High-resolution remote sensing-based non-administrative pressure and covering evaluation method for mineral resources and related device

Abstract

The application discloses a non-examination and approval overlying assessment method for mineral resources based on high-resolution remote sensing and related devices, relates to the technical field of mineral resource assessment, and comprises the following steps: acquiring a remote sensing image; performing coordinate conversion on the remote sensing image to obtain a converted image; each point on the converted image is consistent with a field coordinate; automatically extracting construction project information from the converted image to obtain a construction project vector; calculating a deep-thick ratio and a moving angle based on the construction project vector; determining a safety protection range and a safe operation range according to the deep-thick ratio and the moving angle; determining a non-examination and approval maximum overlying range based on the safety protection range and the safe operation range; determining a non-examination and approval overlying resource quantity according to the non-examination and approval maximum overlying range; and performing mapping based on the non-examination and approval overlying resource quantity to obtain final mapping. The application can realize non-examination and approval overlying assessment of mineral resources.

Description

Technical Field

[0001] This application relates to the field of mineral resource assessment technology, and in particular to a non-approval assessment method and related apparatus for mineral resource overlay assessment based on high-resolution remote sensing. Background Technology

[0002] Mineral resources covered by construction projects can be categorized into approved and unapproved coverage. Approved covered mineral resources have clearly defined coverage areas and resource quantities, while unapproved covered mineral resources are those for which assessments have not been conducted for various reasons and are currently unusable. These unapproved covered mineral resources lack clearly defined coverage areas and resource quantities and require assessment. Assessing unapproved covered mineral resources requires determining two aspects: the spatial scope affected by the construction project and the resource quantity within that spatial scope. Assessing important mineral types in large and medium-sized mining areas (mineral deposits) and identifying unapproved covered mineral resources by major construction projects is crucial for understanding the amount of usable mineral resources in my country.

[0003] Therefore, how to achieve non-approval assessment of mineral resource overlay has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0004] The purpose of this application is to provide a method and related apparatus for assessing non-approval-based mineral resource overlay based on high-resolution remote sensing, which can realize the assessment of non-approval-based mineral resource overlay.

[0005] To achieve the above objectives, this application provides the following solution.

[0006] In a first aspect, this application provides a method for assessing non-approved mineral resource overlay based on high-resolution remote sensing, which includes the following steps.

[0007] Acquire remote sensing images.

[0008] The remote sensing image is subjected to coordinate transformation to obtain the transformed image; each point on the transformed image has the same coordinates as the actual ground coordinates.

[0009] The project information is automatically extracted from the converted image to obtain the project vector.

[0010] Based on the project vector, calculate the depth ratio and the angle of movement.

[0011] The safety protection range and the safe operating range are determined based on the depth ratio and the movement angle.

[0012] Based on the aforementioned safety protection range and the aforementioned safe operation range, the maximum non-approved coverage area is determined.

[0013] The amount of resources to be covered by non-approved materials is determined based on the maximum non-approved coverage area.

[0014] Based on the amount of non-approved overlaid resources, a map was created to obtain the final map.

[0015] In a second aspect, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to implement the non-approval overlay assessment method for mineral resources based on high-resolution remote sensing as described in the first aspect.

[0016] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the non-approval overlay assessment method for mineral resources based on high-resolution remote sensing as described in the first aspect.

[0017] Fourthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the non-approval overlay assessment method for mineral resources based on high-resolution remote sensing as described in the first aspect.

[0018] Based on the specific embodiments provided in this application, the following technical effects are disclosed.

[0019] This application provides a method and related apparatus for non-approval-based assessment of mineral resource overlay based on high-resolution remote sensing. The method includes: acquiring remote sensing images; performing coordinate transformation on the remote sensing images to obtain transformed images; ensuring that each point on the transformed image matches the actual coordinates; achieving a precise one-to-one correspondence between image points and actual coordinates, eliminating the problem of original image coordinate deviation, and establishing a unified and accurate geographic coordinate benchmark for all subsequent spatial analysis operations. The transformed image is then automatically used to extract project information, resulting in project vectors; this can replace traditional manual extraction methods, efficiently and accurately extracting spatial information of construction projects and completing standardized vectorization processing, improving the efficiency and accuracy of spatial information extraction. Based on the project vectors, the depth ratio and movement angle are calculated; the core key parameters required for mineral resource overlay assessment can be accurately calculated based on the spatial vector characteristics of the construction project, providing a quantitative basis for the scientific delineation of subsequent safety zones. The scientifically quantified core parameters can be used to accurately delineate two types of safety zones, clarifying the safety boundary between mineral resource development and construction project operation, and defining the safety threshold for undisturbed mineral resources. Based on the aforementioned safety protection range and safe operation range, the maximum non-approved coverage area is determined. This allows for precise delineation of the mineral resource coverage area in non-approved scenarios by combining dual safety boundary conditions, effectively avoiding overestimation or underestimation of the coverage area and clarifying the actual maximum coverage boundary under the non-approved model. Based on the maximum non-approved coverage area, the amount of resources covered by non-approved resources is determined. The scientific calculation and quantitative statistics of the corresponding mineral resources can be completed based on the precisely delineated coverage area, forming specific data on the covered resources, providing core data support for the assessment of the resource value of non-approved coverage. Mapping is performed based on the non-approved covered resources to obtain the final map. This transforms the quantified covered resource data and spatial range information into visual thematic maps, intuitively and clearly presenting the spatial distribution and resource quantity of mineral resources covered by non-approved resources, facilitating the review, application, and implementation of assessment results and related decisions. This application, through the orderly connection and progressive scientific processing of each step, leverages the advantages of high-resolution remote sensing technology to achieve standardized, precise, and automated assessment of the entire process from basic image acquisition to final visualization. It effectively solves the problems of low efficiency, large coordinate deviations, inaccurate scope delineation, and large errors in resource quantity calculation in traditional assessment methods, ensuring the scientific, objective, and accurate delineation of non-approved mineral resource coverage areas and resource quantity calculation. It also significantly improves the overall efficiency of non-approved mineral resource coverage assessment. Furthermore, through clear safety scope definition and precise resource quantity quantification, it provides comprehensive, reliable, intuitive, and effective technical support and data basis for mineral resource development control, compliant planning and layout of construction projects, and decision-making related to non-approved mineral resource coverage. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart illustrating a non-approval overlay assessment method for mineral resources based on high-resolution remote sensing, provided as an embodiment of this application.

