Collision detection for rigid superellipsoids using the normal parameterization

The normal parameterization as an approach to describe geometries is introduced. The advantages of this description – as compared to other parameterizations or implicit functions – in the context of collision detection are: the possibility to explicitly calculate axis aligned bounding boxes for any convex geometry, an efficient iterative algorithm for collision detection between objects with arbitrary geometry that does not require any (analytical) derivatives. A system of several rigid superellipsoids is used to demonstrate the application and performance of the proposed approach in a multibody simulation.