Korean Journal of Environmental Biology (환경생물)

Korean Society of Environmental Biology (한국환경생물학회)
권호 표지
DOI QR코드

DOI QR Code

Classification of Convolvulaceae plants using Vis-NIR spectroscopy and machine learning

근적외선 분광법과 머신러닝을 이용한 메꽃과(Convolvulaceae) 식물의 분류

  • Yong-Ho Lee (Plant Life & Environmental Science, Hankyong National University) ;
  • Soo-In Sohn (Biosafety Division, National Institute of Agricultural Science, RDA) ;
  • Sun-Hee Hong (Plant Life & Environmental Science, Hankyong National University) ;
  • Chang-Seok Kim (Highland Agriculture Research Institute, National Institute of Crop Science, RDA) ;
  • Chae-Sun Na (Wild Plant seeds Research Division, Baekdudaegan National Arboretum) ;
  • In-Soon Kim (Institute for Future Environmental Ecology Co., Ltd) ;
  • Min-Sang Jang (Institute for Future Environmental Ecology Co., Ltd) ;
  • Young-Ju Oh (Institute for Future Environmental Ecology Co., Ltd)
  • 이용호 (국립한경대학교 식물생태화학연구소) ;
  • 손수인 (국립농업과학원 생물안전성과) ;
  • 홍선희 (국립한경대학교 식물생태화학연구소) ;
  • 김창석 (국립식량과학원 고령지농업연구소) ;
  • 나채선 (국립백두대간수목원 야생식물종자연구실) ;
  • 김인순 ((주)미래환경생태연구소) ;
  • 장민상 ((주)미래환경생태연구소) ;
  • 오영주 ((주)미래환경생태연구소)
  • Received : 2021.12.06
  • Accepted : 2021.12.21
  • Published : 2021.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

Using visible-near infrared(Vis-NIR) spectra combined with machine learning methods, the feasibility of quick and non-destructive classification of Convolvulaceae species was studied. The main aim of this study is to classify six Convolvulaceae species in the field in different geographical regions of South Korea using a handheld spectrometer. Spectra were taken at 1.5 nm intervals from the adaxial side of the leaves in the Vis-NIR spectral region between 400 and 1,075 nm. The obtained spectra were preprocessed with three different preprocessing methods to find the best preprocessing approach with the highest classification accuracy. Preprocessed spectra of the six Convolvulaceae sp. were provided as input for the machine learning analysis. After cross-validation, the classification accuracy of various combinations of preprocessing and modeling ranged between 43.4% and 98.6%. The combination of Savitzky-Golay and Support vector machine methods showed the highest classification accuracy of 98.6% for the discrimination of Convolvulaceae sp. The growth stage of the plants, different measuring locations, and the scanning position of leaves on the plant were some of the crucial factors that affected the outcomes in this investigation. We conclude that Vis-NIR spectroscopy, coupled with suitable preprocessing and machine learning approaches, can be used in the field to effectively discriminate Convolvulaceae sp. for effective weed monitoring and management.

본 연구는 메꽃과 6종의 식물에 대해 신속하고 비파괴적으로 분류하기 위해 근적외선(Vis-NIR) 스펙트럼을 이용하였고 데이터의 전처리와 머신러닝 기술을 적용하였다. 전국적으로 분포하는 메꽃과 6종에 대해 야외에서 휴대용 분광기를 이용하여 판별하였다. 식물의 잎의 표면에서 400~1,075 nm의 근적외선 스펙트럼(1.5 nm)을 수집하였다. 수집된 스펙트럼 데이터는 3가지의 전처리와 raw데이터를 이용하였고 4종류의 머신러닝 모델을 적용하여 높은 판별 정확도를 확인하였다. 전처리와 머신러닝 모델의 조합을 통해 분석된 판별의 정확도는 43~99%의 범위로 분석되었고, standard normal variate 전처리와 support vector machine 머신러닝 모델의 조합에서 판별 정확도가 98.6%로 가장 높게 나타났다. 본 연구에서 수집된 스펙트럼은 식물의 성장단계, 다양한 측정 지역 및 잎에서의 측정 위치 등과 같은 요인과 더불어 데이터 분석을 위한 조건으로 최적의 전처리와 머신러닝 기술을 적용한다면 메꽃과 식물의 야외에서의 정확한 분류가 가능하고 이들 식물의 효과적인 관리와 모니터링에 활용할 수 있을 것으로 판단되었다.

Keywords

Acknowledgement

본 연구는 농촌진흥청 공동연구사업(과제번호: PJ013855032021)의 지원에 의해 수행되었음.