[0022] Figure 2 This is a flowchart illustrating the overall technical process of a non-approval overlay assessment method for mineral resources based on high-resolution remote sensing, as provided in one embodiment of this application.

[0023] Figure 3 This is a schematic diagram of projection transformation provided in an embodiment of this application.

[0024] Figure 4 This is a schematic diagram illustrating the use of band information provided in an embodiment of this application.

[0025] Figure 5 A typical structure diagram of a convolutional neural network provided in an embodiment of this application.

[0026] Figure 6 This is a schematic diagram of a neural network interpretation method provided in an embodiment of this application.

[0027] Figure 7 This is a schematic diagram illustrating the determination of the slope angle in an open-pit mining operation, as provided in one embodiment of this application.

[0028] Figure 8 This is a schematic diagram illustrating the determination of the safe operating range for open-pit mining, provided in one embodiment of this application.

[0029] Figure 9 This is a schematic diagram illustrating the determination of the underground mining movement angle pair according to an embodiment of this application.

[0030] Figure 10 This is a schematic diagram illustrating the determination of the safe operating range for underground mining, provided in one embodiment of this application.

[0031] Figure 11 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0032] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0033] The purpose of this application is to determine a reasonable coverage area by spatially overlaying high-resolution imagery and mineral resource reserve estimation range, based on the safety protection and safe operation range of the construction project, and then estimate the amount of mineral resources covered by the overlying area. This provides a new method for non-approved overlying, mainly solving the problems of estimating the coverage area and the amount of resources covered by non-approved overlying, and providing an estimate for temporarily unusable mineral resources in the evaluation of mineral resource availability. This is of great significance for the evaluation of the amount of usable mineral resources in my country.

[0034] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0035] In one exemplary embodiment, such as Figure 1 As shown, a method for non-approval assessment of mineral resource overlay based on high-resolution remote sensing is provided, which includes the following steps.

[0036] S1: Acquire remote sensing images.

[0037] S2: Perform coordinate transformation on the remote sensing image to obtain the transformed image; each point on the transformed image has the same coordinates as the actual ground coordinates.

[0038] S3: Automatically extract construction project information from the converted image to obtain a construction project vector.

[0039] S4: Calculate the depth ratio and movement angle based on the project vector.

[0040] S5: Determine the safety protection range and the safe operating range based on the depth ratio and the movement angle.

[0041] S6: Based on the safety protection range and the safety operation range, determine the maximum non-approved overlay range.

[0042] S7: Determine the amount of non-approved overlay resources based on the maximum non-approved overlay range.

[0043] S8: Based on the amount of non-approved overlaid resources, a map is created to obtain the final map.

[0044] Implementing steps S1 to S8 above can provide a new method for non-approved overburden coverage, mainly solving the problems of estimating the scope of non-approved overburden coverage and the amount of overburdened resources. It provides an estimate for mineral resources that cannot be used temporarily in the evaluation of mineral resource availability, which is of great significance for the evaluation of the amount of available mineral resources in my country.

[0045] As an optional implementation, in step S2, the remote sensing image is subjected to coordinate transformation to obtain the transformed image, which specifically includes the following steps.

[0046] S21: Perform geometric correction on the remote sensing image to obtain the corrected image; the geometric correction includes: image-to-image geometric correction, image control point geometric correction, and image-to-image automatic image registration.

[0047] S22: Perform coordinate transformation and orthorectification on the corrected image to obtain the transformed image.

[0048] As an optional implementation, in step S3, the project information is automatically extracted from the converted image to obtain the project vector, which specifically includes the following steps.

[0049] S31: Based on the converted image, determine the high-resolution image; the high-resolution image is visible light-near infrared 4-band or 8-band data.

[0050] S32: Based on the high-resolution image, determine the type of construction project.

[0051] S33: Perform band information transformation on the original bands of the high-resolution image to obtain a band combination; the band information transformation includes: band ratio, principal component analysis, independent component analysis and minimum noise transformation.

[0052] S34: Combine the bands with the original bands to form a band encapsulation to obtain a virtual spectrum.

[0053] S35: Perform constraint processing on the virtual spectrum to obtain the final signal; the constraint processing includes: angle constraint and intensity constraint.

[0054] The expression for the angle constraint is shown below.

[0055] .

[0056] The expression for the strength constraint is shown below.

[0057] .

[0058] The expression for the final signal is shown below.

[0059] .

[0060] in, It is the spectral angle; For high-intensity pipe flow; For the final signal; As a reference virtual spectral vector; For image virtual spectral vectors, For vectors sum vector The inner product; For vectors Length; For vectors Length; It is a limit value.

[0061] S36: Input the final signal into a convolutional neural network to obtain preliminary project information.

[0062] S37: Based on the project type, perform accuracy verification and validation on the preliminary project information to obtain the final project information.

[0063] S38: The final construction project information is vectorized to obtain the construction project vector.

