Hello,
I want to implement a collision detection routine. In the narrow-phase, I will detect collision between 3D cubic curve (with thickness) and polyhedron (convex and in future, concave). -- (I will use V-clip as my fine CD alg.)
My questions:
1- Generally (and in my case), How to detect collision between 3D cubic curves? and between 3D cubic curve and a polyhedron? [in the Narrowphase]
2- Should I simplify the curve geometry (to make a collision geometry)? (the interpolation used is the Hermite interpolation, in case it will have effect!)
-- a. I mean by simplify: Subdivide the curve
-- b. After subdividing the curve -> use OBBTree or AABBs as a hierarchical structure for the collision geometry.
---> Is that a right approach? any more accurate ones? any other better methods/algorithms?
Please, guide/hint me to any papers/algorithms/approaches.
(Feel free to discuss/advice me about anything)
Thanks!
Fine (Narrowphase) Collision Detection with 3D cubic curves
-
- Posts: 6
- Joined: Mon Nov 29, 2010 3:42 pm
- Location: Mother Earth, Milky Way Galaxy!
-
- Posts: 6
- Joined: Mon Nov 29, 2010 3:42 pm
- Location: Mother Earth, Milky Way Galaxy!
Re: Fine (Narrowphase) Collision Detection with 3D cubic cur
I really need some thoughts here! anyone?
-
- Posts: 231
- Joined: Tue Feb 20, 2007 4:56 pm
Re: Fine (Narrowphase) Collision Detection with 3D cubic cur
1) I don't know what approach is generally used for that type of collision. I have used quadric curved surfaces as height maps (with only vertical feelers), but that's much easier than the general 3-D problem, because colliding a curved surface vs. a triangle can result in multiple contact points.
2a) Even without knowing your accuracy & performance requirements, my guess is that you'll want to simplify the collision geometry into some sort of polygon-based format.
2b) Those choices for bounding volume hierarchies are probably fine (I use AABBs), just understand the tradeoffs when you make your choice. OBBs provide tighter fits but take more storage. If you haven't picked up Ericson's Real-Time Collision Detection book, you probably should.
2a) Even without knowing your accuracy & performance requirements, my guess is that you'll want to simplify the collision geometry into some sort of polygon-based format.
2b) Those choices for bounding volume hierarchies are probably fine (I use AABBs), just understand the tradeoffs when you make your choice. OBBs provide tighter fits but take more storage. If you haven't picked up Ericson's Real-Time Collision Detection book, you probably should.
-
- Posts: 26
- Joined: Mon Jun 01, 2009 2:21 pm
- Location: Sweden
Re: Fine (Narrowphase) Collision Detection with 3D cubic cur
1) The only way I can think of is to subdivide your cubic curves into several convex objects.
If there exist a general method for convex polyhedron vs concave surfaces I have never heard of it.
2) Like I wrote in 1 you should subdivide it and for efficiency it would probably be good to put some bounding volume on top.
If there exist a general method for convex polyhedron vs concave surfaces I have never heard of it.
2) Like I wrote in 1 you should subdivide it and for efficiency it would probably be good to put some bounding volume on top.