Reconstructs a real-valued signal from ridges found by crazy climbers on a Gabor transform.
gcrcrec(siginput, inputgt, beemap, nvoice, freqstep, scale, compr,
bstep=5, ptile=0.01, epsilon=0, fast=TRUE, para=5, minnbnodes=3,
hflag=FALSE, real=FALSE, plot=2)
original signal.
Gabor transform.
occupation measure, output of crc
.
number of frequencies.
sampling step for frequency axis.
size of windows.
compression rate to be applied to the ridges.
size (in the time direction) of the steps for chaining.
threshold of ridge
constant in front of the smoothness term in penalty function.
if set to TRUE, uses trapezoidal rule to evaluate \(Q_2\).
scale parameter for extrapolating the ridges.
minimal number of points per ridge.
if set to FALSE, uses the identity as first term in the kernel. If not, uses \(Q_1\) instead.
if set to TRUE
, uses only real constraints.
displays signal,components, and reconstruction one after another.
displays signal, components and reconstruction.
Returns a structure containing the following elements:
reconstructed signal.
image of the ridges (with different colors).
2D array containing the signals reconstructed from ridges.
When ptile
is high, boundary effects may appear. para
controls extrapolation of the ridge.
See discussions in the text of “Practical Time-Frequency Analysis”.