Musings on Volumetric Level of Detail for Virtual Environmentsy

This paper considers the extension of a real-time graphics renderer to support fovea enhancement|the technique of localising visual detail to the particular region of the display which the user is looking towards. This is performed by extending the standard notion of distance Level of Detail (LOD) into three dimensions to give volu-metric LOD; whereby the LOD of an object is related to its presence within a three-dimensional volume which is aligned with the user's gaze. Before introducing this technique , some background details are discussed regarding the relevant characteristics of the human visual system and current solutions for eeective gaze tracking. Subsequently, a brief cross-section of relevant research is presented and a conceptual model of volumetric LOD is formulated. Finally , implementation factors for such a system are considered and a theoretical evaluation is prooered.