DESIGNING AND BUILDING A MODEL FOR VISUAL SERVOING ROBOTIC ARM

Abstract

Visual Servoing has been a focus of intensive research form the last three decades due to its various robotic applications, mainly to allow performing autonomous tasks with the help of image sensors. This thesis presenting a visual servoing control system for a 3DOF robotic arm which is visual approach of servo using a vision sensor as a feedback, this approach introduce a better flexibility more than usual used passive sensors allow the robot arm to track the object without depending on a predefined location In this thesis a 3DOF robotic arm built and servo motors and Arduino Board used to control the moving of the robotic arm. An analytic calculations made for the robot to define the DH parameters and to get the forward and inverse kinematics equations for the robot. An Object detection algorithm developed to extract an object center from an image. A calibration method developed to find the scales between the image pixels and the dimensions on the robot real world. And to perform the visual servoing an algorithm developed which initialize the system, calibrate it, finding the object in the photo, convert it to a position and send it to the robot to track the object. The system successfully built and tested and could track the object