Surface roughness analysis from GLCM features: a surrogate speckle interferometric method

Abstract The development of novel sensitive techniques for surface roughness analysis has always been of great interest in manufacturing industries, as surface roughness can significantly alter the efficiency of moving parts due to friction. Besides several techniques, optical interferometric techniques stand unique because of their sensitivity and accuracy. This paper proposes speckle interferometry as a surrogate technique for the detection of surface roughness, taking sandpaper as an example. For this, specklegrams of standard sandpapers are recorded, and the 2D and 3D contour plot analysis is carried out to get information about the surface morphology. The specklegram image features—inertia moment, energy, contrast, homogeneity, and correlation—are extracted from the gray level co-occurrence matrix (GLCM), and a linear regression equation is set relating each feature with the root mean square roughness ( β ). Out of the sixteen data extracted corresponding to each feature, fifteen are used to set up the relation, and the remaining one is used for validation. The observed computed accuracy in the estimated β value reveals the possibility of industrial application of speckle interferometry in surface roughness monitoring from GLCM features.