Issues on Displaying 3D Data for Scientific Visualization Thomas Theußl theussl@cg.tuwien.ac.at Institute of Computer Graphics Vienna University of Technology Vienna, Austria

Ray-casting Vector Fields

Ray-casting [8] is a direct volume rendering method, where rays are cast into a 3D array of values and for each ray a vector of sample colors and opacities is computed by re-sampling the voxel database at k evenly spaced locations.
When ray-casting a vector field [3], there is exactly one stream line crossing the viewing ray at each sampling point. To shade the sampling color accordingly, there is again the problem that lines in 3D do not have a unique normal vector. Since there is one plane to which the local velocity vector is normal, the vector which points mostly towards the viewer while lying in this plane is chosen as the normal vector. If the light source is located at the viewing position, the shading computation is simplified and flow towards or away from the viewer appears dark and flow normal to the viewing direction appears brightest. However, this approach leads to visualizations that are difficult to interpret if only a still image is used (see Figure 2 on the left). The spatial orientation of the flow is much better perceived if viewpoint animation is used.
To enhance the spatial impression in single images, it makes sense to use pseudo-color to express velocity magnitude (see Figure 2 on the right) or even vector field direction. Therefore, the angle between the streamline normal and the light source is mapped to sample hue, so we can see where the flow is directed towards or away from the light source, i.e., the viewing direction here. The use of pseudo-color yet for the visualization of vector field direction is something to which the user first has to become accustomed to.

  
Figure 2: Ray-casted vector fields. On the left streamline shading without color mapping, on the right color is mapped to velocity magnitude. Light source in both pictures at viewing position.