This interface provide functionalities of blob analysis. More...

Inheritance diagram for ILBlobAnalysis:

Public Member Functions
void	AddBackgroundRange (int grayLB, int grayUB)

void	AddBackgroundRangeColor (LColor colorLB, struct LColor colorUB)

void	AddBlobFilter (ILBlobFilter *filter)

void	AddBlobRange (int grayLB, int grayUB)

void	AddBlobRangeColor (LColor colorLB, struct LColor colorUB)

LPVErrorCode	Build (ILImage img, ILRegion region, ILBlobResults **results)

int	CountBackgroundRange ()

int	CountBackgroundRangeColor ()

int	CountBlobFilter ()

int	CountBlobRange ()

int	CountBlobRangeColor ()

void	ItemBackgroundRange (int index, int grayLB, int grayUB)

void	ItemBackgroundRangeColor (int index, LColor colorLB, struct LColor colorUB)

ILBlobFilter *	ItemBlobFilter (int index)

void	ItemBlobRange (int index, int grayLB, int grayUB)

void	ItemBlobRangeColor (int index, LColor colorLB, struct LColor colorUB)

void	Preview (ILImage img, ILImage outputImg)

void	ResetBackgroundRange ()

void	ResetBackgroundRangeColor ()

void	ResetBlobFilter ()

void	ResetBlobRange ()

void	ResetBlobRangeColor ()

void	SelectRangeAuto (ILImage image, int x, int y, double zoomX, double zoomY, double panX, double panY, int tolerance, int grayLB, int grayUB, ILImage rangeMask)

void	SelectRangeAutoColor (ILImage image, int x, int y, double zoomX, double zoomY, double panX, double panY, int tolerance, LColor colorLB, struct LColor colorUB, ILImage rangeMask)

void	SetRangeThresholdAdaptGlobal (LPVPolarity blobPolarity)

void	SetRangeThresholdAdaptLocal (LPVPolarity blobPolarity, int blockSize, int bias)

void	SetSoftThresholdRange (int softRange)

Public Member Functions inherited from ILObject
ILObject *	Copy ()

LPVErrorCode	Load (LString filename)

void	Reset ()

LPVErrorCode	Save (LString filename)

BOOL	Valid ()

Properties
LPVPolarity	AdaptBlobPolarity `[get]`
	The blob's polarity defined for the global or local adaptive thresholding in segmentation. More...

LPVImageFormat	ColorMode `[get, set]`
	The mode of the blob analysis tool, either handling grayscale or color image.

LPVBlobContourType	ContourType `[get, set]`
	The type of the result blob contours, see LPVBlobContourType. More...

BOOL	EnableMerge `[get, set]`
	Whether to enable blob merging.

LPVExcludeBoundaryMode	ExcludeBoundaryMode `[get, set]`
	The mode how to handle blobs that intersect the region boundary (exclude the entire blob, exclude only those outset pixels, or do nothing), see LPVExcludeBoundaryMode.

BOOL	FillHole `[get, set]`
	Whether to enable the contribution of holes in the overall blob statistics. By default, it's off.

int	Hierarchy `[get, set]`
	The hierarchy of the result blob contours. It's 0-based, negative value means all hierarchies. More...

int	LocalBias `[get]`
	The revised value to generate a local adaptive threshold. More...

int	LocalBlockSize `[get]`
	The size of the local region used in local adaptive thresholding. More...

int	MaxCount `[get, set]`
	The maximum of blobs to find. If there are more than the setting counts of blobs in the detect region, sort the candidates then pick the tops. More...

int	MergeAreaLB `[get, set]`
	The minimum acceptable area for the merged blobs. Area is the optional filtering property for the merged results and is not used to control the merging process. To constrain the size of the merged result, you should set the merge size bounds: MergeWidthLB, MergeWidthUB, MergeHeightLB, MergeHeightUB.

int	MergeAreaUB `[get, set]`
	The maximum acceptable area for the merged blobs.

int	MergeHeightLB `[get, set]`
	The minimum acceptable height (length of shorter axis) for the merged blobs.

int	MergeHeightUB `[get, set]`
	The maximum acceptable height (length of shorter axis) for the merged blobs.

LPVBlobMergeType	MergeType `[get, set]`
	The way to do the blob merging.

int	MergeWidthLB `[get, set]`
	The minimum acceptable width (length of longer axis) for the merged blobs.

int	MergeWidthUB `[get, set]`
	The maximum acceptable width (length of longer axis) for the merged blobs.

LPVBlobFeatures	SortBy `[get, set]`
	The method used to sort all blob candidates, the top ones are reported as final results. More...

BOOL	SortByDescending `[get, set]`
	Works together with SortBy. True to sort all blob candidates descending(from large value to small), false for sorting in ascending order. More...

