Package qupath.opencv.ml.objects.features

Class Preprocessing

java.lang.Object

qupath.opencv.ml.objects.features.Preprocessing

public class Preprocessing
extends Object

Helper class for preprocessing input for machine learning algorithms using OpenCV Mats.

Author:

Pete Bankhead

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	Preprocessing.PCAProjector	Helper class to apply PCA projection.

Constructor Summary

Constructors

Constructor	Description
Preprocessing()

Method Summary

Modifier and Type	Method	Description
static Normalizer	createNormalizer(Normalization normalization, Mat samples, double missingValue)	Create a simple normalizer to rescale input data.
static Preprocessing.PCAProjector	createPCAProjector(Mat data, double retainedVariance, boolean normalize)	Create a principle components analysis projection to reduce features.
static void	normalize(Mat samples, Normalizer normalizer)	Apply a Normalizer to new training data samples.

Methods inherited from class java.lang.Object

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

Constructor Details

Preprocessing

public Preprocessing()

Method Details

createPCAProjector

public static Preprocessing.PCAProjector createPCAProjector(Mat data,
 double retainedVariance,
 boolean normalize)

Create a principle components analysis projection to reduce features.

Parameters:

data - input data used to create the projector

retainedVariance - variance to retain (determines the number of output features)

normalize - if true, normalize the output features

Returns:

a Preprocessing.PCAProjector that can be applied to new data

createNormalizer

public static Normalizer createNormalizer(Normalization normalization,
 Mat samples,
 double missingValue)

Create a simple normalizer to rescale input data.

Parameters:

normalization - the method of normalization to apply

samples - the input samples used to determine the normalization parameter

missingValue - an optional value that may be used to replace non-finite (i.e. missing) feature values

Returns:

a Normalizer that may be applied to new data

normalize

public static void normalize(Mat samples,
 Normalizer normalizer)

Apply a Normalizer to new training data samples. Features may be either columns or channels.

Parameters:

samples - the input data

normalizer - the normalizer to apply