References

  1. Bergo MC, TC Pastore, VT Coradin, AC Wiedenhoeft and JW Braga. 2016. NIRS identification of Swietenia macrophylla is robust across specimens from 27 countries. IAWA J. 37:420-430. https://doi.org/10.1163/22941932-20160144
  2. Buitrago MF, TA Groen, CA Hecker and AK Skidmore. 2017. Identifying leaf traits that signal that signal stress in TIR spectra. ISPRS-J. Photogramm. Remote Sens. 125:132-145. https://doi.org/10.1016/j.isprsjprs.2017.01.014
  3. Delwiche SR and JB Reeves. 2004. The effect of spectral pre-treatments on the partial least squares modelling of agricultural products. J. Near Infrared Spectrosc. 12:177-182. https://doi.org/10.1255/jnirs.424
  4. Durgante FM, N Higuchi, A Almeida and A Vicentini. 2013. Species spectral signature: Discriminating closely related plant species in the Amazon with Near-Infrared Leaf-Spectroscopy. For. Ecol. Manage. 209:240-248.
  5. Fernandez-Cabanas VM, A Garrido-Varo, D Perez-Marin and P Dardenne. 2006. Evaluation of pretreatment strategies for near-infrared spectroscopy calibration development of unground and ground compound feeding stuffs. Appl. Spectrosc. 60:17-23. https://doi.org/10.1366/000370206775382839
  6. Gaye B, D Zhang and A Wulamu. 2021. Improvement of Support Vector Machine Algorithm in Big Data Background. Math. Probl. Eng. 2021:ID5594899.
  7. Hadlich HL, FM Durgante, J Dos Santos, N Higuchi, JQ Chambers and A Vicentini. 2018. Recognizing Amazonian tree species in the field using bark tissues spectra. For. Ecol. Manag. 427:296-304. https://doi.org/10.1016/j.foreco.2018.06.002
  8. Harrison D, B Rivard and A Sanchez-Azofeifa. 2018. Classification of tree species based on longwave hyperspectral data from leaves, a case study for a tropical dry forest. Int. J. Appl. Earth Obs. Geoinf. 66:93-105.
  9. Jayanthi MG and DR Shashikumar. 2020. Automatic tomato plant leaf disease classification using multi-kernel support vector machine. Int. J. Eng. Adv. Technol. 9:560-565.
  10. Judd WS, CS Campbell, EA Kellog, PF Stevens and MJ Dongoghus. 2008. Plant Systematics: A Phylogenetic Approach, 3rd edition. Sinauer Associates. Sunderland, MA. p. 620.
  11. Lang C, FRC Costa, JLC Camargo, FM Durgante and A Vicentini. 2015. Near infrared spectroscopy facilitates rapid identification of both young and mature Amazonian tree species. PLoS One 10:e0134521.
  12. Luz BR and JK Crowley. 2007. Spectral reflectance and emissivity features of broad leaves plants: Prospects for remote sensing in the thermal infrared (8.0-14.0 μm). Remote Sens. Environ. 109:393-405. https://doi.org/10.1016/j.rse.2007.01.008
  13. Meerdink SK, DA Roberts, JY King, KL Roth, PE Dennison, CH Amaral and SJ Hook. 2016. Linking seasonal foliar traits to VSWIR-TIR spectroscopy across California ecosystems. Remote Sens. Environ. 186:322-338. https://doi.org/10.1016/j.rse.2016.08.003
  14. Park SH. 2009. New Illustrations and Photographs of Naturalized Plants of Korea. Ilchokak Inc. Seoul.
  15. Patil SP and RS Zambre. 2014. Classification of cotton leaf spot disease using support vector machine. J. Eng. Res. Appl. 4:92-97.
  16. Perumal P. 2021. Guava leaf disease classification using support vector machine. Turk. J. Comput. Math. Edu. 12:1177-1183.
  17. Porker K, M Zerner and D Cozzolino. 2017. Classification and authentication of barley (Hordeum vulgare) malt varieties: Combining attenuated total reflectance mid-infrared spectroscopy with chemometrics. Food Anal. Meth. 10:675-682. https://doi.org/10.1007/s12161-016-0627-y
  18. Soares-Filho BS, AS Oliveira, RG Rajao, U Oliveira, LRS Santos and AC Assuncao. 2017. Economic Valuation of Changes in the Amazon Forest Area: Economic Losses by Fires to Sustainable Timber Production. Center for Remote Sensing. Belo Horizonte, Brazil.
  19. Ullah AK, A Skidmore, TA Ramoelo, M Groen and A Naeem. 2014. Retrieval of leaf water content spanning the visible to thermal infrared spectra. ISPRS-J. Photogramm. Remote Sens. 93:56-64. https://doi.org/10.1016/j.isprsjprs.2014.04.005