Machine Vision Based Cotton Recognition for Cotton Harvesting Robot

A new cotton recognition method is proposed in this paper. It provides parameters for motion of the manipulator so that it can acquire precise location information of cotton, identify cotton from surroundings correctly, and accordingly pick up them automatically. This method is based on color subtraction information of different parts of cotton. Furthermore, in order to increase accuracy rate of cotton recognition, dynamic Freeman chain coding is used to remove noise. Experimental results show that the proposed method has good performance for cotton identification.