Journal of the Korean Association of Geographic Information Studies (한국지리정보학회지)

The Korean Association of Geographic Information Studies (한국지리정보학회)
권호 표지
DOI QR코드

DOI QR Code

Spatial Replicability Assessment of Land Cover Classification Using Unmanned Aerial Vehicle and Artificial Intelligence in Urban Area

무인항공기 및 인공지능을 활용한 도시지역 토지피복 분류 기법의 공간적 재현성 평가

  • Geon-Ung, PARK (Dept. of Smart Ocean Environmental Energy, Changwon National University) ;
  • Bong-Geun, SONG (Institute of Industrial Technology, Changwon National University) ;
  • Kyung-Hun, PARK (School of Smart & Green Engineering, Changwon National University) ;
  • Hung-Kyu, LEE (School of Smart & Green Engineering, Changwon National University)
  • 박건웅 (창원대학교 스마트환경에너지공학과정) ;
  • 송봉근 (창원대학교 산업기술연구원) ;
  • 박경훈 (창원대학교 스마트그린공학부) ;
  • 이흥규 (창원대학교 스마트그린공학부)
  • Received : 2022.10.26
  • Accepted : 2022.11.22
  • Published : 2022.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

As a technology to analyze and predict an issue has been developed by constructing real space into virtual space, it is becoming more important to acquire precise spatial information in complex cities. In this study, images were acquired using an unmanned aerial vehicle for urban area with complex landscapes, and land cover classification was performed object-based image analysis and semantic segmentation techniques, which were image classification technique suitable for high-resolution imagery. In addition, based on the imagery collected at the same time, the replicability of land cover classification of each artificial intelligence (AI) model was examined for areas that AI model did not learn. When the AI models are trained on the training site, the land cover classification accuracy is analyzed to be 89.3% for OBIA-RF, 85.0% for OBIA-DNN, and 95.3% for U-Net. When the AI models are applied to the replicability assessment site to evaluate replicability, the accuracy of OBIA-RF decreased by 7%, OBIA-DNN by 2.1% and U-Net by 2.3%. It is found that U-Net, which considers both morphological and spectroscopic characteristics, performs well in land cover classification accuracy and replicability evaluation. As precise spatial information becomes important, the results of this study are expected to contribute to urban environment research as a basic data generation method.

현실의 공간을 가상의 공간으로 구현하여 문제를 분석하고 예측하는 기술이 개발되면서, 복잡한 도시 내의 정밀한 공간정보를 취득하는 것이 중요해지고 있다. 본 연구는 복잡한 경관을 가진 도시지역을 대상으로 무인항공기를 통해 영상을 취득하고 고해상도 영상에 적합한 영상분류 기법인 객체기반 영상분석 기법과 의미론적 분할 기법을 적용하여 토지피복 분류를 수행하였다. 또한, 동일시기에 수집된 영상을 바탕으로 인공지능이 학습하지 않은 지역에 대해 각 인공지능 모형의 토지피복 분류 재현성을 확인하고자 하였다. 학습 지역을 대상으로 인공지능 모형을 학습하였을 때, 토지피복 분류 정확도가 OBIA-RF는 89.3%, OBIA-DNN은 85.0%, U-Net의 경우 95.3%로 분석되었다. 재현성을 평가하기 위해 검증 지역에 인공지능 모형을 적용하였을 때, OBIA-RF는 7%, OBIA-DNN은 2.1%, U-Net은 2.3%의 정확도가 감소하였다. 형태학적인 특성과 분광학적인 특성을 모두 고려한 U-Net이 토지피복 분류 정확도 및 재현성 평가에서 우수한 성능을 보이는 것으로 나타났다. 본 연구의 결과는 정밀한 공간정보가 중요해짐에 따라 기초자료 생성 방법으로써 도시환경 연구분야에 기여할 수 있을 것으로 판단된다.

Keywords

Acknowledgement

본 결과물은 2022년도 정부(과학기술정보통신부)의 재원으로 한국연구재단의 지원을 받아 수행된 연구이며(No. NRF-2022R1C1C2009639), 2021년도 교육부의 재원으로 한국연구재단의 지원을 받아 수행된 지자체-대학 협력기반 지역 혁신 사업의 결과입니다(NTIS 과제고유번호: 1345341781, NRF 과제관리번호: 2021RIS-003).

