DOI QR코드

DOI QR Code

Spectral Band Selection for Detecting Fire Blight Disease in Pear Trees by Narrowband Hyperspectral Imagery

초분광 이미지를 이용한 배나무 화상병에 대한 최적 분광 밴드 선정

  • Kang, Ye-Seong (Department of Agro-System Engineering, GyeongSang National University (Institute of Agriculture & Life Science)) ;
  • Park, Jun-Woo (Department of Agro-System Engineering, GyeongSang National University (Institute of Agriculture & Life Science)) ;
  • Jang, Si-Hyeong (Department of Agro-System Engineering, GyeongSang National University (Institute of Agriculture & Life Science)) ;
  • Song, Hye-Young (Department of Agro-System Engineering, GyeongSang National University (Institute of Agriculture & Life Science)) ;
  • Kang, Kyung-Suk (Department of Agro-System Engineering, GyeongSang National University (Institute of Agriculture & Life Science)) ;
  • Ryu, Chan-Seok (Department of Agro-System Engineering, GyeongSang National University (Institute of Agriculture & Life Science)) ;
  • Kim, Seong-Heon (Granduate School of Bioresource and Bioenvironmental Science, Kyushu Universtiy) ;
  • Jun, Sae-Rom (Hortizen Co. Ltd.) ;
  • Kang, Tae-Hwan (Bio Industry Mechanical Engineering, Kongju National Universtiy) ;
  • Kim, Gul-Hwan (National Academy of Agricultural Science, Rural Development Administration)
  • 강예성 (경상대학교 농업생명과학대학 애그로시스템공학전공) ;
  • 박준우 (경상대학교 농업생명과학대학 애그로시스템공학전공) ;
  • 장시형 (경상대학교 농업생명과학대학 애그로시스템공학전공) ;
  • 송혜영 (경상대학교 농업생명과학대학 애그로시스템공학전공) ;
  • 강경석 (경상대학교 농업생명과학대학 애그로시스템공학전공) ;
  • 유찬석 (경상대학교 농업생명과학대학 애그로시스템공학전공) ;
  • 김성헌 (큐슈대학교) ;
  • 전새롬 (주식회사 호티젠) ;
  • 강태환 (공주대학교 농업생명과학대학 생물산업기계공학과) ;
  • 김국환 (농촌진흥청 농업공학부)
  • Received : 2021.03.02
  • Accepted : 2021.03.22
  • Published : 2021.03.30

Abstract

In this study, the possibility of discriminating Fire blight (FB) infection tested using the hyperspectral imagery. The reflectance of healthy and infected leaves and branches was acquired with 5 nm of full width at high maximum (FWHM) and then it was standardized to 10 nm, 25 nm, 50 nm, and 80 nm of FWHM. The standardized samples were divided into training and test sets at ratios of 7:3, 5:5 and 3:7 to find the optimal bands of FWHM by the decision tree analysis. Classification accuracy was evaluated using overall accuracy (OA) and kappa coefficient (KC). The hyperspectral reflectance of infected leaves and branches was significantly lower than those of healthy green, red-edge (RE) and near infrared (NIR) regions. The bands selected for the first node were generally 750 and 800 nm; these were used to identify the infection of leaves and branches, respectively. The accuracy of the classifier was higher in the 7:3 ratio. Four bands with 50 nm of FWHM (450, 650, 750, and 950 nm) might be reasonable because the difference in the recalculated accuracy between 8 bands with 10 nm of FWHM (440, 580, 640, 660, 680, 710, 730, and 740 nm) and 4 bands was only 1.8% for OA and 4.1% for KC, respectively. Finally, adding two bands (550 nm and 800 nm with 25 nm of FWHM) in four bands with 50 nm of FWHM have been proposed to improve the usability of multispectral image sensors with performing various roles in agriculture as well as detecting FB with other combinations of spectral bands.

