本論文發展一套以數位訊號處理( DSP )為基礎架構的視覺感知智慧型輔助車輛駕駛系統，使用CCD攝影機擷取車輛行進與道路的影像。在此系統中依天色狀況整合日間和夜間兩系統架構。本文系統分成兩個子系統，包括路面標線偵測與車輛偵測。透過道路的影像所偵測出的路面標線所形成的車行區域，在此區域中辨識前方和鄰近車道是否有車輛存在、估測相對距離以及本車是否偏離車道，提供智慧型自動導航車輛的決策依據，達到安全的輔助車輛駕駛。

The purpose of this thsis is to develop a vision perception based Intelligent Vehicle Driving Assistant System ( IVDAS ), which utilizes CCD camera to capture the movement of vehicle and road image on DSP-Based . According to daytime and night time, we analyzed the full information in the image to acquire the important and proper characteristics about lane mark and vehicle.

There are two sub-systems in our system , including Lane Mark Detection and Vehicle Detection. The main goal is to identify if there are existing vehicles in the front of or near our vehicle. This system can provide information for the Intelligent Vehicle to make decision to avoid accident happening and assisted driver in driving safely.

中文摘要……………………………………………………………………………I
英文摘要…………………………………………………………………………II
目錄…………………………………………………………………………………III
圖目錄……………………………………………………………………………VII
表目錄………………………………………………………………………………XI

第一章 緒論………………………………………………………………………1
1.1 研究動機…………………………………………………………1
1.2 研究目的…………………………………………………………………2
1.3 系統架構…………………………………………………………………3
1.4 論文組織…………………………………………………………………5

第二章 文獻回顧與影像處理……………………………………………………6
2.1 文獻回顧…………………………………………………………6
2.1.1 感測器的文獻回顧………………………………………………7
2.1.2 CCD路面標線偵測的相關文獻回顧………………………10
2.1.3 CCD車輛影像辨識的相關文獻回顧………………………15
2.1.4 CCD定位的相關文獻回顧……………………………………17
2.1.5 論文歸納及總結……………………………………………………19
2.2 影像處理技術理論……………………………………………………20
2.2.1 數位影像取樣原理…………………………………………………20
2.2.2 數位影像的表示法…………………………………………………21
2.2.3 以灰階表示的數位影像……………………………………………21
2.2.4 對比強化(Contrast Enhancement)……………………………22
2.2.5 二值化(Binarize)………………………………………………23
2.2.6 二值化影像的投影處理…………………………………………23
2.2.7 二維影像上尋找邊緣特徵………………………………………24
2.2.8 投影圖上的最大梯度變化………………………………………25
2.2.9二維影像的座標…………………………………………………25
2.2.10 動態影像特性介紹……………………………………………26

第三章 路面標線偵測…………………………………………………………28
3.1 前言………………………………………………………………………28
3.2 路面標線偵測架構………………………………………………………30
3.2.1 日間路面標線偵測架構…………………………………………30
3.2.2 夜間路面標線偵測架構…………………………………………31
3.3 路面標線的靜態影像資訊…………………………………………32
3.4 路面標線的動態影像特性………………………………………36
3.5 路面標線方程式………………………………………………………39

第四章 車輛偵測與距離估測…………………………………………………41
4.1 車輛特性分析…………………………………………………………41
4.1.1 日間車輛特性分析………………………………………………41
4.1.2 夜間車輛特性分析………………………………………………43
4.1.3 日間與夜間車輛特性的比較……………………………………44
4.2 車輛偵測系統架構……………………………………………………45
4.3 二值化處理……………………………………………………………47
4.3.1 動態調整二值化區域…………………………………………48
4.3.2 平均灰度值求閥值……………………………………………49
4.4 日間車輛辨識…………………………………………………………51
4.4.1 水平投影尋找前車底部陰影…………………………………51
4.4.2 雜訊濾除………………………………………………………53
4.4.3 標示出車輛左右邊界…………………………………………55
4.4.4 多車道車輛辨識………………………………………………56
4.5 夜間車輛辨識…………………………………………………………58
4.5.1 車輛尾燈的擷取…………………………………………………59
4.5.2 車輛可能位置……………………………………………………62
4.5.3 標示出車輛左右邊界……………………………………………65
4.5.4 雜訊濾除…………………………………………………………66
4.6光學距離估測…………………………………………………………68

