Structure-Preserving Random Noise Attenuation Method for Seismic Data Based on a Flexible Attention CNN
The noise attenuation of seismic data is an indispensable part of seismic data processing, directly impacting the following inversion and imaging. This paper focuses on two bottlenecks in the AI-based denoising method of seismic data: the destruction of structural information of seismic data and the...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-10-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/20/5240 |
| _version_ | 1827648089954975744 |
|---|---|
| author | Wenda Li Tianqi Wu Hong Liu |
| author_facet | Wenda Li Tianqi Wu Hong Liu |
| author_sort | Wenda Li |
| collection | DOAJ |
| description | The noise attenuation of seismic data is an indispensable part of seismic data processing, directly impacting the following inversion and imaging. This paper focuses on two bottlenecks in the AI-based denoising method of seismic data: the destruction of structural information of seismic data and the inferior generalizability. We propose a flexible attention-CNN (FACNN) and realized the denoising work of seismic data. This paper’s main work and advantages were concentrated on the following three aspects: (i) We propose attention gates (AGs), which progressively suppressed features in irrelevant background parts and improved the denoising performance. (ii) We added a noise level map M as an additional channel, making a single CNN model expected to inherit the flexibility of handling noise models with different parameters, even spatially variant noises. (iii) We propose a mixed loss function based on <i>MS_SSIM</i> to improve the performance of FACNN further. Adding the noise level map can improve the network’s generalization ability, and adding the attention structure with the mixed loss function can better protect the structural information of the seismic data. The numerical tests showed that our method has better generalization and can better protect the details of seismic events. |
| first_indexed | 2024-03-09T19:31:34Z |
| format | Article |
| id | doaj.art-4631d674038b4a1aa1b3dc4a4711da51 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T19:31:34Z |
| publishDate | 2022-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-4631d674038b4a1aa1b3dc4a4711da512023-11-24T02:21:56ZengMDPI AGRemote Sensing2072-42922022-10-011420524010.3390/rs14205240Structure-Preserving Random Noise Attenuation Method for Seismic Data Based on a Flexible Attention CNNWenda Li0Tianqi Wu1Hong Liu2Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100864, ChinaInstitute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100864, ChinaInstitute of Geology and Geophysics, Chinese Academy of Sciences, Beijing 100864, ChinaThe noise attenuation of seismic data is an indispensable part of seismic data processing, directly impacting the following inversion and imaging. This paper focuses on two bottlenecks in the AI-based denoising method of seismic data: the destruction of structural information of seismic data and the inferior generalizability. We propose a flexible attention-CNN (FACNN) and realized the denoising work of seismic data. This paper’s main work and advantages were concentrated on the following three aspects: (i) We propose attention gates (AGs), which progressively suppressed features in irrelevant background parts and improved the denoising performance. (ii) We added a noise level map M as an additional channel, making a single CNN model expected to inherit the flexibility of handling noise models with different parameters, even spatially variant noises. (iii) We propose a mixed loss function based on <i>MS_SSIM</i> to improve the performance of FACNN further. Adding the noise level map can improve the network’s generalization ability, and adding the attention structure with the mixed loss function can better protect the structural information of the seismic data. The numerical tests showed that our method has better generalization and can better protect the details of seismic events.https://www.mdpi.com/2072-4292/14/20/5240random noiseattentionnoise levelstructure-preservingdenoising |
| spellingShingle | Wenda Li Tianqi Wu Hong Liu Structure-Preserving Random Noise Attenuation Method for Seismic Data Based on a Flexible Attention CNN Remote Sensing random noise attention noise level structure-preserving denoising |
| title | Structure-Preserving Random Noise Attenuation Method for Seismic Data Based on a Flexible Attention CNN |
| title_full | Structure-Preserving Random Noise Attenuation Method for Seismic Data Based on a Flexible Attention CNN |
| title_fullStr | Structure-Preserving Random Noise Attenuation Method for Seismic Data Based on a Flexible Attention CNN |
| title_full_unstemmed | Structure-Preserving Random Noise Attenuation Method for Seismic Data Based on a Flexible Attention CNN |
| title_short | Structure-Preserving Random Noise Attenuation Method for Seismic Data Based on a Flexible Attention CNN |
| title_sort | structure preserving random noise attenuation method for seismic data based on a flexible attention cnn |
| topic | random noise attention noise level structure-preserving denoising |
| url | https://www.mdpi.com/2072-4292/14/20/5240 |
| work_keys_str_mv | AT wendali structurepreservingrandomnoiseattenuationmethodforseismicdatabasedonaflexibleattentioncnn AT tianqiwu structurepreservingrandomnoiseattenuationmethodforseismicdatabasedonaflexibleattentioncnn AT hongliu structurepreservingrandomnoiseattenuationmethodforseismicdatabasedonaflexibleattentioncnn |