臺灣博碩士論文加值系統

許多人都有物品遺留在抽屜的經驗，本智慧書桌在使用完畢時，能夠自動檢測抽屜是否有物品遺留，若有物品遺留就提醒使用者。本論文提出的智慧書桌結合壓力感測器(Force-Sensing Resistor, FSR)、Arduino、raspberry pi 3 model b(簡稱PI 3)、CCD攝像頭來完成此智慧裝置。壓力感測器搭配Arduino感測有無人使用，在使用者離開時使用CCD攝像頭對抽屜拍照，把取得的照片引用PI3的opencv之函式做影像處理。先由濾波器去雜訊後灰階化，最後再和已知的空抽屜的照片(同樣經過濾波、灰階化)做背景相減法，再將兩者相減完的結果做二值化，用以判斷是否有物品在抽屜內，若有物品遺留就會透過PI3的GIPO埠輸出電壓，使LED閃爍提醒使用者。智慧書桌經過實驗得知，外來物佔照片的5.7%以上可被偵測。從壓力感測器判斷有無人使用到將資訊傳達給PI3的時間為0.013秒，而PI3運算完到LED呈現結果的時間為5.329秒。

This paper proposes a device for detecting the forgotten objects in a desk drawer. The device consists of the pressure sensing system and the object detecting system. The pressure sensing system detects the appearance of a seated person. The object detection system detects the foreign objects with the following operations (1) Software on computer receives the information from the pressure sensing system; (2) when information indicates that user leaves the chair, CCD camera takes a picture from computer in the desk drawer; (3) the picture is named "current-image"; (4) converting "current-image" to grayscale image; (5) calculating the difference between the "current-image" and the image of empty desk based on background subtraction; (6) converting the background subtraction result to binary image; (7) get the number of white pixels in the binary image; (8) and if the white pixels in the binary image exceeds a certain threshold, computer flashes LEDs. Experiment results show: (a) When different pixels take up over 5.7% of the “current-image”, the object detection system determines that there is foreign objects in the desk drawer; (b) the pressure sensing system sends data to the object detection system in 0.013 seconds, and the object detection system presents calculation result in 5.329 seconds.

誌謝 iii
摘要 iv
ABSTRACT v
目次 vi
表次 ix
圖次 x
第壹章 緒論 1
1.1 研究動機 1
1.2 研究背景 1
1.3 貢獻 1
1.4 論文架構 1
第貳章 背景知識與文獻探討 2
2.1 背景知識 2
2.1.1 壓力感測器 2
2.2 影像處理文獻探討 5
2.2.1 灰階化 5
2.2.2 物體探測文獻探討 6
2.2.3 二值法文獻探討 8
第參章 研究構想與開發 11
3.1 研究流程 11
3.2 研究設計概念 11
3.3 智慧書桌系統開發 15
3.3.1 研究構想 15
3.3.2 系統規劃設計 16
3.3.3 壓力感測系統規劃 16
3.3.4 物體偵測系統規劃 20
第肆章 研究方法與結果 24
4.1 運算時間 24
4.2 物體大小 24
4.3 其他物體量測 26
4.4 角度量測 27
4.5 廣角鏡測量 27
4.6 教室量測 28
第伍章 結論與建議 29
5.1 結論 30
5.2 建議 30
參考文獻 31

[1]A. H. A. Razak, A. Zayegh, R. K. Begg, “A wireless system-on-chip for MEMS biomedicail plantar pressure sensor,” in Intelligent and Advanced Systems(ICIAS), 2012 4th International Conference on , Volume 1, pages 78-81, June 2012
[2]私立大同大學,戴宗霖” 穿戴式足部壓力感測裝置之系統研究”
[3]“ FSR 402 Data Sheet, ” in Interlink Electronics, [Online]. Retrieved fromhttp://www.interlinkelectronics.com/datasheets/Datasheet_FSR.pdf
[4]吳裕弘，應用熱壓嵌入技術製作電容式壓力感測器，大同大學機械工程研究所碩士論文，2008
[5]Podpora, Michal, Grzegorz Pawel Korbas, and Aleksandra Kawala-Janik. "YUV vs RGB-Choosing a Color Space for Human-Machine Interaction." FedCSIS Position Papers. 2014.
[6]Li, Congcong, et al."Joint Optimization Of Background Subtraction AND OBJECT DETECTION FOR NIGHT SURVEILLANCE."
[7]Piccardi, Massimo. "Background subtraction techniques: a review." Systems, man and cybernetics, 2004 IEEE international conference on. Vol. 4. IEEE, 2004.Novel.de, “ emed®-systems, ” [Online].
Retrieved from http://www.novel.de/novelcontent/emed
[8]鍾宜岑, and 林進燈. 應用於動態背景中的移動物體影像之偵測與即時追蹤系統. Diss. 2007.
[9]http://46aae4d1e2371e4aa769798941cef698.devproxy.yunshipei.com/d14665/article/details/41990481
[10]Otsu, Nobuyuki. "A threshold selection method from gray-level histograms." Automatica 11.285-296 (1975): 23-27.resolution, ” in Journal of Biomechanics 26(3), pages 301-301, March 1993.
[11]“大同大學生輔組”,http://assistance.ttu.edu.tw/files/15-1040-10076,c1067-1.php?Lang=zh-tw
[12]https://www.bnext.com.tw/article/34043/BN-ARTICLE-34043
[13]http://www.damanwoo.com/node/72676
[14]李政裕,”車牌辨識系統”,靜宜大學資訊管理研究論文2003.
[15]中國機械工程學會第三十一屆全國學術研討會論文集, “應用於車輛安全之行人偵測系統設計”, 朱振緯、張國清, 中華民國一百零三年十二月六、七日
[16]國立台北科技大學,盧晨皓”及時多重物體追蹤之系統設計”,2009
[17]國立中山大學,葉家宏,Edge detection