[0064] As an optional implementation, in step S37, the accuracy of the preliminary construction project information is checked and verified based on the construction project type to obtain the final construction project information, which specifically includes the following steps.

[0065] S371: Determine whether the preliminary construction project information belongs to the construction project type, and obtain the first determination result.

[0066] S372: If the first judgment result is yes, then retain it.

[0067] S373: If the result of the first judgment is negative, then delete.

[0068] S374: If the first judgment result is uncertain, then virtual spectral acquisition is performed on the preliminary construction project information to obtain the image virtual spectrum.

[0069] S375: Compare the spectral angles of the image virtual spectrum with the reference virtual spectrum, and determine whether the spectral angle of the lithium-containing pegmatite is smaller than that of the lithium-free pegmatite to obtain a second judgment result.

[0070] S376: If the second judgment result is yes, then retain it.

[0071] S377: If the result of the second judgment is negative, then delete.

[0072] As an optional implementation, in step S6, the method for determining the maximum non-approved coverage area based on the safety protection range and the safety operation range includes: vertical profile method, plumb line method and digital elevation projection method.

[0073] As an optional implementation, the calculation formula for the amount of non-approved overlay resources is as follows.

[0074] .

[0075] in, These are unapproved resource reserves that have been covered. This represents the total number of overlaid block segments; For the first Volume of each overlay block segment; For the first Apparent density of each overburden block segment.

[0076] This application mainly includes four aspects: remote sensing data processing, delineation of the maximum non-approved overburden area, assessment of non-approved overburden resources, and final mapping. Remote sensing data processing primarily utilizes high-resolution remote sensing data after coordinate correction and machine learning methods to automatically extract the project details. Delineation of the maximum non-approved overburden area mainly determines the maximum overburden area based on the project's safety protection and operational safety zones, ensuring that mineral resource extraction does not affect the project's safe operation. Assessment of non-approved overburden resources mainly estimates the amount of non-approved overburden resources by calculating the depth ratio and the angle of movement, and then maps the mineral resources overburdened by the project. The overall technical flowchart is shown below. Figure 2 .

[0077] (a) Remote sensing data processing.

[0078] 1. Geometric correction and coordinate transformation of high-resolution remote sensing images.

[0079] ① Geometric correction.

[0080] High-resolution remote sensing imagery includes 4-band and 8-band high-resolution satellite imagery. This application uses a general method, performing geometric correction and coordinate transformation on 4-band high-resolution satellite imagery. For the domestically produced GF2 imagery used, correction using the satellite's built-in geolocation files is not recommended; this application offers three main geometric correction transformation methods.

[0081] 1) Image-to-image geometric correction.

[0082] Image correction uses a geometrically corrected remote sensing image as a reference. The same ground features are selected on both the image to be corrected and the reference image, ensuring that these same features appear in the same locations on the corrected image. This is the method used in most geometric corrections.

[0083] 2) Geometric correction of image control points.

[0084] Geometric correction of remote sensing imagery is performed using ground control points. These control points can be obtained from keyboard input, vector files, or raster files, such as for map correction.

[0085] 3) Automatic image registration between images.

[0086] Automatically find corresponding points between two images based on image pixel gray values ​​or ground feature characteristics to complete the geometric correction of the images.

[0087] ② Coordinate transformation.

[0088] The coordinate transformation in this application mainly involves map projection coordinate transformation, such as... Figure 3 As shown, the coordinates of a raster map differ from those of a vector map, necessitating coordinate transformation. First, considering two coordinate systems, V and B, the coordinates of a fixed point P differ between the two systems. In coordinate system B, the coordinates are... P=[ x, y], in the V coordinate system P=[ x, The two can be transformed through matrix rotation.

[0089] .

[0090] .

[0091] The above formula is a rotation formula. Adding translation, that is, rotating coordinate system B to V and then translating it to A, the coordinate transformation formula is as follows.

[0092] .

[0093] .

[0094] in, t This represents the translation amount.

[0095] This application requires processing data from different regions. For data ranging from 1 million to 5 million units nationwide, the data should be converted to the National 2000 coordinate system, Lambert projection, with a central meridian of 105°, first latitude of 25°, second latitude of 47°, and origin latitude of 18°. For data ranging from 500,000 to 50,000 units for a specific region, the data should be converted to the National 2000 coordinate system, Gauss projection, and 6-degree zone. For data ranging from 5,000 to 25,000 units for a specific region, the data should be converted to the National 2000 coordinate system, Gauss projection, and 3-degree zone.

[0096] ③Orthorectification.

[0097] Orthorectification of high-resolution satellite imagery is performed based on RPC (rational polynomial coefficient) information. Conventional remote sensing software can be used to perform orthorectification.

[0098] 2. Automatic extraction of project information from high-resolution remote sensing imagery.

[0099] The automatic extraction of project information from high-resolution remote sensing imagery employs a method combining spectral analysis and a 3D-CNN convolutional neural network. The first step involves selecting high-resolution imagery, typically 4-band or 8-band visible-near-infrared data. Domestic satellites such as Gaofen-2 use 4-band data, and this application utilizes 4-band high-resolution satellite imagery for automatic extraction of project information.

[0100] ① Determining the type of construction project.

[0101] Major construction projects include highways, high-speed railways, railways, oil and gas pipelines, the South-to-North Diversion Project, airports, reservoirs and dams, bridges, tunnels, pipelines, and urban development.

[0102] ②Use of band information.

[0103] For 4 bands ( B 1, B 2, B 3, B 4) High-resolution satellite imagery undergoes four types of transformations: band ratio analysis, principal component analysis, independent component analysis, and minimum noise transformation. Based on the results of these four transformations, a 22-channel "spectral" information (original bands + transformed bands, each band forming one channel) is constructed from the original four bands. A virtual "spectrum" is built for the project using these 22 channels, and an information extraction method based on the mutual constraints of angle and intensity is developed using this virtual "spectrum." The overall technical flowchart for this step is shown below. Figure 4 As shown.

