Remote sensing alteration zone-based potential mineral resource prediction area rapid delineation method

Abstract

The application discloses a method for quickly delineating a potential mineral resource prediction area based on remote sensing alteration zones, and relates to the technical field of mineral resource analysis. The method comprises the following steps: based on remote sensing data, using atmospheric correction and mineral mapping to extract alteration zone information; performing rasterization processing on mineral investigation data to obtain a mineral investigation raster image; performing registration analysis and data fusion on the alteration zone information and the mineral investigation raster image to obtain fused data; generating a vector cloud space according to coordinates based on the mineralization zone and the fused data, and performing mineralization zone restriction analysis based on the vector cloud space; and determining the potential mineral resource prediction area based on the mineralization zone restriction analysis result. The application can improve the delineation efficiency and precision of the potential mineral resource prediction area by superimposing remote sensing data and mineral investigation data.

Description

Technical Field

[0001] This application relates to the field of mineral resource analysis technology, and in particular to a method for rapidly delineating potential mineral resource prediction areas based on remote sensing alteration zones. Background Technology

[0002] Potential mineral resources are the unverified portion of solid mineral resources predicted based on geological evidence and geophysical and geochemical anomalies. Therefore, quantitative prediction of undiscovered potential mineral resources in a region is an important task in mineral resource evaluation. The analysis of mineral resources in existing potential mining areas generally involves first using the volumetric method to estimate undiscovered mineral resources through Vickers, and then estimating mineral resources in areas where resources have been discovered. This method is inefficient and the accuracy of the analysis results is low. Summary of the Invention

[0003] The purpose of this application is to provide a rapid delineation method for potential mineral resource prediction areas based on remote sensing alteration zones, which can improve the delineation efficiency and accuracy of potential mineral resource prediction areas.

[0004] To achieve the above objectives, this application provides the following solution: This application provides a method for rapidly delineating potential mineral resource prediction areas based on remote sensing alteration zones, including: Acquire remote sensing data and mineral survey data for the target area; Based on the remote sensing data, alteration zone information was extracted using atmospheric correction and mineral mapping. The mineral survey data is rasterized to obtain a mineral survey raster image. The alteration zone information and the mineral survey raster image are registered, analyzed, and fused to obtain fused data. The metallogenic belt and fused data are used to generate a vector cloud space according to coordinate correspondence, and the metallogenic belt constraint analysis is performed based on the vector cloud space; Based on the results of the metallogenic belt constraint analysis, potential mineral resource prediction areas were identified.

[0005] Optionally, based on the remote sensing data, alteration zone information is extracted using atmospheric correction and mineral mapping, specifically including: The remote sensing data is subjected to quality assessment to obtain remote sensing data that meets the quality assessment requirements. The indicators used in the quality assessment include: spectral angle, root mean square error, peak signal-to-noise ratio, structural similarity, average structural similarity, and image homogenization error. Atmospheric correction processing is performed on remote sensing data that meets the quality assessment requirements to obtain corrected remote sensing data; The corrected remote sensing data is subjected to border removal processing to obtain border-removed remote sensing data; The borderless remote sensing data is processed using interference removal methods to obtain interference-free remote sensing data; the interference removal methods include: ratio method, segmentation method, Q-value method and spectral angle method; Mineral mapping is carried out based on the aforementioned de-interference remote sensing data; Based on the comprehensive overlay analysis of the mineral mapping information, the alteration zone information is determined by the combination of alteration minerals.

[0006] Optionally, the corrected remote sensing data is subjected to border removal processing to obtain border-removed remote sensing data, specifically including: Determine any band as the current band; Determine the binary image corresponding to the current band: For each point in the target area, assign a value of 1 if there is corrected remote sensing data for the current band, and assign a value of 0 if there is no corrected remote sensing data for the current band, thus obtaining the binary image corresponding to the current band. Update the current band and return to the step "determine the binary image corresponding to the current band" until all bands are traversed to obtain the binary image corresponding to each band; Multiply the binary images corresponding to all bands to obtain a debounded binary image; The corrected remote sensing data is processed by removing the borders using a de-bordered binary image to obtain de-bordered remote sensing data.

