Motion Control Of Robotic Arm Based On Visual Positioning

The emergence of the robotic arm has brought great changes to the human production and life,and can replace human beings in the harsh conditions to accomplish a lot of work.At the same time,it has a higher efficiency in the field of production.Therefore,it has become one of the hot spots in the world.At present,the traditional mechanical arm can no longer meet the production needs.Therefore,scholars at home and abroad are committed to studying the mechanical arm combined with vision.The object is detected and positioned through vision,then the robot arm is moved to the target position,making the whole process more intelligent and easy to control.The vision guided mechanical arm will no longer use the fixed point position to move back and forth.The whole process will be directly controlled by the host computer for easy setting and changing.In this paper,vision based SCARA manipulator motion control is mainly studied in 3 aspects.The kinematic model of a four axis plane joint(SCARA)type manipulator is established by the standard D-H method,and the kinematics and inverse kinematics are solved to determine the transformation relationship between the joint space of the manipulator and the Cartesian space.In order to verify the correctness of the kinematics model,a simulation model is built in the Robotics Toolbox for MATLAB toolbox to verify the correctness of the solution.Then preprocess the image captured by camera and then using MSER to detect the target area.For detecting the maximum stable extreme region,we first select the initial size according to the aspect ratio and the average pixel threshold,then merge the duplicate region.Then a screening is carried out according to the statistical characteristics,then the connected domain is marked,each candidate region is extracted and the contour is detected.The contour of the extracted region is matched with the template contour to get the most similar candidate region,whicn is the matched target area.After getting the location of the object in the image,we need to calculate the position of the location in the world coordinate system.Therefore,camera calibration is carried out to get the transformation relation from image coordinates to 3D spatial coordinates.The speed control of the joint space of the manipulator and the trajectory planning of the Cartesian space are also studied.On the basis of the Cartesian space linear interpolation,the path planning of multiple target points is carried out,and a genetic algorithm is used to determine an optimal path from the starting point,traversing each position,and finally returning to the starting point.The best condition is the shortest path.Finally,the hardware experiment platform is built,and the system software is designed to realize the motion control of the robot arm based on the visual guidance,which verifies the effectiveness of the algorithm.