第五章 硬體架構………………………………………………………………71
5.1 車輛辨識系統硬體架構簡介…………………………………………71
5.2 取樣設備簡介…………………………………………………………75
5.2.1 以PC為基礎之取樣設備…………………………………………75
5.2.2 以DSP為基礎之取樣設備…………………………………………76
5.2.3視訊標準與複合式端子介面………………………………………78
5.3 Blackfin ADSP-BF533……………………………………………79
5.3.1 ADSP-BF533 EZ-KIT Lite 相關規格……………………………80
5.3.2 ADSP-BF533 EZ-KIT Lite 介紹…………………………………81
5.4 Visual DSP++……………………………………………………………86
5.5 LCD……………………………………………………………………87
5.6 系統DSP架構流程……………………………………………………88


第六章 實驗方法與結果………………………………………………………89
6.1 錄製資料庫平台…………………………………………………………89
6.2 以PC為基礎驗證的環境………………………………………………90
6.3 以DSP為基礎應用的環境………………………………………………92
6.4 路面標線的實驗方法與結果……………………………………………95
6.4.1以PC為基礎之架構……………………………………………96
6.4.2 以DSP為基礎之架構……………………………………………99
6.5 車輛偵測的實驗方法與結果…………………………………………101
6.5.1以PC為基礎之架構……………………………………………102
6.5.2 以DSP為基礎之架構……………………………………………103
6.5.3錯誤判別率與辨識率之討論……………………………………105
6.6 系統效能評估…………………………………………………………108
6.6.1以PC為基礎之架構………………………………………108
6.6.2 以DSP為基礎之架構………………………………………109

第七章 結論與未來研究方向…………………………………………………110
7.1結論……………………………………………………………………110
7.2未來展望………………………………………………………………112

參考文獻…………………………………………………………………………114
圖 目 錄

圖(1-1) 論文系統架構圖……………………………………………………………3

圖(2-1) 防追撞系統的基本發展流程與架構………………………………………6
圖(2-2) RALPH系統………………………………………………………………11
圖(2-3) LOIS系統…………………………………………………………………11
圖(2-4) GLOD系統………………………………………………………………12
圖(2-5) 以假定的圓心進行投影…………………………………………………13
圖(2-6) 對300x40 的偵測區域進行標線偵測…………………………………13
圖(2-7) 彩色示意圖……………………………………………………………21
圖(2-8) 灰階示意圖……………………………………………………………22
圖(2-9) 對比強化示意圖………………………………………………………23
圖(2-10) 二值化示意圖…………………………………………………………23
圖(2-11) 水平與垂直投影………………………………………………………24
圖(2-12) 邊緣偵測遮罩…………………………………………………………24
圖(2-13) 消失點於正前方的三維透視圖………………………………………25

圖(3-1) 道路種類和天色差異……………………………………………………29
圖(3-2) 車行區域範圍……………………………………………………………29
圖(3-3) 日間路面標線偵測系統流程……………………………………………30
圖(3-4) 夜間路面標線偵測系統流程……………………………………………31
圖(3-5) 道路上某列的灰階值大小………………………………………………32
圖(3-6) 路面標線系統-圖片座標定義與名詞說明…………………………33
圖(3-7) 日間標示出在擷取影像中，可能的左右路面標線點……………………34
圖(3-8) 夜間路面資訊……………………………………………………………35
圖(3-9) 動態二值化的結果………………………………………………………35
圖(3-10) 夜間標示出在擷取影像中，可能的左右路面標線點……………………35
圖(3-11) 虛線路面標線在連續擷取的圖像中，移動的變化………………………36
圖(3-12) 路面標線特徵點…………………………………………………………37
圖(3-13) 虛線標線經過一定數量的Frame後，路面標線特徵會連續……………38
圖(3-14) 路面標線偵測結果………………………………………………………40

