臺灣博碩士論文加值系統

本文以即時外觀建圖演算法(Real-Time Appearance-Based Mapping, RTAB-MAP)，搭配彩色深度相機(RGB-D)和慣性量測單元感測器(Inertial Measurement Unit, IMU)，實現無人機之室內定位。
基於圖像特徵的迴環檢測法(Feature-based Loop-closure Detection)是RTAB-MAP演算法的主要功能之一，在地圖最佳化的過程中經由限制地圖的特徵點數量以節省記憶體空間，使得迴環檢測能在固定的時間內完成，滿足長期和大規模環境即時建圖的要求。故本文採用RTAB-MAP並以RGB-D深度相機為基礎，根據相鄰圖像間的特徵點，估計出相機運動軌跡、里程以及局部地圖，再進一步最佳化以取得全局一致的地圖資訊與定位。然而單純使用相機進行即時定位與建圖(Simultaneous localization and mapping, SLAM) 時，當運動過快，相機會出現運動模糊，或者影像兩幀重疊區域太少，造成無法進行特徵匹配，故本研究搭配IMU以解決此問題。
本研究使無人機與地面站建立通訊後，地面站蒐集無人機之感測資料，以RTAB-MAP搭配IMU完成SLAM建立地圖資訊後，預先規劃出一條可安全飛行之路徑，再發送航點給無人機，以實現室內定位。模擬及實驗結果顯示，在室內環境之航點飛行，軌跡平均誤差平均在5公分以內。

This work implements the Real-Time Appearance Map (RTAB-MAP) algorithm on a unmanned aerial vehicle (UAV) to perform indoor localization task.
Feature-based loop-closure detection is one of the main functions of the RTAB-MAP algorithm. Basing on image features, RTAB-MAP manages memory space by limiting the number of feature points for map optimization. Such that the loop-closure detection can be processed in fixed time, and meet the requirements of real-time mapping in long-term and large-scale environments. The RTAB-MAP, equipping with a RGB-D camera, estimates the camera moving trajectory, mileage and local map according to the feature points between adjacent images to obtain globally consistent map information and localization. However, when using the camera for simultaneous localization and mapping (SLAM), images are prone to blurred by the fast motion of the vehicle, or the overlapping area of two consecutive image frames is too few, causing the feature matching failed. Therefore, this study combines an IMU to solve the problem.
In this study, when the communication between the experimental UAV and ground station is well established, the ground station collects the sensory data from the UAV and build the map of an indoor environment through the RTAB-MAP method. According to the map, the system plans a path and sends waypoints to test the localization of the drone. Simulation and experimental results reveal that average trajectory errors is within ± 5 cm.

摘要...i
Abstract...ii
誌謝...iii
目錄...iv
表目錄...vi
圖目錄...viii
第一章 緒論...1
1.1 簡介...1
1.2 文獻探討...2
1.3 研究動機...3
1.4 研究方法...4
1.5 論文架構...5
第二章 RTAB-MAP演算法...6
2.1 Robot Operating System (ROS)...6
2.2 感測器數據...9
2.3 視覺里程表(VIO)...9
2.4 閉環檢測...15
2.5 地圖最佳化...17
第三章 RTAB-MAP之無人機定位模擬...18
3.1 Gazebo...18
3.2 URDF...19
3.3 Rviz...20
3.4 Robot_localization...21
3.5 RTAB-MAP之室內定位模擬結果...22
3.6 模擬結果討論...28
第四章 四旋翼無人機機構與系統設計...29
4.1 四旋翼無人機整體機構...29
4.2 硬體設備...30
4.3 地面站...33
4.4 系統整合...34
第五章 RTAB-MAP之無人機定位實驗...36
5.1 PID控制器原理...36
5.2 PID控制器調整...36
5.3 RTAB-MAP之室內定位實驗結果...38
5.4 實驗結果討論...41
第六章 結論與未來展望...42
6.1 討論與結論...42
6.2 本研究之貢獻...42
6.3 未來展望...43
參考文獻...44
Extended Abstract...48

[1]C. J. Hegarty and E. Chatre, “Evolution of The Global Navigation SatelliteSystem (GNSS),” Proceedings of the IEEE, Vol. 96, Iss. 12, Dec. 2008
[2]Z. Biskaduros and R. M. Buehrer, “Collaborative Localization Enhancement to The Global Positioning System,” 2012 9th Workshop on Positioning, Navigation and Communication, Mar. 2012
