On the subject of underappreciated GPU vector graphics papers: If you take the basic idea behind Valve-style SDFs for fonts—encode geometric information in the texture—and push it as far as you can, you get “Silhouette maps”, Sen 2004: https://graphics.stanford.edu/papers/silmap_tex/ …
-
-
Anyway, I think something like silhouette maps is probably the most promising approach for static, baked vector graphics assets in 3D (including VR). For dynamic vector graphics, different story, of course—that’s where Pathfinder’s approach shines.
Show this threadThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I assume you know the existing work in this area? E.g. vector textures based on biarc splines are my favorite variation: https://cs.uwaterloo.ca/research/tr/2007/CS-2007-41.pdf …
-
The most annoying issue with vector textures has always seemed to be minification. You end up having to do a nontrivial geometric simplification problem unless you're okay with an insane per-pixel rendering cost.
- 2 more replies
New conversation -
-
-
You can extend/generalize a voronoi diagram (constrained by a rectangular modulo lattice for peformance), by not having one DistanceToPoint per rectangle, but by having DistanceToQuadraticSpline2d per rectangle, and have 4 rectangles overlap with 50% kerning over 2 domains.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Google image search "gereralized voronoi" and you may just end up in the foldAndCut theorem with motorcycleGraphs , and all its open questions for some lacking general proofs.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.