近年來工業已邁向新形態的產業模式，自動化生產的趨勢提高生產的效率，傳統人力已逐步被機械取代。機器的穩定性高，運作時間長， 能確保產線的生產效能與品質。
傳統曲面加工大多數以CNC多軸工具機為主，透過CNC多軸工具機程式的設定，使刀具能夠以曲線的路徑對工件加工。但還是需要以靠人工的方式將工件固定於CNC多軸工具機加工機中，而人工擺放的位置會影響加工精度。若是工件已被安裝在機台上，CNC多軸工具機無法對工件進行二次加工。所以需要發展一套自動化曲面掃描與辨識的系統，能即時在任意空間對物件掃描。
本論文將利用立體視覺的技術結合機械手臂，能改變拍攝視角從各方向進行掃描且縫合掃描結果，解決視線被遮擋的問題，同時避免人工重新安裝物件產生的安裝誤差。並且從掃描物件的結果分析物件邊緣毛邊的位置，同時使用神經網路提高毛邊檢測的精度。未來此技術能夠應用於其他自動化光學檢測，如工廠管線檢修、渦輪引擎葉片修補、產線自動分類瑕疵品。

Automatic machines have gradually replaced humans in various activities; this has improved both the stability and effectiveness of manufacturing operations.
Five-axis computer numerical control machines are commonly used for traditional surface milling machining. However, manually repositioning the workpiece for the computer numerical control machines is still necessary, and it inevitably lowers position accuracy. Therefore, developing an integrated fusion system that can automatically scan objects to rectify position error is crucial.
The primary objective of this research is to integrate a stereo vision system into an existing industrial manipulator. With our proposed method and system, position error can be reduced by avoiding workpiece movement during operations. In addition, reconstructing the scan results from multiple viewpoints avoids blocking vision problem. The scan result can also be used for deburring. An artificial convolution neural network is proposed to increase the positioning accuracy of burr detection. In the future, this method can be integrated with other optical inspection systems and implemented for automated optical inspection, such as in pipeline maintenance, turbine engine maintenance, and automatic classification of defects.

摘要 I
Abstract II
致謝 III
目錄 IV
圖目錄 VII
表目錄 XI
第一章 緒論 1
1.1 前言 1
1.2 研究動機 2
1.3 文獻回顧 2
1.3.1 立體視覺系統之相關應用 3
1.3.2 毛邊檢測 9
1.4 研究目標與方法 11
1.5 預期結果 12
第二章 自動化掃描系統之建模 13
2.1 相機相關參數 13
2.1.1 相機內部參數 14
2.1.2 相機外部參數 16
2.1.3 失真參數 17
2.1.4 粒子群最佳化 18
2.2 立體視覺系統 20
2.2.1 立體視覺原理 21
2.2.2 Epipolar Geometry 22
2.2.3 Homography 24
2.2.4 影像配對 26
2.3 毛邊辨識 29
2.3.1 迭代最近點算法 29
2.3.2 卷積神經網路 31
2.4 機械手臂路徑規劃 32
2.4.1 順向運動解 33
2.4.2 逆向運動解 34
第三章 自動化掃描系統整合 37
3.1 系統架設 37
3.1.1 硬體架設 38
3.1.2 參數分析 39
3.1.3 重複精度測試 41
3.2 立體視覺系統校正 45
3.2.1 相機內部參數與失真校正 45
3.2.2 立體視覺影像校正 47
3.2.3 影像失真與立體視覺校正 51
3.2.4 粒子群最佳化立體視覺校正 52
3.3 量測金屬工件 61
3.3.1 量測金屬平面 61
3.3.2 量測金屬曲面 64
3.4 Eye-on-hand 系統整合 66
第四章 毛邊掃描應用 69
4.1 鋁塊毛邊掃描實驗架設 69
4.2 立體視覺系統掃描毛邊 72
4.3 建立卷積神經網路 74
4.4 毛邊定位結果比較 78
第五章 結論 82
5.1 總結 82
5.2 本文貢獻 83
5.3 未來展望 85
參考文獻 86

