Enhanced Texture Advection Algorithm

While texture-based methods are used to visualize 3D vector fields, one major issue is the perception of the resulting images that might suffer from problems of low contrast and rough-textured flow pattern. In order to enhance the 3D vector field visualization, this paper presents a high-quality texture advection algorithm. Pseudo-gradients are calculated with the position relation between the current noise point and its next neighbors, and the maximum descending degree is selected to participate in the texture advection. The weight of the noise texture is adjusted adaptively according to the contribution from every advection position. Higher-order filter kernel is obtained by recursively convolving box-type filters. The output texture is produced by convolving texels along the calculated particle tracing trajectory. The validity of the proposed algorithm is examined with the objective quantization function on visual contrast. Experiment results demonstrate that this algorithm can effectively increase the contrast between the advection trajectories, improve the rendering effect, and achieve high-quality visualization.