Rendering Transitions in a Geometric Level of Detail Framework

In computer graphics, the geometric complexity of models increases with time. As scanner and other hardware technologies improve, individual models may contain billions of polygons. Interactively rendering a single polygonal model of this size is not possible with current hardware. Thus for complex scenes that contain many of these models, rendering slows dramatically. One rendering technique, called displacement mapping, introduces even more polygons into a scene. This dissertation develops a novel technique to enhance rendering speeds of displaced models. By utilizing view-dependent algorithms and appearance preserving techniques, interactively rendering complex scenes with many displaced polygons is possible. Focusing on the user’s view of a scene facilitates the removal of large amounts of irrelevant data. This data removal allows efficient rendering of displaced surfaces.