Composite rigid body construction for fast and compact dynamic data compression

Compression of 3D dynamic datasets in remote visualization still remains two challenges. One is low time performance due to the grown data and complex computation of compression algorithm. Another is small compression factor because of dynamic scenes without known equations of their motions. In this paper, we propose a fast and compact compression for 3D dynamic datasets. It accelerates compression with KD-tree construction and node-grid mapping for the dynamic data, which allow parallel rigid body decomposition and merging with disjoint union method. To increase the compression factor, composite rigid body is introduced with consideration of temporary motion consistency among rigid bodies. The results of the experiments show that our algorithm can compress dynamic datasets quickly and obtain high compression factor to reduce limitation of bandwidth.