dl1dnoise package is used to store the reproducible and open-access scripts for Single-channel microseismic denoising using dictionary learning, which can be straightforwardly applied to improve arrival picking and source location.
If you find this package useful, please do not forget to cite the following paper.
Chen, Y., Savvaidis, A., and Fomel, S., (2023). Dictionary learning for single-channel passive seismic denoising, Seismological Research Letters, 94, 2840-2851.
BibTeX:
@article{dl1dnoise,
author={Yangkang Chen and Alexandros Savvaidis and Sergey Fomel},
title = {Dictionary learning for single-channel passive seismic denoising},
journal={TBD},
year=2023,
volume={94},
number={6},
issue={6},
pages={2840-2851},
doi={10.1785/0220230169},
}
Authors of the dl1dnoise paper, 2021-present
GNU General Public License, Version 3
(http://www.gnu.org/copyleft/gpl.html)
Using the latest version
git clone https://github.com/chenyk1990/dl1dnoise
cd dl1dnoise
addpath(genpath('./')); #in Matlab command line
The "main" directory contains all runnable scripts test_figNO.m to reproduce all figures in the paper.
- MATseisdl (https://github.com/chenyk1990/MATseisdl)
The development team welcomes voluntary contributions from any open-source enthusiast.
If you want to make contribution to this project, feel free to contact the development team.
Regarding any questions, bugs, developments, or collaborations, please contact
Yangkang Chen
[email protected]
-
To run the reproducible scripts (test_xxxx.m), please first download the required package from: https://github.com/chenyk1990/MATseisdl.
-
Please put the MATseisdl package in the main directory and add its sub-directories to the Matlab toolbox path.
-
The scripts beginning with "test_" are runnable scripts.
-
The directory subroutines stores all the required subroutines.
-
The current version is based on Matlab. Future versions may also support Python and be optimized regarding computational efficiency.
-
All figures (except for fig1, which is a schematic plot, and fig11, which is based on Madagascar and powerpoint) in the dl1dnoise paper are in the following directory for a quick look (https://github.com/chenyk1990/dl1dnoise/tree/main/gallery/).
-
The label fonts are different across different Matlab versions and platforms (Linux, Max). The figures presented in the paper are from Mac-Pro Matlab 2022b.
The gallery figures of the dl1dnoise package can be found at https://github.com/chenyk1990/gallery/tree/main/dl1dnoise Each figure in the gallery directory corresponds to a test_figNO.m script in the "main" directory with the exactly the same file name (figNO.png).
These gallery figures are also presented below.
Figure 2 Generated by test_figs2_3.m
Figure 3 Generated by test_figs2_3.m
Figure 4 Generated by test_fig4.m
Figure 5 Generated by test_fig5.m
Figure 6 Generated by test_fig6.m
Figure 7 Generated by test_figs7_8.m
Figure 8 Generated by test_figs7_8.m
Figure 9 Generated by test_fig9.m
Figure 10 Generated by test_fig10.m