화상병이란 erwinia amylovora라는 강한 전염성을 보유하고 있어 감염 시 1년 내에 과수를 고사시키며 그 중심으로 반경 500m이내에 과수 재배를 불가능하게 만드는 세균성 바이러스이다. 이 화상병은 과수의 잎과 가지를 진한 갈색 또는 검은색으로 변색시키기 때문에 분광학적으로 검출이 가능하다고 판단되며 이는 다중분광센서를 탑재한 무인기를 이용하는 것이 효율적이다. 그러나 다중분광센서는 적은 중심 파장과 함께 넓은 반치전폭(FWHM)을 가지고 있어 화상병에 가장 민감하게 반응하는 파장 대역을 파악하기 어렵다. 그렇기 때문에, 본 논문에서는 화상병에 감염된 잎과 가지와 비감염된 잎과 가지의 초분광 이미지를 5 nm FWHM으로 취득한 후 각각 10 nm, 25 nm, 50 nm와 80 nm FWHM로 평준화한 후 샘플을 7:3, 5:5와 3:7의 비율로 훈련데이터와 검증데이터로 나누어 의사결정트리 기법으로 최적의 파장을 선정하고 overall accuracy (OA)와 kappa coefficient (KC)를 이용한 분류 정확도 평가를 통해 배나무 화상병 검출가능성을 확인하였다. 화상병에 감염 및 비감염된 잎과 가지의 초분광 반사율을 비교한 결과, green, red edge 및 NIR 영역에서 차이가 두드러지게 나타났으며 첫 번째 분류 노드로 선택된 파장 영역은 대체로 750 nm와 800 nm였다. 잎과 가지 영역의 영상데이터를 의사결정트리 기법을 이용하여 분류정확도를 종합적으로 비교한 결과, 50nm FWHM 인 4개 대역(450, 650, 750, 950nm)은 10nm FWHM인 8개 대역(440, 580, 660, 680, 680, 710, 730, 740nm)의 분류 정확도 차이가 OA에서 1.8%와 KC에서 4.1%로 나타나 더 낮은 비용의 밴드패스필터인 50nm FWHM을 이용하는 것이 더 유리하다고 판단된다. 또한 기존의 50nm FWHM 파장대역들에 25nm FWHM파장대역들(550, 800nm)을 추가하는 것을 통해 화상병 검출뿐만 아니라 농업에서 다양한 역할을 수행할 수 있는 다중분광센서를 개발할 수 있다고 판단된다.

Keywords

References

  1. Acimovic, S. G., Q. Zeng, G. C. McGhee, G. W. Sundin, and J. C. Wise, 2015: Control of fire blight (Erwinia amylovora) on apple trees with trunk-injected plant resistance inducers and antibiotics and assessment of induction of pathogenesisrelated protein genes. Frontiers in plant science 6, 16. https://doi.org/10.3389/fpls.2015.00016
  2. Apan, A., A. Held, S. Phinn, and J. Markley, 2004: Detecting sugarcane 'orange rust'disease using EO-1 Hyperion hyperspectral imagery. International journal of remote sensing 25(2), 489-498. https://doi.org/10.1080/01431160310001618031
  3. Bagheri, N., H. Mohamadi-Monavar, A. Azizi, and A. Ghasemi, 2018: Detection of Fire Blight disease in pear trees by hyperspectral data. European Journal of Remote Sensing 51(1), 1-10. https://doi.org/10.1080/22797254.2017.1391054
  4. Bahadou, S. A., A. Ouijja, A. Karfach, A. Tahiri, and R. Lahlali, 2018: New potential bacterial antagonists for the biocontrol of fire blight disease (Erwinia amylovora) in Morocco. Microbial pathogenesis 117, 7-15. https://doi.org/10.1016/j.micpath.2018.02.011
  5. Berni, J. A. J., P. J. Zarco-Tejada, G. Sepulcre-Canto, E. Fereres, and F. Villalobos, 2009: Mapping canopy conductance and CWSI in olive orchards using high resolution thermal remote sensing imagery. Remote Sensing of Environment 113(11), 2380-2388. https://doi.org/10.1016/j.rse.2009.06.018
  6. Breiman, L., J. Friedman, R. Olshen, and C. Stone, 1984: Classification and regression trees-crc press. Boca Raton, Florida.
  7. Buchanan, G. E., and M. P. Starr, 1980: Phytotoxic material from associations betweenErwinia amylovora and pear tissue culture: Possible role in necrotic symptomatology of fireblight disease. Current Microbiology 4(2), 63-68. https://doi.org/10.1007/BF02602894
  8. Broggini, G. A., B. Duffy, E. Holliger, H. J. Scharer, C. Gessler, and A. Patocchi, 2005: Detection of the fire blight biocontrol agent Bacillus subtilis BD170 (Biopro®) in a Swiss apple orchard. European journal of plant pathology 111(2), 93-100. https://doi.org/10.1007/s10658-004-1423-x
  9. Calderon, R., J. A. Navas-Cortes, C. Lucena, and P. J. Zarco-Tejada, 2013: High-resolution airborne hyperspectral and thermal imagery for early detection of Verticillium wilt of olive using fluorescence, temperature and narrow-band spectral indices. Remote Sensing of Environment 139, 231-245. https://doi.org/10.1016/j.rse.2013.07.031
  10. Chan, J. C. W., and D. Paelinckx, 2008: Evaluation of random forest and adaboost tree-based ensemble classification and spectral band selection for ecotope mapping using airborne hyperspectral imagery. Remote Sensing of Environment 112, 2999-3011. https://doi.org/10.1016/j.rse.2008.02.011
  11. Deng, X., Q. Liu, Y. Deng, and S. Mahadevan, 2016: An improved method to construct basic probability assignment based on the confusion matrix for classification problem. Information Sciences 340, 250-261. https://doi.org/10.1016/j.ins.2016.01.033
  12. Du, H., H. Qi, X. Wang, R. Ramanath, and W. E. Snyder, 2003: Band selection using independent component analysis for hyperspectral image processing. In: 32nd Applied Imagery Pattern Recognition Workshop, 2003. Proceedings. IEEE, 93-98.
  13. Foody, G. M., 2002: Status of land cover classification accuracy assessment. Remote sensing of environment 80(1), 185-201. https://doi.org/10.1016/S0034-4257(01)00295-4
  14. Friedl, M. A., D. Sulla-Menashe, B. Tan, A. Schneider, N. Ramankutty, A. Sibley, and X. Huang, 2010: MODIS Collection 5 global land cover: Algorithm refinements and characterization of new datasets. Remote sensing of Environment 114, 168-182. https://doi.org/10.1016/j.rse.2009.08.016
  15. Goel, P. K., S. O. Prasher, R. M. Patel, J. A. Landry, R. B. Bonnell, and A. A. Viau, 2003: Classification of hyperspectral data by decision trees and artificial neural networks to identify weed stress and nitrogen status of corn. Computers and Electronics in Agriculture 39(2), 67-93. https://doi.org/10.1016/S0168-1699(03)00020-6
  16. Gusberti, M., U. Klemm, M. S. Meier, M. Maurhofer, and I. Hunger-Glaser, 2015: Fire blight control: the struggle goes on. A comparison of different fire blight control methods in Switzerland with respect to biosafety, efficacy and durability. International journal of environmental research and public health 12(9), 11422-11447. https://doi.org/10.3390/ijerph120911422
  17. Hasler, T., H. J. Schaerer, E. Holliger, J. Vogelsanger, A. Vignutelli, and B. Schoch, 2001: Fire blight situation in Switzerland. IX International Workshop on Fire Blight 590, 73-79.
