| 摘要 |
A method for estimating the motion vector for a current macroblock within a current frame is provided by reference to a number of reference macroblocks within the previous frame. The method includes the following steps. First, an error between the current macroblock within a current frame and each of the reference macroblocks is determined according to a weighted sum of absolute differences (SAD) operation. The weighted SAD operation is to emphasize the absolute differences for high-frequency pixels within the current macroblock. Next, the reference macroblock having the lowest error may be determined as a matched macroblock. Finally, the motion vector for the current macroblock can be defined as the displacement between the macroblock and the matched macroblock.
|
