org.openimaj.image.processing.algorithm

Class FourierTransform

java.lang.Object

org.openimaj.image.processing.algorithm.FourierTransform

public class FourierTransform
extends Object

Perform forward and inverse Fast Fourier Transforms on image data. This class computes the result of the transform in polar form (i.e. in terms of phase and magnitude). If you need the result of the transform in the raw complex form, then use the FourierTransformComplex class.

Author:

Jonathon Hare (jsh2@ecs.soton.ac.uk)

Constructor Summary

Constructors

Constructor and Description
FourierTransform(FImage image, boolean centre) Construct Fourier Transform by performing a forward transform on the given image.
FourierTransform(FImage magnitude, FImage phase, boolean centre) Construct Fourier Transform object from the given magnitude and phase images in the frequency domain.

Method Summary

Modifier and Type	Method and Description
FImage	getLogNormalisedMagnitude() Get a log-normalised copy of the magnitude image suitable for displaying.
FImage	getMagnitude()
FImage	getPhase()
FImage	inverse() Perform the inverse FFT using the underlying magnitude and phase images.
boolean	isCentre()
static float[][]	prepareData(FImage input, int rs, int cs, boolean centre) Prepare data for a input to the FFT, padding if necessary.
static float[][]	prepareData(float[][] input, int rs, int cs, boolean centre) Prepare data for a input to the FFT, padding if necessary.
static float[]	prepareData1d(FImage input, int rs, int cs, boolean centre) Prepare data for a input to the FFT, padding if necessary.
static float[]	prepareData1d(float[][] input, int rs, int cs, boolean centre) Prepare data for a input to the FFT, padding if necessary.
static void	unprepareData(float[][] prepared, FImage output, boolean centre) Extract the actual data from prepared data.
static void	unprepareData(float[][] prepared, float[][] output, boolean centre) Extract the actual data from prepared data.
static void	unprepareData(float[] prepared, FImage output, boolean centre) Extract the actual data from prepared data.
static void	unprepareData(float[] prepared, float[][] output, boolean centre) Extract the actual data from prepared data.

Methods inherited from class java.lang.Object

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

Constructor Detail

FourierTransform

public FourierTransform(FImage image,
                        boolean centre)

Construct Fourier Transform by performing a forward transform on the given image. If the centre option is set, the FFT will be re-ordered so that the DC component is in the centre.

Parameters:

image - the image to transform

centre - should the FFT be reordered so the centre is DC component

FourierTransform

public FourierTransform(FImage magnitude,
                        FImage phase,
                        boolean centre)

Construct Fourier Transform object from the given magnitude and phase images in the frequency domain. The resultant object can then be used to construct the image using the inverse() method.

Parameters:

magnitude - the magnitude image

phase - the phase image

centre - is the DC component in the image centre?

Method Detail

prepareData

public static float[][] prepareData(FImage input,
                                    int rs,
                                    int cs,
                                    boolean centre)

Prepare data for a input to the FFT, padding if necessary.

Parameters:

input - input data

rs - desired number of rows

cs - desired number of columns

centre - if true, then the data will be prepared so that the DC component is centred.

Returns:

prepared data

prepareData1d

public static float[] prepareData1d(FImage input,
                                    int rs,
                                    int cs,
                                    boolean centre)

Prepare data for a input to the FFT, padding if necessary. The data is prepared as a packed 1D array.

Parameters:

input - input data

rs - desired number of rows

cs - desired number of columns

centre - if true, then the data will be prepared so that the DC component is centred.

Returns:

prepared data

prepareData

public static float[][] prepareData(float[][] input,
                                    int rs,
                                    int cs,
                                    boolean centre)

Prepare data for a input to the FFT, padding if necessary.

Parameters:

input - input data

rs - desired number of rows

cs - desired number of columns

centre - if true, then the data will be prepared so that the DC component is centered.

Returns:

prepared data

prepareData1d

public static float[] prepareData1d(float[][] input,
                                    int rs,
                                    int cs,
                                    boolean centre)

Prepare data for a input to the FFT, padding if necessary. The data is prepared as a packed 1D array.

Parameters:

input - input data

rs - desired number of rows

cs - desired number of columns

centre - if true, then the data will be prepared so that the DC component is centered.

Returns:

prepared data

unprepareData

public static void unprepareData(float[][] prepared,
                                 FImage output,
                                 boolean centre)

Extract the actual data from prepared data. The output image must have the same number of rows as the prepared data, and half the number of columns.

Parameters:

prepared - the prepared data

output - the output

centre - if true, then the data will be prepared so that the DC component is centered.

unprepareData

public static void unprepareData(float[] prepared,
                                 FImage output,
                                 boolean centre)

Extract the actual data from prepared data. The output image must have the same number of rows as the prepared data, and half the number of columns.

Parameters:

prepared - the prepared data

output - the output

centre - if true, then the data will be prepared so that the DC component is centered.

unprepareData

public static void unprepareData(float[][] prepared,
                                 float[][] output,
                                 boolean centre)

Extract the actual data from prepared data. The output array must have the same number of rows as the prepared data, and half the number of columns.

Parameters:

prepared - the prepared data

output - the output

centre - if true, then the data will be prepared so that the DC component is centered.

unprepareData

public static void unprepareData(float[] prepared,
                                 float[][] output,
                                 boolean centre)

Extract the actual data from prepared data. The output array must have the same number of rows as the prepared data, and half the number of columns.

Parameters:

prepared - the prepared data

output - the output

centre - if true, then the data will be prepared so that the DC component is centered.

inverse

public FImage inverse()

Perform the inverse FFT using the underlying magnitude and phase images. The resultant reconstructed image may need normalisation.

Returns:

the reconstructed image

getLogNormalisedMagnitude

public FImage getLogNormalisedMagnitude()

Get a log-normalised copy of the magnitude image suitable for displaying.

Returns:

a log-normalised copy of the magnitude image

getPhase

public FImage getPhase()

Returns:

the phase image

getMagnitude

public FImage getMagnitude()

Returns:

the magnitude image

isCentre

public boolean isCentre()

Returns:

true if the DC component is in the centre; false otherwise