论文部分内容阅读
本文提出一种基于块稀疏贝叶斯学习的多任务压缩感知重构算法,利用块稀疏的单测量矢量模型求解多任务重构问题.通过对信号统的计特性和稀疏块内的结构特性进行联合数学建模,将稀疏重构问题转贝叶斯框架下的特征参数的迭代更新问题.本文算法不需要信号稀疏度和噪声强度的先验信息,是一种高效的盲重构算法.仿真实验表明,本文算法能有效利用信号的统计特性和结构信息,在重构精度和收敛速率方面能够很好地折衷.
In this paper, we propose a multiscreen compressed perceptual reconstruction algorithm based on sparsity Bayesian learning, which solves the multitask reconfiguration problem by using a sparse single-vector vector model. By considering the statistical properties of signals and the structural features of sparse blocks Joint mathematical modeling, the sparse reconstruction problem to the iterative updating of the characteristic parameters under the Bayesian framework, is a highly efficient blind reconstruction algorithm which does not require prior information of signal sparsity and noise intensity. Experiments show that the proposed algorithm can make good use of the statistical properties and structural information of the signal, and can well compromise the reconstruction accuracy and the convergence rate.