3D Machine Vision Improves Quality Control

Measuring the three-dimensional shape and color of an object, more reliable decisions can be made during the inspection and production of parts, improving quality.

By Jim Anderson, Sick July 14, 2010

Customers and machine builders always look for ways to cut component counts, reduce complexity, and increase throughput in a manufacturing process. New 3D camera technology can help improve quality and reduce costs by measuring the shape and color of an object during production.

Users can improve system up time, increase throughput of quality parts, and scrap bad parts sooner. 3D imaging is ideal where height, shape, volume, and even color is important. This is common in grading fruits and vegetables, baked goods, lumber, electronic assemblies, cosmetics, and pharmaceuticals. With 3D and color, more reliable inspections improve quality and reduce waste.

One such product, Sick ColorRanger 3D Camera, can replace several cameras for the same task, reducing cost and complexity by combining functionality of a 3D camera and a linescan color camera. The camera can be configured to provide several measurements at the same time, such as 3D shape, laser scatter, color, and monochrome, according to the needs of each application.

The camera works on the general principle of laser triangulation for the 3D data acquisition, collecting up to 35,000 3D profiles per second. Chip design parses the imager into separate cameras so multiple algorithms can be run simultaneously, allowing acquisition of the color data. Color data is acquired from three rows, creating a special offset among the three color segments.

A combination of algorithms, color, and 3D has been used in the baking industry and construction materials, such as bricks and clay, for years. Manufacturers often used large, complex systems to measure the overall product shape and compare it to the desired part. A second color camera would image and analyze the items for baking degree or “doneness.” With one system, machine size can be reduced, and throughput of the line can be increased due to the camera’s high-speed imaging.

This article appears in June/July 2010 issue of Control Engineering. Learn more about this 3D imaging system below.

-Jim Anderson is vision product manager for Sick, www.sickusa.com.

ONLINE extra – The following information didn’t appear in print, but does appear with this online posting.

Software helps the hardware

Sick’s “iCon” PC software handles that offset. The filtering function adapts for the variation in speed using mark data from an encoder. Improved structure for getting encoder data eases color image registration.

The camera can take advantage of the recently implemented “Coordinator” tool, calibrating the 3D data. The calibration and scalable target for different fields-of-view, reduces to minutes a job that could have taken hours or days.

Calibration look-up file is stored in the camera’s flash drive, making real-world calibrated points available for application programming.

See other machine vision articles in this issue.