计算机工程2019，Vol.45Issue(2)：245-249,5.DOI:10.19678/j.issn.1000-3428.0050093
基于双稀疏优化的空域错误隐藏
Spatial Error Concealment Based on Coupled Sparse Optimization
摘要
Abstract
Linear interpolation algorithm or conventional sparse representation algorithm are used to recover the lost pixels currently. However, linear interpolation restores image blur due to inconsistent neighborhood information when restoring unsmooth images. For the conventional sparse representation algorithm, improper dictionary construction will result in a poor recovered image quality. To solve these problem, an improved spatial error concealment algorithm is proposed. The proposed algorithm optimizes the process of potential set and template set search by means of dynamic threshold searching, which improves the precision of the constructed dictionary. It can obtain the value of double sparse optimization using Canonical Correlation Analysis (CCA), and recover lost pixels by sparse reconstruction. Experimental results show that the proposed algorithm improves the Peak Signal to Noise Ratio (PSNR) by at least 1. 23 dB compared with the current mainstream algorithm, and has a good error hiding effect.关键词
空域错误隐藏/线性插值/I帧丢失/字典构建/稀疏优化Key words
spatial error concealment/linear interpolation/I-frame loss/dictionary construction/sparse optimization分类
信息技术与安全科学引用本文复制引用
严静文,肖晶,高戈..基于双稀疏优化的空域错误隐藏[J].计算机工程,2019,45(2):245-249,5.基金项目
国家自然科学基金 (61471271). （61471271）
