Measurement and
characterization of
surface texture is an important aspect of precision metrology. Historically this has
involved partitioning a profile into different wavelength regimes referred to as roughness, waviness and form followed by numerical quantization. Parameters computed are then inspected for tolerance compliance to ensure a part performs its intended function. This approach is satisfactory when the specification has been carefully determined and the process is stable. However, when the manufacturing process is under development or when instability or modifications to the process invalidate specifications, there is a need to study surface finish parameters in relation to functional performance or process measures. In this context, the problem of
surface texture classification and recognition are discussed. Advanced techniques developed for this purpose along with applications are presented. Also, the techniques discussed here will be useful across large bandwidth, from the
characterization of nano scale to traditional micro scale surfaces.