Wavelet Transforms¶
This chapter describes functions for performing Discrete Wavelet
Transforms (DWTs). The library includes wavelets for real data in both
one and two dimensions. The wavelet functions are declared in the header
files gsl_wavelet.h
and gsl_wavelet2d.h
.
Definitions¶
The continuous wavelet transform and its inverse are defined by the relations,
and,
where the basis functions are obtained by scaling and translation from a single function, referred to as the mother wavelet.
The discrete version of the wavelet transform acts on equallyspaced samples, with fixed scaling and translation steps (, ). The frequency and time axes are sampled dyadically on scales of through a level parameter .
The resulting family of functions constitutes an orthonormal basis for squareintegrable signals. The discrete wavelet transform is an algorithm, and is also referred to as the fast wavelet transform.
Initialization¶

type gsl_wavelet¶
This structure contains the filter coefficients defining the wavelet and any associated offset parameters.

gsl_wavelet *gsl_wavelet_alloc(const gsl_wavelet_type *T, size_t k)¶
This function allocates and initializes a wavelet object of type
T
. The parameterk
selects the specific member of the wavelet family. A null pointer is returned if insufficient memory is available or if a unsupported member is selected.
The following wavelet types are implemented:

type gsl_wavelet_type¶

gsl_wavelet_type *gsl_wavelet_daubechies¶

gsl_wavelet_type *gsl_wavelet_daubechies_centered¶
This is the Daubechies wavelet family of maximum phase with vanishing moments. The implemented wavelets are , with
k
even.

gsl_wavelet_type *gsl_wavelet_haar¶

gsl_wavelet_type *gsl_wavelet_haar_centered¶
This is the Haar wavelet. The only valid choice of for the Haar wavelet is .

gsl_wavelet_type *gsl_wavelet_bspline¶

gsl_wavelet_type *gsl_wavelet_bspline_centered¶
This is the biorthogonal Bspline wavelet family of order . The implemented values of are 103, 105, 202, 204, 206, 208, 301, 303, 305 307, 309.

gsl_wavelet_type *gsl_wavelet_daubechies¶
The centered forms of the wavelets align the coefficients of the various subbands on edges. Thus the resulting visualization of the coefficients of the wavelet transform in the phase plane is easier to understand.

const char *gsl_wavelet_name(const gsl_wavelet *w)¶
This function returns a pointer to the name of the wavelet family for
w
.

void gsl_wavelet_free(gsl_wavelet *w)¶
This function frees the wavelet object
w
.

type gsl_wavelet_workspace¶
This structure contains scratch space of the same size as the input data and is used to hold intermediate results during the transform.

gsl_wavelet_workspace *gsl_wavelet_workspace_alloc(size_t n)¶
This function allocates a workspace for the discrete wavelet transform. To perform a onedimensional transform on
n
elements, a workspace of sizen
must be provided. For twodimensional transforms ofn
byn
matrices it is sufficient to allocate a workspace of sizen
, since the transform operates on individual rows and columns. A null pointer is returned if insufficient memory is available.

void gsl_wavelet_workspace_free(gsl_wavelet_workspace *work)¶
This function frees the allocated workspace
work
.
Transform Functions¶
This sections describes the actual functions performing the discrete wavelet transform. Note that the transforms use periodic boundary conditions. If the signal is not periodic in the sample length then spurious coefficients will appear at the beginning and end of each level of the transform.
Wavelet transforms in one dimension¶

int gsl_wavelet_transform(const gsl_wavelet *w, double *data, size_t stride, size_t n, gsl_wavelet_direction dir, gsl_wavelet_workspace *work)¶

int gsl_wavelet_transform_forward(const gsl_wavelet *w, double *data, size_t stride, size_t n, gsl_wavelet_workspace *work)¶

