Machine Vision & Web Inspection Clarity of vision
By John Cusack, business development specialist for Baldwin Technology Company’s Baldwin Vision Systems business unit.
I
f you ask any printer what are the key tools that help guarantee print quality, most will list 100% print Inspection technology in their top five. But simply having 100% inspection does not always guarantee that the ever-increasing quality demands of brand owners are being met. So, let’s look at some of the critical area’s printers should understand to achieve that high level of quality. It sounds obvious but fundamental performance is often not understood. Can your system detect the defects you expect? With any “100%” inspection product, the key point is “100%” and what that really means for you as a printer. There are many factors that impact an inspection system’s ability to find the defects you care about – resolution and inspection algorithms are two of the most important ones.
NOT ALL SYSTEMS ARE EQUAL Two systems can inspect the same product, but if their resolution varies, so will the results. As resolution goes up, so too does inspection performance.
At Baldwin, we focus on what the printer’s needs are and then specify the correct system configuration to meet those needs. Our print quality verification (PQV) 100% system has been optimised for the needs of demanding Label printers with resolution and inspection algorithms designed for very detailed artwork inspection. The PQV uses an ‘objects based processing’ technique to automatically identify and inspect every text or code element within the artwork. It automatically applies a high sensitivity to these key components and while applying an alternative sensitivity to all other non-critical areas of the artwork. This approach allows the PQV to find small, but critical errors in locations such as health warnings, nutritional panels or even in variable data codes in real time.
Inspection system providers often specify performance based on the smallest detectable defect. While this is driven partially by resolution (how many pixels represent an ink splash for example), it is also driven by the artwork itself. Just like for humans, it is easy for an inspection system to find a black ink splash in a white background as it is very high contrast and a new ‘feature’ that appears in a featureless landscape. But just as often, this same black ink splash can fall onto
a more critical area like a barcode or nutritional panel causing a potentially fatal error. In this case, the defect may not have much contrast as it is against a background of other black features. Using the powerful objects based processing technique, the PQV system automatically identifies and tracks every text and code element in the artwork independently. This allows the system to be completely immune to subtle, acceptable process variations such as print registration allowing the system to detect the same, tiny ink splash in those critical areas of the artwork. As the print industry continues its migration towards mass customisation and serialisation supported by the growing impact of digital print technology, it is causing a seismic shift in how print is inspected. Most 100% systems in use today are not equipped with the intelligence to support inspection of codes and variable content. While they can detect printability issues they cannot read and verify the content. The PQV has been future proofing printers with support for reading and verification of fixed and variable content for years. These optional software features may not be initially purchased by users but over time, as their work has evolved, these advanced software tools have been enabled to great effect. Take this set of codes and text strings. The PQV automatically finds and identifies each
requiring a more strategic approach to quality and defect management. With an ever-increasing trend toward low skilled and untrained labour, staff turnover is often high, and training can be a challenge.
To address these skill and training issues it is important the tools operators use are simple, intuitive and quick to learn. With the introduction of its 4th generation packaging inspection system called Defender, Baldwin addresses these issues head on. Using a simple step by step training wizard Defender guides the user through the setup process with helpful hints and cues to ensure no step is missed along the way. It also uses state of the art, multi-touch UI that users will find familiar and extremely efficient to use. The system can be set up by a new user in 30s, making it ideal for a low skilled, high turnover environment. It also removes the need for complex integrations to workflows or MIS that might help automate more complex user interfaces.
The other key area where many inspection system solutions fail to deliver is defect management. Many state they can track defects into a finishing step for removal, but in most cases, flexible packaging printers need to find only the most significant waste. To achieve this, it is critical to have high quality data to drive business decisions.
of these codes using the same Objects Based Processing used on traditional content. It can than apply a variety of software tools such as OCR/ OCV and data file matching for variable content ensuring complete conformance. Where codes are misprinted or simply incorrect, the PQV can provide the printer a report on the missing codes so they can be quickly reprinted to achieve 100% yield.
Print applications vary wildly in their technical needs, workflow, and end user requirements. For example, Flexible packaging printers often face a very different set of inspection challenges to those in the Labels space. With high volumes of relatively low value product, margins can be very tight,
Obvious datapoints are defect location and image, but what are also very critical are the more fundamental descriptors of the nature of the defect that can facilitate fast filtering and navigation of the data to optimise the defect management process. Baldwin’s defect management database – DataCentral leverages the powerful set of defect classification tools on the Defender 100% system and an SQL database to catalog and store defect information for easy review.
DataCentral provides the user tools that allow them to filter defects by size and type – such as splash, streak, color defect, etc. – to quickly drill down into the real areas of concern. The following image shows a production roll with the raw defect data displayed. This roll has over 1000 defects. From this view, it is impossible to know what material should to be removed.
Xwww.baldwinvisionsystems.com
36
May 2022
www.convertermag.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46
