臺灣博碩士論文加值系統

本論文提出的人機互動系統讓使用者配戴頭戴式攝影機以取得接近使用者視點的影像，再從此影像中追蹤投影機投射在投影幕上的四邊形區塊。一旦影像中使用者的手指遮蓋住部份的投影區塊，利用系統本身已知的顯示內容截圖與變換後的投影區塊進行背景相減，便能得到使用者的手部區塊。找出手部區塊中的指尖位置及數量來模擬滑鼠的游標位置與按鍵動作，便能讓使用者對著螢幕內容進行選取拖放等手勢操作。
由於本系統適合在單一投影幕的教學環境中進行即時的團體合作，我們設計了一套能讓多位學生共同進行互動式的英文學習實驗活動，待未來本系統完成多攝影機的影像並行處理之後，便能評估此系統在互動性、學習動機或學習成效上相對於其他遠距指向介面的差異。

In this thesis, we propose a human-computer interaction system for a user, who wears a head-mounted camera to obtain live images from the user’s viewpoint. We subsequently tracked the real projection screen area in front of the user in the captured image sequence. Once the user points to the targeted area with fingers, the system extracts the hand region through background subtraction with screenshot and transformed projection area. By determining the positions and the number of fingertips, users can manipulate the content on the screen using intuitive finger gestures, such as selecting, dragging, and dropping.
Because our system is suitable for real-time group cooperation and interaction on a projection screen, we proposed multi-user interactive English-learning activities with scrambled word ordering. After we implement multi-camera image processing for our system in the future, we will evaluate the learning interactivity, the participants’ motivation, and learning effectiveness of the system relative to other remote pointing interfaces

致謝 I
摘要 II
Abstract III
目錄 IV
圖目錄 VI
第一章 緒論 １
1.1 研究動機 １
1.2 研究目的 ２
1.3 系統概觀 ３
第二章 文獻探討 ５
2.1 指向識別分類 ５
2.1.1 傳統指向裝置 ５
2.1.2 三維定位指向裝置 ５
2.2 視覺式指向識別 ６
2.3 綜合比較 ７
第三章 研究方法 ９
3.1 環境配置 ９
3.2 校正階段 ９
3.2.1 投影區塊定位 １１
3.2.2 計算射影變換矩陣 １２
3.2.3 擷取外框顏色 １２
3.2.4 找出系統平均延遲時間 １３
3.2.5 等距顏色取樣 １３
3.3 操作階段 １７
3.3.1 投影區塊追蹤 １９
3.3.2 背景預估與前景區塊分割 ２０
3.3.3 指尖偵測 ２１
3.3.4 手勢事件判別 ２２
第四章 英文單字學習實驗設計 ２５
第五章 結果討論與未來展望 ２９
5.1 結果討論 ２９
5.2 未來展望 ３０
參考文獻 ３１

[1] Google Project Glass, https://plus.google.com/+projectglass/posts
[2] Interactive whiteboard, http://en.wikipedia.org/wiki/Interactive_whiteboard
[3] Miguel A. Nacenta, Samer Sallam, Bernard Champoux, Sriram Subramanian, and Carl Gutwin. Perspective cursor: perspective-based interaction for multi-display environments. CHI, 289-298, ACM, 2006.
[4] Nintendo Wii Remote, http://www.nintendo.com/wii/what-is-wii/#/controls
[5] WiiYourself!, http://wiiyourself.gl.tter.org/
[6] Smoothboard, http://www.smoothboard.net/
[7] Mi-Suen Lee, Daphna Weinshall, and Eric Cohen-Solal. A Computer Vision System for On-Screen Item Selection by Finger Pointing. CVPR 1, 1026-1033, IEEE Computer Society, 2001.
[8] Roland Kehl, and Luc J. Van Gool. Real-Time Pointing Gesture Recognition for an Immersive Environment. FGR, 577-582, IEEE Computer Society, 2004.
[9] Andrew Wu, Mubarak Shah, and Niels da Vitoria Lobo. A Virtual 3D Blackboard: 3D Finger Tracking Using a Single Camera. FG, 536-543, IEEE Computer Society, 2000.
[10] Sylvia M. Dominguez, Trish Keaton, and Ali H. Sayed. A Robust Finger Tracking Method for Multimodal Wearable Computer Interfacing. IEEE Transactions on Multimedia, (8)5:956-972, 2006.
[11] Yi-Ping Hung, Yao-Strong Yang, Yong-Sheng Chen, Ing-Bor Hsieh, and Chiou-Shann Fuh. Free-Hand Pointer by Use of an Active Stereo Vision System. ICPR, pp. 1244-1246, 1998.
[12] Zhengyou Zhang, Ying Wu, Ying Shan, and Steven Shafer. Visual Panel: Virtual Mouse, Keyboard and 3D Controller with an Ordinary Piece of Paper. In Proc. ACM Workshop on Perceptive User Interfaces, 1-8, 2001.
[13] Masaaki Fukumoto, Yasuhito Suenaga, and Kenji Mase. "Finger-Pointer": Pointing interface by image processing. Computers &; Graphics, (18)5:633-642, 1994.
[14] Miguel Nacenta. Computer Vision approaches to solve the screen pose acquisition problem for Perspective Cursor. Tech. Rep. HCI-TR-06-01, 2006.
[15] Ramer–Douglas–Peucker algorithm, http://en.wikipedia.org/wiki/Ramer-Douglas-Peucker_algorithm
[16] Online Word Scramble Puzzles, http://www.proprofs.com/games/player/wardraw/
[17] GlovePIE, http://glovepie.org/glovepie.php
[18] ARToolKit, http://www.hitl.washington.edu/artoolkit/