An Interactive Control System for 3d Display Using Hidden Markov Model

Liujing Xu,Xun Cao,Xiangdong Deng,Yuanqing Wang,Jing Jin
DOI: https://doi.org/10.2991/icmt-13.2013.95
2013-01-01
Abstract:This paper describes a system fulfilling interactive 3D-TV control. The system captures the position of the users' right hand, and then records and feature-extracts the trajectory of the gestures. After that, the movement of the users' input gestures can be recognized for interactive 3D-TV operations such as rotating a 3D object. In the system, the recognition of trajectory consists of three stages: the training of models, the pre-process of data and the estimation of gestures. Hidden Markov Model (HMM) with Left-Right Bounded (LRB) structure is used in the training and estimation while an improved direction coding is proposed to extract features. Besides, the methods of two-dimension filtering and corner detection are designed for the segmentation of gestures. The system can be applied to many applications (e.g. 3D game) with the gesture control section.
What problem does this paper attempt to address?