Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Pattern Recognition and devices

Research on Hand Tracking and Application Platform for Hand Gesture Recognition

Author ChenBangMin
Tutor LiWenJun
School Sun Yat-sen University
Course Software Engineering
Keywords Hand gesture recognition Skin color model Hands tracking Hand-hand occlusion Hand gesture application platform
CLC TP391.4
Type Master's thesis
Year 2011
Downloads 52
Quotes 0
Download Dissertation

There are more and more ways for Human-Computer Interaction (HCI) with the development of computer science technology. The vision-based hand gesture recognition is an important part of new generation HCI. It makes the computer know the meaning of a hand gesture and response for the hand gesture, which compensates for the shortcomings of the traditional interaction ways, by using the keyboard and mouse. However, there still exist some problems in vision-based hand gesture recognition research. First, while the skin color thresholds for hands segmentation and tracking, the occlusion between hands will lead to the wrong tracking of the mass centers of the hands. Furthermore, in the context of complex environment, how to develop a robust hand gesture recognition system.In this paper, a Gaussian distribution-based skin color statistics model is created in YCrCb color space, which initializes the color and area parameters for hand segmentation automatically. With the regard of the problem of hands occlusion, a method based on mass center estimation is proposed. Empirical experiments and comparisons have been conducted to evaluate the effectiveness of mass centers estimation for hands occlusion.A hand gesture recognition platform named G-Platform has been designed and developed for research and application system development. Two lost-cost cameras are utilized in G-Platform. To facility and accelerate researchers experiment and application development, we can use the platform to set up the hand parameters for the signer, record the hand gesture videos, segment the videos and extract the features for Hidden Markov Model (HMM) training, generate the HMM hand gesture model files, manage the hand gesture model files. To demonstrate the functions of the platform, comments are collected by questionnaires.

Related Dissertations
More Dissertations