Package boofcv.alg.background

Class BackgroundGmmCommon

java.lang.Object

boofcv.alg.background.BackgroundGmmCommon

public class BackgroundGmmCommon
extends Object

Common code for all implementations of BackgroundAlgorithmGmm. This is where most of the important mathematics is contained.

Field Summary

Fields

Modifier and Type	Field	Description
float	decay
int	gaussianStride
int	imageHeight
int	imageWidth
float	initialVariance
GImageGray	inputWrapperG
GImageMultiBand	inputWrapperMB
float	learningRate
float	maxDistance
int	maxGaussians
RArray2D_F32	model
int	modelStride
int	numBands
float	significantWeight
GrowArray<float[]>	storagePixels
int	unknownValue

Constructor Summary

Constructors

Constructor	Description
BackgroundGmmCommon(float learningPeriod, float decayCoef, int maxGaussians, ImageType imageType)

Method Summary

Modifier and Type	Method	Description
int	checkBackground(float[] pixelValue, float[] dataRow, int modelIndex)	Checks to see if the pixel value refers to the background or foreground
int	checkBackground(float pixelValue, float[] dataRow, int modelIndex)	Checks to see if the the pivel value refers to the background or foreground
float	getInitialVariance()
float	getMaxDistance()
float	getSignificantWeight()
void	setInitialVariance(float initialVariance)
void	setLearningPeriod(float period)
void	setMaxDistance(float maxDistance)
void	setSignificantWeight(float significantWeight)
int	updateMixture(float[] pixelValue, float[] dataRow, int modelIndex)	Updates the mixtures of gaussian and determines if the pixel matches the background model
int	updateMixture(float pixelValue, float[] dataRow, int modelIndex)	Updates the mixtures of gaussian and determines if the pixel matches the background model
void	updateWeightAndPrune(float[] dataRow, int modelIndex, int ng, int bestIndex, float bestWeight)	Updates the weight of each Gaussian and prunes one which have a negative weight after the update.

Methods inherited from class java.lang.Object

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

Field Details

model

public RArray2D_F32 model

imageWidth

public int imageWidth

imageHeight

public int imageHeight

numBands

public int numBands

modelStride

public int modelStride

gaussianStride

public int gaussianStride

learningRate

public float learningRate

decay

public float decay

maxGaussians

public int maxGaussians

maxDistance

public float maxDistance

significantWeight

public float significantWeight

initialVariance

public float initialVariance

unknownValue

public int unknownValue

inputWrapperMB

public GImageMultiBand inputWrapperMB

storagePixels

public GrowArray<float[]> storagePixels

inputWrapperG

public GImageGray inputWrapperG

Constructor Details

BackgroundGmmCommon

public BackgroundGmmCommon(float learningPeriod,
 float decayCoef,
 int maxGaussians,
 ImageType imageType)

Method Details

updateMixture

public int updateMixture(float[] pixelValue,
 float[] dataRow,
 int modelIndex)

Updates the mixtures of gaussian and determines if the pixel matches the background model

Returns:

true if it matches the background or false if not

updateWeightAndPrune

public void updateWeightAndPrune(float[] dataRow,
 int modelIndex,
 int ng,
 int bestIndex,
 float bestWeight)

Updates the weight of each Gaussian and prunes one which have a negative weight after the update.

updateMixture

public int updateMixture(float pixelValue,
 float[] dataRow,
 int modelIndex)

Updates the mixtures of gaussian and determines if the pixel matches the background model

Returns:

true if it matches the background or false if not

checkBackground

public int checkBackground(float[] pixelValue,
 float[] dataRow,
 int modelIndex)

Checks to see if the pixel value refers to the background or foreground

Returns:

true for background or false for foreground

checkBackground

public int checkBackground(float pixelValue,
 float[] dataRow,
 int modelIndex)

Checks to see if the the pivel value refers to the background or foreground

Returns:

true for background or false for foreground

setLearningPeriod

public void setLearningPeriod(float period)

getMaxDistance

public float getMaxDistance()

setMaxDistance

public void setMaxDistance(float maxDistance)

getSignificantWeight

public float getSignificantWeight()

setSignificantWeight

public void setSignificantWeight(float significantWeight)

getInitialVariance

public float getInitialVariance()

setInitialVariance

public void setInitialVariance(float initialVariance)