跳到主要內容

臺灣博碩士論文加值系統

(3.235.174.99) 您好!臺灣時間:2021/07/24 20:20
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

我願授權國圖
: 
twitterline
研究生:林群展
研究生(外文):Lloyd
論文名稱:使用單應性做基於影像的繪圖
論文名稱(外文):Image-Based Rendering Using Homography
指導教授:唐政元
學位類別:碩士
校院名稱:華梵大學
系所名稱:資訊管理學系碩士班
學門:電算機學門
學類:電算機一般學類
論文種類:學術論文
論文出版年:2007
畢業學年度:95
語文別:中文
論文頁數:43
中文關鍵詞:單應性基礎矩陣區塊分割區塊合併
外文關鍵詞:HomographyFundamental matrix
相關次數:
  • 被引用被引用:4
  • 點閱點閱:329
  • 評分評分:
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:1
本篇論文的目的在於根據不同視角的影像,合成出視角間的影像。我們把它分成兩個部份:一個是利用兩張影像找對應點的問題,另一個是透過對應點的取得產生出視角間的影像。
首先,找對應點的部份,使用SIFT(Scale Invariant Feature Transform)找出為數不多的對應點當作種子(Seed),我們使用Least Median of Squares (LMedS)來挑選、計算來獲得一個較好的基礎矩陣,再透過所估測出的基礎矩陣(Fundamental matrix)來刪除不好的對應點。接著使用ZNCC與propagation產生出密集的對應點,然而,密集對應點的部份仍然存在著一些問題。因此,再透過單應性與基礎矩陣的幾何關係來刪除不好的對應點。
接著,由於有了密集的對應關係,我們使用LMedS去估測出兩個8*8大小的區塊的單應性矩陣(Homography matrix)。我們提出了區塊的分割與合併的方法來找出不同視差區域的邊緣線。使用基於影像的繪圖(IBR)的方法來產生出兩個視角間的影像,本論文將展示一些初步的結果。

關鍵字:單應性、基礎矩陣、基於影像的繪圖(IBR)、區塊分割、區塊合併
The purpose of this thesis is to synthesize new images of scenes from novel viewpoints. There are two parts in this paper. The first part is to find the corresponding point pairs from two images. The second one is to generate new images from novel viewpoints using the corresponding point pairs obtained.
In the first part, the Scale-Invariant Feature Transform (SIFT) is used to find a few corresponding point pairs as seeds. A robust estimation, Least Median of Squares (LMedS) is used to estimate the fundamental matrix. The estimated fundamental matrix is used to delete the wrong corresponding point pairs. And then the propagation procedure by the Zero-mean Normalized Cross-Correlation (ZNCC) is used to obtain quasi-dense corresponding point pairs. However, there still exist some wrong corresponding point pairs. Therefore, the homography and the fundamental matrix are used to delete the wrong corresponding point pairs.
Given quasi-dense corresponding point pairs, in the second part, the robust method, least median of squares (LMedS) is used to estimate the homography matrix within two small patches size of 8x8 pixels. The methods of patch split and patch merge are proposed to locate the regions of disparity discontinuities. The Image-Based Rendering (IBR) method is used to generate new views of scenes from novel viewpoints. Some preliminary results are shown in this paper.

