Hi!
I'm using SAT, and swept collision, only that in the next frame when there will be a collision, for both moving objects with their angular velocity, I'm not able to use angular swept(only linear); how do we do that? so how we can calculate the time of impact? never seen someone did until now!
continuous collision detection(point of collision;
-
- Posts: 20
- Joined: Wed Sep 02, 2015 10:00 pm
continuous collision detection(point of collision;
Last edited by Isawyou on Wed Sep 16, 2015 5:11 pm, edited 1 time in total.
-
- Posts: 197
- Joined: Sat Aug 19, 2006 11:52 pm
Re: continuous collision detection(point of collision;
Erin Catto's GDC2013 presentation on "bilateral advancement" might be worth checking out: http://box2d.org/files/GDC2013/ErinCatto_GDC2013.zip
(he uses distance queries rather than SAT, but maybe the general idea is useful..)
(he uses distance queries rather than SAT, but maybe the general idea is useful..)
-
- Posts: 20
- Joined: Wed Sep 02, 2015 10:00 pm
Re: continuous collision detection(point of collision;
it's too hairy, I found where when it is a linear velocity.(t=relativedist/relavel);
sampling is used to calculate the right teta to impact. is there any other way. I make my teta too small and a small timer, :/ that not genuis;(I saw this http://www.wildbunny.co.uk/blog/2011/04 ... r-dummies/;)
making object rotate due gjk and detect time of collision using CCD(is that sampling), how he've done this,more explanations??
sampling is used to calculate the right teta to impact. is there any other way. I make my teta too small and a small timer, :/ that not genuis;(I saw this http://www.wildbunny.co.uk/blog/2011/04 ... r-dummies/;)
making object rotate due gjk and detect time of collision using CCD(is that sampling), how he've done this,more explanations??