I thought this would never end. Finally, on November 04, 2004 I became an engineer (at least qualified to be one).
The TFC was about stereo reconstruction, with a slight twist. An interface (for example air/water) was added between the cameras and the scenery. Snell's law had to be dealt with. As a "bonus", a way of estimating the interface's surface was developed.
You can have a look at the TFC Description Page (Local cache - Portuguese)
Thesis This is a slightly corrected version from the one turned in. For one thing a latex command that didn't work correctly with latex2html was circumvented. The content is exactly the same though. Please note that I received a few negative comments on my writing style. Beware. :)
Code (not yet available). I hope to clean the matlab and C code and make it available here someday. Meanwhile if anyone is interested, drop me a note.
Computer Vision Test Images - Many links to other sites.
CMU VASC Image Database - From the page: "The goal of the VASC Image Database is to share image data sets with researcheres around the world. To facilitate this, we have created this site, which contains over 5000 images split up over nearly 200 different data sets."