Post
by Dinoguy1000 » Mon Aug 06, 2012 10:04 pm
That's a case of skinning an existing model. As long as the model is carefully designed, and the algorithm to match the texture to the model is clever enough, you'll get passable results.
finale00 was asking about taking a picture and generating a (high-quality) model from it, which is another matter entirely. Even if there were software to do it (and, outside of security/government applications, I don't know if there is), it wouldn't generate very high-resolution models, and it would have to extrapolate (read: guess, based on averages) for the areas that aren't in the picture. This, of course, assumes that you're talking specifically about generating human models. If you broaden the question to generating high-res models of arbitrary objects, that requires strong AI.
Welcome to Xentax!
Rules |
Requests |
Wiki |
Discord
If you run across a post that breaks the rules, please
report the post - a mod or admin will handle it from there.