Abstract:
This paper deals with surface profilometry, where we try to detect a periodic structure,
hidden in randomness using the matched filter method of analysing the intensity of light, scattered
from the surface. From the direct problem of light scattering from a composite rough surface of the
above type, we find that the detectability of the periodic structure can be hindered by the randomness,
being dependent on the correlation function of the random part. In our earlier works, we had
concentrated mainly on the Cauchy-type correlation function for the rough part. In the present work,
we show that this technique can determine the periodic structure of different kinds of correlation
functions of the roughness, including Cauchy, Gaussian etc. We study the detection by the matched
filter method as the nature of the correlation function is varied.
