CN113848191A - Intelligent sandstone classification method based on spectrum - Google Patents
Intelligent sandstone classification method based on spectrum Download PDFInfo
- Publication number
- CN113848191A CN113848191A CN202111242822.7A CN202111242822A CN113848191A CN 113848191 A CN113848191 A CN 113848191A CN 202111242822 A CN202111242822 A CN 202111242822A CN 113848191 A CN113848191 A CN 113848191A
- Authority
- CN
- China
- Prior art keywords
- sandstone
- constructing
- layer
- spectrum
- parameters
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001228 spectrum Methods 0.000 title claims abstract description 28
- 238000000034 method Methods 0.000 title claims abstract description 24
- 238000013528 artificial neural network Methods 0.000 claims abstract description 18
- 230000003595 spectral effect Effects 0.000 claims abstract description 13
- 238000002310 reflectometry Methods 0.000 claims description 17
- 238000005259 measurement Methods 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 3
- 102220052459 rs200562715 Human genes 0.000 claims description 3
- 102220042973 rs35903915 Human genes 0.000 claims description 3
- 102220101353 rs61735044 Human genes 0.000 claims description 3
- 102220208211 rs767215118 Human genes 0.000 claims description 3
- 230000003287 optical effect Effects 0.000 claims 2
- 238000010521 absorption reaction Methods 0.000 abstract description 4
- 229910052500 inorganic mineral Inorganic materials 0.000 description 4
- 239000011707 mineral Substances 0.000 description 4
- 239000011435 rock Substances 0.000 description 4
- 230000000694 effects Effects 0.000 description 2
- 208000035126 Facies Diseases 0.000 description 1
- 235000013339 cereals Nutrition 0.000 description 1
- 238000005553 drilling Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 235000013312 flour Nutrition 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 230000033116 oxidation-reduction process Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000035699 permeability Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/55—Specular reflectivity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Evolutionary Biology (AREA)
- Software Systems (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention belongs to the field of geological environment, and particularly discloses an intelligent sandstone classification method based on a spectrum. The method comprises the following steps: step 1, classifying and numbering sandstone samples; step 2, measuring the spectrum of the sample; step 3, constructing parameters I1-I7; step 4, constructing a parameter I8; step 5, constructing parameters I9-I11; step 6, constructing parameters I12-I14; step 7, constructing a 3-layer neural network; step 8, constructing a tag data set; step 9, the tag data set is scrambled; step 10, inputting the disordered label data set into a 3-layer neural network for training; step 11, measuring a reflection spectrum of unknown sandstone and calculating I1-I14; and step 12, inputting I8-I14 of unknown sandstone into the trained 3-layer neural network, and determining the sandstone type according to the output result. The method combines the spectral shape and the absorption characteristics, has few input parameters, high training speed and more sandstone type resolution.
Description
Technical Field
The invention belongs to the field of geological environment, and particularly discloses an intelligent sandstone classification method based on a spectrum.
Background
Sandstone is an important sedimentary rock, and the petrological characteristics of the sandstone have very important indicating significance for environment, resources and mineral products. For example, the deep oxidation-reduction environment can be judged according to the color of sandstone, the ancient depositional environment and the sedimentary facies type can be reflected by utilizing the granularity characteristics, and the porosity represents the permeability of underground water resources and the storage capacity of oil and gas resources. Therefore, it is very important to know the type of sandstone.
The conventional sandstone type identification is completed by a manual mode, and is completed by collecting sandstone samples and combining laboratory analysis or by utilizing field observation and combining a drilling and recording mode, so that the efficiency is low, the sandstone type identification is easily influenced by subjective factors, and errors are easily caused. The method has the advantages that the acquisition speed of the ground object reflection spectrum is high, the method is widely applied to rock classification and mineral identification, the method is influenced by spectral resolution, the type of the identified rock is limited, and the accuracy is low. Although the ground object spectrum with high spectral resolution can improve the capability of distinguishing rock types, the number of wave bands is large, the data processing difficulty is high, the identification working efficiency is low, manual intervention is still needed, and the labor cost is high.