Keywords: Homography、Fundamental matrix、IBR (Image-Based Rendering), Patch Split, Patch Merge
誌 謝 I
摘要 II
ABSTRACT III
目錄 V
一、緒論 - 1 -
1.1研究背景與動機 - 1 -
1.2研究目的 - 1 -
1.3論文架構 - 2 -
二、文獻探討 - 3 -
2.1 對應點相關研究 - 3 -
2.2 基於影像的繪圖 - 4 -
2.3 投影幾何關係 - 5 -
三、研究方法 - 8 -
3.1 建立QUASI-DENSE DISPARITY MAP - 8 -
3.1.1較好對應點之取得(Seed selection) - 8 -
3.1.2 Propagation - 9 -
3.1.3 刪除錯誤對應點 - 11 -
3.2 區塊的分割與合併 - 12 -
3.2.1 區塊的分割 - 12 -
3.2.2區塊的合併 - 21 -
3.3 基於影像的繪圖 - 23 -
四、實驗結果 - 24 -
4.1 較好對應點之取得(SEED SELECTION) - 24 -
4.2 區塊的分割及合併 - 31 -
4.2.1 用disparity做區塊的分割 - 33 -
4.2.2 用Homography做區塊的分割 - 37 -
4.2.3用disparity做區塊的合併 - 38 -
4.3 基於影像的繪圖 - 39 -
五、結論與未來方向 - 41 -
參考文獻 - 42 -
[1]T. Beier, S. Neely. “Feature- Based Image Metamorphosis,” ACM SIGGRAPH. Vol. 26, pp. 35-42 Jul. 1992.
[2]A. Barla, F. Odone and A. Verri. “Hausdorff Kernel for 3D Object Acquisition and Detection,” In European Conference on Computer Vision, Vol. 2353. pp. 20-33. May. 2002.
[3]M. Z. Brown, D. Burschka, “Advances in Computational Stereo,” IEEE Trans. Pattern Analysis and Machine Intelligence, Vol. 25, No. 8, pp. 993-1008. Aug. 2003.
[4]M. P. Dubuisson, A. K. Jain. “A Modified Hausdorff Distance for Object Matching,” Proc. of IAPR Int. Conf. on Pattern Recognition, Vol. 1 pp. 566-568. 1994.
[5]B. Georgescu, P. Meer, “Point Matching under Large Image Deformations and Illumination Changes,” IEEE Trans. Pattern Analysis and Machine Intelligence, Vol. 26, No. 6, pp. 674-688. June 2004.
[6]S. J. Gortler, R. Grzeszczuk, R. Szeliski, and M. F. Cohen, “The Lumigraph,” SIGGRAPH'96, pp. 43-54, 1996.
[7]C. Harris and M. Stephens, “A Combined Corner and Edge Detector,” Alvey Vision Conference, pp. 147–152. 1988.
[8]R. Hartley and A. Zisserman, Multiple View Geometry in Computer Vision, Cambridge University Press. 2003.
[9]H. Huang, S. Nain, Y. Hung and T. Cheng, “Disparity-Based View Morphing-A New Technique for Image-Based Rendering”, Proceedings of the ACM Symposium on Virtual reality software and technology, page 9-16. 1998.
[10]D. Huttenlocher, G. Klanderman and W. Rucklidge, “Comparing Images Using the Hausdorff Distance, ” IEEE Trans. Pattern Analysis and Machine Intelligence, Vol. 15, No. 9, pp. 850-863. Sep. 1993.
[11]M. Lhuillier and L. Quan, “Match Propagation for Image-Based Modeling and Rendering,” IEEE Transcations on Circuits and Systems for Video Technology, Vol. 24, No. 8, pp. 1140-1146. Aug. 2002.
[12]M. Lhuillier and L. Quan, “Image-Based Rendering by Joint View Triangulation,” IEEE Transcations on Circuits and Systems for Video Technology, Vol.13, No.11, pp. 1051-1063. Nov. 2003.
[13]M. Lhuillier, “Towards Automatic Interpolation for Real and Distant Image Pairs,” INRIA technical report RR-3619, Feb. 1999.
[14]D.G. Lowe, “Distinctive Image Features from Scale-Invariant Keypoints,” International Journal of Computer Vision, Vol.60, pp.91-110. Nov. 2004.
[15]M. Okutomi and T. Kanade, “A Multiple-Baseline Stereo,” IEEE Trans. Pattern Analysis and Machine Intelligence, Vol. 15, No. 4, pp. 353-363. Apr. 1993.
[16]F. Odone, A. Barla and A. Verri, “Building Kernels From Binary Strings for Image Matching,” IEEE Trans. Image Processing, Vol. 14, No. 2, pp. 169-180. Feb. 2005.
[17]C. Y. Tang, H. L. Chou, Y. H. Yeh, M. W. Lin, W. H. Kuo and Y. N. Liao, “Kernel-based Stereo Correspondence for Color Images,” CVGIP, 2004.
[18]C. Zhang and T. Chen, “Active Rearranged Capturing of Image-Based Rendering Scenes-Theory and Practice,” IEEE Trans. Multimedia, Vol. 9 Issue.3. April. 2007.
[19]洪莘逸,「使用多視角影像合成新視點影像」,華梵大學資訊管理學系,碩士論文,九十五學年度。
[20]王文宏,「特徵擷取與特徵點對應之研究與應用」,華梵大學資訊管理學系,碩士論文,九十五學年度。
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top