AUTHOR: Thomas Diewald
Date: 07.02.2013 last edited: 07.02.2013
################################################################################
source code of my first experiments with space-partitioning techniques:
it was a study to compare performance and usage for raytracing/pathtracing. most important for me was the traversing-performance for ray-triangle-intersection, which i needed for the opencl-implementation. in the end i decided to use the BVH for that, and so the octree functionality might be not that "perfect".
a lot of things should/could be improved during building the data-structure. at the moment it is way to slow for dynamic changes in the scene. if you notice any errors, feel free to contact me.
for more information, i'd like to refer to my blog-post: http://thomasdiewald.com/blog/?p=1488
thomas