VisualServoingBaxter

Vision servoing based object following and obstacle avoidance using the Baxter robot.

Contents

• Introduction
• Usage
• Result

Introduction

A visual servoing control method enhanced by an obstacle avoidance strategy using potential based function. The solution require relative bearing measurements, which acquire from the RGB-camera by color-based detection method through OpenCV. Eye-in-hand sensor arrangement can provide a relative high accuracy and avoid vision collision. We also validate our approach by implementation on the Baxter robot.

Usage

1. Save this package into catkin_ws/src/.

2. Verify the robot is enabled.

rosrun baxter_tools enable_robot.py -e

3. Start the joint trajectory controller.

rosrun baxter_interface joint_trajectory_action_server.py --mode position

4. Open the camera on the left hand.

rosrun baxter_tools camera_control.py -o left_hand_camera -r 960x600

rosrun image_view image_view image:=/cameras/left_hand_camera/image

5. Run the tracking part.

rosrun VisualServoingBaxter tracking.py

6. Run the following part.

rosrun VisualServoingBaxter following.py

Result

Trajectories of the object and the Baxter robot.