H.264 比起以往的視訊壓縮標準來說，在相同畫面品質下，H.264 只需要一半左右的位元率即可，即H.264 有著較佳的壓縮率，其原因是H.264運用了許多以往視訊標準沒有使用的技術，大致上有多幅參考畫面(multi-reference frame)、可變區塊大小(variable block size)以及四分之一像素精確度搜尋，雖然這些技術讓H.264 編碼效益(coding efficiency)提升許多，但同時也帶來了龐大的運算複雜度，故H.264 在即時系統應用實現上有著相當大的難度。在本論文中，首先提出了快速區塊模式決策，以及快速多幅參考畫面決策，藉此降低其移動估測時的運算複雜度，加快其編碼時間，並在最後再進一步地去做編碼加速，使其盡可能地達到即時運算，從實驗數據顯示，我們所提出的即時演算法最快可以達到每秒播放六十張畫面左右，大概減少96％左右的複雜度，同時維持畫面品質在一定可接受的失真範圍之內，故此即時演算法在即時系統應用上應該會有著相當不錯的表現。 Compared to previous video coding standards, H.264/AVC has better coding efficiency with numerous powerful techniques such as multi-reference frame, variable block size,quarter pixel resolution. It can save about 50% bit rate under the same image quality. But these structures also give rise to a heavy loading of the H.264 encoder to be applied in real-time applications. Therefore,we propose some methods to reduce the computational complexity as possible. First, we design a fast decision schemes for mode and multi-eference frames. And then we accelerate the search pattern. Finally, we integrate these ideals into a fast algorithm. According to the experimental results, our proposed algorithm can improve the coding speed efficiently. It can roughly reduce 96% of coding time compared to full search in reference software with negligible degradation.