[0007] Optionally, based on the mineral mapping and comprehensive overlay analysis of alteration mineral information, alteration zone information is determined through alteration mineral assemblages, specifically including: Based on the mineral mapping, determine whether there are zonations; When there are zones, select k minerals as the inner zone alteration mineral assemblage and select h minerals as the outer zone alteration mineral assemblage; When there is no zoning, p minerals are selected as the alteration mineral assemblage; where the values ​​of k, h, and p are determined according to the deposit type of a typical deposit. Alteration mineral assemblage is used as information about alteration zones.

[0008] Optionally, the mineral survey data is rasterized to obtain a mineral survey raster image, specifically including: List the coordinates and resource values ​​of the mineral survey data to form a vector layer; Interpolation processing is performed on the vector points in the vector layer to obtain a mineral survey raster image: for vector points in the vector layer with a range greater than the range threshold, the multifractal method is used for interpolation processing; for vector points in the vector layer with a range less than or equal to the range threshold, the multi-kriging method is used for interpolation processing.

[0009] Optionally, the alteration zone information and the mineral survey raster image are registered, analyzed, and fused to obtain fused data, specifically including: Based on the alteration zone information, the mineral assemblage traversed by each alteration zone is determined; Identify any alteration zone as the current alteration zone; After performing minimum noise separation transformation on the images of multiple bands corresponding to the current alteration zone, a multidimensional image of the current alteration zone is obtained; the multidimensional image includes multiple single-channel images; the channels of the multidimensional image correspond one-to-one with the mineral combinations traversed by the alteration zone; and the multiple bands correspond one-to-one with the mineral combinations traversed by the alteration zone. Select any channel as the current channel; Determine the structural similarity coefficient between the mineral survey raster image and the single-channel image of the current channel; If the structural similarity coefficient is greater than the structural similarity coefficient threshold, the single-channel image of the current channel in the mineral survey raster image is replaced with the mineral survey raster image. Update the current channel and return to the step "Determine the structural similarity coefficient between the mineral survey raster image and the single-channel image of the current channel" until all channels are traversed to obtain the fused data of the current alteration zone; Update the current alteration zone and return to step "After performing minimum noise separation transformation on the images of multiple bands corresponding to the current alteration zone, obtain the multidimensional image of the current alteration zone", to obtain the fused data of each alteration zone.

[0010] Optionally, a vector cloud space is generated by mapping the metallogenic belt and the fused data to coordinates, and a metallogenic belt constraint analysis is performed based on the vector cloud space, specifically including: The metallogenic belt and fused data are used to generate corresponding data according to coordinates, forming a vector cloud space; Based on the aforementioned vector cloud space, construct a vector co-occurrence matrix: take a vector in the vector cloud space. and another vector offset along a certain direction and by an offset distance Forming a matrix If the calculation involves movement within a spatial vector cloud, then multiple matrices are obtained. Statistical matrix The number of times an opportunity arises is the first opportunity number. Dividing the first opportunity number by the total number of opportunities yields the first opportunity probability density, forming a vector co-occurrence matrix. Based on the vector co-occurrence matrix, co-occurrence constraint indices are calculated; the co-occurrence constraint indices include: entropy, inertia, and energy; Based on the aforementioned vector cloud space, construct a vector difference matrix: take a vector in the vector cloud space. and another vector offset along a certain direction and by an offset distance The spatial angles are calculated. If the calculation involves movement within a spatial vector cloud, the spatial angles of multiple vectors are obtained. The number of outflows of the spatial angles is counted as the second outflow count. The second outflow count is divided by the total number of outflows to obtain the second outflow probability density, forming a vector difference matrix. The number of rows in the vector difference matrix is ​​equal to the number of rows in the vector difference matrix. The number of columns in the vector difference matrix is ​​equal to the number of columns in the vector difference matrix. Based on the vector difference matrix, the difference constraint index is calculated; the difference constraint index includes: curvature, tensile strength, and enthalpy; Based on the symbiotic constraint index and the differential constraint index, an analysis of the constraints on metallogenic belts is conducted.

