臺灣博碩士論文加值系統

近年來，超速駕駛為導致交通意外的主要原因之一，相信若能夠在車輛尚未超速前就警告駕駛人，即可有效降低事故的發生率。本研究以隱藏式馬可夫模型(Hidden Markov Model，HMM)作為預測車速之方法，以駕駛人過去行駛的車速資料作為模型訓練資料，並將訓練好的HMM模型建立車速預測模組，分析即時車速資料序列以預測未來之車速，若預測車速會超過路段限速，則以警訊提醒駕駛人。本研究背景設定在車輛上無配備任何可擷取車速設備的前提下，建立車速估算模組，利用已安裝在車輛上的行車記錄器之影像來估算出當時之車輛的車速，並送至車速預測模組中以進行下一個時間點之車速的預測。本研究所提出兩種影像測速法，一種是參考路面上的車道分隔線之線段在連續影像中的像素移動量來估算車速，另一種則是使用立體視覺計算方式以求得道路兩旁景象之景深，再以連續影像中的景象之景深變化來估算車速。為了驗證系統預測車速之準確率，本研究利用OBD2(On-Board Diagnostics Phase 2)外接設備以擷取車輛之實際車速，並用以和本研究所提出之方法比較，結果顯示在高速公路與市區道路的平均預測誤差值分別約0.6 km/h和1.2 km/h。最後，利用預測之車速作為超速警示之依據，以達到預警之目的。

In recent years, vehicle over-speeding is one of the major sources of car accidents. The accident rate can be effectively reduced if we can warn the drivers before over-speeding occurred. In this report, under the premise of without vehicle’s speed measurement devices, two methods are proposed to compute the vehicle’s speed from the images provided by digital video recorder in the vehicle, and use the Hidden Markov Model(HMM) is used as a predictor by analyzing real-time speed data series to predict vehicle’s future speed. The HMM is trained from the recorded driving speed data of the driver in the past, then the trained HMM is used to predict the vehicle speed in the next time instance from the contiguous collected speed of the driving. With the predicted data, a speed warning system can be established to inform the driver that the vehicle will exceed the speed limit. Since the HMM predicts the vehicle speed variance by using the real-time vehicle speed, this research proposes two methods to compute the vehicle speed. One is based on the lane separating segments in the image to estimate the movement of vehicle speed; the other uses stereo vision to estimate depth difference of environment to compute the vehicle speed. To evaluate correctness of the proposed methods, the estimated speed is compared with the real speed which grasped via an OBD2 system. The experimental results show that the proposed method exhibits good performance. The mean absolute errors of estimated speed are 0.6 km/h and 1.2 km/h in highway and urban road respectively. Finally, based on the vehicle’s future speed, the speed warning system can be setup.

摘要 ii
Abstract iii
誌謝 iv
目錄 v
表目錄 vii
圖目錄 viii
第1章 緒論 1
1.1 研究動機與目的 1
1.2 系統架構 4
1.3 論文架構 6
第2章 相關背景之研究 7
2.1 相機校正(Camera Calibration) 7
2.2 鳥瞰轉換(Top View Image Transformation) 12
2.3 影像測速 13
2.4 隱藏式馬可夫模型之相關研究 14
2.5 相關的影像處理技術 16
2.5.1 形態學運算(Morphology) 16
2.5.2 連通標記(Connected-Component Labeling) 18
2.5.3 雙線性內插法(Bilinear Interpolation) 20
第3章 車速預警系統 22
3.1 基於單一行車記錄器之車道線段測速法 23
3.1.1 平面投影轉換法(Homography Mapping) 24
3.1.2 車道線段偵測及追蹤 27
3.1.3 實測像素點的距離比例 31
3.2 基於雙行車記錄器之立體視覺測速法 33
3.2.1 影像深度 34
3.2.2 立體視覺測速法之流程 35
3.2.3 參考區塊的比對及追蹤 38
3.2.4 像差景深對應表之實測 43
3.3 測速法切換模型 47
3.4 隱藏式馬可夫模型(Hidden Markov Model，HMM) 50
3.4.1 隱藏式馬可夫模型的參數 51
3.4.2 隱藏式馬可夫模型的估算程序 53
3.4.3 隱藏式馬可夫模型的訓練程序 57
3.4.4 利用隱藏式馬可夫模型建立車速預測模組 61
第4章 實驗結果與討論 72
4.1 實驗設備及環境架設 72
4.2 實驗結果 76
4.2.1 高速公路測速結果 76
4.2.2 市區道路測速結果 78
4.2.3 測速法切換模型之實驗結果 83
4.2.4 車速預測模組之實驗結果 84
4.2.5 車速預警系統之實驗結果 88
第5章 結論與未來研究方向 90
5.1 結論 90
5.2 未來研究方向 91
參考文獻 92

