- » View All
No Crawling Allowed When It Comes to Automated Web Inspection
by Winn Hardin, Contributing Editor - AIA Posted 06/26/2014
The world isn’t getting bigger, but it is getting more complex.
The Earth is still roughly 24,902 miles around its middle. However, the number of people on the Earth’s surface has grown from 1 billion in 1800 to more than 7 billion in 2013. In other words, while the physical size of our planet hasn’t increased, the density of information as represented by sentient beings, the primary drivers of complexity, has increased considerably. (True, not all brothers-in-law qualify as sentient beings, and one might question the 1 in 6 people who use Facebook on a daily basis, but even subtracting these two groups, there is still a lot more going on than when Napoleon was named First Consul.)
What does this rumination on the world’s limited size and growing complexity have to do with machine vision? Like the world at large, web inspection applications — which use machine vision technology to inspect and grade the quality of “continuous” products such as paper, textiles, plastics, and glass, to name a few — have to analyze more information in the same amount of space (and in even less time). Printing presses and automated looms are not necessarily producing wider rolls of material, but they are producing them faster as customers demand higher quality and greater accuracy.
Machine vision web inspection systems are responding by using advances in hardware, software, and the engineering genius that binds them together to analyze more data faster and with greater accuracy — whether it is verifying that the package of your favorite bread clearly says “gluten free” or that your IRA statement doesn’t accidentally place Bill Gates’ balance where yours should be, leading to an unfortunate decision to retire prematurely.Web Inspection Applications
Just as a rip in a spider’s web means a hungry belly and extra work for Charlotte and her offspring, a rip in a printing press, poorly registered print pattern, bubbles in plastic or glass sheets, and missing threads from fabrics all spell trouble for their manufacturers.
To capture these defects with a web inspection system, machine vision designers often have to push the technology to the limit as part of an application-specific machine vision (ASMV) solution. For example, Boulder Imaging (Louisville, Colorado) recently landed a web inspection project at the United States’ most prestigious printer, Crane Currency, which produces the paper for the U.S. dollar, passports, and other important government documents.
Printers like Crane cannot afford to have security features embedded in their specialty paper products be out of specification or they risk losing one of the largest printing contracts on the planet. Boulder Imaging recently installed its SPX Series inline inspection system that allowed Crane to increase production speeds from 250 feet per minute with six false positives per hour to 750 feet per minute with one false positive every two to three days, all the while enabling 100% micron-level inspection at the higher production speed.
In a similar application, Videk recently developed a print inspection system for financial, insurance, healthcare, government, utility, and other customers that produce high quantities of printed materials with personalized information on each page, such as credit card statements and explanation of benefits statements.
These industries traditionally have had to depend on manual inspection of roll-to-fan-fold documents rather than automated inspection of roll-to-roll printing processes due to the need for high accuracy of the printed data. Roll-to-roll offers faster operations, which translates to operational cost savings but cannot be manually inspected since the paper is “rolled" when it exits the printer.
Videk developed its DocuVision 8600 inline inspection system using Teledyne DALSA’s (Waterloo, Ontario) Piranha 4K color line-scan cameras and proprietary encoding electronics that allow the camera trigger to be accurate at web speeds up to 1000 feet per minute. Videk also chose high-speed Karbon and Neon frame grabbers from BitFlow (Woburn, Massachusetts) to handle the high-bandwidth image data streams and custom dome lights from Metaphase Technologies (Bristol, Pennsylvania) to avoid any shadows caused by ripples in the paper.
Line-scan cameras have dominated web applications for many years because of the nature of the moving target, as well as their ability to create continuous images by stacking one line of pixels on top of the next. However, this approach places extreme requirements on the lighting system to keep pace with the growing speed of web inspection systems.
According to Steve Kinney, Director of Technical Pre-Sales at JAI, Inc. (San Jose, California), this is part of the drive behind its new SW-2000 2k line-scan camera with 20-micron pixels versus the industry standard 7-micron pixels. “At high speeds for print inspection, we can offer considerably higher signal-to-noise ratios for light-limited applications,” says Kinney.
While packaging represents one of the largest markets for machine vision technology, the printing industry itself has struggled in recent years due to competition from digital media and the after effects of the 2007 recession. The pain is made worse because tight margins haven’t greatly reduced the number of printing companies fighting for business in what remains a highly fragmented industry.
To maintain healthy operations, printing and other web-based production industries have to use automation to keep costs low while continuing to improve quality. As machine vision cameras and frame grabbers continue to increase the speed at which they can acquire images and perform quality inspections, the growth of industries that depend on web production will not have to slow to a crawl but instead can jet ahead to higher margins and productivity.
There are currently no comments for this article.
Leave a Comment:
All fields are required, but only your name and comment will be visible (email addresses are kept confidential). Comments are moderated and will not appear immediately. Please no link dropping, no keywords or domains as names; do not spam, and please do not advertise.