readme

STFT相关函数

The Time-Frequency Toolbox (TFTB) is a collection of about 100 scripts
for GNU Octave and Matlab (R) developed for the analysis of
non-stationary signals using time-frequency distributions. It is
primary intended for researchers and engineers with some basic
knowledge in signal processing.

The toolbox contains numerous algorithms performing time-frequency
analysis with a special emphasis on quadratic energy distributions of
the Cohen and affine classes, and their version enhanced by the
reassignment method. The toolbox also includes signal generation
procedures and processing/post-processing routines (with display
utilities).

Important note for GNU Octave users'
------------------------------------

The TFTB includes a Graphical User Interface for adjusting the layout
of the time-frequency plots. This GUI is currently _not_ supported by
GNU Octave (version 2.70). This is linked to several limitations in
GNU Octave graphics. GNU Octave's development team is preparing a
major release (upcoming version 3) which will much improve the
graphical functionalities. We expect that this will allow us to design
a GUI working with GNU Octave as well.

As of today, the scripts for TFR computations and signal generations
are expected to be fully compatible with GNU Octave, but the user has
to write a few lines of codes to produce the time-frequency plots.

Example:

%% generate signal
fs=100; % Hz
N=128;
x=fmlin(N);

%% define time and frequency axes
t=1:N;
Nf=N;
f=(0:Nf/2)/Nf;

%% compute spectrogram
tfr=tfrsp(x,t,Nf);

%% select positive frequencies
tfr=tfr(1:Nf/2+1,:);

%% make image plots
colormap(jet);
imagesc(t/fs,fs*f,flipud(tfr));

%% make contour plots
contour(t/fs,fs*f,tfr);
axis('xy');