長久以來,電腦人機介面的操作一直以滑鼠、鍵盤、手寫板為主。近年的人機介面逐漸朝直覺更人性化的操控模式發展。藉由更貼近人類自然行為,改善操作介面的效率。關於電腦視覺的技術日益成熟,已經發展出許多不同的應用技術,在各種領域中協助人類提升工作績效。本研究以電腦視覺為基礎,發展非接觸式人機介面操作技術,可辨識移動物體,讓使用者無需穿戴任何感測設備,即可與電腦進行互動。研究中,設計四種不同操作情境應用案例,驗證技術之可行性,包括手勢辨識、物體移動辨識,以及多人同時使用。結果証實所發展之互動介面,可讓使用者即時與系統進行互動,而且能應用在多種不同的操作情境。以非接觸式的操控,設備將不易損壞,運用在互動介面上,可有效降低人機介面硬體裝置的成本。介面所需之硬體設備相對簡易低廉。
Traditionally, the computer system is always driven by the mouse and the keyboard. In recent year, the man-machine interface has progressed rapidly, by which users can manipulate the system more friendly and intuitively with nature act of people. The technology of computer vision has been developed maturely to apply to the various fields. This paper presents the technique of interactive interface based on computer vision, which can tract and position the moving objects and recognize the gesture. Through the proposed method, users can move their body or hand to interact with a machine system real time without contacting any controlling device or expensive sensor. In this paper, four demonstrated media systems have been implemented to prove the proposed interactive interface is effective in different modes and environment.