A formal approach to quantitative image quality assessment in visualizing medical and scientific data is practically absent in the literature to date despite the abundance of approaches to optimizing the volume rendering algorithm in terms of quality and performance. A method to evaluate volume rendering artifacts related to an insufficient sampling rate of volume ray-casting (RC) algorithms is proposed. Different RC-quality optimization approaches are experimentally compared on the basis of the proposed assessment method and for different visualization cases the optimal ones are identified by the quality/performance criterion.
|
