適用於H.264之動態搜尋範圍編碼機制

Dynamic Search Range Decision for H.264

徐俊傑(Jun-Jie Shu)

指導教授 ： 郭天穎

國立臺北科技大學/電資學院/電機工程系所/碩士(2007年)

https://doi.org/10.6841/NTUT.2007.00181

摘要

H.264為目前廣泛所採用的最新視訊壓縮標準。有別於之前的壓縮標準， H.264採用了多張參考圖框、可變區塊大小及最高提供1/4像素精確度的技術，使得在運動估測方面，能獲得更精準的運動向量。在運動估測時，搜尋範圍越大，運算複雜度也隨之提高，而運動向量也會較準確，反之亦然。然而，我們則是希望在運算複雜度與運動向量的精確度上取得平衡。由於H.264使用的是固定的搜尋範圍，但是這是缺乏效率的。因為當編碼影像是屬於靜態的畫面時，較小的搜尋範圍其實就已足夠，不需要用到像動態畫面所需要之較大的搜尋範圍。因此，本論文提出適用於H.264之動態搜尋範圍編碼機制，利用時間域和空間域的關係，針對不同的影像內容，動態地提供較小且有效的搜尋範圍，避免對大範圍內所有像素進行無效率的運動估測。實驗結果顯示，我們的方法不僅有效地減少運算複雜度，縮短運動估測時間，並維持原來的壓縮品質。

關鍵字

H.264 ； 多張參考圖框 ； 可變區塊大小 ； 運動估測 ； 動態搜尋範圍

並列摘要

H.264 is the latest and widely adopted video coding standard. Comparing to previous coding standard, H.264 adopts multiple reference frames, variable block size, and provides as high as quarter resolution accuracy. When performing motion estimation, the larger search range, the more accurate motion vectors we can get, which is resulted by heavier computational complexity, and vice versa. This is the trade-off between the motion accuracy and computational complexity. Fixed search range is used in H.264 but it is inefficient. Because the smooth moving sequences need smaller search range while high motion sequences need larger one. For this reason, we proposed Dynamic Search Range Decision for H.264 to reduce the computational complexity. We dynamically determine the search range based on the relation between temporal and spatial video content to avoid inefficient motion estimation on useless search positions. Experimental results show that our method not only efficiently reduces the computational complexity, motion estimation time, but also keeps the original great performance.

並列關鍵字

H.264 ； multiple reference frames ； variable block size ； motion estimation ； dynamic search range

參考文獻

[6] L.M. Po and W.C. Ma, "A novel four-step search algorithm for fast block motion estimation," IEEE Trans. Circuits Syst. Video Technol., vol. 6, no. 3, June 1996, pp. 313-317.

[7] S. Zhu and K.K. Ma, "A new diamond search algorithm for fast block-matching motion estimation," IEEE Trans. Image Processing, vol. 9, no. 2, Feb. 2000, pp. 287-290.

[8] C. Zhu, X. Lin, L.P. Chau, K.P. Lim, H.A. Ang, and C.Y. Ong, "A novel hexagon-based search algorithm for fast block motion estimation," in Proc. IEEE Int. Conf. Acoustics, Speech, and Signal Processing, vol. 3, May 2001, pp. 1593-1596.

[9] C. Zhu, X. Lin, L.P. Chau, and L.M. Po, "Enhanced hexagonal search for fast block motion estimation," IEEE Trans. Circuits Syst. Video Technol., vol. 14, no. 10, Oct. 2004, pp. 1210-1214.

[10] Z. Chen, P. Zhou, and Y. He, "Fast Integer Pel and Fractional Pel Motion Estimation for JVT," ITU-T Q6/SG16, Doc#JVT-F017, 5-12 Dec. 2002.