Pornpisut Lohitharn and Viboon Sangveraphunsiri
Visual Servo Control, CAMShift algorithm, Real-Time Tracking
This paper focuses on the real-time target tracking for a Cartesian robot in 2 dimensions by using visual servo control. It is the process of controlling a system in servo-loop by using visual image data as a referenced command to control the motion of a robot. The visual servo control can be divided into 2 loops. The first loop is image processing loop for tracking a target in video frame and the second loop is servo loop for feedback control of a Cartesian robot. A target tracking is based on color probability image using a Continuously Adaptive Mean Shift (CAMShift) algorithm that finds the most probable target position from color histogram in the current frame with a single stationary camera. The experimental results illustrate that a Cartesian robot is automatically follow a moving target and record it. The visual servo control can be adapted to more complex control applications.
Important Links:
Go Back
