Abstract:
In medical visualization,transfer function design is an efficient means to emphasize the interested tissues and organs while suppressing other regions.It is challenging,however,to get an appropriate transfer function for characterizing arbitrary shapes.This paper introduces a shape-based transfer function design scheme that highlights similarly shaped structures using pre-determined shape features such as 3D moment invariants,fractal features and linear features.We verified our approach on several medical datasets to show its efficiency and convenience.