[3]J. Li, J. Zhang, B. Zhang and B. Shen, “Operation and Development of BeiDou Navigation Satellite System,” 2015 International Association of Institutes of Navigation World Congress (IAIN), Dec. 2015
[4]D. Whyte and T. Bailey, “Simultaneous Localization and Mapping (SLAM): Part I The Essential Algorithms,” IEEE Robotics & Automation Magazine, Vol. 13, Iss. 2, Jun. 2006
[5]X. Liang, H. Chen, Y. Li and Y. Liu, “Visual Laser-SLAM in Large-scale Indoor Environments,” 2016 IEEE International Conference on Robotics and Biomimetics (ROBIO), Mar. 2017
[6]C. Tao, Z. Gao, J. Yan, C. Li and G. Cui, “Indoor 3D Semantic Robot VSLAM Based on Mask Regional Convolutional Neural Network,” IEEE Access, Vol. 8, Mar. 2020
[7]R. M. Artal and J. D. Tardós, “ORB-SLAM2: An Open-Source SLAM System for Monocular, Stereo, and RGB-D Cameras,” IEEE Transactions on Robotics, Vol. 33,Iss. 5, Jun. 2017
[8]D. Cremers, J. Hess, J. Sturm, R. M. Artal and O. Burgard, “3-D Mapping with an RGB-D Camera,” IEEE Transactions on Robotics, Vol. 30, Iss. 1, Sep. 2013
[9]S. Ramachandran and F. Sahin, “Smart Walker V: Implementation of RTAB-Map Algorithm,” 2019 14th Annual Conference System of Systems Engineering (SoSE), Jul. 2019
[10]A. Bachrach, A. d. Winter, R. He, G. Hemann, S. Prentice and N. Roy, “RANGE - Robust Autonomous Navigation in GPS-Denied Environments,” 2010 IEEE International Conference on Robotics and Automation, May 2010
[11]G. Chowdhary, D. M. Sobers, C. Pravitra, H. C. Christmann and A. Wu, “Self-Contained Autonomous Indoor Flight with Ranging Sensor Navigation,” AIAA J. Guid. Control Dyn, Vol. 35, No. 6, Oct. 2012
[12]M. Achtelik, M. Achtelik, S. Weiss and R. Siegwart, "Onboard IMU and Monocular Vision Based Control for MAVs in Unknown In- and Outdoor Environments," 2011 IEEE International Conference on Robotics and Automation, May 2011
[13]S. J. Haddadi and E. B. Castelan, "Visual-Inertial Fusion for Indoor Autonomous Navigation of a Quadrotor Using ORB-SLAM," 2018 Latin American Robotic Symposium, 2018 Brazilian Symposium on Robotics (SBR) and 2018 Workshop on Robotics in Education (WRE), Nov. 2018
[14]P. Zhao, R. Li, Y. Shi and L. He, "Design of 3D Reconstruction System on Quadrotor Fusing LiDAR and Camera," 2020 39th Chinese Control Conference (CCC), Jul. 2020
[15]W. G. Aguilar, G. A. Rodríguez, L. Á. and S. Sandoval, "Visual SLAM with a RGB-D Camera on a Quadrotor UAV Using on-Board Processing," International Work-Conference on Artificial Neural Networks, May 2017
[16]K. Kamarudin, S.M. Mamduh, A.S.A Yeon, R. Visvanathan, A.Y.M Shakaff, A. Zakaria, L.M. Kamarudin and N.A. Rahim, "Improving Performance of 2D SLAM Methods by Complementing Kinect with Laser Scanner," 2015 IEEE International Symposium on Robotics and Intelligent Sensors (IRIS), Oct. 2015
[17]N. Ragot, R. Khemmar, A. Pokala, R. Rossi and J. Y. Ertaud, "Benchmark of Visual SLAM Algorithms: ORB-SLAM2 vs RTAB-Map*," 2019 Eighth International Conference on Emerging Security Technologies (EST), Jul. 2019
[18]Robot Operating System，https://reurl.cc/0O8Mpx
