近年來，在輔助駕駛系統中，行人偵測是重要的研究領域之一。本系統利用裝置於後照鏡後方之行車記錄器，判斷拍攝畫面中是否出現行人。我們所提出的行人偵測系統可分為三部分：影像前處理、行人偵測以及警示系統。在影像前處理中，首先將行車記錄器擷取之影像設定感興趣區域，縮小偵測範圍，將連續兩張影像做相減得到其差值影像，並將其轉換為二值化差值影像。在行人偵測中，我們利用一組模板影像，對目前所擷取影像之邊緣偵測圖以及二值化差值影像，進行模板比對，找出高度相似於行人之候選區域，並分析候選區域之直方圖成分，去除不含行人成份之區域。最後使用梯度方向直方圖計算候選區域之描述子，將其結果利用LIBSVM判斷此區域之描述子是否高度相似於正樣本，並利用追蹤演算法更新行人移動後的位置。在警示系統中，我們使用光流法估測車速，並使用一公式評估影像中行人車之真實距離，再根據評估的車速與人車距離，判斷此行人是否位於碰撞危險範圍內；若人車距離小於應有之安全距離時，本系統將框選影像中行人位置，並發出警示音提醒駕駛。最後我們於八個不同之實際道路下測試，實驗數據顯示本系統能精確地判斷行人是否出現及其所在位置。

Pedestrian detection is an important research field in advanced driver assistance system (ADAS). We mounted a vehicle video recorder behind the rear-view mirror to capture the scene for detecting pedestrians. The proposed system is composed of three components. In the first component, we determine the region of interest (ROI) from the captured image and transform the difference image between the previous and the current frames into a binary image. In the second component, the template matching is performed on the edge image of the current frame and the binary difference image to coarsely detect candidate pedestrians by using a set of template images. The histogram analysis again roughly filters out the candidate pedestrians. Histogram of Oriented Gradients (HOG) combined with library support vector machine (LIBSVM) is used to verify those candidate pedestrians. If we find a pedestrian, the tracking algorithm would determine the new location of the pedestrian. In the final component, we use the optical flow to estimate the automobile speed, and we calculate the distance from the pedestrian on the image with a known model. The system will give the driver visual and audio notifications if the distance from the pedestrian is smaller than the safety distance. In the experiment, the proposed system is evaluated with eight videos. The experimental results show that the proposed system has high accuracy of the pedestrian detection.

[1] WHO Regional Office for the Western Pacific, “Road Safety in the Western Pacific Region: Call for Action,” 2009.

[2] U.S. Census Bureau, “The 2012 Statistical Abstract: Section 23 Transportation,” Oct. 2011.

[3] Ministry of Transportation and Communications, R.O.C., “The Statistics of Accident Events in 1998~2012,” 2012.

[4] K.C. Fuerstenberg, K.C.J. Dietmayer, and V. Willhoeft, “Pedestrian Recognition in Urban Traffic Using A Vehicle based Multilayer Laserscanner,” Intelligent Vehicle Symposium, 2002. IEEE , vol.1, pp. 31- 35 vol.1, 17-21 June 2002

[5] G. Gate and F. Nashashibi, “Fast algorithm for pedestrian and group of pedestrians detection using a laser scanner,” Proc. 2009 IEEE Intelligent Vehicles Symposium, pp. 1322-1327, 2009.

[6] K. Fuerstenberg, J. Hipp, and A.Liebram, “A Laser Scanner for Detailed Traffic Data Collection and Traffic Control,” Proc. ITS 2000, 7th World Congress on Intelligent Transport Systems, ITS 2000 Turin, Paper 2335.

[7] S. Sato, M. Hashimoto, M. Takita, K. Takagi, and T. Ogawa, “Multilayer Lidar-Based Pedestrian Tracking in Urban Environments,” IEEE Intelligent Vehicles Symposium, pp. 849-854, 2010.

[8] L. S. Jin, Q. N. Niu, H. J. Hou, S. X. Hu, and F. R. Wang, “Study on vehicle front pedestrian detection based on 3D laser scanner,” Proc. International Conference on Transportation, Mechanical, and Electrical Engineering (TMEE '11), pp. 735-738, Changchun, China, December 2011.

[9] F. Liu, C. Sun, and B. Li, “A New Weighted Way of Linear Discriminant Function,” Computer Engineering, pp. 17-18, 2004.

[10] S.-B. Marco, C. Marco, and C. Marco, “Recursive Segmentation Based on Higher Order Statistics in Thermal Image Pedestrian Detection,” IEEE Communications Control and Signal Processing, pp. 1-4, 2012.

[11] Q. Liu, J. Zhuang, and S. Kong, “Detection of Pedestrians at Night Time Using Learning-based Method and Head Validation,” IEEE Imaging Systems and Techniques, pp. 398-402, 2012.

[12] M. Bertozzi, A. Broggi, P. Grisleri T. Graf, and M. Meinecke, “Pedestrian Detection in Infrared Images,” IEEE Intelligent Vehicles Symposium, pp. 662-667, 2003.

[13] D. Geronimo, A.M. Lopez, A.D. Sappa, and T. Graf, “Survey of pedestrian detection for advanced driver assistance systems,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol.32, pp. 1239-1258, 2010.

[14] L. Zhao and C. Thorpe, “Stereo and Neural Network-Based Pedestrian Detection,” IEEE Trans. Intelligent Transportation Systems, vol. 1, no. 3, pp. 148-154, Sept. 2000.

[15] S.J. Krotosky and M.M. Trivedi, “On Color-, Infrared-, and Multimodal-Stereo Approaches to Pedestrian Detection,” Intelligent Transportation Systems, IEEE Transactions on , vol.8, no.4, pp. 619-629, Dec. 2007.

