在3D視訊的壓縮中，考慮到位元率及編碼時間的限制，常常使用非對稱多視角視訊編碼技術，意即在壓縮端將偶數號碼的相機所拍攝的影片畫面進行降頻取樣，以降低在壓縮後的位元率及編碼的時間，並在解碼端將進行超解析度還原成原始畫面大小。在這篇論文中，我們提出了一套有效的方法來提高超解析度後畫面的品質。在所提出的方法中，我們除了利用了未知像素的鄰近像素的資訊，還利用奇數號碼相機所拍攝的影片透過深度影像彩現技術合成產生偶數號碼相機的虛擬視角來進行未知像素的預測。此外，我們亦提出一套植基於虛擬視角的改良式高頻紋理補償技術以提高邊緣及紋理區域的畫面品質。實驗結果顯示與現存多個的超解析度方法相比，我們的方法能使超解析度後視訊畫面的訊號雜訊比平均可提升超過0.31 dB。

For the 3D video compression applications, the asymmetric multiview video coding is often employed due to the restrictions on the bitrate and compression time. The asymmetric multiview video coding down-samples the video frame captured by each second view to reduce the bit-rate and compression time at the encoding side and up-samples each decoded the low resolution video frame individually to recover the original video resolution at the decoding side. In this thesis, we propose a novel super resolution method to enhance the quality of up-sampled video frames. In the proposed method, in addition to the neighborhoods of current pixel, the pixels in the corresponding virtual view, which is constructed by the depth-image-based rendering, is also exploited to assist in predicting the pixel value of current pixel. Furthermore, we develop a novel improved virtual-view-based high frequency texture compensation technique to enhance the quality of texture and edge regions. Experimental results demonstrate that the proposed method can yield better quality performance in terms of peak signal-to-noise ratio (PSNR) measure when compared with the existing methods, and the average PSNR gain of the proposed method can achieve more than 0.31 dB.

[1] L. Zhang and W. Tam, “Stereoscopic image generation based on depth images for 3D TV,” IEEE Trans. Broadcasting, Vol. 51, No. 2, pp. 191-199 (2005).
[2] A. Vetro, S. Yea, M. Zwicker, W. Matusik, and H. Pfister, “Overview of multiview video coding and anti-aliasing for 3D displays,” IEEE Trans. Image Processing, Vol. 1, pp. I-17-I-20 (2007).
[3] P. Merkle, A. Smolic, K. Muller and T. Wiegand, “Efficient prediction structures for multiview video coding,” IEEE Trans. Circuits Systems for Video Technology, Vol. 17, No. 11, pp. 1461-1473 (2007).
[4] Y. Chen, S. Liu, Y. K. Wang, M. M. Hannuksela, H. Li, and M. Gabbouj, “Low-complexity asymmetric multiview video coding,” IEEE Int. Conf. Multimedia and Expo, Hannover, Germany, pp. 773-776, June. 23-26 (2008).
[5] X. Li and M. T. Orchard, “New edge-directed interpolation,” IEEE Trans. Image Processing, Vol. 10, No. 10, pp.1521-1527 (2001).
[6] Y. Zhang, D. Zhao, J. Zhang, R. Xiong and W. Gao, “Interpolation Dependent Image Downsampling,” IEEE Trans. Image Processing, Vol. 20, No. 11, pp. 3291-3296 (2011).
[7] K. L. Chung, Y. H. Huang, T. H. Shen and W. C. Liu, “Quality-Efficient Up-Sampling Method for Asymmetric Stereoscopic Video Coding Using Inter-View Motion Compensation and Error Compensation Schemes,” in Proc. of IPPR Conf. on Computer Vision, Graphics, and Image Processing, Nantou, Taiwan, session A1, A1-6, Aug. 12-14 (2012).
[8] D. C. Garcia, C. Dorea and R. L. de Queiroz, “Super Resolution for Multiview Images Using Depth Information,” IEEE Trans. Circuits Systems for Video Technology, Vol. 22, No. 9, pp.1249-1256 (2012).
[9] C. Fehn, “Depth-image-based rendering (DIBR), compression and transmission for a new approach on 3D-TV,” in Proc. of SPIE Conf. Stereoscopic Displays and Virtual Reality Systems XI, CA, USA, Vol. 5291, pp. 93-104, Jan. 19-12 (2004).
[10] T. Richter, J. Seiler, W. Schnurrer and A. Kaup, “Robust Super-resolution in a multiview Setup based on Refined High-frequency Synthesis,” in Proc. of IEEE Int. Workshop on Multimedia Signal Processing (MMSP), Banff, Canada, pp.7-12, Sep.
17-19 (2004).
[11] M. D. Rehman and K. B. Yu, “Total least squares approach for frequency estimation using linear prediction,” IEEE Trans. Speech and Signal Processing, Vol. 35, No. 10, pp. 1440-1454 (1987).