• 14 FEB 17

    Reading the Not-So-Fine Print

    Cast your minds back to January 2017, if you can, and remember the incident this very site reported on – the recall of Spar’s egg mayonnaise sandwich filler due to an incorrectly printed ‘use by’ date.

    We mentioned at the time that vision equipment was a good way to ensure products had the correct information – very true – but it is also very true that reading this information at high speeds has historically been difficult to perform accurately. Indeed, the development of a software tool that can accurately interpret dot matrix inkjet printing at high speeds has been something of a Holy Grail for vision inspection.

    Award Winning

    That is probably why any effort by a vision inspection company to more effectively read this text will turn some heads. METTLER TOLEDO’s CIVCore Software recently did just that with the debut of its Dot Print Tool, which just won the Packaging Digest WestPack 2017 Innovation Award. With the Dot Print Tool, CIVCore is able to detect and interpret variable inkjet printing at previously unattainable levels of accuracy by essentially teaching the software to play connect the dots.

    As a quick review, here’s how inkjet printers work:

    A jet of ink is fired through a nozzle (or nozzles) on to the product surface. Most expiration dates and lot numbers printed in-line use a dot matrix font, in which evenly-spaced dots give the appearance of a particular character. The problem with dot matrix fonts is that any missed dot caused by a clogged nozzle, low ink or movement of the target surface can mean the difference between, for example, a Q, O, G or 0. For critical information such as a lot code, that confusion can be a real problem.

    The Dot Print Tool utilizes technology from Matrox Imaging SureDotOCR. This allows users to specify the size of the ink jet dots and, rather than specifying an exact location for each individual character, instead gives a general expected size and region. The tool then searches for collections of dots which match the expected results. Because precise locations for each character are not programmed, characters can be identified even if they are on an uneven surface. The software looks for the characters it expects to be present, and if it cannot find them the package is rejected.

    This allows manufacturers to program the acceptable expiration dates for a production run into the system and know for certain that each product coming off the line has the correct expiration date. It certainly beats having to conduct a product recall, doesn’t it?

    For more information on how vision systems read text and codes, you can read our white paper Making Codes Readable.

    Leave a reply →

Leave a Reply

Your email address will not be published. Required fields are marked *