LatLIC

DESCRIPTION

This module implements the line integral convolution algorithm for visualizing 2D vector fields. The result is a one-band grayscale image, which can be displayed directly by DisplayImg.

INPUTS

Port: Vector Field
Type: Lattice
Constraints: 2-D.
2-vector.
float.
uniform.
The vector field to be visualized.

Port: Tolerance
Type: Parameter
Optional: This port is optional
Specifies accuray of streamline integration.

Port: hTry
Type: Parameter
Optional: This port is optional
Initial stepsize used by the adaptive integrators.

Port: hMin
Type: Parameter
Optional: This port is optional
Minimum allowed stepsize of an adaptive integrator.

Port: hMax
Type: Parameter
Optional: This port is optional
Maximum allowed stepsize of an adaptive integrator.

Port: Seed
Type: Parameter
Optional: This port is optional
Seed of random number generator used for computing random texture.

WIDGETS

Port: Integrator
Type: Option Menu
Menu Item: Cash-Karp
Menu Item: Dormand-Prince
Menu Item: RK4(3)
Menu Item: Euler
Menu Item: Midpoint
Specifies type of streamline integrator. Cash-Karp, Dormand-Prince, and RK4(3) are Runge-Kutta integrators with adaptive stepsize control and error monitoring. Euler and Midpoint simple fixed size methods.

Port: L Filter
Type: Slider
Specifies length of convolution filter (in one direction). The number corresponds to pixel lengths in the output image.

Port: L Streamline
Type: Slider
Maximum distance used for tracing and updating convolution integral along a streamline. This parameter might be changed to optimize the performance of the module.

Port: Scale
Type: Slider
Scaling factor applied to the vector field. Might be used to get large images from a low-res vector field.

OUTPUTS

Port: Output Image
Type: Lattice
Constraints: 2-D.
byte.
uniform.
The final LIC image.

PROBLEMS

The random texture used for convolution cannot be scaled or changed by the user in this version of the module.

SEE ALSO

Literature: "Fast and Resolution Independent Line Integral Convolution" Detlev Stalling and Hans-Christian Hege Proceeding of SIGGRAPH '95 pp. 249-256
[ Documentation Home ]
© The Numerical Algorithms Group Ltd, Oxford UK. 1996