论文部分内容阅读
只要信号是可压缩的或在某个变换域是稀疏的,压缩感知理论便可在信号采样的同时对其进行高效的压缩,该理论中选择合适的测量矩阵对信号的获取和重建精度起着关键作用。文章通过阈值处理的方法在高斯随机测量矩阵中引入零元,形成一定稀疏结构的高斯随机测量矩阵,使得非零元个数减少到原高斯矩阵的1/2~1/16,甚至更少,有利于数据的存储和传输。仿真实验表明,优化后的测量矩阵不仅保证了信息原有的重建效果,而且降低了程序运行时间,使得信息的重建速度加快。
As long as the signal is compressible or sparse in a transform domain, compressed sensing theory can efficiently compress the signal while sampling it. Selecting the appropriate measurement matrix for this theory plays a significant role in signal acquisition and reconstruction accuracy Key role. In this paper, by introducing the zero element into Gaussian random measurement matrix by means of thresholding method, a Gaussian random measurement matrix with a certain sparse structure is formed, reducing the number of nonzero elements to 1/2 to 1/16 of the original Gaussian matrix or even less, Conducive to data storage and transmission. The simulation results show that the optimized measurement matrix not only guarantees the original reconstruction effect, but also reduces the running time of the program and accelerates the reconstruction of information.