An improved HMM/SVM dynamic hand gesture recognition algorithm

Yi Zhang; Yuanyuan Yao; Yuan Luo

doi:10.1117/12.2197328

15 October 2015 An improved HMM/SVM dynamic hand gesture recognition algorithm

Yi Zhang, Yuanyuan Yao, Yuan Luo

Proceedings Volume 9672, AOPC 2015: Advanced Display Technology; and Micro/Nano Optical Imaging Technologies and Applications; 96720D (2015) https://doi.org/10.1117/12.2197328
Event: Applied Optics and Photonics China (AOPC2015), 2015, Beijing, China

Abstract

In order to improve the recognition rate and stability of dynamic hand gesture recognition, for the low accuracy rate of the classical HMM algorithm in train the B parameter, this paper proposed an improved HMM/SVM dynamic gesture recognition algorithm. In the calculation of the B parameter of HMM model, this paper introduced the SVM algorithm which has the strong ability of classification. Through the sigmoid function converted the state output of the SVM into the probability and treat this probability as the observation state transition probability of the HMM model. After this, it optimized the B parameter of HMM model and improved the recognition rate of the system. At the same time, it also enhanced the accuracy and the real-time performance of the human-computer interaction. Experiments show that this algorithm has a strong robustness under the complex background environment and the varying illumination environment. The average recognition rate increased from 86.4% to 97.55%.

Citation Download Citation

Yi Zhang, Yuanyuan Yao, and Yuan Luo "An improved HMM/SVM dynamic hand gesture recognition algorithm", Proc. SPIE 9672, AOPC 2015: Advanced Display Technology; and Micro/Nano Optical Imaging Technologies and Applications, 96720D (15 October 2015); https://doi.org/10.1117/12.2197328

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available