Automatic on-line vision inspection is becoming increasingly important for many food manufacturers, especially for the growing need to ensure that products are allergen free and to ensure that date codes, product origin details, etc. printed on-line are both legible and accurate. Using camera-based solutions, component matching and label verification can be achieved without needing to redesign artwork to introduce identifiers.
Where a human eye easily misses a defect or two, a vision system can fully inspect each individual product to minimise costly batch rejections, rework and waste.
Optical Character Recognition (OCR) is used for printed information that is both machine-readable and human-readable, in contrast to barcodes and 2D symbols. In cases where a printer cannot be configured directly via Harford Autocoding, OCR may be used to verify that the printed data is legible and correct.
Usually, OCR is just one of many inspections carried out by a Harford vision inspection system. The ability to read printed information heavily relies on the quality of the print and the font used. However, even in cases where an inkjet printer is used, we have been able to successfully ensure that print is present and of acceptable quality.
If the packaging or the labels on a product cannot be verified through the use of a barcode or 2D code, e.g. due to artwork restrictions, it may be worth considering a vision inspection system. An in-line camera, appropriately chosen for the application, takes photos of the product and uses a reference or template image to find similar shapes within a new image. This way, regardless of location, rotation, or scaling of the template, we can verify if the correct packaging is being used.
Such systems often comprise a short conveyor with a rejection mechanism and a cover to ensure consistent lighting and minimal environmental influence on the vision inspection process.