|
One of the first successful applications of WAY-2C was to
sort unfinished red oak for flooring applications. The speed with which
WAY-2C can be trained makes it possible to select representative
samples representing the range of colors in a shipment,
instantly train the system to recognize this range, and then
automatically sort and/or label the entire shipment by color based on this training.
WAY-2C's powerful ability to differentiate objects, such as these cabinet doors, based on their color distributions usually produces results consistent with, but more reliable and much faster than human inspectors. Even orientation and natural variations, such as those due to wood grain or irregular outlines, which can defeat traditional mean and/or variance based matching, or pixel thresholding methods, are seldom, if ever, a problem.
|
