Responsive image
博碩士論文 etd-0805116-180529 詳細資訊
Title page for etd-0805116-180529
論文名稱
Title
基於立體視覺與雷射測距儀融合之車前物件辨識系統設計
Design of a Front Object Recognition System for Vehicles Based on Sensor Fusion with Stereo Vision and Laser Range Finder
系所名稱
Department
畢業學年期
Year, semester
語文別
Language
學位類別
Degree
頁數
Number of pages
98
研究生
Author
指導教授
Advisor
召集委員
Convenor
口試委員
Advisory Committee
口試日期
Date of Exam
2016-08-10
繳交日期
Date of Submission
2016-09-06
關鍵字
Keywords
類神經網路、感測器融合、支持向量機、雷射測距儀、立體視覺、物件辨識
objects recognition, sensor fusion, stereo vision, neural network, laser range finder, support vector machine
統計
Statistics
本論文已被瀏覽 5634 次,被下載 59
The thesis/dissertation has been browsed 5634 times, has been downloaded 59 times.
中文摘要
在智慧車輛發展的過程中,許多國內外學術單位與相關企業已經陸續投入智慧車輛研發積極發展。隨著Google無人駕駛車的出現,人工智慧與物聯網之技術亦趨成熟,使得智慧交通系統的應用更加廣泛。
對於智慧交通系統,行車安全仍是十分重要的一環,本論文提出以立體視覺與雷射測距儀之感測器融合技術,建構一套即時車輛前方障礙物偵測辨識系統,來保障用路人的行車安全。本研究以立體視覺偵測系統為主軸,由於立體視覺演算法較耗費計算時間,本論文透過整合雷射測距儀之資訊,能夠有效加快立體視覺偵測演算法之演算速度以及偵測正確率,並透過支持向量機作為系統後端的物件辨識。
本論文先以雷射測距儀掃描車輛前方環境,當有物件出現在車輛前方道路的時候,先利用倒傳遞類神經網路將雷射測距儀之資料點轉換至立體視覺視差影像的座標上,再透過對應之視差區間進行處理並偵測物件。偵測到前方物件之後,再以偵測到的物件特徵當作支持向量機分類器之輸入,利用線性支持向量機進行物件辨識,並判定前方物件是行人或是車輛。此系統經由實驗後,驗證系統可達95%以上的正確率。
Abstract
Following the development of intelligent vehicles, the safety of vehicles is becoming an important issue. Many academic research institutions and industries related to vehicles are developing the technology and applications of intelligent vehicles. The advance of technology about Google autonomous vehicles, artificial intelligence (AI) and the internet of things (IoT) expand the applications of advanced driving assist system (ADAS) in the intelligent transportation systems.
The study proposes a front object recognition system based on sensor fusion with laser range finder and stereo vision for the safety of driving. The study focuses on the stereo vision system for the objects detection. Because the calculation time of stereo vision algorithm is too long to detect the objects immediately, the laser range finder will be integrated to improve the calculation time and the detection accuracy. After the detection stage completed, the support vector machine (SVM) be used to recognize the objects.
The proposed system scans the environment around the vehicles with laser range finder. When the objects appear on the road, the points of laser range finder are transformed to the coordinates of disparity map by the back propagation neural network. Next, the system searches the region of corresponded disparity for the objects detection. After that, it extracts the features of objects as the input of SVM classifiers. Finally, the recognition system can distinct the pedestrians or vehicles from the objects by using linear SVM classifier. The experimental results show the accuracy rate reaches 95% for the objects detection system.
