臺灣博碩士論文加值系統

事故傷害一直居於國人十大死因之中，其中又以機動車交通事故為主要的原因。因此要如何降低交通事故的發生，儼然成為目前政府與產學界所努力的目標，而在智慧型車輛上加裝先進駕駛者輔助系統(Advanced Driver Assistance Systems, ADAS)為現階段發展較成熟的作法。
本論文研究目的為避免車輛與行人發生碰撞，作法如下，首先以影像式方法搜尋輸入影像中所存在的行人並獲取位置，接著判別該行人移動速度與方向，隨後預測該行人未來移動向量，判斷有無與車輛發生碰撞的可能，如果有則對駕駛發出警示。
根據上述本論文提出的系統分為三個部分，第一部分為使用行人偵測技術辨識行人所在位置，藉此縮減後續計算範圍，第二部分為使用光流法獲取行人之移動向量，最後將求得之移動向量使用灰色理論(Grey theory)預測該行人後續行進路徑，建立預測碰撞點、脫離危險範圍位置，計算可能發生碰撞的時間，判斷是否可能發生碰撞。
本論文將應用於各種交通場景中，使用單一攝影機在車輛行駛時獲取前方的影像，在檢測到行人時將行人框取，同時對有移動的行人進行追蹤，獲取該行人的移動向量，判斷該行人在未來是否可能與車輛發生碰撞，以利提醒駕駛，避免駕駛因疲倦、疏忽或不當行駛而造成意外事故。此系統是以加強車輛行駛安全和減少交通事故發生率為目的。

The accidental injury is one of reasons for top ten leading causes of death, and the main cause is motor vehicle accidents; therefore, Government, Industry and Academia are trying to make an effort to decrease traffic accident rate.At the present stage, to install Advanced Driver Assistance Systems (ADAS) on intelligent vehicle is the good way.
The purpose of this paper is to avoid colliding between pedestrian and vehicle. As following: First, the systems will use image processing method to search inputting image in order to get pedestrian’s position. Then, it can determine pedestrian’s speed and direction of movement and can predict the pedestrian’s motion vector. Through these results, it will judge whether the collision may occur or not. Otherwise, it will alert the driver to be careful.
According to the above, this paper system has divided into three parts. First part is using pedestrian detection technique which can shrink subsequent calculation range to check where the pedestrian is; second part is using the method of optical flow to get pedestrian’s motion vector that use into Grey theory to predict pedestrian’s subsequent route. second part is using the method of optical flow to get pedestrian’s motion vector that use into Grey theory to predict pedestrian’s subsequent motion path. To build predicted collision point, out of dangerous position, also calculate and estimate the time that collision may happen.
This paper will be applied to various traffic scenes. When vehicle is underway, using single camera gets front image to frame grabbing pedestrian who is be detected, and to trace moving pedestrian at the same time for obtaining pedestrian’s motion vector that can judge it may occur collision or not. It can remind driver to avoid accident because of fatigue, carelessness or improper driving. The purpose of this system is to strengthen driving safety and decrease the rate of traffic accident.

