博碩士論文 983203107 詳細資訊


姓名 伍展佑(zhan-you Wu)  查詢紙本館藏   畢業系所 機械工程學系
論文名稱 高效能駕駛狀態監控系統之研發
檔案 [Endnote RIS 格式]    [Bibtex 格式]    至系統瀏覽論文 ( 永不開放)
摘要(中) 在我國,疲勞駕駛、分心而未注意車前狀況等人為疏失情況是交通事故發生的重要原因,因此本研究目的在開發一套即時且非接觸式的高效能駕駛狀態監控系統,透過影像處理方式即時偵測駕駛人的面朝方向與眼睛開闔狀態,進而精確判斷出駕駛人的疲勞與分心狀態。本系統將攝影機輸入的畫面,透過移動偵測、邊緣偵測、區域亮度偵測、樣版比對、連通域標記演算法、區域特徵分析等影像處理方法,在動態畫面中即時追蹤頭部和眼睛的所在位置,再利用頭部中心與兩眼相對位置來判斷駕駛者面朝方向,並在眼睛區域內分析其特徵,判斷眼睛開闔狀態。最後,本系統以面朝方向角度資訊來判斷駕駛者的分心狀態,並以單位時間內的閉眼時間比例(PERCLOS)來判斷駕駛者的疲勞狀態。
本研究除了完成系統驗證外,更進行多種臉部外型(眼鏡、膚色、髮型、頭部大小)及實車環境(背景複雜度、環境亮暗、對向車眩光)等因素的測試,發現本系統在多數臉部外型及環境情況下,判斷面朝方向的角度誤差都在4°之內,而閉眼辨識率也都在99%以上。最差的情況下,面朝方向角度誤差也在8.5°以內,閉眼辨識率則在78.85%以上,證實本系統於不同臉部外型和背景環境下都能有良好的辨識能力。未來本系統將結合車內影音或座椅及方向盤振動等警示裝置,有效的提醒駕駛人,以避免事故的發生。
摘要(英) Some human errors such as driver fatigue and distraction are the main reason for traffic accidents in Taiwan. In this thesis, we develop a real-time, non-contact and high-Performance driver status detection system. Through image analysis, the system can estimate the state of the driver, such as distraction and fatigue. The system tracks the driver’’s head and eyes by motion detection, edge detection, brightness detection, template matching, connected component labeling algorithm and feature analyzing. Then determine the direction of the driver’’s face by Position of head and eyes. Finally, the system calculates the direction of the driver’’s face and PERCLOS to determine the status of the driver.
In this study, we test the system in many factors including eyeglasses, skin color, hair, head size and complexity of the background, brightness and backlight on a real car. We found that in most cases, the errors of facing direction are within 4°, and the detection rate of eyes closed in more than 99%. In the worst case, the errors of facing direction are within 8.5°, and the detection rate of eyes closed in more than 78.85%. It confirmed that the system has good recognition ability. In the future, we will combine audio or vibration alert device to avoid traffic accidents.
關鍵字(中) ★ 影像辨識
★ 駕駛狀態偵測
★ 人臉辨識
★ 駕駛安全輔助
★ PERCLOS
關鍵字(英) ★ PERCLOS.
★ Driving safety support systems
★ Face recognition
★ Driver state detection
★ Image recognition
論文目次 摘要 i
ABSTRACT ii
誌謝 iii
目錄 iv
圖目錄 vii
表目錄 x
符號說明 xi
第一章 緒論 1
1-1 研究動機 1
1-2 研究目的 2
1-3 論文架構 4
第二章 相關文獻 5
2-1 影像辨識之應用 5
2-2 影像辨識之方法 5
第三章 影像處理方法 10
3-1 系統架構流程 10
3-2 頭部搜尋與追蹤 10
3-2-1 灰階處理 10
3-2-2 移動偵測 13
3-2-3 邊緣偵測 15
3-2-4 建立頭部橢圓樣板 17
3-2-5 搜尋與追蹤模式之橢圓樣板比對 19
3-3 眼睛搜尋與追蹤 22
3-3-1 自動測光與二值化處理 22
3-3-2 連通域標記演算法 24
3-3-3 建立眼睛樣板 27
3-3-4搜尋與追蹤模式之眼睛樣板比對 31
3-3-5 眼睛開闔狀態判斷 34
3-4 分心與疲勞之判斷 35
3-4-1 面朝方向估計與分心判斷 35
3-4-2 閉眼時間比例估計與疲勞判斷 38
3-5 實驗方法 39
3-5-1 系統軟硬體 39
3-5-2 系統驗證 44
3-5-3 臉部外型 46
3-5-4 實車環境 48
第四章 結果與討論 51
4-1 系統驗證 51
4-1-1 面朝方向判斷 51
4-1-2 閉眼判斷 54
4-2 臉部外型之影響 55
4-2-1 眼鏡之影響 56
4-2-2 膚色之影響 57
4-2-3 髮型之影響 57
4-2-4 頭部大小之影響 58
4-3 實車環境之影響 59
4-3-1 背景複雜度之影響 59
4-3-2 環境亮暗之影響 59
4-3-3 對向車眩光之影響 60
4-4 最差因素組合 61
第五章 結論與未來發展方向 63
5-1 結論 63
5-2 未來發展方向 64
參考文獻 66
參考文獻 [1] 內政部警政署,內政部警政署統計通報,上網日期:民國100 年7 月,網址:http://www.npa.gov.tw/NPAGip/wSite/lp?ctNode=12594&xq_xCat=16&nowPage=2&pagesize=15&mp=1。
[2] 林家吉, “利用人臉辨識追蹤系統於鎖定物體之保護”, 逢甲大學資訊電機工程學系碩士論文, 2008.
