exercicio: Noir rendering vs. Contour rendering
|Reported by:||ricardofabbri||Owned by:||wakkuu|
On a blog post I mentioned a curious fact that many artistic videos are rendered only in two colors (black and white) while still giving a nearly perfect notion of depth. This is called Noir rendering:
(read also the comments in there)
The goal of this exercise is to see how the video would look if only the edges between black and white were shown. You must generate a video just like the one from the Crystal Method (or any similarly rendered video you like best), but showing the contours. I would call that "Contour rendering".
Before and after you generate the video, ask yourself: can the contours represent the scene just as well as a Noir rendering?
The steps to do this would be
- first learn how to do this for a single image in Scilab
- then do this for a video, perhaps in C/C++
- generate a final movie with sync'd audio and everything.
Fefo: would you like to take on this exercise? If so, the other two can start on the airhacktable exercise and you can join them afterwards.
Change History (6)
comment:1 Changed 6 years ago by ricardofabbri
- Component changed from #gt-default to #gt-video
- Milestone changed from default-end-of-august to learn-cpp-vision
comment:4 Changed 6 years ago by ricardofabbri
- Resolution done deleted
- Status changed from closed to reopened