One of the crucial elements of image based inspection system development is the lighting conditions. It directly defines the quality of the image which in turn affects the accuracy and robustness of an inspection procedure using machine vision system. The common image characteristics change with variation in lighting leading to large image differences. In recent years, evaluation of surface roughness of a work piece by machine vision has received a great deal of attention. However, practical surface roughness instruments based on machine vision are still difficult to develop for application specific online assessment in particular. This is due to the fact that the images taken from the machined surfaces are affected by illumination, reflectivity and ambience during the image acquisition process. This lighting inhomogeneity is considered to be a disturbing signal component, which should be suppressed to achieve consistency in surface roughness quantification. In this paper, the illumination compensated images are used for surface roughness evaluation. The homomorphic filtering and Discrete Cosine Transform (DCT) based normalization techniques are utilized to remove the illumination inhomogeneity and the performance of these techniques were compared. The results clearly indicate that it is important to consider the lighting conditions when the machine vision approach is used to quantify the surface texture parameters. Copyright © 2016 by ASME.