CN113128404A - Intelligent mineral identification method and system - Google Patents
Intelligent mineral identification method and system Download PDFInfo
- Publication number
- CN113128404A CN113128404A CN202110423576.9A CN202110423576A CN113128404A CN 113128404 A CN113128404 A CN 113128404A CN 202110423576 A CN202110423576 A CN 202110423576A CN 113128404 A CN113128404 A CN 113128404A
- Authority
- CN
- China
- Prior art keywords
- characteristic
- mineral
- information
- component information
- training
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- 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
- G06N3/045—Combinations of networks
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
- G06V10/464—Salient features, e.g. scale invariant feature transforms [SIFT] using a plurality of salient features, e.g. bag-of-words [BoW] representations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Spectrometry And Color Measurement (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to an intelligent mineral identification method and system, which are characterized in that the classification and identification of a mineral to be detected are realized by acquiring principal component information, trace component information, optical characteristics, crystal characteristics, color characteristics and scratch characteristics of the mineral to be detected, calculating the weight values corresponding to the principal component information, the trace component information, the optical characteristics, the crystal characteristics, the color characteristics and the scratch characteristics respectively to obtain a characteristic vector, training a mineral identification model according to the characteristic vector, and inputting the mineral to be detected into the trained mineral identification model. The invention trains the mineral identification model through multiple identification characteristics, thereby improving the accuracy of mineral identification.
Description
Technical Field
The invention relates to the technical field of intelligent mineral identification, in particular to an intelligent mineral identification method and system.
Background
With the development of detection technology, mineralogy research increasingly depends on large-scale instrument analysis and test technology, and the results of mineral identification and rock component identification are more accurate and reliable. However, for geological researchers who do not analyze and test, the detection report is only of various chemical compositions, and the mineral names wanted by the researchers can be obtained only by referring to relevant books of mineral chemical analysis. The process of converting the chemical composition of the mineral into the name of the mineral only depends on manual checking, and the process is not only extremely efficient but also has inaccurate results. The major element components alone cannot be distinguished accurately from the major element components of a large variety of identical minerals or minerals with identical element types, and accurate mineral names can be obtained only by combining trace elements, optical characteristics under a rock slice microscope and identification characteristics (such as crystal forms, colors, scratches and the like) of hand specimens.
In the prior art, a digital intelligent mineral identification and classification technology of an image identification method is generally adopted. The technology mainly uses a deep learning algorithm to identify and classify hyperspectral pictures or object pictures of minerals. Usually, hyperspectral pictures or object pictures of a large number of minerals are collected firstly and used for training convolutional neural networks in a deep learning algorithm, and after training is completed, the task of mineral identification can be completed.
Disclosure of Invention
The invention aims to provide an intelligent mineral identification method and system, which train a mineral identification model through multiple identification characteristics and intelligently identify minerals through the trained mineral identification model, so that the accuracy of mineral identification is improved.
In order to achieve the purpose, the invention provides the following scheme:
an intelligent mineral identification method comprises the following steps:
acquiring characteristic information data of a mineral to be detected; the characteristic information data comprises principal component information, micro component information, optical characteristics, crystal characteristics, color characteristics and scratch characteristics;
calculating the weight values corresponding to the major component information, the minor component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic;
multiplying the principal component information, the minor component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic by the respective corresponding weight values to obtain a characteristic vector;
inputting the characteristic vector into an intelligent mineral identification model to obtain a predicted mineral name and a predicted confidence coefficient;
and determining the classification result of the mineral to be detected according to the predicted mineral name and the prediction confidence coefficient.
