| 基于小波树模型的改进SP算法 |
| |
| 引用本文： | 袁静. 基于小波树模型的改进SP算法[J]. 电声技术, 2014, 38(12): 61-64 |
| |
| 作者姓名： | 袁静 |
| |
| 作者单位： | 宿迁学院,江苏宿迁223800;南京邮电大学信号处理与传输研究院,江苏南京210003 |
| |
| 基金项目： | 宿迁市科技创新基金项目，国家自然科学基金项目 |
| |
| 摘 要： | 针对传统的子空间追踪算法(SP)只利用了信号在某个字典下是稀疏的或者可压缩的这个简单的先验知识,没有将信号的内在模型考虑进去,因此重构效率较低的问题。根据一般信号的小波树系数的特点,提出了一种基于小波树模型的改进子空间追踪算法。由于引入了信号的小波树内在模型,使得改进后算法中得到的最佳K项小波树结构稀疏逼近比子空间追踪算法中的最佳K项稀疏逼近更加接近于原信号,实验仿真证明基于小波树模型的SP算法的重构性更好。
|
| 关 键 词： | 压缩感知 小波树形 子空间追踪算法 |
Improved Subspace Pursuit Algorithm Based on Wavelet Tree Model |
| |
| YUAN Jing. Improved Subspace Pursuit Algorithm Based on Wavelet Tree Model[J]. Audio Engineering, 2014, 38(12): 61-64 |
| |
| Authors: | YUAN Jing |
| |
| Affiliation: | YUAN Jing(1. Suqian College ,Suqian Jiangsu 223800 ,China;2. Institute of Signal Processing and Transmission, Nanjing University of Posts and Telecommunications,Nanjing 210003, China) |
| |
| Abstract: | Because traditional subspace pursuit algorithm （SP） hasn＇ t taken the signal＇ s inherent model into consideration, it has a low efficiency. An improved subspace pursuit algorithm based on wavelet tree model is prooposed,using the character- istic of signal ＇ s wavelet tree＇ s coefficients. Owing to having introduced the inherent model of the signal into reconstruction al- gorithm, the best K term wavelet tree structure sparse approximation from the improved algorithm is much closer to the original signal than SP arithmetic. Experimental results show the better performance than the subspace pursuit algorithm. |
| |
| Keywords: | compressed sensing wavelet tree subspace pursuit algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
