Statistical analysis of surface roughness measurements using laser speckle images

Stylus profilers are still used as a successful method for surface roughness measurement in spite of its stylus tip diameter that acts as a low pass filter on steep valley on rough surfaces. The setup and operation time for surface measurements using a stylus profiler is considerably high. Hence a reliable non-contact optical technique for surface measurements has good potential for surface measurements based on the availability of a powerful CCD camera and fast processing digital computers. When a rough surface is illuminated with a coherent laser source, a speckle image is formed due to the scattering of light rays on the rough surface. The speckle pattern thus obtained can be used for surface roughness measurements. The contrast of the speckle image is processed to evaluate the surface roughness using the surface image parameters. Statistical parameters such as mean, variance, standard deviation, skew and kurtosis were used to analyze surface roughness using the pixel intensity of the surface images. Milled and ground surface specimens were used, and the images obtained were processed using MATLAB software.