Responsive image
博碩士論文 etd-0920117-153723 詳細資訊
Title page for etd-0920117-153723
論文名稱
Title
於機器人作業系統使用多個3D深度感測器實現接近感測
Approaching Detection with Multiple 3D Depth Sensor in Robot Operating System
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
55
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
2017-08-28
繳交日期
Date of Submission
2017-10-30
關鍵字
Keywords
Octree、固定半徑近鄰、ROS、點雲、隨機抽樣一致
ROS, Octree, fixed-radius near neighbors, pointcloud, random sample consensus
統計
Statistics
本論文已被瀏覽 5641 次,被下載 1358
The thesis/dissertation has been browsed 5641 times, has been downloaded 1358 times.
中文摘要
本論文以使用3D深度感測器獲得包含一個固定姿態的機器人和該機器人周遭環境的外觀點雲資訊為底。從中確定機器人所在的位置以及分離屬於機器人的點雲之後,我們偵測機器人後續的點雲變化,進而發展出一套對機器人是否即將與周遭物體發生碰撞的第三方接近感測系統。整個系統大致可分為三個部分,第一部分是將多台3D深度感測器所獲得的點雲資訊轉換到世界座標系後合併出世界點雲,另外再透過機器人模擬軟體,使用已知的機器人模型檔案生成理想的機器人模型點雲。第二部分是使用隨機抽樣一致演算法,從世界點雲中找到並分離出與理想模型點雲形狀一致的機器人點雲,而其餘的世界點雲則會被視為環境點雲。最後一個部分是將兩點雲資料結構轉換成Octree 資料結構,並對之後的世界點雲進行固定半徑近鄰搜索,以搜索的結果分類點雲,之後系統更新Octree以及判斷機器人是否過於接近環境中的物體。最後,以精密機械研究發展中心之七軸工業型機器人實現本研究之方法來驗證可行性,以達到對機器人進行接近感測的目的。為了讓整個系統容易與其他機器人或感應器結合使用,我們選擇在機器人作業系統(Robot Operating System)底下實現這整個系統。
Abstract
This thesis is based on using 3D depth sensors to obtain a surface pointcloud data including a fixed-pose robot and the surroundings of the robot. After determining where the robot is located and separating the pointcloud that belongs to the robot, we detects the subsequent pointcloud change of the robot. Further, develop a set of third-party approaching detection system for whether the robot is about to collide with the surrounding objects. The whole system can be divided into three parts, the first part is to transform pointclouds obtained from multiple 3D depth sensors into world coordinate system, and to merge transformed pointclouds into world pointcloud. In addition, through the robot simulation software, we use the known robot model file to generate the ideal robot model pointcloud. The second part is to use random sample consensus algorithm to find robot pointcloud whose shape is consistent with ideal model pointcloud and separate it from world pointcloud, and the rest of world pointcloud is regarded as the environment pointcloud. The last part is to convert the two pointclouds from pointcloud data structure into Octree data structure. And after that, the following world pointclouds are used to perform fixed-radius near neighbors search. Pointclouds are classified by the results of the search, and then system updates Octrees and determine if the robot is too close to the object in environment. Finally, we verify the feasibility of the method of this thesis by implementing this system at PMC 7-Axis manipulator to achieve the purpose of implementing approaching detection. In order to make the entire system easy to use with other robots or sensors, we choose to implement the entire system under Robot Operating System (ROS).
目次 Table of Contents
[論文審定書+i]
[誌謝+ii]
[中文摘要+iii]
[Abstract+iv]
[目錄+v]
[圖目錄+vii]
[表目錄+ix]
[第一章 緒論+1]
[1-1 研究動機與目的+1]
[1-2 論文架構+2]
[第二章 背景介紹+3]
[2-1 機器人作業系統(ROBOT OPERATING SYSTEM,ROS)+3]
[2-1-1 Rviz+5]
[2-2 建立世界點雲+5]
[2-2-1 Microsoft Kinect+5]
[2-2-2 Open Natural Interaction+6]
[2-2-3 Point Cloud Library+6]
[2-3 機器人點雲模型+7]
[2-3-1 Unified Robot Description Format+7]
[2-3-2 Gazebo+7]
[2-4 隨機抽樣一致演算法(RANDOM SAMPLE CONSENSUS,RANSAC)+8]
[2-5 OCTREE+9]
[第三章 方法設計與分析+10]
[3-1 系統架構+10]
[3-2 建立世界點雲+12]
[3-3 建立機器人點雲模型+13]
[3-4 從世界點雲分離模型點雲+14]
[3-5 轉換成OCTREE來實現接近感測+22]
[第四章 實驗設計與結果+26]
[4-1 實驗環境+26]
[4-2 實驗方法+28]
[4-3 實驗結果+30]
[4-3-1 融合點雲次數對於執行時間的影響+34]
[4-3-2 融合前降取樣與否對於執行時間的影響+35]
[4-3-3 Octree解析度對更新時間之影響+36]
[4-3-4 kinect點雲距離與實際量測值之比較+37]
[第五章 結論與未來展望+42]
[5-1 結論+42]
[5-2 未來展望+42]
[參考文獻+43]
參考文獻 References
[1] R. Rusu, S. Cousins, "3D is here: Point cloud library (PCL)", Int. Conf. Robot. Autom. (ICRA), pp. 1-4, 2011.
