Skip to main content
ARS Home » Southeast Area » Tifton, Georgia » Crop Genetics and Breeding Research » Research » Publications at this Location » Publication #347903
Title: High throughput phenotyping of tomato spotted wilt disease in peanuts using unmanned aerial systems and multispectral imaging

Author
item PATRICK, AARON - University Of Georgia
item PELHAM, SARA - University Of Georgia
item CULBREATH, A - University Of Georgia
item Holbrook, Carl - Corley
item DE GODOY, IGNACIO JOSE - Agronomical Institute Of Campinas (IAC)
item LI, CHANGYING - University Of Georgia

Submitted to: IEEE Instrumentation & Measurement Magazine
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/30/2017
Publication Date: 12/15/2017
Citation: Patrick, A., Pelham, S., Culbreath, A.K., Holbrook Jr, C.C., De Godoy, I., Li, C. 2017. High throughput phenotyping of tomato spotted wilt disease in peanuts using unmanned aerial systems and multispectral imaging. IEEE Instrumentation & Measurement Magazine. p. 4-12.

Interpretive Summary: It may be possible to more rapidly assess the health of plants by measuring the amount of reflected visible and near infrared light. In this study, multispectral images were acquired by quadcopter for detecting tomato spotted wilt virus amongst twenty genotypes of peanuts. The plants were also visually assessed to acquire ground truth ratings of disease progression. Multispectral images were processed into several vegetation indexes using mathematical formulas. Ultimately the best vegetation indices for disease detection were determined and correlated with manual ratings and yield. The relative resistance of each genotype was then compared. We concluded that the image based disease rating can be used to rapidly, and accurately separate resistant from susceptible peanut genotypes. This should be a useful tool for breeding and genetic research on diseases resistance in peanut.

Technical Abstract: The amount of visible and near infrared light reflected by plants varies depending on their health. In this study, multispectral images were acquired by quadcopter for detecting tomato spot wilt virus amongst twenty genetic varieties of peanuts. The plants were visually assessed to acquire ground truth ratings of disease progression. Multispectral images were processed into several vegetation indexes. Then features from the vegetation index images and manually acquire data were correlated to develop a model for accessing the percentage of plots diseased. Ultimately the best vegetation indices for disease detection were determined and correlated with manual ratings and yield. The relative resistance of each genotype was then compared; revealing tha the image based disease rating could separate resistant from susceptible cultivars.