하였다. 북한 대홍단에서 수집된 다중시기 Landsat-8 OLI 영상을 이용한 밭작물 구분의 사례 연구를 통해 제안된 분류 방법론의 적용 가능성을 검토하였다. 사례 연구 결과, 초기 분류 결과에서 나타난 작물과 산림의 오분류와 고립된 화소가 제안 분류 방법론에서 완화되었다. 또한 인접 화소의 분류 결과를 고려한 훈련 자료 추출을 통해 이러한 오분류 완화 효과가 더욱 두드러지게 나타났으며, 초기 분류 결과와 기존 준감독 학습에 비해 고립된 화소도 감소되었다. 따라서 비접근 지역으로 인해 훈련 자료의 확보가 어려울 경우 이 연구에서 제안된 방법론이 작물 분류에 유용하게 적용될 수 있을 것으로 기대된다.

Keywords

References

  1. Bernstein, L.S., X. Jin, B. Gregor, and S.M. Adler-Golden, 2012. Quick atmospheric correction code: algorithm description and recent upgrades, Optical Engineering, 51(11): 1-11. https://doi.org/10.1117/1.OE.51.11.114002
  2. Breiman, L., 2001. Random forests, Machine Learning, 45(1): 5-32. https://doi.org/10.1023/A:1010933404324
  3. Bruzzone, L., M. Chi, and M. Marconcini, 2006. A novel transductive SVM for semisupervised classification of remote-sensing images, IEEE Transactions on Geoscience and Remote Sensing, 44(11): 3363-3373. https://doi.org/10.1109/TGRS.2006.877950
  4. Bruzzone, L., and C. Persello, 2009. A novel contextsensitive semisupervised SVM classifier robust to mislabeled training samples, IEEE Transactions on Geoscience and Remote Sensing, 47(7): 2142-2154. https://doi.org/10.1109/TGRS.2008.2011983
  5. Camps-Valls, G., T.V.B. Marsheva, and D. Zhou, 2007. Semi-supervised graph-based hyperspectral image classification, IEEE Transactions on Geoscience and Remote Sensing, 45(10): 3044-3054. https://doi.org/10.1109/TGRS.2007.895416
  6. FAO/WFP, 2012. FAO/WFP Crop and Food Security Assessment Mission to the Democratic People's Republic of Korea, FAO/WEP, Rome, Italy.
  7. Hong, S.Y., B.-K. Min, J.-M. Lee, Y. Kim, and K. Lee, 2012. Estimation of paddy field area in North Korea using RapidEye images, Korean Journal of Soil Science and Fertilizer, 45(6): 1194-1202 (in Korean with English Abstract). https://doi.org/10.7745/KJSSF.2012.45.6.1194
  8. Hong, S.Y., H.-J. Park, K. Jang, S.-I. Na, S.-C. Baek, K.-D. Lee, and J.-B. Ahn, 2015. Status of rice paddy field and weather anomaly in the spring of 2015 in DPRK, Korean Journal of Soil Science and Fertilizer, 48(5): 361-371 (in Korean with English Abstract). https://doi.org/10.7745/KJSSF.2015.48.5.361
  9. KIGAM, 2007. Spectral Library for Land Cover Materials, KIGAM, Daejeon, Korea (in Korean).
  10. Kim, Y., K.-D. Lee, S.-I. Na, S.-Y. Hong, N.-W. Park, and H.Y. Yoo, 2016. MODIS data-based crop classification using selective hierarchical classification, Korean Journal of Remote Sensing, 32(3): 235-244 (in Korean with English Abstract). https://doi.org/10.7780/kjrs.2016.32.3.3
  11. Kwak, G.-H., H.Y. Yoo, and N.-W. Park, 2016. Classification of crop areas in North Korea using multi-temporal Landsat images: a case study in Daehongdan, Proc. of the KSRS Fall Conference 2016, Chungju, Korea, Nov. 3-4, pp. 368-371 (in Korean).
  12. Liu, B., X. Yu, P. Zhang, X. Tan, A. Yu, and Z. Xue, 2017. A semi-supervised convolutional neural network for hyperspectral image classification, Remote Sensing Letters, 8(9): 839-848. https://doi.org/10.1080/2150704X.2017.1331053
  13. Ma, X., H. Wang, and J. Wang, 2016. Semisupervised classification for hyperspectral image based on multi-decision labeling and deep feature learning, ISPRS Journal of Photogrammetry and Remote Sensing, 120: 99-107. https://doi.org/10.1016/j.isprsjprs.2016.09.001
  14. Mathur, A., and G.M. Foody, 2008. Crop classification by support vector machine with intelligently selected training data for an operational application, International Journal of Remote Sensing, 29(8): 2227-2240. https://doi.org/10.1080/01431160701395203
  15. Romaszewski, M., P. Glomb, and M. Cholewa, 2016. Semi-supervised hyperspectral classification from a small number of training samples using a cotraining approach, ISPRS Journal of Photogrammetry and Remote Sensing, 121: 60-76. https://doi.org/10.1016/j.isprsjprs.2016.08.011
  16. Samat, A., J. Li, S. Liu, P. Du, Z. Miao, and J. Luo, 2016. Improved hyperspectral image classification by active learning using pre-designed mixed pixels, Pattern Recognition, 51: 43-58. https://doi.org/10.1016/j.patcog.2015.08.019
  17. Tan, K., E. Li, Q. Du, and P. Du, 2014. An efficient semisupervised classification approach for hyperspectral imagery, ISPRS Journal of Photogrammetry and Remote Sensing, 97: 36-45. https://doi.org/10.1016/j.isprsjprs.2014.08.003
  18. Tuia, D., F. Ratle, F. Pacifici, M.F. Kanevski, and W.J. Emery, 2009. Active learning methods for remote sensing image classification, IEEE Transactions on Geoscience and Remote Sensing, 47(7): 2218-2232. https://doi.org/10.1109/TGRS.2008.2010404
  19. Tuia, D., E. Pasolli, and W.J. Emery, 2011. Using active learning to adapt remote sensing image classifiers, Remote Sensing of Environment, 115(9): 2232-2242. https://doi.org/10.1016/j.rse.2011.04.022
  20. Wang, Z., B. Du, L. Zhang, and L. Zhang, 2016. A batch-mode active learning framework by querying discriminative and representative samples for hyperspectral image classification, Neurocomputing, 179: 88-100. https://doi.org/10.1016/j.neucom.2015.11.062
  21. Wardlow, B.D., L. S. Egbert, and J.H. Kastens, 2007. Analysis of time-series MODIS 250 m vegetation index data for crop classification in the US Central Great Plains, Remote Sensing of Environment, 108(3): 290-310. https://doi.org/10.1016/j.rse.2006.11.021