臺灣博碩士論文加值系統

這篇論文的目標是發展一個使用影像縫合技術於相機校正設計，加快相機校正速度的演算法。近年來因為相機校正技術需要輔助校正物且機器的精確性誤差的關係，因此影像縫合技術在相機校正領域佔有越來越重要的地位。而在影像縫合的應用上，考慮縫合速度和影像品質的權衡是件非常重要的事。傳統的演算法由於為了達到夠好的影像品質，因此消耗了相當大的運算量，也因為如此，傳統的演算法不適用於硬體的實作。因此本篇論文提出一個快速的演算法，在擷取特徵點的時候由內外往外擴散慢慢刪除，並且使用適應性的方法讓擷取特徵點的步驟停止。接著我們在RANSAC的部份也使用特殊的技巧來擷取matching points做RANSAC，使擷取到的matching points容易呈現兩極化的現象，如此就不需要做大量的RANSAC，此兩個步驟都大大地節省CPU time，加快速度但卻可以得到一樣的影像品質。最後我們也發現在多台Sensor的角度、位置、距離都已經固定的前提下，其影像彼此之間的轉換關係是固定的，因此我們只要跑過一次完整的演算法即可求出彼此之間的轉換關係，所以之後的校正只要透過其固定的轉換關係即可校正，如此的發現非常有利於我們在硬體實現的突破，因此校正的速度將可以大幅地改善。

第一章 簡介..................................................................................................................1
1-1 相機校正(Camera Calibration)與影像縫合(AutoStitch)的應用....1
1-2 相機校正(Camera Calibration)與影像縫合(AutoStitch)演的缺點2
1-3 影像縫合與校正在效能上的考量............................................................5
1-4 章節規劃....................................................................................................6
第二章 背景...........................................................................................................7
2-1 影像縫合流程(The Flow of AutoStitch)............................................7
2-2 特徵點擷取(Feature Extraction)........................................................7
2-2.1 特徵點定義........................................................................................7
2-2.2 ㄧ般Harris Corner取法..................................................................9
2-2.3 適應性非最大化壓縮法(Adaptive Non-Maximal Suppression)......................................................................................................................14
2-3 特徵點匹配(Feature Matching)..........................................................17
2-3.1 最小平方差SSD(Sum of Squared Differences)........................17
2-3.2 次近鄰居法(Second-Closest Neighbor)....................................18
2-4 剔除掉錯誤的匹配點(Delete The False Match)..............................22
2-5 歪斜影像(Warp Image)..........................................................................26
2-6 混合影像(Blending)..............................................................................28
第三章 使用BUMS與PRSC影像縫合快速演算法設計........................................33
3-1 Bottom-Up Maxima Selection(BUMS)..................................................33
3-1.1 傳統適應性非最大化壓縮法之深入討論......................................33
3-1.2 BUMS流程(The flow of Bottom-Up Maxima Selection)......39
3-1.3 特徵點均勻分佈特性驗證(Verification of the Uniform Distribution of the Feature Points)................................................45
3-2 極化型隨機取樣程序(Polarized Random Sample Consensus)........48
3-2.1 RANSAC執行次數之討論(Discussion of RANSAC Times)..........48
3-2.2極化型隨機取樣程序流程(The Flow of Polarized Random Sample Consensus)..................................................................................................51
第四章 模擬結果之探討與比較.........................................................................57
4-1 BUMS運算效率模擬(Simulation of Computational Efficiency for Bottom-Up Maxima Selection)...........................................................57
4-2 PRSC運算效率模擬(Simulation of Computational Efficiency for Polarized Random Sample Consensus).............................................60
4-3 整體運算效率模擬(Simulation of Total Computational Efficiency for the Algorithm).............................................................................62
4-4 轉換矩陣H的重覆使用性質(Reusability of Homography H)..........64
第五章 結論與未來發展.....................................................................................68
參考文獻：..................................................................................................................71

[1] C. Rother and S. Carlsson. Linear multi view reconstruction and camera recovery using a reference plane. International Journal of Computer Vision, 49(2/3):117–141, 2002.
[2] B. Bascle, A. Blake, and A. Zisserman. Motion deblurring and super-resolution from and image sequence. In Proceedings of the 4th European Conference on Computer Vision (ECCV96), pages 312–320. Springer-Verlag, 1996.
[3] D. Capel and A. Zisserman. Automated mosaicing with super-resolution zoom. In Proceedings of the Interational Conference on Computer Vision and Pattern Recognition (CVPR98), pages 885–891, June 1998.
