New technique adds speed and ease - Ohio State Univ researchers are using a computer vision sensor to speed up digitizing process for designing manufactured products - Brief Article

USA Today (Society for the Advancement of Education), June, 1997

Researchers at Ohio State University. Columbus, have found a way to shorten a key process used in designing many manufactured products, from car parts to detergent bottles. In the process, called digitizing, a computer scas an object -- a clay prototype of an auto part, for example -- to obtain its exact dimensions. This data is used to mass produce the part. The problem is that digitizing is a slow, laborious process. Using a laser sensor, manufacturers have to measure many thousands of "digitizing points" on an object's surface to get precise data.

Yuan Zheng, professor of elective engineering, points out that a method using two types of sensors canneeded to digitize a small car part was cut from nine hours to about three and a half.

Currently, many manufacturers use a laser displacement sensor (LDS) to digitize objects. The LDS works by bouncing laser beams off thousands of points on an object's surface. A sensor picks up the reflected beams and measures the distance they travel. These measurements, taken over the entire surface of the object, help provide an extremely precise representation of the surface. The LDS is very accurate, hut also very slow.

Under the new method, objects first are digitized with a computer vision sensor. One version -- called a photometric stereo device -- works by shining a light on an object in three different positions. Comparisons of the lighting intensity under these dimensions.

Zheng indicates that the computer vision approach is much faster than, but not as accurate as, the LDS system. However, a computer vision device can give a global picture of an object's shape. Using that global picture, manufacturers can determine which areas of the object need a through scan with the LDS and which portions need only a more cursory scan. "By combining use of these two sensors, we can improve scan. "by combining use of these two sensors, we can improve both efficiency and precision in the digitizing process."