近年來，室內機器人的運用越趨廣泛，然而機器人首先必須克服自主定位的問題，而利用相機作為感測器是不錯的出路，可以達到較低成本的同步定位及建圖(SLAM）。而在機器人作動時，可能會因為某些因素，造成機器人脫離主要路徑，導致定位失效；或是因為某些特殊工作需求，使得初始絕對位置與原始路徑不同，皆會造成即時定位及建圖(SLAM)無法延續。
本論文以輪型機器人為實驗載具，搭載一向上單鏡頭，運用擴展型卡爾曼濾波器(EKF)，整合馬達編碼器及廣角相機，利用加速穩健特徵(SURF)，作為量測更新，達到即時定位及建圖(SLAM)。
而經由即時定位及建圖(SLAM）得出的各個階段機器人位置，及收斂的影像特徵點資訊，集結成資料庫，將定位失效，或重新啟動的機器人，藉由此資料庫，利用影像定位方法，反推機器人位置，達到機器人自身位置復原和路徑整併，改善即時定位及建圖(SLAM）無法延續的問題。

The application of indoor mobile robots has been increasingly prevalent in recent years, but mobile robots have a problem which is self-localization. Integrating vision sensors and inertial sensors seems a good solution to achieve a low-cost SLAM (Simultaneously Localization and Mapping). However, there are factors could cause SLAM unsuccessfully located itself, factors such as, leaving the original paths, or restarting SLAM from different origin points.
EKF (Extended Kalman Filter) with integrated encoder sensors and an upward-looking wide-field camera has been used on a two wheels robot. In the Furthermore, by using SURF (Speeded up robust features) to acquire updating measurements to achieve SLAM. In order to enable a robot to automatically recognize its location and get back on its original path to fix the segmented SLAM issue, the experiment collected robot positions, images, and converged features data of all steps from SLAM to build a database. A database could be utilized to provide information to derive a robot position onto its original path.

[1] Riisgaard, Sren, and M.R. Blas, SLAM for Dummies: A Tutorial Approach to Simultaneous Localization and Mapping. 2012.
[2] Djugash, J., et al., Range‐Only SLAM for Robots Operating Cooperatively with Sensor Networks. Robot Motion Planning, 2007: p. 16‐735.
[3] Brown, R.G. and P.Y.C. Hwang, Introduction to Random Signals and Applied Kalman Filtering. 1997.
[4] Smith, R.C., Self, M. and P. Cheeseman, Estimating Uncertain Spatial Relationships in Robotics. USA: Elsevier, 1986: p. 435-461.
[5] Bay, H., et al., Speeded-Up Robust Features (SURF). Computer Vision and Image Understanding, 2008. 110(3): p. 346-359.
[6] Baker, K., Singular Value Decomposition Tutorial. 2005 p. 14-21.
[7] Smith, R.C. and P. Cheeseman, On the Representation and Estimation of Spatial Uncertainty. The International Journal of Robotics Research, 1986: p. 56–68.
[8] Smith, R.C., M. Self, and P. Cheeseman, Estimating Uncertain Spatial Relationships in Robotics. Proceedings of the Second Annual Conference on Uncertainty in Artificial Intelligence, 1986: p. 435–461.
[9] Leonard, J.J. and H.F. Durrant-whyte, Simultaneous map building and localization for an autonomous mobile robot. Intelligent Robots and Systems' 91.'Intelligence for Mechanical Systems, Proceedings IROS'91. IEEE/RSJ International Workshop on., 1991.
[10] Davison, A.J., Real-time simultaneous localisation and mapping with a single camera. Ninth Ieee International Conference on Computer Vision, Vols I and Ii, Proceedings, 2003: p. 1403-1410.
[11] Davison, A.J., et al., MonoSLAM: Real-time single camera SLAM. Ieee Transactions on Pattern Analysis and Machine Intelligence, 2007. 29(6): p. 1052-1067.
[12] Civera, J., A.J. Davison, and J.M.M. Montiel, Inverse Depth Parametrization for Monocular SLAM. Ieee Transactions on Robotics, 2008. 24(5): p. 932-945.
[13] Jeong, W. and K.M. Lee, CV-SLAM: A new ceiling vision-based SLAM technique. 2005 IEEE/RSJ International Conference on Intelligent Robots and Systems, Vols 1-4, 2005: p. 3070-3075.
[14] Morevec, H.P., Towards automatic visual obstacle avoidance. Proceedings of 5th International Joint Conference on Artificial Intelligence, 1977: p. 584.
[15] Harris, C. and M. Stephens, A combined corner and edge detector. In Alvey Vision Conference, 1988: p. 147-151.
[16] Lowe, D.G., Distinctive image features from scale-invariant keypoints. International Journal of Computer Vision, 2004. 60(2): p. 91-110.
[17] Bay, H., T. Tuytelaars, and L. Van Gool, SURF: Speeded up robust features. Computer Vision - Eccv 2006 , Pt 1, Proceedings, 2006. 3951: p. 404-417.
[18] Fischler, M.A. and R.C. Bolles, Random sample consensus：a paradigm for model fitting with applications to image analysis and automated cartography. Comm. of the ACM, 1981: p. 381-395.
[19] Fox, D., et al., Monte Carlo localization: Efficient position estimation for mobile robots. Sixteenth National Conference on Artificial Intelligence (Aaai-99)/Eleventh Innovative Applications of Artificial Intelligence (Iaai-99), 1999: p. 343-349.
[20] Lee, S., S. Lee, and S. Baek, Vision-Based Kidnap Recovery with SLAM for Home Cleaning Robots. Journal of Intelligent & Robotic Systems, 2012. 67(1): p. 7-24.
[21] Richard, H. and Z. Andrew, Multiple View Geometry in computer vision Second Edition. 2003.
[22] Zhang, Z.Y., A flexible new technique for camera calibration. Ieee Transactions on Pattern Analysis and Machine Intelligence, 2000. 22(11): p. 1330-1334.
[23] Dereniowski, D. and M. Kubale, 5th International Conference on Parallel Processing and Applied Mathematics. 2004: p. 985–992.
[24] 邱敬洲, 以深度影像資料庫為基礎的嵌入式全向輪機器人同步定位與建圖. 國立臺灣科技大學機械工程系碩士論文, 2014.
[25] Evans, C., Notes on the OpenSURF Library. 2009.
[26] 張博詠, 基於超寬頻無線測距與運動感測器之輪型機器人定位與建圖. 國立臺灣科技大學機械工程系碩士論文, 2015.