摘要…………………………………………………………………………………….i
Abstract………………………………………………………………………………..ii
致謝…………………………………………………………………………………...iii
目錄…………………………………………………..……………………………….iv
圖目錄…………………………………………………………………….………….vii
表目錄…………………………………………………………………...…………….x
第一章 緒論………………………………………………………………….……….1
1.1 研究背景………………………………………………………………………….1
1.2 研究動機與目的………………………………………………………………….4
1.3 系統架構………………………………………………………………………….6
1.4 論文貢獻………………………………………………………………………….7
1.5 論文架構………………………………………………………………………….8
第二章 相關研究…………………………………………………………….……….9
2.1 概述……………………………………………………………………………….9
2.2 現階段先進駕駛輔助系統之產品 ………………………………...…………….9
2.3 相關研究參考文獻……………………………………………………………...10
第三章 行人偵測系統 ……………………………………………….…………….19
3.1 概述………………………………………….…………………………………..19
3.2 行人偵測資料庫………………………………………………………..……….21
3.3 Haar-like特徵……………………...…………………………………………….23
3.4 Haar-like特徵模板介紹………………………………………...……………….24
3.5 行人偵測之萃取Haar-like特徵範圍……………………………….………….26
3.6 共變異矩陣描述子…………………………………………………..………….31
3.7 共變異矩陣描述子用於行人偵測的特徵向量……………………………..….32
3.8 積分影像………………………………………………………………………...35
3.9 積分影像應用於計算特徵……………………………………………………...37
3.10 行人偵測之系統架構………………………………………………………….40
第四章 獲取行人移動向量………………………………………………...……….46
4.1 概述…………………………………………………………..………………….46
4.2 Harris角點偵測………………………………………………………………….47
4.3 實際角點偵測之作法………………………………………………..………….51
4.4 光流法…………………………………………..……………………………….53
4.5 光流法之疊代計算步驟…………………………………………..…………….59
4.6 建立自主運動光流模型…………………………………..…………………….62
第五章 預測碰撞時間……………………………………...……………………….68
5.1 概述…………………………………………………………………..………….68
5.2 計算行人所在位置………………………………………………………..…….68
5.3 建立危險區域範圍……………………………………………………..……….70
5.4 預測行人移動路徑……………………………………..……………………….72
5.5 預測碰撞點之計算……………………………..……………………………….76
5.6 預測碰撞時間之計算………………………………………………………..….79
第六章 實驗結果與分析………………………………………...………………….83
6.1 概述……………………………………………………..……………………….83
6.2 行人偵測系統分析結果………………………………………..……………….83
6.3 獲取行人移動向量結果………………………………..……………………….84
6.4 預測碰撞時間……………………………………………..…………………….89
6.5 效能分析………………………..……………………………………………….96
第七章 結論及未來展望……………….………………………………………….101
7.1 結論………………………………………...………………………………….101
7.2 未來展望……………………………………………………...……………….101
參考文獻…………………………………………………………...……………….103

[1] 台灣區車輛工業同業公會, http://www.ttvma.org.tw/cht/information.php
[2] 衛生福利部國民健康署, http://health99.hpa.gov.tw/txt/Default.aspx
[3] 交通部高速公路局, http://www.freeway.gov.tw/Publish.aspx?cnid=516&p=2849
[4] Frost & Sullivan, European Market for Advanced Driver Assistance Systems – ACC, LDW, BSD, Night Vision and Ultrasonic Park Assist Systems, 2010/03
[5] TRI拓墣產業研究所, 智慧型車輛引爆車用電子新發展, TRI產業專題報告-132, Dec. 2008
[6] 內政部統計處, http://www.moi.gov.tw/stat/news_content.aspx?sn=8057
[7] 經濟部標準檢驗局, http://www.bsmi.gov.tw/wSite/mp?mp=1
[8] 顏志鴻,“使用Fisher線性判別分析之快速行人偵測系統”, 國立勤益科技大學, 碩士論文, 2011.
[9] P. Viola and M. J. Jones, “Robust real-time face detection”, Int. J. Comput. Vis., vol. 57, no. 2, pp. 137–154, 2004.
[10] P. Viola and M. J. Jones, “Rapid object detection using a boosted cascade of simple features” , IEEE Conf. on Computer Vision and Pattern Recognition, Kauai, Hawaii, 2001.
[11] C. P. Papageorgiou, M. Oren, and T. Poggio, “A general framework for object detection” , In International Conference on Computer Vision, 1998.
[12] R. A. Fisher, “The use of multiple measures in taxonomic problems”, Ann. Eugenics, vol. 7, pp. 179 - 188, 1936.
[13] O. Tuzel, F. Porikli, and P. Meer, “Region covariance: A fast descriptor for detection and classification”, Proc. Eur. Conf. Comput. Vis., vol. 2, pp. 589–600, Graz, Austria, 2006.
[14] B. D. Lucas, T. Kanade, “An Iterative Image Registration Technique with an Application to Stereo Vision”, From Proceedings of Imaging Understanding Workshop, pp. 121-130 (1981).
