臺灣博碩士論文加值系統

魚眼相機所提供超過180度的廣角特性，提供我們能在限有的硬體設備中，得到最多的影像資料量，在現今許多監控系統、車輛駕駛輔助系統、機器人電腦視覺當中廣為使用，然而魚眼鏡頭的特性造成影像扭曲現象與人眼視覺習慣相悖，造成視覺上的不適應，需要進行魚眼影像解扭曲(Fisheye Distortion Correction)，觀察傳統方法所提供的魚眼修正圖結果仍有邊緣比例過度拉升的現象，做為準確電腦視覺的距離測量會造成嚴重誤差，必須做更精確的魚眼相機校正，才能將魚眼影像解扭曲成正常相機透視角(Perspective Camera)的影像。
本研究對於傳統的魚眼相機校正法進行準確度的改善，依據拍攝實際世界的棋盤格校正板角點，觀察影像像素位置與實際校正板角點位置之間的關係，並且測量魚眼相機垂直與水平的FOV(Field of View)資訊，依據魚眼垂直與水平FOV範圍的差異性，加入魚眼模型的範圍調整。本文另外加入多項式分布調整聯合殘差(Polynomial Weight Joint Residual)的方式做曲線擬合(Curve Fitting)，以更準確的魚眼相機校正參數，獲得最終魚眼解扭曲的影像結果。實驗顯示，本研究可改善魚眼影像的扭曲失真，以及解扭曲圖邊緣拉升的現象，提供簡單快速的魚眼相機校正演算法與流程。

Due to the wide angle (over 180 degrees) characteristic of fisheye camera, it provides maximum image data in limited hardware resources. This comes convenient and is widely used in modern surveillance systems, driver auxiliary systems, and robot computer vision. However, fisheye camera introduces radial distortion which is irregular to normal human visual habits. This results in visual disorder and some might find it distracting. Fisheye distortion correction is essential in order to undo curve distortion to obtain normal perspective camera image. Traditional fisheye calibration still suffers from stretched boundary results of the undistorted fisheye image, which is abnormal and may cause severe error when executing metric measuring in computer vision. Therefore, a more accurate fisheye camera calibration is needed.
In this thesis we make improvement to traditional fisheye camera calibration methods. Chessboard corners are used as eigenpoints to calculate the relation between 2D image points and 3D world points. We also measure the horizontal and vertical FOV of the fisheye camera and alternate the range in the fisheye model. In addition, we perform a polynomial distribution weight in the joint residual while we do curve fitting refinement. With our fisheye camera calibration method, we can attain better camera parameters and produce a more accurate fisheye undistortion image. Experiment results shows that our method can correct radial distortion and stretched boundary in fisheye images. Our approach is easy to use and we improve the overall accuracy for fisheye camera calibration.

摘　要 i
ABSTRACT ii
目　錄 iii
表目錄 v
圖目錄 vi
第一章 緒論 1
1.1 研究動機與目的 1
1.2 研究方法 3
1.3 論文組織架構 5
第二章 相關文獻回顧 6
2.1 非度量校正法 7
2.1.1 共線點的平行性與垂直性 7
2.1.2 FOV 模型 8
2.2 度量校正法 9
2.2.1 3D物體依據的校正法 9
2.2.2 純旋轉校正法 11
2.2.3 2D平面依據的校正法 13
第三章 提出方法 17
3.1 魚眼數學模型 19
3.1.1 定義魚眼數學模型 19
3.1.2 考慮h.FOV與v.FOV範圍調整的魚眼模型 21
3.2 魚眼相機校正 23
3.2.1 相機的內部參數與外部參數估測 23
3.2.2 相機曲線擬合與精化 26
3.3 魚眼影像扭曲修正 30
3.3.1 製作相機透視角查找表 30
3.3.2 魚眼影像解扭曲 30
第四章 實驗結果與討論 32
4.1 實驗環境 32
4.2 評估標準 33
4.3 提出方法分析 35
4.3.1 加入h.FOV與v.FOV範圍調整的魚眼模型 35
4.3.2 加入多項式權重做曲線擬合修正 37
4.4 提出方法比較 40
4.4.1 提出方法曲線擬合分析 40
4.4.2 提出方法與文獻比較 41
4.5 不同魚眼相機硬體解扭曲圖 44
4.6 運算時間分析 50
第五章 結論 52
參考文獻 53

