Research: RGB cameras to detect diesases

By tracking subtle changes in plant size, canopy structure, and color, researchers in Italy demonstrate that RGB-based imaging can distinguish drought from disease and identify stress-resistant tomato varieties.

Farmers often discover crop stress only after visible damage appears—when yield losses are already inevitable. Precision agriculture aims to boost productivity while reducing inputs such as water, fertilizers, and pesticides. High-throughput phenotyping (HTP)—the automated measurement of plant traits using imaging platforms—has become central to this effort. While multispectral and hyperspectral sensors are powerful, they are also costly and complex. RGB cameras, by contrast, are affordable and widely available, but their ability to interpret different types of plant stress has remained uncertain. In tomatoes, HTP has been widely used to study drought responses, yet much less attention has been paid to biotic stresses caused by viruses, fungi, and nematodes, especially using simple RGB data.

A study published in Plant Phenomics on 30 September 2025 by Fabrizio Cillo’s team, Istituto per la Protezione Sostenibile delle Piante, Consiglio Nazionale delle Ricerche, highlights a practical, low-cost path toward smarter crop monitoring and more sustainable precision agriculture.

Using a Scanalyzer 3D platform for RGB high-throughput phenotyping (HTP), researchers challenged multiple tomato genotypes with five stressors—tomato spotted wilt virus (TSWV), corky root rot (CRR), root-knot nematode (RKN), and drought across two seasons. They extracted 20 image-derived indices and focused on four key traits: plant height, projected shoot area, shoot area solidity, and senescence index. Visual disease scoring and targeted measurements, including symptom progression, pathogen pressure, and biomass traits, were used to validate HTP data. Results showed distinct stress signatures: TSWV symptoms appeared at 10 days post-inoculation (dpi), with susceptible lines showing bronzing, yellowing, and necrosis, while resistant ‘Dobler F1’ plants showed no symptoms. TSWV RNA levels increased more than 100-fold in susceptible ‘UC82’ plants. Under CRR pressure, the susceptible ‘UC82’ showed significant root symptoms, escalating from 28 dpi to 75% severity by 91 dpi. RKN caused a steady increase in root galls from 20–60 dpi, indicating successful nematode development. Drought resulted in biomass and yield reductions, with shoot fresh weight down by 16–34% and fruit weight by 39–88%, while leaf proline increased during drought and returned to control levels post-rewatering. HTP effectively tracked these outcomes, with size indices like height and shoot area marking stress, especially for TSWV and drought. Senescence increased under biotic stresses and peaked with RKN, while drought caused slower growth and partial recovery. PCA revealed clear separation between biotic and abiotic stresses, with color indices indicating greener, more compact plants under drought and faster senescence under biotic stresses. Genotype discrimination was strongest for TSWV, where resistant varieties showed near-control profiles, while responses to CRR and drought were more variable.

These findings show that low-cost RGB imaging can support precision farming decisions, from early stress detection to variety selection. Growers could identify whether crops are suffering from drought or disease and respond with targeted irrigation or pest management. For breeders, RGB-based HTP offers a scalable tool to screen large populations for stress tolerance without expensive sensors.

Reference: Maria Isabella Prigigallo, Giovanni Bubici, Giorgia Batelli, Antonello Costa, Monica De Palma, Maria Teresa Melillo, Angelo Petrozza, Alessandra Ruggiero, Giorgia Sportelli, Stephan Summerer, Pasqua Veronico, Francesco Cellini, Marina Tucci, Livia Stavolone, Stefania Grillo, Fabrizio Cillo, High-throughput plant phenotyping identifies and discriminates biotic and abiotic stresses in tomato, Plant Phenomics, Volume 7, Issue 4, 2025, 100124, ISSN 2643-6515,

Full article at: https://doi.org/10.1016/j.plaphe.2025.100124