目次 Table of Contents
論文審定書 i
致 謝 ii
摘 要 iii
Abstract iv
目 錄 v
圖 次 vii
表 次 xi
第一章 緒 論 1
1-1 研究動機 1
1-2 文獻回顧 3
1-3 主要貢獻 7
1-4 章節介紹 7
第二章 系統概述 8
2-1 立體視覺與雷射測距儀融合之物件偵測辨識系統 8
2-2系統功能 9
2-3 系統架構流程 10
第三章 系統實現 12
3-1 實驗平台 12
3-1-1 實驗載具平台 12
3-1-2 實驗計算平台 13
3-1-3 系統開發軟體與相關套件 13
3-2 雷射測距儀偵測系統 15
3-2-1 雷射測距儀 15
3-2-2 資料聚類演算法 18
3-2-3物件特徵萃取與偵測 21
3-3 立體視覺偵測系統 23
3-3-1 網路攝影機 23
3-3-2 相機校正 24
3-3-3 立體視覺演算法 33
3-4 雙感測器融合偵測系統 47
3-4-1 感測器融合 47
3-4-2 障礙物偵測 52
3-5 智慧型物件辨識系統 53
3-5-1 建立訓練樣本 53
3-5-2 支持向量機 55
第四章 實驗結果 59
4-1 實驗場景 59
4-2 雷射測距儀偵測系統 60
4-3 立體視覺偵測系統 62
4-4 感測器融合偵測系統 67
4-5 偵測系統測試結果 68
4-6 物件辨識系統與辨識結果 77
第五章 結論與未來展望 81
5-1 結論 81
5-2 未來展望 81
參考文獻 82
參考文獻 References
[1]中華民國內政部統計處,機動車輛及道路交通事故統計,2016。取自:http://www.moi.gov.tw/stat/index.aspx
[2]ETSI, “Intelligent Transport Systems.” Retrieved from http://www.etsi.org/technologies-clusters/technologies/intelligent-transport
[3]黃榮章,前方防撞輔助系統評價方法,車輛研測資訊,107,pp. 2-7,2009。取自:https://www.artc.org.tw/
[4]T. Cao, Z. Ziang, and J. Liu, “Perception in disparity: an efficient navigation framework for autonomous vehicles with stereo cameras,” IEEE Trans. Intell. Transp. Syst., vol. 16, pp. 2935-2948, 2015.
[5]M. A. Fischler and R. C. Bolles, “Random sample consensus: A paradigm for model fitting with applications to image analysis and automated cartography,” Commun. ACM, vol. 24, no. 6, pp. 381–395, Jun. 1981.
[6]S. Kim, H. Kim, W. Yoo, and K Huh, “Sensor fusion algorithm design in detecting vehicles using laser scanner and stereo vision,” IEEE Trans. Intell. Transp. Syst., vol. 17, pp. 1072-1084, 2016.
[7]J. Ćesić, I. Marković, I. Cvišić, and I. Petrović, “Radar and stereo vision fusion for multitarget tracking on the special Euclidean group,” Robotics and Autonomous Systems, vol. 83, pp. 338-348, 2016.
[8]Y. Wang, Y. Gao, A. Achim, and N. Dahnoun, “Robust obstacle detection based on a novel disparity calculation method and G-disparity,” Computer Vision and Image Understanding, vol.123, pp. 23-40, 2014.
[9]Club Car, “DS golf car electric vehicles.” Retrieved from http://www.clubcar.com/
[10]Acer, “Acer Aspire E15 E5-572G-591D.” Retrieved from http://www.acer.com/
[11]SICK Sensor Intelligence, “Technical description of LMS-291 laser measurement systems.” Retrieved from https://www.sick.com/tw/
[12]MOXA, “UPort™ 1250/1250I.” Retrieved from http://www.moxa.com.tw/
[13]M. Ester, H. Kriegel, J.Sander, and X. Xu, “A density-based algorithm for discovering clusters in large spatial databases with noise,” in Proc. Int. Conf. Knowledge and Data Mining, pp. 226-231.
[14]S.P. Lloyd, “Least squares quantization in PCM,” IEEE Trans. Information Theory, vol. 28, no. 2, pp. 129-137, 1982.
