wavelet
Class Wavelet

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

public class Wavelet

extends java.lang.Object

Main Wavelet Transform Class Description: Performs the wavelet transform on array of doubles

Since:

JDK 1.2.2

Field Summary

javax.sound.sampled.AudioFormat	audioFormat
int	depthT
int	inSignalCutoff
double[][]	wavTree

Constructor Summary

Wavelet(javax.sound.sampled.AudioInputStream audioIn)
Constructs an Wavelet object using AudioInputStream

Wavelet(javax.sound.sampled.AudioInputStream audioIn, int length)
Constructs an Wavelet object using AudioInputStream

Wavelet(double[] inputSig)
Constructs an Wavelet object

Wavelet(double[] inputSig, int inputLength)
Constructs an Wavelet object with length different than array size.

Method Summary

static WaveletFilter	chooseFilter(int n) choose a wavelet filter from an ordered list.
double	detCoefCost(int level) computes the l1 norm of a level of detail coefficients
double[]	detCoefs(int level) returns a copy of one level's detail coefficients
static void	fa1d(double[] a, int aStart, int nA, boolean isFwd, WaveletFilter wfltr, double[] aXf, int aXfStart) intermediate routine to compute discrete wavelet wavelet transform just calls oneDVarStep with a step size of 1
void	forwardTransform(int levels, int inSignalStart, WaveletFilter wFilter) perform a n level wavelet transform on the data in inDataArray
byte[]	getByteData(double[] dat) Utility routine: conversion routine from double to bytes
void	getByteData(double[] dat, byte[] byteArray) Utility routine: conversion routine from double to bytes
double[]	getDoubleData() Get a new copy of the input array.
double[]	getDoubleData(javax.sound.sampled.AudioInputStream audioIn) Get a new copy of the input array.
double[]	inverseMallatTree(WaveletFilter wFilter, double[] mtree) performs an inverse wavelet transform of a mallat tree.
static void	main(java.lang.String[] args)
double[]	mallatTree(WaveletFilter wfilter) Performs the wavelet transform, and returns an array with all of the detail coef's, starting from the root
double[]	mix(Wavelet other, WaveletFilter wfilter) mixes two mallat trees.
void	mixDoub(double[] signalA, double[] signalB, double[] mixOut) mixes two arrays of doubles.
static void	Othermain(java.lang.String[] args) miscallaneous old code
void	printTree(int level) debug utility: prints array to standard out
double[]	smoothCoefs(int level) returns a copy of one level's wavelet (smooth) coefficients

Methods inherited from class java.lang.Object

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

Field Detail

wavTree

public double[][] wavTree

depthT

public int depthT

inSignalCutoff

public int inSignalCutoff

audioFormat

public javax.sound.sampled.AudioFormat audioFormat

Constructor Detail

Wavelet

public Wavelet(double[] inputSig)

Constructs an Wavelet object

Parameters:

intputSignal - The one-dimensional input signal

Wavelet

public Wavelet(double[] inputSig,
               int inputLength)

Constructs an Wavelet object with length different than array size.

Parameters:

intputSignal - The one-dimensional input signal

intputLength - Length of signal to analyze (<= inputSignal.length)

Wavelet

public Wavelet(javax.sound.sampled.AudioInputStream audioIn)

Constructs an Wavelet object using AudioInputStream

Parameters:

audioIn - inputSignal to analyze

Wavelet

public Wavelet(javax.sound.sampled.AudioInputStream audioIn,
               int length)

Constructs an Wavelet object using AudioInputStream

Parameters:

audioIn - inputSignal to analyze

length - Length of signal to analyze (<= audioIn length)

Method Detail

getDoubleData

public double[] getDoubleData()

Get a new copy of the input array.

Returns:

array copy of input signal

getDoubleData

public double[] getDoubleData(javax.sound.sampled.AudioInputStream audioIn)

Get a new copy of the input array.

Returns:

array copy of input signal

getByteData

public void getByteData(double[] dat,
                        byte[] byteArray)

Utility routine: conversion routine from double to bytes

Parameters:

dat - (double array to convert (for instance, the inverse wavelet inverse transform)

byteArray - input double array converted to bytes.

getByteData

public byte[] getByteData(double[] dat)

Utility routine: conversion routine from double to bytes

Parameters:

dat - (double array to convert (for instance, the inverse wavelet inverse transform)

Returns:

input double array converted to new array of bytes.

mallatTree

public double[] mallatTree(WaveletFilter wfilter)

Performs the wavelet transform, and returns an array with all of the detail coef's, starting from the root

Parameters:

wfilter - wavelet filter to use.

Returns:

input double array converted to new array of bytes.

inverseMallatTree

public double[] inverseMallatTree(WaveletFilter wFilter,
                                  double[] mtree)

performs an inverse wavelet transform of a mallat tree.

Parameters:

wfilter - wavelet filter to use

mtree - mallat tree

Returns:

reconstitued signal

forwardTransform

public void forwardTransform(int levels,
                             int inSignalStart,
                             WaveletFilter wFilter)

perform a n level wavelet transform on the data in inDataArray

Parameters:

levels - number of levels of the wavelet transform to do.

inSignalStart - place to start in the input signal

wFilter - wavelet filter to use.

printTree

public void printTree(int level)

debug utility: prints array to standard out

Parameters:

level - level of wavelet transform to print out.

smoothCoefs

public double[] smoothCoefs(int level)

returns a copy of one level's wavelet (smooth) coefficients

Parameters:

level - specify level wavelet transform to return

Returns:

copy of specified level's average coefficients

detCoefs

public double[] detCoefs(int level)

returns a copy of one level's detail coefficients

Parameters:

level - specify level wavelet transform to return

Returns:

copy of specified level's average coefficients

detCoefCost

public double detCoefCost(int level)

computes the l1 norm of a level of detail coefficients

Parameters:

level - specify level wavelet transform to use

Returns:

l1 norm of a level of detail cofficients

fa1d

public static void fa1d(double[] a,
                        int aStart,
                        int nA,
                        boolean isFwd,
                        WaveletFilter wfltr,
                        double[] aXf,
                        int aXfStart)

intermediate routine to compute discrete wavelet wavelet transform just calls oneDVarStep with a step size of 1

Parameters:

a - original array

aStart - index to start on in original array

nA - size of portion of input array to use

isFwd - TRUE <=> forward transform FALSE - inverse

wfltr - wavelet filter to use

aXf - space for transformed array

aXfStart - index to start from in transformed array

mix

public double[] mix(Wavelet other,
                    WaveletFilter wfilter)

mixes two mallat trees.

Parameters:

other - oher mallat tree to mix with (one is implicit)

wflilter - wavelet filter to use

Returns:

new mallat tree which is a mix of the first two

mixDoub

public void mixDoub(double[] signalA,
                    double[] signalB,
                    double[] mixOut)

mixes two arrays of doubles. just a linear crossfade right now.

Parameters:

signalA - first signal

signalB - second signal

mixOut - space to put the result

chooseFilter

public static WaveletFilter chooseFilter(int n)

choose a wavelet filter from an ordered list.

Parameters:

n - wavelet filter number

Returns:

the wavelet filter associated with this number

Othermain

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

miscallaneous old code

Parameters:

args - input args

main

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