The inside track on reading codes and labels

23 February, 2009

Coding, marking and labelling have become an essential component of modern production technology. Correct packaging labelling is critical for consumer safety, especially in the food, beverage, pharmaceutical and medical industries. In addition, ‘mark and read’ applications are used for product identification and traceability, especially in safety-critical industries such as aerospace and automotive.

Tracking a component and all the processes it has gone through, from manufacturing and assembly right through to end-user requirements for spare parts replacement, is becoming an essential requirement as ERP, MRP and quality assurance systems become more widespread throughout the manufacturing supply chain.

Coding and labelling usually takes the form of alphanumeric codes (such as lot details and best-before information), bar codes and 2D codes. Products can be tagged either by a stick-on label or by information printed directly onto them or onto the packaging. Label-free direct marking methods are intended to survive the most unfavorable of production and operational processes, as well as environmental factors.

Reading of these codes is generally carried out using industrial vision systems and dedicated products are available, both for 1D and 2D code reading. More often, however, systems using an illumination system, one or more cameras and lenses to produce an image of the item to be inspected and some digital image processing/analysis capability are required. Industrial vision systems can also be used for other label applications, such as checking the presence or absence of labels, as well as looking for consistency in colours and logo positioning for brand image purposes.

Code reading challenges

In a factory environment, the vision system faces challenges such as:

• random product orientation giving random code or label orientation
• position of the code on the product (irrespective of orientation)
• abnormal codes or characters (such as stretched or skewed characters)
• damage to the code
• contrast between the code or characters and the background
• unsuitable code/background colour combinations
• reflections from surrounding surfaces
• space in the line for the vision system
• interface to a reject mechanism and/or the rest of the process
• speed of measurement

Generally, 1D bar codes are less robust than 2D codes. Although 1D bar code scanners can generally identify damaged codes as ‘unreadable’, developments such as the recently introduced CLV620 compact 1D bar code laser scanner from Sick IVP AB, offer improved reading algorithms in its SMART decoding system for rapid and reliable reading of even badly damaged or partially covered bar codes. The 2D Datamatrix format includes built-in ECC220 error correction to allow the recognition of codes that are up to 60% damaged, making it particularly popular for direct part marking. Since every production line is different, it is advisable that end users work closely with vision specialists to get a solution which delivers the required inspection capabilities.

Choosing a vision system

The UK Industrial Vision Association (UKIVA) is a not-for-profit organisation, concerned primarily with promoting the use of vision technology in industry and science. Members include vision component suppliers, system integrators, consultancy services and academic research groups. The Association is happy to discuss any application of code and label reading and will be able to suggest possible suppliers.

Practical applications

Some practical examples supplied by UKIVA members illustrate the diversity of code reading applications.

Scorpion Vision has established a high speed code reading application for beer cans. This runs at up to 20 cans per second and reads and verifies the 10 character batch code and four character sell by date on separate lines, achieving false rejects of better than 0.005%. Among the problems to be overcome was distortion of the code due both to slight jumping of the cans during coding and the convex shape of the can’s base. With a variety of factors slowing the processing, a dual core processing system was implemented to deliver the required speed.

Many applications involve multiple code reading systems. For example, Cognex installed a system for automotive part manufacturer, Borg Warner, which initially reads the two-line alphanumeric characters engraved on individual components. This data is transferred to another system where it is converted to 2D Datamatrix code which is then engraved on the part as well. A second code reading system verifies the 2D code. A third code reader checks the code after the part has been subjected to hydraulic testing. Machine Vision Technology has developed a system to prevent mis-labelling of cans for Premier Foods. The system essentially consists of dual bar code readers to read the labels on the can and a separate PC-based vision system to read the alphanumeric codes on each can using optical character recognition (OCR) techniques from the Common Vision Blox imaging toolkit from Stemmer Imaging. The alphanumeric code is compared to the bar code and if they do not match, the PC initiates a standard reject or stop line mechanism for automated removal of the can from the line.