LPVThresholdType	ThresholdType `[get]`
	The thresholding method used in segmentation. More...

Detailed Description

This interface provide functionalities of blob analysis.

Blob analysis is a useful approach to segment object from background into connected components of similar grayscale values. Various geometric features, such as area, perimeter or circularity, can be computed on these blobs. In addition, image-related features, such as average luminance or contrast, can be computed based on the image attached.

To use this interface, you should create a LBlobAnalysis object.

Example Code

Member Function Documentation

◆ AddBackgroundRange()

void AddBackgroundRange	(	int	grayLB,
		int	grayUB
	)

Add grayscale value range for background. Pixels in this range will be treated as background.
It indicates this is a grayscale-mode blob analysis tool, all the previous color range setting will be cleaned up.

Parameters

[in]	grayLB	The lower bound of the grayscale value range in 0 ~ 255.
[in]	grayUB	The upper bound of the grayscale value range in 0 ~ 255.

See also: AddBackgroundRangeColor(), AddBlobRange(), SelectRangeAuto().

◆ AddBackgroundRangeColor()

void AddBackgroundRangeColor	(	LColor	colorLB,
		struct LColor	colorUB
	)

Add color value range for background. Pixels in this range will be treated as background.
It indicates this is a color-mode blob analysis tool, all the previous grayscale range setting will be cleaned up.

Parameters

[in]	colorLB	The lower bound of the color value range. It's in the same channel order and color space as the source image.
[in]	colorUB	The upper bound of the color value range. It's in the same channel order and color space as the source image.

See also: AddBackgroundRange(), AddBlobRangeColor(), SelectRangeAutoColor().

◆ AddBlobFilter()

void AddBlobFilter ( ILBlobFilter * filter )

Add an alternative blob filter.

Parameters

[in] filter The new blob filter

◆ AddBlobRange()

void AddBlobRange	(	int	grayLB,
		int	grayUB
	)

Add grayscale value range for blob objects. Pixels in this range will be treated as foreground, and connected to be blobs.
It indicates this is a grayscale-mode blob analysis tool, all the previous color range setting will be cleaned up.

Parameters

[in]	grayLB	The lower bound of the grayscale value range in 0 ~ 255.
[in]	grayUB	The upper bound of the grayscale value range in 0 ~ 255.

See also: AddBlobRangeColor(), AddBackgroundRange(), SelectRangeAuto().

◆ AddBlobRangeColor()

void AddBlobRangeColor	(	LColor	colorLB,
		struct LColor	colorUB
	)

Add color value range for blob objects. Pixels in this range will be treated as foreground, and connected to be blobs.
It indicates this is a color-mode blob analysis tool, all the previous grayscale range setting will be cleaned up.

Parameters

[in]	colorLB	The lower bound of the color value range. It's in the same channel order and color space as the source image.
[in]	colorUB	The upper bound of the color value range. It's in the same channel order and color space as the source image.

See also: AddBlobRange(), AddBackgroundRangeColor(), SelectRangeAutoColor().

◆ Build()

LPVErrorCode Build	(	ILImage *	img,
		ILRegion *	region,
		ILBlobResults **	results
	)

Build the blobs using the input image and the provided region.

Parameters

[in]	img	The input image. You should pass in the images of the matching format, aka. grayscale images for grayscale blob analysis mode. Otherwise, we will do the conversion automatically according to the current blob analysis mode.
[in]	region	The input region.
[out]	results	Return the blob results.

Return values

error Return error code if anything is wrong.

◆ CountBackgroundRange()

int CountBackgroundRange ( )

Get count of the grayscale value ranges for background.

◆ CountBackgroundRangeColor()

int CountBackgroundRangeColor ( )

Get count of the color value ranges for background.

◆ CountBlobFilter()

int CountBlobFilter ( )

Get count of the blob filters.

◆ CountBlobRange()

int CountBlobRange ( )

Get count of the grayscale value ranges for blob objects.

◆ CountBlobRangeColor()

int CountBlobRangeColor ( )

Get count of the color value ranges for blob objects.

◆ ItemBackgroundRange()

void ItemBackgroundRange	(	int	index,
		int *	grayLB,
		int *	grayUB
	)

Get the grayscale value ranges for background, with the given index.

Parameters

[in]	index	The input index, it should be a 0-based number smaller than total count.
[out]	grayLB	Return the lower bound of the grayscale value range.
[out]	grayUB	Return the upper bound of the grayscale value range.

◆ ItemBackgroundRangeColor()

void ItemBackgroundRangeColor	(	int	index,
		LColor *	colorLB,
		struct LColor *	colorUB
	)

Get the color value ranges for background, with the given index.

Parameters

[in]	index	The input index, it should be a 0-based number smaller than total count.
[out]	colorLB	Return the lower bound of the color value range.
[out]	colorUB	Return the upper bound of the color value range.

