| With the rapid development of agricultural intelligence,machine vision technology is widely used in the field of agricultural automated picking,which can assist humans in completing the corresponding picking operations.In mechanical control systems,machine vision has become the mainstream of today’s manipulator control systems with its advantages of high efficiency and high positioning accuracy.This paper is based on machine vision to study the six-degree-of-freedom robot hand to realize the capture and placement of the target of selenium-enriched green tea buds.Use Open MV as the machine vision sensor to study the problems of manipulator kinematics analysis,trajectory planning,and shoot target recognition and positioning,realize the intelligent tea picking manipulator control system based on machine vision,and successfully complete the shoot picking.First,determine the specific function of each sub-module through the design of the overall scheme,and explain the control principle of each module.The basic model of the six-degree-of-freedom manipulator is established through D-H parameters,and its kinematics is analyzed,the difference of its trajectory planning is analyzed under different coordinates,and the orderly study of its work space is carried out.Compare the performance of different picking path planning algorithms,and select the PSO algorithm to complete the robot path planning.Secondly,a simulation study is carried out on the control strategy and picking of the manipulator,the three most common control methods are compared and the simulation analysis is carried out,and the fusion control scheme is adopted to complete the research of this paper.Finally,based on the Arduino IDE software platform,the picking test of the intelligent tea picking manipulator based on machine vision is completed,and the modular manipulator steering gear control program is written to drive the steering gear to complete the corresponding picking tasks.The platform environment must be configured before the test.The bud images are preprocessed,and the improved K-means algorithm is used to complete the identification of the buds.The coordinates of the buds are output based on the April Tag visual positioning.The picking process is specifically designed and the simulation test is completed.The test results show that the designed intelligent tea picking manipulator control system based on machine vision can complete the task of picking buds.The average recognition success rate of the simulation test is 94.57%,and the average picking time is10.28 s,which is useful for the implementation of automated agriculture.A good driving effect is of great practical significance to the development of intelligent agriculture. |
