Package gov.nih.mipav.model.algorithms.utilities

Class AlgorithmSubsample

java.lang.Object

java.lang.Thread

gov.nih.mipav.model.algorithms.AlgorithmBase

gov.nih.mipav.model.algorithms.utilities.AlgorithmSubsample

All Implemented Interfaces:

ActionListener, WindowListener, Runnable, EventListener

public class AlgorithmSubsample
extends AlgorithmBase

Takes in an image and subsamples it to a new set of dimensions. The subsample image is a Gaussian weighted average of neighboring pixels (voxels) of the original image. For 4D images only the first 3 dimensions are subsampled

Author:

Evan McCreedy

Nested Class Summary

Nested classes/interfaces inherited from class java.lang.Thread

Thread.Builder, Thread.State, Thread.UncaughtExceptionHandler

Field Summary

Fields

Modifier and Type	Field	Description
private ModelImage	paddedImage	Padded image.
private double[]	posRatios	Ratios used to translate points in the result image to points in the source image.
private boolean	processIndep	DOCUMENT ME!
private int[]	resultExtents	Extents the resultImage should have.
private ModelImage	resultImage	Result image.
private float[]	sigmas	Std deviation of the Gaussian function.
private ModelImage	srcImage	Image to subsample from.
private boolean	transformVOI	DOCUMENT ME!
private TransMatrix	transMatrix	DOCUMENT ME!
private boolean	trilinearFlag	Whether to perform trilinear filtering when getting intensities of points on the source image.

Fields inherited from class gov.nih.mipav.model.algorithms.AlgorithmBase

destFlag, destImage, image25D, mask, maxProgressValue, minProgressValue, multiThreadingEnabled, nthreads, progress, progressModulus, progressStep, runningInSeparateThread, separable, threadStopped

Fields inherited from class java.lang.Thread

MAX_PRIORITY, MIN_PRIORITY, NORM_PRIORITY

Constructor Summary

Constructors

Constructor	Description
AlgorithmSubsample(ModelImage src, ModelImage result, int[] newExtents, int[] padExtents, float[] _sigmas, boolean indep, boolean transformVOI, TransMatrix transMatrix, boolean doPad)	Setup for a later subsampling of an image.

Method Summary

Modifier and Type	Method	Description
private double	computeGaussian2D(double locX, double locY, double sigmaX, double sigmaY)	Compute the value of the 2D gaussian at a point.
private double	computeGaussian3D(double locX, double locY, double locZ, double sigmaX, double sigmaY, double sigmaZ)	Computes the gaussian weights for a particular point.
private void	computeRatios()	Computes the image scaling ratios.
void	finalize()	Prepares this class for destruction.
private float[]	getNearestNeighborPos2D(float x, float y)	Translates a point from the result image to the source image.
private float[]	getNearestNeighborPos3D(float x, float y, float z)	Translates a point from the result image to the source image.
ModelImage	getResultImage()	Return the result image (or null if not created yet).
private float	getTriLinearVal3D(float x, float y, float z, int t, int channelIndex, ModelSimpleImage src)	Version of get that performs tri-linear interpoloation.
private float	getTriLinearVal3D(float x, float y, float z, int channelIndex, ModelSimpleImage src)	Version of get that performs tri-linear interpoloation.
private float	getValue2D(float x, float y, float z, int channelIndex, ModelSimpleImage src)	Get the value of a 2D point.
private float	getValue3D(float x, float y, float z, int t, int channelIndex, ModelSimpleImage src)	Get the value of a 4D point.
private float	getValue3D(float x, float y, float z, int channelIndex, ModelSimpleImage src)	Get the value of a 3D point.
private static String	makeImageName(String image_name, String ext)	Helper method for making the result image's name.
private ModelImage	padImage(ModelImage kImage, int[] padExtents)
void	runAlgorithm()	Subsample the image.
private ModelSimpleImage	subsample2D(int nImages)	Subsample a 2D image.
private ModelSimpleImage	subsample3D()	Perform the subsampling.
private ModelSimpleImage	subsample4D()	Perform the subsampling.
private void	transform25DVOI(ModelImage image, float[] imgBuffer, TransMatrix xfrm)	Transforms and resamples a 2D or 25D VOI using nearest neighbor interpolation.
private void	transform3DVOI(ModelImage image, float[] imgBuffer, TransMatrix xfrm)	Transforms and resamples a 3D VOI using nearest neighbor interpolation.

