GRADING DIEFFENBACHIA PLANTS AT DIFFERENT GROWTH STAGES USING DIGITAL IMAGE PROCESSING

The main objective of this research was to investigate the possibility of increasing the uniformity in groups of Dieffenbachia plants by grading the plants at different growth stages using Digital Image Processing (DIP). Therefore growth experiments were performed in order to test whether it is possible to grade plants using DIP and which grading features should be measured by DIP. The plants were measured at the unrooted stage, after six weeks (half-grown), after nine weeks (half-grown), and after eleven weeks of growth (full-grown). At the eleven week old stage they were also judged by an expert.
The corrected leaf area proved to be the feature at the unrooted stage which is best related to expert judgement at the full-grown stage.
At the six and nine week stage the volume and the projected area from top-and side-view showed to be the features with the highest relationship to expert judgement.
From the grading experiments it was seen that the uniformity at the full-grown stage in ordered blocks is higher than in random placed blocks.
The ordering feature for the ordered block was the corrected leaf area.
The uniformity can even be improved if regarding in the six week old stage or the nine week old stage is applied.
The ordering feature at these stages is the projected leaf area from the top-view.

II International Symposium On Sensors in Horticulture

J. Dijkstra, J. Meuleman

https://doi.org/10.17660/ActaHortic.1998.421.9