"基于压缩感知的三维视频错误隐藏技术研究"

Based on the research of "Error Concealment of Three-Dimensional Video using Compressive Sensing" (The full document is not provided), the paper explores the effectiveness of error concealment techniques at the decoding end for 3D video, utilizing the low-frequency characteristics of natural images to restore damaged images and reduce visual artifacts and errors caused by misinterpretation. In recent years, the emergence of compressive sensing theory has brought new ideas to error concealment techniques, presenting a new trend in the development of future video applications. The paper proposes a new method of error concealment for three-dimensional video based on compressive sensing, using the sparse properties of video image compression to create a local dictionary and perform error concealment through sparse representation. This method does not rely on the recovery of motion vectors, instead dividing the lost blocks into foreground and background blocks. The foreground blocks are restored using compressive sensing-based error concealment, while the background blocks are recovered using a direct copy from the temporal reference frame or inter-view reference frame. Experimental results demonstrate that this method effectively addresses the issue of damaged video image restoration, with good subjective and objective quality of the reconstructed images. The keywords for this approach include error concealment, compressive sensing, sparse representation, and dictionary construction. The study provides valuable insights into the application of compressive sensing theory in error concealment for three-dimensional video, contributing to the advancement of video processing and compression technologies. Overall, the research highlights the potential of compressive sensing in enhancing error concealment techniques and improving the quality of 3D video transmission and reception.