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!
-
-