Porous medium image reconstruction method based on generative network

A technique for porous media and image reconstruction, applied in the field of image processing, can solve problems such as difficulties, missing, incomplete 3D reconstruction and analysis of training images, and achieve fast and accurate reconstruction

Active Publication Date: 2020-09-29
SICHUAN UNIV
View PDF9 Cites 4 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

For non-stationary images, they are often difficult to reconstruct more realistic results
In addition, in practice, two-dimensional images usually do not allow users to choose, they may be partially missing and incomplete
The incompleteness of training images undoubtedly brings difficulties to subsequent 3D reconstruction and analysis based on them.

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
  • Porous medium image reconstruction method based on generative network
  • Porous medium image reconstruction method based on generative network
  • Porous medium image reconstruction method based on generative network

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0037] In order to make the reconstruction method described in the present invention easier to understand and closer to the real application, the following is a detailed overall description of each step of the reconstruction method based on deep learning. The specific operation steps are as follows:

[0038] (1) For different reconstructed images, make a data set for deep neural network training. image 3 Several samples from the dataset of rubber / silica materials are given (where white is silica and black is rubber). The data set has a total of 800 image pairs, and each image pair consists of two 128×128 images to be reconstructed and the target image.

[0039] (2) The calculation of the mode-based constraint function is as follows: Figure 4 shown. In the experiment, in order to balance the accuracy and speed of reconstruction, a template size of 3×3 is used. That is to say, there are at most 2 types of patterns 3×3 = 512 patterns. Count the frequency of these patterns ...

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 porous medium image reconstruction method based on a generative network. The invention provides a reconstruction method based on a generative network in order to solve the problem that a two-dimensional image is missing or incomplete. The main innovation of the method comprises the following steps: proposing to learn a mapping relationship between a local image and a complete image by using a deep generation network; designing a mode-based loss function and a porosity-based loss function; combining GAN loss and L1 loss, and setting different weights to jointly constrain the reconstruction process; and introducing gaussian noise to realize diversity of reconstruction results. For different porous media, corresponding data sets are made, and the effectiveness of themethod is verified by adopting a visual effect and a statistical function. The method is rapid, accurate and high in expansibility, can be used for reconstructing a multiphase medium, processing anisotropic image reconstruction, accelerating other reconstruction algorithms in combination with user-defined data and the like, and has a relatively good application value.

Description

technical field [0001] The present invention relates to a method for reconstructing a porous medium image based on a generative network, in particular to a method for reconstructing a complete image of a porous medium using extremely limited information, which belongs to the technical field of image processing. Background technique [0002] Porous media such as rocks, soils, and composite materials exist in nature and people's lives in large quantities, and are widely used in practical engineering applications. The microstructure of porous media directly determines the external macroscopic properties, so it is particularly important to understand their internal structure. [0003] Taking rocks as an example, there are two main ways to obtain images of their internal microstructure. One is to use 3D imaging techniques, such as computed tomography (CT), scanning electron microscope (SEM) and other technologies for imaging, to directly obtain 3D images of the core; the second ...

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): G06T5/50
CPCG06T5/50G06T2207/20081G06T2207/20084
Inventor 滕奇志冯俊羲何小海卿粼波吴小强吴晓红
Owner SICHUAN 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