本論文發展一套機器人室內定位系統，克服SURF(Speeded Up Robust Features)演算法僅能進行靜態特徵點提取的問題，本研究的貢獻是以模糊理論為基礎，應用Fuzzy邊緣檢測結合SURF，以兩種方法的整合作為本論文的改良法，透過改良演算法使影像輪廓更加清晰與貼近現實生活中物體的樣貌，SURF演算法於動態檢測的問題也獲得改善，並且藉由網路攝影機以及多項不同的程式軟體，將各項軟體與硬體進行合適的系統整合後，實現即時獲取室內機器人之座標位置。具體的研究架構包含以下內容：第一，藉由一張棋盤格之不同角度形成多角度照，多角度照形成的棋盤圖像代表校正的模式，再藉由相機校正以及針孔相機模型參數的轉換運算方法獲得機器人於空間中的所有座標點，座標記錄著機器人的移動狀態，這樣的紀錄方式將使得空間中的機器人無所遁形，並且有利於後續的機器人定位研究之實現。第二，先使用Canny邊緣檢測法或Sobel邊緣檢測法投影之機器人於環境中的樣子後，再藉由模糊邊緣檢測法使得邊緣檢測的效果得以提升並且經由改良後也能使SURF演算法於動態狀態下的問題獲得改善，抓取到足夠數量的特徵點進行特徵點的匹配，如此一來SURF演算法透過網路攝影機將能準確的標示輪型機器人之確切的位置。第三，SURF演算法藉由網路攝影機投影後，即時定位環境中之移動機器人位置並且藉由座標與改良的定位方法的整合，了解機器人於室內的動態軌跡與位置變化。整體而言，本論文的F-SURF能更快速並且更準確的定位移動機器人之位置，有效降低定位需花費之時間，同時改善SURF演算法於動態狀態標示位置跑掉之問題，提升最大有效範圍，以及將焦點鎖定在移動機器人身上，從起點到終點，改良的演算法都能於雜亂環境中準確快速地標示移動機器人的正確位置並且記錄著動態的每一個座標，最終實現了機器人之室內定位的系統。

This Thesis develops a robot indoor positioning system to overcome the problem that the SURF (Speeded Up Robust Features) algorithm can only extract static feature points. The contribution of this research is based on fuzzy theory, using Fuzzy edge detection combined with SURF. The integration of methods is an improved method. Through the improved method, the image is made clearer and closer to the appearance of real objects. The problem of SURF in dynamic detection is also improved. After the software and hardware are integrated, the coordinates of the indoor robot can be obtained in real time. The architecture includes the following contents: First, a checkerboard image formed by a multi-angle photo of a checkerboard represents the correction mode, and then the coordinate points of the robot are obtained through the camera correction and the calculation of the pinhole camera model parameters, which is conducive to the subsequent positioning Research realization. Second, first use Canny or Sobel projection of the robot in the environment, and then use the fuzzy edge detection method to improve the effect of edge detection. After the improvement, the problem of SURF in the dynamic state can be improved, and enough The feature points are matched so that SURF can accurately mark the position of the wheeled robot through the webcam. Third, SURF locates the robot's position in the environment in real time after projection from a webcam and integrates coordinates and improved positioning methods to understand the dynamic trajectory and position changes of the robot in the room. On the whole, the improved method of this thesis can locate the position of the robot more quickly and accurately, effectively reducing the time spent in positioning, and at the same time improve the SURF to run away at the dynamic state marking position, increase the maximum effective range, and lock the focus to the robot. From the start point to the end point, the improved method can accurately and quickly mark the position of the mobile robot in a messy environment and record all coordinates, realizing an indoor positioning system for the robot.

[1] A. Morar, “A Comprehensive Survey of Indoor Localization Methods Based on Computer Vision,” State-of-the-Art Sensors Technology in Romania 2020, vol. 20, May 2020.
