
Windowed Fourier Transforms ◾ 41
fringe signal according to Figure2.1b, as well as that of the fringe signal
according to Figure2.6. Comparing the distribution of the spectra, it is
evident that the noise has an eect on the whole spectrum of the sig-
nal, and it is dicult to lter out such a noise spectrum in the frequency
domain.
2.3 PHASE DEMODULATION BASED ON
WINDOWED FOURIER TRANSFORM
WFT is an important method for nonstationary signal analysis. To over-
come the deciency of FT with poor localization in the spatial frequency
domains, Dennis Gabor adapted the FT in 1946 to analyze only a local
section of the signal at a time and p ...