AN improved approach to side information extraction in distributed video coding

In the distributed video coding, side information is involved not only in the decoding of the quantized source, but also in the reconstruction of the images. So the quality of side information takes an important role in the distributed video codec. In this paper, a refined approach to SI extraction is proposed. Based on the high correlation between the motion vectors of adjacent blocks, the motion vector of each block is estimated considering the predicted motion vector of the neighboring blocks as initial M V. Besides, extrapolation beyond boundary is applied in the unrestricted motion vectors estimation for the case that there is an object moving in and out of the picture or a camera lens shaking. The performance of the proposed approach is evaluated in the context of Wyner-Ziv video coding, and compared with existing methods. The results verify the effectiveness of the proposed approach.