為了取得最佳的特徵點，在實驗中採用了測試影像的深度值與之間的梯度大 小作為評判標準。但此方法雖然簡單易懂，但也導致仍有特徵點無法有效地形成 最佳的三角形。在演算法技術成熟的今日，我們或許可以測試其他不同的評判擷 取節點的方式，來找出一個兼顧影像品質和時間效能的演算法。
特徵點個數對於 PSNR 的整體表現有相關的現象。也就是說，三角形的切割 數目越多對於 PSNR 上的表現越好。然而，三角形的切割數目若是過於精細則與 VSRS 2.0 的結果是差不多的。所以能夠有個好的評判準則判定整體畫面切割的 三角形個數是非常重要的研究議題。
傳統的任意視點合成(FTV)的演算法架構通常都是針對空間域上的一致性提 出改善的方法[8][9][16][17][28]，目前也有許多針對景深預估的演算法架構加強 了時間域上的一致性，且實驗結果有明顯的改善效果。然而，現有的場景還原技 術而言，並沒有太多演算法設計討論並改善時間域上的一致性。或許在視點合成 演算法日漸成熟的今後，能夠更好的演算法可以提出改善影像品質的新架構。
134
參考文獻
[1] D. Scharstein and R. Szeliski, “A taxonomy and evaluation of dense two-frame stereo correspondence algorithms,” International Journal of Computer Vision, vol. 47, pp. 7–42, Apr. 2002.
[2] Z. Zhang, “A Flexible New Technique for Camera Calibration,” IEEE Trans. on
Pattern Analysis and Machine Intelligence, Vol.22, pp.1330-1334, Nov., 2000.
[3] T. Fujii, T. Kimoto, M. Tonimoto, "A New Flexible Acquisition System of Ray-space Data for Arbitrary Objects," IEEE trans. on circuits and systems for
video technology (CSVT), vol. 10, no.2, pp.218-224, 2000.
[4] S. E. Chen and L. Williams, "View Interpolation for Image Synthesis," ACM
SIGGRAPH, pp. 279-288, 1996.
[5] M. Levoyand P. Hanrahan, "Light Field Rendering," ACM SIGGRAPH, pp.
31-42, 1996.
[6] S. J. Gortler, R. Grzesczuk, R. Szeliski, and M. F. Cohen, "The Lumigraph,"
ACM SIGGRAPH, pp. 43-54, 1996.
[7] R. Szeliski, "Video Mosaics for Virtual Environments," IEEE Computer Graphic
Applic., vol. 16, pp. 22-30, Mar., 1996.
[8] C. L. Zitnick, S. B. Kang, M. Uyttendaele, S. Winder, R. Szeliski, "High-quality Video View Interpolation using a Layered Representation," ACM SIGGRAPH, pp. 600-608, 2004.
[9] Y. Mori, N. Fukushima, T. Yendo, T. Fujii, and M. Tanimoto, "View Generation with 3D Warping using Depth Information for FTV," Signal Processing: Image
Communication , pp. 6572, 2009.
[10] D. Min, D. Kim, S. Yun, K. Shon, “2D/3D freeview video generation for 3DTV system, “ Signal Processing: Image Communication, vol. 24, pp. 31-48, 2009.
[11] W. Mark, L. Mcmillan, G. Bishop, "Post-rendering 3D Warping," Proc.
Symposium Interactive on 3D Graphics ,pp. 7-1, 1997.
[12] C. Fehn, “3D-TV using depth-image-based rendering (DIBR),” in Proc. Picture
Coding Symp., pp.307-312, Dec. 2004
.[13] M. Brown, and D. G. Lowe, “Automatic panoramic image stitching using invariant features,” International Journal of Computer Vision (IJCV),vol. 74(1), pp.59-73, 2007.
[14] M. Tanimoto, "Overview of Free Viewpoint Television,” Signal Processing:
Image Communication, pp. 454-461, 2006.
