臺灣博碩士論文加值系統

本論文提出一個嵌入式駕駛者安全輔助系統。此系統架構是利用一個固定位置的攝影機，透過兩面鏡子反射，取得前方道路影像及駕駛者影像，進行影像分析。透過影像分析取得駕駛者的視線、道路中線的方向以及行車方向。最後再結合前面路況，判斷是否為安全駕駛。
CDriver為駕駛者視線方向與道路線的方向所得之相關係數， CLane為行車方向與道路線的方向所得之相關係數，我們結合這兩個相關係數來區分駕駛情況是否安全。當前方有車輛或行人時，安全判斷標準必須更加嚴格，藉由調整相關係數安全範圍的臨界值，能提早提醒駕駛者注意前方路況，減低駕駛者因注意力不足而造成意外的情況。
整個系統在TI DaVinci平台上執行，此平台為雙核心設計，包含ARM9及TMSC64x+。藉由兩個CPU合作及有效率的演算法，使得處理速度在嵌入式系統也能達到即時的偵測效果。每張影像處理時間為92ms，大約11fps。實驗結果顯示這個系統能正確判斷是否為安全駕駛。

This paper presents an embedded system for the purpose of driving safety assistance. The system utilizes one fixed camera and two mirrors to capture image of road and the driver’s image for image analysis. By image analysis the system extracts driver’s viewing direction, lane direction and driving direction Finally, these orientation information is combined with the road information to judge if it’s safe driving.
More specifically, CDriver is the factor obtained from the driver’s viewing direction and lane direction. CLane is the factor derived from the lane direction and driving direction. We combine the two factors to judge if the driving situation is safe or not.
The proposed system runs on a TI DaVinci platform. This platform is dual-core design, ARM926EJ-S and TMSC64x+. By the cooperation of the two CPUs, we got an average processing time of 92ms about 11fps. The experimental results show that the performance of the proposed system is satisfied for real application.

目錄
指導教授推薦書...............................................i
口試委員會審定書............................................ii
長庚大學博碩士論文著作授權書...............................iii
國家圖書館博碩士論文電子檔案上網授權書......................iv
誌謝.........................................................v
中文摘要....................................................vi
Abstract...................................................vii
目錄......................................................viii
圖目錄.......................................................x
表目錄....................................................xiii
第一章 緒論..............................................1
1.1 研究動機.....................................1
1.2 文獻探討.....................................1
1.3 論文架構.....................................2
第二章 軟硬體架構與系統作業流程..........................4
2.1 硬體架構.....................................4
2.2 軟體架構.....................................6
2.3 系統作業流程................................11
第三章 前方道路影像分析.................................13
3.1 道路線偵測..................................13
3.1.1 影像標籤化..................................16
3.1.2 霍夫轉換....................................16
3.2 前車偵測....................................17
3.2.1 邊緣偵測....................................18
3.2.2 左右邊界搜尋................................19
3.2.3 候選區域確認................................20
3.2.4 AdaBoost演算法..............................21
3.3 行人偵測....................................24
3.3.1 垂直邊緣資訊擷取............................25
3.3.2 候選區域確認................................26
第四章 駕駛者影像分析...................................27
4.1 眼睛及鼻子偵測......................................28
4.2 轉頭角度............................................29
第五章 安全性判斷.......................................32
5.1 彈性形狀脈絡................................32
5.2 相似性量測..................................34
5.3 安全性判斷..................................35
第六章 實驗結果分析.....................................36
6.1 車道線偵測結果..............................36
6.2 前車偵測結果................................38
6.3 行人偵測結果................................40
6.4 駕駛者影像實驗結果..........................42
6.5 駕駛狀況監控實驗結果........................44
第七章 結論與未來展望...................................50
7.1 結論........................................50
7.2 未來研究方向................................50
參考文獻....................................................52



圖目錄
圖1-1 系統流程圖..........................................3
圖2-1 DaVinci硬體平台.....................................5
圖2-2 DM6446處理器架構圖..................................5
圖2-3 取像系統示意圖......................................6
圖2-4 實際系統............................................6
圖2-5 軟體系統............................................7
圖2-6 軟體發展步驟........................................8
圖2-7 VISA create/delete流程說明圖.......................10
圖2-8 ARM-DSP 資料傳遞...................................11
圖2-9 影像處理流程.......................................12
圖3-1 道路線偵測流程圖...................................14
圖3-2 Sobel遮罩..........................................14
圖3-3 道路線偵測結果.....................................15
圖3-4 形態學運算子.......................................15
圖3-5 影像標籤化示意圖...................................16
圖3-6 霍夫直線轉換示意圖.................................17
圖3-7 前車偵測流程圖.....................................18
圖3-8 車尾水平邊緣偵測...................................19
圖3-9 車尾垂直邊緣偵測...................................19
圖3-10 車尾垂直邊緣投影...................................20
圖3-11 車輛偵測結果.......................................21
圖3-12 矩形特徵...........................................22
圖3-13 矩形特徵計算.......................................22
圖3-14 45°斜方向矩形特徵計算..............................23
圖3-15 行人偵測流程圖.....................................24
圖3-16 垂直邊緣資訊擷取...................................25
圖3-17 行人偵測...........................................26
圖4-1 駕駛者視線偵測流程圖...............................27
圖4-2 人臉偵測...........................................28
圖4-3 眼睛偵測...........................................28
圖4-4 眼睛區域二值化結果.................................29
圖4-5 鼻子區域二值化結果.................................29
圖4-6 鼻孔偵測...........................................29
圖4-7 人臉模型...........................................30
圖5-1 形狀脈絡示意圖.....................................32
圖5-2 形狀脈絡計算.......................................33
圖5-3 形狀脈絡直方圖.....................................33
圖5-4 彈性形狀脈絡.......................................33
圖5-5 視線相似度欄位.....................................34
圖5-6 車道線相似度欄位...................................34
圖6-1 直線路段1..........................................36
圖6-2 直線路段2..........................................37
圖6-3 直線路段3..........................................37
圖6-4 直線路段4..........................................37
圖6-5 左轉彎影像之行車線與道路方向.......................38
圖6-6 右轉彎影像之行車線與道路方向.......................38
圖6-7 前車測試影像.......................................39
圖6-8 前車影像測試結果...................................40
圖6-9 行人測試影像.......................................40
圖6-9(續) 行人測試影像......................................41
圖6-10 行人影像測試結果...................................41
圖6-10(續) 行人影像測試結果..........................42
圖6-11 人臉五官偵測結果...................................43
圖6-12 區間閥值定義.......................................44
圖6-13 駕駛狀態偵測結果(1)................................44
圖6-13(續) 駕駛狀態偵測結果(1)..............................45
圖6-14 駕駛狀態偵測結果(2)................................45
圖6-15 駕駛狀態偵測結果(3)................................46


表目錄
表4-1 轉頭角度區域分類...................................31
表5-1 駕駛狀態範圍.......................................35
表5-2 駕駛狀態範圍修正...................................35
表6-1 駕駛者影像處理時間.................................47
表6-2 車道線偵測處理時間.................................47
表6-3 前車偵測處理時間...................................48
表6-4 行人偵測處理時間...................................48
表6-5 系統執行時間.......................................48
表6-6 人臉與五官偵測率...................................49
表6-7 前車偵測率.........................................49
表6-8 行人偵測率.........................................49

