隨著電子科技的發展，家庭自動化系統的對於使用者提供非常大的幫助，以往各項電器設備的操作介面多半是以現場啟動或由個別的遙控器操作，本實驗室發展出一套輔助性家電控制系統，可將生活上必須的電器設備整合至個人電腦。並發展出一輔助性電腦輸入裝置，提供身心障礙者經由操作嘴控開關，以摩斯碼輸入方式來控制鍵盤及滑鼠，達到操作各項電器用品的目的。唯使用者仍需由他人將嘴控開關放置於使用者嘴上，尚無法達到全自動化功能。因此本研究提出一套自動電腦操控開關載運系統，讓使用者不須經由他人幫忙，便可自行取用或收回嘴控開關，以提高身心障礙者使用整個系統的完整性。

系統架構採用三軸線性機器人，搭載伺服馬達及齒型軌道，以提供可控制的動力來源，配合CCD及個人電腦進行影像處理軟體，可在25 cm * 25 cm 的面積內搜尋使用者臉部所在的位置。手臂的動作流程可選擇自動及手動模式，以圖控介面控制伺服馬達六向運動. 並可學習(更新)使用者圖庫資料，將嘴控開關準確送至使用者的嘴巴內。

於工廠初步以模型樣品測試系統平均成功率為100%，平均辨識相似度為96.5%，平均每次耗費時間為1分32秒，於個案測試時因測試場地不同而使得成功率降低，調整各項參數後辨識成功率約為75%。此系統之可行性及實用性仍有改善空間，另外整體體積也是後續需研究可改進之處。

There are many inconveniences with the disabled in the daily life, especially for quadriplegics. They almost have to be helped with everything by others and their lives are quite boring. As the reasons, we developed an assistive computer input device to help the disabled control a keyboard and a mouse by a nipple switch with Morse code encoding. In addition, an appliance control system was developed to help the disabled control appliances by themselves including TV, AC,…, and so on. However, before the disabled operate the computer, the assistive computer input device still has to be placed into the mouth of the user by others. Therefore, we propose an automatic transportation system to place/take the nipple into/from the mouth to make the disabled use the whole system more convenient without others’ help.

The hardware was constructed from mechanical supports, servo motors and gears. The software was programmed by LabVIEW. The system can learn and update the pattern data of user’s face to make the nipple switch put into the user’s mouth correctly.

The experiment in the factory, the average accuracy is 100%, the recognition similar rate is 96.5%, and the average time of transporting the nipple switch to the mouth is 96 seconds. For case study, the average accuracy is reduced because of environment differentness. After adjusting each parameter, the average accuracy is about 75%. The proposed system is a practical system, which can help the disabled operate the entire system more convenient. But the feasibility and the usability of the system still can be improved in the future research, especially the size of system.

[1] 吳崇民著，”應用於重度及脊髓損傷患者之摩斯碼模糊控制辨識嘴控系統” ，國立成功大學電機工程學系博士論文，中華民國93年11月。
[2] 王健宇撰，”可變形樣板於即時人臉追蹤系統之應用”，國立中山大學機械與機電工程學系碩士論文，中華民國96年7月。
[3] 陳宣銘撰，”Non-Photo-Realistic Illustrations with Artistic Style”，國立中央資訊工程學系碩士論文，中華民國92年12月。
[4] Baogang Wei, Yonghuai Liu, Member, IEEE, and Yunhe Pan，“Using Hybrid Knowledge Engineering and Image Processing in Color Virtual Restoration of Ancient Murals”，IEEE Transactions On Knowledge and Data Engineering, vol. 15, no. 5, pp. 213-318, September/October 2003.
[5] 曾郁展著，”利用人臉五官為特徵之人臉辨識系統”，國立中山大學電機工程學系碩士論文，pp. 13-18，中華民國94年6月。
[6] 鍾慶豐著，” 探究影像辨識裡的臉部偵測技術”，電子與電腦資訊網。
[7] M. Propp and A. Samal, “Artificial Neural Network Architectures for Human Face Detection,” Intelligent Eng. Systems through Artificial Neural Networks, vol. 2, 1992.
[8] H. A. Rowley, S. Baluja, and T. Kanade, 'Neural network-based face detection,' IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 20, no. 1, pp. 23-38, 1998.
[9] NI-IMAQ Concept User Guide”2005 edition National Instruments Corporation.
[10] MITSUBISHI ELECTRIC MELSERVO MR-J2-SUPER user manual。
[11] LabVIEW user manual 2001 edition National Instruments Corporation。
[12] LabVIEW Toolkit Mind storms NXT Programming Guide 2006 edition National Instruments Corporation.
[13] 林朝枝、葉建廷、劉木坤著，”An algorithm for solving convex hull problem 向外凸殼問題的演算法” ，吳鳳科技大學資訊工程系，pp. 32-35。。
[14] 陳秋旺、朱銘祥、謝孟達、張慧怡、陳家進”肘關節復健用機械人之研究”中華民國八十八年醫學工程科技研討會論文集，pp. 150-151。