Hello everyone,
I am looking to implement 3D collision detection with concave 3D shapes. Ideally, the implementation/algorithm would be as optimal and as quick as possible. The collision detection for my program is between a cylindrical tube robot, which can assume concave shapes, and an STL model of the cavity inside an ear.
I have already done some research into the GJK algorithm, which only works for convex shapes. One solution to this problem is breaking up the concave shape into multiple convex shapes that represent the concave shape as a whole. I have also found a number of other implementation that use the triangular faces of the models and a ray casting algorithm to detect intersection.
My overall question is: What kind of algorithm would work well for this application? I would also like to hear any relevant experiences you might have had with implementing 3D collision detection in general.
