s373.flob
Class Flob

java.lang.Object
  s373.flob.Flob

public class Flob
extends java.lang.Object

flob is a flood-fill multi-blob detector,
tracks blobs in image streams,their centroids and bounding boxes

Author:

andre sier
url: http://s373.net/code/flob
version 0.0.1m

Field Summary

int[]	backgroundPixels
int	blobpixmax
int	blobpixmin
static int	BLUE
java.lang.String	bluestr
int	blur
int	colormode
static int	CONTINUOUS_DIFFERENCE
boolean	coordsmode
static int	GREEN
java.lang.String	greenstr
ImageBlobs	imageblobs
static int	LUMA601
java.lang.String	luma601str
static int	LUMA709
java.lang.String	luma709str
static int	LUMAUSER
float[]	lumausercoefs
java.lang.String	lumausrstr
boolean	mirrorX
boolean	mirrorY
int	numPixels
int	om
int	presence
int	pvideotexmode
static int	RED
java.lang.String	redstr
static int	STATIC_DIFFERENCE
boolean[]	trackfeatures
java.lang.String	VERSION
int	videofade
processing.core.PImage	videoimg
int	videoresh
int	videoresw
processing.core.PImage	videotex
processing.core.PImage	videotexbin
boolean	videotexchange
processing.core.PImage	videoteximgmotion
int	videotexmode
processing.core.PImage	videotexmotion
int	videothresh
int	worldheight
int	worldwidth

Constructor Summary

Flob(processing.core.PApplet theParent)

Flob(processing.core.PApplet theParent, processing.core.PImage video)
calling the constructor with a PApplet, PImage parameters inits the image vars

Flob(processing.core.PApplet theParent, processing.core.PImage video, int w, int h)
calling the constructor with a PApplet, PImage, yourWidth, yourHeight parameters inits the image vars yourWidth will be default coords return on flob for each blob's x pos, the same for height

Method Summary

processing.core.PImage	binarize(processing.core.PImage video) first pass of the flob engine. revised in version 001l to allow different color channel tracking. transforms the input image in a black and white only image (binary image). optionally insert a fastblur in the image.
java.util.ArrayList<ABlob>	calc(processing.core.PImage img) calcs with current PImage.
java.util.ArrayList<trackedBlob>	calcsimple(processing.core.PImage img) calcsimple is naive tracking.
processing.core.PImage	fastblur(processing.core.PImage img, int radius) PImage img = fastblur(PImage img, int radius); Super Fast Blur v1.1 by Mario Klingemann http://incubator.quasimondo.com
ABlob	getABlob(int i) getABlob returns the nth calc'ed blob of the tracker returns one ABlob element
boolean	getAnyFeatureActive() getAnyFeatureActive true if any feature points on
int[]	getBackground() gets the background image
int	getBlur() get the blur amount on the image.
int[]	getBox(int i) getBox int i.
float[]	getCentroid(int i) getCentroid int i.
float[]	getCentroidPixelcount(int i) getCentroidPixelcount int i.
float[]	getCentroids() getCentroids.
java.lang.String	getColorMode() get selected colormode
boolean	getCoordsMode() get the coords mode for the blobs returns.
float[]	getDim(int i) getDim int i.
int	getFade(int t) get the fade value to continuos_difference mode
processing.core.PImage	getImage() getImage gets the current video image worked inside flob
float[]	getLumaUserCoefs() get lumausercoefs
int	getMaxNumPixels() get max numpixels to be considered a blob
int	getMinNumPixels() get min numpixels to be considered a blob
boolean[]	getMirror() get mirror in XY axis
int	getNumBlobs() getNumBlobs.
int	getNumTrackedBlobs()
int	getNumTrackedSimpleBlobs()
int	getOm() get the om either CONTINUOUS_DIFFERENCE (1) or STATIC_DIFFERENCE (0)
int	getPresence() getPresence.
float	getPresencef() getPresencef.
ABlob	getPreviousABlob(int i) getPreviousABlob returns the nth calc'ed previous blob of the tracker returns one ABlob element
float[]	getPreviousCentroids()
float[]	getPreviousCurrentCentroid(int i) getPreviousCurrentCentroid int i.
float[]	getPreviousCurrentCentroidMass(int i) getPreviousCurrentCentroidMass int i.
trackedBlob	getPreviousTrackedBlob(int i) getPreviousTrackedBlob returns the nth tracked previous blob of the tracker returns one trackedBlob element
float	getSmooth() get smooth of blob speeds
processing.core.PImage	getSrcImage() returns the updated videotex (in case it needs updating)
int	getThresh() get the threshold value to the image binarization
trackedBlob	getTrackedBlob(int i) getTrackedBlob returns the nth tracked blob of the tracker returns the arraylist of trackedBlob elements a tracked blob holds: // pos & vel & dim results are local world coords // int tb.id; // float tb.cx; // float tb.cy; // float tb.velx; // float tb.vely; // float tb.prevelx; // float tb.prevely; // int tb.presencetime; // float tb.dimx; // float tb.dimy; // int tb.birthtime;
float[]	getTrackedBlobf(int i) getTrackedBlobf returns the data of the nth tracked blob of the tracker as float[] returns the arraylist of trackedBlob elements
float[]	getTrackedSimpleBlob(int i)
boolean[]	getTrackFeatures() getTrackFeatures gets the boolean array with on/off's for searching feature points: armleft, armright, head, bottom for each blob
float	getTrackingMinDist() get tracking min dist to be the same blob
void	mirrorX(boolean m) mirror video data along X axis
void	mirrorY(boolean m) mirror video data along Y axis
void	setBackground(processing.core.PImage video) sets the background to compare to to this PImage
void	setBlur(int blur) set the blur amount on the image.
void	setColorMode(int t) set the colormode for the binarization stage.
void	setCoordsMode(boolean t) set the coords mode for the blobs returns.
void	setFade(int t) set the fade value to continuos_difference mode
void	setImage(int t) setImage sets the videotex returned by flob.videotex or flob.getSrcImage
void	setLumaUserCoefs(float[] data) set lumausercoefs
void	setMaxNumPixels(int t) set max numpixels to be considered a blob
void	setMinNumPixels(int t) set min numpixels to be considered a blob
void	setMirror(boolean m0, boolean m1) set mirror in XY axis with two booleans
void	setOm(int t) set the om either CONTINUOUS_DIFFERENCE (1) or STATIC_DIFFERENCE (0)
void	setSmooth(float s) set smooth of blob speeds
void	setSrcImage(int t) setSrcImage sets the videotex returned by flob.videotex or flob.getSrcImage
void	setThresh(int t) set the threshold value to the image binarization
void	setTrackFeatures(boolean[] tf) setTrackFeatures turns on/off searching for feature points: armleft, armright, head, bottom for each blob
void	setTrackingMinDist(float s) set tracking min dist to be the same blob
void	setTresh(int t) set the threshold value to the image binarization.
void	setVideoTex(int t) set the videotex returned by flob.videotex case 0: videotex = videoimg; break; case 1: videotex = videotexbin; break; case 2: videotex = videotexmotion; break; case 3: videotex = videoteximgmotion; break;
boolean	testPos(float x, float y) testPos float x, float y.
boolean	testPos(int x, int y) testPos int x, int y.
java.util.ArrayList<trackedBlob>	track(processing.core.PImage img) calcs with current PImage.
java.util.ArrayList<trackedBlob>	tracksimple(processing.core.PImage img) tracksimple is good tracking code, maintains id's, speed's, presencetime for each trackedBlob returns the arraylist of trackedBlob elements
processing.core.PImage	updateVideoTex()
java.lang.String	version() return the version of the library.

