An automated growth measurement system for leafy vegetables

Abstract

In plant science, the fundamental information for research and related applications is derived from the measurement of plant features. It is especially useful for applications in plant growth modelling and climate control in greenhouses or plant factories. Standard, direct measurement methods are generally simple and reliable, but they are time consuming and laborious. In contrast, vision-based methods are non-destructive and an efficient way to describe external plant features and plant growth. In this study, a stereo-vision system, using two off-the-shelf cameras with parallel optical axes, integrated a self-developed image processing algorithm to monitor the growth of Boston lettuce in a plant factory. The system was mounted on a sliding rail to extend the field of vision of planting beds. Images were continuously recorded to determine the plants' features and construct panoramic images. The image processing algorithms, that calculated geometric features such as the projected leaf area, plant height, volume and diameters were developed and incorporated into the automated measurement system. Subsequently, plant growth curves were determined from calculations of the plant features data. This automated vision-based system showed promising results when put into practice.