This paper describes and demonstrates the effectiveness of several metrics for data level comparison of direct volume rendering (DVR) algorithms. The focus is not on speed ups from approximations or implementations with parallel or specialized hardware, but rather on means for comparing resulting images. However, unlike image level comparisons, where the starting point is 2D images, the main distinction of data level comparison is the use of intermediate 3D information to produce the individual pixel values during the rendering process. In addition to identifying the location and extent of differences in DVR images, these data level comparisons allow us to explain why these differences arise from different DVR algorithms. Because of the rich variety of DVR algorithms, finding a common framework for developing data level comparison metrics is one of the main challenges and contribution of this paper. In this paper, we report on how ray tracing can be used as a common framework for comparing a class of DVR algorithms. While this paper focuses on comparing different DVR algorithms, we believe that similar metrics and comparison techniques are also useful for volumetric data comparisons. For example, comparison of experimental versus simulated data sets, or forecasted versus observed data sets, etc.
Ray sampling using cell face intersection. (23k)
Ray sampling at regular intervals. (22k)
Ray sampling using cell face intersections (27k),
VRML 2.0 file (360k)
Ray sampling at regular intervals (10k),
VRML 2.0 file (355k)
Return to AVIS home page.