org.openimaj.image.feature.local.detector.mser

Class MSERFeatureGenerator

java.lang.Object

org.openimaj.image.feature.local.detector.mser.MSERFeatureGenerator

@Reference(type=Article,
           author={"J Matas","O Chum","M Urban","T Pajdla"},
           title="Robust wide-baseline stereo from maximally stable extremal regions",
           year="2004",
           journal="Image and Vision Computing",
           pages={"761 "," 767"},
           url="http://www.sciencedirect.com/science/article/pii/S0262885604000435",
           number="10",
           volume="22",
           customData={"issn","0262-8856","doi","10.1016/j.imavis.2004.02.006","keywords","Robust metric"})
public class MSERFeatureGenerator
extends Object

Detector for MSER features.

Author:

David Dupplaw (dpd@ecs.soton.ac.uk)

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	MSERFeatureGenerator.MSERDirection A way of representing how the MSER should be processed.

Constructor Summary

Constructors

Constructor and Description
MSERFeatureGenerator(Class<? extends ComponentFeature>... featureClasses) Default constructor
MSERFeatureGenerator(int delta, int maxArea, int minArea, float maxVariation, float minDiversity, Class<? extends ComponentFeature>... featureClasses) Constructor that takes all the parameters for the MSER process.

Method Summary

Modifier and Type	Method and Description
int	countMSERs(TreeNode<Component> tree) Returns a count of the number of components in the tree that are marked as MSERs.
List<Component>	generateMSERs(FImage img) Performs a watershed then an MSER detection on the given image and returns the MSERs.
List<Component>	generateMSERs(FImage img, MSERFeatureGenerator.MSERDirection dir) Performs a watershed then an MSER detection on the given image and returns the MSERs.
int	getDelta()
int	getMaxArea()
float	getMaxVariation()
int	getMinArea()
float	getMinDiversity()
List<Component>	performMSERDetection(List<MergeTreeBuilder> mtbs, MSERFeatureGenerator.MSERDirection dir) Performs MSER detection on the trees provided.
List<MergeTreeBuilder>	performWatershed(FImage img) Perform the watershed algorithm on the given image.
void	setDelta(int delta)
void	setMaxArea(int maxArea)
void	setMaxVariation(float maxVariation)
void	setMinArea(int minArea)
void	setMinDiversity(float minDiversity)

Methods inherited from class java.lang.Object

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

Constructor Detail

MSERFeatureGenerator

@SafeVarargs
public MSERFeatureGenerator(Class<? extends ComponentFeature>... featureClasses)

Default constructor

Parameters:

featureClasses - features to generate for each mser

MSERFeatureGenerator

@SafeVarargs
public MSERFeatureGenerator(int delta,
                                         int maxArea,
                                         int minArea,
                                         float maxVariation,
                                         float minDiversity,
                                         Class<? extends ComponentFeature>... featureClasses)

Constructor that takes all the parameters for the MSER process.

Parameters:

delta -

maxArea -

minArea -

maxVariation -

minDiversity -

featureClasses - features to generate for each mser

Method Detail

generateMSERs

public List<Component> generateMSERs(FImage img)

Performs a watershed then an MSER detection on the given image and returns the MSERs.

Parameters:

img - The image to analyse.

Returns:

A list of Components

generateMSERs

public List<Component> generateMSERs(FImage img,
                                     MSERFeatureGenerator.MSERDirection dir)

Performs a watershed then an MSER detection on the given image and returns the MSERs.

Parameters:

img - The image to analyse.#

dir - The direction which to process the MSERS

Returns:

A list of Components

performWatershed

public List<MergeTreeBuilder> performWatershed(FImage img)

Perform the watershed algorithm on the given image.

Parameters:

img - The image to perform the watershed on

Returns:

A tuple of MergeTreeBuilders (down first, up second)

performMSERDetection

public List<Component> performMSERDetection(List<MergeTreeBuilder> mtbs,
                                            MSERFeatureGenerator.MSERDirection dir)

Performs MSER detection on the trees provided. The input list must be a list containing MergeTreeBuilders, the first being the downward watershed, the second being the upward watershed.

Parameters:

mtbs - The list of MergeTreeBuilders

dir - The direction to detect MSERs from

Returns:

A list of Components

countMSERs

public int countMSERs(TreeNode<Component> tree)

Returns a count of the number of components in the tree that are marked as MSERs.

Parameters:

tree - The tree to count MSERs in

Returns:

the count

getDelta

public int getDelta()

Returns:

the delta

setDelta

public void setDelta(int delta)

Parameters:

delta - the delta to set

getMaxArea

public int getMaxArea()

Returns:

the maxArea

setMaxArea

public void setMaxArea(int maxArea)

Parameters:

maxArea - the maxArea to set

getMinArea

public int getMinArea()

Returns:

the minArea

setMinArea

public void setMinArea(int minArea)

Parameters:

minArea - the minArea to set

getMaxVariation

public float getMaxVariation()

Returns:

the maxVariation

setMaxVariation

public void setMaxVariation(float maxVariation)

Parameters:

maxVariation - the maxVariation to set

getMinDiversity

public float getMinDiversity()

Returns:

the minDiversity

setMinDiversity

public void setMinDiversity(float minDiversity)

Parameters:

minDiversity - the minDiversity to set