Movie #1 (564k)
Movie #2 (2.47M)
Movie #3 (2.31M)
Movie #4 (668k)
Movie #5 (629k)
Movie #6 (327k)


The extension of LIC to volume flows is pretty straightforward. The major difference lies in the use of volume rendering in the pipeline (Figure 1) to display a 3D scalar texture of flow. Although volume rendering has been widely adopted for medical data visualization, its application to volume flows, in particular dense 3D LIC textures, is still limited due to view occlusion, depth ambiguity, direction vagueness, and lack of any physical meaning of a LIC texture value. Without effective transfer function design, flow patterns just could not be revealed out of a dense volume. In general, sparse noise is advantageous over white noise in addressing the aforementioned perceptual issues (Figure 2).
Figure 1. The pipeline of Volume LIC (© Zhanping Liu).
(a) white noise used
(b) sparse noise used
(c) sparse noise used
Figure 2. VolumeLIC images.
Line Integral Convolution
Flow Visualization