跳帧转换编码中的运动矢量修正

Motion vector refinement for frame-skipping transcoding

在跳帧转换编码中,利用前向向量法进行运动矢量合成可以得到较好的图像信噪比。本文提出了用HAVS快速搜索算法对合成的运动矢量进行修正的方法,同时给出了一个最佳搜索窗口。实验结果表明,在±2像素内,用HAVS算法对合成的运动矢量修正,既能明显提高图像的信噪比,同时又能最大限度的降低计算量。

In frame-skipping transcoding, the forward vector method to compose motion vectors from the incoming bit stream can improve PSNR better than others. We propose a fast-search algorithm of HAVS and opitimal search window for refinement of composed motion vectors in frame-skipping transcoding. Experimental results show that HAVS can improve PSNR greatly in 2 pixels,while needing less additional computation.