Package gov.nih.mipav.model.algorithms.levelset

Class LseParameters

  • public class LseParameters
extends java.lang.Object

    This class is a simple wrapper for the parameters that occur in the level-set evolution, which is controlled by the partial differential equation:

        du/dt = -a*A(x).grad(u) - b*P(x)*|grad(u)| + c*Z(x)*K(u)*|grad(u)| + d*Laplacian(u)

    where u(x,t) is the evolved image at time t, du/dt is the time derivative of u, grad(u) is the gradient of u with respect to the x-variable, A(x) is the advection term with advection weight a, P(x) is the propagation term with propagation weight b, Z(x) is the curvature term with curvature weight c, K(u) is the mean curvature of level curves (in 2D) or surfaces (in 3D), and Laplacian(u) is the sum of second-order unmixed derivatives with Laplacian weight d. The notation A(x).grad(u) indicates the dot product of the two vectors.

    Version:
    0.1 November 7, 2006
    Author:
    David Eberly

    • Field Detail

      • m_fAdvectionWeight

        private float m_fAdvectionWeight
        The advection weight 'a'.

      • m_fPropagationWeight

        private float m_fPropagationWeight
        The propagation weight 'b'.

      • m_fCurvatureWeight

        private float m_fCurvatureWeight
        The curvature weight 'c'.

      • m_fLaplacianWeight

        private float m_fLaplacianWeight
        The Laplacian weight 'd'.

      • m_fGradientEpsilon

        private float m_fGradientEpsilon
        A small, positive constant added to the gradient magnitude when it is used in a division. This guards against divisions by zero.

      • m_fMaxAdvPropTimeStep

        private float m_fMaxAdvPropTimeStep
        The stability of the numerical PDE solver depends heavily on the data being processed. The time step must be carefully controlled to be small enough to ensure stability. These parameters are modified during the PDE solver update to reflect the current state of the image data. The modified values are used in computeTimeStep() to calculate the time step to be used by the next PDE solver update.

      • m_fMaxCurvatureTimeStep

        private float m_fMaxCurvatureTimeStep
        The stability of the numerical PDE solver depends heavily on the data being processed. The time step must be carefully controlled to be small enough to ensure stability. These parameters are modified during the PDE solver update to reflect the current state of the image data. The modified values are used in computeTimeStep() to calculate the time step to be used by the next PDE solver update.

      • m_fMaxAdvectionChange

        private float m_fMaxAdvectionChange
        The stability of the numerical PDE solver depends heavily on the data being processed. The time step must be carefully controlled to be small enough to ensure stability. These parameters are modified during the PDE solver update to reflect the current state of the image data. The modified values are used in computeTimeStep() to calculate the time step to be used by the next PDE solver update.

      • m_fMaxPropagationChange

        private float m_fMaxPropagationChange
        The stability of the numerical PDE solver depends heavily on the data being processed. The time step must be carefully controlled to be small enough to ensure stability. These parameters are modified during the PDE solver update to reflect the current state of the image data. The modified values are used in computeTimeStep() to calculate the time step to be used by the next PDE solver update.

      • m_fMaxCurvatureChange

        private float m_fMaxCurvatureChange
        The stability of the numerical PDE solver depends heavily on the data being processed. The time step must be carefully controlled to be small enough to ensure stability. These parameters are modified during the PDE solver update to reflect the current state of the image data. The modified values are used in computeTimeStep() to calculate the time step to be used by the next PDE solver update.

    • Constructor Detail

      • LseParameters

        public LseParameters()
        Creates a new LseParameters object.

    • Method Detail

      • setAdvectionWeight

        public final void setAdvectionWeight​(float fAdvectionWeight)
        Set the advection weight 'a'.
        Parameters:
        fAdvectionWeight - The desired advection weight.

      • getAdvectionWeight

        public final float getAdvectionWeight()
        Get the advection weight 'b'.
        Returns:
        The current advection weight.

      • setPropagationWeight

        public final void setPropagationWeight​(float fPropagationWeight)
        Set the propagation weight 'b'.
        Parameters:
        fPropagationWeight - The desired propagation weight.

