Font Size: a A A

Research And Implementation Of Gesture Interactive System For Interactive TV

Posted on:2014-02-08Degree:MasterType:Thesis
Country:ChinaCandidate:X J LiuFull Text:PDF
GTID:2208330434472809Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
With the fusion of telecom network, Internet and broadcasting network the country calling for and the development of digital television and interactive television networks, new media industry has received much attention. As the center of family entertainment, TV gradually becomes customization and substantive interaction. It has also brought challenges to human-computer interaction research. Gesture interaction is an important technical direction to build a more comprehensive human-computer interaction in the future. How to quickly and efficiently manipulate TV has become a key issue.Interactive TV has gradually formed a unique hierarchical menu-based navigation principle, which utilizes up, down, left and right buttons, and some form of selection. This is effective in most remote controls, but we still do not know whether it represents an optimal paradigm and how far away from this paradigm. Whether to use numbers, color-coded shortcuts or other dedicated button is still in discussion. Consistent with the Convention has not yet appeared.Aiming at the needs of interaction with iTV, an accelerometer-based gesture interaction system is presented. Gestures are defined through research on television functions frequently used. To adapt to the smart mobile devices with limited memory resources and computing capacity, it utilizes a simple but efficient method to extract features of acceleration of the gestures, including noise-reduction, redundancy reduction and normalization. And SVM is used to for gesture recognition. The gestures are implemented on an Android-based STB running TV Launcher. Users can control the Launcher by pre-defined gestures. Experimental results demonstrate the effectiveness of the system, with accuracy rate reaching96%, and that the proposed gesture-based interaction system has certain practical value.
Keywords/Search Tags:iTV, HCI, Accelerometer, SVM, Gesture recognition
PDF Full Text Request
Related items