[2] D. Alexiadis, D. Zarpalas, P. Daras, "Real-time full 3-D reconstruction of moving foreground objects from multiple consumer depth cameras", IEEE Trans. Multimedia, vol. 15, no. 2, pp. 339-358, 2013.
[3] P.J. Besl, N.D. McKay, "A method for registration of 3-d shapes", IEEE Trans. pattern analysis and machine intelligence, vol. 14, no. 2, pp. 239-256, 1992.
[4] H. Wang, C. Schmid, "Action recognition with improved trajectories", Computer Vision (ICCV) 2013 IEEE International Conference on, pp. 3551-3558, 2013.
[5] Martin A. Fischler, Robert C. Bolles, "Random Sample Consensus: A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography", Comm. of the ACM, vol. 24, no. 6, pp. 381-395, 1981.
[6] X. Wu, Q. Zhao, "Deformed palmprint matching based on stable regions", IEEE Trans. Image Process., vol. 24, no. 12, pp. 4978-4989, 2015.
[7] P. H. S. Torr and A. Zisserman. MLESAC: A new robust estimator with application to estimating image geometry. CVIU, 78:138-156, 2000.
[8] O. Chum, J. Matas, "Optimal randomized RANSAC", IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 30, no. 8, pp. 1472-1482, 2008.
[9] J. Matas, O. Chum, "Randomized ransac with sequential probability ratio test", International Conference on Computer Vision, 2005.
[10] J. Elseberg, D. Borrmann, A. Nüchter, "One billion points in the cloud - an octree for efficient processing of 3D laser scans", ISPRS J. of Photogramm. and Rem. Sens., vol. 76, pp. 76-88, 2013.
[11] A. Hornung, K. M. Wurm, M. Bennewitz, C. Stachniss, W. Burgard, "OctoMap: An efficient probabilistic 3D mapping framework based on octrees", Autonomous Robots, 2013.
[12] M. Quigley, B. Gerkey, K. Conley, J. Faust, T. Foote, J. Leibs, E. Berger, R. Wheeler, A. Ng, "ROS: An open-source robot operating system", Proc. ICRA Workshop Open Source Software, 2009.
[13] C. Rich, B. Ponsler, A. Holroyd, C. L. Sidner, "Recognizing engagement in human-robot interaction", Human-Robot Interaction (HRI) 2010 5th ACM/IEEE International Conference, pp. 375-382, 2010.
[14] ROS.org | Powering the world's robots, [online] Available: http://www.ros.org/.
[15] T. Foote, "tf: The transform library", Technologies for Practical Robot Applications (TePRA) 2013 IEEE International Conference on., pp. 1-6, 2013.
[16] Kinect - Windows app development, [online] Available: https://developer.microsoft.com/en-us/windows/kinect.
[17] PCL - Point Cloud Library (PCL), [online] Available: http://pointclouds.org/.
[18] urdf - ROS Wiki, [online] Available: http://wiki.ros.org/urdf.
[19] Gazebo, [online] Available: http://gazebosim.org/.
[20] D. J. Meagher, “Octree Encoding: A New Technique for the Representation Manipulation and Display of Arbitrary Three-Dimensional Objects by Computer”, 1980.
[21] P. Torr, A. Zisserman, "MLESAC: A new robust estimator with application to estimating image geometry", Computer Vision and Image Understanding, vol. 78, pp. 138-156, 2000.
[22] O. Chum, J. Matas, "Randomized RANSAC with T(d,d) Test", Proc. British Machine Vision Conf., pp. 448-457, 2002.
[23] S. Umeyama, "Least-Squares Estimation of Transformation Parameters Between Two Point Patterns", IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 13, no. 4, pp. 376-380, 1991.
[24] J. L. Bentley, "A Survey of Techniques for Fixed Radius Near Neighbor Searching", Stanford University Tech. Rep., 1975.
[25] W. G. Aref, D. Barbará, S. Johnson, S. Mehrotra, "Efficient processing of proximity queries for large databases", Proc. of the 11th ICDE, 1995.
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:校內校外完全公開 unrestricted
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