为了降低多视点视频压缩的复杂度，根据密集相机阵列系统与Wyner—Ziv编码的特点，提出了基于Wyner—Ziv编码的密集相机阵列低复杂度视频压缩算法。该算法首先在各相机之间相互独立地采用基于感兴趣区提取的低复杂度来进行Wyner—ziv编码，然后在中心解码端利用各视点间的相关性进行联合解码。该算法是通过对DCT量化系数进行相似性判断来提取感兴趣区，以有效地避免对背景和平坦等区域进行编码，从而降低了编码复杂度。实验证明，该算法具有极低的编码复杂度特性和良好的率失真性能，编码复杂度仅为H．264帧间预测编码的1／22。

In this paper, a new low complexity video coding method in camera array is proposed, based on the ROI-based Wyner-Ziv coding, The proposed method encodes each view independently at the senders, and performs joint decoding with side information interpolated from adjacent views at the receiver. Through comparing the quantized DCT coefficients, the proposed ROI-based Wyner-Ziv coding method can avoid to encoding the regions with little changes. Experimental results show that, the average encoding time of the proposed scheme is only 1/22 of the H. 264 P coding.