A fundamental task in computer graphics is the generation of images. Given a geometric model of an object or scene the user wants a visual representation for a given viewpoint. This visualization process is also known as rendering images of a model.
An active research area in rendering is the generation of images with high visual realism. Important visual cues like highlights on surfaces and shadows cast by light sources aid the viewer in understanding an image. Images with high realism are produced with global illumination methods, which approximate the light distribution in an environment to varying degrees.
The major challenge in real-time rendering is to generate images as quickly as possible. Frame rates higher than 20 frames per second are needed to ensure that users perceive smooth motion. Although image generation hardware is widely available, it is hard to visualize very large structures at interactive rates even on high-end platforms. Furthermore, model sizes are growing quicker than graphics hardware capabilities. Several classes of techniques for optimized rendering have been devised to achieve high frame rates.
This text places most publications of the author in the context of these two research areas in computer graphics.