int gsl_wavelet_transform_inverse(const gsl_wavelet *w, double *data, size_t stride, size_t n, gsl_wavelet_workspace *work)¶
These functions compute inplace forward and inverse discrete wavelet transforms of length
n
with stridestride
on the arraydata
. The length of the transformn
is restricted to powers of two. For thetransform
version of the function the argumentdir
can be eitherforward
() orbackward
(). A workspacework
of lengthn
must be provided.For the forward transform, the elements of the original array are replaced by the discrete wavelet transform in a packed triangular storage layout, where
j
is the index of the level andk
is the index of the coefficient within each level, . The total number of levels is . The output data has the following form,where the first element is the smoothing coefficient , followed by the detail coefficients for each level . The backward transform inverts these coefficients to obtain the original data.
These functions return a status of
GSL_SUCCESS
upon successful completion.GSL_EINVAL
is returned ifn
is not an integer power of 2 or if insufficient workspace is provided.
Wavelet transforms in two dimension¶
The library provides functions to perform twodimensional discrete wavelet transforms on square matrices. The matrix dimensions must be an integer power of two. There are two possible orderings of the rows and columns in the twodimensional wavelet transform, referred to as the “standard” and “nonstandard” forms.
The “standard” transform performs a complete discrete wavelet transform on the rows of the matrix, followed by a separate complete discrete wavelet transform on the columns of the resulting rowtransformed matrix. This procedure uses the same ordering as a twodimensional Fourier transform.
The “nonstandard” transform is performed in interleaved passes on the rows and columns of the matrix for each level of the transform. The first level of the transform is applied to the matrix rows, and then to the matrix columns. This procedure is then repeated across the rows and columns of the data for the subsequent levels of the transform, until the full discrete wavelet transform is complete. The nonstandard form of the discrete wavelet transform is typically used in image analysis.
The functions described in this section are declared in the header file
gsl_wavelet2d.h
.

int gsl_wavelet2d_transform(const gsl_wavelet *w, double *data, size_t tda, size_t size1, size_t size2, gsl_wavelet_direction dir, gsl_wavelet_workspace *work)¶

int gsl_wavelet2d_transform_forward(const gsl_wavelet *w, double *data, size_t tda, size_t size1, size_t size2, gsl_wavelet_workspace *work)¶

int gsl_wavelet2d_transform_inverse(const gsl_wavelet *w, double *data, size_t tda, size_t size1, size_t size2, gsl_wavelet_workspace *work)¶
These functions compute twodimensional inplace forward and inverse discrete wavelet transforms in standard form on the array
data
stored in rowmajor form with dimensionssize1
andsize2
and physical row lengthtda
. The dimensions must be equal (square matrix) and are restricted to powers of two. For thetransform
version of the function the argumentdir
can be eitherforward
() orbackward
(). A workspacework
of the appropriate size must be provided. On exit, the appropriate elements of the arraydata
are replaced by their twodimensional wavelet transform.The functions return a status of
GSL_SUCCESS
upon successful completion.GSL_EINVAL
is returned ifsize1
andsize2
are not equal and integer powers of 2, or if insufficient workspace is provided.

int gsl_wavelet2d_transform_matrix(const gsl_wavelet *w, gsl_matrix *m, gsl_wavelet_direction dir, gsl_wavelet_workspace *work)¶

int gsl_wavelet2d_transform_matrix_forward(const gsl_wavelet *w, gsl_matrix *m, gsl_wavelet_workspace *work)¶

int gsl_wavelet2d_transform_matrix_inverse(const gsl_wavelet *w, gsl_matrix *m, gsl_wavelet_workspace *work)¶
These functions compute the twodimensional inplace wavelet transform on a matrix
m
.

int gsl_wavelet2d_nstransform(const gsl_wavelet *w, double *data, size_t tda, size_t size1, size_t size2, gsl_wavelet_direction dir, gsl_wavelet_workspace *work)¶

int gsl_wavelet2d_nstransform_forward(const gsl_wavelet *w, double *data, size_t tda, size_t size1, size_t size2, gsl_wavelet_workspace *work)¶

int gsl_wavelet2d_nstransform_inverse(const gsl_wavelet *w, double *data, size_t tda, size_t size1, size_t size2, gsl_wavelet_workspace *work)¶
These functions compute the twodimensional wavelet transform in nonstandard form.