[0011] Optionally, the entropy is: ; in, ; In the formula, Entropy; This represents the total number of rows in the vector co-occurrence matrix; This represents the total number of columns in the vector co-occurrence matrix; Let be the element in the i-th row and j-th column of the vector co-occurrence matrix; Let i be the i-th vector in the vector cloud space; Let j be the j-th vector in the vector cloud space; For matrix The first number of times one can achieve success; Total number of times; The inertia is: ; In the formula, For inertia; The energy is: ; In the formula, For energy; The curvature is: ; ; In the formula, Curvature; Let be the element in the i-th row and j-th column of the vector difference matrix; Vector before and after movement The spatial angle; Vector before and after movement The spatial angle; This is the second highest number of times a student can earn a living. The tensor is: ; In the formula, For the degree of tension; The enthalpy is: ; In the formula, It is enthalpy.

[0012] Optionally, based on the results of the metallogenic belt constraint analysis, potential mineral resource prediction areas are determined, specifically including: When both the symbiotic constraint index and the differential constraint index are within a preset range, the ore-forming zone is determined to be a forced ore-forming zone; All alteration zones within the forced ore zone are identified as undetermined alteration zones. By deleting known alteration zones from the undetermined alteration zones, multiple prediction areas are obtained; the known alteration zones are based on mineral survey data. A multiple linear regression model is constructed by defining a random variable as the dependent variable and multiple prediction regions as independent variables. A set of equations for a multiple linear regression model is constructed using multiple sets of observation data, and the coefficients of the multiple linear regression model are obtained by solving the set of equations using the least squares estimation method. Determine any one of the independent variables as the current independent variable; Analysis of deviations was used to test the hypotheses in a multiple linear regression model before and after removing the current independent variable. The absolute value of the difference between the coefficients of determination before and after removing the current independent variable in a multiple linear regression model is used as the discriminant; the coefficients of determination are: ;in, is the coefficient of determination for the multiple linear regression model; SS is the sum of squares of the regressions for the multiple linear regression model. The total variance of the multiple linear regression model is represented by MS, which is the sum of squared residuals. When the discriminant is less than the discriminant threshold, the prediction region corresponding to the current independent variable is deleted; Update the current independent variable and return to the step "Use deviation analysis to perform hypothesis testing on the multiple linear regression model before and after deleting the current independent variable" until all independent variables are traversed to obtain the potential mineral resource prediction area.

[0013] Optionally, after determining the potential mineral resource prediction area based on the results of the metallogenic belt constraint analysis, the method further includes: Selecting the optimal band combination using the optimal index method; Based on the optimal band combination, the grid and vector are superimposed using coordinate layering to generate a schematic diagram of the potential mineral resource prediction area: the base map is a false color map with the band combination with the highest information entropy, and the vectors are represented by points, lines and surfaces with the same projection.

[0014] According to the specific embodiments provided in this application, the following technical effects are disclosed: This application provides a rapid delineation method for potential mineral resource prediction areas based on remotely sensed alteration zones. The method involves fusing remotely sensed alteration zones with mineral survey data. First, hyperspectral satellite remote sensing data is acquired, and alteration zone information is extracted. Based on this, national survey data is analyzed and fused with the remotely sensed alteration zones. By considering the constraints of metallogenic belts, potential mineral resource prediction areas are rapidly delineated. Finally, the predicted areas are overlaid with image maps. This provides a new technical method for potential mineral resource prediction and is of great significance for comprehensively utilizing mineral survey data in this field. Attached Figure Description

[0015] 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.

[0016] Figure 1 This is a general technical flow diagram of one embodiment of this application; Figure 2 This is a flowchart of a method for rapidly delineating potential mineral resource prediction areas based on remote sensing alteration zones, according to one embodiment of this application. Figure 3 This is a schematic diagram of border information in one embodiment of this application; Figure 4 A comparison image showing the removal of boundary information in one embodiment of this application; Figure 5 This is a schematic diagram of mineral survey data in one embodiment of this application; Figure 6 This is a schematic diagram of the predicted region defined in one embodiment of this application. Detailed Implementation

[0017] 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.