Methods inherited from class gov.nih.mipav.model.algorithms.AlgorithmBase

actionPerformed, addListener, addProgressChangeListener, calculateImageSize, calculatePrincipleAxis, computeElapsedTime, computeElapsedTime, convertIntoFloat, delinkProgressToAlgorithm, delinkProgressToAlgorithmMulti, displayError, errorCleanUp, fireProgressStateChanged, fireProgressStateChanged, fireProgressStateChanged, fireProgressStateChanged, fireProgressStateChanged, generateProgressValues, getDestImage, getElapsedTime, getMask, getMaxProgressValue, getMinProgressValue, getNumberOfThreads, getProgress, getProgressChangeListener, getProgressChangeListeners, getProgressModulus, getProgressStep, getProgressValues, getSrcImage, isCompleted, isImage25D, isMultiThreadingEnabled, isRunningInSeparateThread, isThreadStopped, linkProgressToAlgorithm, linkProgressToAlgorithm, makeProgress, notifyListeners, removeListener, removeProgressChangeListener, run, setCompleted, setImage25D, setMask, setMaxProgressValue, setMinProgressValue, setMultiThreadingEnabled, setNumberOfThreads, setProgress, setProgressModulus, setProgressStep, setProgressValues, setProgressValues, setRunningInSeparateThread, setSrcImage, setStartTime, setThreadStopped, startMethod, windowActivated, windowClosed, windowClosing, windowDeactivated, windowDeiconified, windowIconified, windowOpened

Methods inherited from class java.lang.Thread

activeCount, checkAccess, clone, countStackFrames, currentThread, dumpStack, enumerate, getAllStackTraces, getContextClassLoader, getDefaultUncaughtExceptionHandler, getId, getName, getPriority, getStackTrace, getState, getThreadGroup, getUncaughtExceptionHandler, holdsLock, interrupt, interrupted, isAlive, isDaemon, isInterrupted, isVirtual, join, join, join, join, ofPlatform, ofVirtual, onSpinWait, resume, setContextClassLoader, setDaemon, setDefaultUncaughtExceptionHandler, setName, setPriority, setUncaughtExceptionHandler, sleep, sleep, sleep, start, startVirtualThread, stop, suspend, threadId, toString, yield

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, wait, wait, wait

Field Details

posRatios

private double[] posRatios

Ratios used to translate points in the result image to points in the source image.

processIndep

private boolean processIndep

DOCUMENT ME!

resultExtents

private int[] resultExtents

Extents the resultImage should have.

resultImage

private ModelImage resultImage

Result image.

paddedImage

private ModelImage paddedImage

Padded image.

sigmas

private float[] sigmas

Std deviation of the Gaussian function.

srcImage

private ModelImage srcImage

Image to subsample from.

transformVOI

private boolean transformVOI

DOCUMENT ME!

transMatrix

private TransMatrix transMatrix

DOCUMENT ME!

trilinearFlag

private boolean trilinearFlag

Whether to perform trilinear filtering when getting intensities of points on the source image.

Constructor Details

AlgorithmSubsample

public AlgorithmSubsample(ModelImage src,
 ModelImage result,
 int[] newExtents,
 int[] padExtents,
 float[] _sigmas,
 boolean indep,
 boolean transformVOI,
 TransMatrix transMatrix,
 boolean doPad)

Setup for a later subsampling of an image.

Parameters:

src - image to be subsampled

result - image created by subsampling

newExtents - dimensions of the result image to be sampled into

_sigmas - x, y, and z scales for the gaussian function

indep - process slices independently (cannot change the z dimension)

transformVOI - if true transform the VOIs

transMatrix - transformation matrix

Method Details

finalize

public void finalize()

Prepares this class for destruction.

Overrides:

finalize in class AlgorithmBase

getResultImage

public ModelImage getResultImage()

Return the result image (or null if not created yet).

Returns:

the subsampled image

runAlgorithm

public void runAlgorithm()

Subsample the image.

Specified by:

runAlgorithm in class AlgorithmBase

computeGaussian2D

private double computeGaussian2D(double locX,
 double locY,
 double sigmaX,
 double sigmaY)

Compute the value of the 2D gaussian at a point.

Parameters:

locX - x location of the value to compute

locY - y location of the value to compute

sigmaX - standard deviation of the guassian in the x dimension

sigmaY - standard deviation of the guassian in the y dimension

Returns:

the value of the 2D gaussian at the given point

computeGaussian3D

private double computeGaussian3D(double locX,
 double locY,
 double locZ,
 double sigmaX,
 double sigmaY,
 double sigmaZ)

Computes the gaussian weights for a particular point. Written by Paul Hemler in AlgorithmVesselSegmenation.

Parameters:

locX - distance to go in the x dir from the point

locY - distance to go in the y dir from the point

locZ - distance to go in the z dir from the point

sigmaX - stddev of the gaussian function in the x dir

sigmaY - stddev of the gaussian function in the y dir

sigmaZ - stddev of the gaussian function in the z dir

Returns:

the gaussian weight coeff.

computeRatios

private void computeRatios()

Computes the image scaling ratios.

getNearestNeighborPos2D

private float[] getNearestNeighborPos2D(float x,
 float y)

Translates a point from the result image to the source image.

