Reconstruct single branch from 2-D wavelet coefficients
wrcoef2 is a two-dimensional wavelet analysis function.
wrcoef2 reconstructs the coefficients of an image.
returns the matrix of reconstructed coefficients of type
x = wrcoef2(
on the wavelet decomposition structure
[c,s] of an image (see
wavedec2 for more information) using
the wavelet specified by
wname. The coefficients at the maximum
decomposition level are reconstructed. The size of
x is equal to
the size of the original image.
Save the current extension mode. Load an image.
origMode = dwtmode("status","nodisp"); load woman imagesc(X) title("Original") colormap gray
dwtmode to change the extension mode to zero-padding. Obtain the 2-level wavelet decomposition of the image using the
dwtmode("zpd","nodisp") [c,s] = wavedec2(X,2,"sym5");
Reconstruct the approximation coefficients at levels 1 and 2. Display the results.
a1 = wrcoef2("a",c,s,"sym5",1); a2 = wrcoef2("a",c,s,"sym5",2); subplot(1,2,1) imagesc(a1) title("Level 1") subplot(1,2,2) imagesc(a2) title("Level 2") colormap gray
Reconstruct the horizontal, vertical, and diagonal detail coefficients at level 2.
h2 = wrcoef2("h",c,s,"sym5",2); v2 = wrcoef2("v",c,s,"sym5",2); d2 = wrcoef2("d",c,s,"sym5",2);
Confirm all the reconstructions are the same size as the original image.
sX = size(X); sa1 = size(a1); sa2 = size(a2); sh2 = size(h2); sv2 = size(v2); sd2 = size(d2); [sX;sa1;sa2;sh2;sv2;sd2]
ans = 6×2 256 256 256 256 256 256 256 256 256 256 256 256
Restore the extension mode to the original setting.
type— Coefficients to reconstruct
Coefficients to reconstruct, specified as follows:
"a" — Approximation coefficients
"h" — Horizontal detail coefficients
"v" — Vertical detail coefficients
"d" — Diagonal detail coefficients
Wavelet, specified as a character vector or string scalar.
wrcoef2 supports only Type 1 (orthogonal) or Type 2
(biorthogonal) wavelets. See
wfilters for a list of orthogonal
and biorthogonal wavelets.
LoR,HiR— Wavelet reconstruction filters
Wavelet reconstruction filters, specified as a pair of even-length real-valued
LoR is the lowpass reconstruction filter, and
HiR is the highpass reconstruction filter. The lengths of
HiR must be equal. See
wfilters for additional
x— Reconstructed coefficients
Reconstructed coefficients, returned as a matrix. The size of
x is equal to the size of the original image