A Parallel Volume Visualization Algorithm for Unstructured Grid

Volume rendering is one of the most important visualization methods for unstructured grid data. However, the existing serial volume rendering algorithms for unstructured meshes are not efficient enough to meet the needs of large-scale data volume visualization. In order to solve the performance bottleneck, this paper proposes a parallel volume rendering algorithm. Firstly, we design a parallel KD-tree algorithm to split volume data and track the transparent relationship. Secondly, each process uses an independent visual pipeline to calculate the volume rendering image. Finally, a tree synthesis strategy is used to synthesize the final image. Experimental results show that the algorithm can be efficiently applied to the visualization of large-scale unstructured grid data.