在最新的標準MPEG-4 AVC/JVT/H.264 中，動態頻估可以利用多重參考以及多種型態的區塊來改良其壓縮率。然而全域搜尋（full search-FS）將搜尋所有的區塊型態和參考圖像，因此其運算複雜度隨著參考的張數以及區塊型態而線性增加。因此在本篇論文當中，一個新的搜尋方式，預測六角型演算法(predict hexagon search-PHS)將被提出。預測六角型演算法是改良自六角型演算法，其可預測六角型的形狀，並且在±2的十字範圍內做重點式的搜尋。經由實驗結果可得知，預測六角型演算法與鑽石型演算法(diamond search–DS)相比較可減少58％的搜尋點數，和六角型演算法(hexagon based search-HEXBS)比較可減少53％的搜尋點數。另一方面來說，預測六角型演算法可以預測相同的動態向量而花較少的搜尋點數。 The upcoming video coding standard, MPEG-4 AVC/JVT/H.264, motion estimation allowed to use multiple references and multiple block sizes to improve the rate-distortion performance. However, full exhaustive search of all block sizes is computational intensive with complexity increasing linearly to the number of allowed reference frame and block size. In this thesis, a novel search algorithm, predict hexagon search (PHS), is proposed. The PHS pattern is applied by the comparison with the hexagon based search pattern. It can predict the hexagon search pattern in horizontal or vertical direction and search the ±2 rood side first. Analysis shows that the speed improvement of the PHS over the diamond search (DS) and the hexagon based search (HEXBS) is about 58% and 53% respectively. On the other hand, the proposed PHS algorithm can find a same motion vector with fewer search points than the DS and HEXBS algorithm.
