net.imglib2.algorithm.metrics.segmentation

Class SEGMetrics

java.lang.Object

net.imglib2.algorithm.metrics.segmentation.SEGMetrics

public class SEGMetrics
extends Object

The SEG metrics computes the IoU metrics between ground-truth labels and the prediction labels that overlap with a percentage greater than 0.5. In the context of instance segmentation, the labels are the integer pixel values. Pixels with value 0 are considered background and are ignored during the metrics calculation. If the ground-truth image is only background (no labels), then the metrics returns NaN.

The metrics expect images of dimensions XYZT, where Z and T can be of depth 1. The metrics score is calculated for each ground-truth label in each XYZ volume (XY if dimension Z is of depth 1), and averaged over the total number of ground-truth labels in the XYZT volume. If the Z dimension has depth greater than 1, then the labels are considered 3D and pixels of equal values at different depths are considered to be part of the same labeling.

Finally, if the image stack does not fit in memory, you can use LazySEGMetrics to compute a running SEG score for which images can be added one at time.

Reference: Ulman, V., Maška, M., Magnusson, K. et al. An objective comparison of cell-tracking algorithms. Nat Methods 14, 1141–1152 (2017).

Author:

Joran Deschamps

See Also:

Original implementation by Martin Maška and Vladimír Ulman

Constructor Summary

Constructors

Constructor and Description
SEGMetrics()

Method Summary

Modifier and Type	Method and Description
protected static double[][]	computeCostMatrix(ConfusionMatrix cM)
static <T,I extends net.imglib2.type.numeric.IntegerType<I>,U,J extends net.imglib2.type.numeric.IntegerType<J>> double	computeMetrics(net.imglib2.roi.labeling.ImgLabeling<T,I> groundTruth, net.imglib2.roi.labeling.ImgLabeling<U,J> prediction) Compute a global SEG score between labels from a ground-truth and a predicted image.
static <I extends net.imglib2.type.numeric.IntegerType<I>,J extends net.imglib2.type.numeric.IntegerType<J>> double	computeMetrics(net.imglib2.RandomAccessibleInterval<I> groundTruth, net.imglib2.RandomAccessibleInterval<J> prediction) Compute a global SEG score between labels from a ground-truth and a predicted image.
protected static double	getLocalIoUScore(ConfusionMatrix cM, int i, int j)
protected static <I extends net.imglib2.type.numeric.IntegerType<I>,J extends net.imglib2.type.numeric.IntegerType<J>> net.imglib2.util.Pair<Integer,Double>	runSingle(net.imglib2.RandomAccessibleInterval<I> groundTruth, net.imglib2.RandomAccessibleInterval<J> prediction)

Methods inherited from class java.lang.Object

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

Constructor Detail

SEGMetrics

public SEGMetrics()

Method Detail

computeMetrics

public static <T,I extends net.imglib2.type.numeric.IntegerType<I>,U,J extends net.imglib2.type.numeric.IntegerType<J>> double computeMetrics(net.imglib2.roi.labeling.ImgLabeling<T,I> groundTruth,
                                                                                                                                              net.imglib2.roi.labeling.ImgLabeling<U,J> prediction)

Compute a global SEG score between labels from a ground-truth and a predicted image. The method expects images of dimension XYZT. The score is computed for each ground-truth label over each XYZ volume (or XY if Z is of depth 1) and averaged over all ground-truth labels in XYZT. If both images are empty (only pixels with value 0), then the metrics score is NaN.

This method is not compatible with ImgLabeling with intersecting labels.

Type Parameters:

T - Label type associated to the ground-truth

I - Ground-truth pixel type

U - Label type associated to the prediction

J - Prediction pixel type

Parameters:

groundTruth - Ground-truth image

prediction - Predicted image

Returns:

Metrics score

computeMetrics

public static <I extends net.imglib2.type.numeric.IntegerType<I>,J extends net.imglib2.type.numeric.IntegerType<J>> double computeMetrics(net.imglib2.RandomAccessibleInterval<I> groundTruth,
                                                                                                                                          net.imglib2.RandomAccessibleInterval<J> prediction)

Compute a global SEG score between labels from a ground-truth and a predicted image. The method expects images of dimension XYZT. The score is computed over the XYZ volume (or XY if Z is of depth 1) and averaged over all ground-truth labels. If both images are empty (only pixels with value 0), then the metrics score is NaN.

Type Parameters:

I - Ground-truth pixel type

J - Prediction pixel type

Parameters:

groundTruth - Ground-truth image

prediction - Predicted image

Returns:

Metrics score

runSingle

protected static <I extends net.imglib2.type.numeric.IntegerType<I>,J extends net.imglib2.type.numeric.IntegerType<J>> net.imglib2.util.Pair<Integer,Double> runSingle(net.imglib2.RandomAccessibleInterval<I> groundTruth,
                                                                                                                                                                       net.imglib2.RandomAccessibleInterval<J> prediction)

computeCostMatrix

protected static double[][] computeCostMatrix(ConfusionMatrix cM)

getLocalIoUScore

protected static double getLocalIoUScore(ConfusionMatrix cM,
                                         int i,
                                         int j)