본문 바로가기
KERD
논문KCI 등재

UAV 영상 기반 벼 양분 결핍 조기 진단 기계학습 모델 구축

Development of a Machine Learning Model for Early Diagnosis of Nutrient Deficiency in Rice Based on UAV Images

이상훈|정회정|상완규|백재경|이지현|박혁진|장성율
농업생명과학연구원 (Institute of Agriculture & Life Science, Gyeongsang National University)|2024|농업생명과학연구 (Journal of Agriculture & Life Science)|제58권 제2호|pp. 53-64|DOI 10.14397/jals.2024.58.2.53

2

조회

0

다운로드

1

인용

0

참고문헌

키워드

무인항공기기계학습Machine learningRice벼Unmanned aerial vehicle양분 결핍 진단Nutrient deficiency diagnosis

초록

작물의 스트레스 조기 진단은 농업에 있어 빠른 대응을 가능하게 해 피해를 경감시킬 수 있어 중요한 기술이다. 기존의 스트레스 진단이 가진파괴적인 형식의 시료 채집과 양분 분석에 많은 노동력을 필요로 한다는 단점 극복을 위해 새로운 기술 개발이 필요하다. 미래에는 대단위 영상을이용한 생육 진단 기술에 대한 수요가 높아질 것으로 예상되어 이를 이용한 연구를 진행하였다. 본 연구는 2023년 경상남도 밀양시에 위치한국립식량과학원 실험 포장에서 수행되었으며, 무인항공기(UAV)를 이용하여 양분 결핍 처리(관행시비, 질소 결핍, 인 결핍, 칼륨 결핍, 무비)에따른 벼의 생육을 조사했다. UAV를 이용해 생육 기간 중 총 6회에 걸쳐 포장을 촬영하였고, 영상을 기반으로 11개의 식생 지수를 산출하여기계학습을 통해 양분 결핍을 진단하는 모델을 구축하여 평가했다. 연구 결과에 따르면, 엽록소 함량과 관련된 지수인 NDRE (NormalizedDifference Red Edge)가 가장 높은 중요도를 나타내어 벼의 양분 상태를 효과적으로 진단하는 데 유용하다는 것을 확인하였다. 6개의 각 단계별로모델을 평가하였을 때 모든 단계에서 accuracy가 0.7 이상으로 나타났다. 조기 진단을 위해 첫 촬영 날짜인 7월 5일의 자료로 모델을 만들어다른 회차에 적용하여 모델의 성능을 평가한 결과, 5개의 모든 단계에서 0.9 이상의 accuracy를 얻었다. 종합적으로, UAV 영상 기반의 식생지수를 활용한 양분 결핍 진단은 벼의 생육을 조기에 예측하는 데 효과적이며, 이는 정밀 농업 분야에서 시간과 노동을 절감하고 양분 관리를개선하는 데 도움이 될 것으로 기대된다.

Abstract

Early diagnosis of crop stress is a crucial technology in agriculture as it enables rapid responses to mitigate damages. Traditionalstress diagnosis requires destructive sampling and nutrient analysis, which consumes significant labor. To overcome these limitations,there is a need for new technological developments. It is anticipated that there will be a growing demand for diagnostic technologiesbased on large-scale imagery in the future. This study was conducted at the National Institute of Crop Science experimental field inMiryang-si, Gyeongsangnam-do in 2023. We investigated the growth of rice under nutrient deficiency treatments (conventional fertilization,nitrogen deficiency, phosphorus deficiency, potassium deficiency and no fertilization) using UAV. Throughout the growth period, thefield was imaged six times using UAV, and based on these images, 11 vegetation indices were calculated. A machine learning modelfor diagnosing nutrient deficiencies was developed and evaluated. According to the research results, the index related to chlorophyllcontent, NDRE (Normalized Difference Red Edge), showed the highest importance, proving to be effective in diagnosing the nutrientstatus of rice. When evaluating the model for each of the six stages, an accuracy of over 0.7 was consistently achieved. For early diagnosis, the model was built using data from the first imaging date, July 5th, and applied to subsequent sessions. The performanceof the model was evaluated, and an accuracy of over 0.9 was obtained for all five stages. Overall, the UAV imagery-based vegetationindices for diagnosing nutrient deficiencies are effective in predicting the growth of rice early on. This is expected to help save timeand labor in precision agriculture and improve nutrient management.

참고문헌 (49)