  18. Hinze, M., L. Kohl, S. Kunz, S. WeiBhaupt, M. Ernst, A. Schmid, and R. T. Voegele, 2016: Real-time PCR detection of Erwinia amylovora on blossoms correlates with subsequent fire blight incidence. Plant Pathology 65(3), 462-469. https://doi.org/10.1111/ppa.12429
  19. Holasek, R., K. Nakanishi, L. Ziph-Schatzberg, J. Santman, P. Woodman, R. Zacaroli, and R. Wiggins, 2017: The selectable hyperspectral airborne remote sensing kit (SHARK) as an enabler for precision agriculture. Hyperspectral Imaging Sensors: Innovative Applications and Sensor Standards 10213, 1021304
  20. Jaillais, B., P. Roumet, L. Pinson-Gadais, and D. Bertrand, 2015: Detection of Fusarium head blight contamination in wheat kernels by multivariate imaging. Food Control 54, 250-258. https://doi.org/10.1016/j.foodcont.2015.01.048
  21. Jarolmasjed, S., S. Kostick, Y. Si, J. Quiros, A. Marzougui, K. Evans, and S. Sankaran, 2019: High-Throughput Phenotyping of Fire Blight Disease Symptoms Using Sensing Techniques in Apple. Frontiers in plant science 10, 576. https://doi.org/10.3389/fpls.2019.00576
  22. Jock, S., V. Donat, M. M. López, C. Bazzi, and K. Geider, 2002: Following spread of fire blight in Western, Central and Southern Europe by molecular differentiation of Erwinia amylovora strains with PFGE analysis. Environmental Microbiology 4(2), 106-114. https://doi.org/10.1046/j.1462-2920.2002.00277.x
  23. Kang, Y. S., C. S. Ryu, S. R. Jun, S. H. Jang, J. W. Park, H. Y. Song, T. K. Sarkar, S. H. Kim, and W. S. Lee, 2018: Distinguishing between closely related species of Allium and of Brassicaceae by narrowband hyperspectral imagery. Biosystems Engineering 176, 103-113. https://doi.org/10.1016/j.biosystemseng.2018.10.003
  24. Kim, W. S., M. Hildebrand, S. Jock, and K. Geider, 2001: Molecular comparison of pathogenic bacteria from pear trees in Japan and the fire blight pathogen Erwinia amylovora. Microbiology 147 (11), 2951-2959. https://doi.org/10.1099/00221287-147-11-2951
  25. Kimura, R., S. Okada, H. Miura, and M. Kamichika, 2004: Relationships among the leaf area index, moisture availability, and spectral reflectance in an upland rice field. Agricultural Water Management 69(2), 83-100. https://doi.org/10.1016/j.agwat.2004.04.009
  26. Kumar, A., W. S. Lee, R. J. Ehsani, L. G. Albrigo, C. Yang, and R. L. Mangan, 2012: Citrus greening disease detection using aerial hyperspectral and multispectral imaging techniques. Journal of Applied Remote Sensing, 6(1), 063542. https://doi.org/10.1117/1.JRS.6.063542
  27. Landis, J. R., and G. G. Koch, 1977: The measurement of observer agreement for categorical data. Biometrics, 159-174.
  28. Leucker, M., A. K. Mahlein, U. Steiner, and E. C. Oerke, 2015: Improvement of lesion phenotyping in Cercospora beticola-sugar beet interaction by hyperspectral imaging. Phytopathology 106(2), 177-184. https://doi.org/10.1094/PHYTO-04-15-0100-R
  29. Li, S., Z. Zheng, Y. Wang, C. Chang, and Y. Yu, 2016: A new hyperspectral band selection and classification framework based on combining multiple classifiers. Pattern Recognition Letters 83, 152-159. https://doi.org/10.1016/j.patrec.2016.05.013
  30. Mahlein, A. K., 2016: Plant disease detection by imaging sensors-parallels and specific demands for precision agriculture and plant phenotyping. Plant disease 100(2), 241-251. https://doi.org/10.1094/PDIS-03-15-0340-FE
  31. Mahlein, A. K., U. Steiner, C. Hillnhutter, H. W. Dehne, and E. C. Oerke, 2012: Hyperspectral imaging for small-scale analysis of symptoms caused by different sugar beet diseases. Plant methods 8(1), 3. https://doi.org/10.1186/1746-4811-8-3
  32. Marlin, B., K. Swersky, B. Chen, and N. Freitas, 2010: Inductive principles for restricted Boltzmann machine learning. In Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, 509-516.
  33. Medjahed, S. A., T. A. Saadi, A. Benyettou, and M. Ouali, 2016: Gray wolf optimizer for hyperspectral band selection. Applied Soft Computing 40, 178-186. https://doi.org/10.1016/j.asoc.2015.09.045
  34. Norelli, J. L., A. L. Jones, and H. S. Aldwinckle, 2003: Fire blight management in the twenty-first century: using new technologies that enhance host resistance in apple. Plant Disease 2003 87(7), 756-765. https://doi.org/10.1094/PDIS.2003.87.7.756
  35. Pu, Y. Y., and D. W. Sun, 2016: Prediction of moisture content uniformity of microwave-vacuum dried mangoes as affected by different shapes using NIR hyperspectral imaging. Innovative Food Science & Emerging Technologies 33, 348-356. https://doi.org/10.1016/j.ifset.2015.11.003
  36. Qiu, R., C. Yang, A. Moghimi, M. Zhang, and B. Steffenson, 2019: Detection of Fusarium Head Blight in Wheat Using a Deep Neural Network and Color Imaging. Preprints.
