Control Design and Implementation of Autonomous 2-DOF Wireless Visual Object Tracking System
ABSTRACT: Due to large scale
implementation of visual detection and tracking as a mean of sensor andnavigation
tool, target detection and tracking using image manipulation for autonomousrobotic
system becomes an interesting object of study for many researchers. In
addition, there have been attempts to develop a system that can detect and
track a moving target by using an image or video processing in a real time
condition. Despite that, visual objecttracking can be a subject of noise
because of image manipulation. The noise can createuncertainty on state and
observation model that can lead to control instability, especiallythat in
remote operation. Therefore, an effective filter that can tackle or reduce this
noise isneeded in developing a visual object tracking system. In this work, a
2-degree of freedom (2-DOF) visual object tracking system was developed with an
information filter. The systemconsists of an image capture unit, an image
processing unit, a wireless communication unit, and a manipulator. Then to
observe the filter effectiveness on real time visual object tracking in remote
operation, performances of this visual object tracking system with and without
the filter were tested based on video simulation and real time tracking. In the
live streaming test, the information filter can reduce the error of the
measurement about 30% than that withoutit.
Keywords: Visual detection,
wireless object tracking, Denavit-Hartenberg parameter, information filter,
two-degree of freedom
Author: Radius Bhayu Prasetiyo
Journal Code: jptmesingg170006
