為提升行車安全，目前之車輛紛紛引入各種駕駛輔助系統，以協助車輛駕駛注意行車時之各種突發事件。本研究提出一個基於車用影像之目標物碰撞預警系統，透過雙鏡頭行車視訊紀錄器之動態影像內容，建構出一套車輛前方防撞預警系統。首先我們使用AdaBoost分類器偵測車用影像中含有目標物之區域，並透過樣板比對取得目標物在左影像和右影像之間的像差值，再以立體視覺原理計算出目標物與車輛的相對位置。為了能監控車輛前方的目標物，並即時判斷出發生碰撞的危險程度，預警系統會持續追蹤目標物在動態影像的移動資訊，一旦偵測到目標進入了車輛正前方所預設的危險區域內，預警模組會依據目標物的深度距離和接近速度來判定其危險程度，最後系統依照危險程度提出不同等級的預警通知。本研究設計之碰撞預警系統可作為駕駛人的第三隻眼，輔助駕駛人在車輛行駛當中可以避免掉因意外所造成的車禍損害。

To improve the safety of driving, variant driver assistance systems have been introduced to prevent the car driver from emergent accidents. In this study a real-time pedestrian collision warning system based on vehicle video image has been proposed. The proposed system was implemented via the captured dynamic images of a binocular vehicle video recorder installed in the car. Firstly, the AdaBoost classifier is used to detect the interested targets in the captured vehicle video images, and then the disparity of the targets between the left image and right image is conducted through template matching method. The relative distance of the targets and the vehicle are estimated based on the conducted disparity. Different dangerous regions are pre-defined in front of the vehicle according to the relative distance between the targets and the car. Different alarming information is issued according to the dangerous region the target located in. The designed system can process 5 fps. The results show that the designed anti-collision warning system can operate well under real-time driving test data.

[1] D. Geronimo, et al. "Survey of pedestrian detection for advanced driver assistance systems," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 32, no. 7, pp. 1239-1258, 2010.
[2] D. M. Gavrila, "Pedestrian detection from a moving vehicle," in European Conference on Computer Vision, 2000.
[3] D. M. Gavrila, J. Giebel, and S. Munder, "Vision-based pedestrian detection: The protector system," in IEEE International Conference on Intelligent Vehicles Symposium, 2004.
[4] 范俊海(2004年)。動態車輛周圍資訊偵測系統之研究。行政院國家科學委員會專題研究計畫成果報告(民93年10月31日).
[5] N. Dalal, B. Triggs, "Histograms of oriented gradients for human detection," in IEEE International Conference on. Computer Vision and Pattern Recognition, vol. 1, pp. 886-893, 2005.
[6] X. Wang, T. X. Han, S. Yan, "An HOG-LBP human detector with Partial occlusion handling, " in IEEE International Conference on. Computer Vision, pp.32-39, 2009.
[7] P. Viola, M. Jones, "Robust Real-time Object Detection, " ICCV 2nd Internaltion Workshop on Statistical and Computational Theories of Vision: Modeling, Learning, Computing and Sampling, Jul. 2001.
[8] Y. Freund, R.E. Schapire, "A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting," Journal of Computer and System Sciences, vol. 55, pp. 119-139, 1997.
[9] Q. Zhu, M. C. Yeh and K. T. Cheng, "Fast human detection using a cascade of histograms of oriented gradients." In IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2006.
[10] X. B. Cao, Q. Hong and J. Keane, "A low-cost pedestrian-detection system with a single optical camera." IEEE Transactions on Intelligent Transportation Systems, vol. 9, no.1, pp. 58-67, 2008
[11] P. Cerri, et al. "Day and night pedestrian detection using cascade AdaBoost system." In 13th International IEEE Conference on Intelligent Transportation Systems (ITSC), 2010.
[12] 方國意, "基於特徵自動定位之人臉表情辨識系統之實現," 國立成功大學工程科學系碩士論文, Jul. 2009.
[13] 楊棠鈞, "結合Adaboost分類器和支援向量機的路標辨識系統之實現," 國立成功大學工程科學系碩士論文, Jul. 2009.
[14] Z. Zhang, "A Flexible New Technique For Camera Calibration, " IEEE Trans. on Pattern Analysis and Machine Intelligence, vol.22, no.11, pp.1330-1334, 2000.
[15] Computational Vision at Caltech, Camera Calibration Toolbox for Matlab, Available: http://www.vision.caltech.edu/bouguetj/calib_doc/
[16] National Institute for Research in Computer Science and Control(n.d.), INRIA Person Dataset, Available: http://pascal.inrialpes.fr/data/human/
[17] G. Woods 原著, 繆紹綱譯, 數位影像處理 Digital image processing, 3rd Ed. 普林斯頓國際有限公司, May 2009.
[18] R. E. Kalman, "A new approach to linear filtering and prediction problems," Journal of basic Engineering, vol. 82, no. 1, pp. 35-45, 1960.
[19] G. Welch, G. Bishop, "An Introduction to the Kalman Filter", September 17, 1997.
[20] N. Peterfreund, "Robust tracking of position and velocity with Kalman snakes." In IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 21, no. 6, pp. 564-569, 1999.
[21] G. Welch, et al. "High-performance wide-area optical tracking: The hiball tracking system."Presence: Teleoperators and Virtual Environments vol. 10, no. 1, 2001.
[22] 財團法人車輛研究測試中心(2009年)。安全車距的拿捏與應用情境。網址：http://www.artc.org.tw/chinese/03_service/03_02detail.aspx?pid=839
[23] 交通部運輸研究所(2004年)。我國汽車煞車距離與行車速度關係之測試與研究。網址：http://www.iot.gov.tw/ct.asp?xItem=5028&ctNode=622
[24] C. G. Keller, et al. "The benefits of dense stereo for pedestrian detection," In IEEE Transactions on Intelligent Transportation Systems, vol. 12, no. 4, pp. 1096-1106, 2011.