Experimental Study on Image Filtering Algorithm.

With the continuous development of intelligent assembly industry, the research on robotic assembly systems has became a hot and cutting-edge issue concerned by powers. During the development of the system, robot vision is mainly used for tracking and positioning. In the establishment of robot vision system, image preprocessing technology is one of the key steps. This paper is aiming at the robot system assembled on the ground of the satellite, which involves machine vision part, the median filtering method and mean filtering method are analyzed through PSNR of peak noise ratio and fuzzy coefficient based on Matlab platform. The noise reduction ability of the five filtering algorithms, including gaussian filtering, guided filtering and bilateral filtering. Then, comparing the best algorithm with the operation time, bilateral filtering has the best effect on image noise reduction and the longest time.