Design and Implementation of Real-time Object Tracking System based on Viola-Jones Algorithm for Supporting Video Conference

Abstract

Object tracking system is a system that can detect pre-defined object from frame to frame in a sequence automatically. In this research, the system is designed for supporting multimedia conferencing system for distance learning. Viola-Jones algorithm is used mainly related to relative simplicity, execution speed, and minimum memory requirement for image processing. The development of the algorithm is done by adding a template matching feature when Viola-Jones fails to detect the face area. With a fairly simple function, this algorithm can be used on system with low computing power processor, one of them is Odroid C2. In this research, an object tracking device is developed and featured with a pan and tilt bracket camera drive with two stepper motors and integrated circuit electronics. The remote controller component is added to the system to add control over the system remotely.