交談式多媒體展現的同步 - 以 OCPN 模型為基礎

Abstract

多媒體的展現不僅需要顯示與時間無關的媒體 (例如文字, 影像或圖形) , 也需要顯示與時間有關的媒體 (例如聲音, 動畫或視訊) . 換句話說, 多媒體展現是由一群有時間先後順序的動作所組成, 而這些動作則是依照 使用者所預定的時間順序來執行. 針對這些動作所作的時間先後順序的安 排, 稱為這些動作的同步. 為了解決這些動作的同步問題, T. D. C. Little提出了以 OCPN (Object Composition Petri Net)模型為基礎的儲 存模型 (Storage Model)和展現演算法 (Presentation Algorithm) . 然 而, Little所提出的方法並不適用於交談式的多媒體展現.在本論文中, 我們修改 Little 所提出的儲存模型和展現演算法, 使其能夠適用於交談 式的多媒體展現. 我們並且提出一種語言 (MPL)來描述交談式的多媒體展 現, 而 MPL解譯器的實作則是在 SUN工作站上的 Open Window環境下完 成. Presenting a multimedia object not only means the displaying of some time-independent media, but also involves time-dependent media. That means the presentation of a multimedia object consists of a set of temporally related actions, which have to be executed in a special intended sequence defined by the user. This scheduling is generally called synchronization of actions. To solve the problem of synchronization of actions, T. D. C. Little presented a storage model and a presentation algorithm based on Object Composition Petri Net model. However, the approaches proposed by Little can not be applied to the presentation with user interaction. In this thesis, we present a modified storage model and a modified presentation algorithm which are suitable for the interactive presentation. We also propose a script language, MPL , to describe the synchronization of actions for the interactive presentation of multimedia objects. The implementation of the MPL interpreter is executed under OpenWindow environment on SUN workstation.