boofcv.alg.feature.detect.interest

Class GeneralFeatureDetector<I extends boofcv.struct.image.ImageSingleBand,D extends boofcv.struct.image.ImageSingleBand>

java.lang.Object

boofcv.alg.feature.detect.interest.GeneralFeatureDetector<I,D>

Type Parameters:

I - Input image type.

D - Image derivative type.

public class GeneralFeatureDetector<I extends boofcv.struct.image.ImageSingleBand,D extends boofcv.struct.image.ImageSingleBand>
extends java.lang.Object

Detects features which are local maximums and/or local minimums in the feature intensity image. A list of pixels to exclude as candidates can be provided. Image derivatives need to be computed externally and provided as needed. The passed in GeneralFeatureIntensity is used to determine if local maximums or minimums should be detected.

If a maximum number of features is specified then the N most intense features are returned. By default all found features are returned. Set to a value <= 0 to detect all features.

Field Summary

Fields

Modifier and Type	Field and Description
protected boofcv.struct.QueueCorner	excludeMaximum
protected boofcv.struct.QueueCorner	excludeMinimum
protected NonMaxSuppression	extractor
protected boofcv.struct.QueueCorner	foundMaximum
protected boofcv.struct.QueueCorner	foundMinimum
protected GeneralFeatureIntensity<I,D>	intensity
protected int	maxFeatures
protected SelectNBestFeatures	selectBest

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	GeneralFeatureDetector()
	GeneralFeatureDetector(GeneralFeatureIntensity<I,D> intensity, NonMaxSuppression extractor) Specifies which algorithms to use and configures the detector.

Method Summary

Methods

Modifier and Type	Method and Description
boofcv.struct.image.ImageFloat32	getIntensity()
boofcv.struct.QueueCorner	getMaximums() Returns a list of all the found maximums.
boofcv.struct.QueueCorner	getMinimums() Returns a list of all the found maximums.
boolean	getRequiresGradient() If the image gradient is required for calculations.
boolean	getRequiresHessian() If the image Hessian is required for calculations.
float	getThreshold() Returns the current feature extraction threshold.
boolean	isDetectMaximums() true if it will detector local maximums
boolean	isDetectMinimums() true if it will detector local minimums
void	process(I image, D derivX, D derivY, D derivXX, D derivYY, D derivXY) Computes point features from image gradients.
void	setExcludeMaximum(boofcv.struct.QueueCorner exclude) Specify points which are excluded when detecting maximums
void	setExcludeMinimum(boofcv.struct.QueueCorner exclude) Specify points which are excluded when detecting maximums
void	setMaxFeatures(int numFeatures) Turns on select best features and sets the number it should return.
void	setSearchRadius(int radius) Species the search radius for the feature
void	setThreshold(float threshold) Changes feature extraction threshold.

Methods inherited from class java.lang.Object

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

Field Detail

foundMaximum

protected boofcv.struct.QueueCorner foundMaximum

foundMinimum

protected boofcv.struct.QueueCorner foundMinimum

excludeMaximum

protected boofcv.struct.QueueCorner excludeMaximum

excludeMinimum

protected boofcv.struct.QueueCorner excludeMinimum

selectBest

protected SelectNBestFeatures selectBest

maxFeatures

protected int maxFeatures

extractor

protected NonMaxSuppression extractor

intensity

protected GeneralFeatureIntensity<I extends boofcv.struct.image.ImageSingleBand,D extends boofcv.struct.image.ImageSingleBand> intensity

Constructor Detail

GeneralFeatureDetector

public GeneralFeatureDetector(GeneralFeatureIntensity<I,D> intensity,
                      NonMaxSuppression extractor)

Specifies which algorithms to use and configures the detector.

Parameters:

intensity - Computes how much like the feature the region around each pixel is.

extractor - Extracts the corners from intensity image

GeneralFeatureDetector

protected GeneralFeatureDetector()

Method Detail

process

public void process(I image,
           D derivX,
           D derivY,
           D derivXX,
           D derivYY,
           D derivXY)

Computes point features from image gradients.

Parameters:

image - Original image.

derivX - image derivative in along the x-axis. Only needed if getRequiresGradient() is true.

derivY - image derivative in along the y-axis. Only needed if getRequiresGradient() is true.

derivXX - Second derivative. Only needed if getRequiresHessian() ()} is true.

derivXY - Second derivative. Only needed if getRequiresHessian() ()} is true.

derivYY - Second derivative. Only needed if getRequiresHessian() ()} is true.

setMaxFeatures

public void setMaxFeatures(int numFeatures)

Turns on select best features and sets the number it should return. If a list of excluded features is passed in, then the maximum number of returned features is 'numFeatures' minus the number of excluded features.

Parameters:

numFeatures - Return at most this many features, which are the best.

getRequiresGradient

public boolean getRequiresGradient()

If the image gradient is required for calculations.

Returns:

true if the image gradient is required.

getRequiresHessian

public boolean getRequiresHessian()

If the image Hessian is required for calculations.

Returns:

true if the image Hessian is required.

getIntensity

public boofcv.struct.image.ImageFloat32 getIntensity()

setThreshold

public void setThreshold(float threshold)

Changes feature extraction threshold.

Parameters:

threshold - The new feature extraction threshold.

getThreshold

public float getThreshold()

Returns the current feature extraction threshold.

Returns:

feature extraction threshold.

setExcludeMaximum

public void setExcludeMaximum(boofcv.struct.QueueCorner exclude)

Specify points which are excluded when detecting maximums

Parameters:

exclude - List of points being excluded

getMaximums

public boofcv.struct.QueueCorner getMaximums()

Returns a list of all the found maximums.

Returns:

found point features

setExcludeMinimum

public void setExcludeMinimum(boofcv.struct.QueueCorner exclude)

Specify points which are excluded when detecting maximums

Parameters:

exclude - List of points being excluded

getMinimums

public boofcv.struct.QueueCorner getMinimums()

Returns a list of all the found maximums.

Returns:

found point features

isDetectMinimums

public boolean isDetectMinimums()

true if it will detector local minimums

isDetectMaximums

public boolean isDetectMaximums()

true if it will detector local maximums

setSearchRadius

public void setSearchRadius(int radius)

Species the search radius for the feature

Parameters:

radius - Radius in pixels