Illustrating Dynamics of Time-Varying Volume Datasets in Static Images

Abstract

We present visualization techniques for timevarying volume datasets, i.e., time series of 3D data volumes. Instead of sequentially displaying the consecutive 3D volumes contained in such a time series, our method allows to depict the dynamics in a single static image. We extract the motion dynamics inherent to a time-varying volume dataset automatically during a preprocessing step by using three-dimensional block matching. The final rendering of a 3D volume associated with a specific point in time incorporates the extracted information to depict dynamics in a single static image by nonphotorealistic overlays in the form of glyphs and other graphical elements. In addition the preceding and following volumes of the time series may be rendered and combined with the final image.