Figure 3From: Wyner-Ziv video coding for wireless lightweight multimedia applicationsGraphical representation of the motion estimation algorithm. All the overlapping B × B blocks, that contain the current motion-compensated pixel position in the hash frame W ̃ , are designated W ̃ m i , where m i = m 1 i , m 2 i are the top-left coordinates of the block. For every block W ̃ m i the best matching block in the preprocessed reference frames R ̃ k = 0 , R ̃ k = 1 is found by minimizing the SAD, yielding the motion vectors v k = 0 i , v k = 1 i . The co-located pixels in the estimated blocks R ̃ k = 0 , m i - v k = 0 i , R ̃ k = 1 , m i - v k = 1 i serve as potential temporal predictors for the current motion-compensated pixel position.Back to article page