抽象的

Real Time Hand Tracking System

Aisha Meethian, Dr. B.M.Imran

This paper investigates a real time gesture recognition system. Hand tracking is first step for gesture recognition and hand tacking system is based on video acquisition, skin colour segmentation, foreground detection and background elimination. The proposed Gesture Recognition Hand Tracking (GR-HT) system take less training time compared to existing system. Moreover, other feature of GR-HT system is less memory requirement. Disadvantage of GR-HT system is it is affected by environmental condition, to reduce the effects, we use Gaussian Mixture Model (GMM) for background elimination and foreground detection. The experimental result shows that the proposed GR-HT system achieves the results up to 98% in recognition and our work is a novel method for real time gesture recognition applications.