[0104] 1) Band ratio.

[0105] The main task is to perform ratio processing on the four bands.

[0106] ; ; ; ; ; .

[0107] 2) Principal component analysis.

[0108] Principal component analysis (PCA) is used to transform the original data. The principle of conventional PCA is as follows: First, the origin of the coordinate system is moved so that the mean is zero. After this step, the coordinates are rotated so that one axis aligns with the direction of the data's maximum distribution. This new axis is the first principal component, which accounts for the largest share of the total variance. The other axis perpendicular to it represents the direction of the remaining variance; this is the second principal component. In multidimensional spaces (two or more dimensions), this process continues to determine a set of rectangular coordinate axes. These axes gradually distribute (consume) all the variance. It is not entirely contained within a single secondary principal component; rather, the number of principal components corresponds to the number of original parameters. The sum of the variance values ​​of all principal components is equal to the sum of the variance values ​​before the transformation; this is the conservation of information.

[0109] The original data with several bands is mapped to several new principal components. Each principal component is formed by the linear summation of its eigenvectors. In mathematics, this means identifying some new variables. ( i =1, 2, ..., p ), making them linear functions of X and uncorrelated with each other, as shown below.

[0110] .

[0111] in, These are the eigenvector components. It is a vector composed of the original data. In essence, it's about finding... p 2 constants ( i , k =1,…, p (Represented in matrix form)

[0112] .

[0113] .

[0114] in, L For the eigenma matrix, each These are the components of this eigenvector; for C The eigenvalues ​​of a matrix. and L It has the following characteristics.

[0115] This is called a trace, or total variation, corresponding to different of L (That is, each principal component) is linearly uncorrelated; and orthogonal.

[0116] From linear algebra, we know the covariance matrix.C The eigenpolynomial is det( I - C The roots of this eigenpolynomial Both are covariance matrices C The eigenvalues ​​of .

[0117] The calculation process is shown below.

[0118] Find the covariance moments C .

[0119] .

[0120] Find the eigenvalues .

[0121] | I |- C =0.

[0122] I = .

[0123] in, I It is an identity matrix.

[0124] Find the eigenvectors L .

[0125] ( IC ) L =0.

[0126] exist N When the coordinate axes of the band data are transposed, the covariance matrix will also be transformed, and after the transformation, the covariance between each band becomes zero.

[0127] The sum of the squares of the distances from each point to its centroid is the sum of the eigenvalues, which can be represented as S. In a sense, it can be said that the ratio of the variation "constituted" by the first component to the total variation is... 1 / S, the ratio of the variation "composed" by the first two components to the total variation is ( 1+ 2) / S, and so on. Sometimes, for convenience, we can say, for example, "the first four components constitute the p% of the variation".

[0128] The eigenvalue of a principal component is the mean square error introduced into the corresponding eigenvector after the principal component is eliminated.

[0129] Ultimately, four principal components are formed.

[0130] .

[0131] 3) Independent component analysis.

[0132] The original data is transformed using Independent Component Analysis (ICA). ICA is a statistical signal processing method that decomposes observed data into statistically independent non-Gaussian signal sources through linear transformation. It is assumed that the observed random signal X follows a model.

[0133] .

[0134] in, It is a known data matrix. For an unknown mixture matrix, [Unknown] IC The matrix has independent components. E Let be the residual matrix. X Each component can be derived from S The components are linearly represented. The basic principle is based on the known data matrix. X Estimate independent components S and mixture matrix M .

[0135] The first step is to zero out the value. X Subtract the mean from each eigenvector to ensure that the mean is zero in each dimension.

[0136] .

[0137] in, E [ X [A known matrix] X The mean of the eigenvectors, This is the matrix after being zeroed out.

[0138] The second step is whitening, which transforms the matrix into a new space so that its covariance matrix is ​​the identity matrix.

[0139] .

[0140] Perform eigenvalue decomposition on the covariance matrix.

[0141] .

[0142] in, E The eigenvector matrix, D The diagonal eigenvector matrix is ​​shown below after whitening transformation.

[0143] .

[0144] in, This is the matrix after whitening transformation.

[0145] The third step is to maximize the non-Gaussianity.

[0146] Choose a non-Gaussian metric (such as negative entropy) and find the projected weights that maximize the metric. Use a non-linear function g(·) to approximate the maximization of negative entropy. The weights are as follows.

[0147] .

[0148] For the new weight vector Normalization is performed.

[0149] .

[0150] The fourth step, orthogonalization and convergence, requires that the independent component transformations be mutually orthogonal. This is achieved using the Gram-Schmidt process. i Each component, update rules considered i -1 component.

[0151] .

[0152] right After normalization, check for convergence. and If the difference between the components is less than a certain threshold, then an independent component is considered to have been found.

[0153] Step 5: Independent component analysis.

[0154] Once all weight vectors are found Independent components can be estimated using the following formula. S .

[0155] .

[0156] The original data is a high-resolution image of 4 bands, so 4 independent components can be obtained.

[0157] .

[0158] 4) Minimum noise transformation.

[0159] The original data is transformed using a minimum noise separation transform. This minimum noise separation transform separates noise from the data, reducing computational requirements in subsequent processing. Essentially, it involves two cascaded principal component transformations. The first transformation separates and readjusts the noise in the data, while the second step is a standardized principal component transformation of the noise-whitened data.

[0160] The first step is to use the covariance matrix. Diagonalization is performed to form a diagonalized matrix. .

[0161] .