The neural network technology can be used for learning different types of surface feature spectra through sample training and automatically identifying unknown samples according to learning results, so that manual intervention is reduced, labor cost is reduced, and the efficiency of identification work is improved. However, in addition to the requirement of sufficient training samples, neural network training has a great dependence on the input parameters of the samples. The number of spectral bands of the ground object spectrum with high spectral resolution is large, but many spectral bands have insignificant characteristics, and the training efficiency and effect are seriously reduced if all the spectral bands are used as input parameters. Therefore, the neural network can be trained better only by selecting a representative spectrum segment, specially processing the input parameters and enhancing the representativeness of the input parameters to the sandstone types, and a better recognition effect can be obtained in the future.
Sandstone types can be classified in terms of both color, including red, magenta, dark red, yellow, white, gray, green, etc., and grain size, including coarse, medium, fine, flour, mud, etc. The particle size of sandstone influences the overall morphological characteristics of the reflection spectrum, while the color of sandstone is mostly related to the type of minerals contained in sandstone, and the type of minerals influences the absorption peak characteristics of the reflection spectrum. Therefore, in order to better distinguish the sandstone types, the two aspects of the line shape and the absorption peak characteristics must be considered.
Disclosure of Invention
The invention aims to provide a sandstone intelligent classification method based on a spectrum, which is characterized in that the spectrum is found to have high reflection, medium reflection and low reflection characteristics by measuring the spectrum of approximately 10000 sandstone samples, the curve form is characterized by left slant, right slant, peace, left slant and right slant, the curve is stable at 400nm, 600nm, 750nm and 950nm after removing the envelope line, and has the maximum change at 520nm, 650nm and 860nm, so that the reflectivity of a specific wavelength position is selected, and characteristic input parameters are formed by calculating the mutual relation among the reflectivities, so that the form and the absorption characteristics of the spectrum curve are highlighted to the greatest extent under the condition of ensuring that the input parameters are reduced.
The technical scheme for realizing the purpose of the invention is as follows: an intelligent sandstone classification method based on spectrum is characterized in that: the method comprises the following steps:
step 1, classifying sandstone samples according to color and granularity, and numbering each type;
step 2, performing spectral measurement on different types of sandstone samples by using a spectrometer to obtain a reflectivity spectral curve of the sandstone samples within the range of 400-1000 nm;
step 3, constructing reflectivity values of parameters I1, I2, I3, I4, I5, I6 and I7 which are respectively equal to the reflectivity values at 400nm, 520nm, 600nm, 650nm, 750nm, 860nm and 950 nm;
step 4, constructing a parameter I8, wherein I8 is the maximum value of I1, I3, I5 and I7;
step 5, constructing parameters I9, I10, I11, I9= I3/I1, I10= I5/I3, and I11= I7/I5;
step 6, constructing parameters I12, I13, I14, I12= (I1+ I3)/(2 × I2), I13= (I3+ I5)/(2 × I4), I14= (I5+ I7)/(2 × I6);
step 7, constructing a 3-layer neural network, wherein the 3 layers are an input layer, a hidden layer and an output layer respectively, the number of input parameters of the input layer is 7, the input parameters correspond to I8, I9, I10, I11, I12, I13 and I14 respectively, the number of output parameters of the output layer is consistent with the type number of the sandstone samples, and the number of neurons of the hidden layer is also consistent with the type number of the sandstone samples;
step 8, constructing a tag data set;
step 9, the label data set in the step 8 is disturbed;
step 10, inputting the label data set disturbed in the step 9 into the layer 3 neural network in the step 7 for neural network training;
step 11, measuring the reflectivity spectrum of an unknown sandstone sample, and calculating I1-I14 according to the method in the steps 3-6;
and 12, inputting the I8-I14 in the step 11 into the 3-layer neural network trained in the step 10, and determining the type of the sandstone according to the output result of an output layer.
