Romero Galvan, F. E. | bioRxiv Science

bioRxiv2022

Towards Cloud-Native, Machine Learning Based Detection of Crop Disease with Imaging Spectroscopy

Developing actionable early detection and warning systems for agricultural stakeholders is crucial to reduce the annual $200B USD losses and environmental impacts associated with crop diseases. Agricultural stakeholders primarily rely on labor-intensive, expensive scouting and molecular testing to detect disease. Spectroscopic imagery (SI) can improve plant disease management by offering decision-makers accurate risk maps derived from Machine Learning (ML) models. However, training and deploying ML requires significant computation and storage capabilities. This challenge will become even greater as global scale data from the forthcoming Surface Biology & Geology (SBG) satellite becomes available. This work presents a cloud-hosted architecture to streamline plant disease detection with SI from NASAs AVIRIS-NG platform, using grapevine leafroll associated virus complex 3 (GLRaV-3) as a model system. Here, we showcase a pipeline for processing SI to produce plant disease detection models and demonstrate that the underlying principles of a cloud-based disease detection system easily accommodate model improvements and shifting data modalities. Our goal is to make the insights derived from SI available to agricultural stakeholders via a platform designed with their needs and values in mind. The key outcome of this work is an innovative, responsive system foundation that can empower agricultural stakeholders to make data-driven plant disease management decisions, while serving as a framework for others pursuing use-inspired application development for agriculture to follow that ensures social impact and reproducibility while preserving stakeholder privacy. Key PointsO_LICloud-based plant disease detection system, easily accommodates newly developed and/or improved models, as well as diverse data modalities. C_LIO_LIEmpower agricultural stakeholders to use hyperspectral data for decision support while preserving stakeholder data privacy. C_LIO_LIOutline framework for researchers interested in designing geospatial/remote sensing applications for agricultural stakeholders to follow. C_LI

ecology↗

bioRxiv2022

Scalable early detection of grapevine virus infection with airborne imaging spectroscopy

Viral diseases, including Grapevine Leafroll-associated Virus Complex 3 (GLRaV-3), cause $3 billion in damages and losses to the United States wine and grape industry annually. GLRaV-3 has a well-studied, year-long latent period in which vines are infectious but do not yet display visible symptoms, making it an ideal model pathosystem to evaluate the scalability of symptomatic and asymptomatic imaging spectroscopy-based disease detection. Plant disease causes physiological and chemical changes to occur locally and systemically throughout a plant, which imaging spectroscopy can detect both directly and indirectly. Reliable and scalable disease detection during the latent period would greatly reduce management costs, as current detection methods are entirely ground-based, labor-intensive, and expensive. Here, we use data collected in September 2020 by the NASA Airborne Visible/Infrared Imaging Spectrometer Next Generation (AVIRIS-NG) to detect GLRaV-3 in Cabernet Sauvignon grapevines in Lodi, CA. During September 2020 and 2021, industry collaborators scouted 317 acres of Vitis vinifera winegrapes for visible disease symptoms, and collected a subset for confirmation molecular testing at a commercial facility. Grapevines identified as visibly diseased in 2021 were assumed to have been latently infected (asymptomatic) during the September 2020 AVIRIS-NG data collection. We combined random forest with synthetic minority oversampling technique (SMOTE) to train multiple spectral models able to distinguish between non-infected (NI) and GLRaV-3-infected grapevines. We observed clear spectral differences that allowed for differentiation between NI and GLRaV-3 infected vines both pre- and post-symptomatically at 1m through 5m resolution. Our two best performing models had 87% accuracy (0.73 Kappa) distinguishing between NI and asymptomatic (aSy), and 85% accuracy (0.71 Kappa) distinguishing between NI and (aSy + symptomatic [Sy]) respectively. We hypothesize these spectral differences are linked to changes in overall plant physiology induced by disease, as visible foliar symptoms were restricted to the lower canopy. HighlightsO_LIAirborne imaging spectroscopy allows for scalable early-detection models of grapevine leafroll-associated virus complex 3 (GLRaV-3). C_LIO_LIRandom Forest based models trained with scouting ground data and imaging spectroscopy are accurate up to 5 meter but perform best at 3 meter spatial resolution. C_LIO_LIGLRaV-3 detection via imaging spectroscopy will not replace existing field scouting strategies or molecular testing but supplement by allowing for more strategic resource deployment to improve the overall financial, environmental, and societal sustainability of winegrape production. C_LI

pathology↗