An intuitive way to visualize a flow is to watch particles or textures move in the flow. In this paper, the authors show how texture mapping hardware can produce near-real-time texture motion, using a polygon grid, and one fixed texture. However, the authors make no attempt to indicate the flow direction in a still frame. As discussed here, any anisotropic stretching comes from the velocity gradient, not the velocity itself. The basic idea is to advect the texture by the flow field. In a cited paper, they gave an indication of the wind velocity by advecting the 3D texture coordinates on the polygon vertices of a cloudiness contour surface in a climate simulation. This was slow, because the 3D texture was rendered in software, and because advecting the texture was difficult for time-varying flows. In this paper, they replace the 3D textures by 2D texture maps compatible with hardware rendering, and give techniques for handling time-varying flows more efficiently. The next section gives their technique for the case of 2D steady flows, and the following one discusses the problems of texture distortion. Then they discuss the problems with extending method to time-varying flows, and two solutions. Next they develop compositing methods for visualizing 3D flows. The final section gives their results and conclusions.
[1]
Lisa K. Forssell.
Visualizing flow over curvilinear grid surfaces using line integral convolution
,
1994,
Proceedings Visualization '94.
[2]
Tom Duff,et al.
Compositing digital images
,
1984,
SIGGRAPH.
[3]
Richard Shoup.
Color table animation
,
1979,
SIGGRAPH '79.
[4]
Nelson Max,et al.
Texture splats for 3D vector and scalar field visualization
,
1993
.
[5]
Jarke J. van Wijk.
Flow visualization with surface particles
,
1993,
IEEE Computer Graphics and Applications.
[6]
Nelson L. Max,et al.
Visualizing wind velocities by advecting cloud textures
,
1992,
Proceedings Visualization '92.
[7]
Brian Cabral,et al.
Imaging vector fields using line integral convolution
,
1993,
SIGGRAPH.
[8]
David A. Lane.
UFAT-a particle tracer for time-dependent flow fields
,
1994,
Proceedings Visualization '94.
[9]
Hans-Christian Hege,et al.
Fast and resolution independent line integral convolution
,
1995,
SIGGRAPH.