Optical near-field data analysis through time-frequency distributions: application to the characterization and separation of the image spectral content by reassignment

The near-field optical images have been traditionally analyzed by Fourier a nalysis and, recently, by wavelet analysis. Those data are nonstationary, w hich means that their spectral content varies with time, owing to the scann ing-probe recording process; therefore time-frequency representations are, potentially, powerful tools for local characteristics extraction or shape s eparation, since they distribute the energy of the analyzed signal over the time and frequency variables and faithfully depict the signal local behavi or. In this study we show that Cohen's class time-frequency distributions a nd their modified version by the reassignment method are appropriate tools for the analysis of near-field optical data. We demonstrate this by using t hese tools first on simulated data and second on experimental near-field op tical images. Within this context we observe that time-frequency analysis a llows one to easily characterize local frequencies, which involves a possib le separation of relevant optical signal fi om artifacts. (C) 2000 Optical Society of America [S0740-3232(00)01312-0] OCIS codes: 070.0070, 100.0100, 180.0180, 180.5810.