[16] S. Nedevschi, R. Danescu, C. Pocol, and M. M. Meinecke, “Stereo Image Processing for ADAS and Pre-Crash Systems”, International Workshop on Intelligent Transportation, pp. 55-60, 2008.

[17] A. R. Tilley and H. D Associates, “The Measure of Man and Woman: Human Factors in Design,” Wiley, 1993.

[18] D. E. Rumelhart, G. E. Hinton, and R. J. Williams, “Learning Internal Representations by Error Propagation,” in Parallel Distribution Processing: Explorations in the Microstructure of Cognition. Cambridge, MA: MIT Press, vol. I, Foundations, pp. 318-362, 1986.

[19] M. Enzweiler and D. M. Gavrila, “Monocular Pedestrian Detection: Survey and Experiments,” IEEE Transactions on Pattern Analysis and Machine Intelligence 31(12) pp. 2179-2195.

[20] L. Bombini, P. Cerri, P. Grisleri, S. Scaffardi, and P. Zani, “An Evaluation of Monocular Image Stabilization Algorithms for Automotive Applications,” IEEE Int’l Conf. Intelligent Transportation Systems, pp. 1562-1567, 2006.

[21] D. Gavrila, J. Giebel, and S. Munder, “Vision-Based Pedestrian Detection: The PROTECTOR System,” IEEE Intelligent Vehicles Symp., pp. 13-18, 2004.

[22] D. Gavrila, “Pedestrian Detection from a Moving Vehicle,” Proc. European Conf. Computer Vision, vol. 2, pp. 37-49, 2000.

[23] D. Gavrila and S. Munder, “Multi-Cue Pedestrian Detection and Tracking from a Moving Vehicle,” Int’l J. Computer Vision, vol. 73, no. 1, pp. 41-59, 2007.

[24] P. Viola, M. J. Jones, and D. Snow. Detecting pedestrians using patterns of motion and appearance. The 9th ICCV, Nice, France, vol. 1, pp. 734-741, 2003.

[25] N. Dalal and B. Triggs, “Histograms of oriented gradients for human detection,” IEEE Computer Vision and Pattern Recognition, vol. 2, pp. 886-893, Jun. 2005.

[26] D. G. Lowe. “Distinctive Image Features from Scale-Invariant Keypoints,” IJCV, 60(2):91-110, 2004.

[27] J. Maller, “FXScript Reference: RGB and YUV Color,” 2004, [Online] http://www.fxscriptreference.org/rgb-yuv-color.html

[28] R. Brunelli, “Template Matching Techniques in Computer Vision: Theory and Practice,” Wiley, 2009.

[29] J. Canny, “A Computational Approach to Edge Detection,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. Pami-8, no. 6, Nov. 1986.

[30] C. F. Gauss, “Theoria Motus Corporum Coelestium. In Sectionibus Conicis Solem Ambientium,” 1809.

[31] D. Coomans and D.L. Massart, “Alternative K-Nearest Neighbour Rules in Supervised Pattern Recognition : Part 1. K-Nearest Neighbour Classification by Using Alternative Voting Rules,” Analytica Chimica Acta, vol. 136, pp. 15-27, 1982.
[32] J. J. Hopfield, “Neural Networks and Physical Systems with Emergent Collective Computational Abilities,” Proc. Natl. Acad. Sci. U.S.A., vol. 79, pp. 2554 -2558, 1982.

[33] J.R. Quinlan, “Induction of Decision Trees,” Machine Learning, vol. 1, pp. 81-106, 1986.

[34] C. Cortes, and V. Vapnik, “Support-Vector Network,” Machine Learning, pp. 273-297, 1995.

[35] C.C. Chang and C.J Lin, “LIBSVM: a library for support vector machines,” ACM Transactions on Intelligent Systems and Technology, vol. 2, 27:1--27:27, 2011.

[36] J. G. Allen, R. Y. D. Xu and J. S. Jin, “Object Tracking Using Camshift Algorithm and Multiple Quantized Feature Spaces,” Pan-Sydney area workshop on Visual information processing, ACM International Conference Proceeding Series vol. 100, 2004.

[37] G.H. Joblove and D. Green, “Color Spaces for Computer Graphics,” Computer Graphics, pp. 20-25, 1978.

[38] J. J. Gibson, “The Perception of the Visual World,” Houghton Mifflin, Boston, Massachusetts, 1950.

[39] J. Shi, C. Tomasi, “Good Feature to Track,” Computer Vision and Pattern Recognition, 1994.

[40] B. D. Lucas and T. Kanade, “An Iterative Image Registration Technique with An Application to Stereo Vision,” International Joint Conferences on Artificail intellignece, 1981.

[41] J. Y. Lin, J. R. Huang, and C.P. Fung, “Using Stopping Distance to Estimate Vehicle Speed for Accident Arbitration,” Publication Abstracts of Research Projects Institute of Transportation Ministry of Transportation and Communications, 2002.

[42] L. Livne and D. Shinar, “Effects of Uncertainty, Transmission Type, Driver Age and Gender on Brake Reaction and Movement Time,” Journal of Safety Research, 2002.
[43] C. Papageorgiou and T. Poggio, “A Trainable System for Object Detection,” journal of Computer Vision, vol. 38, no. 1, pp. 15-33, 2000.

[44] D. Geroimo, A. Sappa, A. Lopez, and D. Ponsa, “Adaptive Image Sampling and Windows Classification for On-Board Pedestrian Detection,” International Conference on Computer Vision Systems, 2007.