圖(4-1) 日間系統在不同的條件下，觀察前方的車底陰影………………………42
圖(4-2) 夜間系統在不同的條件下，觀察前方的車底陰影………………………44
圖(4-3) 日間車輛辨識系統架構…………………………………………………45
圖(4-4) 夜間車輛辨識系統架構…………………………………………………46
圖(4-5) 二值化處理………………………………………………………………47
圖(4-6) 動態二值化示意圖………………………………………………………48
圖(4-7) 日間動態二值化閥值，偵測可能陰影特徵………………………………49
圖(4-8) 夜間動態二值化閥值，偵測可能陰影特徵………………………………50
圖(4-9) 水平投影偵測車輛可能位置……………………………………………51
圖(4-10) 日間路面標線所產生的雜訊……………………………………………54
圖(4-11) 日間標示路面標線與車輛的寬度………………………………………54
圖(4-12) 日間車輛寬度與 ………………………………………………………55
圖(4-13) 日間車輛偵測辨識結果…………………………………………………55
圖(4-14) 日間左右車道的車輛辨識區域…………………………………………56
圖(4-15) 日間左右車道陰影區域…………………………………………………57
圖(4-16) 日間三車道車輛辨識……………………………………………………57
圖(4-17) 夜間車輛辨識的概念圖…………………………………………………58
圖(4-18) 夜間車輛尾燈燈光與距離所造成之差異………………………………59
圖(4-19) 車行範圍內的前車尾燈偵測……………………………………………59
圖(4-20) 利用水平投影尋找到車燈位置…………………………………………60
圖(4-21) 各類車種的寬和高………………………………………………………61
圖(4-22) 夜間車輛陰影位置………………………………………………………63
圖(4-23) 單純使用動態二值化……………………………………………………64
圖(4-24) 使用式(4-15)所得到的結果………………………………………………64
圖(4-25) 利用水平投影所找到的車底列 ………………………………………64
圖(4-26) 夜間車輛偵測的動態閥值區域範圍、車輛寬度與 …………………65
圖(4-27) 夜間車輛偵測辨識結果……………………………………………………65
圖(4-28) 夜間路面標線所產生的雜訊……………………………………………67
圖(4-29) 夜間標示路面標線與車輛的寬度………………………………………67
圖(4-30) 光學成相模型……………………………………………………………68
圖(4-31) 標示出前方道路與本車的相對距離……………………………………69
圖(4-32) 估測距離與實際距離比較………………………………………………69

圖(5-1) 以PC為基礎之系統硬體架構圖…………………………………………71
圖(5-2) 以PC為基礎之系統驗證架構……………………………………………72
圖(5-3) 以DSP為基礎之系統硬體架構圖………………………………………73
圖(5-4) 以DSP為基礎之系統驗證架構…………………………………………74
圖(5-5) 系統在PC based上所使用的擷取器……………………………………75
圖(5-6) CCD的構造………………………………………………………………76
圖(5-7) 感光元件電子示意圖………………………………………………………77
圖(5-8) 24Bit Color CCD…………………………………………………………77
圖(5-9) ADSP-BF533系統核心架構圖…………………………………………79
圖(5-10) ADSP-BF533發展板架構圖……………………………………………81
圖(5-11) ADSP-BF533用戶開發板實體圖與各界面接頭………………………81
圖(5-12) Visual DSP++3.1執行圖…………………………………………………86
圖(5-13) 設定DSP輸入影像流程…………………………………………………88

圖(6-1) 錄製資料庫平台………………………………………………………89
圖(6-2) PC驗證之人機介面………………………………………………………91
圖(6-3) DSP驗證平台……………………………………………………………92
圖(6-4) GPIO輸出辨識結果圖……………………………………………………93
圖(6-5) Debug Mode-Image Viewer………………………………………………94
圖(6-6) 各種路面標線的辨識結果………………………………………………96
圖(6-7) 分析路面警告標示對標線偵測系統的影響……………………………97
圖(6-8) 標線辨識系統針對路面標線消失時的處理……………………………98
圖(6-9) 啟動雨刷後，連續擷取的影像…………………………………………99
圖(6-10) 以DSP為基礎，利用動態二值化所找到的可能標線…………………99
圖(6-11) 以DSP為基礎，可能路面標線結果……………………………………100
圖(6-12) 日間車輛辨識結果………………………………………………………102
圖(6-13) 夜間車輛辨識結果………………………………………………………102
圖(6-14) 夜間車輛偵測之DSP架構平台………………………………………103
圖(6-15) 在上圖( 6-14 )中( a )部分的放大結果…………………………………104
圖(6-16) 在上圖( 6-14 )中( b )部分的放大結果…………………………………104









