簡易檢索 / 詳目顯示

回結果列表
研究生： 陳維濤
Chen, Wei-Tao
論文名稱： 計算複雜度可變動之MPEG-4 AVC Encoder
Computational power sensitive MPEG-4 AVC encoder
指導教授： 蘇文鈺
Su, Wen-Yu
學位類別： 碩士
Master
系所名稱： 電機資訊學院 - 資訊工程學系
Department of Computer Science and Information Engineering
論文出版年： 2004
畢業學年度： 92
語文別： 中文
論文頁數： 54
中文關鍵詞： 控制計算 、分配
外文關鍵詞： MPEG-4, Computation, MMX, Motion Estimation, H.264
相關次數： 點閱：63  下載：1 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 　　隨著壓縮標準不斷的進步與改良，使得我們得到越來越好的影音品質以及壓縮比。如同大家所知道的，壓縮比與計算複雜度成反比，相較於以往的video encoder , MPEG-4 AVC Encoder多出數倍的計算量而獲得較好的壓縮效能。但是由於必須耗費的計算量太大，使得可以應用的方面相對變少。因此我們希望能夠控制AVC Encoder 的計算量，並且在有限的計算量之下得到最好的影像品質。

    　　與以往的video encoder 一樣，AVC encoder 大部分的計算量仍然來自於動態估測(Motion Estimation, ME)。因此我們藉由控制ME 的計算複雜度達到控制計算量的目的。在這篇論文中，我們除了修改程式的流程以求提昇Encoder 的執行速率之外，並藉由統計了解AVC Encoder 的特性，決定計算量的分配方式，並實際規劃出根據不同的frame-rate 決定Encoder 計算複雜度的演算法。最後我們實際模擬我們所提出的演算法，並提供模擬數據加以驗證。

    　　Better Audio-Video compression quality is achieved with the progress of new standards. As we know, the better compression ratio, the more computation complexity we need. AVC Encoder has many times complexity than other encoders before to achieve better compression performance. The huge computation complexity limits its practical applications. Therefore, we want to control the encoder to achieve better quality with limited computation power of the available computer.

    　　Similar to previous encoders, most computation power of AVC encoder is spent on “Motion Estimation”. Therefore, we control the computation complexity of AVC encoder through controlling “Motion Estimation” part. In this paper, we not only modify the coding procedure to improve the performance of AVC Encoder, but also design an algorithm that allocates the available computing power to different coding functions under a preset frame-rate. Computer Simulations are presented for various standard test sequences to show the effect of the algorithms proposed in this thesis.

    摘要.................................................i 目錄...............................................iii 圖目錄...............................................v 表目錄.............................................vii 第一章 AVC Encoder introduction......................1 1.1 AVC Encoder 簡介.................................1 1.2 AVC Encoder 壓縮流程.............................1 1.3 研究背景與動機...................................2 1.4 章節概要.........................................3 第二章 AVC Encoder 加速與改進的相關研究..............4 2.1 Motion Estimation................................4 2.1.1 完整搜尋法(Full Search)........................7 2.1.2 三步搜尋法(Three-Step Search)..................9 2.1.3 菱形搜尋法(Diamond Search)....................11 2.2 Intra Prediction................................13 2.2.1 Fast Intra Prediction Mode Decision...........14 2.3 Transform.......................................17 2.3.1 Method for detecting all-zero coefficient.....17 第三章 AVC Encoder 的相關統計.......................19 3.1 AVC Encoder 計算量的分佈........................19 3.2 MacroBlock 分割模式的比例分佈...................21 3.3 參考影像的分佈..................................23 3.4 移動向量的分佈..................................25 3.5 三步搜尋法與完整搜尋法的比較....................29 第四章 AVC Encoder 的改進與控制計算量的實作.........31 4.1 AVC Encoder 的修改..............................31 4.1.1 MMX optimized.................................31 4.1.2 Sub-pel Motion Estimation Computation Reduced.32 4.2 Computation allocation..........................38 4.2.1 Reference Frame-level computation allocation..39 4.2.2 Mode-level computation allocation.............40 4.3 Implementation..................................40 4.4 Simulation Result...............................44 第五章 結論與未來研究方向...........................53 參考文獻............................................54

    [1] “Draft ITU-T Recommendation and Final Draft International Standard of Joint Video Specification (ITU-T Rec. H.264 | ISO/IEC 14496-10 AVC)”, ISO/IEC JTC1 /SC29/WG11, 2003, May.

    [2] T. Koga, K. Iinuma, A. Iijima, and T. Ishiguro, “Motion-compensated interframe coding for video conferencing,” in Proc. NTC81, New Orleans, LA, 1981, pp. C9.6.1–9.6.5.

    [3] R. Li, B. Zeng, and M. L. Liou, “A new three-step search algorithm for block motion estimation,” IEEE Trans. Circuits Syst. Video Technol. ,vol. 4, pp. 438–442, Aug. 1994.

    [4] S. Zhu and K.-K. Ma, “A new diamond search algorithm for fast block matching motion estimation,” IEEE Trans. Image Processing, vol. 9, pp. 287–290, Feb. 2000.

    [5] I-Ming Po, Ming-Ting Sun, “Modeling DCT coefficients for fast video encoding,” IEEE Trans. Circuit Syst. Video Technol., vol. 9, pp. 608–616, June 1999.

    [6] I-Ming Po, Ming-Ting Sun, “Computation Reduction for Discrete Cosine Transform,” IEEE Trans. Consumer Electron., vol. 45, pp.417–426, May 1999.

    [7] X. Zhou, Z. H. Yu, and S. Y. Yu, “Method for detecting all-zero DCT coefficients ahead of discrete cosine transformation and quantization,” Electron. Lett., vol. 34, pp.1839–1840, Sept. 1998.

    [8] Bojun Meng, Oscar C. Au, “Fast Intra-Prediction Mode Decision for 4x4 Blocks in H.264,” IEEE Acoustics, Speech, and Signal Processing Int. Conf., ICASSP '03 pp. 389-392,Vol.3 ,April 2003.

    下載圖示 校內：2014-08-27公開
    校外：2014-08-27公開
    QR CODE