Preferably, the determination method of the mineral type identification model comprises the following steps:
acquiring characteristic information data of a mineral to be trained;
carrying out category marking on the characteristic information data of the mineral to be trained to obtain marking information;
determining a first training set and a second training set; the first training set comprises optical characteristics and corresponding labeling information in the characteristic information data of the mineral to be trained; the second training set comprises principal component information, trace component information, crystal form characteristics, color characteristics, scratch characteristics and corresponding labeling information in the characteristic information data of the mineral to be trained;
calculating a weight value of the first training concentration optical characteristic, and multiplying the first training concentration optical characteristic by the weight value of the optical characteristic to obtain a first characteristic vector;
calculating a weight value of the principal component information, a weight value of the minor component information, a weight value of the crystal form characteristic, a weight value of the color characteristic and a weight value of the scratch characteristic in the second training set, and multiplying the principal component information, the minor component information, the crystal form characteristic, the color characteristic and the scratch characteristic by the corresponding weight values to obtain a second feature vector;
constructing a convolutional neural network module and a full-connection network module;
inputting the first feature vector and the labeling information in the first training set into the convolutional neural network module for training;
and inputting the second feature vector and the label information in the second training set into the fully-connected network module for training, combining the trained convolutional neural network module and the fully-connected network module, and inputting the combined result into a fully-connected neural network, wherein the output of the fully-connected neural network is used as the output of a prediction result.
Preferably, the inputting the first feature vector and the label information in the first training set into the convolutional neural network module for training specifically includes:
determining a loss function according to the first feature vector and the labeling information in the first training set;
and training the convolutional neural network module by adopting a gradient descent optimization algorithm with the minimum loss function as a target to obtain the trained convolutional neural network.
Preferably, the acquiring the feature information data of the mineral to be trained specifically includes:
and extracting characteristic information data from the mineral data characteristic library to serve as the characteristic information data of the mineral to be trained.
Preferably, the principal component information and the minor component information are numerical data; the optical characteristic is picture type data; the crystal form characteristics, the color characteristics and the scratch characteristics are text type data.
An intelligent mineral identification system comprising:
the acquisition unit is used for acquiring characteristic information data of the mineral to be detected; the characteristic information data comprises principal component information, micro component information, optical characteristics, crystal characteristics, color characteristics and scratch characteristics;
the calculating unit is used for calculating weight values corresponding to the principal component information, the minor component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic;
the vector acquisition unit is used for multiplying the main component information, the micro component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic by the corresponding weight values to obtain a characteristic vector;
the prediction unit is used for inputting the feature vector into an intelligent mineral identification model to obtain a predicted mineral name and a prediction confidence coefficient;
and the determining unit is used for determining the classification result of the mineral to be detected according to the predicted mineral name and the prediction confidence coefficient.
Preferably, a model determination unit is further included; the model determining unit specifically includes:
the acquisition module is used for acquiring characteristic information data of the mineral to be trained;
the marking module is used for carrying out category marking on the characteristic information data of the mineral to be trained to obtain marking information;
a training set determining module for determining a first training set and a second training set; the first training set comprises optical characteristics and corresponding labeling information in the characteristic information data of the mineral to be trained; the second training set comprises principal component information, trace component information, crystal form characteristics, color characteristics, scratch characteristics and corresponding labeling information in the characteristic information data of the mineral to be trained;
the first vector determination module is used for calculating a weight value of the first training concentration optical characteristic and multiplying the first training concentration optical characteristic by the weight value of the optical characteristic to obtain a first feature vector;
the second vector determining module is used for calculating a weight value of the principal component information, a weight value of the minor component information, a weight value of the crystal form characteristic, a weight value of the color characteristic and a weight value of the scratch characteristic in the second training set, and multiplying the principal component information, the minor component information, the crystal form characteristic, the color characteristic and the scratch characteristic by the corresponding weight values to obtain a second feature vector;
the building module is used for building the convolutional neural network module and the full-connection network module;
the first model determining module is used for inputting the first feature vector and the labeling information in the first training set into the convolutional neural network module for training;
and the second model determining module is used for inputting the second feature vector and the labeling information in the second training set into the fully-connected network module for training, combining the trained convolutional neural network and the fully-connected network module and inputting the combined result into a fully-connected neural network, and outputting the fully-connected neural network as the output of the prediction result.
Preferably, the first model determining module specifically includes:
a loss function determining submodule, configured to determine a loss function according to the first feature vector and the label information in the first training set;
and the training submodule is used for training the convolutional neural network module by adopting a gradient descent optimization algorithm with the minimum loss function as a target to obtain a trained convolutional neural network.
Preferably, the obtaining module includes:
and the acquisition submodule is used for extracting characteristic information data from the mineral data characteristic library to serve as the characteristic information data of the mineral to be trained.
