One of the most computationally intensive processes in graphical displays is ray tracing, where the intersection between a line of sight (or a light source) and a 3D surface (represented by a large database of triangles) is calculated. In real time rendered images (e.g., video games) this is so much so that only primary ray tracing is performed. Global illumination, where diffuse light sources and multiple reflections are calculated, are only performed for high resolution movie quality, as each image can take hours to render. The more dynamic the image, the more computationally intensive it will be.
This invention describes how to identify the intersection in a much more efficient manner, so it can introduce the concept of movie quality for real time rendered images. It is also deterministic with respect to the complexity of the calculation, allowing exact memory requirements to be defined and, being faster; it is less power hungry and therefore suitable for mobile applications. The algorithm is available as a library in a format compatible with OpenGL or DirectX, and is eminently suitable for committing to hardware as a graphics accelerator. A lower level in memory usage is highly advantageous to the computer systems that create 3D graphics as internal memory storage is limited and of critical importance. This newly patented software process also promises to allow for the production of real time 3D images for scenes in computer games and films that are comparable in quality to that of static 3D images.
This novel system operates at an extremely fast rate which makes it highly suitable for mobile applications which can be viewed from smart phones and tablets. The manufacture of 3D games, films and architecture applications will ultimately improve from this interactive technology.
A US patent has been granted to Swansea University under US 8,797,324.