[0162] U Let be an orthogonal matrix composed of eigenvectors. Further transformation yields the following formula.

[0163] .

[0164] .

[0165] in, I It is the identity matrix. W For the transformation matrix, put W Application to high-resolution satellite data X It can be done Y=WX The original image is transformed into a new space.

[0166] The second step is to perform standardized principal component transformation on the noisy data.

[0167] .

[0168] in, Let the image be the covariance matrix. For the process W The transformed matrix is ​​further diagonalized to form a diagonalized matrix.

[0169] .

[0170] in, for A diagonal matrix composed of the eigenvalues. Q The orthogonal matrix composed of eigenvectors is used to obtain the minimum noise transformation matrix through the above two steps. M。

[0171] .

[0172] 5) Band encapsulation.

[0173] The eigenvectors obtained from various transformations are encapsulated with the original wavebands to ultimately form a virtual spectrum.

[0174] SH =[ B 1, B 2, B 3, B 4, 1, 2, 3, 4, 5, 6, P 1, P 2, P 3, P 4, I 1, I 2, I 3, I 4, 1, 2, 3, 4).

[0175] 6) Angle and intensity constraints.

[0176] The virtual spectrum after band encapsulation is subjected to angle and intensity constraints. The angle constraint uses the spectral angle, and the intensity constraint uses the intensity tube current.

[0177] Spectral angle , As a reference virtual spectral vector; For image virtual spectral vectors, For vectors sum vector The inner product; For vectors Length; For vectors The length.

[0178] Intensity Pipe Flow , It is a limit value.

[0179] The final signal is shown below.

[0180] .

[0181] ③ Use band information to build a neural network to automatically extract the project type.

[0182] In 1986, Rumelhart et al. proposed the backpropagation algorithm for artificial neural networks, which outperformed shallow machine learning models such as support vector machines, bootsting, and logistic regression based on statistical learning theory. In 2006, Hinton et al. proposed that multi-hidden-layer deep learning artificial neural networks possess excellent feature learning capabilities. They can overcome training difficulties through layer-by-layer unsupervised training, utilize nonlinear operators in deep structures to obtain higher-level features, and improve recognition rate and accuracy. They mainly consist of an input layer, convolutional layers, sampling layers, connection layers, and an output layer. A typical structure diagram of a convolutional neural network is shown below. Figure 5 As shown.

[0183] Convolutional filtering is performed on each channel of the input image, and features from the image training area are extracted to obtain feature maps with the same number of convolutional kernels. These features are then passed through a resampling layer to obtain feature maps that are small in size but numerous in number. The feature maps are then unfolded in a certain way and concatenated into a one-dimensional vector connection layer. After passing through several connection layers, the output is obtained, thereby achieving the recognition task.

[0184] A convolutional layer can be represented by the following formula, the first... l Layer j Feature matrices It can be obtained by weighted convolution of several feature maps from the previous layer.

[0185] .

[0186] in, For neuron activation functions, This represents a combination of input feature maps. Represents convolution operation. The convolution kernel matrix, The bias matrix, For the first i -1st floor i The feature matrix; commonly used neuron activation functions include: Sigmoid function, Tanh function, ReLU function, etc.

[0187] The sampling layer, also known as the "pooling layer," is used to perform pooling sampling based on the principle of local correlation, thereby reducing the amount of data while retaining useful information.

[0188] .

[0189] in, This refers to sampling functions, including maximum sampling and mean sampling. For the first l -1st floor j A feature matrix.

[0190] After convolutional and sampling layers, they are connected to a linking layer. Each neuron is connected to every neuron in the next layer, and so on. l The feature vector of a layer connection layer can be represented by the following formula.

[0191] .

[0192] in, The weight matrix, This is the bias vector. For the first l -1 layer connection layer feature vector.

[0193] When the final output layer of the model is a logistic regression layer, each node output by the convolutional neural network represents a certain category of the input data. i The probability of.

[0194] .

[0195] in, For the last layer i Individual weight parameters, b For the corresponding bias parameters.

[0196] ④ Accuracy verification and validation of project type extraction.

[0197] like Figure 6 As shown, the output results of various types of construction project information obtained by the neural network are used as the input layer of the judgment device. If it is construction project information of this type, it is retained; if it is not construction project information of this type, it is deleted. If it is uncertain, the virtual spectrum of this information is collected and the spectral angle is calculated with the reference virtual spectrum. The size of the spectral angle is judged. If the spectral angle of lithium-containing pegmatite is smaller than that of lithium-free pegmatite, it is retained; otherwise, it is deleted.

[0198] The spectral angle method represents each multidimensional spatial point with its spatial vector and compares the similarity of the spatial vector angles. It is a supervised classification method. It requires a known reference spectrum for each category. This reference spectrum can be obtained from ground measurements and stored in a reference spectrum library, or it can be statistically analyzed from regions of interest (ROIs) of map cells with known conditions and stored in the reference spectrum library. The formula is shown below.

[0199] .

[0200] In the formula, For reference spectrum; α , β )for n dimensional vector α , β The inner product is defined according to the inner product definition.

[0201] ( α , β )= α 1 β 1+ α 2 β 2+…+ αnβn .

[0202] when α , β When it is a column vector ( α , β )= β= α.

[0203] ; .

[0204] .

[0205] | α |、| β | represents the lengths of vectors α and β.

[0206] |α|= = = .

[0207] By finding the inner product and length of α and β, cos can be calculated. The included angle can be obtained by looking up the table, and finally, manual verification is carried out.

[0208] ⑤ Vectorize the scope of the construction project.

[0209] The construction projects obtained through remote sensing are vectorized.

[0210] (ii) The maximum scope of non-approved overlay is delineated.

