Research on Manipulator Control Based on Visual Feature Constraints

With the deepening of research in machine vision and other related fields, vision-based robotic arm control has always been a research hotspot in the field of robotics. For the automatic grasping control of robot arm, a control method based on visual feature constraints is proposed in this paper. The linear spatial interpolation method is used to plan the linear trajectory of the end of the robot arm in Cartesian space, and then control the end of the robot arm to pass the planned trajectory points successively, so as to grasp the target successfully. In order to solve the problem that the object exceeds the camera field of view, the visual feature constraint function is introduced to limit the movement of the robot arm. The simulation results show that this method can effectively control the position of the end of the robot arm in the process of automatic grasping, and ensure that the grasping target is always in the field of view of the camera.