[1] N. A. Borghese, G. Ferrigno, G. Baroni, A. Pedotti, S. Ferrari, and R. Savarè, "Autoscan: A flexible and portable 3D scanner," IEEE Computer Graphics and Applications, vol. 18, no. 3, pp. 38-41, 1998.
[2] J. Cooper, M. Azhar, T. Gee, W. Van Der Mark, P. Delmas, and G. Gimel'farb, "A raspberry pi 2-based stereo camera depth meter," in 2017 Fifteenth IAPR International Conference on Machine Vision Applications (MVA), 2017, pp. 274-277: IEEE.
[3] A. Fusiello, E. Trucco, and A. Verri, "A compact algorithm for rectification of stereo pairs," Machine Vision and Applications, vol. 12, no. 1, pp. 16-22, 2000.
[4] C. H. Esteban and F. Schmitt, "Silhouette and stereo fusion for 3D object modeling," Computer Vision and Image Understanding, vol. 96, no. 3, pp. 367-392, 2004.
[5] H. Kieu, T. Pan, Z. Wang, M. Le, H. Nguyen, and M. Vo, "Accurate 3D shape measurement of multiple separate objects with stereo vision," Measurement Science and Technology, vol. 25, no. 3, p. 035401, 2014.
[6] T.-H. Lin, "Resolution adjustable 3D scanner based on using stereo cameras," in 2013 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, 2013, pp. 1-5: IEEE.
[7] Z. Lv and Z. Zhang, "Build 3D scanner system based on binocular stereo vision," J. Comput, vol. 7, pp. 399-404, 2012.
[8] T. Weise, B. Leibe, and L. Van Gool, "Fast 3d scanning with automatic motion compensation," in 2007 IEEE Conference on Computer Vision and Pattern Recognition, 2007, pp. 1-8: IEEE.
[9] R. Sagawa, K. Sakashita, N. Kasuya, H. Kawasaki, R. Furukawa, and Y. Yagi, "Grid-based active stereo with single-colored wave pattern for dense one-shot 3D scan," in 2012 Second International Conference on 3D Imaging, Modeling, Processing, Visualization & Transmission, 2012, pp. 363-370: IEEE.
[10] C. Hernández, G. Vogiatzis, G. J. Brostow, B. Stenger, and R. Cipolla, "Non-rigid photometric stereo with colored lights," in 2007 IEEE 11th International Conference on Computer Vision, 2007, pp. 1-8: IEEE.
[11] D. Scharstein et al., "High-resolution stereo datasets with subpixel-accurate ground truth," in German conference on pattern recognition, 2014, pp. 31-42: Springer.
[12] W. Jang, C. Je, Y. Seo, and S. W. Lee, "Structured-light stereo: Comparative analysis and integration of structured-light and active stereo for measuring dynamic shape," Optics and Lasers in Engineering, vol. 51, no. 11, pp. 1255-1264, 2013.
[13] G. K. Cheung, S. Baker, and T. Kanade, "Visual hull alignment and refinement across time: A 3d reconstruction algorithm combining shape-from-silhouette with stereo," in 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2003. Proceedings., 2003, vol. 2, pp. II-375: IEEE.
[14] H. Rebecq, G. Gallego, E. Mueggler, and D. Scaramuzza, "EMVS: Event-based multi-view stereo—3D reconstruction with an event camera in real-time," International Journal of Computer Vision, vol. 126, no. 12, pp. 1394-1414, 2018.
[15] M. Roberts et al., "Submodular trajectory optimization for aerial 3d scanning," in Proceedings of the IEEE International Conference on Computer Vision, 2017, pp. 5324-5333.
[16] M. Dinham and G. Fang, "Autonomous weld seam identification and localisation using eye-in-hand stereo vision for robotic arc welding," Robotics and Computer-Integrated Manufacturing, vol. 29, no. 5, pp. 288-301, 2013.
[17] Y. Liu, H. Liu, F.-l. Ni, and W.-f. Xu, "New self-calibration approach to space robots based on hand-eye vision," Journal of Central South University of Technology, vol. 18, no. 4, pp. 1087-1096, 2011.
[18] K.-C. Lee, H.-P. Huang, and S.-S. Lu, "Burr detection by using vision image," The International Journal of Advanced Manufacturing Technology, vol. 8, no. 5, pp. 275-284, 1993.
[19] R. Sharan and G. C. Onwubolu, "Measurement of end-milling burr using image processing techniques," Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, vol. 225, no. 3, pp. 448-452, 2011.
[20] A. Tellaeche and R. Arana, "Robust 3D object model reconstruction and matching for complex automated deburring operations," Journal of Imaging, vol. 2, no. 1, p. 8, 2016.
[21] X. Chen, G. Shi, C. Xi, L. Zhong, X. Wei, and K. Zhang, "Design of Burr Detection Based on Image Processing," in Journal of Physics: Conference Series, 2019, vol. 1237, no. 3, p. 032075: IOP Publishing.
[22] J. Weng, P. Cohen, and M. Herniou, "Camera calibration with distortion models and accuracy evaluation," IEEE Transactions on Pattern Analysis & Machine Intelligence, no. 10, pp. 965-980, 1992.
[23] J. Kennedy and R. Eberhart, "Particle swarm optimization," in Proceedings of ICNN'95-International Conference on Neural Networks, 1995, vol. 4, pp. 1942-1948: IEEE.
[24] C. Loop and Z. Zhang, "Computing rectifying homographies for stereo vision," in Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149), 1999, vol. 1, pp. 125-131: IEEE.
[25] Y. Boykov, O. Veksler, and R. Zabih, "Fast approximate energy minimization via graph cuts," IEEE Transactions on pattern analysis and machine intelligence, vol. 23, no. 11, pp. 1222-1239, 2001.
[26] P. J. Besl and N. D. McKay, "Method for registration of 3-D shapes," in Sensor fusion IV: control paradigms and data structures, 1992, vol. 1611, pp. 586-606: International Society for Optics and Photonics.
[27] J. Denavit, "A kinematic notation for low pair mechanisms based on matrices," ASME J. Appl. Mech., vol. 22, pp. 215-221, 1955.
[28] M. Brandstötter, A. Angerer, and M. Hofbaur, "An analytical solution of the inverse kinematics problem of industrial serial manipulators with an ortho-parallel basis and a spherical wrist," in Proceedings of the Austrian Robotics Workshop, 2014, pp. 7-11.
[29] D. Gallup, J.-M. Frahm, P. Mordohai, and M. Pollefeys, "Variable baseline/resolution stereo," in 2008 IEEE Conference on Computer Vision and Pattern Recognition, 2008, pp. 1-8: IEEE.
[30] E. Olson, "AprilTag: A robust and flexible visual fiducial system," in 2011 IEEE International Conference on Robotics and Automation, 2011, pp. 3400-3407: IEEE.
[31] K. Simonyan and A. Zisserman, "Very deep convolutional networks for large-scale image recognition," arXiv preprint arXiv:1409.1556, 2014.