Package qupath.process.gui.commands.ml

Class PixelClassifierTraining

java.lang.Object

qupath.process.gui.commands.ml.PixelClassifierTraining

public class PixelClassifierTraining
extends Object

Helper class for training a pixel classifier.

Author:

Pete Bankhead

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	PixelClassifierTraining.ClassifierTrainingData	Wrapper for training data.

Constructor Summary

Constructors

Constructor	Description
PixelClassifierTraining(ImageDataOp featureCalculator)	Create a new pixel classifier helper, to support generating training data.

Method Summary

Modifier and Type	Method	Description
PixelClassifierTraining.ClassifierTrainingData	createTrainingData(Collection<ImageData<BufferedImage>> imageData)	Create training data, using a label map automatically generated from the available classifications.
PixelClassifierTraining.ClassifierTrainingData	createTrainingData(ImageData<BufferedImage> imageData)	Create training data, using a label map automatically generated from the available classifications.
PixelClassifierTraining.ClassifierTrainingData	createTrainingDataForLabelMap(Collection<ImageData<BufferedImage>> imageData, Map<PathClass,Integer> labels)	Get a classifier training map, using a predefined label map (which determines which classifications to use).
BoundaryStrategy	getBoundaryStrategy()	Get the strategy for handling the boundaries of area annotations.
ImageDataOp	getFeatureOp()	Get an ImageDataOp used for feature calculation.
PixelCalibration	getResolution()	Get the resolution at which the training should occur.
void	setBoundaryStrategy(BoundaryStrategy strategy)	Set the strategy for handling the boundaries of area annotations.
void	setFeatureOp(ImageDataOp featureOp)	Set the ImageDataOp used to calculate features.
void	setResolution(PixelCalibration cal)	Set the resolution at which the training should occur.

Methods inherited from class java.lang.Object

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

Constructor Details

PixelClassifierTraining

public PixelClassifierTraining(ImageDataOp featureCalculator)

Create a new pixel classifier helper, to support generating training data.

Parameters:

featureCalculator -

Method Details

getFeatureOp

public ImageDataOp getFeatureOp()

Get an ImageDataOp used for feature calculation.

Returns:

getResolution

public PixelCalibration getResolution()

Get the resolution at which the training should occur.

Returns:

setResolution

public void setResolution(PixelCalibration cal)

Set the resolution at which the training should occur.

Parameters:

cal -

setFeatureOp

public void setFeatureOp(ImageDataOp featureOp)

Set the ImageDataOp used to calculate features.

Parameters:

featureOp -

setBoundaryStrategy

public void setBoundaryStrategy(BoundaryStrategy strategy)

Set the strategy for handling the boundaries of area annotations.

Parameters:

strategy -

getBoundaryStrategy

public BoundaryStrategy getBoundaryStrategy()

Get the strategy for handling the boundaries of area annotations.

Returns:

createTrainingData

public PixelClassifierTraining.ClassifierTrainingData createTrainingData(ImageData<BufferedImage> imageData)
                                                                  throws IOException

Create training data, using a label map automatically generated from the available classifications.

Parameters:

imageData -

Returns:

Throws:

IOException

createTrainingData

public PixelClassifierTraining.ClassifierTrainingData createTrainingData(Collection<ImageData<BufferedImage>> imageData)
                                                                  throws IOException

Create training data, using a label map automatically generated from the available classifications.

Parameters:

imageData -

Returns:

Throws:

IOException

createTrainingDataForLabelMap

public PixelClassifierTraining.ClassifierTrainingData createTrainingDataForLabelMap(Collection<ImageData<BufferedImage>> imageData,
 Map<PathClass,Integer> labels)
                                                                             throws IOException

Get a classifier training map, using a predefined label map (which determines which classifications to use).

Parameters:

imageData - collection of ImageData used for training

labels - map linking classifications to labels in the output; may be null, in which case a label map will be generated from the data

Returns:

a PixelClassifierTraining.ClassifierTrainingData object representing training data

Throws:

IOException