Short Time Fourier Transform Probability Distribution for Time-Frequency Segmentation

Taking as signal model the sum of a non-stationary deterministic part embedded in a white Gaussian noise, this paper presents the distribution of the coefficients of the short time Fourier transform (STFT), which is used to determine the maximum likelihood estimator of the noise level. We then propose an automatic segmentation algorithm of the real and imaginary parts of the STFT based on statistical features, which is an alternative to the spectrogram segmentations considered as image segmentations. Examples of segmented time-frequency space are presented on a simulated signal and on a dolphin whistle

[1]  Richard O. Duda,et al.  Pattern classification and scene analysis , 1974, A Wiley-Interscience publication.

[2]  Nadine Martin,et al.  Maximum likelihood noise estimation for spectrogram segmentation control , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[3]  J. Kenney,et al.  Mathematics of statistics , 1940 .

[4]  Nadine Martin,et al.  Spectrogram segmentation by means of statistical features for non-stationary signal interpretation , 2002, IEEE Trans. Signal Process..

[5]  David G. Long,et al.  The probability density of spectral estimates based on modified periodogram averages , 1999, IEEE Trans. Signal Process..