Preferably, the principal component information and the minor component information are numerical data; the optical characteristic is picture type data; the crystal form characteristics, the color characteristics and the scratch characteristics are text type data.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention provides an intelligent mineral identification method and system, which are characterized in that the classification and identification of a mineral to be detected are realized by acquiring the principal component information, the minor component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic of the mineral to be detected, calculating the weight values corresponding to the principal component information, the minor component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic respectively to obtain a characteristic vector, training a mineral identification model according to the characteristic vector, and inputting the mineral to be detected into the trained mineral identification model. The invention trains the mineral identification model through multiple identification characteristics, thereby improving the accuracy of mineral identification.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
FIG. 1 is a flow chart of a method for intelligent mineral identification according to the present invention;
FIG. 2 is a schematic diagram of computing a feature vector according to an embodiment of the present invention;
fig. 3 is a module connection diagram of an intelligent mineral identification system according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide a method for training a mineral recognition model through multiple recognition characteristics and intelligently recognizing minerals through the trained mineral recognition model, so that the accuracy of mineral recognition is improved.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Fig. 1 is a flowchart of an intelligent mineral identification method according to the present invention, and as shown in fig. 1, the intelligent mineral identification method according to the present invention includes:
step 100: acquiring characteristic information data of a mineral to be detected; the characteristic information data comprises principal component information, micro-component information, optical characteristics, crystal characteristics, color characteristics and scratch characteristics.
Step 200: and calculating the weight values corresponding to the principal component information, the minor component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic respectively.
Step 300: and multiplying the principal component information, the minor component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic by the respective corresponding weight values to obtain a characteristic vector.
Step 400: and inputting the characteristic vector into an intelligent mineral identification model to obtain a predicted mineral name and a predicted confidence coefficient.
Optionally, the intelligent mineral identification model comprises a mineral type identification model and a confidence prediction model.
Step 500: and determining the classification result of the mineral to be detected according to the predicted mineral name and the prediction confidence coefficient.
Fig. 2 is a schematic diagram of calculating a feature vector in the embodiment of the present invention, and as shown in fig. 2, the feature vector is obtained by multiplying the identification features by the respective weights. The principal component information and the micro component information in the feature vector are numerical data; the optical characteristic is picture type data; the crystal form characteristic, the color characteristic and the scratch characteristic are text type data.
Optionally, the main characteristics of the invention mainly comprise main components and main components of the mineral, trace components, optical characteristics and text information such as crystal form, color, scratch and the like observed by naked eyes. The principal component and minor component characteristics are organized primarily numerically, the phototropic characteristics are presented graphically, and the information observed by the naked eye is presented in natural sentences. For picture data, image features and natural text features are extracted in a CNN (common noise network) mode and the like, and are represented and extracted in a word vector mode.
Specifically, the expert mineral data analysis system is adopted to analyze each characteristic to obtain the corresponding weight.
Optionally, the feature vector may include one feature vector, or may include a plurality of feature vectors.
Preferably, the determination method of the mineral type identification model comprises the following steps:
and acquiring characteristic information data of the mineral to be trained.
And carrying out category marking on the characteristic information data of the mineral to be trained to obtain marking information.
Determining a first training set and a second training set; the first training set comprises optical characteristics and corresponding labeling information in the characteristic information data of the mineral to be trained; the second training set comprises principal component information, trace component information, crystal form characteristics, color characteristics, scratch characteristics and corresponding labeling information in the characteristic information data of the mineral to be trained.
And calculating a weight value of the first training concentration optical characteristic, and multiplying the first training concentration optical characteristic by the weight value of the optical characteristic to obtain a first characteristic vector.
And calculating the weight value of the principal component information, the weight value of the minor component information, the crystal form characteristic, the color characteristic and the scratch characteristic in the second training set, and multiplying the principal component information, the minor component information, the crystal form characteristic, the color characteristic and the scratch characteristic by the respective corresponding weight values to obtain a second feature vector.
And constructing a convolutional neural network module and a fully-connected network module.
And inputting the first feature vector and the labeling information in the first training set into the convolutional neural network module for training.
Specifically, the trained convolutional neural network module is determined as the mineral type identification model.
And inputting the second feature vector and the labeling information in the second training set into the full-connection network module for training.
As an alternative embodiment, the trained fully-connected network module is determined as the confidence prediction model.
And combining the trained convolutional neural network and the fully-connected network module and inputting the combined signals into a fully-connected neural network, wherein the output of the fully-connected neural network is used as the output of the prediction result.
As an optional implementation manner, the classifier pre-training unit of the classifier pre-training unit before the convolutional neural network module includes a data set generation module for neural network training, the data set generation module for neural network training pre-processes the phototropic characteristics in the characteristic information data of the mineral to be trained for training the neural network and creates a first training set, and the convolutional neural network training module performs calculation of the neural network by taking the first training set and the labeling information output by the data set generation module for neural network training as inputs.
Optionally, after the training of the convolutional neural network module and the fully-connected network module is completed, a picture of a mineral to be detected is input into the convolutional neural network module to obtain a one-dimensional eigenvalue vector or a non-normalized one-dimensional probability value vector, the fully-connected network module is provided with a plurality of layers, a second eigenvector is input into two or more layers of fully-connected networks to output the non-normalized one-dimensional probability value vector, then the one-dimensional eigenvalue vector or the non-normalized one-dimensional probability value vector obtained from the convolutional neural network and the non-normalized one-dimensional probability value vector obtained from the fully-connected networks are spliced into a new one-dimensional vector, the new one-dimensional vector is input into one or more layers of fully-connected networks, and a final result, namely, the final result belongs to different types of probability sizes and mineral names, is output.
Preferably, the inputting the first feature vector and the label information in the first training set into the convolutional neural network module for training specifically includes:
and determining a loss function according to the first feature vector and the labeling information in the first training set.
And training the convolutional neural network module by adopting a gradient descent optimization algorithm with the minimum loss function as a target to obtain the mineral type identification model.
Preferably, the acquiring the feature information data of the mineral to be trained specifically includes:
and extracting characteristic information data from the mineral data characteristic library to serve as the characteristic information data of the mineral to be trained.
Fig. 3 is a module connection diagram of an intelligent mineral identification system according to the present invention, and as shown in fig. 3, the intelligent mineral identification system according to the present invention includes:
the acquisition unit is used for acquiring characteristic information data of the mineral to be detected; the characteristic information data comprises principal component information, micro-component information, optical characteristics, crystal characteristics, color characteristics and scratch characteristics.
And the calculating unit is used for calculating the weight values corresponding to the principal component information, the minor component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic.
And the vector acquisition unit is used for multiplying the main component information, the micro component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic by the corresponding weight values to obtain a characteristic vector.
And the prediction unit is used for inputting the feature vector into an intelligent mineral identification model to obtain a predicted mineral name and a prediction confidence coefficient.
And the determining unit is used for determining the classification result of the mineral to be detected according to the predicted mineral name and the prediction confidence coefficient.
Preferably, a model determination unit is further included; the model determining unit specifically includes:
and the acquisition module is used for acquiring the characteristic information data of the mineral to be trained.
And the marking module is used for carrying out category marking on the characteristic information data of the mineral to be trained to obtain marking information.
A training set determining module for determining a first training set and a second training set; the first training set comprises optical characteristics and corresponding labeling information in the characteristic information data of the mineral to be trained; the second training set comprises principal component information, trace component information, crystal form characteristics, color characteristics, scratch characteristics and corresponding labeling information in the characteristic information data of the mineral to be trained.
And the first vector determination module is used for calculating the weight value of the first training concentration optical characteristic and multiplying the first training concentration optical characteristic by the weight value of the optical characteristic to obtain a first feature vector.
And the second vector determining module is used for calculating a weight value of the principal component information, a weight value of the minor component information, a crystal form characteristic, a color characteristic and a scratch characteristic in the second training set, and multiplying the principal component information, the minor component information, the color characteristic and the scratch characteristic by the corresponding weight values to obtain a second feature vector.
And the building module is used for building the convolutional neural network module and the full-connection network module.
And the first model determining module is used for inputting the first feature vector and the labeling information in the first training set into the convolutional neural network module for training.
And the second model determining module is used for inputting the second feature vector and the labeling information in the second training set into the fully-connected network module for training, combining the trained convolutional neural network module and the fully-connected network module and inputting the combined result into a fully-connected neural network, and outputting the fully-connected neural network as the output of a prediction result.
Preferably, the first model determining module specifically includes:
and the loss function determining submodule is used for determining a loss function according to the first feature vector and the labeling information in the first training set.
