(18.210.12.229) 您好!臺灣時間:2021/02/26 09:57
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:張為彥
研究生(外文):Wei-Yan Chang
論文名稱:基於模糊演算法之車道偏離預警系統
論文名稱(外文):Lane Departure Warning System Based on Fuzzy Algorithm
指導教授:郭英哲
指導教授(外文):Ying-Che Kuo
學位類別:碩士
校院名稱:國立勤益科技大學
系所名稱:電機工程系
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2011
畢業學年度:99
語文別:中文
論文頁數:92
中文關鍵詞:車道偵測偏離預警影像處理霍夫轉換模糊演算法卡爾曼濾波器
外文關鍵詞:Lane DetectionLane Departure WarningImage ProcessingHough TransformFuzzy AlgorithmsKalman Filter
相關次數:
  • 被引用被引用:3
  • 點閱點閱:321
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:70
  • 收藏至我的研究室書目清單書目收藏:1
根據交通部的統計顯示,台灣每年的交通事故死亡人數高達7,300人左右、導致23萬人受傷、車輛肇事直接的損失即已達千億以上。其中有90%以上的交通事故涉及人為的失誤。因此,為了減少人為疏失所帶來的社會成本的消耗,許多國家相近的研究先進駕駛者輔助系統(Advanced Driver Assistance Systems, ADAS)來減少交通事故的發生。其中以電腦視覺為基礎的獨立式輔助系統已被視為重點的發展方向。且據統計在許多的交通事故原因中,都是因駕駛的車輛未行駛於固定的車道上,進而發生車禍。於此原因本論文運用電腦視覺來偵測車道標線,並提出一個方法來判別駕駛的車輛是否偏離,早先一步的警示駕駛者,以達到減少車禍發生機率。
本論文所進行的研究主要是將單一CCD攝影機架設於車輛擋風玻璃的正中央,向車輛前方拍攝,擷取車輛前方的道路影像,將擷取的影像輸入於本論文所建構的系統中進行處理。此系統主要分為三個部分:第一部份為車道模型的建立,此部分是離線(Off-Line)的運算,其主要的功用在於以參數向量X以及共變異矩陣C_x來描述一個置信區間(confidence interval),此區間為車道標線可能出現的位置,目的在於縮小系統的偵測範圍,加快偵測速度以及減少車道標線的錯誤判斷。
第二部分為車道標線的偵測,以霍夫轉換(Hough transform)在置信區間內的每一區塊偵測車道標線,並以遞迴運算方式使用擴展型卡爾曼濾波器(Extended Kalman Filter)來更新參數向量X以及共變異矩陣C_x的數值,最後以參數向量X與每一區塊內所偵測出來的車道標線資訊使用最小平方法(Least squares)做曲線擬合,找出最佳的左右車道線位置。
最後一部分運用模糊演算法(Fuzzy Algorithm)來做車道偏離的判別,在這裡運用了在第二部分中所獲得的左、右兩條車道標線位置點集合,進一步估算出一個虛擬的車道中線位置,並以此車道中線位置之平均值與其偏角作為模糊隸屬函數,以模糊演算法來判別車輛是否行駛於當前車道上或需要變換車道,並在駕駛者有危險駕駛情況時提供預警。由這三個部分所建構出的車道偏離預警系統,將使得駕駛者在行駛於國道上能減少車禍的發生率。