[0211] 1. Assessment and delineation of the safety protection scope of the construction project.

[0212] It should be noted that where other laws, regulations, or standards do not provide clear stipulations, the scope of protection shall be determined according to the documents or standards of each province. If the documents or standards of each province do not provide clear definitions, the main body to be covered shall be classified according to the actual situation, and the safety protection scope shall be determined according to the special grade, first grade, second grade, third grade, and fourth grade. (1) Underground mining area: special grade 50 meters, first grade 20 meters, second grade 15 meters, third grade 10 meters, and fourth grade 5 meters. (2) Open-pit mining and the upper part of the main body to be covered: special grade 1000 meters, first grade 600 meters, second grade 500 meters, third grade 200 meters, and fourth grade 100 meters. The safety protection scope of major construction projects is shown in Table 1.

[0213] Table 1 Safety Protection Scope for Major Construction Projects

[0214] If there is overlap, it is directly interpreted as overlapping; if there is no overlap, an estimation is required for interpretation.

[0215] Based on the boundary of reserve estimation and the distance from the safety protection zone of major construction projects L and the minimum burial depth of the ore body h The ratio between them, the angle of movement / angle of slope α The following formula is used for a rough estimate.

[0216] L ≥ h ×tan( αIt does not cover mineral resources, so there is no need to estimate the coverage.

[0217] L < h ×tan( α If a mineral resource is likely to be covered, it needs to be determined, and if there is coverage, the coverage needs to be estimated.

[0218] 2. Assessment and delineation of the safe operation scope of the construction project.

[0219] The safe operating range is determined according to laws, regulations, standards, and norms. The overburden area includes the area above the overburdened main body, open-pit mining areas, and underground mining areas. For solid minerals, the safe operating range is determined by the safety protection zone above the overburdened main body; for open-pit mining areas, the safe operating range is mainly determined by calculating the slope angle; for underground mining areas, the safe operating range is mainly determined by calculating the movement angle; and for liquid minerals, the safe operating range is determined by the safety protection zone (generally considered as complete overburden or no overburden treatment is required). For deep mineral resources with a large depth ratio, the safe operating range is determined by the safety protection zone. In this case, if the landslide caused by mining does not affect the surface, no overburden treatment is required.

[0220] 3. Delineate the maximum area to be covered by the construction project.

[0221] After determining the protection boundary, spatial overlay analysis is performed on the reserve assessment map, ore body distribution map, and geological profile map using methods such as vertical profile, plumb line (for extended structures), and digital elevation projection (for varying dip angles), based on slope angles and movement angles, or by using vertical cross-sections, vertical line methods, and digital elevation projection (for dipping structures). The vertical cross-section method is commonly used. On a large-scale plan view, vertical cross-sections are drawn with the center point of the safety protection boundary, indicating the direction and dip. Based on the topography, stratigraphic column, and the selected movement angle and slope angle (90° for vertical cross-sections), a large-scale cross-section is drawn. The protection boundary point of the protected pillar is estimated on the cross-section, and the maximum coverage area and elevation range are determined. The coverage area and elevation range of each ore body (layer) are also determined. For multi-layered ore bodies, the coverage area of ​​each layer must be calculated and then transferred to the plan view. Connecting the lines can estimate the coverage area of ​​the protected pillar. The maximum value is taken for multi-layered ore bodies.

[0222] (iii) Assessment of the amount of resources covered by non-approved infrastructure.

[0223] 1. Depth ratio and movement angle estimation.

[0224] ① Determine the slope angle or the angle of movement.

[0225] The slope angle is a reserved slope in open-pit mining. The boundary is determined by the stripping ratio and the boundary stripping ratio, and the slope angle is ultimately determined by the bench slope surface, safety platform, cleaning platform, and transport platform. The determination of the slope angle in open-pit mining is as follows: Figure 7 As shown in the figure, mFor safety protection scope; h For platform height; α The slope angle of the steps; β The slope angle; H The depth of the mine; L for H (Deep safe operating range half-width), the safe operating range of open-pit mining is determined as follows: Figure 8 As shown in the figure, m For safety protection scope; L for H The safe operating range is half-width. Slope selection employs analogy, verification design, and optimization design methods. Influenced by rock physical properties, geological structure, hydrogeology, and mining technology conditions, the step slope angle and side slope angle are shown in Tables 2 and 3. When the geological profile is not perpendicular to the slope direction, the slope angle should be converted. . The apparent dip angle of the slope. For the slope angle, The angle between the cross section and the orthogonal cross section on the plane.

[0226] Table 2 Reference Data Table for Step Slope Angle

[0227] Table 3. Rock classification and approximate data on open-pit mine slope angles based on slope stability.

[0228] The movement angle is determined by analogy to the "Specifications for Coal Pillar Retention and Coal Mining in Buildings, Water Bodies, Railways and Main Roadways," and is comprehensively determined based on the rock type, mechanical characteristics, and geological characteristics of the overlying geological body. The determination of the movement angle in underground mining is as follows... Figure 9 As shown in the figure, L for H Deep security operating range half-width; HH The depth of the ore body; hh The thickness of the loose layer; n The thickness of the bedrock; m For safety protection scope; γ For moving uphill; β For the angle of movement downhill; α 1, α 2 represents the bedrock movement angle; Ф 1, Ф 2 represents the angle of movement of the loose layer. The safe operating range for underground mining is determined as follows: Figure 10 As shown in the figure, L for H Deep security operating range half-width; J for H The depth-adaptive direction has a wide safe operating range;m For safety protection scope; (The angle of inclination).

[0229] Generally, a movement angle is designed for loose layers and a movement angle is designed for bedrock. If the mechanical properties of bedrock stratification are particularly obvious (such as the presence of loose layers in the bedrock), a separate movement angle is set for bedrock.

