An optical/digital approach to the classification of rough surfaces is described. The setup consists of a coherent optical system to generate the Fourier irradiance of a rough sample and a digital computer to process data. The sampling of Fourier spectra is achieved with a wedge ring detector. The extraction and measurement of specific statistical parameters of Fourier spectrum are done for two typical mached samples. An analysis of the efficiency of these parameters in roughness classification is also carried out.