← Back to Sessions

Seismic Data Denoising Using Multi-Scale Mathematical Morphological Filtering

Description: 

Recorded seismic data are generally contaminated by noise from different sources, which masks the signals of interest. In the seismology community, frequency filtering is the standard method for noise suppression. However, frequency filtering is inadequate when the signal of interest and noise share the same frequency band. We implemented a new data denoising technique based on the mathematical morphology theorem. The method uses morphological opening and closing operations with respect to predefined structuring elements of varying scales, and decomposes an input noisy waveform into several time series with differing characteristics. Using weighted stacking the denoised waveform is constructed from the time series. The reconstruction factors used to perform the weighted stacking are chosen carefully based on the assessment of which scales are likely to enhance the signal of interest and which ones are likely to amplify the noise. We will discuss the implementation of the approach and its initial application to process high-frequency local and near-regional data.

Session: Advancements in Forensic Seismology and Explosion Monitoring [Poster Session]

Type: Poster

Date: 5/2/2024

Presentation Time: 08:00 AM (local time)

Presenting Author: Rigobert

Student Presenter: No

Invited Presentation: 

Authors