Methods inherited from class java.lang.Object

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

Field Detail

imageblobs

public ImageBlobs imageblobs

videoimg

public processing.core.PImage videoimg

videotex

public processing.core.PImage videotex

videotexmotion

public processing.core.PImage videotexmotion

videotexbin

public processing.core.PImage videotexbin

videoteximgmotion

public processing.core.PImage videoteximgmotion

backgroundPixels

public int[] backgroundPixels

numPixels

public int numPixels

videoresw

public int videoresw

videoresh

public int videoresh

presence

public int presence

videotexmode

public int videotexmode

pvideotexmode

public int pvideotexmode

videotexchange

public boolean videotexchange

videothresh

public int videothresh

videofade

public int videofade

mirrorX

public boolean mirrorX

mirrorY

public boolean mirrorY

worldwidth

public int worldwidth

worldheight

public int worldheight

coordsmode

public boolean coordsmode

blur

public int blur

trackfeatures

public boolean[] trackfeatures

om

public int om

STATIC_DIFFERENCE

public static final int STATIC_DIFFERENCE

See Also:

Constant Field Values

CONTINUOUS_DIFFERENCE

public static final int CONTINUOUS_DIFFERENCE

See Also:

Constant Field Values

colormode

public int colormode

RED

public static final int RED

See Also:

Constant Field Values

GREEN

public static final int GREEN

See Also:

Constant Field Values

BLUE

public static final int BLUE

See Also:

Constant Field Values

LUMA601

public static final int LUMA601

See Also:

Constant Field Values

LUMA709

public static final int LUMA709

See Also:

Constant Field Values

LUMAUSER

public static final int LUMAUSER

See Also:

Constant Field Values

lumausercoefs

public float[] lumausercoefs

blobpixmin

public int blobpixmin

blobpixmax

public int blobpixmax

redstr

public java.lang.String redstr

greenstr

public java.lang.String greenstr

bluestr

public java.lang.String bluestr

luma601str

public java.lang.String luma601str

luma709str

public java.lang.String luma709str

lumausrstr

public java.lang.String lumausrstr

VERSION

public java.lang.String VERSION

Constructor Detail

Flob

public Flob(processing.core.PApplet theParent)

Flob

public Flob(processing.core.PApplet theParent,
            processing.core.PImage video)

calling the constructor with a PApplet, PImage parameters inits the image vars

Parameters:

PApplet, - PImage

Flob

public Flob(processing.core.PApplet theParent,
            processing.core.PImage video,
            int w,
            int h)

calling the constructor with a PApplet, PImage, yourWidth, yourHeight parameters inits the image vars yourWidth will be default coords return on flob for each blob's x pos, the same for height

Parameters:

PApplet, - PImage

Method Detail

binarize

public processing.core.PImage binarize(processing.core.PImage video)

first pass of the flob engine.
revised in version 001l to allow different color channel tracking.
transforms the input image in a black and white only image (binary image).
optionally insert a fastblur in the image. (if setBlur > 0, blur has that radius)
// nice fastblur insertion, thanks to fausto fonseca for showing the code,
// and to eduardo pinto for pissing me about it. it's fast and great!
// fast blur code by Mario Klingemann
returns a binary image suitable for the calc engine.

built in fastblur filter if blurRadius > 0.

Returns:

PImage

getSrcImage

public processing.core.PImage getSrcImage()

returns the updated videotex (in case it needs updating)

Returns:

PImage

updateVideoTex

public processing.core.PImage updateVideoTex()

setVideoTex

public void setVideoTex(int t)

set the videotex returned by flob.videotex
case 0: videotex = videoimg; break;
case 1: videotex = videotexbin; break;
case 2: videotex = videotexmotion; break;
case 3: videotex = videoteximgmotion; break;

setImage

public void setImage(int t)

setImage sets the videotex returned by flob.videotex or flob.getSrcImage

getImage

public processing.core.PImage getImage()

getImage gets the current video image worked inside flob

Returns:

void

setSrcImage

public void setSrcImage(int t)

setSrcImage sets the videotex returned by flob.videotex or flob.getSrcImage

setTrackFeatures

public void setTrackFeatures(boolean[] tf)

setTrackFeatures turns on/off searching for feature points: armleft, armright, head, bottom for each blob

getTrackFeatures

public boolean[] getTrackFeatures()

getTrackFeatures gets the boolean array with on/off's for searching feature points: armleft, armright, head, bottom for each blob

