net.imglib2.algorithm.morphology.distance

Class EuclidianDistanceAnisotropic

java.lang.Object

net.imglib2.algorithm.morphology.distance.EuclidianDistanceAnisotropic

All Implemented Interfaces:

Distance

public class EuclidianDistanceAnisotropic
extends Object
implements Distance

Implementation of weighted anisotropic Euclidian distance: D( p ) = min_q f(q) + sum_i w_i*(p_i - q_i)*(p_i - q_i).

Author:

Philipp Hanslovsky

Constructor Summary

Constructors

Constructor and Description
EuclidianDistanceAnisotropic(double... weights) When accounting for anisotropic image data, the ratios of the weights should be equal to the squared ratios of the voxel sizes, e.g. if voxel is twice as long along the y-axis, the weight for the y-axis should be four times as big as the weight for the x-axis.
EuclidianDistanceAnisotropic(int nDim, double weight) Use the same weight for each dimension

Method Summary

Modifier and Type	Method and Description
double	evaluate(double x, double xShift, double yShift, int dim) Evaluate function with family parameters xShift and yShift at position x in dimension dim.
double	intersect(double xShift1, double yShift1, double xShift2, double yShift2, int dim) Determine the intersection point in dimension dim of two members of the function family.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

EuclidianDistanceAnisotropic

public EuclidianDistanceAnisotropic(double... weights)

When accounting for anisotropic image data, the ratios of the weights should be equal to the squared ratios of the voxel sizes, e.g. if voxel is twice as long along the y-axis, the weight for the y-axis should be four times as big as the weight for the x-axis.

Parameters:

weights - multiply squared difference between x and xShift with weight (one weight per dimension)

EuclidianDistanceAnisotropic

public EuclidianDistanceAnisotropic(int nDim,
                                    double weight)

Use the same weight for each dimension

Parameters:

nDim - number of dimensions

weight - multiply squared difference between x and xShift with weight

Method Detail

evaluate

public double evaluate(double x,
                       double xShift,
                       double yShift,
                       int dim)

Description copied from interface: Distance

Evaluate function with family parameters xShift and yShift at position x in dimension dim.

Specified by:

evaluate in interface Distance

Parameters:

x - position on x-axis at which function is evaluated

xShift - shift along the x-axis

yShift - shift along y-axis

dim - dimension in which to evaluate

Returns:

Function value at position x

intersect

public double intersect(double xShift1,
                        double yShift1,
                        double xShift2,
                        double yShift2,
                        int dim)

Description copied from interface: Distance

Determine the intersection point in dimension dim of two members of the function family. The members are parameterized by xShift1, yShift1, xShift2, yShift2, with xShift1 > xShift2.

Specified by:

intersect in interface Distance

Parameters:

xShift1 - shift along the x-axis

yShift1 - shift along the y-axis

xShift2 - shift along the x-axis

yShift2 - shift along the y-axis

dim - dimension in which to evaluate

Returns:

intersection point along axis defined by dim