Homography-based visual servo regulation of mobile robots

Area of Science:

• Robotics
• Computer Vision
• Control Theory

Background:

• Mobile robot navigation often requires accurate position and orientation estimation.
• Camera-in-hand configurations present unique challenges due to uncalibrated depth information.
• Existing methods may rely on object models or depth sensors, limiting applicability.

Purpose of the Study:

• To develop an adaptive control strategy for mobile robot pose regulation using a monocular camera.
• To address the challenge of unknown depth parameters in camera-based navigation.
• To enable robust robot localization without prior object knowledge.

Main Methods:

• Utilizing a monocular camera in a camera-in-hand configuration on a mobile robot.
• Exploiting geometric relationships between target points in sequential camera images.
• Applying Lyapunov-based techniques to design an adaptive controller for pose estimation and regulation.
• Developing an adaptive estimate to compensate for unmeasurable depth parameters.

Main Results:

• A novel adaptive controller was designed for mobile robot position and orientation regulation.
• The controller successfully compensates for unknown constant depth parameters.
• Experimental results demonstrate the effectiveness of the proposed control strategy.
• The system achieves asymptotic regulation of the mobile robot's pose.

Conclusions:

• Lyapunov techniques can be effectively used to create adaptive controllers for mobile robots.
• Accurate robot pose regulation is achievable with monocular vision, even without depth information or object models.
• The proposed method offers a robust solution for camera-based mobile robot navigation in unknown environments.