Returns:

boolean[]

getAnyFeatureActive

public boolean getAnyFeatureActive()

getAnyFeatureActive true if any feature points on

Returns:

boolean

setOm

public void setOm(int t)

set the om either CONTINUOUS_DIFFERENCE (1) or STATIC_DIFFERENCE (0)

getOm

public int getOm()

get the om either CONTINUOUS_DIFFERENCE (1) or STATIC_DIFFERENCE (0)

Returns:

int

setCoordsMode

public void setCoordsMode(boolean t)

set the coords mode for the blobs returns. if true, will scale to global world coordinates, if false, each blob returns normalized coordinates

getCoordsMode

public boolean getCoordsMode()

get the coords mode for the blobs returns. if true, will scale to global world coordinates, if false, each blob returns normalized coordinates

Returns:

boolean

setColorMode

public void setColorMode(int t)

set the colormode for the binarization stage. how to consider a diff pix from background on which channel. red, green, blue, luma

getColorMode

public java.lang.String getColorMode()

get selected colormode

Returns:

int

setLumaUserCoefs

public void setLumaUserCoefs(float[] data)

set lumausercoefs

getLumaUserCoefs

public float[] getLumaUserCoefs()

get lumausercoefs

Returns:

float[]

setBackground

public void setBackground(processing.core.PImage video)

sets the background to compare to to this PImage

getBackground

public int[] getBackground()

gets the background image

Returns:

int[]

setTresh

public void setTresh(int t)

set the threshold value to the image binarization. missing h for backwards compatibility

setThresh

public void setThresh(int t)

set the threshold value to the image binarization

getThresh

public int getThresh()

get the threshold value to the image binarization

Returns:

int

setFade

public void setFade(int t)

set the fade value to continuos_difference mode

getFade

public int getFade(int t)

get the fade value to continuos_difference mode

Returns:

int

mirrorX

public void mirrorX(boolean m)

mirror video data along X axis

mirrorY

public void mirrorY(boolean m)

mirror video data along Y axis

setMirror

public void setMirror(boolean m0,
                      boolean m1)

set mirror in XY axis with two booleans

getMirror

public boolean[] getMirror()

get mirror in XY axis

Returns:

boolean[]

setMinNumPixels

public void setMinNumPixels(int t)

set min numpixels to be considered a blob

setMaxNumPixels

public void setMaxNumPixels(int t)

set max numpixels to be considered a blob

getMinNumPixels

public int getMinNumPixels()

get min numpixels to be considered a blob

Returns:

int

getMaxNumPixels

public int getMaxNumPixels()

get max numpixels to be considered a blob

Returns:

int

setTrackingMinDist

public void setTrackingMinDist(float s)

set tracking min dist to be the same blob

getTrackingMinDist

public float getTrackingMinDist()

get tracking min dist to be the same blob

Returns:

float

setSmooth

public void setSmooth(float s)

set smooth of blob speeds

getSmooth

public float getSmooth()

get smooth of blob speeds

Returns:

float

setBlur

public void setBlur(int blur)

set the blur amount on the image. 0 = off, > 5 high blur

getBlur

public int getBlur()

get the blur amount on the image. 0 = off, > 5 high blur

Returns:

int

calc

public java.util.ArrayList<ABlob> calc(processing.core.PImage img)

calcs with current PImage. PImage must be binary image by this stage. returns the arraylist of the blobs

Returns:

ArrayList

track

public java.util.ArrayList<trackedBlob> track(processing.core.PImage img)

calcs with current PImage. PImage must be binary image by this stage. returns the arraylist of trackedBlob elements

Returns:

ArrayList

tracksimple

public java.util.ArrayList<trackedBlob> tracksimple(processing.core.PImage img)

tracksimple is good tracking code, maintains id's, speed's, presencetime for each trackedBlob returns the arraylist of trackedBlob elements

Returns:

ArrayList

calcsimple

public java.util.ArrayList<trackedBlob> calcsimple(processing.core.PImage img)

calcsimple is naive tracking. works good in stable configs. returns the arraylist of trackedBlob elements

Returns:

ArrayList

getTrackedBlob

public trackedBlob getTrackedBlob(int i)

getTrackedBlob returns the nth tracked blob of the tracker
returns the arraylist of trackedBlob elements

a tracked blob holds:

// pos & vel & dim results are local world coords

// int tb.id;
// float tb.cx;
// float tb.cy;
// float tb.velx;
// float tb.vely;
// float tb.prevelx;
// float tb.prevely;
// int tb.presencetime;
// float tb.dimx;
// float tb.dimy;
// int tb.birthtime;

Returns:

trackedBlob

getPreviousTrackedBlob

public trackedBlob getPreviousTrackedBlob(int i)

getPreviousTrackedBlob returns the nth tracked previous blob of the tracker
returns one trackedBlob element

getABlob

public ABlob getABlob(int i)

getABlob returns the nth calc'ed blob of the tracker
returns one ABlob element

Returns:

ABlob

getPreviousABlob

public ABlob getPreviousABlob(int i)

getPreviousABlob returns the nth calc'ed previous blob of the tracker
returns one ABlob element

Returns:

ABlob

getTrackedSimpleBlob

public float[] getTrackedSimpleBlob(int i)

getNumBlobs

public int getNumBlobs()

getNumBlobs. should be called after calc.

Returns:

int

getNumTrackedBlobs

public int getNumTrackedBlobs()

getNumTrackedSimpleBlobs

public int getNumTrackedSimpleBlobs()

getTrackedBlobf

public float[] getTrackedBlobf(int i)

getTrackedBlobf returns the data of the nth tracked blob of the tracker as float[] returns the arraylist of trackedBlob elements

Returns:

float[12]

getPresence

public int getPresence()

getPresence. returns the number of active pixels

Returns:

int

getPresencef

public float getPresencef()

getPresencef. returns the normalized number of active pixels

Returns:

float

getCentroids

public float[] getCentroids()

getCentroids. returns all coordinates as normalized floats

Returns:

float[]

getPreviousCentroids

public float[] getPreviousCentroids()

getCentroid

public float[] getCentroid(int i)

getCentroid int i. returns coordinates of this centroid as float[2]

Returns:

float[]

getCentroidPixelcount

public float[] getCentroidPixelcount(int i)

getCentroidPixelcount int i. returns coordinates of this centroid + pixelcount as float[3]

Returns:

float[]

getPreviousCurrentCentroid

public float[] getPreviousCurrentCentroid(int i)

getPreviousCurrentCentroid int i. returns previous and current coordinates of this centroid as normalized float[4]

Returns:

float[]

getPreviousCurrentCentroidMass

public float[] getPreviousCurrentCentroidMass(int i)

getPreviousCurrentCentroidMass int i. returns previous and current coordinates of this centroid as normalized float[4]

Returns:

float[]

getDim

public float[] getDim(int i)

getDim int i. returns dimensions of the bounding box of this centroid as normalized float[2]

Returns:

float[]

getBox

public int[] getBox(int i)

getBox int i. returns coordinates of this centroid's box as int[4] box min x + box min y + box max x + box max y

Returns:

int[]

testPos

public boolean testPos(int x,
                       int y)

testPos int x, int y. tests a point coords x + y in the image map returns true if on a blob, false if not on a blob x and y should be constrained to src video dimensions

Returns:

boolean

testPos

public boolean testPos(float x,
                       float y)

testPos float x, float y. tests a point normalized coords x + y in the image map returns true if on a blob, false if not on a blob

Returns:

boolean

fastblur

public processing.core.PImage fastblur(processing.core.PImage img,
                                       int radius)

PImage img = fastblur(PImage img, int radius); Super Fast Blur v1.1 by Mario Klingemann http://incubator.quasimondo.com

Returns:

PImage

version

public java.lang.String version()

return the version of the library.

Returns:

String