[15] J. Y. Bouguet, “Pyramidal Implementation of the Lucas Kanade Feature Tracker Description of the algorithm”, Intel Corporation Microprocessor Research Labs 2000.
[16] 陳璿文,“以視覺為基礎之全日嵌入式車輛偵測系統”,國立勤益科技大學電機工程系,碩士論文,2010..
[17] B. F. Lin, Y. M. Chan, L. C Fu, et al, “Incorporating Appearance and Edge Features for Vehicle Detection in the Blind-spot Area”, IEEE International Conference on Intelligent Transportation Systems, pp. 869-874, 2010.
[18] H. Weigel, P. Lindner, G. Wanielik, “Vehicle Tracking with Lane Assignment by Camera and Lidar Sensor Fusion”, IEEE Intelligent Vehicles Symposium, pp. 513-520, 2009.
[19] J. K. Suhr, H. G. Jung, “Fully-automatic Recognition of Various Parking Slot Markings in Around View Monitor (AVM) Image Sequences”, IEEE International Conference on Intelligent Transportation Systems, pp. 1294-1299, 2012..
[20] Y.C. Lin, C.T. Lin, W.C. Liu, L.T. Chen, “A Vision-Based Obstacle Detection System for Parking Assistance”, Mechanical and Systems Research Laboratories Industrial Technology Research Institute Hsinchu, Taiwan, R.O.C.
[21] .D. Langer, T. Jochem, “Fusing Radar and Vision for Detecting, Classifying and Avoiding Roadway Obstacles”, The Robotics Institute, Carnegie Mellon University, Pittsburgh, PA 15213.
[22] C.G. Keller, T. Dang, “Active Pedestrian Safety by Automatic Braking and Evasive Steering”, IEEE Transactions On Intelligent Transportation Systems, VOL. 12, NO. 4, DECEMBER 2011.
[23] 洪肇澤, “An Improved Obstacle Detection Using Optical Flow Adjusting Based on Inverse Perspective Mapping for the Vehicle Safety”, 國立台北科技大學, 碩士論文,2012.
[24] Q.Y. Chen, Z.P. Sun, X.J. An, D. Liu, J. Song, B. Zhang, “Prediction of Unintended Lane Departure based on Virtual Driver”, IEEE International Conference on Intelligent Transportation Systems, pp. 1727~1734, 2011.
[25] Q.Y. Chen, Z.P. Sun, T.T. Chen, L. Cao, “Prediction of Unintended Lane Departure based on Detection of Lane Boundary”, IEEE International Conference on Vehicular Electronics and Safety, pp.65~70, 2011.
[26] H. Dahmani, M. Chadli, A. Rabhi, A.E. Hajjaji, “Fuzzy uncertain observer with unknown inputs for Lane departure detection”, IEEE American Control Conference, pp. 688~693, 2010.
[27] J.D. Alonso, E.R. Vidal, A. Rotter, M. Mühlenberg, “Lane-Change Decision Aid System Based on Motion-Driven Vehicle Tracking”, IEEE Transactions on Vehicular Technology, VOL. 57, NO. 5, SEPTEMBER 2008.
[28] 焦點擴展, https://classconnection.s3.amazonaws.com/932/flashcards/1253932/png/flow_of_environment_711366759435046.png
[29] S. Paisitkriangkrai, C. Shen, and J. Zhang, “Performance evaluation of local features in human classification and detection”, IET Comput. Vis., vol. 2, no. 4, pp.236-246, 2008.
[30] 行人影像資料庫, http://pascal.inrialpes.fr/data/human/
[31] C. Harris, M. Stephens, “A combined edge and corner detector”, Proceedings of the 4th Alvey Vision Conference. 1988: pp. pages 147--151.
[32] D. Julong, “Control problems of grey system”, System and Control Letters, pp. 288-294, 1982.
[33] Cramer's rule, http://zh.wikipedia.org/wiki/%E5%85%8B%E8%90%8A%E5%A7%86%E6%B3%95%E5%89%87
[34] 李彥鋒,“以視覺為基礎之嵌入式車輛偵測系統”,國立勤益科技大學電機工程系,碩士論文,2009