And the training submodule is used for training the convolutional neural network module by adopting a gradient descent optimization algorithm with the minimum loss function as a target to obtain the trained convolutional neural network module.
Preferably, the obtaining module includes:
and the acquisition submodule is used for extracting characteristic information data from the mineral data characteristic library to serve as the characteristic information data of the mineral to be trained.
Preferably, the principal component information and the minor component information are numerical data; the optical characteristic is picture type data; the crystal form characteristics, the color characteristics and the scratch characteristics are text type data.
The invention has the following beneficial effects:
(1) the invention trains the mineral recognition model through a plurality of recognition characteristics and intelligently recognizes minerals by the trained mineral recognition model, thereby improving the accuracy of mineral recognition.
(2) According to the invention, by utilizing the mineral data feature library (mineral chemical composition, crystal growth characteristics, optical characteristics and hand specimen visual identification characteristics) accumulated by an applicant for many years, the unknown mineral is intelligently identified by utilizing a machine big data learning method, the purposes of automatically detecting and identifying the mineral name are realized, and the automation degree of the identification process is improved.
The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (10)
1. An intelligent mineral identification method is characterized by comprising the following steps:
acquiring characteristic information data of a mineral to be detected; the characteristic information data comprises principal component information, micro component information, optical characteristics, crystal characteristics, color characteristics and scratch characteristics;
calculating the weight values corresponding to the major component information, the minor component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic;
multiplying the principal component information, the minor component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic by the respective corresponding weight values to obtain a characteristic vector;
inputting the characteristic vector into an intelligent mineral identification model to obtain a predicted mineral name and a predicted confidence coefficient;
and determining the classification result of the mineral to be detected according to the predicted mineral name and the prediction confidence coefficient.
2. The intelligent mineral identification method according to claim 1, wherein the determination method of the mineral type identification model is as follows:
acquiring characteristic information data of a mineral to be trained;
carrying out category marking on the characteristic information data of the mineral to be trained to obtain marking information;
determining a first training set and a second training set; the first training set comprises optical characteristics and corresponding labeling information in the characteristic information data of the mineral to be trained; the second training set comprises principal component information, trace component information, color characteristics, scratch characteristics and corresponding labeling information in the characteristic information data of the mineral to be trained;
calculating a weight value of the first training concentration optical characteristic, and multiplying the first training concentration optical characteristic by the weight value of the optical characteristic to obtain a first characteristic vector;
calculating a weight value of the principal component information, a weight value of the minor component information, a weight value of the crystal form characteristic, a weight value of the color characteristic and a weight value of the scratch characteristic in the second training set, and multiplying the principal component information, the minor component information, the crystal form characteristic, the color characteristic and the scratch characteristic by the corresponding weight values to obtain a second feature vector;
constructing a convolutional neural network module and a full-connection network module;
inputting the first feature vector and the labeling information in the first training set into the convolutional neural network module for training;
and inputting the second feature vector and the label information in the second training set into the fully-connected network module for training, combining the trained convolutional neural network module and the fully-connected network module, and inputting the combined result into a fully-connected neural network, wherein the output of the fully-connected neural network is used as the output of a prediction result.
3. The intelligent mineral identification method according to claim 1, wherein the inputting the first feature vector and the labeling information in the first training set into the convolutional neural network module for training specifically comprises:
determining a loss function according to the first feature vector and the labeling information in the first training set;
and training the convolutional neural network module by adopting a gradient descent optimization algorithm with the minimum loss function as a target to obtain the trained convolutional neural network.
4. The intelligent mineral identification method according to claim 2, wherein the acquiring of the feature information data of the mineral to be trained specifically comprises:
and extracting characteristic information data from the mineral data characteristic library to serve as the characteristic information data of the mineral to be trained.
5. The intelligent mineral identification method according to claim 1, wherein the principal component information and the minor component information are numerical data; the optical characteristic is picture type data; the crystal form characteristics, the color characteristics and the scratch characteristics are text type data.
