在這篇論文中我們提出了一個新穎的深度推估演算法應用於兩台水平架設的攝影機所拍攝之立體序列影像。以往立體匹配演算法針對單張的左右影像找出正確的深度圖，但若對於每張新畫格都用此演算法重新計算，由於物體移動使計算時參考的周圍資訊不同、且有遮蔽的情況發生，讓畫面容易出現跳動與不一致，在使用者的舒適度上將產生重大的影響。我們認為立體電視技術成果要好，深度圖前後的穩定性是一個關鍵的因素，本篇論文將根據此重點提出新的概念:我們考慮畫面內容的運動量、時間軸上的關係、空間上的關係來推估新的深度圖。

此演算法包括三個部分，第一部分中，為了同時考慮畫面中相似的點，我們使用四分樹的結構把畫面分為一塊塊相似的區塊，再利用雙向濾波器與對比上下文直方圖 (Contrast Context Histogram) 演算法給予每個區塊在下個時間點的粗略深度;第二部分將運動量、前後畫格的關係、左右對應的關係定義為機率公式，進而推算修正這個區塊最有可能的深度。第三部分為精煉算出的深度圖。

In this paper, we propose a novel algorithm to propagate depth map using on stereo video. Stereo matching algorithm used to calculate the depth value according to the clues given by the left and right frames at the same time. However, if we recalculate every new frame by these algorithms, due to the movement of the objects, surrounding information will be different and some objects may be covered by others when calculating, the output of depth map will easily vibrate without coincidence. What mentioned above will generate big influences and bring the users trouble. If we want better outcomes, the stability of depth maps is a crucial point. This paper will focus on this issue and bring out a new notion, considering the motion of objects, the frames in different time, and the relationship between left and right frames to propagate a new depth map.

This algorithm include three parts. First, to consider the similar pixels at the same time, we use Quadtree to divide the frame into many blocks, and give the rough depth value of the blocks through CCH and Bilateral Filter. Second, using moving value, the relationships between different times and correspondence of left and right to modify the most possible depth value. The last one, to refine the depth map.

[1] Liang Wang, Mingwei Gong, Minglun Gong and Ruigang Yang, “How Far Can We Go with Local Optimization in Real-Time Stereo Matching,” IEEE International Symposium on 3D Data Processing, Visualization, and Transmission (3DPVT 2006), pp.129-136, 2006.

[2] Ke Zhang, Jiangbo Lu, G. Lafruit, “Cross-based local stereo matching using orthogonal integral images,” IEEE Transactions on Circuits and Systems for Video Technology, pp.1073-1079, 2009.

[3] A. Hosni, M. Bleyer, M. Gelautz, and C. Rhemann, “Local stereo matching using geodesic support weights,” IEEE International Conference on Image Processing, pp.2093-2096, 2009.

[4] Dengfeng Chai, Qunsheng Peng, “Bilayer Stereo Matching,” IEEE International Conference on Computer Vision, pp.1-8, 2007.

[5] Jong Dae Oh, Siwei Ma and C.-C.J. Kuo, “Stereo Matching via Disparity Estimation and Surface Modeling,” IEEE Conference on Computer Vision and Pattern Recognition, pp.1-8, 2007.

[6] Yi Deng, Qiong Yang, Xueyin Lin and Xiaoou Tang, “Stereo Correspondence with Occlusion Handling in a Symmetric Patch-Based Graph-Cuts Model,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol.29 (6) pp.1068 - 1079, 2007.

[7] Li Hong and George Chen, “Segment-based stereo matching using graph cuts,” IEEE Computer Society Conference on Computer Vision and Pattern Recognition, pp.I-74-I-81 vol.1, 2004.

[8] Yuri Boykov, Olga Veksler and Ramin Zabih, “Fast approximate energy minimization via graph cuts,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 23 (11) pp.1222-1239, 2001.

[9] QingXiong Yang, Liang Wang, Ruigang Yang, H Stewenius and David Nister, “Stereo Matching with Color-Weighted Correlation, Hierarchical Belief Propagation, and Occlusion Handling,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol.31 (3) pp.492 - 504, 2009.

[10] Ping Li, Rene Klein Gunnewiek, “On Creating Depth Maps from Monoscopic Video using Structure from Motion,” Proc. of IEEE Workshop on Content Generation and Coding for 3D-television, pp.508-515, 2006.

[11] IanirIdeses, Leonid Yaroslavsky, Barak Fishbain, “Depth Map Manipulation for 3D Visualization,” 3DTV Conference: The True Vision - Capture, Transmission and Display of 3D Video, pp.337-340, 2008

[12] S. Battiato, S. Curti, M. La Cascia, E. Scordato, and M. Tortora, “Depth Map Generation By Image Classification,” Proc. of SPIE IS&T/SPIE's 16th Annual Symposium on Electronic Imaging, pp.95-104, 2004.

[13] Ludovic J. Angot, Wei-Jia Huang and Kai-Che Liu, “A 2D to 3D video and image conversion technique based on a bilateral filter,” Proc. of SPIE-IS&T Electronic Imaging, Vol. 7526, doi:10.1117/12.838571, 2010

[14] Chao-Chung Cheng; Chung-Te Li; Liang-Gee Chen, “A Novel 2D-to-3D conversion system using edge information,” Int’l Conf. on Consumer Electronics (ICCE), pp.377-378, 2010.

[15] P. Harman, J. Flack, S. Fox and M. Dowley, “Rapid 2D to 3D Conversion,” Proc. SPIE, Vol. 4660, pp.78-86, 2002.

[16] Chun-Rong Huang , Chu-Song Chen, and Pau-Choo Chung, “Contrast Context Histogram – A Discriminating Local Descriptor for Image Matching,” In Proceedings of 18 th IAPR International Conference on Pattern Recognition, ICPR'06, vol.4, pp. 53-56, Hong Kong, Aug., 2006.

[17] C. Varekamp, B. Barenbrug, “Improved depth propagation for 2D to 3D videoconversion using key-frames,” Proc. of 4th European Conference on VisualMedia Production (IETCVMP), pp.1 – 7, 2007

[18] Xing Mei, Xun Sun, Mingcai Zhou, Shaohui Jiao, Haitao Wang, Xiaopeng Zhang, “On building an accurate stereo matching system on graphics hardware,” 2011 IEEE International Conference on Computer Vision Workshops (ICCV Workshops), pp. 467-474, 2011.

[19] Min Ki Park, Ji-Ho Cho, In yeop Jang, Seung Joo Lee and Kwan H. Lee, "An Iterative Joint Bilateral Filtering for Depth Refinement of a 3D Model", ACM SIGGRAPH Asia 2011 Poster, 12-15 December 2011, Hong Kong.

[20] Wan-Yu Chen, Yu-Lin Chang, Shyh-Feng Lin, Li-Fu Ding and Liang-Gee Chen, "Efficient Depth Image Based Rendering with Edge Dependent Depth Filter and Interpolation," IEEE International Conference on Multimedia and Expo, 2005. ICME 2005, pp. 1314-1317, 2005

[21] L. Wang, M. Liao, M. Gong, R. Yang, and D. Nistér, “High-quality real-time stereo using adaptive cost aggregation and dynamic programming,” 3rd International Symposium on 3D Data Processing, Visualization, and Transmission, 3DPVT 2006, pp. 798-805, 2006