A volume rendering algorithm using an analytic approach for curvilinear volumes

Abstract

A new approach is presented to visualize curvilinear volume data produced using the CFD or FEM Methods. Ray, casting methods are effective and renders images with good quality. However, these methods suffer from a tremendous number of computations and can not deal with the curvilinear volume efficiently. The marching cubes method is a popular approach when the ambiguity problem reduces the reliability of the rendered image. We propose a new method to solve these problems. Our method is a cell-by-cell process which constructs sosurfaces using trilinear functions. Several speed up techniques are proposed to reduce the number of computations in our algorithm. In addition to this, error analysis and error minimization are also discussed. Our method is also easy to implement parallelized.