6. An intelligent mineral identification system, comprising:
the acquisition unit is used for acquiring characteristic information data of the mineral to be detected; the characteristic information data comprises principal component information, micro component information, optical characteristics, crystal characteristics, color characteristics and scratch characteristics;
the calculating unit is used for calculating weight values corresponding to the main component information, the micro component information, the optical characteristic, the crystal form characteristic, the color characteristic and the scratch characteristic;
a vector obtaining unit, configured to multiply the principal component information, the minor component information, the optical characteristic, the color characteristic, and the scratch characteristic by their respective corresponding weight values to obtain a feature vector;
the prediction unit is used for inputting the feature vector into an intelligent mineral identification model to obtain a predicted mineral name and a prediction confidence coefficient;
and the determining unit is used for determining the classification result of the mineral to be detected according to the predicted mineral name and the prediction confidence coefficient.
7. The intelligent mineral identification system of claim 6, further comprising a model determination unit; the model determining unit specifically includes:
the acquisition module is used for acquiring characteristic information data of the mineral to be trained;
the marking module is used for carrying out category marking on the characteristic information data of the mineral to be trained to obtain marking information;
a training set determining module for determining a first training set and a second training set; the first training set comprises optical characteristics and corresponding labeling information in the characteristic information data of the mineral to be trained; the second training set comprises principal component information, trace component information, crystal form characteristics, color characteristics, scratch characteristics and corresponding labeling information in the characteristic information data of the mineral to be trained;
the first vector determination module is used for calculating a weight value of the first training concentration optical characteristic and multiplying the first training concentration optical characteristic by the weight value of the optical characteristic to obtain a first feature vector;
the second vector determining module is used for calculating a weight value of the principal component information, a weight value of the minor component information, a weight value of the color feature and a weight value of the scratch feature in the second training set, and multiplying the principal component information, the minor component information, the crystal form feature, the color feature and the scratch feature by the corresponding weight values to obtain a second feature vector;
the building module is used for building the convolutional neural network module and the full-connection network module;
the first model determining module is used for inputting the first feature vector and the labeling information in the first training set into the convolutional neural network module for training;
and the second model determining module is used for inputting the second feature vector and the labeling information in the second training set into the fully-connected network module for training, combining the trained convolutional neural network module and the fully-connected network module and inputting the combined result into a fully-connected neural network, and outputting the fully-connected neural network as the output of a prediction result.
8. The intelligent mineral identification system according to claim 6, wherein the first model determination module specifically comprises:
a loss function determining submodule, configured to determine a loss function according to the first feature vector and the label information in the first training set;
and the training submodule is used for training the convolutional neural network module by adopting a gradient descent optimization algorithm with the minimum loss function as a target to obtain the trained convolutional neural network module.
9. The intelligent mineral identification system of claim 6, wherein the acquisition module comprises:
and the acquisition submodule is used for extracting characteristic information data from the mineral data characteristic library to serve as the characteristic information data of the mineral to be trained.
10. The intelligent mineral identification system according to claim 6, wherein the principal component information and the minor component information are numerical data; the optical characteristic is picture type data; the crystal form characteristics, the color characteristics and the scratch characteristics are text type data.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110423576.9A CN113128404A (en) | 2021-04-20 | 2021-04-20 | Intelligent mineral identification method and system |
| AU2021102380A AU2021102380A4 (en) | 2021-04-20 | 2021-05-06 | Intelligent Mineral Identification Method and System |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110423576.9A CN113128404A (en) | 2021-04-20 | 2021-04-20 | Intelligent mineral identification method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113128404A true CN113128404A (en) | 2021-07-16 |
