Hyperspectral imaging
Approximately 5,000 images from control and inoculated Alternaria and Botrytis groups are being quality-checked and prepared for AI analysis.
HEALTHYTOMATO
A multi-layer research programme connecting images, sensors, greenhouse conditions and field observations.
Approximately 5,000 images from control and inoculated Alternaria and Botrytis groups are being quality-checked and prepared for AI analysis.
RGB, Green, Red, Red Edge and NIR imagery is used for temporal plot monitoring and five vegetation indices.
An approximately 2 GB image repository is being cleaned, segmented and prepared for expert-validated labelling.
Temperature, humidity, soil moisture, light, CO₂, leaf wetness, EC and pH will support one-, three- and seven-day risk models.
Exploratory AS7265x VIS-NIR studies investigate the potential of low-cost spectral measurements; they are not calibrated quantitative NPK devices.
Scientific note: project prototypes and ongoing analyses must not be interpreted as completed commercial products or final diagnostic results.