Wageningen UR Greenhouse Horticulture has started a public-private partnership this year, for a project focusing on automatic inspection of the quality of vine tomatoes at the packaging line. Currently, quality assessment of vine tomatoes takes place through visual inspection by employees and quality controllers.
Higher quality assurance camera technique
The first phase consists of a feasibility study, which has to answer the question whether, and with what camera technique, it's possible to realize higher quality assurance than with the current visual scans by employees. Different rejection characteristics, such as fruit defects (spots, blossom end rot, tearing, collapsing, etc.), defects on stem and crown (yellowing, blight, crown mould and rot), and detection of ripeness (green-red colouration) are researched.
Image analysis vine tomatoes
Visual data of vine tomatoes has been collected using a number of different camera techniques, including colour, hyperspectral and in the visible spectrum (400nm-1000nm) and the infrared spectrum (900nm-1700nm) and chlorophyll fluorescence. Initial analysis of the data has shown that it should be possible to develop algorithms for many of the defects. A particularly interesting wavelength region for detection of stems and fruit affected by fungi is the (near) infrared region, which can't be perceived by the human eye.At the moment, in Wageningen they're working on a lab environment to record a large amount of tomatoes with the selected cameras. Afterwards, image analysis algorithms will be developed for real-time analysis. If the classification score is high enough, realization of an inspection module will follow, which can be inserted into the packaging line.
Source: WageningenUR
Source: WageningenUR