Statistics reported by Ministry of Transportation and Communications shows that traffic accidents caused up to about 7,300 deaths and some 230 thousand injured every year, and the loss directly caused by vehicle accidents come to more than one hundred billion dollars. Of which more than 90% traffic accidents were caused by human errors. In order to reduce depletion in social cost attributed to human negligence, many countries started to research Advanced Driver Assistance Systems (ADAS) to reduce traffic accidents. Of which independent auxiliary system based on computer vision has been highlighted as a major developmental orientation. Statistics displays that many accidents happened because those drivers failed to have the cars run on certain traffic mankings. In terms of this reason, computer vision was applied in this study to detect lane lines and a method for judging whether the driver is deviating from the traffic lane was proposed in order to warn the driver previously and so reduce incidence of accidents.
The investigation conducted in this paper was that a CCD camera was mounted at center of a car’s windshield and images in front of the car were taken and extracted. The extracted images were then input to the system constructed in this study for processing. The system is mainly composed of three part: the first part is construction of lane model which is an off-line computation and whose major function lies in expressing a confidence interval with parameter vector X and covariance matrix C_x where lane mankings might appear and which aims at reducing detection area for the system to accelerate speed of detection and reduce wrong judgement on traffic lanes.
The second part is detection of traffic lanes where Hough transform was applied to detect traffic lanes in each block within the confidence interval, then recursive computation was conducted and Extended Kalman Filter was adopted to renew the values of the parameter vector X and the covariance matrix C_x, and finally curve fitting was conducted between the parameter vector X and the data of lane lines detected in each block by using least squares in order to find the best location of the left lane line.
In the final part, fuzzy algorithm was employed to judge the car’s deviation. Here the point set of the left and the right lane lines obtained in the second part were utilized to estimate a virtual center line of the traffic lane, and then the mean value and deflection of such center line were used as fuzzy membership function to judge whether the car was driving on the current lane or needed to change lane by using fuzzy algorithm, so the system could warn the driver if he/she was driving in a danger condition. A warning system for car deviation constructed by these three parts will reduce accident incidence for drivers driving on highway.

摘要 I
Abstract III
誌謝 V
章節目錄 VI
圖目錄 VIII
表目錄 XI
第一章 緒論 1
1.1 研究背景 1
1.2 研究動機與目的 3
1.3 系統功能 4
1.4 論文貢獻 6
1.5 論文架構 7
第二章 相關研究 8
2.1 車道偏離演算法的架構 8
2.2 文獻回顧 9
第三章 車道標線偵測 15
3.1 建立車道模型 17
3.1.1 卡爾曼濾波器 19
3.1.2 擴展卡爾曼濾波器 22
3.1.3 建立初始置信區間 24
3.2 影像前處理 27
3.2.1色彩空間轉換 27
3.2.2 影像二值化 28
3.2.3 邊緣提取 30
3.3 車道標線特徵資訊提取 36
3.3.1 霍夫轉換(Hough transform) 38
3.3.2 偵測區塊內之線段特徵 40
3.4 共變異矩陣與參數向量更新 42
3.5車道標線位置標定 44
第四章 模糊演算法判別偏離 47
4.1 模糊控制系統 48
4.2 模糊隸屬函數 49
4.3 模糊化與模糊規則庫 54
4.4 模糊推論 55
4.5解模糊化 56
第五章 實驗結果及分析 61
5.1車道模型間距設定的探討 63
5.2車道標線偵測結果 64
5.3特殊狀況的車道標線偵測結果 66
5.4模糊控制器判別結果 68
5.5結果探討 70
第六章 結論 72
6.1 結論 72
6.2 未來研究方向 74
參考文獻 75

[1] 工研院IEK ITIS計畫,2010年第四季及全年度台灣汽車產業回顧與展望報告,2011年 二月。
[2] 拓墣產業研究所,智慧型車輛引爆車用電子新發展,2008年 十二月。
[3] P. Lindner, E. Ricter, G. Wanielik, K. Takagi, A Isogi, 2009, “Multi-channel lidar processing for lane detection and estimation ”, pp 1-6, IEEE Intelligent Transportation System Conference
[4] J. Sparbert, K. Dietmayer, D. Strelle, 2001, “Lane detection and Street Type Classification using laser range images”, pp. 25-29, IEEE Intelligent Transportation Systems Conference Proceedings.
[5] M. Bertozzi, A. Broggi, A. Fascioli, A. Tibaldi, 2002, “An Evolutionary Approach to Lane Markings Detection in Road Environments”, pp. 627-736, In Atti del 6 convegno dell’Associazione Italiana per I’Intelligenza Artificiale.
[6] D. Pomerleau, 1995, “RALPH: rapidly adapting lateral position handler”, pp.506-511, Proceedings ofthe Intelligent Vehicles 95 Symposium.
[7] T. Ito, K. Yamada, 1994, “Preceding Vehicle and Road Lanes Recognition Methods for RCAS Using Vision System”, pp. 85-90, IEEE Symposium on Intelligent Vehicles.
[8] S.G. Foda, A.K. Dawoud, 2001, “Highway Lane Boundary Determination For Autonomous Navigation”, pp. 698-702, IEEE Pacific Rim Conf. Communications.
[9] S.Y. Choi, J.M. Lee, 2001, “Optimal moving windows for real-time road image processing”, pp. 1220-1225, IEEE International Conference on Robotics and Automation.
[10] S. Kim, J. Park, S.I. Cho, S. Park, K.H. Choi, 2007, “A Framework for the Development of Robust Lane Recognition Systems”, pp. 754 - 759, Intelligent Transportation Systems Conference.
[11] K. Kluge, S. Lakshmanan, 1996, “A Deformable-Template Approach to Lane Detection”, pp. 54-59, Proceedings of the IEEE Intelligent Vehicles Symposium.
[12] R. Chapuis, 2002, “Accurate Road Following and Reconstruction by Computer Vision”, IEEE transactions on intelligent transportation systems.
[13] R.E. Kalman, 1960, “A New Approach to Linear Filtering and Prediction Problems”, pp. 35-45, Transaction of the ASME-Journal of Basic Engineering.
[14] 交通部台灣區高速公路局,道路交通標誌標線號誌設置規則(97.4.14修訂),2008年 4月。
[15] V. Leavers, 1993,“Which Hough Transform?” ,pp. 250-264, CVGIP: Image Understanding.
[16] 王文俊,“認識Fuzzy(第三版)”,全華科技圖書公司,2008年
[17] 孫宗瀛、楊英魁,“Fuzzy控制理論、實作與應用”,全華科技圖書,台北2005年,張欣儀整理2004年。
[18] M. Bertozzi, A. Broggi, 1996, “Real-time lane and obstacle detection on the GOLD system”, pp. 213-218, IEEE Intelligent Vehicles Symposium.
[19] M. Bertozzi, A. Broggi, 1997, “A real-time oriented system for vehicle detection,”, pp.317-325, Journal of Systems Architecture.
[20] M. Bertozzi, A. Broggi, 1998, “GOLD: a parallel real-time stereo vision system for generic obstacle and lane detection”, pp. 62-81, IEEE Trans. on Image Processing.
[21] W.S. Wijesoma, K.R.S. Kodagoda, A.P. Balasuriya, E.K. Teoh, 2001, “Road edge and lane boundary detection using laser and vision”, pp.1440-1445, Proc. IEEE/RSJ Int. Conf. on Intelligent Robots and Systems.
[22] B. Ma, S. Lakshmanan, A.O. Hero, 2000, “Pavement boundary detection via circular shape models”, pp.644-649, Proc. IEEE Intelligent Vehicles Symposium.
[23] A. Gern, U. Franke, P. Levi, 2000, “Advanced lane recognition-fusing vision and radar”, pp.45-51, Proc. IEEE Intelligent Vehicles Symposium.
[24] L.A. Zadeh, 1965, “Fuzzy Sets”, pp. 338-353, Inform. Control.
[25] R. Aufrere, F. Marmotiton, R. Chapuis, F. Collange, J.P. Derutin, 2000, “Road detection and vehicles tracking by vision for an on-board ACC system in the VELAC vehicle”, pp 11-18, IEEE Int’l Conf. on Information Fusion.

連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
無相關期刊
 
系統版面圖檔 系統版面圖檔