Probe-based hyperspectral imager for crop monitoring

Automated crop monitoring techniques help in better management of growing conditions in order to improve quality and yield, and to reduce the impact on the environment. Various stages of a crop in its life cycle is noticeable with a change in its leaf color. The yellowing of leaf is considered as an important quality defect in green leafy vegetables. Yellowing is initiated at the end of maturity stage and continues in the senescence stage. Most of the methods used for monitoring leaf quality requires the detachment of the leaf from the plant, which is destructive in nature. Among the several nondestructive techniques available, hyperspectral imaging (HSI) modality offers the possibility to address this problem by monitoring the reflection spectrum of the leaf in situ. When the freshness of the leaf reduces, the chlorophyll content in the leaf decreases. This results in an increase in its reflectance in the visible region as the absorption of light by chlorophyll reduces. Hence, the reflection spectrum can be used as a measure for the freshness of the leaf. However, monitoring large areas usually require translation of the whole imaging system or removal of the leaf from the plant. In this context, we propose to use a flexible probe-based HSI system to mitigate these issues. We demonstrate the adoption of a probe based HSI modality to enable in situ live plant monitoring. The classification of the leaves from the HSI data is performed using principal component analysis (PCA) technique.