Evaluation of machined surface texture using a vision sensor

A non-contact measurement technique was developed for the in-cycle assessment of ground surface texture. The system utilises off-the-self vision components to capture the pattern of scattered light from the surface, which is illuminated by a white light source at a shallow grazing angle. Three parameters, related to the surface texture, were derived from the pattern and correlation curves were established by plotting the three textural parameters against the corresponding average surface roughness R-a obtained from a stylus profilometer. A set of test samples, ground to a pre-determined roughness. were used to estimate the efficacy of the measurement technique. The proposed technique is fast and reliable means of extracting textural information from ground surfaces and, furthermore, it can be easily adapted to use in a production environment.