[1]Berto Autonomous Valet Parking and Charging for e-Mobility (V-Charge) Project, http://www.volkswagenag.com/content/vwcorp/info_center/en/news/2015/07/charge.html
[2]M. Zzi, “360°Detection and tracking algorithm of both pedestrian and vehicle using fisheye images,” IEEE Intelligent Vehicles Symposium, June 2015.
[3]C.F. Hsu, “Automatic Parking Based on Bird’s –eye View Image,” Dept. of Comput. Sci. National Taiwan Normal University Univ, 2015.
[4]Y.T. Chen, “Obstacle Detection in A Wide-scope Top-view Monitoring System,” Dept. of Comput. Sci. National Central University Univ, 2015.
[5]Volvo BLIS (Blind Spot Information System) Operation page, http://support.volvocars.com/uk/cars/Pages/owners-manual.aspx?mc=Y555&my=2015&sw=14w46&article=4da5e00011e70e1bc0a801e800212255
[6]F. Devernay, O. Faugeras, “Straight lines have to be straight: automatic calibration and removal of distortion from scenes of structured enviroments,” Machine Vision and Applications, vol 13, pp.14-24, 2001.
[7]K. Kanatani, “Calibration of Ultrawide Fisheye Lens Cameras by Eigenvalue Minimization”, Dept. of Comput. Sci. Okayama Univ, Feb 2013.
[8]J. Kannala, S. Sami, “A Generic Camera Model and Calibration Method for Conventional, Wide-Angle, and Fish-Eye Lenses,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 28, no. 8, Aug 2006.
[9]M. Glavin, “Accuracy of fish-eye lens models,” Connaught Automotive Research Group, Dept. of Electric Engineering. National Univ. of Ireland, June 2010.
[10]O. Faugeras, “Three-Dimensional Computer Vision a Geometric Viewpoint,” MIT Press, 1993.
[11]M. Pollefeys, “3D Modelling from Images,” ECCV Tutorial notes, June 2000.
[12]P. Debevec, “Reconstructing and Augmenting Architecture with Image-Based Modelling, Rendering and Lighthing,” VAA International Symposium on Virtual Architecture, vol. 21-22, pp. 1-10, June 2001.
[13]A.W. Fitzgibbon, A. Zisserman, “Automatic 3D Model Acquisition and Generation of New Images from Video Sequences,” EUSIPCO European Signal Processing conference, pp. 1261-1269, 1998.
[14]R. Hartley, “Self-Calibration from Multiple Views with a Rotating Camera,” Proc. Third European Conf. Computer Vision, J.-O. Eklundh, ed., vol. 800-801, pp. 471-478, May 1994
[15]G.P.Stein, “Accurate Internal Camera Calibration using Rotation, with Analysis Sources of Error,” IEEE Fifth International Conference on Computer Vision Proceedings, 1995.
[16]Z. Zhang, “Camera calibration with one-dimensional objects,” IEEE Pattern Analysis and Machine Intelligence, vol. 26 Jul 2004
[17]R.Y. Tsai, “A Versatile Camera Calibration Technique for High-Accuracy 3D Machine Vision Metrology Using Off-the-Shelf TV Cameras and Lenses,” IEEE Journal of Robotics and Automation, vol. 3, no. 4, pp. 323-344, Aug 1986
[18]R. K. Lenz, R. Y. Tsai, “Techniques for calibration of the scale factor and image center for high accuracy 3-D machine vision metrology,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 10, no. 5, pp. 713-720, 1988.
[19]R. G. Willson, “Modeling and Calibration of Automated Zoom Lenses,” Ph.D. thesis, Dept. of Electrical and Computer Engineering, Carnegie Mellon Univ, January 1994.
http://www.cs.cmu.edu/afs/cs.cmu.edu/user/rgw/www/TsaiCode.html
[20]J. Heikkila, O. Silven, “A Four-Step Camera Calibration Procedure with Implicit Image Correction,” IEEE Computer Vision and Pattern Recognition, pp. 1106-1112, 1997.
[21]Z. Zhang, “Flexible Camera Calibration By Viewing a Plane From Unknown Orientations,” IEEE International Conference on Computer Vision, vol.1, pp.666-673, Jul 1999
[22]Z. Zhang, “A Flexible New Technique for Camera Calibration”, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol.22, no.11, 2000.
[23]OpenCV 2.4.13.0 documentation, “OpenCV Tutorials, calib3d module. Camera calibration and 3D reconstruction,” http://docs.opencv.org/2.4/doc/tutorials/calib3d/camera_calibration/camera_calibration.html
[24]R. Kumar and A. Hanson. “Robust Methods for Estimating Pose and a Sensitivity Analysis,” CVGIP-Image Understanding, vol. 60, no. 3, pp. 313-342, 1994.
[25]B. Micušık, “Two-view geometry of omnidirectional cameras,” Ph.D. thesis, Czech Technical University, 2004.
[26]D. Scaramuzza, A. Martinelli, R. Siegwart, “A flexible technique for accurate omnidirectional camera calibration and structure from motion,” IEEE International Conference on Computer Vision Systems, pp. 45-45, 2006.
[27]D. Scaramuzza, A. Martinelli, R. Siegwart, “A toolbox for easily calibrating omnidirectional cameras,” IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 5695–5701, 2009.
[28]D. Scaramuzza, R. Siegwart, “Automatic Detection of Checkerboards on Blurred and Distorted Images,” IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 3121 – 3126, 2008.
[29]D. Scaramuzza, “OCamCalib: Omnidirectional Camera Calibration Toolbox for Matlab,” https://sites.google.com/site/scarabotix/ocamcalib-toolbox, August 2015.
[30]S. Urban, J. Leitloff, S. Hinz, “Improved wide-angle, fisheye and omnidirectional camera calibration,” ISPRS Journal of Photogrammetry and Remote Sensing, vol. 108, pp. 72–79, 2015.
[31]Chen-Feng Li, Shi-Min Hu, R. Ralph “Fisheye Video Correction,” IEEE Visualization and Computer Graphics, vol. 18, no. 10, Aug. 2012.
[32]Jie-Shou Lu et al, “A novel calibration method based on heuristic B-spline model for fish-eye lenses,” IEEE International Conference on System Science and Engineering, 2013.
[33]M. Bronstein, “Solving Polynomial Equations: Foundations, Algorithms, and Applications,” Springer, vol. 14, 2005.
[34]M. Hazewinkel, “Normal distribution”, Encyclopedia of Mathematics, 2001.
[35]M.Humphrys, “Continuous output, the sigmoid function,” School of Computing. Dublin City University, 1987.
[36]M. Zoltowski, “Equations for the Raised Cosine and Square-Root Raised Cosine Shapes,” Communication Systems division, Linköping University, 2013.
[37]J. Charmant, Estimating a camera’s field of view page, http://xn--1-2fa.fr/?tag=fov