臺灣博碩士論文加值系統

如何讓機器人自己透過感測器感知環境，並且加以確認環境中自己的位置，這個議題在自主性行動機器人裡是相當重要的，現今已有相當不錯的方法-蒙地卡羅定位來解決這個問題，在提供正確資訊的前提下，蒙地卡羅定位能表現出很好的穩定性與準確度，但是實際應用上，蒙地卡羅定位會受限於機器人使用的感測器與機器人自身運算能力。一般研究常使用於蒙地卡羅定位的感測器分為兩大類，提供影像資訊的攝影機和提供距離資訊的感測器(如:雷射測距儀、紅外線…等等)，使用影像資訊作為蒙地卡羅定位的輸入，因為資訊量較多，所以能夠快速收斂且得到更高準確度，但需要花費的計算時間也會增加；使用距離資訊作為輸入的話，因為資訊較少，花費的計算時間少，相對的，在特徵不夠突出的話，收斂速度會變慢且準確度也會降低。本篇論文結合距離與影像資訊，利用距離資訊來限制影像的範圍，並且利用影像作為特徵的比對，不僅能如影像做到較好的準確度與收斂速度，也能擁有距離資訊的優勢，花費較少的運算時間。此外，使用的RGB-D攝影機為微軟Kinect感測器，雖然精準度較低，但比起一些感測器較為便宜，也符合實際系統應用上，可用較低的成本來完成目的。

How to make robot sense the environment and localize its position in the workspace is a very important issue in the field of autonomous mobile robots. Now there is a good method – Monte Carlo Localization to solve this problem. If sensors can provide the correct information, Monte Carlo Localization can exhibit good stability and accuracy. In the actual application, Monte Carlo Localization has restriction on the sensors and computing ability of robots. In much recent research, two kinds of sensors are applied to Monte Carlo Localization. One is camera which can provide image information, and the other is distance sensor which can provide distance information (ex. laser range finder, infrared, and etc.). The use of camera can get more information, so the convergence can be sped up and get higher accuracy. However, more computation time is required. Using the distance information as sensory input, in contrast, the computation time will be less. If the feature of distance information is not prominent, the convergence speed and localization accuracy will be reduced. In this paper, a novel approach to integrate distance and image information for Monte Carlo Localization is proposed. The distance information is used for locating the regions of interest to perform feature matching. In this way, Monte Carlo Localization is able to achieve higher localization accuracy and convergence speed. The RGB-D camera applied in our work is Microsoft Kinect sensor. Although the accuracy of Microsoft Kinect sensor is not fully satisfactory, it is cheaper than the other sensors. In actual system application, the cost can be reduced.

中文摘要 I
Abstract II
致謝 III
目錄 IV
表目錄 VI
圖目錄 VII
第一章 緒論 1
1.1 研究背景與動機 1
1.2 研究目的 2
1.3 論文貢獻 3
1.4 論文架構 4
第二章 相關文獻 5
2.1 定位演算法 5
2.1.1 卡爾曼濾波器 5
2.1.2 蒙地卡羅定位 6
2.2 感測器 6
2.2.1 距離感測器 6
2.2.2 攝影機 7
2.2.3 RGB-D攝影機 8
第三章 特徵擷取及比對 10
3.1 資料前處理 11
3.2 區域成長法 12
3.3 特徵比對 14
3.3.1 樣板與候選區域長寬比比對 14
3.3.2 顏色統計圖比對 15
第四章 蒙地卡羅定位 17
4.1 定位原理 17
4.2 模擬器 18
4.3 預測階段 18
4.4 更新階段 19
4.4.1 資料減量 19
4.4.2 重新取樣 20
第五章 實驗結果 22
5.1 環境與參數設定 22
5.2 實驗結果 24
5.2.1 順向的行進路線 24
5.2.2 逆向的行進路線 30
5.3 定位準確度與運算時間之探討 34
5.3.1 順向移動之準確度比較與運算時間評估 34
5.3.2 逆向移動之準確度比較與運算時間評估 36
5.4 實驗總結 38
第六章 結論與未來展望 39
6.1 結論 39
6.2 未來展望 39
參考文獻 40