The tag data set in the step 8 comprises data and tags, the data part is I8-I14, the tag part is a binary number sequence with the same number of elements as the number of sandstone types, and the first sandstone in the number sequence is represented in the following way: {1,0,0,0, … }, the second class being: {0,1,0,0, … }, with the third being: {0,0,1,0, … }, and so on, when constructing the label dataset, sequentially calculating the I8-I14 and the label binary number sequence of each sandstone sample from the reflectivity spectrum of the first sandstone sample, and finally forming the label dataset of all sandstone samples.
The method for determining the sandstone category in the step 12 comprises the following steps: and (3) according to the number sequence output by the output layer, wherein the sequence number corresponding to the maximum element is the type number of the unknown sandstone, and the type of the unknown sandstone can be determined according to the number and the sandstone type corresponding to the number in the step (1).
Detailed Description
The present invention will be described in further detail with reference to examples.
The invention provides a spectrum-based intelligent sandstone classification method, which comprises the following steps:
step 1, dividing sandstone samples into 9 types of red coarse sandstone, red middle sandstone, red fine sandstone, yellow coarse sandstone, yellow middle sandstone, yellow fine sandstone, white coarse sandstone, white middle sandstone and white fine sandstone, and numbering the types of the sandstone samples as 1-9;
step 2, performing spectral measurement on different types of sandstone samples by using a spectrometer to obtain a reflectivity spectral curve of the sandstone samples within the range of 400-1000 nm;
step 3, constructing reflectivity values of parameters I1, I2, I3, I4, I5, I6 and I7 which are respectively equal to the reflectivity values at 400nm, 520nm, 600nm, 650nm, 750nm, 860nm and 950 nm;
step 4, constructing a parameter I8, wherein I8 is the maximum value of I1, I3, I5 and I7;
step 5, constructing parameters I9, I10, I11, I9= I3/I1, I10= I5/I3, and I11= I7/I5;
step 6, constructing parameters I12, I13, I14, I12= (I1+ I3)/(2 × I2), I13= (I3+ I5)/(2 × I4), I14= (I5+ I7)/(2 × I6);
step 7, constructing a 3-layer neural network, wherein the 3 layers are an input layer, a hidden layer and an output layer respectively, the number of input parameters of the input layer is 7, the input parameters correspond to I8, I9, I10, I11, I12, I13 and I14 respectively, the number of output parameters of the output layer is 9, and the number of neurons of the hidden layer is 9;
step 8, sequentially calculating I8-I14 and a label binary number sequence of each sandstone sample from the reflectivity spectrum of the first sandstone sample, wherein the red coarse sandstone is {1,0,0,0,0,0,0,0,0}, the red medium sandstone is {0,1,0,0,0,0,0,0,0}, and so on, and finally forming a label data set of all sandstone samples;
step 9, the label data set in the step 8 is disturbed;
step 10, inputting the label data set disturbed in the step 9 into the layer 3 neural network in the step 7 for neural network training;
step 11, measuring the reflectivity spectrum of an unknown sandstone sample, and calculating I1-I14 according to the method in the steps 3-6;
and 12, inputting the I8-I14 in the step 11 into the 3-layer neural network trained in the step 10, wherein the output result of an output layer is {0.43,0.52,0.74,0.93,0.82,0.25,0.67,0.79 and 0.89}, the maximum value is 0.93, the corresponding sequence number is 4, the unknown sandstone is represented as 4 th sandstone, and the 4 th sandstone is known to be yellow sandstone by combining the step 1, so that the type of the unknown sandstone is yellow sandstone.
The present invention has been described in detail with reference to the embodiments, but the present invention is not limited to the embodiments, and various changes can be made without departing from the gist of the present invention within the knowledge of those skilled in the art. The prior art can be adopted in the content which is not described in detail in the invention.
