net.imglib2.roi.labeling

Class OverlappingLabels<T extends Comparable<T>>

java.lang.Object

net.imglib2.roi.labeling.OverlappingLabels<T>

public class OverlappingLabels<T extends Comparable<T>>
extends Object

Constructor Summary

Constructors

Constructor and Description
OverlappingLabels(RandomAccessibleInterval<LabelingType<T>> labeling)

Method Summary

Modifier and Type	Method and Description
List<T>	getIndexedLabels() Get the list of labels for this confusion matrix.
RandomAccessibleInterval<UnsignedIntType>	getMatrix() Get the confusion/co-occurrence matrix.
RandomAccessibleInterval<DoubleType>	getNormalizedMatrix() Get the normalized directional confusion matrix.
double	getPartialOverlap(T ofLabel, T overlappingLabel) The fraction of ofLabel that overlaps with overlappingLabel.
double	getPartialOverlapForIndex(int ofLabelIndex, int overlappingLabelIndex) The fraction of the label at ofLabelIndex that overlaps with the label at overlappingLabelIndex.
long	getPixelOverlap(T label1, T label2) The number of pixels overlapping between two labels.
long	getPixelOverlapForIndex(int index1, int index2) The number of pixels overlapping between two labels at the specified indices.

Methods inherited from class java.lang.Object

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

Constructor Detail

OverlappingLabels

public OverlappingLabels(RandomAccessibleInterval<LabelingType<T>> labeling)

Method Detail

getMatrix

public RandomAccessibleInterval<UnsignedIntType> getMatrix()

Get the confusion/co-occurrence matrix. Values indicate the number of pixels that are in common between the label at position x and the label at position y. For the mapping of dimension indices to labels, see getIndexedLabels().

Returns:

the co-occurrence matrix

getNormalizedMatrix

public RandomAccessibleInterval<DoubleType> getNormalizedMatrix()

Get the normalized directional confusion matrix. Values indicate the proportion of the label at position x that is co-occurring with the label at position y. For the mapping of dimension indices to labels, see getIndexedLabels().

Returns:

the normalized co-occurrence matrix

getIndexedLabels

public List<T> getIndexedLabels()

Get the list of labels for this confusion matrix.

Returns:

the list of labels in order of the matrix indices

getPixelOverlap

public long getPixelOverlap(T label1,
                            T label2)

The number of pixels overlapping between two labels.

Parameters:

label1 - The first label

label2 - The second label

Returns:

the size of the overlapping region (in pixels)

getPixelOverlapForIndex

public long getPixelOverlapForIndex(int index1,
                                    int index2)

The number of pixels overlapping between two labels at the specified indices.

Parameters:

index1 - The index of the first label

index2 - The index of the second label

Returns:

the size of the overlapping region (in pixels)

getPartialOverlap

public double getPartialOverlap(T ofLabel,
                                T overlappingLabel)

The fraction of ofLabel that overlaps with overlappingLabel. If ofLabel has size 2 and overlappingLabel has size 100, and they overlap by 1 pixel, the resulting fraction is 0.5

Parameters:

ofLabel -

overlappingLabel -

Returns:

the overlapping fraction

getPartialOverlapForIndex

public double getPartialOverlapForIndex(int ofLabelIndex,
                                        int overlappingLabelIndex)

The fraction of the label at ofLabelIndex that overlaps with the label at overlappingLabelIndex. See getPartialOverlap(Comparable, Comparable).

Parameters:

ofLabelIndex -

overlappingLabelIndex -

Returns:

the overlapping fraction