Object Tracking and Motion Capturing in Hardware-Accelerated Multi-camera System

  • Authors:

  • Sirisak Leephokhanon;Theerayod Wiangtong

  • Affiliations:

  • Department of Electronic Engineering, Mahanakorn University of Technology, Bangkok, Thailand 10530;Department of Electronic Engineering, Mahanakorn University of Technology, Bangkok, Thailand 10530

  • Venue:
  • ARC '09 Proceedings of the 5th International Workshop on Reconfigurable Computing: Architectures, Tools and Applications
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper presents a multi-camera system where reconfigurable hardware and video codec are used to assist real-time video processing. Video inputs from all analog cameras are decoded and processed by FPGA to accelerate overall processing time. Two applications including multiple-object tracking and motion capturing are implemented in this system to verify real-time capability. In multiple-object tracking, video images are rearranged and combined into a single image that profits the tracking algorithm called multiple-CAMshift running in software. In motion capturing, every frame from all cameras, located in different angles, is processed to find the coordinates of 11 reference color-objects. Results are sent to a computer for constructing 3-D image in real-time.