[0018] 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.

[0019] In one exemplary embodiment, a method for rapidly delineating potential mineral resource prediction areas based on remotely sensed alteration zones is provided. This method rapidly delineates potential mineral resource prediction areas using remotely sensed alteration zones. It primarily relies on information from national surveys regarding mining areas, mining rights, exploration rights, minimum prediction zones, and mineral deposits. This information is overlaid with remotely sensed alteration zones, and after analyzing the metallogenic belts from the national surveys, remotely sensed alteration zones within the same metallogenic belt that are correlated with information such as mining areas, mining rights, exploration rights, minimum prediction zones, and mineral deposits are selected as potential mineral resource prediction areas. The core technology is [see overall technical flowchart]. Figure 2 The method for rapidly delineating potential mineral resource prediction areas based on remote sensing alteration zones includes the following steps: data acquisition (step 201), remote sensing alteration zone information extraction (step 202), mineral (national conditions) survey data analysis (step 203), data fusion (step 204), metallogenic belt constraint analysis (step 205), prediction area delineation (step 206), and final mapping (step 207).

[0020] like Figure 1 As shown, a method for rapidly delineating potential mineral resource prediction areas based on remote sensing alteration zones is provided, including: Step 201: Acquire remote sensing data and mineral survey data for the target area. The mineral survey data is based on national survey data; national survey data is only based on the most accurate data from field surveys. Combining national survey data with remote sensing alteration zones will improve efficiency and accuracy in forecasting.

[0021] Step 202: Based on remote sensing data, extract alteration zone information using atmospheric correction and mineral mapping.

[0022] Step 202 specifically includes: Step 202-1: Perform a quality assessment on the remote sensing data to obtain remote sensing data that meets the quality assessment requirements. The indicators used in the quality assessment include: spectral angle, root mean square error, peak signal-to-noise ratio, structural similarity, average structural similarity, and image homogenization error.

[0023] Step 202-2: Perform atmospheric correction on the remote sensing data that meets the quality assessment requirements to obtain corrected remote sensing data.

[0024] Step 202-3: Remove the borders from the corrected remote sensing data to obtain borderless remote sensing data.

[0025] Step 202-3 specifically includes: Step 202-3-1: Determine any band as the current band.

[0026] Step 202-3-2: Determine the binary image corresponding to the current band: For each point in the target area, assign a value of 1 if there is corrected remote sensing data for the current band, and assign a value of 0 if there is no corrected remote sensing data for the current band, thus obtaining the binary image corresponding to the current band.

[0027] Step 202-3-3: Update the current band and return to step 202-3-2 until all bands have been traversed, obtaining the binary image corresponding to each band.

[0028] Step 202-3-4: Multiply the binary images corresponding to all bands to obtain the debounded binary image.

[0029] Step 202-3-5: Use the debounded binary image to perform border removal processing on the corrected remote sensing data to obtain border-removed remote sensing data.

[0030] Step 202-4: Use interference removal methods to process the borderless remote sensing data to obtain interference-free remote sensing data. Interference removal methods include: ratio method, segmentation method, Q-value method, and spectral angle method.

[0031] Step 202-5: Conduct mineral mapping based on interference-free remote sensing data.

[0032] Step 202-6: Based on the comprehensive overlay analysis of mineral mapping information, determine the alteration zone information through the combination of alteration minerals.

[0033] Step 202-6 specifically includes: Step 202-6-1: Determine whether there is zoning based on mineral mapping.

[0034] Step 202-6-2: When there are zonation, select k minerals as the inner zone alteration mineral assemblage and select h minerals as the outer zone alteration mineral assemblage.

