When not in season, most of the corn-on-the-cob consumed in the UK is imported, predominantly from Southern African countries. The corn is shipped whilst still "in the leaf", as this acts as a natural protective wrapper. Corn-pickers feel the cobs prior to picking to gauge whether they are of sufficient size/quality, but without being able to strip the leaf and see the cob, the grading is far from perfect. The result is that large quantities of bad product are shipped 1000s of miles, only to be rejected at the packaging plant.

The aim of the project was to develop a system that could automatically detect damaged or undersize fruits whilst they were still in the leaf. These could then be used locally to where they are harvested - for animal feeds, maize products or biofuel production - rather than being needlessly transported.

The solution used a monoblock X-ray source and area-scan detection to collect digital images. These were then analysed using an inspection routine developed in Scorpion Vision, which measured the cobs, identified breaks/contaminants and issued a pass/fail signal.

Key to the inspection system was understanding how to optimise the performance of the system, by presenting the best possible image to the analysis algorithms. Careful selection of X-ray and imaging parameters enabled the analysis to focus on the fruit body, rather than the surrounding leaf.