非對稱交叉混合多邊型式移動估測演算法之研究

The Study of Unsymmetrical-Cross Hybrid-Polygon Search of Motion Estimation Algorithm

周麟(Lin Chou)

指導教授 ： 林登彬

大同大學/電機資訊學院/電機工程學系所/碩士(2011年)

摘要

H.264/AVC是一個非常有效率的影像壓縮標準。本論文利用ITU-T/ISO的JM software，來進行H.264/AVC的非對稱交叉混合多邊形移動估測演算法的實現。在影像壓縮的過程中，移動估測扮演著重要的角色，不同的演算法將會對壓縮時間、搜尋點有不同的影響，假使採用高效率的演算法便可以有效的節省編碼所需的時間。
本篇提出以UMHexagonS為基礎改良的非對稱交叉混合多邊形移動估測演算法 UHPolygonS(Unsymmetrical-Cross Hybrid-Polygon Search)，改變亞像素搜尋樣板以及增加四分之一像素的精確度減少計算時間。實驗結果顯示，與其他移動估測演算法比較，此演算法可有效減少計算時間和增加精確度。以計算時間來說，與全域演算法比較最大差距可至511.08秒，相當於快了8.51分鐘;而以精確度來說，與非對稱交叉多重六邊型演算法來比較，PSNR值至多可增至1.897，因此演算法能夠有效的增進整個系統的效能。

關鍵字

H.264/MPEG-4 ； 視訊壓縮標準 ； motion estimation ； UMHexgonS

並列摘要

H.264 is a very efficient video compression standard. In the process of H.264 video compression, motion estimation plays an important role in video compression for reducing the complexity of calculation. There are different impacts on the performance while using various search algorithms for motion estimation. Efficient algorithm can save more computing time and more search points. A search algorithm is based on Unsymmetrical-cross multiple-Hexagon Search algorithm (UMHexagonS) for motion estimation called Unsymmetrical-cross Hybrid-Polygon Search algorithm (UHPolygonS) is proposed. It changed the sub-pixel search pattern and increased the accuracy of quarter-pixel search to reduce computation time. The experimental result shows that this algorithm can reduce computation time and increase the accuracy than other motion estimation algorithms. For time consumption, compared with Full Search algorithm, this algorithm almost faster about 511.08 seconds than FS, equally to 8.51 minutes; for accuracy, the PSNR values increased up to 1.897 compared with UMHexagonS algorithm. Hence this algorithm can effectively improve the system performance.

並列關鍵字

H.264/VAC ； Video Compression Standard ； Motion Estimation ； UMHexagonS

參考文獻

Standard”.IEEE Transactions on Circuits and System for Video Technology,

[2] Brunig, M., Niehsen, W. “Fast Full-Search Block Matching”. IEEE Trans. on

Inst.of Commun. Eng., pp. 241-247, Feb 2001

[3] J.Jain, A.Jain, “Displacement measurement and its application in inter-frame

[4] R. Li; B. Zeng and M.L Liou; “A new three-step search algorithm of block