Resources
/
News

Hyperspectral Disease Detection in Wine Grapes

January 22, 2021

New research has demonstrated the power of hyperspectral cameras to see into the chemical profile of plants.

Early Detection of Plant Viral Disease Using Hyperspectral Imaging and Deep Learning

Crop yield is affected by diseases, which have the ability to inflict significant economic loss. Apart from pesticide and herbicide use, early detection is the regime to monitor and control plant diseases. One such condition is caused by a DNA virus, the grapevine vein-clearing virus (GVCV). Grapevine is a highly profitable crop, but it is also prone to many diseases. Risks are high particularly amid changing climate and erratic weather. Viral infection, however, does not immediately affect the phenotype, or the observable physical traits of the plant leaves. But chemical changes inside the leaves occur. The virus causes damaging reactions with unique spectral signatures identifiable by spectral sensors. New research has demonstrated the power of hyperspectral cameras to see into the chemical profile of plants. The main fluorescence constituents analyzed in this study are chlorophyll, anthocyanin, and pheophytin. The concentrations of these constituents are affected by the viral infection. Researchers investigated a hyperspectral push-broom imaging technology, similar to the Opsyne VNIR-1384 camera, to analyze the grapevine fields remotely. This portable spectral imaging technology is non-invasive and generates 204 bands in a wavelength range of 397-1004nm. As the infection spreads, the pigmentation of leaves is changed. Hyperspectral images are classified using Random Forest (RF) machine learning algorithm. Further commercial adoption would benefit from greater optical sensitivity and fewer spectral bands to increase performance, sensitivity and specificity.

Book a consultation to learn more

We can walk you through how hyperspectral imaging can improve your workflows.