以基要畫框及可調適細節扭曲作超解析度視訊重建

Video Super-resolution Reconstruction Using Key Frames and Adaptive Detail Warping

陳畇蓁(Chen, Yunjhen)

指導教授 ： 柳金章

國立中正大學/工學院/資訊工程學系/碩士(2013年)

摘要

科技日新月異，許多大螢幕的顯示器紛紛推出，而在相同的訊號播放之條件下，在顯示於螢幕的解析度變高變成一種新的問題，所以超解析度(super-resolution)的方法被提出來解決這個需求。
本論文中，我們提出一種使用基要畫框和適應性細節扭曲的視訊超解析的方法，首先將基要畫框與非基要畫框做動作估計 (motion estimation) ，找出在非機要畫框在基要畫框內對應點，將對應點在基要畫框內的細節擷取出來，再利可適性細節扭曲的方法來估計非基要話框的細節，使用非局部均值濾波的遮罩來做細節的估計，估計出直接用內插得到的模糊視訊所失去的細節，最後將細節與模糊影像生成出高解析度的視訊結果。根據實驗結果顯示，本研究所提方法比其他系統有叫好得效能。

關鍵字

超解析

並列摘要

The technique has greatly advanced day by day, and many big monitors are springing up all over the shop recently. It is a problem for displaying high resolution in a monitor on condition that the signals are smaller, and a method to solve this problem by super-resolution.
In this thesis, we proposed a super-resolution approach using key frames and adaptive detail warping. At the first, find the corresponding pixels between key frames and non-key frames by motion estimation, and extract the detail of each corresponding pixels. Then adaptive detail warping based on non-local mean to evaluate the details which are missing in the smooth video, and then add the details evaluated by the proposed approach into the smooth video to reconstruct the high resolution video results. Based on the experimental results obtained in this thesis, the proposed system has better performance than two comparison systems.

並列關鍵字

super resolution

參考文獻

[1] K. H. Yap, Y. He, Y. Tian, and L. P. Chau, “A nonlinear L1-norm approach for joint image registration and super-resolution,” IEEE Signal Processing Letters, vol. 16, pp. 981-984, 2009.

[3] W. T. Freeman, T. R. Jones, and E. C. Pasztor, “Example-based super-resolution,” Computer Graphics and Applications, vol. 22, no. 2, pp. 56-65, 2002.

[4] J. Yang, J. Wright, T. Huang, and Y. Ma, “Image super-resolution as sparse representation of raw image patches,” in Proc. of 2008 IEEE Int. Conf. on Computer Vision and Pattern Recognition, 2008, pp. 1–8.

[5] X. Yang, G. D. Su, J. Chen, and Y. S. Moon, “Restoration of low resolution car plate images using PCA based image super-resolution,” in Proc. of 17th IEEE Int. Conf. on Image Processing, 2010, pp. 2789-2792.

[6] H. Takeda, S. Farsiu, and P. Milanfar, “Kernel regression for image processing and reconstruction,” IEEE Trans. on Image Processing, vol. 16, no. 2, pp. 349-366, Feb. 2007.