[15] T. Fujii, M. Tanimoto"Free-viewpoint Television Based on the Ray-Space Representation," Proc. SPIE : 3-D TV, Video, and Display, vol.4864, pp. 175-189, 2002.
[16] L. Yang, T. Yendo, M. P. Tehrani, T. Fujii, M. Tanimoto,” Artifact reduction using reliability reasoning for image generation of FTV,” Journal of Visual
Communication on Image Representation(JVCI), vol.21, Issues 5-6, pp.542-560,
Sep., 2010.
[17] N. Fukushima, T. Yendo, T. Fujii, M. Tanimoto, “Real-time Arbitrary View Interpolation and Rendering System using Ray-Space,” Proc. SPIE
Three-Dimensional TV. Video. and Display IV, vol. 6016, pp. 250-261, 2005.
[18] M. Tanimoto, "Overview of Free Viewpoint Television," Signal Processing:
Image Communication, pp. 454-461, 2006.
[19] M. Tonimoto, "Reference Software for Depth Estimation and View Synthesis,"
ISO/IEC/WG11/MPEG2008, M15377 , Apirl, 2008.
[20] M. Tonimoto, "1D Parallel Test Sequences for MPEG-FTV,"
ISO/IEC/WG11/MPEG2008, M15378, Apirl, 2008.
[21] M. Tonimoto, "Experimental Framework for 3D Video Coding,"
136
ISO/IEC/WG11/MPEG2010, W11477, July, 2010.
[22] I. Feldmann et al., "HHI Test Material for 3D Video,"
ISO/IEC/WG11/MPEG2010, M15413, July, 2008.
[23] O. Stankiewicz et al., "A soft–segmentation matching in Depth Estimation Reference Software (DERS) 5.0," ISO/IEC JTC1/SC 29/WG11, M17049, October, 2009.
[24] R. Hartley, A. Zisserman, Multiple View Geometry in Computer Vision, 2nd ed., Cambridge University Press, Mar., 2004.
[25] H. H. Baker, R. C. Bolles, “Epipolar-plane Image Analysis: A Technique for Analysis Motion Sequences," IEEE 3rd workshop Computer Vision:
Representation and Control, pp. 168-178, Oct. 1985.
[26] J. H. Park and H. W. Park, " Fast View Interpolation of Stereo Images using Image Gradient and Disparity Triangulation," Signal Processing : Image
Communication , vol. 18, pp. 401-416, 2003.
[27] J. H. Park, H. W. Park, "A Mesh-Based Disparity Representation Method for View Interpolation and Stereo Image Compression," IEEE Trans. on Image
Processing, vol. 15, no.7, July, 2006.
[28] B. Choi, T. Kim, K.-J. Oh, Y.-S. Ho, J.-S. Choi, "Intermediate View Synthesis Algorithm using Mesh Clustering for Rectangular Multiview Camera System,"
Proc. SPIE, vol.49, Feb., 2010.
[29] D. Farin, R. Peerlings, and H. N. Peter , “Depth-image Representation Employing Meshes for Intermediate-view Rendering and Coding,” IEEE 3DTV
Conference 2007, pp.1-4, May ,2007.
[30] A. Criminisi, P. Perez, K. Toyama, “Object Removal by Exemplar-based Inpainting,” IEEE Computer Society Conference on Computer Vision and
Pattern Recognition (CVPR), vol.2, pp.721-728, June, 2003.
自傳
李兆軒，1987 年 2 月 10 日出生於南投市，台中市人。2009 年 6 月畢業於國 立中興大學電機工程學系，之後於 2009 年 8 月攻讀國立交通大學電子研究所碩 士學位，承蒙杭學鳴教授的指導，進入通訊電子與訊號處理實驗室(CommLab)，
主要研究主題為利用攝影機陣列還原任意視點合成演算法設計。論文題目為「基 於三角轉換映射之任意視點合成演算法」。