Real-Time Capturing and Visualization of Human Hand Poses Using Low Cost Web Cameras

Majid A. Al-Taee, Tarek Abu-Jabarah, Ihab Sultan, Ashraf Al-Karmi

Abstract


This paper presents a real-time tracking system for three-dimensional (3D) hand poses using a personal computer and two low-cost web cameras. A new computer algorithm is developed and implemented to capture and visualize a wide range of 3D hand poses. The proposed algorithm employs efficient 3D modeling and digital image processing techniques to map the extracted image features to an articulated 3D object like human hand poses. The fundamental components of this algorithm include a set of image acquisition functions, image processing and features selection functions, mapping functions, and a number of 3D modeling functions. The image processing and features selection functions are based on intensity information of the acquired images with static background. Neither machine training nor Cyber-Gloves or color information is required in the proposed system and therefore no restrictions are imposed on users except following a few simple predefined steps. The overall performance of the proposed system is examined experimentally and a robust real-time motion tracking of various hand poses is demonstrated.

Keywords


Computer vision, Real-time Systems, Motion tracking, 3D human-hand modeling, Digital image Processing

Full Text:

PDF

Refbacks

  • There are currently no refbacks.