  1. Akbarzadeh S, "Plant discrimination by Support Vector Machine classifier based on spectral reflectance", Comput. Electron. Agric, 148, pp.250~258 (2018)
  2. Bannari A, "A review of vegetation indices", Remote Sensing Reviews, 13(1-2), pp.95~120 (1995)
  3. Boiarskii B, "Comparison of NDVI and NDRE indices to detect differences in vegetation and chlorophyll content", J. Mech. Contin. Math. Sci, 4, pp.20~29 (2019)
  4. Candiago S, "Evaluating multispectral images and vegetation indices for precision farming applications from UAV images", Remote Sens, 7(4), pp.4026~4047 (2015)
  5. Cartelat A, "Optically assessed contents of leaf polyphenolics and chlorophyll as indicators of nitrogen deficiency in wheat (Triticum aestivum L.)", Field Crops Res, 91(1), pp.35~49 (2005)
  6. Chaudhary A, "A hybrid ensemble for classification in multiclass datasets : An application to oilseed disease dataset", Comput. Electron. Agric, 124, pp.65~72 (2016)
  7. Chen L, "Diagnosing of rice nitrogen stress based on static scanning technology and image information extraction", J. Soil Sci. Plant Nutr, 14(2), pp.382~393 (2014)
  8. Chen L, "Identification of nitrogen, phosphorus, and potassium deficiencies in rice based on static scanning technology and hierarchical identification method", PloS One, 9(11), pp.e113200~ (2014)
  9. Cuevas CMC, "Assessment of Nutrient Status of Lowland Irrigated Rice (Oryza sativa L.) using Low Altitude Remote Sensing", Philippine J. of Agricultural and Biosystems Engineering, 17(1), pp.40~48 (2021)
  10. Dobermann A, "Potash and phosphate institute, international rice research institute", Rice:Nutrient Disorders & Nutrient Management, 72~ (2000)
  11. Fu Y, "An overview of crop nitrogen status assessment using hyperspectral remote sensing : Current status and perspectives", Eur. J. Agron, 124, pp.126241~ (2021)
  12. Gitelson AA, "Relationships between leaf chlorophyll content and spectral reflectance and algorithms for non-destructive chlorophyll assessment in higher plant leaves", J. Plant Physiol, 160(3), pp.271~282 (2003)
  13. Gitelson AA, "Use of a green channel in remote sensing of global vegetation from EOS-MODIS", Remote Sens. Environ, 58(3), pp.289~298 (1996)
  14. Gitelson AA, "Remote estimation of leaf area index and green leaf biomass in maize canopies", Geophys. Res. Lett, 30(5) (2003)
  15. Jordan CF, "Derivation of leaf-area index from quality of light on the forest floor", Ecology, 50(4), pp.663~666 (1969)
  16. Kaufman YJ, "Atmospherically resistant vegetation index(ARVI)for EOS-MODIS", IEEE Trans. Geosci. Remote Sens, 30(2), pp.261~270 (1992)
  17. Kim MH, "Response of grain protein content to nitrogen topdress rate at panicle initiation stage under different growth and plant nitrogen status of rice", Korean J. Crop Sci, 52(1), pp.104~111 (2007)
  18. Kurbanov RK, "Application of vegetation indexes to assess the condition of crops", Agric. Mach. Technol, 14(4), pp.4~11 (2020)
  19. Li S, "Combining color indices and textures of UAV-based digital imagery for rice LAI estimation", Remote Sens, 11(15), pp.1763~ (2019)
  20. Li W, "Integrated diagnosis and time-series sensitivity evaluation of nutrient deficiencies in medicinal plant (Ligusticum chuanxiong Hort.) based on UAV multispectral sensors", Front. Plant Sci, 13, pp.1092610~ (2023)
  21. Liu C, "A review of plant spectral reflectance response to water physiological changes", Chin. J. Plant Ecol, 40(1), pp.80~ (2016)
  22. Liu T, "Estimates of rice lodging using indices derived from UAV visible and thermal infrared images", Agric. For. Meteorol, 252, pp.144~154 (2018)
  23. Mahajan GR, "Monitoring nitrogen, phosphorus and sulphur in hybrid rice (Oryza sativa L.) using hyperspectral remote sensing", Precis. Agric, 18(5), pp.736~761 (2017)
  24. Matese A, "Intercomparison of UAV, aircraft and satellite remote sensing platforms for precision viticulture", Remote Sens, 7(3), pp.2971~2990 (2015)
  25. Onoyama H, "Estimation of nitrogen contents in rice plant at the panicle initiation stage using ground-based hyperspectral remote sensing", IFAC Proceedings, 43(26), pp.166~171 (2010)
  26. Pantazi XE, "Wheat yield prediction using machine learning and advanced sensing techniques", Comput. Electron. Agric, 121, pp.57~65 (2016)
  27. Qi J, "A modified soil adjusted vegetation index", Remote Sens. Environ, 48(2), pp.119~126 (1994)
  28. Rouse JW, "Monitoring vegetation systems in the Great Plains with ERTS", NASA Spec. Publ, 351(1), pp.309~ (1974)
  29. Analysis Standard of Agricultural Test and Research, "Rural Development Administration(RDA)", RDA, ~ (2003)
  30. Saberioon MM, "Assessment of rice leaf chlorophyll content using visible bands at different growth stages at both the leaf and canopy scale", Int. J. Appl. Earth Obs. Geoinf, 32, pp.35~45 (2014)
  31. Sharma M, "Ensemble averaging of transfer learning models for identification of nutritional deficiency in rice plant", Electronics, 11(1), pp.148~ (2022)
  32. Shrestha J, "Role of nutrients in rice (Oryza sativa L.): A review", Agrica, 9(1), pp.53~62 (2020)
  33. Sims DA, "Relationships between leaf pigment content and spectral reflectance across a wide range of species, leaf structures and developmental stages", Remote Sens Environ, 81(2), pp.337~354 (2002)
  34. Sripada RP, "Aerial color infrared photography for determining late-season nitrogen requirements in corn", Agron. J, 97(5), pp.1443~1451 (2005)
  35. Talukder MSH, "Nutrients deficiency diagnosis of rice crop by weighted average ensemble learning", Smart Agric. Technol, 4, pp.100155~ (2023)
  36. Tucker CJ, "Red and photographic infrared linear combinations for monitoring vegetation", Remote Sens. Environ, 8(2), pp.127~150 (1979)
  37. Vigneau N, "Potential of field hyperspectral imaging as a non destructive method to assess leaf nitrogen content in Wheat", Field Crops Res, 122(1), pp.25~31 (2011)
  38. Walsh OS, "Assessment of UAV based vegetation indices for nitrogen concentration estimation in spring wheat", Adv. Remote Sens, 7(2), pp.71~90 (2018)
  39. Wang YP, "Estimation of nitrogen status of paddy rice at vegetative phase using unmanned aerial vehicle based multispectral imagery", Precis Agric, 23(1), pp.1~17 (2022)
  40. Xue J, "Significant remote sensing vegetation indices : A review of developments and applications", J Sens (2017)
  41. Yin G, "A broadband green-red vegetation index for monitoring gross primary production phenology", J. Remote Sens (2022)
  42. Yoosefzadeh-Najafabadi M, "Application of machine learning algorithms in plant breeding : predicting yield from hyperspectral reflectance in soybean", Front. Plant Sci, 11, pp.624273~ (2021)
  43. Yu FH, "Combining the critical nitrogen concentration and machine learning algorithms to estimate nitrogen deficiency in rice from UAV hyperspectral data", J. Integr. Agric, 22(4), pp.1216~1229 (2023)
  44. Zhang JH, "Predicting nitrogen status of rice using multispectral data at canopy scale", Pedosphere, 16(1), pp.108~117 (2006)
  45. Zhang K, "Predicting Rice grain yield based on dynamic changes in vegetation indexes during early to mid-growth stages", Remote Sens, 11(4), pp.387~ (2019)
  46. Zhang X, "Monitoring vegetation phenology using MODIS", Remote Sens. Environ, 84(3), pp.471~475 (2003)
  47. Zhang Y, "Advances in hyperspectral remote sensing of vegetation traits and functions", Remote Sens. Environ, 252, pp.112121~ (2021)
  48. Zhou X, "Predicting grain yield in rice using multi-temporal vegetation indices from UAV-based multispectral and digital imagery", ISPRS J. Photogramm. Remote Sens, 130, pp.246~255 (2017)
  49. Zhu H, "UAV-based hyperspectral analysis and spectral indices constructing for quantitatively monitoring leaf nitrogen content of winter wheat", Appl. Opt, 57(27), pp.7722~7732 (2018)

