Seismic noise suppression method based on unbalanced depth expectation block logarithm likelihood network

A technology of logarithmic likelihood and noise suppression, applied in neural learning methods, biological neural network models, image data processing, etc., to achieve the effects of avoiding errors, improving block denoising effect, and good denoising intensity

Active Publication Date: 2021-04-02
JILIN UNIV
View PDF6 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Solving the Desert Noise Suppression Problem with Unequalized Deep Desired Block Log-Likelihood Networks

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Seismic noise suppression method based on unbalanced depth expectation block logarithm likelihood network
  • Seismic noise suppression method based on unbalanced depth expectation block logarithm likelihood network
  • Seismic noise suppression method based on unbalanced depth expectation block logarithm likelihood network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0069] 1. Working conditions

[0070] Experimental platform of the present invention adopts Intel (R) Core (TM) i5-7500 CPU@3.40GHz 3.40GHz, internal memory is 8GB, runs the PC of Windows 7, and language is python language. The operating environment is python==3.7, torch==1.0.1, scipy==1.3.1 and matplotlib.

[0071] 2. Experimental content and result analysis

[0072] The experimental effect of the present invention is illustrated below by experiments on synthetic data and actual data in the field:

[0073] Such as figure 2 As shown, 100 channels of synthetic clean seismic data contain 4 signal axes, which are respectively generated by the main frequency of [19Hz18Hz 17Hz 16Hz] Reker wavelet, and the synthetic desert random noise is as follows image 3 shown. Figure 4 for will image 3 join in figure 2 The desert seismic data polluted by desert noise, the signal-to-noise ratio is -4dB. In this example, the denoising result of the method of the present invention is te...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a seismic noise suppression method based on an unbalanced depth expectation block logarithm likelihood network, belongs to the technical field of machine learning and seismic image processing, and aims to solve the problem that strong noise cannot be suppressed thoroughly due to the fact that regular term parameters in an expectation block logarithm likelihood algorithm only change along with overall noise variance. The invention provides the seismic noise suppression method based on an unbalanced depth expected block logarithm likelihood network, and the end-to-end denoising network consists of an expected block log-likelihood denoising main network and an unbalanced multilayer perceptron parameter estimation network, and takes a noisy seismic image as an input end. The clean seismic image is used as an output-end learning network parameter. Futhermore the multilayer perception paramter estimation parameter with unbalanced block signal-to-noise ratio is firstlyadopted. According to the method, accurate regular term parameters can be estimated for each block in the seismic image, the denoising intensity of each block can be better controlled, the block denoising effect is improved. The method is superior to a traditional seismic denoising algorithm in the aspects of desert strong noise suppression and signal detail maintenance.

Description

technical field [0001] The invention belongs to the technical field of machine learning and seismic image processing, and in particular relates to a method for suppressing seismic noise based on an unbalanced depth expectation block logarithmic likelihood network. Background technique [0002] Exploration of resources such as oil and natural gas has always been a hot spot for a long time. Seismic exploration is currently the main means of exploration and exploitation of underground energy such as oil and gas. However, due to the influence of the environment of the seismic exploration area, the collected seismic images are often mixed with a large amount of random noise. These noises seriously destroy the effective signal and increase the difficulty of extracting reflected seismic signals. Therefore, seismic random noise suppression is the fundamental problem in seismic data processing to improve seismic quality and extract underground structure information from noise-interfe...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06T5/00G06N3/08
CPCG06T5/002G06N3/08G06T2207/20081G06T2207/20084
Inventor 林红波马阳叶文海
Owner JILIN UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap