It’s a far cry from the long-established technique of using low-energy X-ray imaging to detect foreign objects in food items during processing. That works by detecting material that is substantially higher in atomic number than the food matrix. But contrasting densities on their own are not enough to guard against cosmetic issues.
Enhanced detection capabilities and higher resolution are required to look at levels of contrast – the crevices in the surface of a dessert created by a cutting tool, for example. This involves detecting differential greyscales in the inspection image, which are considerably smaller than the variation in the normal range of greyscales in a standard product.
However, in a tart or pie with a semi-glutinous filling, the crevices will tend to close up once the blade has passed through the product – due to the fluid nature of the filling. So the variability in the image is likely to be within the normal range for the tart or pie. A similar issue would occur when trying to check that a tarte au framboise has enough raspberries on the top.
Conventional rule-based algorithms are no help because of the irregularity of the various aspects of the desserts. That is why the latest generation of inspection technology uses specially developed tools that allow large image sets to be acquired, with defects in differing positions and orientations. The image sets can be processed automatically – and associated diagnostic data collected and statistically analysed – to allow a fast and efficient means of analysing the effectiveness of the algorithms being deployed.
The result is that the technology can detect when a multi-bladed cutting tool on the production line is offset – resulting in uneven slices – by measuring the centre of the slices relative to the outside diameter of the tart. Tarts which fail to meet the quality criterion will be rejected and – if there is a series of rejects – the inspection system will shut down the production line until the tooling is adjusted.
Fragile items damaged during production can also be identified by X-ray inspection. Products such as fruit turnovers are quite fragile when baked and chunks of pastry can easily be knocked off on the production line. X-ray technology can monitor the overall shape of each part of the turnovers and reject those with bits missing – even after they are packaged.
Meanwhile, to ensure a tarte au framboise has a consistent covering of raspberries to look good, an X-ray image of the product is taken – either within or outside its carton – and an image processing algorithm applied to determine the coverage. A similar approach can be used to detect the presence of the small plastic tripod designed to protect the fragile top of items such as lemon meringue pies so that they do not get crushed. It is important to ensure that a tripod is present on each pie – but only one.
Quality control is particularly crucial for luxury bakery products as they are often in transparent packaging which means the consumer can see the items on the supermarket shelf. Damaged and non-uniform products will be left behind and often thrown away – adding to food waste and leading to lost revenue for manufacturers. That’s why the latest generation of X-ray inspection technology is hitting the sweet spot.
Dr Richard Parmee, founder and CEO Sapphire Inspection Systems
Virgin Atlantic’s Flight100 saved 95 tonnes of CO2 in first SAF flight
Good comment. I think these reports are different from many others , in that they were prepared outside the government and the issues, they raised, of...