관련 논문AI추천

무인비행체 영상 기반 연차 간 벼 생육 및 흰잎마름병 병해 추정김상민, 나상일, 박찬원, 소규호, 안호용, 이경도, 홍석영한국농공학회 · 한국농공학회논문집2020 · 제62권 제4호무인비행체 영상을 활용한 벼 수량 분포 추정나상일, 박찬원, 소규호, 안호용, 이경도, 장수용한국농공학회 · 한국농공학회논문집2019 · 제61권 제4호무인기 기반 다중분광 영상을 이용한 콩 경태 추정 머신러닝 회귀 모델강예성, 권호준, 박창혁, 유찬석, 제강인농업생명과학연구원 · 농업생명과학연구2025 · 제59권 제6호Estimation of Leaf Dry Mass and Nitrogen Content for Soybean using Multi-spectral Camera Mounted on Unmanned Aerial Vehicle강예성, 강정균, 김성헌, 류찬석, 홍영기, Tapash Kumar Sarkar농업생명과학연구원 · 농업생명과학연구2016 · 제50권 제6호드론 영상 기반 노지 콩 품종 분류를 위한 딥러닝 모델의 적용 및 평가박 재성, 김설민, 박진기, 송철민, 안치용한국농공학회 · 한국농공학회논문집2025 · 제67권 제6호무인비행장치를 이용한 작물 생육 라이브러리 구축김남용, 민관식, 이창복, 정인준한국지적학회 · 한국지적학회지2018 · 제34권 제2호

이 간행물의 다른 논문 (14건)

Development of a Specific SNP Marker to Identify the Medicinal Plants P. grandiflorum김윤희, 김주환pp. 1-6In vitro Micropropagation of Water Lettuce (Pistia stratiotes L.) via Direct Organogenesis임채오, 이영민, 이재혁, 강창호pp. 7-14낫또용 Penta Null 유전자형을 가진 소립 노란콩 계통 육종정권문, 리사랏, 정종일pp. 15-20치유농업 5년간(2018년~2022년) 동향 분석을 통한 프로그램 운영기준 모색문지원, 이상미, 홍인경pp. 21-32대학 캠퍼스 내 정원 인식에 대한 요인 구조화 및 이용 만족도와의 관계 분석김용현, 허무룡pp. 33-41
원문 보기