Machine Vision Takes a Global Look

Complex Inspection and Error-Proofing Applications Have Traditionally Required Expensive Components That Are Difficult to Install and Use

By Loren Shaum, contributing editor

With vision installations growing at double-digit rates annually, a few smaller suppliers are gaining notice because they’ve added value to a few well-niched vision solution offerings that compete well in industrial applications.

“Most [EU suppliers] offer premium products in applications that provide cost efficiencies,” says Patrick Schwarzkopf, general secretary of the European Machine Vision Association. “The challenge is to stay ahead of the innovation curve to keep an edge over competitors.”

It’s clear that machine vision systems are getting more compact, faster, with better resolutions, and can be adapted quicker to production changes without much knowledge. “And most importantly,” says Schwarzkopf, “through the progressing standardization initiatives, it’s getting easier to integrate vision systems into complex production environments.”

Vision Components GmbH (VC) founder Michael Engel claims that he developed the first smart camera in 1995. “Surely, everybody is talking about processor performance, but speed isn’t everything,” says Engel. “The product needs to be easily programmable, and there must be competent support.”

VC has introduced a smart camera system with stereo-imaging capability, which lets one image capture seek one feature of interest, while the second looks for a different feature.

Pepperl+Fuchs introduced an approach that combines camera, illumination, digital outputs, process data and five evaluation methods in one smart-camera housing. The result, says Gary Frigyes, P+F’s product marketing manager, is a solution that’s configured easily and operates without detailed image processing experience. “Complex inspection and error-proofing applications traditionally required expensive components that are difficult to install and use,” says Frigyes. “The hardware and software are standardized in a fully integrated package, eliminating the need for more complex integration of individual components.”

In a drink-box application, says Frigyes, the system could determine if the straw is glued to the box, tamper-proof seal is in place, bar code is printed and package is labeled correctly and closed. Ethernet is used to deliver information to host controllers.

Dalsa offers vision solutions from frame-grabbers to turnkey solutions. Dalsa claims recent advances in color vision are pushing into mainstream inspection—comparing it to human vision inspection, to some extent.

“Of course, vision actually can’t see in color,” says Robert Howison, project leader at Dalsa’s OEM application group. “Machines use mathematical models to approximate human color detection. A machine can be calibrated against average human response to color and give consistent responses to colors observed in a controlled setting. This calibrated color vision is useful for matching colorants in paint, plastics and fabrics. It’s important to make the distinction between the relative measurements made with a machine vision system versus absolute measurements from devices such as photospectrometers.”

To expand color inspection capabilities, Dalsa offers a variety of distinct color imaging technologies, including Bayer color filter array, beam splitter prism and tri-linear sensor. These recent advances allow Dalsa and its partners to take on highly complex bank-note inspection.

Bank-note quality has been determined largely by means of human inspection that focuses on visible aspects of the product and hence yielded subjective observations and lack of consistency. Today, automatic inspection and sorting systems, such as those from Dalsa’s partner, system integrator Parvis Systems and Services, Milan, Italy, are used in many printing applications to verify 100% of the final product. Defective bank notes identified according to specified parameters are sorted and removed for destruction. The inspection system can identify process deviations and warn the operator that intervention is required to minimize downtime.

Show Comments
Hide Comments

Join the discussion

We welcome your thoughtful comments.
All comments will display your user name.

Want to participate in the discussion?

Register for free

Log in for complete access.

Comments

No one has commented on this page yet.

RSS feed for comments on this page | RSS feed for all comments