表 目 錄

表(1-1) 論文組織圖…………………………………………………………………5
表(2-1) 常見的車輛主動式警示系統採用的感測元件…………………………9
表(2-2) 影像定位方法的比較……………………………………………………18
表(2-3) 取樣像素以及有效的取樣解析度………………………………………20
表(2-4) 取樣速度與移動距離關係表……………………………………………27
表(4-1) 實地測量各式車輛的寬( Width )和高( Height ) ………………………62
表(5-1) 各國視訊標準……………………………………………………………78
表(5-2) EZ-KIT Lite Evaluation Board Memory Map……………………………82
表(5-3) PPI Connections…………………………………………………………83
表(5-4) Programmable Flag Connections…………………………………………84
表(5-5) Flash A Configuration Registers for port A, B……………………………85
表(5-6) Flash B Configuration Registers for port A, B……………………………85
表(5-7) Flash A Port B Controls…………………………………………………85
表(6-1) 以 PC為基礎環境介紹…………………………………………………90
表(6-2) 以DSP為基礎環境介紹………………………………………………91
表(6-3) GPIO輸出LED示意表………………………………………………94
表(6-4) 車輛偵測可能發生的情況……………………………………………105
表(6-5) 日間車輛偵測系統測試影片的辨識結果……………………………106
表(6-6) 夜間車輛偵測系統測試影片的辨識結果……………………………106
表(6-7) PC-Based車輛偵測系統耗費的辨識時間…………………………108
表(6-8) DSP-Based車輛偵測系統耗費的辨識時間…………………………109

