3D vision for mobile robot manipulator on detecting and tracking target

Abstract

In this study, an investigation of mobile robot manipulation applied stereo vision system, tracking and gripping the object with different disciplines to produce a robotics system with abilities to perform a complex real task is conducted. Stereo cameras for pose-based visual servoing and construction 3D map of the object to the manipulator is generated. Algorithms for mobile robot controlling, stereo visions, transformation functions and inverse kinematics are provided to make the study more comprehensive. The study has successfully integrated computer vision technique to mobile robot manipulator, define the target base on two cameras and solving the inverse kinematic functions for 4 DOF elbow manipulator. Camera system has succeeded tracking measuring the distance; the object is inside the workspace of the manipulator, computer computes the inverse kinematic equations to find out the best angle for the arm reaching the target. Experiment results are presented to demonstrate the ability of the proposed control method.