[19]TF，http://wiki.ros.org/tf
[20]ROS的TF-transform，https://robot-ros.com/robot/35127.html
[21]ORB，https://reurl.cc/7ovzY1
[22]漢明距離，https://reurl.cc/A8evMp
[23]SIFT，https://reurl.cc/6l3nqk
[24]SURF，https://reurl.cc/k0jKkK
[25]FAST，https://reurl.cc/3LlqmX
[26]Ebrahim, S. Prasad and M. shehata, “Image Matching Using SIFT, SURF, BRIEF and ORB: Performance Comparison for Distorted Images,” 2015 Newfoundland Electrical and Computer Engineering Conference, Nov. 2015
[27]Brief，https://reurl.cc/WLQyek
[28]RANSAC，https://reurl.cc/j5Zp71
[29]單應性矩陣，https://reurl.cc/GrylV3
[30]homography, essential and fundamental matrix，https://reurl.cc/q8R5xR
[31]Homogeneous coordinates，https://reurl.cc/A8rODZ
[32]RANSAC前後匹配結果，https://reurl.cc/e9vAm7
[33]S. A. S. Mohamed, M. Haghbayan, T. Westerlund, J. Heikkonen, H. Tenhunen and J. Plosila, "A Survey on Odometry for Autonomous Navigation Systems," IEEE Access(Volume:7), Jul. 2019
[34]PNP演算法，https://reurl.cc/od2Gr3
[35]Bundle adjustment，https://reurl.cc/3LE87L
[36]最小平方法，https://reurl.cc/N6d5Rm
[37]Gauss–Newton algorithm，https://reurl.cc/4mVyaV
[38]Levenberg–Marquardt algorithm，https://reurl.cc/zznbWe
[39]視覺詞袋模型，https://reurl.cc/n03XRd
[40]D.Fox, J.Hightower, L.Liao, D.Schulz and G.Borriello, "Bayesian filtering for local estimation," IEEE Pervasive Computing 2(3):24-33, Aug. 2003
[41]R. Kümmerle, G. Grisetti, H. Strasdat, K. Konolige and W. Burgard, "G2o: A general framework for graph optimization," 2011 IEEE International Conference on Robotics and Automation, May 2011
[42]F. Dellaert, "Factor graphs and GTSAM: A hands-on introduction," Tech. Rep. GT-RIM-CP&R-2012-002, Georgia Institute of Technology, Sep. 2012
[43]G. Grisetti, S. Grzonka, C. Stachniss, P. Pfaff and W. Burgard, "Efficient estimation of accurate maximum likelihood maps in 3d," In Proc. of the IEEE/RSJ International Conference on Intelligent Robots and Systems, Nov. 2007
[44]Ferrão, J. M. Miranda, "Study of SLAM algorithms for autonomous navigation in unstructured environments," Mar. 2018
[45]Gazebo，http://wiki.ros.org/gazebo
[46]URDF，https://reurl.cc/e8NvmL
[47]Rviz，http://wiki.ros.org/rviz
[48]Mavlink，https://reurl.cc/N6WA6e
[49]Mavros，https://reurl.cc/R1Znb9
[50]機器人模型架構，https://reurl.cc/4m08ND
[51]Intel Aero Ready to Fly Drone，https://ardupilot.org/plane/docs/common-intel-aero-rtf.html
[52]Intel Aero compute board，https://reurl.cc/m9gxe7
[53]Pixhawk，https://reurl.cc/2gnKxn
[54]Intel Dual Band Wireless-AC 8260，https://reurl.cc/bRvNGo
[55]FPS，https://reurl.cc/0O3aRl
[56]IMU，https://reurl.cc/8nkLAo
[57]MPU-6500，https://reurl.cc/3LzpQV
[58]HMC5883L，https://reurl.cc/A85YnK
[59]Spektrum，https://reurl.cc/k06gAq
[60]Spektrum DXe transmitter，https://reurl.cc/7oamz9
[61]Spektrum* DSMX serial receiver，https://reurl.cc/4mxnVY
[62]鋰電池，https://reurl.cc/zzVLY7
[63]QGroundControl，http://qgroundcontrol.com/
[64]TCP/IP，https://reurl.cc/WL6R85
[65]SSH，https://reurl.cc/Z7K92V
[66]PID，https://reurl.cc/N6kpXQ
[67]無人機的三軸，https://reurl.cc/7ogeyN
[68]C. S. Gurel, "Real-Time 2D And 3D SLAM Using RTAB-MAP, GMapping, And Cartographer Packages," Aug. 2018