Distributed Global Illumination Method Based on Photon Mapping

Abstract

As an excellent global illumination algorithm, photon mapping requires a large amount of memory to store photons, and it has obvious spatial locality in the step of photon searching. In this paper, we propose a distributed photon mapping algorithm. Firstly, a new delivering strategy of rendering tasks is proposed based on the ray intersection, while path tracing is used to calculate the direct illumination. Before the rendering, our work automatically synchronizes the acceleration of geometries and photons to realize data partitioning. Secondly, an overlapped KD-Tree is proposed to store photons in order to solve the problem of boundary data merging. The method of this paper is tested using multiple scenes. With ensuring good rendering results, both the rendering speed and storage load obtain the linearly optimization.