Advertisement
HANNOVER MESSE 2018, 23 - 27 April
Switch to:
Industrie 4.0

The Ever-Watchful Computer Eye

Smart factories are virtually unthinkable without industrial image processing. Justifiably so, as automatic computer systems can detect things invisible to the human eye.

19 Apr. 2016
Bildverarbeitung

Whether for inspection and quality assessment, measuring gaps, position sensing for robots or simply for scanning codes, industrial image processing has gained a foothold in almost every area of factories. That is no surprise, as contact-free measurement of components via image processing offers a real advantage. Measurement is extremely fast – even capable of keeping pace with manufacturing cycles.

Instead of taking individual samples and feeding them into measuring systems, the image processing system is integrated in the manufacturing process, and measures every single workpiece without slowing production. The size of the components is largely irrelevant to the measurement precision. Vision systems can assess the dimensional accuracy of electronic modules just micrometer in sizes, as well as inspecting the gap dimensions of entire car bodies.

From welded seams to irregular peperoni

An increasing number of 3D methods, like light section or stereometric methods, have established themselves. These methods can measure everything from welded seams and shaped sheet metal parts, to irregularly shaped peperoni sausages. For example, the solution by Keyence (Hall 7, Stand A34) monitors the production and packaging processes of cookies , to ensure that no broken cookies end up in the package.

Another advantage of computerized inspection, besides the speed, is its absolute objectivity. One sample application for this is in industrial labelling. Human inspectors are influenced by subjective impressions when assessing the quality of text. Completeness, contrast, sharpness and homogeneity – we are far more tolerant of various errors than the unerring electronic eye.

Typeface control systems compare the test specimen with a selected error-free specification, but require complex set-up procedures. As a result, the Fraunhofer Institute for Factory Operation and Automation (Hall 2, Stand C22) developed an image-based method that assesses the most important features of the text automatically and objectively. It uses a combination of contrast, homogeneity, completeness and font width, and therefore does not need a target specimen, making complex set-up processes a thing of the past.

Tracking down invisible defects

However, digital image processing not only allows the right position to be detected – three-dimensional criteria are also becoming increasingly important. When combined with robotics, automated systems that allow automatic gripping of unsorted parts in boxes can be created, for example. The cameras can even look under the surface: With the latest dynamic heat flow measurement methods, they can even detect invisible defects .

At HANNOVER MESSE, the Industrial Automation area will showcase the latest developments in industrial image processing, for example at Association for Sensors and Measurement (AMA) in Hall 3, Stand F33, or in the Industrial Automation Forum in Hall 14, Stand L17 .

Advertisement