Evaluation of Hand Pointing System Based on 3-D Computer Vision

  • P. Serafinavičius Kaunas University of Technology
  • S. Sajauskas Kaunas University of Technology
  • G. Daunys Šiauliai University

Abstract

Evaluation of hand pointing system is presented. The system is based on 3D computer vision and implemented using open source computer vision library (OpenCV). The system is able to detect and track user’s head and hand and return 3D coordinates in real time. The detection of head and hand is based on Viola-Jones detector applying human skin color model information. The detection rate is about 90%. The tracking is based on Kanade-Lucas-Tomasi iterative algorithm, customized for 3D computer vision case. High tracking robustness (90%) is achieved due to skin color model and variable size search window according to depth information. The speed of the system is 15–23 frames per second on 3 GHz desktop PC. Ill. 3, bibl. 11 (in English; summaries in English, Russian and Lithuanian).

Published
2015-03-30
How to Cite
Serafinavičius, P., Sajauskas, S., & Daunys, G. (2015). Evaluation of Hand Pointing System Based on 3-D Computer Vision. Elektronika Ir Elektrotechnika, 88(8), 95-98. Retrieved from http://eejournal.ktu.lt/index.php/elt/article/view/11359
Section
T 125 AUTOMATION, ROBOTICS