[1]交通部統計查詢網, http://stat.motc.gov.tw/mocdb/stmain.jsp?sys=100, 2013
[2]Cheng, W., Han, C., ＆ Xi, J. (2010, May). Mechanism Analysis of Over-speeding Impact on the Road Safety, In Intelligent Computation Technology and Automation (ICICTA), 2010 International Conference on (Vol. 2, pp. 652-656). IEEE.
[3]內政部統計處, http://www.moi.gov.tw/stat/news_content.aspx?sn=6885, 2013
[4]Chou, H. L., ＆ Tsai, W. H. (1986). A new approach to robot location by house corners, Pattern Recognition, 19(6), 439-451.
[5]Chen, C. S., Yu, C. K., ＆ Hung, Y. P. (1999). New calibration-free approach for augmented reality based on parameterized cuboid structure, In Computer Vision, 1999. The Proceedings of the Seventh IEEE International Conference on (Vol. 1, pp. 30-37). IEEE.
[6]Tsai, R. (1987). A versatile camera calibration technique for high-accuracy 3D machine vision metrology using off-the-shelf TV cameras and lenses, Robotics and Automation, IEEE Journal of, 3(4), 323-344.
[7]Zhang, Z. (2000). A flexible new technique for camera calibration, Pattern Analysis and Machine Intelligence, IEEE Transactions on, 22(11), 1330-1334.
[8]Zhang, Z. (2004). Camera calibration with one-dimensional objects. Pattern Analysis and Machine Intelligence, IEEE Transactions on, 26(7), 892-899.
[9]Maybank, S. J., ＆ Faugeras, O. D. (1992). A theory of self-calibration of a moving camera, International Journal of Computer Vision, 8(2), 123-151.
[10]Bertozzi, M., ＆ Broggi, A. (1998). GOLD: A parallel real-time stereo vision system for generic obstacle and lane detection, Image Processing, IEEE Transactions on, 7(1), 62-81.
[11]Bertozzi, M., Broggi, A., Medici, P., Porta, P. P., ＆ Sjogren, A. (2006, June). Stereo vision-based start-inhibit for heavy goods vehicles, In Intelligent Vehicles Symposium, 2006 IEEE (pp. 350-355). IEEE.
[12]Jung, H. G., Kim, D. S., Yoon, P. J., ＆ Kim, J. (2006, June). Parking slot markings recognition for automatic parking assist system, In Intelligent Vehicles Symposium, 2006 IEEE (pp. 106-113). IEEE.
[13]He, X. C., ＆ Yung, N. H. (2007, February). A novel algorithm for estimating vehicle speed from two consecutive images, In Applications of Computer Vision, 2007. WACV'07. IEEE Workshop on (pp. 12-12). IEEE.
[14]Hassan, M. R., ＆ Nath, B. (2005, September). Stock market forecasting using hidden Markov model: a new approach, In Intelligent Systems Design and Applications, 2005. ISDA'05. Proceedings. 5th International Conference on (pp. 192-196). IEEE.
[15]余文利, 廖建平, and 馬文龍. 一種新的基於隱藏式馬可夫模型的股票價格時間序列預測方法, 計算機應用與軟體 27.6 (2010): 186-190.
[16]Jung, H. G., Lee, Y. H., Kim, D. S., ＆ Yoon, P. J. (2005). Stereo vision based advanced driver assistance system, In Proceedings of AVEC International Workshop (Vol. 19, pp. 97-104).
[17]Laganiere, R. (2000). Compositing a bird's eye view mosaic, In Vision Interface, 10(3), 382-386
[18]謝明逢, 利用雙攝影機取像模組建構一大型環境監控系統, 國立中央大學資訊工程研究所碩士論文, 2004
[19]Otsu, N. (1975). A threshold selection method from gray-level histograms. Automatica, 11(285-296), 23-27.
[20]Toulminet, G., Bertozzi, M., Mousset, S., Bensrhair, A., ＆ Broggi, A. (2006). Vehicle detection by means of stereo vision-based obstacles features extraction and monocular pattern analysis. Image Processing, IEEE Transactions on, 15(8), 2364-2375.
[21]Lsqcurvefit函數說明, http://www.mathworks.com/help/optim/ug/lsqcurvefit.html, 2013
[22]Baum, L. E., ＆ Petrie, T. (1966). Statistical inference for probabilistic functions of finite state Markov chains, The annals of mathematical statistics, 37(6), 1554-1563.
[23]Juang, B. H., ＆ Rabiner, L. R. (1991). Hidden Markov models for speech recognition, Technometrics, 33(3), 251-272.
[24]Vinciarelli, A., ＆ Luettin, J. (2000, September). Off-line cursive script recognition based on continuous density HMM, In Proc. 7th International Workshop on Frontiers in Handwriting Recognition, Amsterdam.
[25]Baum, L. E., Petrie, T., Soules, G., ＆ Weiss, N. (1970). A maximization technique occurring in the statistical analysis of probabilistic functions of Markov chains, The annals of mathematical statistics, 41(1), 164-171.
[26]Welch, L. R. (2003). Hidden markov models and the baum-welch algorithm, IEEE Information Theory Society Newsletter, 53(4), 1-14.
[27]Shu, H., Deng, L., He, P., ＆ Liang, Y. (2012). Speed Prediction of Parallel Hybrid Electric Vehicles Based on Fuzzy Theory, International Conference on Power and Energy Systems (Vol. 13, pp. 493-498)
[28]車上診斷系統, http://law.epa.gov.tw/search.php?type=att＆sn=439036267＆num=3, 2013
[29]ARTC 財團法人車輛研究測試中心-安全車距的拿捏與應用情境-防禦駕駛哲學 PART 3, http://www.artc.org.tw/chinese/03_service/03_02detail.aspx?pid=839, 2013
[30]林咏漢, 交通標誌辨識系統在PAC Duo嵌入式系統之實現, 國立成功大學工程科學研究所碩士論文, 2011