[1] 川端 昭，“最新超音波工程”，日本工業調查會出版，中華民國八十八年。
[2] International Commission on Non-Ionizing Radiation Protection: Guidelines for limiting exposure to time-varying electric, magnetic and electromagnetic fields.Health Physics 74:494-522, 1998.
[3] Chin-Der Wann and Yi-Ming Chen, “Position tracking and velocity estimation for mobile positioning systems,” Wireless Personal Multimedia communications, 2002. The 5th International Symposium on, vol. 1, pp. 310–314, 2002.
[4] J. K. Kearney, X. Yang, and S. Zhang, “Camera calibration using geometric constraints,” Computer Vision and Pattern Recognition, 1989. Proceedings CVPR '89., IEEE Computer Society Conference on, pp.672-679, 1989.
[5] U. Franke, D. Gavrila, S. Gorzig, F. Lindner, F. Paetzold, and C. Wohler, “Autonomous driving goes downtown,” IEEE Intelligent System, vol. 13, pp. 40–48, 1998.
[6] H. Hattori, “Stereo for 2D visual navigation,” Proceedings of the IEEE Intelligent Vehicles Symposium, pp. 31–38, 2000.
[7] M. B. van Leeuwen and F. C. A. Groen, “"Motion estimation with a mobile camera for traffic applications,” Proceedings of the IEEE Intelligent Vehicles Symposium, pp.58–63, 2000.
[8] C. Knoeppel, A. Schanz, and B. Michaelis, “Robust vehicle detection at large distance using low resolution cameras,” Proceedings of the IEEE Intelligent Vehicles Symposium, pp. 267–272, 2000.
[9] S.G. Jeong, C.S. Kim, K.S. Yoon, J.N. Lee, J.I. Bae, and M.H. Lee, "Real-time lane detection for autonomous navigation," Proceedings of 4th IEEE International Conference on Intelligent Transportation Systems, pp. 508-513, Oakland, California, August 2001.
[10] S.Y. Choi and J.M. Lee, "Optimal moving windows for real-time road image processing," Proceedings of 2001 IEEE International Conference on Robotics and Automation, Vol. 2, pp. 1220-1225, Seoul, Korea, May 2001.
[11] W.S. Wijesoma, K.R.S. Kodagoda, A.P. Balasuriya, and E.K. Teoh, "Laser and camera for road edge and mid-line detection," Proceedings of the Second International Workshop on Robot Motion and Control, 2001, pp. 269-274, Bukowy Dworek, Poland, October 2001.
[12] F. Paetzold, U. Franke, and W.V. Seelen, "Lane recognition in urban environment using optimal control theory," Proceedings of the 2000 IEEE Intelligent Vehicles Symposium, pp. 221-226, Detroit, Michigan, October 2000.
[13] J. Goldbeck and B. Huertgen, "Lane detection and tracking by video sensors," Proceedings of the 1999 IEEE/IEEJ/JSAI International Conference on Intelligent Transportation Systems, pp. 74-79, Tokyo, Japan, October 1999.
[14] M. Bertozzi and A. Broggi, "GOLD: a parallel real-time stereo vision system for generic obstacle and lane detection," Proceedings of IEEE Transactions on Image Processing, Vol. 7 Issue: 1, pp. 62-81, January 1998.
[15] M. Bertozzi and A. Broggi, "Real-time lane and obstacle detection on the GOLD system," Proceedings of the 1996 IEEE Intelligent Vehicles Symposium, pp. 213-218, Tokyo, Japan, September 1996.
[16] K. Yamada, T. Ito, and K. Nishioka, "Road lane recognition system for RCAS", Proceedings of the 1996 IEEE Intelligent Vehicles Symposium, pp. 177-182
[17] D. Pomerleau, "RALPH: rapidly adapting lateral position handler," Proceedings of the Intelligent Vehicles 95 Symposium, pp. 506-511, Detroit, Michigan, September 1995.
[18] J. Hancock and C. Thorpe, "ELVIS: Eigenvectors for Land Vehicle Image System," Proceedings of the 1995 International Conference on Intelligent Robots and Systems. 'Human Robot Interaction and Cooperative Robots', Vol. 1, pp. 35-40, Pittsburg, Pensylvania, August 1995.
[19] K. Kluge and S. Lakshmanan, "A deformable-template approach to lane detection," Proceedings of the 1995 IEEE Intelligent Vehicles Symposium, pp. 54-59, Detroit, Michigan, September 1995.
[20] T. Jochem, D. Pomerleau, and C. Thorpe, "Vision Guided Lane Transition," Proceedings of the 1995 IEEE Symposium on Intelligent Vehicles, pp. 30-35, September 1995.
[21] T. Ito and K. Yamada, "Preceding Vehicle and Road Lanes Recognition Methods for RCAS Using Vision System," Proceedings of the 1994 IEEE Symposium on Intelligent Vehicles, pp. 85-90
[22] J. Sparbert, K. Dietmayer, and D. Streller, "Lane detection and street type classification using laser range images," Proceedings of 4th IEEE International Conference on Intelligent Transportation Systems, pp. 454-459, Oakland, California, August 2001.
[23] A. Gern, U. Franke, and P. Levi, "Advanced lane recognition-fusing vision and
radar," Proceedings of the 2000 IEEE Intelligent Vehicles Symposium, pp. 45-51,
Dearborn, MI, USA, October 2000.
[24] F. Marmoiton, F. Collange, and J. P. Derutin, “Location and relative speed estimation of vehicles by monocular vision,” Proceedings of the IEEE Intelligent Vehicles Symposium, pp. 227–232, 2000.
[25] D. DeMenthon and L. S. Davis, “Exact and approximate solutions of the perspective-three-point problem,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 14, pp. 1100-1105, 1992.
[26] Takeo Kato, Yoshiki Ninomiya , and Ichiro Masaki, “Preceding Vehicle Recognition Based on Learning From Sample Images,” IEEE Trans. On Intelligent Transportation System, vol. 3, no. 4, 2002.

