Statistical estimation of point spread function applied to scanning near-field optical microscopy

Abstract Scanning near-field optical microscopy is a newly-born type of microscopy whose typical resolution reaches a few nanometres. One important step of the characterisation of such an optical device is the measurement of its point spread function (PSF). Up to now, only theoretical considerations about the PSF or results of simulations have been reported. The method used to estimate the PSF consists of determining the theoretical power spectrum of the surface profile corresponding to an experimental one. Using a known profile (randomly distributed) whose theoretical power spectrum can be calculated, allows in Fourier space the deconvolution of the measured profile and the PSF. It yields the estimation of the statistical PSF. The main advantage of this method is that the estimated PSF is experimental and describes the capabilities of such devices easily although the result is statistical.