Robotics Reference
In-Depth Information
-4-
How Computers Recognize
Visual Recognition—How Computers See
Human vision is an amazingly complex process that involves acquiring
and processing massive amounts of visual and other information. Our
eyes and brain collaborate, to pick up the visual information we are seek-
ing and to discard redundant and unwanted information. As you read
these words on the printed page your eyes are focused on the words. You
can see a lot of other information on the page, but for the moment your
brain ignores this extraneous information to some extent, allowing you
to concentrate on the word of the moment and pass on to the next one,
rather than becoming confused by taking in every word that your pe-
ripheral vision can see. This collaboration between eye and brain is a
computational process that has evolved over millions of years and is one
of the marvels of nature that we have come to take for granted. Those of
us who can see rarely stop to think about how fortunate we are—the gift
of sight is an assumption.
The human eye, as shown in Figure 27, is made up of several layers,
of which the innermost is the retina. The eye has muscles that control
the shape of the lens, thus enabling the eye to focus. And, like a camera,
Figure 27.
Cross section of the human eye (from the web site “Computer Vision” by
James West)
