簡易檢索 / 詳目顯示
| 研究生： |
黃瑩發 Huang, Ying-Fa |
|---|---|
| 論文名稱： |
適用於肢體障礙者之電腦操作輔具 Computer Operation Assistance Technology Suitable for Disability |
| 指導教授： |
羅錦興
Luo, Ching-Hsing |
| 學位類別： |
碩士 Master |
| 系所名稱： |
電機資訊學院 - 電機工程學系碩士在職專班 Department of Electrical Engineering (on the job class) |
| 論文出版年： | 2011 |
| 畢業學年度： | 99 |
| 語文別： | 中文 |
| 論文頁數： | 62 |
| 中文關鍵詞： | 輔具 、類神經 、倒傳遞類神經 、臉部偵測 、眼睛偵測 |
| 外文關鍵詞： | Haar feature, Adaboost, BPNN, face detection, eye detection |
| 相關次數： | 點閱：87 下載：10 |
| 分享至: |
| 查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
本論文以一般常見的Webcam改造成可視IR LED光源，並且在鏡頭上貼磁片或底片濾除大部份的環境光使得影像處理的前置作業簡化並將複雜背景影像的干擾降低。以一般隨手可得的吸管浮貼反光片做為滑鼠替代工具，無電源無外接線路非常輕便及無耗電。以反射IR LED的光點移動轉換滑鼠游標移動功能，以嘴巴輕吹氣造成浮貼反光片亮滅做為左鍵、右鍵功能，並搭配免費虛擬鍵盤軟體和Windows系統內建滑鼠設定參數可實現大部分一般滑鼠的功能，例如拖曳。本系統可以閉單眼來控制步進馬達收取吸管，讓使用者可以自行收取工具。經測試，本系統在室內人工照明環境、室內自然光環境、複雜背景下使用，都有很好的動作準確度。眼睛偵測部份，使用Haar feature做為辨識特徵、Cascade Adaboost分類器做為學習機制以偵測人臉位置，為了使閉眼特徵偵測準確度提高再以倒傳遞類神經(BPNN)學習並辨識。經測試後，在室內環境下使用不論在自然光源或人工照明光源環境，使用者在未戴眼鏡的情況下有100%的閉眼辨識準確率，帶眼鏡使用依然有95%以上的閉眼辨識準確率。整套系統的製作成本遠比市面上的電腦輸入輔具便宜並且容易製作與維護，使用者在短時間內就可適應操作和學習，長時間操作造成身體的不適程度也大幅減輕。
In this study, the common webcam was transformed to a webcam that could sense the light source of IR LED, and a floppy disk or negative was attached to the camera lens to filter most of the light from the environment so as to simplify the work of image pre-processing and lower the interference of background image. Commonly found straw attached light reflection sheet was used to replace the mouse, power or external circuit was not necessary, so this alternative is very light, convenient, and power-free. Reflecting the light spot movement of IR LED to replace the cursor moving of mouse, blowing through the straw to light or dim the attached light reflection sheet to serve as the functions of left click and right click, and using free virtual keyboard software and built-in mouse setting parameters of Windows to realize most of the common mouse functions, such as dragging. In this system, blinking single eye would control the stepping motor to collect the straw, so the users could collect the tools by themselves. The test result showed that this system provided very good action accuracy in the indoor artificial illuminating environment, indoor environment with natural light, and complicated background. For eye detection, Haar feature is used to recognize the features and Cascade Adaboost classifier is the learning mechanism to detect face position. To promote the detection accuracy of eye blinking features, backpropagation neural network (BPNN) is learned and recognized. Regardless of using natural light or artificial illumination in the indoor environment, the test result showed that 100% eye blinking detection accuracy could be obtained when the user was not wearing glasses, while 95% eye blinking detection accuracy could be obtained when the user was wearing glasses. The cost of making the whole set of system is way cheaper than the computer-controlled assistance tool sold in the market and it is also easy to make and maintain. The user could adapt to the operation and learn this system in a short period of time, and the discomfort caused by operating it for a long period of time has also be reduced substantially.