Parameters:

x - x coordinate of point to be translated

y - y coordinate of point to be translated

Returns:

the translated point on the source image

getNearestNeighborPos3D

private float[] getNearestNeighborPos3D(float x,
 float y,
 float z)

Translates a point from the result image to the source image.

Parameters:

x - x coordinate of point to be translated

y - y coordinate of point to be translated

z - z coordinate of point to be translated

Returns:

the translated point on the source image

getTriLinearVal3D

private float getTriLinearVal3D(float x,
 float y,
 float z,
 int channelIndex,
 ModelSimpleImage src)

Version of get that performs tri-linear interpoloation. Note: does NOT perform bounds checking.

Parameters:

x - x coordinate

y - y coordinate

z - z coordinate

channelIndex - color channel

src - image to get intensity value from

Returns:

the tri-linear interpolated value

getTriLinearVal3D

private float getTriLinearVal3D(float x,
 float y,
 float z,
 int t,
 int channelIndex,
 ModelSimpleImage src)

Version of get that performs tri-linear interpoloation. Note: does NOT perform bounds checking.

Parameters:

x - x coordinate

y - y coordinate

z - z coordinate

t - t coordinate

channelIndex - color channel

src - image to get intensity value from

Returns:

the tri-linear interpolated value

getValue2D

private float getValue2D(float x,
 float y,
 float z,
 int channelIndex,
 ModelSimpleImage src)

Get the value of a 2D point.

Parameters:

x - x coordinate

y - y coordinate

z - z coordinate (0 if really 2D)

channelIndex - color channel

src - image to sample from

Returns:

the point's value

getValue3D

private float getValue3D(float x,
 float y,
 float z,
 int channelIndex,
 ModelSimpleImage src)

Get the value of a 3D point.

Parameters:

x - x coordinate

y - y coordinate

z - z coordinate

channelIndex - color channel

src - image to sample from

Returns:

the point's value

getValue3D

private float getValue3D(float x,
 float y,
 float z,
 int t,
 int channelIndex,
 ModelSimpleImage src)

Get the value of a 4D point.

Parameters:

x - x coordinate

y - y coordinate

z - z coordinate

t - DOCUMENT ME!

channelIndex - color channel

src - image to sample from

Returns:

the point's value

subsample2D

private ModelSimpleImage subsample2D(int nImages)

Subsample a 2D image.

Parameters:

nImages - the number of 2D images to subsample (1 if really 2D)

Returns:

the subsampled image

subsample3D

private ModelSimpleImage subsample3D()

Perform the subsampling. Create a result image with the extents given into the constuctor. Then, for each point in the result image, find point which corresponds to it in the source image. Add up weighted intensities of the neighbors of this point in the source image and the intensity of the point itself. Store this value in the point of the result image. Once all points in the result image have been visted, return the result image.

Returns:

resampled image

subsample4D

private ModelSimpleImage subsample4D()

Perform the subsampling. Create a result image with the extents given into the constuctor. Then, for each point in the result image, find point which corresponds to it in the source image. Add up weighted intensities of the neighbors of this point in the source image and the intensity of the point itself. Store this value in the point of the result image. Once all points in the result image have been visted, return the result image. Does not subsample in fourth time dimension - 3.5D operation

Returns:

resampled image

transform25DVOI

private void transform25DVOI(ModelImage image,
 float[] imgBuffer,
 TransMatrix xfrm)

Transforms and resamples a 2D or 25D VOI using nearest neighbor interpolation.

1. Export VOIs as a mask image
2. Transform mask
3. Extract VOI contours from mask image and put in new image.

Parameters:

image - Image where VOIs are stored

imgBuffer - Image array

xfrm - Transformation matrix to be applied

transform3DVOI

private void transform3DVOI(ModelImage image,
 float[] imgBuffer,
 TransMatrix xfrm)

Transforms and resamples a 3D VOI using nearest neighbor interpolation.

1. Export VOIs as a mask image
2. Transform mask
3. Extract VOI contours from mask image and put in new image.

Parameters:

image - Image where VOIs are stored

imgBuffer - Image array

xfrm - Transformation matrix to be applied

padImage

private ModelImage padImage(ModelImage kImage,
 int[] padExtents)

makeImageName

private static String makeImageName(String image_name,
 String ext)

Helper method for making the result image's name. Strips the current extension from the original name, adds the given extension, and returns the new name.

Parameters:

image_name - the original image's name

ext - Extension to add which gives information about what algorithm was performed on the image.

Returns:

The new image name.