Mesh Displacement: An Improved Contouring Method for Trivariate Data

Trivariate data arise in a wide range of industrial and medical applications. Some of the most powerful techniques for visualizing these data involve computing piece-wise linear approximations to isocontours of the data WMW86, LC87]. This paper describes a simple optimization that may be applied to the piecewise linear approximations produced by these techniques. The beneets of this optimization include: Reducing the number of triangles in the resulting approximation by 40% to 50% without a signiicant loss of accuracy in the approximation. Improving the shape of the remaining triangles while eliminating badly shaped triangles produced by current methods. Enabling simple computation of the topology of the resulting mesh.