[4] J. Davis. Mosaics of scenes with moving objects. In Proceedings of the Interational Conference on Computer Vision and Pattern Recognition (CVPR98), pages 354–360, 1998.
[5] M. Uyttendaele, A. Eden, and R. Szeliski. Eliminating ghosting and exposure artifacts in image mosaics. In Proceedings of the Interational Conference on Computer Vision and Pattern Recognition (CVPR01), volume 2, pages 509–516, Kauai, Hawaii, December 2001.
[6] A. Agarwala, M. Dontcheva, M. Agarwala, S. Drucker, A. Colburn, B. Curless, D. Salesin, and M. Cohen. Interactive digital photomontage. In ACM Transactions on Graphics (SIGGRAPH’ 04), 2004.
[7] D. Brown. Close-range camera calibration. Photogrammetric Engineering, 37(8):855–866, 1971.
[8] D.B. Goldman and J.H. Chen. Vignette and exposurecalibation and compensation. In Proceedings of the 10th International Conference on Computer Vision (ICCV05), pages I: 899–906, 2005.
[9] P. Debevec and J. Malik. Recovering high dynamic range radiance maps from [1] C. Rother and S. Carlsson. Linear multi view reconstruction and camera recovery using a reference plane. International Journal of Computer Vision, 49(2/3):117–141, 2002.
[2] B. Bascle, A. Blake, and A. Zisserman. Motion deblurring and super-resolution from and image sequence. In Proceedings of the 4th European Conference on Computer Vision (ECCV96), pages 312–320. Springer-Verlag, 1996.
[3] D. Capel and A. Zisserman. Automated mosaicing with super-resolution zoom. In Proceedings of the Interational Conference on Computer Vision and Pattern Recognition (CVPR98), pages 885–891, June 1998.
[4] J. Davis. Mosaics of scenes with moving objects. In Proceedings of the Interational Conference on Computer Vision and Pattern Recognition (CVPR98), pages 354–360, 1998.
[5] M. Uyttendaele, A. Eden, and R. Szeliski. Eliminating ghosting and exposure artifacts in image mosaics. In Proceedings of the Interational Conference on Computer Vision and Pattern Recognition (CVPR01), volume 2, pages 509–516, Kauai, Hawaii, December 2001.
[6] A. Agarwala, M. Dontcheva, M. Agarwala, S. Drucker, A. Colburn, B. Curless, D. Salesin, and M. Cohen. Interactive digital photomontage. In ACM Transactions on Graphics (SIGGRAPH’ 04), 2004.
[7] D. Brown. Close-range camera calibration. Photogrammetric Engineering, 37(8):855–866, 1971.
[8] D.B. Goldman and J.H. Chen. Vignette and exposurecalibation and compensation. In Proceedings of the 10th International Conference on Computer Vision (ICCV05), pages I: 899–906, 2005.
[9] P. Debevec and J. Malik. Recovering high dynamic range radiance maps from photographs. Computer Graphics, 31:369–378, 1997.
[10] H. Seetzen, W. Heidrich, W. Stuerzlinger, G. Ward, L. Whitehead, M. Trentacoste, A. Ghosh, and A. Vorozcovs. High dynamic range display systems. In ACM Transactions on Graphics (SIGGRAPH’04), 2004.
[11] R. Szeliski and S. Kang. Direct methods for visual scene reconstruction. In IEEE Workshop on Representations of Visual Scenes, pages 26–33, Cambridge, MA, 1995.
[12] M. Irani and P. Anandan. About direct methods. In B. Triggs, A. Zisserman, and R. Szeliski, editors, Vision Algorithms: Theory and Practice, number 1883 in LNCS, pages 267–277. Springer-Verlag, Corfu, Greece, September 1999.
[13] H. Sawhney and R. Kumar. True multi-image alignment and its application to mosaicing and lens distortion correction. IEEE Transactions on Pattern Analysis and Machine Intelligence, 21(3):235–243, 1999.
[14] H. Shum and R. Szeliski. Construction of panoramic mosaics with global and local alignment.International Journal of Computer Vision,36(2):101–130, February 2000.
[15] I. Zoghlami, O. Faugeras, and R. Deriche. Using geometric corners to build a 2D mosaic from a set of images. In Proceedings of the International Conference on Computer Vision and Pattern Recognition, Puerto Rico. IEEE, June 1997.
[16] P. McLauchlan and A. Jaenicke. Image mosaicing using sequential bundle adjustment. Image and Vision Computing, 20(9-10):751–759, August 2002.
