以去背相連性為基礎之立體匹配研究

A Study on Matting Affinity Based Stereo Matching

宋秉修(Bing-Shiou Sung)

指導教授 ： 王聖智 ； 簡鳳村

國立交通大學/電機學院/電子工程系所/碩士(2012年)

https://doi.org/10.6842/NCTU.2012.01054

摘要

基於一些技術方面的需求，立體匹配已成為一個討論很久的議題而且現今也有許多既有的演算法，這些演算法大致上可以分成兩種種類：局部演算法和整體演算法，而每一類的演算法都有他們獨特的優缺點。其中，整體演算法不但能實現整體演算法的結果，整體演算法也可以採用一些實現局部演算法的方法。所以在我們的論文中採用了整體演算法的架構去建立一個系統來解立體匹配的問題，而這個系統同時能達到局部演算法跟整體演算法的優點。另一方面，我們也從影像去背的影像處理議題中引進了一個叫做去背相連性的函數到我們的整體演算法中，使我們希望能夠透過使用這個函數，只需要用到整體演算中常用到的兩個計算費用函數的項：資料項和平滑項，就可以達到跟一些現有的論文中外加了許多費用函數項一樣的結果。

關鍵字

去背 ； 相連性 ； 立體匹配

並列摘要

Stereo matching has been a long discussed issue for some technological needs, like 3D scene reconstruction. A large number of algorithms for stereo matching have been developed. Most of the algorithms can be categorized into two types: local and global algorithms. Each category has its own pros and cons. However, global algorithms can not only do the work of global algorithms but also can adopt some methods used in local algorithms, like cost aggregation. Therefore in this thesis, we construct a system based on a global-type algorithm to leverage between the advantages of both local and global algorithms. In addition, we introduce the matting affinity function which comes from image matting processing into the global algorithms of stereo matching in order to achieve a better result by using only traditional term costs: the data term cost and the smoothness term cost.

並列關鍵字

Stereo ； Matching ； Matting

參考文獻

[3] D. Scharstein and R. Szeliski. A taxonomy and evaluation of dense two-frame stereo correspondence algorithms. International Journal of Computer Vision, 47(1/2/3):7-42, April-June 2002

[4] X. Mei, X. Sun, M. Zhou, S. Jiao, H. Wang, and X. Zhang, “On building an accurate stereo matching system on graphics hardware,” International Conference on Computer Vision 2011.

[5] K. Zhang, J. Lu, and G. Lafruit, “Cross-based local stereo matching using orthogonal integral images,” IEEE Transactions on Circuits and Systems for Video Technology, 19(7):1073–1079, 2009.

[6] M. Bleyer, C. Rhemann, and C. Rother, “PatchMatch stereo - stereo matching with slanted support windows,” British Machine Vision Conference 2011

[7] A. Levin, D. Lischinski, Y. Weiss, “A Closed Form Solution to Natural Image Matting,” IEEE T. PAMI, vol. 30, no. 2, pp. 228-242, Feb. 2008.