high performance computing on graphics processing units: hgpu.org

Real-time rendering of complex 3D models is still a very challenging task. Recently, many GPU-based level-of-detail (LOD) algorithms have been proposed to decrease the complexity of 3D models in a parallel fashion. However, LOD approaches alone are not sufficient to reduce the amount of geometry data for interactive rendering of massive scale models. Visibility-based culling, especially occlusion culling, has to be introduced to the rendering pipeline for large models. In this paper, we aim to tackle the challenge of integrated parallel processing for both mesh simplification and occlusion culling. We present a novel rendering approach that seamlessly integrates parallel LOD algorithm, parallel occlusion culling and Out-of-Core method in a unified scheme towards GPU architectures. The result shows the parallel occlusion culling significantly reduces the required complexity of the 3D model and increases the rendering performance.