ISHS Acta Horticulturae 507: III International Workshop on Models for Plant Growth and Control of the Shoot and Root Environments in Greenhouses
INTRODUCTION OF A MARKET ORIENTED INFORMATION SYSTEM FOR GROWTH CONTROL OF FICUS BENJAMINA |
| Authors: | M.W.C. Dijkshoorn-Dekker, B.A. Eveleens-Clark |
| Keywords: | Image processing, neural network, decision-support system, Ficus benjamina, growth control, growth, quality, grading |
| Abstract:
Control of plant development is a major challenge in commercial potplant production. The Market oriented Grading and Control system (MaGiC) provides a new tool in the production process to meet the demand of the market to deliver an uniform product with a preferred quality on a specified delivery date. This system integrates the measurement, grading and control of plant quality development. Such a MaGiC-system has been successfully developed and tested at the Research Station in Aalsmeer (The Netherlands) in 1997 using Ficus benjamina ‘Exotica’ as a pilot crop. Image processing is used as a measuring technique to register plant development. Image features are used as an adequate input for the neural network and the empiric growth control model included in the system. Application of this system reduces labour costs and facilitates the complete automation of internal transport, spacing, grading and marketing. A prototype was used last year at commercial companies in The Netherlands. Two batches of half-grown Ficus plants were formed, a control batch to be controlled by the grower and an experimental batch to be controlled by the MaGiC-system. The neural network graded the experimental batch into uniform groups. Each uniform group was subjected to a different spacing schedule and temperature regime as determined by the growth control model. At the specified delivery date the experimental groups, development and quality were more uniform than the control batch. However not all experimental groups had reached the marketable stage and therefore they could not be combined into one uniform batch. Although the growth control model can accurately predict the development and quality for each group to the specified delivery date, this was not achieved due to the inflexibility of the present logistics and inadequate facilities on the nurseries. | |
Download Adobe Acrobat Reader (free software to read PDF files) | |