◆ ItemBlobFilter()

ILBlobFilter* ItemBlobFilter ( int index )

Reset the blob filters, all settings are cleaned up.

◆ ItemBlobRange()

void ItemBlobRange	(	int	index,
		int *	grayLB,
		int *	grayUB
	)

Get the grayscale value ranges for blob objects, with the given index.

Parameters

[in]	index	The input index, it should be a 0-based number smaller than total count.
[out]	grayLB	Return the lower bound of the grayscale value range.
[out]	grayUB	Return the upper bound of the grayscale value range.

◆ ItemBlobRangeColor()

void ItemBlobRangeColor	(	int	index,
		LColor *	colorLB,
		struct LColor *	colorUB
	)

Get the color value ranges for blob objects, with the given index.

Parameters

[in]	index	The input index, it should be a 0-based number smaller than total count.
[out]	colorLB	Return the lower bound of the color value range.
[out]	colorUB	Return the upper bound of the color value range.

◆ Preview()

void Preview	(	ILImage *	img,
		ILImage *	outputImg
	)

Preview the binarized image the input image with current foreground and background range configuration.

Parameters

[in]	img	The input image. You should pass in the images of the matching format, aka. grayscale images for grayscale blob analysis mode. Otherwise, we will do the conversion automatically according to the current blob analysis mode.
[out]	outputImg	Return the preview image.

◆ ResetBackgroundRange()

void ResetBackgroundRange ( )

Reset the grayscale value ranges for background, all settings are cleaned up.

◆ ResetBackgroundRangeColor()

void ResetBackgroundRangeColor ( )

Reset the color value ranges for background, all settings are cleaned up.

◆ ResetBlobFilter()

void ResetBlobFilter ( )

Remove all blob filters.

◆ ResetBlobRange()

void ResetBlobRange ( )

Reset the grayscale value ranges for blob objects, all settings are cleaned up.

◆ ResetBlobRangeColor()

void ResetBlobRangeColor ( )

Reset the color value ranges for blob objects, all settings are cleaned up.

◆ SelectRangeAuto()

void SelectRangeAuto	(	ILImage *	image,
		int	x,
		int	y,
		double	zoomX,
		double	zoomY,
		double	panX,
		double	panY,
		int	tolerance,
		int *	grayLB,
		int *	grayUB,
		ILImage *	rangeMask
	)

Select a grayscale value range on the provided image. You should then explicitly add the range as blob or background range. It's usually called for a grayscale-mode blob analysis tool.

Parameters

[in]	image	The input image.
[in]	x	The x-coordinate of the cursor(selected) position.
[in]	y	The y-coordinate of the cursor(selected) position.
[in]	tolerance	The maximum relative difference of neighbor pixels that build the selection
[out]	grayLB	Return the lower bound of the grayscale value range in 0 ~ 255.
[out]	grayUB	Return the upper bound of the grayscale value range in 0 ~ 255.
[out]	rangeMask	Return the mask image of the same size as the input image, white pixels represent the ones in selected grayscale value range in 0 ~ 255. Optional, could be null.

See also: Use Draw Functions

Remarks: This function is currently not available on Linux / MacOS platform.

See also: SelectRangeAutoColor(), AddBlobRange(), AddBackgroundRange().

◆ SelectRangeAutoColor()

void SelectRangeAutoColor	(	ILImage *	image,
		int	x,
		int	y,
		double	zoomX,
		double	zoomY,
		double	panX,
		double	panY,
		int	tolerance,
		LColor *	colorLB,
		struct LColor *	colorUB,
		ILImage *	rangeMask
	)

Select a color value range on the provided image. You should then explicitly add the range as blob or background range.

Parameters

[in]	image	The input image.
[in]	x	The x-coordinate of the cursor(selected) position.
[in]	y	The y-coordinate of the cursor(selected) position.
[in]	tolerance	The maximum relative difference of neighbor pixels that build the selection
[out]	colorLB	Return the lower bound of the color value range. It's in the same channel order and color space as the source image.
[out]	colorUB	Return the upper bound of the color value range. It's in the same channel order and color space as the source image.
[out]	rangeMask	Return the mask image of the same size as the input image, white pixels represent the ones in selected grayscale value range in 0 ~ 255. Optional, could be null.

See also: Use Draw Functions

Remarks: This function is currently not available on Linux / MacOS platform.

See also: SelectRangeAuto(), AddBlobRangeColor(), AddBackgroundRangeColor().

◆ SetRangeThresholdAdaptGlobal()

void SetRangeThresholdAdaptGlobal ( LPVPolarity blobPolarity )

Use global-adaptive thresholding for segmentation. The optimal threshold value is calculated base on the entire input image.

Parameters

[in] blobPolarity The object polarity of the blob. LPVBlackOnWhite indicates black pixels with smaller value than global threshold are the blobs, LPVWhiteOnBlack is exactly the opposite.

