Many applications in Computer Graphics require an adequate use of collision detection to deal with the interaction between objects. Incorporating deformable objects in dynamic environments requires the handling of different parameters and constraints to take into account the speed and time response to permit acceptable simulations. In this research physically-based deformable objects and collision detection are studied.

rabbit pictureTwo major computational bottlenecks in many animation systems are the handling of collisions between the objects under the influence of forces (gravity, user input, etc.), and the self-collision detection which tests the object itself to find colliding regions. Accurate self-collision detection is challenging to perform in real time because many adjacent or nearby primitives of a deforming mesh are always in close proximity. Some models suffer large deformations which consequently causes self-collisions. The mesh is deformed in such a way that polygons can touch other polygons during the deformation. In this work 3D objects are considered and a hierarchy of spheres is utilised to approximate the shape of the objects. Spheres are frequently used in computer graphics as approximations to objects mainly because they are rotationally invariant.

Research Team

F. Madera, Dr. S. D. Laycock, Prof. A.M. Day