Main Content

ls2filt

Extract wavelet filters from lifting scheme

Since R2021a

collapse all in page

    Syntax

    [lod,hid,lor,hir] = ls2filt(lscheme)

    Description

    [lod,hid,lor,hir] = ls2filt(lscheme) returns the wavelet decomposition and reconstruction filters associated with the lifting scheme lscheme.

    example

    Examples

    collapse all

    Open Live Script

    Create a lifting scheme associated with the db4 wavelet.

    wv = 'db4';
lsc = liftingScheme('Wavelet',wv);

    Use ls2filt to extract from the lifting scheme the corresponding wavelet filters. Compare with the filters generated by wfilters. Confirm they are equal.

    [lod,hid,lor,hir] = ls2filt(lsc);
[lod2,hid2,lor2,hir2] = wfilters(wv);
tiledlayout(2,1)
nexttile
stem(lod)
hold on
stem(lod2,"x")
hold off
title("Lowpass Decomposition")
legend("ls2filt","wfilters",Location="northwest")
nexttile
stem(hid)
hold on
stem(hid2,"x")
hold off
title("Highpass Decomposition")
legend("ls2filt","wfilters")

    Figure contains 2 axes objects. Axes object 1 with title Lowpass Decomposition contains 2 objects of type stem. These objects represent ls2filt, wfilters. Axes object 2 with title Highpass Decomposition contains 2 objects of type stem. These objects represent ls2filt, wfilters.

    figure
tiledlayout(2,1)
nexttile
stem(lor)
hold on
stem(lor2,"x")
hold off
title("Lowpass Reconstruction")
legend("ls2filt","wfilters")
nexttile
stem(hir)
hold on
stem(hir2,"x")
hold off
title("Highpass Reconstruction")
legend("ls2filt","wfilters",Location="northwest")

    Figure contains 2 axes objects. Axes object 1 with title Lowpass Reconstruction contains 2 objects of type stem. These objects represent ls2filt, wfilters. Axes object 2 with title Highpass Reconstruction contains 2 objects of type stem. These objects represent ls2filt, wfilters.

    Now create a lifting scheme associated with the bior2.2 wavelet.

    wv = 'bior2.2';
lsc = liftingScheme('Wavelet',wv);

    Use ls2filt to extract from the lifting scheme the corresponding wavelet filters. Compare with the filters generated by wfilters. Observe that wfilters includes the missing powers of the associated Laurent series as zeros so that all filters have equal even length. Except for the prepended and appended zeros, the filters coefficients generated by wfilters equal the coefficients returned by ls2filt.

    [lod,hid,lor,hir] = ls2filt(lsc);
[lod2,hid2,lor2,hir2] = wfilters(wv);
fprintf('Lowpass Decomposition\n ls2filt: %s\nwfilters: %s\n', ...
    num2str(lod),num2str(lod2))
    Lowpass Decomposition
 ls2filt: -0.17678     0.35355      1.0607     0.35355    -0.17678
wfilters: 0    -0.17678     0.35355      1.0607     0.35355    -0.17678

    fprintf('Highpass Decomposition\n ls2filt: %s\nwfilters: %s\n', ... 
    num2str(hid),num2str(hid2))
    Highpass Decomposition
 ls2filt: 0.35355    -0.70711     0.35355
wfilters: 0     0.35355    -0.70711     0.35355           0           0

    fprintf('Lowpass Reconstruction\n ls2filt: %s\nwfilters: %s\n', ...
    num2str(lor),num2str(lor2))
    Lowpass Reconstruction
 ls2filt: 0.35355     0.70711     0.35355
wfilters: 0     0.35355     0.70711     0.35355           0           0

    fprintf('Highpass Reconstruction\n ls2filt: %s\nwfilters: %s\n', ...
    num2str(hir),num2str(hir2))
    Highpass Reconstruction
 ls2filt: 0.17678     0.35355     -1.0607     0.35355     0.17678
wfilters: 0     0.17678     0.35355     -1.0607     0.35355     0.17678

    Input Arguments

    collapse all

    Lifting scheme, specified as a liftingScheme object.

    Output Arguments

    collapse all

    Decomposition filters associated with the lifting scheme, returned as vectors. lod is the lowpass decomposition filter. hid is the highpass decomposition filter.

    Data Types: double

    Reconstruction filters associated with the lifting scheme, returned as vectors. lor is the lowpass decomposition filter. hir is the highpass decomposition filter.

    Data Types: double

    Extended Capabilities

    C/C++ Code Generation
    Generate C and C++ code using MATLAB® Coder™.

    Version History

    Introduced in R2021a

    See Also

    | |