[15]羅技,Logitech HD Pro網路攝影機C920R。取自:http://www.logitech.com/
[16]G. Bradski and A. Kaehler, Learning OpenCV: Computer Vision with the OpenCV Library. Sebastopol, CA, USA: O’Reilly, 2008.
[17]J. Bouguet, “Camera calibration toolbox for Matlab.” Retrieved from
http://www.vision.caltech.edu/bouguetj/
[18]Z. Zhang, “A flexible new technique for camera calibration”, IEEE Trans. Pattern Anal. Mach. Intell., vol. 22, no. 11, pp. 1330-1334, 2000.
[19]R. I. Hartley, “Self-calibration of stationary cameras,” International Journal of Computer Vision, vol. 22, no. 1, pp. 5-23, 1997.
[20]H. Hirschmüller, “Accurate and efficient stereo processing by semi-global matching and mutual information,” in Proc. IEEE Conf. Comput. Vision Pattern Recog., vol. 2, pp. 807-814, 2005.
[21]R. Labayrade, D. Aubert, and J. Tarel, “Real time obstacle detection in stereovision on non flat road geometry through ‘v-disparity’ representation,” in Proc. IEEE Intelligent Vehicles Symp., vol. 2, pp. 646-651, 2002.
[22]R. O. Duda and P. E. Hart, “Use of the Hough transformation to detect lines and curves in pictures,” Commun. ACM, vol. 15, no. 1, pp. 11–15, 1972.
[23]Z. Hu and K. Uchimura, “U-V-disparity: an efficient algorithm for stereovision based scene analysis,” in Proc. IEEE Intelligent Vehicles Symp., pp. 48-54, 2005.
[24]邱約聖,以AdaBoost演算法為基礎之興趣區間物件辨識系統設計,碩士論文,國立中山大學機電與機械工程學系,2015。
[25]P. Jain, B. Kulis, J. Davis and I. Dhillon, “Metric and kernel learning using a linear transformation,” J. Mach. Learn. Res., vol. 13, pp. 519-547, 2012
[26]E. W. Weisstein, “Least squares fitting,” MathWorld–A Wolfram Web Resource. Retrieved from http://mathworld.wolfram.com/LeastSquaresFitting.html
[27]S. Haykin, Neural networks: a comprehensive foundation. Prentice Hall, 1999.
[28]王進德,類神經網路與模糊控制理論入門與應用,全華出版社,2011。
[29]N. Dalal and B. Triggs, “Histograms of oriented gradients for human detection”, in Proc. IEEE Conf. Comput. Vision Pattern Recog., vol. 1, pp. 886-893, 2005.
[30]N. Cristianini and J. Shawe-Taylor. An introduction to support vector machines. Cambridge University Press, 2000.
[31]OpenCV, “Introduction to support vector machines.” Retrieved from
http://docs.opencv.org/2.4/doc/tutorials/ml/introduction_to_svm/introduction_to_svm.html
[32]鍾國良,安全車距的拿捏與應用情境—防禦駕駛哲學PART 3,2009。取自:https://www.artc.org.tw/chinese/03_service/03_02detail.aspx?pid=839
[33]F. C. Yang, S. G. Wang, J. L. Li, Z. H. Liu, and Q. B. Sun, “An overview of internet of vehicles”, China Commun. Veh. Netw., vol. 11, no. 10, 2014
電子全文 Fulltext
本電子全文僅授權使用者為學術研究之目的,進行個人非營利性質之檢索、閱讀、列印。請遵守中華民國著作權法之相關規定,切勿任意重製、散佈、改作、轉貼、播送,以免觸法。
論文使用權限 Thesis access permission:自定論文開放時間 user define
開放時間 Available:
校內 Campus: 已公開 available
校外 Off-campus: 已公開 available


紙本論文 Printed copies
紙本論文的公開資訊在102學年度以後相對較為完整。如果需要查詢101學年度以前的紙本論文公開資訊,請聯繫圖資處紙本論文服務櫃台。如有不便之處敬請見諒。
開放時間 available 已公開 available

QR Code