Vision Based Surface Roughness Characterization of Flat Surfaces Machined with EDM

Surface roughness measurement is an important one in any manufacturing next to dimensions. In this investigation, a vision system and image processing tools were used to develop reliable surface roughness characterization technique for Electrical Discharge Machined surfaces. A CMOS camera with red LED light source were used for capturing images of EDMed surfaces. A separate signal vector generated for all the images from its image pixel intensity matrices. The mean, skewness and kurtosis were obtained from the signal vector. The mean, skewness and kurtosis of the images signal vector correlates very well with the stylus measured hybrid roughness parameters R_da and R_dq. Hence the technique may be preferred for online surface roughness characterization of Electrical Discharge Machined (EDMed) surfaces.