Efficient Multiple Occlusion Queries for Scene Graph Systems

Image space occlusion culling is an useful approach to reduce the rendering load of large polygonal models. Like most large model techniques, it trades overhead costs with the rendering costs of the possibly occluded geometry. Meanwhile, modern graphics hardware supports occlusion culling. Unfortunately these hardware extensions consume fillrate and latency costs. In this paper, we propose a new technique for scene graph traversal optimized for efficient use of occlusion queries. Our approach uses several Occupancy Maps to organize the scene graph traversal. During traversal hierarchical occlusion culling, view frustrum culling and rendering is performed. The occlusion information is efficiently determined by asynchronous multiple occlusion queries with hardware-supported query functionality. To avoid redundant results, we arrange these multiple occlusion queries according to the information of several Occupancy Maps. Our presented technique is conservative and benefits from a partial depth order of the geometry.