Options
2019
Conference Paper
Titel
Histogram-Based Nonlinear Transfer Function Edit and Fusion
Abstract
Volume visualization has wide application in science, engineering, biomedicine and other domains to help understand complex observational or simulative data. Transfer function is a traditional volume visualization approach, which is designed to assign different schemes of color and transparency for each voxel in volume data. In this paper, we design a histogram-based nonlinear transfer function editor. The design of nonlinear histogram and non-uniform grids in the background provides visual cues for users to edit transfer function more efficiently. The larger the histogram bin sizes are, the wider the bin widths will be drawn in the background. Then, a wavelet-like short transfer function (we call it TF-let) is designed, which can be serialized and reloaded fast in the subsequent explorations. Furthermore, we design a TF-let fusion approach to fuse multiple TF-lets by simply clicking the corresponding TF-let nodes. Compared with the traditional linear method, two evaluation tests show that the proposed approach is less sensitive and more efficient to edit the control points of the transfer function. Finally, use cases show the proposed approach is capable of achieving some hard-to-find tiny structures and visualizing them much more clearly.