Abstract：By studying the cucumber ridge images captured in greenhouses, a navigation method based on computer vision and fuzzy control was proposed. The chromatic aberration operator （2G- R -B） was used for color-image gray processing. The trajectory position was preliminarily ascertained by vertical his- togram with double peaks, and discrete navigation points were also calculated line by line. The navigation parameters including lateral deviation and angle deviation were obtained from navigation reference line fitted by improved Hough transform. A fuzzy controller was set with lateral deviation and angle deviation as input and differential speed value of mobile robot as output. The results of navigation experiments show that the average time cost of the algorithm is less than 150 ms and the maximum error of path tracking is less than 5 cm with the speed of 0.3 m/s.
2009, Vol. 30 