Family
ID=76442118
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110423576.9A Pending CN113128404A (en) | 2021-04-20 | 2021-04-20 | Intelligent mineral identification method and system |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN113128404A (en) |
| AU (1) | AU2021102380A4 (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115656053A (en) * | 2022-10-19 | 2023-01-31 | 山东大学 | Rock mineral content testing method and system |
| CN115656053B (en) * | 2022-10-19 | 2024-05-31 | 山东大学 | Rock mineral content testing method and system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102053268A (en) * | 2010-11-18 | 2011-05-11 | 中国石油天然气股份有限公司 | Method and device for identifying lithologic character of altered volcanic rock |
| CN110059765A (en) * | 2019-04-26 | 2019-07-26 | 中国地质大学（北京） | A kind of mineral intelligent recognition categorizing system and method |
| CN110119753A (en) * | 2019-01-08 | 2019-08-13 | 长江岩土工程总公司（武汉） | A kind of method of reconstituted texture identification lithology |
| CN111784642A (en) * | 2020-06-10 | 2020-10-16 | 中铁四局集团有限公司 | Image processing method, target recognition model training method and target recognition method |
| AU2020103251A4 (en) * | 2020-03-15 | 2021-01-14 | Jilin University | Method and system for identifying metallic minerals under microscope based on bp nueral network |
-
2021
- 2021-04-20 CN CN202110423576.9A patent/CN113128404A/en active Pending
- 2021-05-06 AU AU2021102380A patent/AU2021102380A4/en not_active Ceased
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102053268A (en) * | 2010-11-18 | 2011-05-11 | 中国石油天然气股份有限公司 | Method and device for identifying lithologic character of altered volcanic rock |
| CN110119753A (en) * | 2019-01-08 | 2019-08-13 | 长江岩土工程总公司（武汉） | A kind of method of reconstituted texture identification lithology |
| CN110059765A (en) * | 2019-04-26 | 2019-07-26 | 中国地质大学（北京） | A kind of mineral intelligent recognition categorizing system and method |
| AU2020103251A4 (en) * | 2020-03-15 | 2021-01-14 | Jilin University | Method and system for identifying metallic minerals under microscope based on bp nueral network |
| CN111784642A (en) * | 2020-06-10 | 2020-10-16 | 中铁四局集团有限公司 | Image processing method, target recognition model training method and target recognition method |
Non-Patent Citations (2)
| Title |
|---|
| 许志华，吴立新: "《地灾与建筑损毁的无人机与地面LiDAR协同观测及评估》", 31 March 2019 * |
| 陈希廉等: "《地质学 修订版》", 31 July 1979 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115656053A (en) * | 2022-10-19 | 2023-01-31 | 山东大学 | Rock mineral content testing method and system |
| CN115656053B (en) * | 2022-10-19 | 2024-05-31 | 山东大学 | Rock mineral content testing method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| AU2021102380A4 (en) | 2021-06-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110287334A (en) | A kind of school's domain knowledge map construction method based on Entity recognition and attribute extraction model | |
| CN109241383B (en) | A kind of type of webpage intelligent identification Method and system based on deep learning | |
| CN114168906B (en) | Mapping geographic information data acquisition system based on cloud computing | |
| CN109978872B (en) | White matter microstructure characteristic screening system and method based on white matter fiber tracts | |
| CN110161480B (en) | Radar target identification method based on semi-supervised depth probability model | |
| CN109858428B (en) | Automatic ANA fluorescent film identification method based on machine learning and deep learning | |
| CN112836067B (en) | Intelligent searching method based on knowledge graph | |
| CN106934055B (en) | Semi-supervised webpage automatic classification method based on insufficient modal information | |
| CN113779260B (en) | Pre-training model-based domain map entity and relationship joint extraction method and system | |
| CN108959474A (en) | Entity relationship extracting method | |
| CN109460471B (en) | Method for establishing fiber category map library based on self-learning mode | |
| CN110377802B (en) | Knowledge point labeling method and system for newly added test questions | |
| CN111400478A (en) | Knowledge graph question-answering method and device based on word embedding matching technology | |
| CN115312183A (en) | Intelligent interpretation method and system for medical inspection report | |
| CN113065356A (en) | IT equipment operation and maintenance fault suggestion processing method based on semantic analysis algorithm | |
| CN109753649A (en) | Text relevant measure and system based on fine granularity matched signal | |
| CN117252842A (en) | Aircraft skin defect detection and network model training method | |
| CN114722204A (en) | Multi-label text classification method and device | |
| CN114153942B (en) | Event time sequence relation extraction method based on dynamic attention mechanism | |
| CN111984790A (en) | Entity relation extraction method | |
| CN107578785A (en) | The continuous emotional feature analysis evaluation method of music based on Gamma distributional analysis | |
| CN113128404A (en) | Intelligent mineral identification method and system | |
| CN104778479A (en) | Image classification method and system based on sparse coding extractor | |
| CN117473096B (en) | Knowledge point labeling method fusing LATEX labels and model thereof | |
| CN116978408B (en) | Depression detection method and system based on voice pre-training model |
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 |