Claims (3)
1. An intelligent sandstone classification method based on spectrum is characterized in that: the method comprises the following steps:
step 1, classifying sandstone samples according to color and granularity, and numbering each type;
step 2, performing spectral measurement on different types of sandstone samples by using a spectrometer to obtain a reflectivity spectral curve of the sandstone samples within the range of 400-1000 nm;
step 3, constructing reflectivity values of parameters I1, I2, I3, I4, I5, I6 and I7 which are respectively equal to the reflectivity values at 400nm, 520nm, 600nm, 650nm, 750nm, 860nm and 950 nm;
step 4, constructing a parameter I8, wherein I8 is the maximum value of I1, I3, I5 and I7;
step 5, constructing parameters I9, I10, I11, I9= I3/I1, I10= I5/I3, and I11= I7/I5;
step 6, constructing parameters I12, I13, I14, I12= (I1+ I3)/(2 × I2), I13= (I3+ I5)/(2 × I4), I14= (I5+ I7)/(2 × I6);
step 7, constructing a 3-layer neural network, wherein the 3 layers are an input layer, a hidden layer and an output layer respectively, the number of input parameters of the input layer is 7, the input parameters correspond to I8, I9, I10, I11, I12, I13 and I14 respectively, the number of output parameters of the output layer is consistent with the type number of the sandstone samples, and the number of neurons of the hidden layer is also consistent with the type number of the sandstone samples;
step 8, constructing a tag data set;
step 9, the label data set in the step 8 is disturbed;
step 10, inputting the label data set disturbed in the step 9 into the layer 3 neural network in the step 7 for neural network training;
step 11, measuring the reflectivity spectrum of an unknown sandstone sample, and calculating I1-I14 according to the method in the steps 3-6;
and 12, inputting the I8-I14 in the step 11 into the 3-layer neural network trained in the step 10, and determining the type of the sandstone according to the output result of an output layer.
2. The intelligent sandstone classification method based on optical spectrum as claimed in claim 1, wherein the method comprises the following steps: the tag data set in the step 8 comprises data and tags, the data part is I8-I14, the tag part is a binary number sequence with the same number of elements as the number of sandstone types, and the first sandstone in the number sequence is represented in the following way: {1,0,0,0, … }, the second class being: {0,1,0,0, … }, with the third being: {0,0,1,0, … }, and so on, when constructing the label dataset, sequentially calculating the I8-I14 and the label binary number sequence of each sandstone sample from the reflectivity spectrum of the first sandstone sample, and finally forming the label dataset of all sandstone samples.
3. The intelligent sandstone classification method based on optical spectrum as claimed in claim 1, wherein the method comprises the following steps: the method for determining the sandstone category in the step 12 comprises the following steps: and (3) according to the number sequence output by the output layer, wherein the sequence number corresponding to the maximum element is the type number of the unknown sandstone, and the type of the unknown sandstone can be determined according to the number and the sandstone type corresponding to the number in the step (1).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111242822.7A CN113848191A (en) | 2021-10-26 | 2021-10-26 | Intelligent sandstone classification method based on spectrum |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111242822.7A CN113848191A (en) | 2021-10-26 | 2021-10-26 | Intelligent sandstone classification method based on spectrum |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113848191A true CN113848191A (en) | 2021-12-28 |
Family
ID=78982862
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111242822.7A Pending CN113848191A (en) | 2021-10-26 | 2021-10-26 | Intelligent sandstone classification method based on spectrum |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113848191A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101040184A (en) * | 2004-09-07 | 2007-09-19 | 彼得罗模型公司 | Apparatus and method for analysis of size, form and angularity and for compositional analysis of mineral and rock particles |
