1
\$\begingroup\$

I have heard of techniques such as overlap testing and intersection testing for collision detection and collision resolution.

I looked up on the two techniques, but I don't see what differences they have. On every frame, the overlap testing will check to see if it overlaps another object by looking if any vertices is inside another object's bounding volume. The intersection testing on the other hand casts a line from its centre point to check for intersection. But since it's only a line, it may not be as accurate. So it has to cast many lines continuously from all over its edges in its moving direction and, hence, is more costly. Is this true and is the only difference between the two techniques?

Otherwise, what are the differences between the two techniques? Also, are there any differences in terms of collision resolution between the 2 techniques/

\$\endgroup\$

1 Answer 1

1
\$\begingroup\$

The terms overlap testing and intersection testing are largely synonyms. Neither does without context signify a specific method. Collision detection could also be said to be a synonym, but therm might in some cases be bit broader.

I'd describe these terms as the art of determining whether or not two figures in any place occupy the same space. Depending on the context this may also include finding the bounds of that space or finding the point where they first touched.

Collision resolution is what comes after collision detection, it is the art of determining how the collision should influence the objects. Usually this means changing the velocities and spin of the objects according to Newtonian rules of physics, but it could also mean breaking the objects or applying some alternate physics rules that fit the gameplay better.

\$\endgroup\$
2
  • \$\begingroup\$ But technically in terms of implementation, aren't there differences at all between overlap testing and intersection testing? \$\endgroup\$
    – xenon
    Apr 28, 2013 at 16:17
  • \$\begingroup\$ @xenon There are no difference between the terms, but both terms cover countless different algorithms. It is just like the words car and automobile, they describe the same group of devices, but that group is very differentiated with many different models serving different purposes. \$\endgroup\$ Apr 28, 2013 at 16:28

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .