Vision-based surface roughness inspection of machined aluminium parts

Abstract

This article presents an experimental research project which aims to control the roughness of aluminium parts produced by high-speed milling using vision data. A CCD camera is used to acquire images and a suitable processing system which uses machining data helps deriving information needed to estimate the surface roughness. A preliminary study on effect of machining parameters on part surface quality shows that roughness is influenced primarily by the feed per tooth, followed by the coating type used, and lastly, by workpiece material. In view of this information, we implemented a data processing tool allowing real-time roughness estimation based on the image captured by a camera and the feed per tooth. This tool takes into account the type of workpiece material and the cutting tool used (coatings and geometry). The results are in good agreement with data obtained using a traditional contact measurement system. The overall results of this study would encourage developments in the field of robust 3D vision system for online measurement of roughness for industrial use.