摘要
为了在稀疏度未知的情况下重构信号,并且解决SAMP框架下的步长选择难题,提出一种新的稀疏度估计方式,以及一种新的压缩感知重构算法——步长自适应匹配追踪算法。该算法通过新的方式估计稀疏度,采用估计出的稀疏度作为初始步长,重构信号间能量差作为改变步长的方法,使得信号能在稀疏度未知的条件下,自适应的重构信号。实验结果表明,本算法能够较好地重构信号,保证重构质量的同时提高重构速度。
In order to reconstruct the signal under the sparse unknown condition and solve the step selection problem for the SAMP framework.A new sparse degree estimation strategy and a new compressive reconstruction algorithm-Estimate Step Adaptive Matching Algorithm (ESAMP) are proposed.The proposed algorithm estimates the sparseness in a new way.Using the estimated sparse degree as the initial step size and reconstructing the energy difference between the signals as the method of changing the step size,so that the signal can adaptively reconstruct the signal under the sparseness unknown condition.Through the experiments,the proposed algorithm can reconstruct the signal better and ensure the reconstruction quality while improving the reconstruction speed.
出处
《青岛大学学报(自然科学版)》
CAS
2017年第3期69-75,共7页
Journal of Qingdao University(Natural Science Edition)
基金
山东省科学技术发展计划(批准号:2012YD01058)资助
关键词
压缩感知
重构算法
稀疏度估计
变步长
自适应
compression sensing
reconstruction algorithm
sparsity estimate
variable step size
adaptive