Abstract—In this paper, we propose a system for Thai alphabet recognition from hand movement trajectory, as a human-computer interaction method. Thai characters drawing by hand movement are analyzed and recognized by our system, which can apply for controlling any specific tasks. The method is based on hand motion analysis combining with Haar-like with a cascade of boost classifiers, as hand detection method. Hand is tracked with skin color using CamShift and Kalman filter. Trajectory features of hand are extracted and used for recognizing 12 Thai alphabet letters though the Hidden Markov Model.

Index Terms—Thai alphabet, hand movement trajectory recognition, hidden markov model

The authors are with the Department of Computer Engineering, Faculty of Engineering, Prince of Songkla University, P.O. Box 2 Kohong, Hatyai,Songkla 90112 Thailand (e-mail: jalego3@hotmail.com,kom@coe.psu.ac.th).

Cite: Kittasil Silanon and Nikom Suvonvorn, "Thai Alphabet Recognition from Hand Motion Trajectory Using HMM," International Journal of Computer and Electrical Engineering vol. 4, no. 3, pp. 312-317, 2012.