RESEARCH ON A SURFACE TEXTURE ANALYSIS BY DIGITAL SIGNAL PROCESSING METHODS

Surface quality has a crucial influence on exploitation properties and durability of machine parts. Hence, measurements and evaluation of surface texture is one of the most significant fields of metrology of geometrical quantities. Taking into account reliability and materials strength of mechanical elements particularly important components of a surface texture are such defects as scratches or cracks. Occurrence of such defects may result in serious damage or even break down of a whole mechanical device. Due to the local character of occurrence of such irregularities they may be easily overlooked during typical measurement and evaluation of surface texture. The paper describes an analysis of applicability of selected digital signal processing techniques, such as the Fourier and wavelet analysis to the detection of surface texture defects. Theoretical considerations are supported by a practical example.