Fiducial Markers Applied for Pose Tracking of a Robotic Manipulator: Application in Visual Servoing Control

2016 ◽  
Author(s):  
Igor Pereira Vieira ◽  
Armando Alves Neto ◽  
Leonardo Amaral Mozelli
Keyword(s):  
Visual Servoing ◽  
Robotic Manipulator ◽  
Fiducial Markers ◽  
Pose Tracking ◽  
Visual Servoing Control

Virtual exoskeleton-driven uncalibrated visual servoing control for mobile robotic manipulators based on human–robot–robot cooperation

2018 ◽  
Vol 40 (14) ◽  
pp. 4046-4062
Author(s):  
Peng Ji ◽  
Hong Zeng ◽  
Aiguo Song ◽  
Ping Yi ◽  
PengWen Xiong ◽  
...  
Keyword(s):  
Control System ◽  
Visual Servoing ◽  
Computational Cost ◽  
Robotic Manipulator ◽  
Switching Algorithm ◽  
Careful Handling ◽  
Uncalibrated Visual Servoing ◽  
Robot Cooperation ◽  
Motion Characteristics ◽  
Visual Servoing Control

This paper presents an uncalibrated visual servoing control system based on the human–robot–robot cooperation (HRRC). In case of malfunctions of the joint sensors of a robotic manipulator, the proposed system enables the mobile robot to continue operating the manipulator to complete the task that requires careful handling. With the aid of a virtual exoskeleton, an operator may use a human–computer interaction (HCI) device to guide the malfunctioning manipulator. During the guiding process, the virtual exoskeleton serves as a connector between the HCI device and the manipulator. However, when using the HCI device to guide the virtual exoskeleton, there could be a risk of a large-residual problem at any time caused by non-uniform guiding. To solve this problem, a residual switching algorithm (RSA) has been proposed that can identify whether the residual should be calculated based on the motion characteristics of the artificial guiding, reducing the computational cost and ensuring the tracking stability. To enhance the virtual exoskeleton’s ability to drive the manipulator, a multi-joint fuzzy driving controller has been proposed, which can drive the corresponding joint of the manipulator in accordance with an offset vector between the virtual exoskeleton and the manipulator. Lastly, the guiding experiments have verified that, compared with the contrast algorithm, the proposed RSA has a better tracking performance. A peg-in-hole assembly experiment has shown that the proposed control system can assist the operator to control efficiently the robotic manipulator with malfunctioning joint sensors.

Binocular vision-based 3-D trajectory following for autonomous robotic manipulation

Robotica ◽  
2007 ◽  
Vol 25 (5) ◽  
pp. 615-626 ◽  
Author(s):  
Wen-Chung Chang
Keyword(s):  
Binocular Vision ◽  
Visual Servoing ◽  
Vision System ◽  
A Priori ◽  
Robotic Manipulator ◽  
Robotic Manipulation ◽  
Control Approach ◽  
Trajectory Following ◽  
Image Planes ◽  
Binocular Vision System

SUMMARYRobotic manipulators that have interacted with uncalibrated environments typically have limited positioning and tracking capabilities, if control tasks cannot be appropriately encoded using available features in the environments. Specifically, to perform 3-D trajectory following operations employing binocular vision, it seems necessary to have a priori knowledge on pointwise correspondence information between two image planes. However, such an assumption cannot be made for any smooth 3-D trajectories. This paper describes how one might enhance autonomous robotic manipulation for 3-D trajectory following tasks using eye-to-hand binocular visual servoing. Based on a novel encoded error, an image-based feedback control law is proposed without assuming pointwise binocular correspondence information. The proposed control approach can guarantee task precision by employing only an approximately calibrated binocular vision system. The goal of the autonomous task is to drive a tool mounted on the end-effector of the robotic manipulator to follow a visually determined smooth 3-D target trajectory in desired speed with precision. The proposed control architecture is suitable for applications that require precise 3-D positioning and tracking in unknown environments. Our approach is successfully validated in a real task environment by performing experiments with an industrial robotic manipulator.

Sign in / Sign up

Export Citation Format

Share Document