[0230] Move towards the angle Downhill movement angle Uphill Movement Angle The movement angles of loose layers are shown in Tables 4 and 5. When estimating the safe operating range, the movement angle, topography, strata, and ore body characteristics are generally considered. Multiple representative profiles are typically created to determine the safe operating range. Steeply dipping ore layers, waterproofing, and fault prevention require consideration of various factors; see the "Regulations for Coal Mine Under Three Levels".

[0231] When the average slope angle of the ore body or bedrock surface is greater than 15°, protective pillars should be provided to accommodate the movement angle. If no measured data is available, the uphill movement angle should be reduced by 5° to 10°; the downhill movement angle should be reduced by 2° to 3°. The influence of faults should be considered if faults are present.

[0232] Move towards the angle Downhill movement angle Uphill Movement Angle The relationship between them is , , This refers to the dip angle of the ore layer. If it's an oblique profile, consider the downhill displacement angle. View of the uphill movement angle Replace the downhill movement angle Uphill Movement Angle Draw a trapezoidal profile. The acute angle between the outer boundary of the safety protection zone and the strike line of the ore layer is... The relationship between them is as follows.

[0233] .

[0234] .

[0235] Table 4. Angle of movement of loose layer Reference value

[0236] Table 5 Reference values ​​for bedrock movement angle

[0237] ② Depth ratio.

[0238] Mining depth ( H ) and normal mining thickness (M The ratio of ) H / M This parameter is widely used in geological hazard prediction, goaf stability assessment, and ground subsidence model selection.

[0239] Shallow mining: If the depth ratio is less than 30, the surface is prone to collapse and should be covered.

[0240] Medium-deep mining: For depth ratios of 30 ≤ depth ≤ 60, deformation risk needs to be monitored, and overburden treatment should be implemented.

[0241] Deep mining: with a depth ratio >60, mining safety is relatively high, and no overburden treatment is required.

[0242] The determination of the depth ratio is as follows.

[0243] ≥70: No significant impact, no overlay treatment required.

[0244] 60-70: Low risk of deformation, no overlay treatment required.

[0245] <60: Subsidence prevention measures must be taken, including burying the affected area.

[0246] 2. Estimation of resources covered by non-approved (actual) conditions.

[0247] All reserve estimation parameters, methods, and block divisions are consistent with the exploration report. The industrial indicators for the deposit should be consistent with current industrial indicators; any inconsistencies between the industrial indicators in the report and current indicators should be corrected to the current indicators.

[0248] Basic parameters for estimating reserves covered by geological structures (these parameters can be determined based on the depth ratio and the angle of movement) include the area, average thickness, dip angle, grade, and average weight of the ore body within the covered area. Sometimes, ore moisture content and ore-bearing coefficient are also included. The most recent exploration report data should be used. Based on factors such as the geological body occurrence, faults, and exploration projects, the location of the profile lines is selected. The vertical profile method is used to create the covered boundaries of each profile line. These boundaries are then projected onto the reserve estimation map, and the boundary points of each profile line are connected to delineate the covered area.

[0249] Depending on the extent of the land cover, geometric methods (arithmetic mean method, geological block method, mining block method, cross-section method, contour line method, linear reserve method, trigonometric method, nearest area method, polygonal method), statistical analysis methods (distance-weighted method, kriging method), and SD method are used to calculate the amount of land cover. The geological block method and cross-section method are commonly used.

[0250] Basic formula: .

[0251] in, These are unapproved resource reserves that have been covered. This represents the total number of overlaid block segments; For the first Volume of each overlay block segment; For the first Apparent density of each overburden block segment.

[0252] Special treatment: If the reserves after deducting the area covered by the cover have no economic value for extraction, they should be treated as the entire area covered. If the covered body is divided into several parts, the part that has no extraction value should be treated as covered.

[0253] ① Overlapping mines.

[0254] In mines covered by overlapping structures, it is necessary to investigate the current status of resource utilization.

[0255] If the area covered by a producing or closed mine includes a goaf, the reserves of the corresponding goaf should be deducted.

[0256] If the remaining mine has been completely mined out and no covering treatment is carried out, or if it includes a goaf, the corresponding goaf reserves should be deducted.

[0257] ② Covering unused mining areas.

[0258] For areas covered by unused mining areas, the slope angle and movement angle are estimated using the vertical profile method, plumb line method, and digital elevation projection method, depending on the level of detail in the exploration data.

[0259] ③ Mineral deposits to be confirmed are covered by the overburden.

[0260] For mineral deposits that are to be confirmed but are covered by other materials, estimates are made based on the level of detail in the data. If a vertical profile cannot be constructed, a planar estimation projection arithmetic mean method is used.

[0261] ④ Treatment of overlapping and overlapping of multiple mining areas (mineral deposits) and mines.

[0262] For overlapping and pressing, one of the methods can be used for estimation.

[0263] ⑤ Treatment of multiple (parallel) linear buildings covering the same mining area.

[0264] When multiple (parallel) linear structures (such as railways, highways, oil and gas pipelines, etc.) cover the same mining area or mine, the protection zone and the boundary of the outermost two structures are taken, and the slope angle and the movement angle are estimated using the vertical profile method, the plumb line method, and the digital elevation projection method.

[0265] ⑥Completely cover.

[0266] If the remaining uncovered resources after the mining area coverage estimate are not feasible for mining, they will be treated as completely covered.

[0267] (iv) Final drawing.

[0268] 1. Vector and raster overlay device.

