| The main problem of the handheld cotton picker is labor-intensive. Through the analysis the movement of the handheld cotton picker’s picking head, the main cause affecting the picking labor intensity is found out. On these bases, improved design for the picking head structure of the original cotton picker is carried out and the intelligent direction control of picking port is implemented by using digital image processing techniques to build simulation experimental platform. The above research establishes a certain foundation for the implement of intelligent cotton picking.The main research contents:(1) Improving the picking head. Firstly, the movement of the handheld cotton picker’s picking head is analyzed to draw a conclusion that labor intensity can be reduced by changing the relative angle between cottons and picking port. Secondly, the picking head structure is redesigned. An image acquisition system and an intelligent control System are established. The relative position between cottons and picking port is got by processing the obtained camera image. On this account, the electrical motor rotation angle can be calculated and then the electric motor rotation is controlled to realize the intelligent direction control of the picking port.(2) Processing the cartoon images. Firstly, Gray level transformation, thresholding and opening and closing operators are applied for taking preprocessing for obtaining cotton images. And then the cartoon image is separated from the background by using means of region labeling and boundary tracking. Finally, with the aims of defining the cotton position, the contour of cotton is approximated seen as a circle and then use randomized Hough transform to find the center of the circle. Under the condition that the cotton is sheltered, location error will be reduced by using randomized Hough transform. And the program execution time will be shortened.(3) Controlling the operation of searching cotton on picking port. First, the strategy of searching target cotton from a number of cottons is presented. The best is that the cotton and picking port is in the ipsilateral, the cotton is closest to the picking port comes second, and final is the minimum amplitude of fluctuation. Afterwards a calculating formula of the electrical motor rotation angle is given. Finally, the rotational response is real-time transferred to MCU from the computer by using serial port communication. And MCU controls the rotation of two electrical motors.(4) The simulation of plucking experiment. The success rate of the judgments of target cottons, the alignment deviation between picking port and cotton and the single adjusted time of the picking port are selected as the evaluation index. Light, the cotton attitude, the number of cottons and cotton overlapping rate are used as the effective factors for orthogonal test. As a result of the test, the success rate of the judgments of target cottons is mainly affected by the light. The light is stronger, the success rate is lower. The alignment deviation between picking port and cotton is mainly affected by cotton attitude. The alignment error is relatively large when the cotton toward below or the side. The single adjusted time of the picking port is mainly considering the image processing time. The image processing time will increase when the number of cottons is larger than five. |