◆ SetRangeThresholdAdaptLocal()

void SetRangeThresholdAdaptLocal	(	LPVPolarity	blobPolarity,
		int	blockSize,
		int	bias
	)

Use local-adaptive threshold value. The optimal threshold value is calculated base on a @math_eq{blockSize \times blockSize} region around individual pixel, then revised by the given bias value.

Parameters

[in]	blobPolarity	The object polarity of the blob. LPVBlackOnWhite indicates black pixels with smaller value than global threshold are the blobs, LPVWhiteOnBlack is exactly the opposite.
[in]	blockSize	Size of a pixel's local region, should be an odd number, say, 3, 5, 7, etc. Set it to 0 for auto-generate block size based in the input image size.
[in]	bias	Revise value to optimal threshold value as \( t' = t - bias \) for LPVBlackOnWhite , \( t' = t + bias \) for LPVWhiteOnBlack .

◆ SetSoftThresholdRange()

void SetSoftThresholdRange ( int softRange )

Set the weighted range of soft thresholding. By default, it's 0 which means soft thresholding is not enabled.
It is useful to locate blobs more robust and more accurately.
For example, if we define blob color range to be [100, 200], and soft thresholding range to be 5, then, [0, 95] and [205, 255] are background, [105, 195] are foreground, [96, 104] are weighted as 0.1 ~ 0.9 to be foreground, [196, 204] are weighted as 0.9 ~ 0.1 to be foreground. Soft thresholding results in soft mask of the result blobs, and will affect the calculation of blob's geometric features, such as area or center. Pixels inside the blob contribute to the final result according to its weight.

Parameters

[in] softRange The soft thresholding range, create a [t - softRange, t + softRange] range.

Property Documentation

◆ AdaptBlobPolarity

LPVPolarity AdaptBlobPolarity

get

The blob's polarity defined for the global or local adaptive thresholding in segmentation.

See also: SetRangeThresholdAdaptGlobal(), SetRangeThresholdAdaptLocal()

◆ ContourType

LPVBlobContourType ContourType

getset

The type of the result blob contours, see LPVBlobContourType.

See also: Hierarchy.

◆ Hierarchy

int Hierarchy

getset

The hierarchy of the result blob contours. It's 0-based, negative value means all hierarchies.

See also: ContourType.

◆ LocalBias

int LocalBias

get

The revised value to generate a local adaptive threshold.

See also: SetRangeThresholdAdaptLocal()

◆ LocalBlockSize

int LocalBlockSize

get

The size of the local region used in local adaptive thresholding.

See also: SetRangeThresholdAdaptLocal()

◆ MaxCount

int MaxCount

getset

The maximum of blobs to find. If there are more than the setting counts of blobs in the detect region, sort the candidates then pick the tops.

See also: SortBy

◆ SortBy

LPVBlobFeatures SortBy

getset

The method used to sort all blob candidates, the top ones are reported as final results.

See also: MaxCount, SortByDescending

◆ SortByDescending

BOOL SortByDescending

getset

Works together with SortBy. True to sort all blob candidates descending(from large value to small), false for sorting in ascending order.

See also: SortBy

◆ ThresholdType

LPVThresholdType ThresholdType

get

The thresholding method used in segmentation.

See also: AddBlobRange(), AddBlobRangeColor(), SetRangeThresholdAdaptGlobal(), SetRangeThresholdAdaptLocal()

Public Member Functions

Properties

Detailed Description

Example Code

Member Function Documentation

◆ AddBackgroundRange()

◆ AddBackgroundRangeColor()

◆ AddBlobFilter()

◆ AddBlobRange()

◆ AddBlobRangeColor()

◆ Build()

◆ CountBackgroundRange()

◆ CountBackgroundRangeColor()

◆ CountBlobFilter()

◆ CountBlobRange()

◆ CountBlobRangeColor()

◆ ItemBackgroundRange()

◆ ItemBackgroundRangeColor()

◆ ItemBlobFilter()

◆ ItemBlobRange()

◆ ItemBlobRangeColor()

◆ Preview()

◆ ResetBackgroundRange()

◆ ResetBackgroundRangeColor()

◆ ResetBlobFilter()

◆ ResetBlobRange()

◆ ResetBlobRangeColor()

◆ SelectRangeAuto()

◆ SelectRangeAutoColor()

◆ SetRangeThresholdAdaptGlobal()

◆ SetRangeThresholdAdaptLocal()

◆ SetSoftThresholdRange()

Property Documentation

◆ AdaptBlobPolarity

◆ ContourType

◆ Hierarchy

◆ LocalBias

◆ LocalBlockSize

◆ MaxCount

◆ SortBy

◆ SortByDescending

◆ ThresholdType