Related Experiment Videos
Nonlinear visual mapping model for 3-D visual tracking with uncalibrated eye-in-hand robotic system
Jianbo Su1, Yugeng Xi, Uwe D Hanebeck
1Department of Automation, Shanghai Jiaotong University, Shanghai, China. jbsu@sjtu.edu.cn
Summary
This study introduces a novel control scheme for uncalibrated robotic visual tracking, balancing computational cost with effective offline modeling and online control for improved performance.
Area of Science:
- Robotics
- Computer Vision
- Control Systems
Background:
- Uncalibrated robotic visual tracking presents challenges due to the lack of prior system calibration.
- Efficient control strategies are needed to manage computational load while maintaining tracking accuracy.
Purpose of the Study:
- To propose a new control scheme for uncalibrated robotic visual tracking.
- To reduce computational expenses through a combination of offline modeling and online control.
Main Methods:
- Developed a nonlinear visual mapping model for uncalibrated hand-eye coordination using artificial neural networks.
- Implemented an online visual tracking controller integrated with a real-time motion planner.
- Incorporated a feedforward controller to compensate for unknown object motions.
Main Results:
- The proposed artificial neural network model effectively handles uncalibrated hand-eye coordination.
- The integrated control scheme demonstrates robust online visual tracking capabilities.
- Feedforward compensation successfully improved system performance by addressing unpredicted object movements.
Conclusions:
- The developed control scheme offers an effective solution for uncalibrated robotic visual tracking.
- The approach balances computational efficiency with high tracking performance.
- Simulations and experiments validate the practical applicability and effectiveness of the proposed system.