## Is there a difference? Computer Holography & Digital Hologra

These are all of the old posts from the first two years of the forum. They are locked.
Updated: 2005-03-28 by HoloM (the god)
Martin_D

### Is there a difference? Computer Holography & Digital Hologra

Hello there,

I've been wondering if there's a difference between "Computer Holography" and "Digital Holography"? Someone told me there was, but he didn't knew the difference. So I thought maybe someone here could tell me...

thanks guys

Martin_D

Dinesh

### Is there a difference? Computer Holography & Digital Hologra

"I've been wondering if there's a difference between "Computer Holography" and "Digital Holography"?"
Both terms cover a multitude of sins. They've both been used, misused and abused.
The fringe structure in a hologram is basically the Fourier Transform of the image. This, if you're not familiar with it, is a mathematical technique that converts an image into a bunch of sin and cosine terms. These individual sin and cosine terms are known as the 'spatial frequencies' of the image and can be plotted. The eye does a similar sort of process in that when you see something, you're brain receives a sort of Fourier Transform of the visual scene, which it 'reverse transforms' back into a visual scene. The reason (1 reason, anyway) you see a hologram in 3D is that the hologram is a photograph of the lines of the Fourier Transform. The plot of sin and cosines. If you can compute the FT of a particular scene and draw them directly onto some kind of medium, this medium will then reconstruct the scene holographically. Simple, in theory!

In the early days (ca 1960's) they used a mainframe computer (pre-Microsoft - they had no Windows!) to calculate the FT curves. they then plotted them on giant paper and photo-reduced the paper. The hologram ended up pretty small. However this technique was obviously expensive in computer power and time consuming. Methods were created to short-circuit the process. One early method was to divide the medium into tiny squares ('pixels' - although not like modern pixels) and calculate what the plotted line would look like only in one particular square. The square ('pixel') was then divided into quarters and, depending on the value of the FT at that square, an appropriate number of quarters were filled in. This is very similar to the way newspapers print pictures. In this sense, 'Compter Generated Hologram' - ie calculation of the Fourier Transform components - was converted and morphed into 'Digital Holography' - where the holographic medium is divided into 'pixels' and each pixel has a value dependant on the FT. In this sense, the hologram is associated with computer graphics in that a computer graphic - 'digital' - image is also composed of pixels.

Today companies like Zebra have a fairly sophisticated setup with fairly sophisticated computers so they can calculate very complex FT's and image them into very small 'pixels'. Here is an example of Zebra's work, done digitally.

Martin_D

### Is there a difference? Computer Holography & Digital Hologra

Ok thanks, got that so far... but what is that technique called where you would record an interference pattern with a CCD camera?

And isn't there something called "Interactive Computer Holography? Is this something different?

Btw, you don't happen to know some good references on this subject? I mean not too mathematical...I know the basics of Fourier Transform, but I'm by no means an expert...

Sergio

### Is there a difference? Computer Holography & Digital Hologra

Interesting holograms and work, those pixel structure is recorded holographically in a H1 master as a FT "pixel" dot hologram? So after that the master convoluted with thousands pixels is recorded on the H2 reflection? Still good holographic materials are necessary for good quality holograms. Seems the still use the DuPont material.

Dinesh

### Is there a difference? Computer Holography & Digital Hologra

"...but what is that technique called where you would record an interference pattern with a CCD camera? "
I don't believe you can record holographic fringes with a CCD camera because a CCD can't take the frequencies. A typical hologram shot at 30 degrees between beams with a HeNe needs a resolution of 800 lines/mm, approx 1.2 microns, much too small to be captured by CCD. However, there is a technique of holographic interferometry or, something similar, shearography ( a sort of reference-less holographic interferometry). In these cases, you shoot a hologram of an object, distort the objet by a very small amount and shoot it again. The small distortion causes a small difference in the object waveform and the hologram shows both objects superimposed but with interference lines across it. These lines are of the order of about a mm or so (depending on the distortion) and can be captured by a CCD. The theory is that by applying a small pressure on the object, it will distort in such a manner that the weakest point will distort slightly more than the rest of the object and you can find possible fracture modes by studying the resulting lines on the so-called interferogram. This analysis is sometimes done by capturing the interference lines on a CCD and transferring them to an analysis software. This used to be a common method of testing airplane tyres, I'm told. The related technique of shearography is also used to test for debonding underneath a surface and for delamination. The exact nature of this delamination/debonding is done by making the double hologram and feeding the lines into the computer via a CCD camera. Here's the method used to examine fluid flow: http://www.physics.uq.edu.au/lp/lasdiag/hi.html. References for CGH's usually occur in books. There's a good explanation of it in Hariharan's book:http://www.amazon.co.uk/exec/obidos/ASI ... 56-6641413
I don't know of any specific book that covers CGH, but you might find one here: http://www.rbookshop.com/engineering/h/Holography/
The SPIE bookshop usually has books on the subject:http://spie.org/app/Publications/

Dinesh

### Is there a difference? Computer Holography & Digital Hologra

I think these days they just image directly onto the medium. The Ft is computed for a view over a pixel area and that small area is recorded with phase and amplitude information.
In the 60's they used what was called a binary detour-phase method. A small area was chosen and a smaller square was recorded. The density of the square represented the amplitude of the FT coefficant while the position of the smaller square inside the larger area gave phase information. The idea was that the position inside the square determined path length and so could code for phase.
Brown & Lohman:
"Complex spatial filtering with binary masks" Applied Optics 1966, 6, 588-9
"Computer Generated Binary Holograms" IBM Journal of R & D 1969, 13, 160-7
Then they had lots of little squares:
Lee W. H. "Sampled Fourier Transform Hologram generated by computer" Applied Optics 1970, 9, 639-43
Then they had the "Kinoform". In this case, they realised that the amplitude information was not as important as the phase information. So they simply had a transparent medium but each cell (pixel) was varied in thickness so each cell had a unit transmision but a varying phase delay:
Kermisch D. "Image reconstruction from phase information only" JOSA 1970, 60, 15-17
MIT was pretty active in this area with Mike Klug.
If you can persuade the people at Zebra to tell you how they do it, you'll get the latest techniques.