Skip to main content
ARS Home » Midwest Area » Wooster, Ohio » Application Technology Research » Research » Publications at this Location » Publication #422260
Research Project: Sustainable Production and Pest Management Practices for Nursery, Greenhouse, and Protected Culture Crops

Location: Application Technology Research

Title: Early detection of tomato bacterial spot in controlled environment using machine learning processed spectral information

Author
item SI, GAOSHOUTONG - The Ohio State University
item LING, PETER - The Ohio State University
item Testen, Anna
item Zhu, Heping

Submitted to: HortScience
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/8/2025
Publication Date: 12/4/2025
Citation: Si, G., Ling, P., Testen, A.L., Zhu, H. 2025. Early detection of tomato bacterial spot in controlled environment using machine learning processed spectral information. HortScience. 61(1):14-22. https://doi.org/10.21273/HORTSCI18892-25.
DOI: https://doi.org/10.21273/HORTSCI18892-25

Interpretive Summary: When plant diseases are detected early, management approaches are more effective, potentially reducing pesticide applications. Traditional plant disease diagnostics requires visually inspecting plants or analyzing samples in a laboratory, which is time consuming. By using sensors to detect light beyond the visual spectrum (hyperspectral), scientists can detect plant diseases before visual symptoms develop. In this study, early detection of tomato bacterial spot was conducted using hyperspectral imaging data which was used to train an algorithm to detect the disease. This approach was successful 92% of times and can potentially be used to detect bacterial spot in greenhouse tomato seedling production.

Technical Abstract: Early detection of tomato bacterial spot is critical for effective disease management in greenhouse environments. This study developed an automated hyperspectral data collection system to capture spectral information from inoculated tomatoes in a walk-in growth chamber, enabling non-invasive monitoring of disease progression. A plant physiological-informed machine learning model, incorporating a genetic algorithm for feature selection, was employed to optimize the classification of healthy and diseased samples. To reduce multicollinearity and enhancing the machine learning model performance, three key vegetation indices (VIs) were selected from a pool of 18 candidate VIs using a genetic algorithm. The model achieved a 92% success rate using the three VIs that are associated with chlorophyll content and photosynthetic efficiency. The result of this study suggests a promising methodology for plant disease detection beyond that of the tomato bacterial spot.