rapsd


Purpose
This function performs the Radially Averaged Power Spectral Density Estimation using the method of Periodograms.

Synopsis
[y, r] = rapsd(x, d)
[y, r] = rapsd(x, d, f)
[y, r] = rapsd(x, d, f, s)
[y, r, c] = rapsd(x, d)
[y, r, c] = rapsd(x, d, f)
[y, r, c] = rapsd(x, d, f, s)

Description
[y, r] = rapsd(x, d) calculates the radially averaged power spectral density estimate, y of type double, for input image x of type double, using periodograms of dims. given by the scalar d = [N] (for square periodograms).
It also returns the radial frequency array r of type double.

[y, r] = rapsd(x, d, f) calculates the radially averaged power spectral density estimate, y of type double, for input image x of type double. f is the frequency domain matrix dimensions, holding the results of the fft of periodogram. It's dimensions cannot be smaller than d. If f is not specified or f = [0], then f = d.

[y, r] = rapsd(x, d, f, s) calculates the radially averaged power spectral density estimate, y of type double, for input image x of type double. s is a matrix holding the shift parameters i.e. the amount by which the sample window (periodogram) must be shifted for the next sample.
s = [SM SN] or s = [SM]. If s is not specified or s = [0], then s = d.

[y, r, c] = rapsd(x, d)
[y, r, c] = rapsd(x, d, f)
[y, r, c] = rapsd(x, d, f, s)
calculates the radially averaged power spectral density estimate, y of type double, for input image x of type double and also returns the count c of the sample windows (periodograms) for the calculation of sample variance.

See Also
psd2, anis, hxpsd2, hxrapsd, hxanis

Reference
[1] R. A. Ulichney, Digital Halftoning, MIT Press, Cambridge, Mass., 1987

Author
Faraz Faheem, Dated: 10/23/99.
Daniel L. Lau, Dated: 04/16/04




Dr. Daniel Leo Lau
Department of Electrical and Computer Engineering
University of Kentucky
453 Anderson Hall
Lexington, KY 40506-0046, USA
(859) 257-2300 x273
dllau@engr.uky.edu

last edited on 4/16/04