wavelet
Class WindowedSegmentation

java.lang.Object
  |
  +--wavelet.WindowedSegmentation

public class WindowedSegmentation

extends java.lang.Object

Breaks up a sound sample into small windows.

Field Summary

int	numWindows
boolean	occupied
java.util.Vector	segments
double	thresholdPercent
Transitions	transitions
Wavelet	waveletWindows
WaveletFilter	wavFilter
static int	windowSize
static int	winIncrement

Constructor Summary

WindowedSegmentation(int minWindows1)
constructor given minWindows

Method Summary

void	analyzeTransitions() calculate which of the transitions between windows can be considered segmentation points.
void	blank() clears out all the currently calculated segmentation and transition data
int[][]	createTransList() creates list of transitions by calculating differences between wavelet coefficient energies.
double[]	doCompat(int inx) returns an array of how compatible each segment is to the one denoted by inx
void	fourFrameEuclideanDistance() routine for determining the difference between windows dependant on their wavelet energies.
double	fourFrameEuclideanDistance(int a, int b) routine for determining the difference between two specfic windows dependant on their wavelet energies.
void	initialize(sound.AudioSample audiosample) initializes class
static void	main(java.lang.String[] args)
void	normalize() normalize each transition over all wavelet levels (6 at present)
void	separate(int waveletFilter, javax.swing.DefaultBoundedRangeModel rangeModel, javax.swing.JProgressBar progressBar) main signal separation routine
void	separateWithTransitions(int waveletFilter, Transitions obj) used if I already know the transitions (ie loading presets)

Methods inherited from class java.lang.Object

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

Field Detail

windowSize

public static final int windowSize

winIncrement

public static final int winIncrement

numWindows

public int numWindows

waveletWindows

public Wavelet waveletWindows

thresholdPercent

public double thresholdPercent

wavFilter

public WaveletFilter wavFilter

transitions

public Transitions transitions

occupied

public boolean occupied

segments

public java.util.Vector segments

Constructor Detail

WindowedSegmentation

public WindowedSegmentation(int minWindows1)

constructor given minWindows

Parameters:

minWindows - how few windows can be in 1 segment

Method Detail

blank

public void blank()

clears out all the currently calculated segmentation and transition data

initialize

public void initialize(sound.AudioSample audiosample)

initializes class

Parameters:

audioSample - audio signal to process

normalize

public void normalize()

normalize each transition over all wavelet levels (6 at present)

createTransList

public int[][] createTransList()

creates list of transitions by calculating differences between wavelet coefficient energies.

main

public static void main(java.lang.String[] args)

separateWithTransitions

public void separateWithTransitions(int waveletFilter,
                                    Transitions obj)

used if I already know the transitions (ie loading presets)

Parameters:

waveletFilter - wavelet filter to use

obj - transitions object to use

separate

public void separate(int waveletFilter,
                     javax.swing.DefaultBoundedRangeModel rangeModel,
                     javax.swing.JProgressBar progressBar)

main signal separation routine

Parameters:

waveletFilter - wavelet filter to use

rangeModel - for use in the update window

progressBar - for use in the update window

analyzeTransitions

public void analyzeTransitions()

calculate which of the transitions between windows can be considered segmentation points.

fourFrameEuclideanDistance

public void fourFrameEuclideanDistance()

routine for determining the difference between windows dependant on their wavelet energies.

fourFrameEuclideanDistance

public double fourFrameEuclideanDistance(int a,
                                         int b)

routine for determining the difference between two specfic windows dependant on their wavelet energies.

Parameters:

a - window number a

b - window number b

Returns:

difference

doCompat

public double[] doCompat(int inx)

returns an array of how compatible each segment is to the one denoted by inx

Parameters:

inx - reference window number

Returns:

array of differences of all other windows to input one.