Splatting Optimizations

Authors: 

Steven Kilthau

Splatting is a method for visualizing point-sampled volume datasets. We present a highly optimized implementation of the splatting algorithm that achieves near real-time rendering rates.

Our optimizations are achieved through numerical approximation techniques, use of triangular shaped splats, the exploitation of current hardware OpenGL extensions, as well as an optimized data structure based on run length encoding. The strength of this work is an exhaustive analysis of the bottlenecks of the splatting algorithm, and our results should therefore be easy to adapt to any specific rendering platform. Using the optimizations presented, we gain speedups on the order of 5-10 times over traditional, recently reported implementations.