[0269] The optimized maximum coverage vector is input, along with the R (red): 669.43nm, G (green): 538.96nm, B (blue): 479.25nm bands from a high-resolution satellite remote sensing image. These are used as inputs for a composite RGB (red, green, blue) color image. The composite color image uses points, lines, and surfaces with the same projection to represent the vectors. Coordinate layering is then used to overlay the raster and vector data.

[0270] vector , For the corresponding coordinates, z For eigenvalues, Vector values, raster , and For the corresponding coordinates, This represents the raster grayscale value.

[0271] make Thus achieving a grid Grayscale values ​​and vectors The superposition of.

[0272] By using corresponding coordinates to overlay raster and vector data, the output is a GIS vector and raster overlay image, thus creating an image that is suitable for human vision.

[0273] 2. Final image.

[0274] Input high-resolution satellite remote sensing imagery and maximum coverage vector, output a superimposed image (GIS) suitable for human visual observation, which can be output as a final image in JPG or TIF format via software.

[0275] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 11As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores remote sensing imagery. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a non-approval overburden assessment method for mineral resources based on high-resolution remote sensing.

[0276] Those skilled in the art will understand that Figure 11 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0277] In one exemplary embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-described method embodiments.

[0278] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the above-described method embodiments.

[0279] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the above-described method embodiments.

[0280] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0281] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0282] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0283] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0284] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for assessing non-approved mineral resource overlay based on high-resolution remote sensing, characterized in that, The method for assessing non-approved mineral resource overlay based on high-resolution remote sensing includes: Acquire remote sensing images; The remote sensing image is transformed to obtain a transformed image; each point on the transformed image has the same coordinates as the real-world coordinates. The project information is automatically extracted from the converted image to obtain the project vector; Based on the project vector, calculate the depth ratio and the angle of movement; The safety protection range and the safe operating range are determined based on the depth ratio and the movement angle. Based on the aforementioned safety protection range and the aforementioned safe operation range, determine the maximum non-approved coverage area; The amount of resources to be covered by non-approved resources is determined based on the maximum non-approved coverage area. Based on the amount of non-approved overlaid resources, a map was created to obtain the final map.

2. The method for non-approval-based assessment of mineral resource overlay based on high-resolution remote sensing as described in claim 1, characterized in that, The remote sensing image is subjected to coordinate transformation to obtain the transformed image, specifically including: The remote sensing image is geometrically corrected to obtain a corrected image; the geometric correction includes: image-to-image geometric correction, image control point geometric correction, and image-to-image automatic image registration; The corrected image is then subjected to coordinate transformation and orthorectification to obtain the transformed image.

3. The method for non-approval-based assessment of mineral resource overlay based on high-resolution remote sensing as described in claim 1, characterized in that, The converted image is then automatically processed to extract project information, resulting in a project vector, specifically including: Based on the converted image, a high-resolution image is determined; the high-resolution image is visible light-near infrared 4-band or 8-band data. Based on the high-resolution imagery, the type of construction project is determined; Band information transformation is performed on the original bands of high-resolution images to obtain band combinations; the band information transformation includes: band ratio, principal component analysis, independent component analysis, and minimum noise transformation; By combining the bands with the original bands and encapsulating them, a virtual spectrum is obtained. The virtual spectrum is subjected to constraint processing to obtain the final signal; the constraint processing includes: angle constraint and intensity constraint. The final signal is input into a convolutional neural network to obtain preliminary project information; Based on the project type, the accuracy of the preliminary project information is checked and verified to obtain the final project information; The final project information is vectorized to obtain the project vector.

4. The method for non-approval-based assessment of mineral resource overlay based on high-resolution remote sensing as described in claim 3, characterized in that, The expression for the angle constraint is: ； The expression for the strength constraint is: ； The expression for the final signal is: ； in, It is the spectral angle; For high-intensity pipe flow; For the final signal; As a reference virtual spectral vector; For image virtual spectral vectors, For vectors sum vector The inner product; For vectors Length; For vectors Length; It is a limit value.

5. The method for non-approval-based assessment of mineral resource overburden based on high-resolution remote sensing according to claim 3, characterized in that, Based on the aforementioned project type, the accuracy of the preliminary project information is checked and verified to obtain the final project information, specifically including: Determine whether the preliminary construction project information belongs to the construction project type to obtain a first determination result; If the first judgment result is yes, then retain it; If the result of the first judgment is negative, then delete it; If the first judgment result is uncertain, then virtual spectral acquisition is performed on the preliminary construction project information to obtain the image virtual spectrum; The image virtual spectrum is compared with the reference virtual spectrum to calculate the spectral angle, and it is determined whether the spectral angle of lithium-containing pegmatite is smaller than that of lithium-free pegmatite to obtain a second judgment result. If the result of the second judgment is yes, then retain it; If the second judgment result is negative, then delete.

6. The method for non-approval-based assessment of mineral resource overlay based on high-resolution remote sensing as described in claim 1, characterized in that, Based on the safety protection range and the safe operation range, the methods for determining the maximum non-approved coverage range include: vertical profile method, plumb line method and digital elevation projection method.

7. The method for non-approval-based assessment of mineral resource overlay based on high-resolution remote sensing as described in claim 1, characterized in that, The formula for calculating the amount of resources covered by non-approved infrastructure is as follows: ； in, These are unapproved resource reserves that have been covered. This represents the total number of overlaid block segments; For the first Volume of each overlay block segment; For the first Apparent density of each overburden block segment.

8. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that the processor executes the computer program to implement the non-approval overlay assessment method for mineral resources based on high-resolution remote sensing as described in any one of claims 1-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the non-approval overlay assessment method for mineral resources based on high-resolution remote sensing, as described in any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, When executed by a processor, the computer program implements the non-approval overlay assessment method for mineral resources based on high-resolution remote sensing, as described in any one of claims 1-7.

Images