[1] Kuo-Yu Chiu, Sheng-Fuu Lin, “Lane detection using color-based segmentation”, Proceedings of the 2005 IEEE Conference on Intelligent Vehicles Symposium, Vol.6-8, pp.706-711, 2005.
[2] Gonzalez J.P., Ozguner U., “Lane detection using histogram-based segmentation and decision trees”, Proceedings of Intelligent Transportation Systems Conference, Vol. 1-3, pp. 346 – 351, 2000.
[3] Seung Gweon Jeong, Chang Sup Kim, Dong Youp Lee, Sung Ki Ha, Dong Hwal Lee, Man Hyung Lee, and Hideki Hashimoto, “Real-Time Lane Detection for Autonomous Vehicle”, Proceedings of the 2001 IEEE Conference on International Symposium on Industrial Electronics (ISIE 2001), pp. 1466-1471, Pusan, Korea, 2001.
[4] T. Liu, N. Zheng, H. Cheng, and Z. Xing, “A novel approach of road recognition based on deformable template and genetic algorithm”, Proceedings of the 2003 IEEE Conference on Intelligent Transportation Systems, vol. 2, pp. 1251-1256, 2003.
[5] R. Chapuis, R. Aufrere, and F. Chausse, “Accurate road following and reconstruction by computer vision”, Proceedings of the 2002 IEEE Conference on Intelligent Transportation Systems, vol. 3, Issue. 4, pp. 261-270, 2002.
[6] Juan M. Collado, Cristina Hilario, Arturo de la Escalera, and Jonse M. Armingol, “Detection and classification of road lanes with a frequency analysis”, Proceedings of the 2003 IEEE International Conference on Intelligent Transportation Systems, volume 2, pp.1251-1256, 2003.
[7] David Schreiber ,Bram Alefs ,Markus Clabian, “Single camera lane detection and tracking”, Proceedings of the 2005 IEEE Conference on Intelligent Transportation Systems, ITSC 05, Vienna, Austria, pp. 1114-1119, 2005
[8] Massimo Bertozzi, Alberto Broggi, and Alessandra Fascioli, “Stereo inverse perspective mapping: theory and applications”, Image and Vision Computing Journal, vol.16, pp. 585-590, 1998.
[9] S.-S. Huang, C.-J. Chen, P.-Y. Hsiao, and L.-C. Fu, “On-Board Vision System for lane Recognition and Front-Vehicle Detection to Enhance Driver's Awareness”, IEEE International Conference on Robotics and Automation, pp. 2456-2461, 2004.
[10] B. Heisele and W. Ritter, “Obstacle detection based on color blob flow”, Proceedings of The Intelligent Vehicles Symposium 1995, pp. 282-286, 1995.
[11] H. Kruppa, M. Castrillon Santana and B. Schiele, “Fast and robust face finding via local context”, Joint IEEE International Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance (VS-PETS), pp. 157-164, 2003.
[12] Y.-J. Wu, et al, “Vision-based Driving Environment Identification for Autonomous Highway Vehicles”, IEEE Conf on Networking, Sensing and Control, vol. 2, pp. 1323-1328, 2004.
[13] Paul Viola and Michael Jones, “Rapid object detection using a boosted cascade of simple features”, Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR), vol. 1, pp. 511-518, 2001.
[14] Y. Fan and M. Paindavoine, “Implementation of an RBF Neural Network on Embedded Systems: Real-Time Face Tracking and Identity Verification”, Proceedings of IEEE Transactions On Neural Networks, Vol. 14, No. 5, pp. 1162- 1175, September 2003.
[15] B. Y. Chen, G. D. Su, and Y.F. Deng, “2-Channel Parallel Face Detection Board Using Two TMS320C6201 DSPS”, Proceedings of International Conference on Machine Learning and Cybernetics, vol. 7, pp. 4446-4451, 2005.
[16] Ming-Hsuan Yang, David J. Kriegman and Narendra Ahuja, “Detecting Faces in Images: A Survey”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 24, No. 1, pp. 34-58, January 2002.
[17] J.-D. Lee, J.-D. Li, and L.-C. Liu, “A Novel Drive Status Monitoring System”, IEEE International Conference on Systems, Man and Cybernetics, pp. 4643-4647, 2006.
[18] 李建德，〈駕駛狀況的監控系統〉，長庚大學電機工程研究所，碩士論文，民國95年。
[19] S. Belongie, J. Malike, and J. Puzicha, “Shape Matching and Object Recognition Using Shape Contexts”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 24, pp. 509-522, 2002.
[20] David Liu, and Tsuhan Chen, “Soft Shape Context for Iterative Closest Point Registration”, Proceedings of IEEE Conference on The International Conference on Image Processing, pp. 1081-1084, 2004.