      • getPropagationWeight

        public final float getPropagationWeight()
        Get the propagation weight 'b'.
        Returns:
        The current propagation weight.

      • setCurvatureWeight

        public final void setCurvatureWeight​(float fCurvatureWeight)
        Set the curvature weight 'c'.
        Parameters:
        fCurvatureWeight - The desired curvature weight.

      • getCurvatureWeight

        public final float getCurvatureWeight()
        Get the curvature weight 'c'.
        Returns:
        The current curvature weight.

      • setLaplacianWeight

        public final void setLaplacianWeight​(float fLaplacianWeight)
        Set the Laplacian weight 'd'.
        Parameters:
        fLaplacianWeight - The desired Laplacian weight. The value must be nonnegative. If it is positive, the Laplacian term adds some diffusion to the equation, which tends to stabilize the numerical calculations.

      • getLaplacianWeight

        public final float getLaplacianWeight()
        Get the Laplacian weight 'd'.
        Returns:
        The current Laplacian weight.

      • setGradientEpsilon

        public final void setGradientEpsilon​(float fGradientEpsilon)
        Set the gradient magnitude epsilon value. This value guards against a division by z zero-valued gradient magnitude in the curvature calculations.
        Parameters:
        fGradientEpsilon - The gradient magnitude epsilon value. The number must be positive.

      • getGradientEpsilon

        public final float getGradientEpsilon()
        Get The gradient magnitude epsilon value.
        Returns:
        The current gradient magnitude epsilon value.

      • setMaxAdvPropTimeStep

        public final void setMaxAdvPropTimeStep​(float fMaxAdvPropTimeStep)
        Set the maximum advection-propagation time step. The default value is 0.25. At this time, there is no reason to change its value.
        Parameters:
        fMaxAdvPropTimeStep - The desired maximum advection-propagation time step.

      • getMaxAdvPropTimeStep

        public final float getMaxAdvPropTimeStep()
        Get the maximum advection-propagation time step.
        Returns:
        The current maximum advection-propagation time step.

      • setMaxCurvatureTimeStep

        public final void setMaxCurvatureTimeStep​(float fMaxCurvatureTimeStep)
        Set the maximum curvature time step. The default value is 0.25. At this time, there is no reason to change its value.
        Parameters:
        fMaxCurvatureTimeStep - The desired maximum curvature time step.

      • getMaxCurvatureTimeStep

        public final float getMaxCurvatureTimeStep()
        Get the maximum curvature time step.
        Returns:
        The current maximum curvature time step.

      • setMaxAdvectionChange

        public final void setMaxAdvectionChange​(float fMaxAdvectionChange)
        Set the maximum advection change allowed in the PDE solver update.
        Parameters:
        fMaxAdvectionChange - The desired maximum advection change.

      • getMaxAdvectionChange

        public final float getMaxAdvectionChange()
        Get the maximum advection change allowed in the PDE solver update.
        Returns:
        The current maximum advection change.

      • setMaxPropagationChange

        public final void setMaxPropagationChange​(float fMaxPropagationChange)
        Set the maximum propagation change allowed in the PDE solver update.
        Parameters:
        fMaxPropagationChange - The desired maximum propagation change.

      • getMaxPropagationChange

        public final float getMaxPropagationChange()
        Get the maximum propagation change allowed in the PDE solver update.
        Returns:
        The current maximum propagation change.

      • setMaxCurvatureChange

        public final void setMaxCurvatureChange​(float fMaxCurvatureChange)
        Set the maximum curvature change allowed in the PDE solver update.
        Parameters:
        fMaxCurvatureChange - The desired maximum curvature change.

      • getMaxCurvatureChange

        public final float getMaxCurvatureChange()
        Get the maximum curvature change allowed in the PDE solver update.
        Returns:
        The current maximum curvature change.

      • computeTimeStep

        public final float computeTimeStep()
        Compute the time step for the next PDE solver update. The time step is based on the current state of the system that is represented by the members of this class. This state is modified in the LseEvolve2 function computeFunction(x,y) and in the LseEvolve3 member function computeFunction(x,y,z).
        Returns:
        The next time step to be used by the PDE solver update.