Greenhouse production systems are increasingly complex, necessitating a data-driven approach with robust, intelligent sensors. Although there is a clear advantage for growers to be able to monitor the physiological status of the crop, current practices mostly involve cumbersome, expensive, and slow laboratory measurements. It is shown that visible and near-infrared imaging spectroscopy allows for the rapid and non-destructive assessment of the concentrations of sugars, starch, pigments, various nutrients, and dry matter in the leaves and fruits of tomato plants. A tailored feature selection algorithm also shows the feasibility of using as few as 8 bands with spectral cameras. This study validates imaging spectroscopy as a rapid tool for assessing crop status and fruit quality in greenhouse horticulture.
|