Synthetic 3D Imaging - Potential Projects List

Your choice of project depends on curiosity, research interests, skills, ideas, resources available, and other practical issues. Anything is possible (if I can help you find a solution.)

Volumetric display
Build a simple volumetric display using whatever hardware you can find. A few points in space is convincing enough. There are a dozen different ways to do this.

Computational holography
Compute holographic fringes and print them out on film using a high-resolution printer, e.g., a 2540 dpi printer. There are many issues to explore: the limitations on image volume and viewing angle, algoithm comparisons, methods for binarization, speed of computing, effectiveness of different types of images.

Holographic imaging
Make some optical holograms. This will require access to a suitably equiped lab with a laser and optics. (I have some film.)

Comparison of depth cues
Compare the relative strengths of depth cues by doing simple experiments on a number of human subjects. This can employ any 3D imaging technique that you have access to, and can also focus on pictorial depth cues.

Web-based hologram formula solver
Using Java or CGI scripts, write an interactive tool for solving the two main equations in holographic imaging: the grating or "sine-theta" equation and the "one-over-R" equation. The optical recording step and the optical reconstruction step are related by these two equations. For a more advanced treatment, make an interactive graphical user interface in which the user can click and drag a source (e.g., object, reference, illumination) and see the resulting effect on the rest of the two-step system.

The earlier list:

3D image relay system
Using two fresnel lenses, create a very realistic 3D image from a real object. This helps us to understand the limitations of a synthetic 3D display. (I can lend the lenses.)

Animated or interactive stereoscope
Displayed on a CRT; viewed with a stereoscope.

LCD plus parallax barrier = autostereoscopic display
Add depth to that laptop or LCTV.

Lenticular hardcopy
Lenticular screens solve the dimness problem that plagues parallax barriers.

LCD plus lenticular screen = autostereoscopic display
Brighter is better.

Interactive test of disparity budget
Using a stereoscope, interactively render stereo pairs to test the limits of the disparity budget, the positioning of cameras, etc.

Interactive or animated anaglyphic display
Use the red-blue glasses that you all have to make an interactive 3D display system.

Two-view renderer
Demonstrate a "speed-hack" or two that increase rendering speed by exploiting the similarities between left and right views.

Shear camera rendering
Implement a shear camera, and compare its performance to the "toed-in" geometry that many have been using.

Interactive Pulfrich effect 3D
Simple but interesting 3D effect that exploits the time-delay associated with reduced brightness in one eye (see the Superbowl halftime in 1989?). Simple create animations (or interactive scenes) in which object rotations and/or lateral camera moves cause objects to appear to have depth, given that the viewer is wearing easy-to-make Pulfrich glass (basically, cheap sunglasses missing one of the lenses!).

Autostereoscopic display
Build one of the simpler display systems described in the literature.

A really great parallax barrier hardcopy
See PS5 and PS6.

Any 3D-related idea that you have.
Be creative!


Synthetic 3D Imaging / Mark Lucente / copyright 1996