ClFastMarching Class Reference

#include <ImFusion/CL/ClFastMarching.h>

OpenCL Fast Marching implementation. More...

Detailed Description

OpenCL Fast Marching implementation.

Public Member Functions
	ClFastMarching (int mode=ClFastMarching2D, ClEnvironment *env=0)
	Creates an OpenCL based FastMarching solver for a specific dimension type.
void	compute (float *phi, int w, int h, const std::vector< vec2 > &startPoints, float *F=0, int maxIter=5000, const vec2 &endPoint=vec2(-1, -1), float maxDist=std::numeric_limits< float >::infinity())
	2D Fast Marching Method for solving the Eikonal equation ||Grad(Phi)||*F = 1.
void	compute (float *phi, int dim[3], const std::vector< vec3 > &startPoints, float *F=0, int maxIter=5000, const vec3 &endPoint=vec3(-1, -1, -1), float maxDist=std::numeric_limits< float >::infinity())
	3D Fast Marching Method for solving the Eikonal equation ||Grad(Phi)||*F = 1.

Static Public Member Functions
static const float	ClMaxFloat ()
	Return maximum float used by OpenCL.

Constructor & Destructor Documentation

ClFastMarching()

ClFastMarching	(	int	mode = ClFastMarching2D,
		ClEnvironment *	env = 0 )

Creates an OpenCL based FastMarching solver for a specific dimension type.

Parameters

mode	gives the dimension type: ClFastMarching2D (default) or ClFastMarching3D

Member Function Documentation

compute() [1/2]

void compute	(	float *	phi,
		int	w,
		int	h,
		const std::vector< vec2 > &	startPoints,
		float *	F = 0,
		int	maxIter = 5000,
		const vec2 &	endPoint = vec2(-1, -1),
		float	maxDist = std::numeric_limits< float >::infinity() )

2D Fast Marching Method for solving the Eikonal equation ||Grad(Phi)||*F = 1.

For F = 1 this produces the distance map of the zero level set. See Osher et al. "Level Set Methods and Dynamic Implicit Surfaces" pp. 69 for details. (If you want to call this method, m_mode has to be set to CLFastMarching2D)

Parameters

phi	level set function
w	level set width
h	level set height
startPoints	vector of start points
F	velocity, if set to zero a constant velocity of one is assumed
maxIter	max iteration count
endPoint	NOT used yet!
maxDist	NOT used yet!

compute() [2/2]

void compute	(	float *	phi,
		int	dim[3],
		const std::vector< vec3 > &	startPoints,
		float *	F = 0,
		int	maxIter = 5000,
		const vec3 &	endPoint = vec3(-1, -1, -1),
		float	maxDist = std::numeric_limits< float >::infinity() )

3D Fast Marching Method for solving the Eikonal equation ||Grad(Phi)||*F = 1.

For F = 1 this produces the distance map of the zero level set. See Osher et al. "Level Set Methods and Dynamic Implicit Surfaces" pp. 69 for details. (If you want to call this method, m_mode has to be set to CLFastMarching3D)

Parameters

phi	level set function
dim	level set dimensions (dim[0] = width, dim[1] = height, dim[2] = depth)
startPoints	vector of start points
F	velocity, if set to zero a constant velocity of one is assumed
maxIter	max iteration count
endPoint	NOT used yet!
maxDist	NOT used yet!

---

The documentation for this class was generated from the following file:

• ImFusion/CL/ClFastMarching.h