한국환경복원기술학회지 (Journal of the Korean Society of Environmental Restoration Technology)

  • 제25권6호
  • /
  • Pages.25-33
  • /
  • 2022
  • /
  • 1229-3032(pISSN)
  • /
  • 2733-5011(eISSN)
한국환경복원기술학회 (The Korea Society of Environmental Restoration Technology)
권호 표지
DOI QR코드

DOI QR Code

해안사구 식생의 보전 및 관리를 위한 딥러닝 기반 모니터링

Deep learning-based monitoring for conservation and management of coastal dune vegetation

  • 김동우 (한국환경연구원) ;
  • 구자운 (한국환경연구원) ;
  • 홍예지 (한국환경연구원) ;
  • 김세민 (룰루랩) ;
  • 손승우 (한국환경연구원)
  • Kim, Dong-woo (Korea Environment Institute) ;
  • Gu, Ja-woon (Korea Environment Institute) ;
  • Hong, Ye-ji (Korea Environment Institute) ;
  • Kim, Se-Min (LULULAB INC) ;
  • Son, Seung-Woo (Korea Environment Institute)
  • 투고 : 2022.10.27
  • 심사 : 2022.11.22
  • 발행 : 2022.12.30
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

In this study, a monitoring method using high-resolution images acquired by unmanned aerial vehicles and deep learning algorithms was proposed for the management of the Sinduri coastal sand dunes. Class classification was done using U-net, a semantic division method. The classification target classified 3 types of sand dune vegetation into 4 classes, and the model was trained and tested with a total of 320 training images and 48 test images. Ignored label was applied to improve the performance of the model, and then evaluated by applying two loss functions, CE Loss and BCE Loss. As a result of the evaluation, when CE Loss was applied, the value of mIoU for each class was the highest, but it can be judged that the performance of BCE Loss is better considering the time efficiency consumed in learning. It is meaningful as a pilot application of unmanned aerial vehicles and deep learning as a method to monitor and manage sand dune vegetation. The possibility of using the deep learning image analysis technology to monitor sand dune vegetation has been confirmed, and it is expected that the proposed method can be used not only in sand dune vegetation but also in various fields such as forests and grasslands.

키워드

과제정보

본 논문은 환경부의 환경기술개발사업(과제번호: 2021003360001)의 지원을 받아 한국환경연구원이 수행한 "ICT 기반 생태계 모니터링 기술 및 동식물 탐지 AI 알고리즘 개발(2022-011R)" 사업의 연구결과로 작성되었으며, 일부 재인용이 되었음을 알립니다.

참고문헌

  1. De Giglio, M., Greggio, N., Goffo, F., Merloni, N., Dubbini, M., & Barbarella, M. (2019). Comparison of pixel-and object-based classification methods of unmanned aerial vehicle data applied to coastal dune vegetation communities: Casal borsetti case study. Remote Sensing, 11(12), 1416.
  2. Ettritch, G., Bunting, P., Jones, G., & Hardy, A. (2018). Monitoring the coastal zone using earth observation: application of linear spectral unmixing to coastal dune systems in Wales. Remote Sensing in Ecology and Conservation, 4(4), 303-319. https://doi.org/10.1002/rse2.79
  3. He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 770-778).
  4. Kahng, T. (2006). The landforms and vegetation of coastal sanddune natural monument at Sindu-ri, Taeangun, South Chungcheong Province. Journal of the Korean Geomorphological Association, 13(3), 35-44.
  5. Kim, D. W., Lee, S. H., Yu, J. J., & Son, S. W. (2021). Image analysis technology with deep learning for monitoring the tidal flat ecosystem-Focused on monitoring the Ocypode stimpsoni Ortmann, 1897 in the Sindu-ri tidal flat. Journal of the Korean Society of Environmental Restoration Technology, 24(6), 89-96.
  6. Laporte-Fauret, Q., Lubac, B., Castelle, B., Michalet, R., Marieu, V., Bombrun, L., ... & Rosebery, D. (2020). Classification of atlantic coastal sand dune vegetation using in situ, UAV, and airborne hyperspectral data. Remote Sensing, 12(14), 2222.
  7. Lee, J. Y., Cheong, J. H., & Kim, H. S., (2020). A Study Monitoring the Changes in Taean Sindu Coastal Sand Dune Vegetation. The Journal of Korean Island, 32(3), 187-202. https://doi.org/10.26840/JKI.32.3.187
  8. Perez-Carabaza, S., Boydell, O., & O'Connell, J. (2021, July). Monitoring Threatened Irish Habitats Using Multi-Temporal Multi-Spectral Aerial Imagery and Convolutional Neural Networks. In 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS (pp. 2556-2559). IEEE.
  9. Ronneberger, O., Fischer, P., & Brox, T. (2015, October). U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention (pp. 234-241). Springer, Cham.
  10. Shin, H. S., Kim, T. G., & Yun, C. W. (2015). Phytosociological vegetation classification and flora for the mid-west coast in Korea. Korean Journal of Apiculture.
  11. Soltani, S., Feilhauer, H., Duker, R., & Kattenborn, T. (2022). Transfer learning from citizen science photographs enables plant species identification in UAVs imagery. ISPRS Open Journal of Photogrammetry and Remote Sensing, 100016.
  12. Song, H. S., & Cho, W. (2007). Diversity and zonation of vegetation related micro-topography in Sinduri coastal dune, Korea-focused on the natural monument area. Korean Journal of Environment and Ecology, 21(3), 290-298.
  13. Song, H., Park, G., Park, H., Seo, E., So, S., & Kim, M. (2005). Vegetation and Soil Properties of the Coastal Sand Dune in Sinduri, Taean Gun. Journal of the Korean Society of Environmental Restoration Technology, 8(6), 59-68.
  14. Villoslada, M., Bergamo, T. F., Ward, R. D., Burnside, N. G., Joyce, C. B., Bunce, R. G. H., & Sepp, K. (2020). Fine scale plant community assessment in coastal meadows using UAV based multispectral data. Ecological Indicators, 111, 105979.
  15. Zhang, C., Atkinson, P. M., George, C., Wen, Z., Diazgranados, M., & Gerard, F. (2020). Identifying and mapping individual plants in a highly diverse high-elevation ecosystem using UAV imagery and deep learning. ISPRS Journal of Photogrammetry and Remote Sensing, 169, 280-291. https://doi.org/10.1016/j.isprsjprs.2020.09.025