[2] R. Pradeep Kumar Reddy, C. Nagaraju, “Improved Canny Edge Detection Technique Using S-Membership Function,” International Journal of Engineering and Advanced Technology, vol. 8, no. 6, pp. 43-49, Aug. 2019.
[3] A. Khunteta, D. Ghosh, "Edge Detection via Edge-Strength Estimation Using Fuzzy Reasoning and Optimal Threshold Selection Using Particle Swarm Optimization," Advances in Fuzzy Systems, vol. 2014, Article ID 365817, 17 pages, Dec. 2014.
[4] H. Bay, T. Tuytelaars, and L. V. Gool, “SURF: Speeded Up Robust Features,” Computer Vision and Image Understanding, vol. 110, no. 3, pp. 346-359, June 2008.
[5] E. Z. Hao, S. Srigrarom, “Development of 3D Feature Detection and on Board Mapping Algorithm from Video Camera for Navigation,” Journal of Applied Science and Engineering, vol. 19, no. 1, pp. 23-39, Mar. 2016.
[6] Z. Zhang, “A Flexible New Technique for Camera Calibration,” in IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 22, no. 11, pp.1330-1334, Nov. 2000.
[7] J. Y. Zhang, Y. Chen, X. X. Huang, “Edge detection of images based on improved Sobel operator and genetic algorithms,” in 2009 International Conference on Image Analysis and Signal Processing, Linhai, Apr. 2009, pp. 31-35.
[8] B. A. Victoria, R. S. A. Jorge and P. H. L. Manuel, “SIFT-SURF commutation using fuzzy logic to image mosaicking,” in 2017 14th International Conference on Electrical Engineering, Computing Science and Automatic Control (CCE), Mexico City, Oct. 2017, pp. 1-6.
[9] P. Jiang, S. Zhao, S. Cheng, “Rotational Invariant LBP-SURF for Fast and Robust Image Matching,” in International Conference on Signal Processing and Communication Systems, Cairns, Dec. 2015, pp.1-7.
[10] H. Bay, A. Ess, T. Tuytelarrs, L. V. Gool, “Speeded-Up Robust Features (SURF),” Computer Vision and Image Understanding, vol. 110, Pages. 346-359, June. 2008.
[11] CSDN博客 (2015)。SURF原理與源碼解析, 上網日期：2021年10月15日。檢自：https://read01.com/zh-tw/QLayA.html#.YbxYKWhBxPa.
[12] L. R. Liang, C. G. Looney. Applied Soft Computing 3 (2003). NOISE REDUCTION using Fuzzy Filtering. Retrieved October 10, 2021, from https://devendrapratapyadav.github.io/Fuzzy_Image_processing/.
[13] L. Juan, O. Gwun, “A Comparison of SIFT, PCA-SIFT and SURF,” International Journal of Image Processing, vol. 3, pp. 143-152, Oct. 2009.
[14] prabhakar C J and Praveen Kumar P U, “LBP-SURF Descriptor with Color Invariant and Texture Based Features for Underwater Images,” ICVGIP, Mumbai, Dec. 2012.
[15] D. Van De Ville, M. Nachtegael, D. Van der Weken, E. E. Kerre, W. Philips and I. Lemahieu, "Noise reduction by fuzzy image filtering," in IEEE Transactions on Fuzzy Systems, vol. 11, no. 4, pp. 429-436, Aug. 2003.
[16] H. K. Kwan, "Fuzzy filters for noisy image filtering," Proceedings of the 2003 International Symposium on Circuits and Systems, 2003. ISCAS '03., May 2003.
[17] F. Russo and G. Ramponi, "A fuzzy filter for images corrupted by impulse noise," in IEEE Signal Processing Letters, vol. 3, no. 6, pp. 168-170, June 1996.
[18] F. Russo and G. Ramponi, "A fuzzy operator for the enhancement of blurred and noisy images," in IEEE Transactions on Image Processing, vol. 4, no. 8, pp. 1169-1174, Aug. 1995.