[17] R. Szeliski. Image Alignment and Stitching: A Tutorial1 Preliminary draft, September 27, 2004 Technical Report.MSR-TR-2004-92.
[18] Chen, S. E. QuickTime VR – an image-based approach to virtual environment navigation. Computer Graphics (SIGGRAPH’95), , 29–38. 1995.
[19] Szeliski, R. Image mosaicing for tele-reality applications. In IEEE Workshop on Applications of Computer Vision (WACV’94), pages 44–53, IEEE Computer Society, Sarasota. 1994.
[20] Szeliski, R. Creating Full View Panoramic Image Mosaics and Environment Maps. Microsoft Research.
[21] irahu http://www.irahul.com/workzone/stitch/index.php
[22] P. Heckbert. Projective Mappings for Image WarpingRendering, 15-862 .26 Sept 1995.
[23] C. Harris and M.J. Stephens. A combined corner and edge detector. In Alvey Vision Conference, pages 147–152, 1988.
[24] H. Moravec. Obstacle avoidance and navigation in the real world by a seeing robot rover. Technical Report CMU-RI-TR-3, Carnegie-Mellon University, Robotics Institute, 1980.
[25] C. Schmid, R. Mohr, and C. Bauckhage. Evaluation of interest point detectors. International Journal of Computer Vision, 37(2):151–172, June 2000.
[26] G. Derpanis. The Harris Corner Detector, October 27, 2004.
[27] M. Brown, R. Szeliski, and S. Winder. Multiimage matching using multi-scale oriented patches. In Proceedings of the Interational Conference on Computer Vision and Pattern Recognition (CVPR05), San Diego, June 2005.
[28] D. Lowe. Distinctive image features from scale-invariant keypoints. International Journal of Computer Vision, 60(2):91–110, 2004.
[29] J. Shi and C. Tomasi. Good features to track. In Proceedings of the Interational Conference on Computer Vision and Pattern Recognition (CVPR94), Seattle, June 1994.
[30] Y. Ke1, R. Sukthankar. PCA-SIFT: A More Distinctive Representation for Local Image Descriptors. Intel Research Pittsburgh In Proceedings of the Interational Conference on Computer Vision and Pattern Recognition(CVPR04).
[31] D. Lowe, "Object recognition from local scale-invariant features," International Conference on Computer Vision, Corfu, Greece (September 1999), pp. 1150-1157.
[32] D. Lowe, "Local feature view clustering for 3D object recognition," IEEE Conference on Computer Vision and Pattern Recognition, Kauai, Hawaii (December 2001), pp. 682-688.
[33] M. Fischler and R. Bolles. Random sample consensus: A paradigm for model fitting with application to image analysis and automated cartography. Communications of the ACM, 24:381–395, 1981.
[34] P. Torr. Bayesian model estimation and selection for epipolar geometry and generic manifold fitting. International Journal of Computer Vision, 50(1):35–61, 2002.
[35] R. Hartley and A. Zisserman. Multiple View Geometry in Computer Vision. Cambridge University Press, ISBN: 0521540518, second edition, 2004.
[36] M. Brown, D. Lowe. Automatic Panoramic Image Stitching using Invariant Features. International Journal of Computer Vision(ijcv),2007.
[37] M. Brown and D. Lowe. Recognising panoramas. In Proceedings of the 9th International Conference on Computer Vision (ICCV03), volume 2, pages 1218–1225, Nice, October 2003.
[38] J. Beis and D. Lowe. Shape indexing using approximate nearest-neighbor search in highdimensional spaces. In Proceedings of the Interational Conference on Computer Vision and Pattern Recognition (CVPR97), pages 1000–1006, 1997.
[39] P. Burt and E. Adelson. A multiresolution spline with application to image mosaics. ACM Transactions on Graphics, 2(4):217–236, 1983.
[40] ROGER Y. TSAI. A Versatile Camera Calibration Technique for 3D Machine Vision. IEEE J. Robotics & Automation, RA-3, No. 4, August 1987, pp. 323-344.
[41] Zhengyou Zhang. A Flexible New Technique for Camera Calibration. December 2, 1998. Technical Report.MSR-TR-98-71.
[42] http://mpac.ee.ntu.edu.tw/~ck/project_panorama/#SIFTTutorial
[43] http://www.csie.ntu.edu.tw/~cyy/courses/vfx/05spring/assignments/proj2/
[44] B. Wilbum and N. Joshi. High Performance Imaging Using Large Camera Arrays.In ACM Transactions on Graphics, Vol 24, No 3, July 2005, pp. 765-776.