采用GPU的遮挡查询功能提出了一种复杂场景的层次遮挡剔除算法,通过交替进行遮挡查询和可见节点的绘制,有效地减少了由于遮挡查询延迟造成的空闲等待时间.为了减少场景中不必要的遮挡测试,将遮挡查询问题描述为最优化决策问题,通过对每一帧遮挡查询的选择进行优化,能够使整个场景绘制的效率近似达到最优.实验结果表明,对于不同复杂度的场景,该算法可以明显地提高场景的绘制速度.
This paper proposes an algorithm for hierarchical occlusion culling using hardware occlusion queries. The algorithm can reduce the idle time of the CPU and GPU due to the latency of occlusion queries by interleaving occlusion queries with rendering of visible objects. Furthermore, in order to minimize the number of unnecessary occlusion queries, we formulate the occlusion queries problem as an optimal decisionmaking and the overall performance is optimized by performing only occlusion queries for objects that are likely to be hidden. The experimental results show that the algorithm can significantly increase rendering speed for arbitrary complex scenes.