References

  1. Belgiu M., and L. Dragut. 2016. Random forest in remote sensing: A review of applications and future directions. ISPRS Journal of Photogrammetry and Remote Sensing 114:24-31. https://doi.org/10.1016/j.isprsjprs.2016.01.011
  2. Berhane, T.M., C.R. Lane, Q. Wu, B.C. Autrey, O.A. Anenkhonov, V.V. Chepinoga, and H. Liu. 2018. Decision-Tree, RuleBased, and Random Forest Classification of High-Resolution Multispectral Imagery for Wetland Mapping and Inventory. Remote Sensing 10:580.
  3. Chew, R., J. Rineer, R. Beach, M. O'Neil, N. Ujeneza, D. Lapidus, and D.S. Temple, 2020. Deep Neural Networks and Transfer Learning for Food Crop Identification in UAV Images. Drones 4(1):7.
  4. Cho, G.S., J.H. Hyun, G.S. Lee. 2020. Erosion and Sedimentation Monitoring of Coastal Region using Time Series UAV Image. Journal of the Korean Association of Geographic Information Studies 23(2):95-105.
  5. Choi, J.W. 2015. Thematic map generation by using multispectral imaery based on UAV. Journal of the Institute of Construction Technology 34(2):53-59.
  6. Choi, S.K., S.K. Lee, Y.B. Kang, S.K. Seong, D.Y. Choi, and G.H. Kim. 2020. Applicability of Image Classification Using Deep Learning in Small Area: Case of Agricultural Lands Using UAV Image. Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography 38(1):23-33. https://doi.org/10.7848/KSGPC.2020.38.1.23
  7. Chung, D., and I. Lee. 2021. The Optimal GSD and Image Size for Deep Learning Semantic Segmentation Training of Drone Images of Winter Vegetables. Korean Journal of Remote Sensing 37(6-1):1573-1587.
  8. De Luca, G., J.M. N. Silva, S. Cerasoli, J. AraŠjo, J. Campos, S. Di Fazio, and G. Modica. 2019. Object-Based Land Cover Classification of Cork Oak Woodlands using UAV Imagery and Orfeo ToolBox. Remote Sensing 11:1238.
  9. Duro, D.C., S.E. Franklin, and M.G. Dub. 2012. A comparison of pixel-based and object-based image analysis with selected machine learning algorithms for the classification of agricultural landscapes using SPOT-5 HRG imagery. Remote Sensing of Environment 118:259-272. https://doi.org/10.1016/j.rse.2011.11.020
  10. Huang, L., X. Wu, Q. Peng, and X. Yu, 2021. Depth Semantic Segmentation of Tobacco Planting Areas from Unmanned Aerial Vehicle Remote Sensing Images in Plateau Mountains, Journal of Spectroscopy, 2021: 1-14. https://doi.org/10.1155/2021/6687799
  11. Jung, J.S., K.W. Jung, S. Kang, and K.H. Hyun. 2018. Evaluation of Water Cycle Improvement in LH Institute Using SWMM -LID Model. Journal of Korean Society on Water Environment 34(3):308-315.
  12. Kilwenge, R., J. Adewopo, Z. Sun, and M. Schut. 2021. UAV-Based Mapping of Banana Land Area for Village-Level Decision-Support in Rwanda. Remote Sensing 13:4985.
  13. Kim, Y.J., J.H. Oh, and C.N. Lee. 2017a. Electric Power Line Dips Measurement Using Drone-based Photogrammetric Techniques. Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography 35(6):453-460. https://doi.org/10.7848/KSGPC.2017.35.6.453
  14. Kim, M.J., H.S. Bang, and J.W. Lee. 2017b. Use of Unmanned Aerial Vehicle for Forecasting Pine Wood Nematode in Boundary Area: A Case Study of Sejong Metropolitan Autonomous City. Journal of Korean Society of Forest Science 106 (1):100-109.
  15. Kim, J.H., and J.H. Kim. 2018. Accuracy Analysis of Cadastral Control Point and Parcel Boundary Point by Flight Altitude Using UAV. Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography 36(4):223-233.
  16. Kim, E.J., J. Won, and J. Kim. 2019a. Is Seoul Walkable? Assessing a Walkability Score and Examining Its Relationship with Pedestrian Satisfaction in Seoul, Korea. Sustainability 11:6915.
  17. Kim, H.J., J.Y. Kim, and Y.I. Kim. 2019b. A Study on the Object-based Building Extraction Using UAV Imagery. Journal of Korean Society for Geospatial Information Science 27(4):21-28. https://doi.org/10.7319/kogsis.2019.27.4.021
  18. Kim, K.L., and J.H. Ryu. 2020. Generation of Large-scale Map of Surface Sedimentary Facies in Intertidal Zone by Using UAV Data and Object-based Image Analysis (OBIA). Korean Journal of Remote Sensing 36(2-2):277-292.
  19. Kim, N.K., M.S. Park, M.J. Jeong, D.H. Hwang, and H.J. Yoon. 2021. A Study on Field Compost Detection by Using Unmanned Aerial Vehicle Image and Semantic Segmentation Technique based Deep Learning. Korean Journal of Remote Sensing 37(3):367-378.
  20. Kim, S.H., K.H. Park, S.A. Lee, and B.G. Song. 2022. Analysis of Thermal Environment Characteristics by Spatial Type using UAV and ENVI-met. Journal of the Korean Association of Geographic Information Studies 25(1):28-43.
  21. Moon, H.G., S.M. Lee, and J.G. Cha. 2017. Land Cover Classification Using UAV Imagery and Object-Based Image Analysis -Focusing on the Maseo-myeon, Seocheon -gun, Chungcheongnam-do-. Journal of the Korean Association of Geographic Information Studies 20(1):1-14.
  22. Myint, S.W., P. Gober, A. Brazel, S. Grossman -Clarke, and Q. Weng. 2011. Per-pixel vs. object-based classification of urban land cover extraction using high spatial resolution imagery. Remote Sensing of Environment 115:1145-1161. https://doi.org/10.1016/j.rse.2010.12.017
  23. No, H.S., and T.K. Baek. 2017. Real-time Underground Facility Map Production Using Drones. Journal of the Korean Association of Geographic Information Studies 20(4):39-54.
  24. Park, C.H., K. Choi, and I. Lee. 2016. Lane Extraction through UAV Mapping and Its Accuracy Assessment. Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography 34(1):11-19. https://doi.org/10.7848/ksgpc.2016.34.1.11
  25. Park, G., K. Park, B. Song, and H. Lee. 2022. Analyzing Impact of Types of UAV -Derived Images on the Object-Based Classification of Land Cover in an Urban Area. Drones 6:71.
  26. Ronneberger, O., P. Fischer, and T. Brox. 2015. U-net: Convolutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention. Springer, Cham, pp.234-241.
  27. Song, W.K. 2019. Application of UAV for Vegetation Monitoring in Urban Green Space. The Korea Society of Environmental Restoration Technology 22(1):61-72.
  28. Song, B.G., and K.H. Park. 2019. A Study on the Relationship between Land Cover Type and Urban Temperature -. focused on Gimhae city - Journal of the Korean Association of Geographic Information Studies 22(2):65-81.
  29. Song, B.G., and B.H. Moon. 2021. Classification of Non-point Source Types Using UAV Images and Object-Based Image Analysis. J. of the Korean Society for Environmental Technology 22(1):50-58.
  30. Song, B., and K. Park. 2021. Temperature trend analysis associated with landcover changes using time-series data (1980-2019) from 38 weather stations in South Korea. Sustainable Cities and Society 65:102615.
  31. Shin, J.S, T.H. Lee, P.M. Jung, and H.S. Kwon. 2015. A study on land cover map of UAV imagery using an object-based classification method. Jorunal of the Korean Society for Geospatial Information Science 23(4):25-33. https://doi.org/10.7319/kogsis.2015.23.4.025
  32. Sze, V., Y. Chen, T. Yang, and J.S. Emer. 2017. Efficient processing of deep neural networks: A tutorial and survey. Proc. IEEE 105:2295-2329. https://doi.org/10.1109/JPROC.2017.2761740
  33. Takahashi Miyoshi, G., N.N. Imai, A.M. Garcia Tommaselli, M.V. Antunes de Moraes, and E. Honkavaara. 2020. Antunes de Moraes, M.V.; Honkavaara, E. Evaluation of Hyperspectral Multitemporal Information to Improve Tree Species Identification in the Highly Diverse Atlantic Forest. Remote Sensing 12:244.
  34. USGS. 2022. Landsat Soil Adjusted Vegetation Index. https://www.usgs.gov/landsat-missions/landsat-soil-adjusted-vegetation-index. (Accessed October 19, 2022)
  35. Xu, C.X., J.H. Lim, X.M. Jin, and H.C. Yun. 2018. Land Cover Mapping and Availability Evaluation Based on Drone Images with Multi-Spectral Camera. Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography 36(6):589-599.