[27] W. Kruger, W. Enkelmann, and S. Rossle, “Real-time estimation and tracking of optical flow vectors for obstacle detection,” Proceedings of the IEEE Intelligent Vehicle ’95 Symposium, pp. 304–309, 1995.
[28] S. M. Smith and J. M. Brady, “ASSET-2: Real-time motion segmentation and shape tracking,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 17, pp. 814–820, 1995.
[29] M. Lutzeler and E. D. Dickmanns, “EMS-vision: recognition of intersections on unmarked road networks,” Proceedings of the IEEE Intelligent Vehicles Symposium, pp. 302–307, 2000.
[30] S. Denasi and G. Quaglia, “Grouping of planar structures for object reconstruction,” Proceedings of the IEEE Digital Signal Processing, 13th International Conference on, vol. 2, pp. 527–530.
[31] A. Bensrhair, M. Bertozzi, A. Broggi, P. Miche, S. Mousset, and G. Toulminet, “A Cooperative Approach to Vision-based Vehicle Detection,” Proceedings of the IEEE Conference on Intelligent Transportation Systems, pp: 207 –212, 2001.
[32] Sholin Kyo, Takuya Koga, and Kazuyuki Sakurai, Shin’ichiro Okazaki, “A Robust Vehicle Detecting and tracking System for Wet Weather Conditions using the IMAP-VISION Image Processing Board,” Proceedings of the IEEEIntelligent Transportation System, pp: 423 –428, 1999.
[33] S. Denasi, C. Lanzone, P. Martinese, G. Pettiti, G. Quaglia, and L. Viglione, “Real-time system for road following and obstacle detection,” Proceedings of the SPIE on Machine Vision Applications, Architectures, and Systems Integration III, vol. 23, no. 4, pp. 349–361, 2001.
[34] M. Bertozzi and A. Broggi, “GOLD: A parallel real-time stereo vision system for generic obstacle and lane detection,” IEEE Trans. on Image Processing, vol. 7, pp. 62–81, 1998.
[35] A. Silberschatz and P. Galvin, "Operating System Concepts," 5th, Addison Wesley,
1997, pp 18-19
[36] Koji Yamaguchi, Yasushi Nagayan, and Koji Ueda, Hiroyuki Nemoto, Makoto Nakagawa, “A Method for Identifying Specific Vehicles Using Template Matching,” Proceedings of the IEEE Intelligent Transportation System, pp: 8 –13, 1999.
[37] 張懿，” 即時路面標線、車輛偵測與距離估計”，淡江大學資訊工程研究所， 碩士論文，2002
[38] 孫正泰，”以電腦視覺為基礎之道路上障礙物偵測”，國立台灣大學資訊工程 研究所，碩士論文，2001
[39] 黃渤泓，”道路車輛偵測與相對距離估測系統”，中山大學電機工程研究所， 碩士論文，2004
[40] M. Hariyama, T. Takeuchi, and M. Kameyama, “Reliable stereo matching for highly-safe intelligent vehicles and its VLSI implementation,” Proceedings of the IEEE Intelligent Vehicles Symposium, pp. 128–133, 2000.
[41] Ping-Cheng Hou”The lane recognition and vehicle detection at night for a camera-assisted car on highway”, Robotics and Automation, 2003.Proceedings. ICRA '03. IEEE International Conference on Volume 2, 14-19 Sept. 2003 Page(s):2110 - 2115 vol.2
[42] ADSP-BF533 EZ-KIT Lite Evaluation System Manual. Rev1.0, April 2003
[43] ADSP-BF533 Blackfin Processor Hardware Reference. Preliminary Revision, March 2003
[44] Visual DSP++3.0 C/C++ Compiler and Library Manual for Blackfin DSPs
[45] Visual DSP++ 3.0 Assember and Preprocessor Manual for Blackfin DSPs Rev2.0,April 2002
[46] Blackfin Processor Instruction Set Reference. Rev 2.0, May 2003
[47] VisualDSP++ 3.0 Getting Started Guide for Blackfin Family DSPs. 2002.4
[48] VisualDSP User’s Guide for Blackfin Family DSPs. 2002.4