Using a Flatbed Scanner

Equipment
In order to take 3D's with a scanner, you need a flatbed scanner, like the one shown here.  I would imagine that just about any flatbed would do, though the quality does vary a bit between brands.

Technique
Taking 3D's with a flatbed scanner is easy.  It is not something I thought up, but it makes sense when you look at the inner workings of a scanner.  My hat's off to the person who came up with this- it is somewhat counter-intuitive.

This works because of two reasons:

1.  The scanner has a good depth of field- i.e., it can scan more than a flat piece of paper.  Take something from around your home, such as a flower or a small carving and place it face down on the scanner.  You will probably be able to see detail at least 1 inch above the glass before it starts fading out.

2.  Images of a non-flat object taken on the left side and right side of a scanner will be different, due to the reasons described below.

I've created several drawings of the inner workings of my scanner- a Microtek ScanMaker E6.  Most flatbeds should work on the same principle, however.  In Figures 1 and 2 below, I show a top view and a side view.  You can click on the top view to show the effective field of view of the optics.

The way it works is that a series of mirrors are used to shorten the distance between the scanning element, a Charged Coupled Device (CCD), and the glass surface of the scanner.  Without these mirrors, the scanner would be either very deep- greater than 1 foot (30 cm)- or the CCD would have to be as wide as the glass scanning surface.  Click here to see how it works.

Since a large CCD (a semiconductor) is not practical, a small CCD (with thousands of pixels lined up in a row) is placed behind a lens, like that on a camera.

The light source is attached to the scanning platform as well.  As the platform moves, the light is reflected off the media being scanned on the glass, and passes back down into the assembly, bouncing off several mirrors and finally entering the optics and landing on an element in the CCD, as shown below.
 

If you think of the lens (#6) as your eye, you can see that if you look at the horizontal mirror (#5), you could see the entire width of the scanning window (albeit a very thin slice of it).

If your eye were at the center, then an object on the left side would look as if you were viewing it with your right eye, since it is to the left of your eye.  Conversely, if you place an object on the right side, it would appear as though you were viewing it with your left eye.

Using this overhead view as a guid, take an object and scan it first on the left side of the scanner (getting the right eye image), and then on the right side of the scanner (getting the left side image).  You will have a perspective shift between the two objects, and hence, you will have depth information when you view the images together.  Here is an example of this particular piece of coral after it has been scanned.  If you look closely at the edges of the coral, you will see that the two images are not exactly alike, and this is where the depth information comes from.

You will have to experiment with how far you move the objects between scans- being careful when you move them not to rotate them any.  I have found that placing an object 1/3 from the left and 1/3 from the right to be sufficient.  Sometimes you don't need that much.  This works quite well, and you probably will not need to use the full width of the glass to separate objects.

Examples of this method can be seen in my Macro gallery (along with photomicrographs taken with a microscope).  Objects around the house could be anything from rocks, geodes, carvings, flowers, or any miscellaneous item with an interesting surface.

Back	Home