int gsl_wavelet2d_nstransform_matrix(const gsl_wavelet *w, gsl_matrix *m, gsl_wavelet_direction dir, gsl_wavelet_workspace *work)¶

int gsl_wavelet2d_nstransform_matrix_forward(const gsl_wavelet *w, gsl_matrix *m, gsl_wavelet_workspace *work)¶

int gsl_wavelet2d_nstransform_matrix_inverse(const gsl_wavelet *w, gsl_matrix *m, gsl_wavelet_workspace *work)¶
These functions compute the nonstandard form of the twodimensional inplace wavelet transform on a matrix
m
.
Examples¶
The following program demonstrates the use of the onedimensional wavelet transform functions. It computes an approximation to an input signal (of length 256) using the 20 largest components of the wavelet transform, while setting the others to zero.
#include <stdio.h>
#include <math.h>
#include <gsl/gsl_sort.h>
#include <gsl/gsl_wavelet.h>
int
main (int argc, char **argv)
{
(void)(argc); /* avoid unused parameter warning */
int i, n = 256, nc = 20;
double *orig_data = malloc (n * sizeof (double));
double *data = malloc (n * sizeof (double));
double *abscoeff = malloc (n * sizeof (double));
size_t *p = malloc (n * sizeof (size_t));
FILE * f;
gsl_wavelet *w;
gsl_wavelet_workspace *work;
w = gsl_wavelet_alloc (gsl_wavelet_daubechies, 4);
work = gsl_wavelet_workspace_alloc (n);
f = fopen (argv[1], "r");
for (i = 0; i < n; i++)
{
fscanf (f, "%lg", &orig_data[i]);
data[i] = orig_data[i];
}
fclose (f);
gsl_wavelet_transform_forward (w, data, 1, n, work);
for (i = 0; i < n; i++)
{
abscoeff[i] = fabs (data[i]);
}
gsl_sort_index (p, abscoeff, 1, n);
for (i = 0; (i + nc) < n; i++)
data[p[i]] = 0;
gsl_wavelet_transform_inverse (w, data, 1, n, work);
for (i = 0; i < n; i++)
{
printf ("%g %g\n", orig_data[i], data[i]);
}
gsl_wavelet_free (w);
gsl_wavelet_workspace_free (work);
free (data);
free (orig_data);
free (abscoeff);
free (p);
return 0;
}
The output can be used with the GNU plotutils graph
program:
$ ./a.out ecg.dat > dwt.txt
$ graph T ps x 0 256 32 h 0.3 a dwt.txt > dwt.ps
Fig. 27 shows an original and compressed version of a sample ECG recording from the MITBIH Arrhythmia Database, part of the PhysioNet archive of publicdomain of medical datasets.
References and Further Reading¶
The mathematical background to wavelet transforms is covered in the original lectures by Daubechies,
Ingrid Daubechies. Ten Lectures on Wavelets. CBMSNSF Regional Conference Series in Applied Mathematics (1992), SIAM, ISBN 0898712742.
An easy to read introduction to the subject with an emphasis on the application of the wavelet transform in various branches of science is,
Paul S. Addison. The Illustrated Wavelet Transform Handbook. Institute of Physics Publishing (2002), ISBN 0750306920.
For extensive coverage of signal analysis by wavelets, wavelet packets and local cosine bases see,
S. G. Mallat. A wavelet tour of signal processing (Second edition). Academic Press (1999), ISBN 012466606X.
The concept of multiresolution analysis underlying the wavelet transform is described in,
S. G. Mallat. Multiresolution Approximations and Wavelet Orthonormal Bases of L^2(R). Transactions of the American Mathematical Society, 315(1), 1989, 69–87.
S. G. Mallat. A Theory for Multiresolution Signal Decomposition—The Wavelet Representation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 11, 1989, 674–693.
The coefficients for the individual wavelet families implemented by the library can be found in the following papers,
I. Daubechies. Orthonormal Bases of Compactly Supported Wavelets. Communications on Pure and Applied Mathematics, 41 (1988) 909–996.
A. Cohen, I. Daubechies, and J.C. Feauveau. Biorthogonal Bases of Compactly Supported Wavelets. Communications on Pure and Applied Mathematics, 45 (1992) 485–560.
The PhysioNet archive of physiological datasets can be found online at http://www.physionet.org/ and is described in the following paper,
Goldberger et al. PhysioBank, PhysioToolkit, and PhysioNet: Components of a New Research Resource for Complex Physiologic Signals. Circulation 101(23):e215e220 2000.