Font Size: a A A

Research And Realization Of Chime-Bell Playing System Driven By Gesture Discrimination

Posted on:2008-06-07Degree:MasterType:Thesis
Country:ChinaCandidate:W J HuFull Text:PDF
GTID:2178360212484919Subject:Digital art and design
Abstract/Summary:PDF Full Text Request
The Chinese nation has a long history and is rich in non-material cultural heritage. The protection and transmission of these non-material cultural heritages usually need a close contact with cultural relic. However, the heritage preservation policy limits the methods of cultural relic, which has made a new interactive approach appeared inevitable. Therefore, the computer gesture recognition technology as the representative for the natural interactive technology gradually becomes the hot area of research.By deeply research the gesture recognition technology of natural human-machine interactive, this paper proposed a way of combining the natural human-machine interactive technology and the protection of non-material cultural heritage to successfully resolve the problem that non-material cultural heritage protection are not directly used for outreach, which uses computer-based visual gestures to plat chime-bells in the 3D environment. This enriches the non-material cultural heritage protection, and thus has a certain practicality and scalability.This paper firstly compares several common computer gesture recognition technologies, and selects the most suitable one-the computer-visual based gesture recognition technology as research priorities. Then, against the invalid distance in gesture recognition process, this paper gives out the corresponding solutions -filtering invalid distance by Computer-vision based three-dimensional distance measurement, and gives the distance measurement formulas are derived. We need to go through several important stages to achieve the real-time user gesture recognition: gestures modeling, gesture segmentation and gestures analysis. Among them, gestures segmentation operation is the key part of this paper. This paper illuminates the technology to extract human region based on human color features dynamically, and further refined the feature data of the gestures segmentation model using Freeman chain principle.Finally, this paper realizes gesture-driven chime-bells playing system, which is based in the five bells excavated from FuHao tomb in Yin dynasty. This system reconstructs the original scene in OpenGL environment and inputs gesture using single-eye camera to drive bells playing. Finally, the paper shows the concrete realization of the system.
Keywords/Search Tags:Chime-bell music, Computer vision, Gesture recognition, Complexion segmentation, Freeman chain
PDF Full Text Request
Related items