Riker子波匹配追踪算法及其改进

匹配追踪（Matching Pursuit）算法的基本思想是基于信号的可分解和重构,是在一个确定的函数集合中自适应地选择一些函数来表示一个信号的计算过程,函数集合中的每个函数都称为原子。本文利用奇异值分解对传统的匹配追踪算法进行了改进,提高收敛速度、计算速度以及重构精度,并将得到的时频分布与其他方法进行对比,测验结果证明了改进算法的高效性和有效性。

Revised Matching Pursuit Algorithm Using Ricker Wavelets

The basic idea of Matching Pursuit algorithm is based on the possibility of decomposition and reconstruction of seismic data.It is a process in which a series of functions can be chosen from a certain function set to represent a given signal,and the each and every function within the function set is called as an atom.This paper uses singular value decomposition method to revise matching pursuit algorithm,and improves the speed of convergence and computation,and then compares the time-frequency distribution ...