Rock identification method, system and apparatus, terminal, and readable storage medium

Patent metadata

Field	Value
Publication number	US-12573179-B2
Application number	US-202118264654-A
Country	US
Kind code	B2
Filing date	Sep 29, 2021
Priority date	Feb 8, 2021
Publication date	Mar 10, 2026
Grant date	Mar 10, 2026

Abstract

A method for rock identification includes the steps of receiving a rock slice image transmitted by an image acquisition device, generating a geometric feature, a mineral feature and a structural feature corresponding to a rock slice based on the rock slice image, and generating an identification result of the rock slice based on the geometric feature, the mineral feature and the structural feature. After the rock slice image is obtained, the rock slice image is subject to feature extraction based on three dimensions, i.e., the geometric feature, the mineral feature and the structural feature, the properties of rock are determined based on multiple dimensional features, and finally the identification result comprising a textual description is generated. After a microscopic visual image corresponding to the rock slice is obtained, feature extraction based on multiple dimensions is performed on the image, and the rock slice is identified with reference to multi-dimensional features.

First claim

The invention claimed is:

1 . A rock identification method, comprising:
receiving a rock slice image obtained by photographing a rock slice obtained by cutting a rock sample, the rock slice image including at least one component region; generating a geometric feature, a mineral feature and a structural feature corresponding to the rock slice based on the rock slice image, wherein the geometric feature is used to indicate division of the component region of the rock slice, the mineral feature is used to indicate distribution of mineral type corresponding to the component region in the rock slice and a mineral identification result obtained therefrom, and the structural feature is used to indicate rock type of the rock sample; and generating an identification result of the rock slice based on the geometric feature, the mineral feature and the structural feature, wherein the structural feature comprises a structural feature of pre-selected rock type and a structural feature of subdivided rock type, wherein the step of generating the geometric feature, the mineral feature and the structural feature corresponding to the rock slice based on the rock slice image comprises: generating the structural feature of pre-selected rock type based on the rock slice image, wherein the structural feature of pre-selected rock type is used to indicate a pre-selected rock type corresponding to the rock sample, the pre-selected rock type comprising sedimentary rock, magmatic rock, and metamorphic rock; determining the pre-selected rock type corresponding to the rock slice based on the structural feature of pre-selected rock type; generating the geometric feature based on the rock slice image according to the pre-selected rock type; and generating the mineral feature based on the rock slice image according to the pre-selected rock type, and determining the structural feature of subdivided rock type based on the geometric feature and the mineral feature, for indicating a subdivided rock type corresponding to the rock sample.

2 . The method according to claim 1 , wherein the rock slice image comprises at least two component images characterizing the component region, wherein the step of generating the geometric feature, the mineral feature and the structural feature corresponding to the rock slice based on the rock slice image comprises:
dividing the rock slice image based on positions of the component images to obtain a rock slice segmentation image comprising at least two segmentation regions, which indicates segmentation of the rock slice image based on the component region; determining the geometric feature corresponding to the rock slice based on the segmentation regions; and determining the mineral feature corresponding to the rock slice based on the segmentation regions.

3 . The method according to claim 2 , wherein the step of dividing the rock slice image based on the positions of the component images to obtain the rock slice segmentation image comprises:
inputting the rock slice image into a rock slice image segmentation model to output the rock slice segmentation image, wherein the rock slice image segmentation model is a Mask-RCNN network model based on machine learning.

4 . The method according to claim 2 , wherein the step of determining the mineral feature corresponding to the rock slice based on the segmentation regions comprises:
determining each component image based on each segmentation region, each component image comprising one component region; inputting the component image into a component type identification model to output a component type feature of the component image for characterizing the mineral feature of the component region, wherein the component type identification model is a neural network model based on machine learning; and determining the mineral feature corresponding to the rock slice based on the component type feature of the component image.

5 . The method according to claim 2 , wherein the step of determining the mineral feature corresponding to the rock slice based on the segmentation regions comprises:
inputting the rock slice segmentation image into a classification and identification model to output a monocrystalline component collection image and a non-monocrystalline component collection image, wherein the monocrystalline component collection image includes at least one monocrystalline component region and a monocrystalline component identification result corresponding thereto, and the non-monocrystalline component region includes at least one non-monocrystalline component region and a non-monocrystalline component identification result corresponding thereto, the classification and identification model being a neural network model based on deep learning; transmitting a scanning instruction to a mineral data acquisition device based on the monocrystalline component collection image; receiving scanning data from the mineral data acquisition device based on the scanning instruction; verifying the monocrystalline component identification result based on the scanning data to obtain a verification result; and determining the mineral feature corresponding to the rock slice based on the non-monocrystalline component identification result, the monocrystalline component identification result and the verification result.

6 . The method according to claim 5 , wherein the step of transmitting a scanning instruction to the mineral data acquisition device based on the monocrystalline component collection image comprises:
inputting the monocrystalline component collection image into a region selecting model to output a region selecting result, which is used to indicate a portion of the monocrystalline component collection image for surface scanning data; generating the scanning instruction based on the region selecting result; and transmitting the scanning instruction to the mineral data acquisition device.

7 . The method according to claim 5 , wherein the step of verifying the monocrystalline component identification result based on the scanning data to obtain the verification result comprises:
determining a scanning identification result based on the scanning data; determining, in response to the scanning identification result which is the same as the monocrystalline component identification result, that the verification result indicates that the verification is passed, and that the monocrystalline component identification result is unchanged; and determining, in response to the surface scanning identification result which is different from the monocrystalline component identification result, that the verification result indicates that the verification fails, and determining the monocrystalline component identification result based on an identification result generation rule including a rule based on a type of the mineral data acquisition device.

8 . The method according to claim 5 , wherein the step of determining the mineral feature corresponding to the rock slice based on the segmentation region comprises:
determining each component image based on each segmentation region, each component image including one component region; transmitting a spectral data acquisition instruction to the mineral data acquisition device based on the component image; receiving the spectral data transmitted by the mineral data acquisition device based on the spectral data acquisition instruction; determining, based on the spectral data, a pre-selected mineral type corresponding to the component region in a mineral spectral data database; determining a mineral type verification rule corresponding to the pre-selected mineral type; determining a mineral type corresponding to the component region based on the mineral t

Assignees

• China Petroleum & Chem Corp
• Sinopec Exploration & Prod Res Inst

Inventors

• Yu Xiaolu
• Zhao Yongqiang
• Ni Chunhua
• Ma Zhongliang
• Zhou Shengyou
• Yang Weili
• Zhang Jun
• Tao Cheng
• Li Kuang
• Wang Qiang
• Zheng Lunju

Classifications

• G06V10/40

  Extraction of image or video features · CPC title

• G06V10/82

  using neural networks · CPC title

• G06V10/26

  Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion · CPC title

• G06V10/14

  Optical characteristics of the device performing the acquisition or on the illumination arrangements · CPC title

• G06V20/60

  Type of objects · CPC title