[3] 曹力仁, “不均勻照度環境的駕駛昏睡偵測與警示”, 國立中央大學資訊工程研究所碩士論文, 2008.
[4] 曾凱群, “應用影像辨識設計疲勞偵測系統”, 國立台北教育大學資訊科學系碩士論文, 2008.
[5] 何肇偉, “人臉辨識應用於電腦使用者眼睛保護之研究”, 國立台北科技大學機電整合研究所碩士論文, 2002.
[6] Davis, J. W. and Keck, M. A., “A two-stage template approach to person detection in thermal Imagery,” Proc. IEEE International Workshop on Applications Computer Vision, vol.1, pp. 364-369, January 2005.
[7] Zhu, Z. and Ji, Q., “Robust Real-Time Eye Detection and Tracking Under Variable Lighting Conditions and Various Face Orientations”, Computer Vision and Image Understanding, Vol. 98, No. 1, 2005, pp.124-154.
[8] Hansen, D.W. and Hammoud, R.I., “An improved likelihood model for eye tracking”, Computer Vision and Image Understanding, Vol. 106, 2007, pp.220-230.
[9] Chai, D. and Ngan, K.N., “Locating Facial Region of a Head-and-Shoulders Color Image,” Proc. Third Int’l Conf. Automatic Face and Gesture Recognition, pp. 124-129, 1998.
[10] Soriano, M., Huovinen, S., Martinkauppi, B. and Laaksonen, M.,“Using the skin Locus to cope with Changing Illumination Conditions in Color-Based Face Tracking,” Proc. of IEEE Nordic Signal Processing Symposium, pp. 383-386, 2000.
[11] Singh, Sanjay Kr., Chauhan, D. S., Vatsa Mayank and Singh Richa, “A Robust Skin Color Based Face Detection Algorithm,” Tamkang Journal of Science and Engineering, Vol.6, No.4, pp. 227-234, 2003.
[12] Chung, Y. C., Wang, J. M. and Chen, S. W., “ Progressive Background Image Generation”, Proceedings of 15th IPPR Conf. on Computer Vision, Graphics and Image Processing, pp 858-865, 2002.
[13] Phung, S. L., Bouzerdoum, A. and Chai, D., “A Novel Skin Color Model in YCbCr Color Space and Its Application to Human Face Detection” , Proceedings of the International Conference on Image Processing, vol. 1, pp 289 – 292, 2002.
[14] Gottumukkal, R. and Asari, V.K., “Real Time Face Detection from Color Video Stream Based on PCA Method”, Proc. of 32nd Applied Imagery Pattern Recognition Workshop, 2003, pp.146-150.
[15] Morris, T., Blenkhorn, P. and Zaidi, F., “Blink detection for real-time eye tracking”, Journal of Network and Computer Applications, Vol. 25, No. 2, 2002, pp.129-143.
[16] Bhaskar, T.N., Keat, F.T., Ranganath, S. and Venkatesh, Y.V., “Blink Detection and Eye Tracking for Eye Localization”, TENCON 2003 Conf. on Convergent Technologies for Asia-Pacific Region, Vol. 2, 2003, pp.821-824.
[17] Chau, M. and Betke, M., “Real Time Eye Tracking and Blink Detection with USB Cameras”, Boston University Computer Science Technical Report No. 2005-12.
[18] Tan, H.C. and Zhang, Y.J., “Detecting eye blink states by tracking iris and eyelids”, Pattern Recognition Letters, Vol. 27, 2006, pp.667-675.
[19] Zhang, S.C. and Liu, Z.Q., “A Real-time Face Detector”, IEEE Int’l Conf. on Systems, Man and Cybernetics, 2004, pp.2197-2202.
[20] Miyakawa, T., Takano, H. and Nakamura, K., “Development of Non-contact Real-time Blink Detection System for Doze Alarm”, SICE Annual Conf., 2004, pp.1626-1631.
[21] 許尹融, “即時眼部狀態偵測系統之研究”, 國立中央大學機械工程學系碩士論文, 2008.
指導教授 黃俊仁(Jiun-ren Hwang) 審核日期 2011-8-24
推文 facebook   plurk   twitter   funp   google   live   udn   HD   myshare   reddit   netvibes   friend   youpush   delicious   baidu   

若有論文相關問題,請聯絡國立中央大學圖書館推廣服務組 TEL:(03)422-7151轉57407,或E-mail聯絡