Research On Robot Control Technology Based On Machine Vision

Multi-DOF manipulator has the features of high-speed and flexibility, and plays animportant role in the industrial production and known environment. But it can not obtain externalinformation independently, which result in that autonomous control is facing new challenges inthe unknown environment. With the development of machine vision technology, vision control ofmanipulator provides the basis to solve the self-control problems of manipulator in unknownenvironment.This paper mainly explored and researched related technology of machine vision androbot control, achieved the detection, localization and autonomy grab for target object inunknown environment. Full work is summarized as follows:Realized functional requirements analysis of the robot system, and established thegeneral scheme of the system. Meanwhile, according to the function of the system to achieve,completed selection work of hardware system, determined the actual application model of thevarious components, established a mathematical model for the manipulator, analysed kinematicscontrol principle of manipulator, discussed the target detection and tracking algorithms related tomanipulator, described control methods of position servo of manipulator in the three-dimensionalspace, completed software system design and coding work, realized target capture function byusing robot vision system control.