| CN106650819A (en) * | 2016-12-30 | 2017-05-10 | 东北农业大学 | Soil classification method through combination of multi-layer perceptron neural networks with spectral characteristic parameters |
| CN110852395A (en) * | 2019-11-15 | 2020-02-28 | 鞍钢集团矿业有限公司 | Ore granularity detection method and system based on autonomous learning and deep learning |
| CN113189021A (en) * | 2021-05-11 | 2021-07-30 | 自然资源实物地质资料中心 | Method for identifying rock color based on spectrum |
-
2021
- 2021-10-26 CN CN202111242822.7A patent/CN113848191A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101040184A (en) * | 2004-09-07 | 2007-09-19 | 彼得罗模型公司 | Apparatus and method for analysis of size, form and angularity and for compositional analysis of mineral and rock particles |
| CN106650819A (en) * | 2016-12-30 | 2017-05-10 | 东北农业大学 | Soil classification method through combination of multi-layer perceptron neural networks with spectral characteristic parameters |
| CN110852395A (en) * | 2019-11-15 | 2020-02-28 | 鞍钢集团矿业有限公司 | Ore granularity detection method and system based on autonomous learning and deep learning |
| CN113189021A (en) * | 2021-05-11 | 2021-07-30 | 自然资源实物地质资料中心 | Method for identifying rock color based on spectrum |
Non-Patent Citations (1)
| Title |
|---|
| 杨柏林 等: "岩石和矿物的反射光谱特征及其应用", 地球化学, no. 1 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108458989B (en) | Terahertz multi-parameter spectrum-based coal rock identification method | |
| CN107727592B (en) | Coal rock interface identification method based on high spectral reflectance characteristics of coal rock | |
| Carranza et al. | Mineral imaging with Landsat Thematic Mapper data for hydrothermal alteration mapping in heavily vegetated terrane | |
| Galvdo et al. | Relationships of spectral reflectance and color among surface and subsurface horizons of tropical soil profiles | |
| CN102183500B (en) | White spirit identification method based on fluorescence characteristic parameter Euclidean distance | |
| CN112766227A (en) | Hyperspectral remote sensing image classification method, device, equipment and storage medium | |
| CN101806898A (en) | Hyperspectral remote sensing image target detecting method based on variable end members | |
| CN105548066A (en) | Method and system for distinguishing colloid types | |
| CN108195771A (en) | A kind of ocean target in hyperspectral remotely sensed image target identification method | |
| Huo et al. | Novel lithology identification method for drilling cuttings under PDC bit condition | |
| Pu | Detecting and Mapping Invasive Plant Species Using Hyperspectral Data | |
| Ishidoshiro et al. | Geological mapping by combining spectral unmixing and cluster analysis for hyperspectral data | |
| CN113408616A (en) | Spectrum classification method based on PCA-UVE-ELM | |
| CN110031414A (en) | A kind of multi-layer perception (MLP) high spectrum mineral classification method based on spectral absorption index | |
| CN113848191A (en) | Intelligent sandstone classification method based on spectrum | |
| Meer | Classification of remotely-sensed imagery using an indicator kriging approach: Application to the problem of calcite-dolomite mineral mapping | |
| CN108827909A (en) | Soil rapid classification method based on visible and near infrared spectrum and multiple targets fusion | |
| Roden et al. | Seismic interpretation with machine learning | |
| CN115908890A (en) | Intelligent lithology recognition method for water-bearing rock by combining image recognition and spectral analysis | |
| Charles et al. | Sparse coding for spectral signatures in hyperspectral images | |
| Prieto et al. | An adaptive approach for the progressive integration of spatial and spectral features when training ground-based hyperspectral imaging classifiers | |
| WO2022261712A1 (en) | A method and system for logging data for a mineral sample | |
| Ilehag et al. | Revisiting existing classification approaches for building materials based on hyperspectral data | |
| CN113640244A (en) | Fruit tree variety identification method based on visible near infrared spectrum | |
| Schneider et al. | Classification of hyperspectral imagery using GPs and the OAD covariance function with automated endmember extraction |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20211228 |