[1]BigDog. Available:
http://www.bostondynamics.com/robot_bigdog.html
[2]ASIMO. Available:
http://world.honda.com/ASIMO/
[3]Nao. Available:
http://www.aldebaran-robotics.com/en/
[4]AIBO. Available:
http://www.sony.co.uk/support/en/hub/ERS
[5]Roomba. Available:
http://www.irobot.com/us/
[6]Kinect. Available:
http://www.microsoft.com/en-us/kinectforwindows/
[7]lenovo thinkpad. Available:
http://www.lenovo.com/tw/zh/
[8]idRobot-EZ. Available:
http://www.idminer.com.tw/product01.php?product_proid=48
[9]OpenCV Ver. 2.0. Available:
http://opencv.org/
[10]OpenNI Ver. 1.5. Available:
http://www.openni.org/
[11]R. Siegwart, I. R. Nourbakhsh, and D. Scaramuzza, Introduction to Autonomous Mobile Robots, 2nd edition, The MIT Press, 2011.
[12]R. E. Kalman, “A new approach to linear filtering and prediction problems”, Transactions of the ASME - Journal of Basic Engineering, vo1. 82, pp. 33-45, 1960.
[13]F. Dellaert, D. Fox, W. Burgard, and S. Thrun, “Monte carlo localization for mobile robots,” in Proceedings of the IEEE International Conference on Robotics and Automation, pp. 1322-1328, 1999.
[14]D. Fox, W. Burgard, and S. Thrun, “Markov Localization for mobile robots in dynamic environments”, Journal of Artificial Intelligence, pp. 391-427, 1999.
[15]S. Thrun, D. Fox, W. Burgard, and F. Dellaert, “Robust monte carlo localization for mobile robots,” Journal of Artificial Intelligence, vol. 128, pp. 99-141, 2001.
[16]I. Kim, N. Kwak, H. Lee, and B. Lee, “Improved particle fusing geometric relation between particles in FastSLAM,” Robotica, vol. 27, pp. 853-859, 2009.
[17]D. Fox, “Adapting the sample size in particle filters through KLD-sampling,” International Journal of Robotics Research, vol. 22, pp. 985-1003, 2003.
[18]B. Qin, Z. J. Chong, T. Bandyopadhyay, M. H. Ang Jr., E. Frazzoli, and D. Rus, “Curb-intersection feature based monte carlo localization on urban roads,” in Proceedings of the IEEE International Conference on Robotics and Automation, pp. 2640-2646, 2012.
[19]O. Wulf, M. Khalaf-Allah, and B. Wagner, “Using 3d data for monte carlo localization in complex indoor environments,” in Proceedings of the 2nd Bi-Annual European Conference on Mobile Robots, pp. 170-175, 2005.
[20]T. Rofer and M. Jungel, “Vision-based fast and reactive monte carlo localization,” in Proceedings of the IEEE International Conference on Robotics and Automation, pp. 856-861, 2003.
[21]J. Wolf, W. Burgard, and H. Burkhardt, “Robust vision-based localization for mobile robots using an image retrieval system based on invariant features,” in Proceedings of the IEEE International Conference on Robotics and Automation, pp. 359-365, 2002.
[22]Y. Fu, S. Tully, G. Kantor, and H. Choset, “Monte carlo localization using 3D texture maps,” in Proceedings of the IEEE International Conference on Intelligent Robots and Systems, pp. 482-487, 2011.
[23]R. Anati, D. Scaramuzza, K. G. Derpanis, and K. Daniilidis, “Robot localization using soft object detection,” in Proceedings of the IEEE International Conference on Robotics and Automation, pp. 4992-4999, 2012.
[24]S. Obwald, A. Hornumg, and M. Bennewitz, “Improved proposals for highly accurate localization using range and vision data,” in Proceedings of the IEEE International Conference on Intelligent Robots and Systems, pp. 1809-1814, 2012.
[25]J. Biswas, B. Coltin, and M. Veloso, “Corrective gradient refinement for mobile robot localization,” in Proceedings of the IEEE International Conference on Intelligent Robots and Systems, pp. 73-78, 2011.
[26]J. Biswas and M. Veloso, “Depth camera based indoor mobile robot localization and navigation,” in Proceedings of the IEEE International Conference on Robotics and Automation, pp. 1697-1702, 2012.
[27]M. F. Fallon, H. Johannsson, and J. J. Leonard, “Efficient scene simulation for robust monte carlo localization using an RGB-D camera,” in Proceedings of the IEEE International Conference on Robotics and Automation, pp. 1663-1670, 2012.
[28]M. A. Fischler and R. C. Bolles, “Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,” Communications of the ACM, vol. 24, pp. 381-395, 1981.
[29]R. C. Gonzalez and R. E. Woods, Digital Image Processing, 3rd edition, Pearson Prentice Hall, 2010.
[30]M. Petrou and P. Bosdogianni, Image Processing the Fundamentals, Wiley, 2004.
[31]J. S. Milton and J. C. Arnold, Introduction to Probability and Statistics, 4th edition, McGraw-Hill, 2002.
[32]J. J. Leonard and H. F. Durrant-Whyte, “Simultaneous map building and localization for an autonomous mobile robot,” in Proceedings of the IEEE International Workshop on Intelligent Robots and Systems, pp. 1442-1447, 1991.