Hierarchical min-max octree applying adaptive blocks

There are several optimization techniques available for improving rendering speed of direct volume rendering. An acceleration method using the hierarchical min-max octree requires little preprocessing and data storage while preserving image quality. However, this method introduces computational overhead because of unnecessary comparison and level shift between blocks. In this paper, we propose an efficient space-leaping method using an hierarchical min-max octree applying adaptive blocks. To determine the size of blocks, our method partitions an image plane into several uniform grids, and computes the minimum and the maximum depth values for each grid. We acquire different-sized blocks suitable for individual rays from on these values. Experimental results show that our method reduces rendering time when compared with the previous hierarchical min-max octree method.