net.imglib2.algorithm.blocks.dfield

Class DisplacementField<T extends net.imglib2.type.NativeType<T> & net.imglib2.type.numeric.RealType<T>>

java.lang.Object

net.imglib2.algorithm.blocks.dfield.DisplacementField<T>

Type Parameters:

T - displacement component type. should be DoubleType or FloatType

All Implemented Interfaces:

net.imglib2.EuclideanSpace, net.imglib2.Typed<T>

public class DisplacementField<T extends net.imglib2.type.NativeType<T> & net.imglib2.type.numeric.RealType<T>>
extends Object
implements net.imglib2.EuclideanSpace, net.imglib2.Typed<T>

A normalized displacement field, and its scale and offset with respect to a source image.

A "normalized" field expresses displacements in units relative to the displacement grid's own pixel spacing. Consequently, downsampling the grid by a factor of N requires scaling the displacement vectors by 1/N to maintain normalization.

Constructor Summary

Constructors

Constructor and Description
DisplacementField(BlockSupplier<T> displacements, double[] scale, double[] translation) Constructs a displacement field with the specified normalized displacements, scale, and translation.

Method Summary

Modifier and Type	Method and Description
BlockSupplier<T>	displacements() Displacement vector field.
T	getType()
int	numDimensions()
double[]	scale() The displacement field coordinates and displacement vectors should be scaled by this factor when looking up intensities in a source image.
double[]	translation() When interpolating displacements() into a position field for value look-up in the source image, this translation vector is added.

Methods inherited from class java.lang.Object

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

Constructor Detail

DisplacementField

public DisplacementField(BlockSupplier<T> displacements,
                         double[] scale,
                         double[] translation)

Constructs a displacement field with the specified normalized displacements, scale, and translation.

A "normalized" field expresses displacements in units relative to the displacement grid's own pixel spacing. Consequently, downsampling the grid by a factor of N requires scaling the displacement vectors by 1/N to maintain normalization.

Parameters:

displacements - provides displacement vectors, with components of the displacement vector in dimension 0. (shifting all other dimensions by 1. That is, the BlockSupplier has numDimensions() + 1 dimensions.)

scale - displacement field coordinates and displacement vectors should be scaled by this factor when looking up intensities in a source image.

translation - when interpolating displacements() into a position field for value look-up in the source image, this translation should be added. (This happens after scaling, so the translation is in units of source image pixels).

Method Detail

numDimensions

public int numDimensions()

Specified by:

numDimensions in interface net.imglib2.EuclideanSpace

getType

public T getType()

Specified by:

getType in interface net.imglib2.Typed<T extends net.imglib2.type.NativeType<T> & net.imglib2.type.numeric.RealType<T>>

displacements

public BlockSupplier<T> displacements()

Displacement vector field.

The components of the displacement vector are in dimension 0, shifting all other dimensions by 1. (The returned BlockSupplier has numDimensions() + 1 dimensions.)

Returns:

the displacement field

scale

public double[] scale()

The displacement field coordinates and displacement vectors should be scaled by this factor when looking up intensities in a source image.

In other words, this is the spacing of the displacement field grid in units of source image pixels.

Returns:

relative scale of the displacement field wrt the source image

translation

public double[] translation()

When interpolating displacements() into a position field for value look-up in the source image, this translation vector is added. (This happens after scaling, so the translation is in units of source image pixels).

For example, this can be used to compensate for half-pixel offsets when applying downsampled displacement fields to full-resolution images.

Returns:

position field offset in source pixels