Dynamic Matching Pursuit Based on Optimized Local Frequency Constraints

YIN Xingyao,XU Lu,ZONG Zhaoyun,LI Kun

DOI: https://doi.org/10.1190/segam2019-3214327.1

2019-01-01

SEG Technical Program Expanded Abstracts

Abstract:Dynamic matching pursuit takes instantaneous characteristics of the signal as a priori information, to reduce the computational complexity of the dictionary indexes and the number of iterations, but the instantaneous frequency result calculated by Hilbert transform exists negative value and is susceptible to seismic noise. For negative frequency sampling points, a global search of the frequency parameters of the atomic library is required, which is contrary to the dynamic search strategy. Considering that there is no negative value in the calculation results of the local frequency and it is more anti-noise, the paper introduces local frequency as frequency parameters to constrain the scanning range of the atomic library. For the problem that the local frequency has the same smoothing scale for all data points, we propose optimized local frequency by constructing a new difference matrix. The change enables it to perform different scale smoothing according to the value of the data. We apply the proposed method to the removal of the strong reflection and the spectral decomposition of the seismic data. The synthetic modeling test and application on field data illustrate the feasibility and stability of our method. Presentation Date: Tuesday, September 17, 2019 Session Start Time: 8:30 AM Presentation Time: 10:10 AM Location: 221B Presentation Type: Oral