Assessment of Areal (Three-Dimensional) Roughness Parameters of Milled Surface Using Charge-Coupled Device Flatbed Scanner and Image Processing

Abstract

A new application of the charge-coupled device (CCD) flatbed scanner for the noncontact assessment of areal three-dimensional (3D) roughness parameters of milled surfaces is proposed in this work. The surfaces were prepared by milling a block of aluminum at five different feed rates using a CNC milling cutter. The specimen was placed on the flatbed scanner with the milled surface facing the scanner array. The scanned images were analyzed using image processing to extract the average pixel intensity, root-mean-square pixel intensity, variance, black-to-white ratio, and bright region average area intensity within a region-of-interest. These parameters were correlated with the 3D areal parameters obtained using a state-of-the-art focus variation 3D metrology system. The highest correlation of 0.9931 was obtained between the average pixel intensity from the scanned image and average areal roughness S a at a scanning height of 7.2 mm from the scanner surface. Compared with the costly 3D metrology system, the flatbed scanner combined with the proposed image processing algorithm has significant potential for roughness assessment due to the high correlation between the image parameters and roughness parameters.