在利用立體匹配演算法計算深度圖時，會面臨到像是單一顏色平面、複雜的遮蓋區域等影像特性所導致的難以正確匹配的問題，在本篇論文中，提出了整合了顏色梯度、顏色強度跟紋理的成本計算方式，加強了成本計算中的紋理重要性，在單一顏色平面的匹配上有更高的準確性。在深度圖的後處理部分，提出了整合左右圖檢查法、雙邊濾波器特性的可遞迴式後處理機制，藉此可以有效修補匹配錯誤和複雜遮蓋的區域。
除此之外，由於圖形處理器的限制，使得演算法無法對高解析度影像的每一個像素點進行平行運算，一般會將高解析度影像降維至可以平行處理的大小後，再去進行立體匹配的計算，而因此則會導致的深度圖解析度不足跟深度值範圍差異的問題。在本篇論文中也有提出了深度值從低解析度深度值範圍更新至高解析度深度值範圍的深度值更新演算法、針對深度圖特性的優化雙線性插值升維演算法以及高解析度影像的局部深度值投票演算法，以此可以得到更正確的高解析度深度圖。

In stereo matching, it is difficult to calculate the right depth value on smooth color plane and occlusion regions, due to the ambiguity of cost values in these regions. In this paper, we propose integrating the color gradient, color intensity and texture for computing the cost. By incorporating the texture in cost calculation, the method can achieve a higher accuracy on corresponding matching on single color plane region. Furthermore, a recursive post-processing mechanism which integrates the characteristics of left-right check method and bilateral filter is also proposed. With the proposed recursive post-processing the matching errors in smooth regions and the occlusion regions can be effectively repaired.
One common problem associated with depth estimation for high-resolution images is the limitations of the graphic processing unit (GPU). Because of this reason in general the high-resolution images are down sized for stereo matching. This approach will lead to the insufficient resolution of depth map and the different depth range between low resolution and high resolution versions of the same image. In this thesis, we also proposed depth value updating algorithm to update depth map from low-resolution depth range to high-resolution depth range. Then an optimized bilinear interpolation is applied followed by a local depth voting algorithm to resize the depth map to high spatial resolution, so that high-resolution depth map can be more correct and in line with high-resolution image.

[1] W. Y. Chen, Y. L. Chang, S. F. Lin and L. F. Di, "Efficient depth image based rendering with edge dependent depth filter and interpolation," IEEE International Conference on Multimedia and Expo., pp. 1314-1317, 2005.
[2] Y. Boykov, O. Veksier, R. Zabih, "Fast approximate energy minimization via graph cuts," IEEE Transactions on Pattern Analysis and Machine Intelligence, pp. 1222-1239, 2001.
[3] V. Kolmogorov, R. Zabih, "Computing visual correspondence with occlusions using graph cuts," Computer Vision, 2001. ICCV 2001. Proceedings. Eighth IEEE International Conference on, 2001.
[4] Yu-Chih Wang, Cheng-Ping Tung, Pau-Choo Chung, "Efficient Disparity Estimation Using Hierarchical Bilateral Disparity Structure Based Graph Cut Algorithm With a Foreground Boundary Refinement Mechanism," IEEE Transactions on Circuits and Systems for Video Technology, pp. 784-801, 2013.
[5] Jian Sun, Nan-Ning Zheng, Heung-Yeung Shum, "Stereo matching using belief propagation," IEEE Transactions on Pattern Analysis and Machine Intelligence, pp. 787-800, 2003.
[6] Qingxiong Yang, Liang Wang, Ruigang Yang, Henrik Stewénius, David Nistér, "Stereo Matching with Color-Weighted Correlation, Hierarchical Belief Propagation, and Occlusion Handling," IEEE Transactions on Pattern Analysis and Machine Intelligence, pp. 492-504, 2009.
[7] A. Klaus, M. Sormann, K. Karner, "Segment-Based Stereo Matching Using Belief Propagation and a Self-Adapting Dissimilarity Measure," 18th International Conference on Pattern Recognition (ICPR'06), 2006.
[8] O. Veksler, "Stereo correspondence by dynamic programming on a tree," 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05), 2005.
[9] X. M. S. J. M. Z. H. W. Xun Sun, "Stereo Matching with Reliable Disparity Propagation," 3D Imaging, Modeling, Processing, Visualization and Transmission (3DIMPVT), 2011 International Conference on, 2011.
[10] Kuk-Jin Yoon, In So Kweon, "Adaptive support-weight approach for correspondence search," IEEE Transactions on Pattern Analysis and Machine Intelligence, pp. 650-656, 2006.
[11] Asmaa Hosni, Michael Bleyer, Margrit Gelautz, Christoph Rhemann, "Local stereo matching using geodesic support weights," Image Processing (ICIP), 2009 16th IEEE International Conference on, 2009.
[12] Ke Zhang, Jiangbo Lu, Gauthier Lafruit, "Cross-Based Local Stereo Matching Using Orthogonal Integral Images," IEEE Transactions on Circuits and Systems for Video Technology, 2009.
[13] Q. Yang, "Stereo Matching Using Tree Filtering," IEEE Transactions on Pattern Analysis and Machine Intelligence , 2014.
[14] H. Hirschmuller, "Stereo Processing by Semiglobal Matching and Mutual Information," IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008.
[15] R. S. R. Z. D. Scharstein, "A taxonomy and evaluation of dense two-frame stereo correspondence algorithms," Stereo and Multi-Baseline Vision(SMBV), 2001.
[16] X. S. M. Z. S. J. H. W. X. Z. Xing Mei, "On building an accurate stereo matching system on graphics hardware," Computer Vision Workshops (ICCV Workshops), 2011.
[17] R. Y. J. D. D. N. Qingxiong Yang, "Spatial-Depth Super Resolution for Range Images," Computer Vision and Pattern Recognition(CVPR), 2007.
[18] "Middlebury," [Online]. Available: http://vision.middlebury.edu/stereo/.
[19] S. B. M. B. M. M. a. L. S. S. Drouyer, "Sparse stereo disparity map densification using hierarchical image sgmentation," in 13th International Symposium on Mathematical Morphology.
[20] C. C. a. A. Alatan, "Information permeability for stereo matching," in Signal Processing: Image Communication 28(9), 2013.
[21] H. Hirschmüller, "Stereo processing by semi-global matching and mutual information," in CVPR, 2006.
[22] Wan-Yu Chen, Yu-Lin Chang, Shyh-Feng Lin, Li-Fu Ding, Liang-Gee Chen, "Efficient Depth Image Based Rendering with Edge Dependent Depth Filter and Interpolation," IEEE International Conference on Multimedia and Expo, 2005.
[23] Q. Yang, R. Yang, J. Davis and D. Nistér, "Spatial-Depth Super Resolution for Range Images," in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2007.