Recently, according to the industrial development, automated system using machine vision technology of based on image processing has been used in many fields. In particular, the more advanced automated, study on the effect of image filters is required because utilization for machine vision system is increasing. So, this study proposed about technique of image process and image filter in machine vision system. In this study, PSNR, edge count by noise removal and filter processing time were measured as outcome indicators, in order to measurement conditions of line detection, processing time and improvement of accuracy. As for the study result, Minimum filter increased to the PSNR and reduced to the edge count but took a long processing time. On the other hand, cross-shaped median filter for all indicators showed the best performance that increased to the PSNR, reduced to the edge count and shorter to processing time.