Advanced Imaging


Advanced Imaging Magazine

Updated: January 12th, 2011 10:01 AM CDT

New Developments Propel Inspection and Measurement

Advances mean increases in speed and quality at lower cost
CyberOptics' WaferSense™, ATS
© CyberOptics
CyberOptics' WaferSense™, ATS (Automatic Teaching System) uses machine vision to see inside semiconductor equipment.
Cognex's OmniView
© Cognex
Cognex's OmniView uses four cameras to inspect the entire cylindrical surface of bottles, cans, vials or cosmetic containers without orientation or accurate positioning.
Allied Vision intelligent sensor
© Allied Vision
Allied Vision worked with EDAG, a German company, to develop an intelligent sensor to use with its Best-Fit image-processing system in the automotive industry.

By Barry Hochfelder

"In the food, beverage and pharmaceutical packaging industries, OmniView allows customers to more cost-effectively avoid mislabeling errors, comply with safety standards and ensure product quality," says Marilyn Matz, Senior VP, PC Vision Products Group.

In one case, the customer, a leading food and beverage manufacturer, was suffering through significant unplanned costs from labeling-related flaws. The primary vision challenge in this application is to ensure that only the expected labels for each batch run are correctly applied to each container just before it is placed in its carton, thereby guarding against package-related quarantines, rework and recalls.

The customer has several container sizes and literally thousands of labels printed for distribution in several countries, along with a zero tolerance quality policy for any false positives (labeled containers accepted as good by the vision system, when they are actually not labeled correctly). OmniView achieves this level of performance by first accurately locating the entire trained part in 3D, allowing the relevant portions of the container surface to be faithfully represented in a single 360-degree 2D image. The included Cognex vision tools then are used to discriminate the brand and flavor logo graphics at speeds of more than 1,000 parts per minute.

As a secondary vision challenge, additional tool regions are used to detect skewed, torn, flapping, dog-eared and bunched label conditions (among others). Vision results are signaled in real time, alerting the factory operators to intervene swiftly when upstream equipment is no longer functioning within allowed limits.

The customer also plans to start printing 2D data matrices on their labels, which can be additionally verified without slowing down the line (at a maximum of 1,200 parts per minute). The datamatrix's ability to store an increased amount of information will give the customer more detailed label discrimination capability than the traditional UPC or EAN barcodes.

Subscribe to our RSS Feeds