Optical Surface Roughness Evaluation of Ground Specimens Using Speckle Line Images

A well-established method of surface roughness measurement is of stylus-based. The filtering effect of the stylus tip is the major lacuna of the process. So in the present study, a vision based 100% inspection procedure is proposed for the characterization of ground specimens. A CMOS camera, and monochromatic line laser source were used for capturing speckle line images of the ground specimens. Signal vectors were generated from each of the surface images of ground specimens using MATLAB software. On the other hand the roughness of the ground specimens, particularly the Arithmetic roughness average (R_a) & Arithmetic mean slope (R_da) were computed using a stylus instrument. It was found that standard deviation and kurtosis having good correlation with the image pixel intensity of the signal vectors with the correlation coefficient of 0.96 & 0.89 for R_a and 0.86 & 0.82 for R_da respectively. So we conclude that the speckle line images can be used for in-situ surface roughness evaluation of the ground specimens.