基于PCA稀疏描述的分布式视频编码

PCA Sparse Representation for Distributed Compressive Video Sampling

在研究视频信号的非局部稀疏模型的基础上,提出了一种适合分布式视频编码的视频信号稀疏描述算法。在解码端,对非关键帧解码时,首先由已重建的关键帧生成边信息SI,然后在这两帧中寻找非关键帧当前块的L个相似块,作为一个类,最后利用这个类生成PCA子字典,并将所有块的子字典集合起来构成非关键帧的字典,进行视频重建。通过与典型的信号稀疏描述方法进行对比,实验结果显示,重建的PSNR值平均提高2 dB,主观视觉质量也有较大的提高。

In this paper, PCA dictionary based on learning is proposed by the video interframe and in- traframe non-local self-similarity. Grouping by matching is realized from key frames which are previously recovered. PCA is then applied to each group （sub-dataset） to compute the principle components, from which the sub-dictionary is constructed. The simulation results show that the proposed algorithm outper- forms many state-of-the-art algorithms in terms of PSNR and visual oerceotion.