Shale three-dimensional super-resolution digital core grading reconstruction method and system

Abstract

The invention provides a shale three-dimensional super-resolution digital core grading reconstruction method and system, and relates to the technical field of oil and gas field development, and the method comprises the steps: obtaining a plurality of horizontal to-be-reconstructed images of a to-be-detected region; inputting the plurality of horizontal to-be-reconstructed images into the image enhancement model to obtain a plurality of horizontal enhanced images and a plurality of vertical enhanced images of the to-be-detected area, wherein the image enhancement model comprises a first image enhancement sub-model and a second image enhancement sub-model; and performing three-dimensional reconstruction on the to-be-detected area according to the plurality of horizontal enhanced images and the plurality of vertical enhanced images to obtain a shale three-dimensional digital core of the to-be-detected area. The vertical direction features and the horizontal direction features of the shale digital core in the to-be-detected area are respectively trained, the vertical direction and the horizontal direction of the to-be-reconstructed image are amplified with different multiplying powers, and the accuracy of shale digital core reconstruction is improved.

Description

technical field [0001] The invention relates to the technical field of oil and gas field development, in particular to a shale three-dimensional super-resolution digital core classification reconstruction method and system. Background technique [0002] Shale reservoirs contain a large number of micro-nano pores, and their structural information is of great significance for determining underground crude oil reserves and crude oil flow capacity. Currently, focused ion beam-scanning electron microscopy (FIB-SEM) is one of the most powerful tools for characterizing the three-dimensional pore structure of shale. Using FIB (Focused Ion beam technology, Focused Ion beam) to continuously cut the surface of the shale sample, and at the same time image it with a scanning electron microscope (SEM); then superimpose multiple SEM images to obtain the three-dimensional pore structure of the core (digital core) . Therefore, the resolution in the horizontal direction (x direction and y d...

AI Technical Summary

Problems solved by technology

Patent CN105006018A relates to the method of super-resolution reconstruction of 3D CT core images. The sparse coding method is used to perform super-resolution reconstruction of 3D cores. This method has a large amount of calculation and takes a long time. It has been replaced by deep learning algorithms

Patent CN108898560A relates to a method for super-resolution reconstruction of rock core CT images based on a three-dimensional convolutional neural network. This method can improve the resolution of three-dimensional images of rock cores. The magnifications in each direction are consistent, and it is impossible to achieve different magnifications in the horizontal and vertical directions, which will lead to insufficient digital core accuracy or image redundancy

Patent CN112308781A relates to a single image 3D super-resolution reconstruction method based on deep learning. This method uses a single image to achieve 3D digital core reconstruction. However, due to insufficient training data, the model often leads to overfitting, so The accuracy of the pore structure of the generated digital core cannot be guaranteed

Images