[1] Y. Tomita, Y. Igarashi, S. Honda, and N. Matsuo, “Electro-Oculography Mouse for Amyotrophic Lateral Sclerosis Patients,” in IEEE Conference Engineering in Medicine and Biology Society, Nov. 1996, vol. 5, pp. 1780-1781.
[2] T. Miyakawa, H. Talano, and K. Nakamura, “Development of Non-contact Real-time Blink Detection System for Doze Alarm,”in Proc. Of SICE Anmual Conference, 2004, vol. 2, pp. 1626-1631.
[3] T. Donnely, P.J. Daver and S. Carlyon, “Laser-Operated Mouse For a Physical Disabled Child,” Institute of Electronic Engineering Colloquium on, pp.8/1~8/March 1997
[4] G.. Yang and T. S. Huang, “ Human Face Detection in Complex Background, Pattern Recognition, ” vol. 27, no. 1, pp. 53-63, 1994.
[5] K.C. Yow and R. Cipolla, “Feature-Based Human Face Detection,” Image and Vision Computing, vol. 15, no. 9, pp. 713-735, 1997.
[6] I. Craw, D. Tock, and A. Bemmett, “Finding Face Features,” Proc. Second European Conf. Computer Vision, pp. 92-96, 1992.
[7] Lin-Lin Huang, Akinobu Shimizu, Yoshihiro Hagihara, Hidefumi Kobatake, “ Face detection from cluttered images using a polynomial neural network,” Neurocomputting vol. 51, 2003, pp. 197-211.
[8] Ming-Hsuang, Yang, David J. Kriegman, narendra Ahuja, “Detecting Faces in Images: A Survey,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 24, 2002, pp. 34-58.
[9] Paul Viola and Michael J. Jones, “Rapid Object Detection using a Boosted Cascade of Simple Features,” IEEE CVPR, 2001.
[10] R. Jain, R. Kasturi, and B.G. Schunck, Machine Vision, McGraw-Hill Science/Engineering/Math, 1995.
[11] 陳弦澤，「改良式紅外線眼控系統之研發與應用」，碩士論文，自動控制工程研究所，逢甲大學，民國九十三年五月。
[12] 蔡金源，「以眼球控制之殘障者人機介面系統:紅外線視動滑鼠」，碩士論文，電機工程研究所，國立台灣大學，民國八十六年。
[13] 郭靜男，「可眼控及頭控之多功能PC Camera之研發與應用」，碩士論文，自動控制工程研究所，逢甲大學，民國九十二年五月。
[14] 周英汶，「微機電重力角度感測式無線頭控滑鼠」，碩士論文，國立成功大學電機工程研究所碩士論文，民國八十九年六月。
[15] 99年度教育部補助技專校院建立特色典範計畫-自編教材”單元三：物件偵測分類器訓練”
[16] 吳上立, 林明德, “C語言數位影像處理,” 全華圖書公司, 2010.
[17] 穆紹綱, “數位影像處理活用Matlab第二版,” 全華圖書公司, 2010.
[18] 洪國勝, 江國軍, 龍國忠, 洪月裡, “C++ Builder 6,” 旗標出版公司, 2002
[19] 劉瑞禎, 于仕琪, “OpenCV 教程”, 北京航空航天大學出版社, 2009
[20] 劉瑞禎, 于仕琪, “學習OpenCV ”, 清華大學出版社, 2009
[21] 范逸之, 陳立元, 賴俊朋, “Visual Basic與RS232串列通訊控制”, 文魁資訊股份有限公司, 2000
[22] 羅華強, “類神經網路-MATLAB的應用”,清蔚科技出版, 2001
[23] 肢體障礙者用中文摩斯碼訓練系統 楊正宏 ; 莊麗月 ; 楊正輝 ; 羅錦興Journal of Medical and Biological Engineering22:增刊 民91.12頁S81-S90
[24] 殘障用中文摩斯碼溝通輔助系統 羅錦興 ; 施清祥 ; 施清添中華醫學工程學刊16:2民85.06頁214-230
校內：2013-08-26公開校外：2013-08-26公開