  37. Ray, S. S., N. Jain, R. K. Arora, S. Chavan, and S. Panigrahy, 2011: Utility of hyperspectral data for potato late blight disease detection. Journal of the Indian Society of Remote Sensing 39(2), 161-169. https://doi.org/10.1007/s12524-011-0094-2
  38. Ray, S. S., J. P. Singh, and S. Panigrahy, 2010: Use of hyperspectral remote sensing data for crop stress detection: ground-based studies. International Archives of Photogrammetry, Remote Sensing and Spatial Information Science 38(Part 8).
  39. Ryu, C., M. Suguri, and M. Umeda, 2011: Multivariate analysis of nitrogen content for rice at the heading stage using reflectance of airborne hyperspectral remote sensing. Field Crops Research 122(3), 214-224. https://doi.org/10.1016/j.fcr.2011.03.013
  40. Sarhrouni, E., A. Hammouch, and D. Aboutajdine, 2012: Dimensionality reduction and classification feature using mutual information applied to hyperspectral images: a filter strategy based algorithm. arXiv preprint arXiv :1210.0052.
  41. Schoofs, H., S. Delalieux, T. Deckers, and D. Bylemans, 2020: Fire Blight Monitoring in Pear Orchards by Unmanned Airborne Vehicles (UAV) Systems Carrying Spectral Sensors. Agronomy 10(5), 615. https://doi.org/10.3390/agronomy10050615
  42. Smail, A. B., O. U. I. J. J. A. Abderrahman, and T. A. H. I. R. I. Abdessalem, 2016: Evaluation of biological control agent Pantoea agglomerans P10c against fire blight in Morocco. African Journal of Agricultural Research 11(18), 1661-1667. https://doi.org/10.5897/AJAR2016.10986
  43. Smits, P. C., S. G. Dellepiane, and R. A. Schowengerdt, 1999: Quality assessment of image classification algorithms for land-cover mapping: A review and a proposal for a cost-based approach. International journal of remote sensing 20(8), 1461-1486. https://doi.org/10.1080/014311699212560
  44. Suarez, L. A., A. Apan, and J. Werth, 2016: Hyperspectral sensing to detect the impact of herbicide drift on cotton growth and yield. ISPRS Journal of Photogrammetry and Remote Sensing 120, 65-76. https://doi.org/10.1016/j.isprsjprs.2016.08.004
  45. Sun, W., L. Zhang, B. Du, W. Li, and Y. M. Lai, 2015: Band selection using improved sparse subspace clustering for hyperspectral imagery classification. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 8, 2784-2797. https://doi.org/10.1109/JSTARS.2015.2417156
  46. Thomson, S. V., S. C. Gouk, and J. P. Paulin, 1998: Efficacy of BION®(Actigard®) to control fire blight in pear and apple orchards in USA, New Zealand and France. In VIII International Workshop on Fire Blight 489, 589-596.
  47. Tomassen, F. H. M., A. De Koeijer, M. C. M. Mourits, A. Dekker, A. Bouma, and R. B. M. Huirne, 2002: A decision-tree to optimise control measures during the early stage of a foot-andmouth disease epidemic. Preventive Veterinary Medicine 54(4), 301-324. https://doi.org/10.1016/S0167-5877(02)00053-3
  48. Tooke, T. R., N. C. Coops, N. R. Goodwin, and J. A. Voogt, 2009: Extracting urban vegetation characteristics using spectral mixture analysis and decision tree classifications. Remote Sensing of Environment 113(2), 398-407. https://doi.org/10.1016/j.rse.2008.10.005
  49. Trilla, G. G., P. Pratolongo, P. Kandus, M. E. Beget, C. Di Bella, and J. Marcovecchio, 2016: Relationship between biophysical parameters and synthetic indices derived from hyperspectral field data in a salt marsh from Buenos Aires Province, Argentina. Wetlands 36(1), 185-194. https://doi.org/10.1007/s13157-015-0715-6
  50. Van Der Zwet, Tom, and Harry Louis Keil, 1979: Fire blight. US Government Printing Office, 21-26.
  51. Yue, J., G. Yang, C. Li, Z. Li, Y. Wang, H. Feng, and B. Xu, 2017: Estimation of winter wheat above-ground biomass using unmanned aerial vehicle-based snapshot hyperspectral sensor and crop height improved models. Remote Sensing 9(7), 708. https://doi.org/10.3390/rs9070708
  52. Zarco-Tejada, P. J., V. Gonzalez-Dugo, and J. A. Berni, 2012: Fluorescence, temperature and narrow-band indices acquired from a UAV platform for water stress detection using a micro-hyperspectral imager and a thermal camera. Remote sensing of environment 117, 322-337. https://doi.org/10.1016/j.rse.2011.10.007
  53. Zhang, M., Z. Qin, X. Liu, and S. L. Ustin, 2003: Detection of stress in tomatoes induced by late blight disease in California, USA, using hyperspectral remote sensing. International Journal of Applied Earth Observation and Geoinformation 4(4), 295-310. https://doi.org/10.1016/S0303-2434(03)00008-4
  54. Zhou, R., S. I. Kaneko, F. Tanaka, M. Kayamori, and M. Shimizu, 2015: Image-based field monitoring of Cercospora leaf spot in sugar beet by robust template matching and pattern recognition. Computers and Electronics in Agriculture 116, 65-79. https://doi.org/10.1016/j.compag.2015.05.020