以Kinect為基礎之人機介面研製
摘要
自2010年11月微軟遊戲機Xbox 360推出不需使用搖桿的「動態感應攝影機(Kinect)」產品來進行遊戲操控,它可提供使用者不需手持或腳踩控制器,而是語音指令或手勢操作Xbox 360的系統介面,它也能捕捉使用者全身上下的動作,將玩遊戲從遙控器的手指運動精進到身體部分關節的律動即可達成控制之目的。因此,引發許多研究者的興趣,嘗試利用Kinect與Windows 7作業系統中進行人機介面研究以便讓使用者達到方便及樂趣性。 因此本論文兩種不同手勢的人機介面來將Kinect與電腦整合,且分別用「網頁瀏覽」、「文字輸入」及「簡報瀏覽」等情境實驗來驗證此兩種手勢介面之效果。試驗結束後,請受測者完成包含朵種問題之問卷,俾了解實驗過程中,使用者可藉由Kinect以不同手勢操控電腦的可用度、功能、接受度與反應為何?另依其意見進行分析與了解,並獲得相關結論,手勢的設計需符合盡量簡單及直覺,也須符合人體工學才能讓使用者輕易操作電腦及快速上手使用。此外,圖示影像須夠大才能讓使用者勿須擔心精準定位之問題。

關鍵字:人機介面、手勢、Kinect

 

 

A Study of a Kinect –based Human-Computer Interface
Abstract
In the end of 2010, the Microsoft Xbox 360 launches the "motion sensor camera (Kinect)" product with which players don’t need hand-held or foot-pedal-based controllers to play games any more. The Kinect provides users a very user-friendly way to control Xbox 360 systems via voice commands and/or body gestures. The Kinect can capture the user's upper and lower body movements. It allows users to play games via body movements without the need of dexterous finger movements on a remote controller. Therefore, many researchers are interested in trying to integrate a Kinect with the Windows 7 operating system to allow users to manipulate computers in a user-friendly and joyful way. This paper presents two different Kinect-based gesture-based interfaces to manipulate computers. The performance of the proposed interfaces was conducted on three experimental scenarios: the "web surfing", the "text typing task", and the "flipping through PowerPoint". At the end of the experiments, subjects were asked to fill out a questionnaire consisted of several statements representing usability, functionality, acceptance and responses of the system. Several observations can be made. First of all, the gestures must be designed to be as simple and intuitive as possible. In addition, they have to meet ergonomics such that users can easily manipulate computers. The icon graphics must be large enough so that users do not need to worry about the precise location problem.

Keywords : human-computer interface, gesture, Kinect.