Surface roughness determination using wavelets

Surface is a finger print of all manufacturing processes and is a quantifying parameter. The determination of the roughness of technical surfaces plays an important role in scientific and industrial applications. Traditional surface texture involves surface profile classified into wavelengths of Roughness, Waviness and Form. Different aspects of manufacturing processes generate different wavelength regimes and these affect the function of part differently. Conventional analysis should not incorporate these wavelength bands while evaluation of the surface roughness values. In the present paper, the surface profile is considered as a signal and using wavelet transforms a coherent approach for evaluating the surface roughness is proposed. The main contribution of the paper is to present wavelength based surface features extracted from the surface profiles by wavelet analysis. By this technique surface roughness can be separated from the actual profile and the `Ra' value can be measured.