The objective of the study is to find out the surface roughness value of the ground components using machine vision system. This study also aims at finding out the influencing parameters of the lighting conditions used in the machine vision system. The surface image of the ground components is captured by a CCD camera and it is processed to calculate a grey scale intensity value called roughness index Ri. A multiple regression equation is developed to correlate the Ri value of the surface image of the specimen and the corresponding roughness value Ra measured by stylus method. The surface roughness values obtained by machine vision are found to be accurate and comparable with the values obtained by conventional measurement method. © 2011 Inderscience Enterprises Ltd.