Dynamic primitive caching for haptic rendering of large-scale models

In this paper we present a software approach to managing complexity for haptic rendering of large-scale geometric models, consisting of tens to hundreds of thousands of distinct geometric primitives. A secondary client-side caching mechanism, exploiting partitioning, is used to dynamically update geometry within the locality of a user. Results show that the caching mechanism performs well, and that graphical rendering performance becomes an issue before the caching mechanism fails. The performance penalty of the caching technique was found to be dependent on the type of partitioning method employed.