[0035] Step 202-6-3: When there is no zoning, select p minerals as the alteration mineral assemblage. The values ​​of k, h, and p are determined based on the typical deposit type; for example, porphyry copper deposits are divided into three zones: the inner zone is a potassium silicification zone, with alteration mineral assemblage consisting of potassium feldspar and silicification; the middle zone is a sericite alteration zone, with assemblage consisting of sericite, high-grade argillaceous material, etc.; and the outer zone is a phyllotitic alteration zone, with alteration mineral assemblage consisting of chlorite, carbonates, and epidote, etc.

[0036] Step 202-6-4: Use alteration mineral assemblages as alteration zone information. Alteration zone information is directly obtained from alteration mineral assemblages. For example, if the assemblages are chlorite and carbonate rocks, then the alteration zone is a chlorite-carbonate rock alteration zone; if it is potassic alteration or silicification, then the alteration zone is a potassic silicification zone.

[0037] Data Acquisition: ① Preliminary assessment of remote sensing data. Acquiring remote sensing data requires a comprehensive evaluation of the data's date and quality. Generally, the first step is to determine whether the remote sensing data was acquired in autumn or winter, whether there are clouds or snow, and the vegetation cover, etc. Autumn / winter data with minimal cloud cover, snow cover, and vegetation cover should be selected.

[0038] ② Evaluation of remote sensing data. After data selection, data quality needs to be evaluated. This application defines six indicators (SAM, RMSE, PSNR, SSIM, MSSIM, ERGAS) for evaluation.

[0039] Hyperspectral remote sensing data AA and BB are obtained by taking the remote sensing spectral vectors A and B of the same point on the image and defining the spectral angle SAM between them.

[0040] .

[0041] Estimate the parameters of AA from two remote sensing datasets, AA and BB. The root mean square error relative to the estimated BB parameter θ is defined as: .

[0042] Peak signal-to-noise ratio (PSNR) is defined by mean square error (MSE) for two images. The mean square error of the remote sensing data I and K is defined as follows: .

[0043] Then the peak signal-to-noise ratio is defined as .

[0044] Among them, MAX I This represents the point with the largest DN value in the image. If B represents the number of bits in the computer, it is generally 1. .

[0045] Structural similarity (SSIM) is used to measure the degree of similarity between two remote sensing images x and y, and is defined as follows: .

[0046] Where l(x,y) compares the brightness of x and y, c(x,y) compares the contrast of x and y, and s(x,y) compares the structure of x and y. , which are parameters for adjusting the relative importance of l(x,y), c(x,y) and s(x,y).

[0047] MSSIM compares the average structural similarity between two hyperspectral images (x, y). This is achieved by calculating the SSIM values ​​for each band and taking the average.

[0048] .

[0049] Image normalization error (ERGAS) is a metric used to evaluate the quality of remote sensing images. It generally reflects the performance of image processing and compression algorithms, as well as the overall image quality, and is defined as follows: .

[0050] Where n represents the number of bands in the image, l represents the grayscale of the image, and MSE i RMSE represents the mean square error of the i-th band. i,j y represents the root mean square error of the j-th pixel in the i-th band. i This represents the average brightness of the i-th band.

[0051] For two images representing the same region, one is a baseline image and the other is an evaluation image. Generally, the smaller the SAM (Significant Amount), the better. The smaller the PSNR value, the larger the SSIM value, the larger the MSSIM value, and the smaller the ERGAS value, the better the image quality.

[0052] Remote sensing alteration zone information extraction Remote sensing alteration zone extraction mainly involves analyzing alteration zones through satellite remote sensing atmospheric correction and mineral mapping.

[0053] ①Atmospheric correction.

[0054] Directly acquiring remote sensing images results in geometric distortions, sensor gain, and offset parameters. Preprocessing is required to obtain a planetary reflectance image with coordinate information. The specific formula is as follows: .

[0055] Where: R is planetary reflectivity; pi = 3.14; d represents the Earth-Sun distance on the day of the image; A is the solar altitude angle; Esun is the solar spectral irradiance at the corresponding wavelength outside the atmosphere; L is the radiance, which can be calculated using the following formula: .

[0056] Where: gain is the gain, and bias is the bias.

[0057] For thermal infrared data, temperature and emissivity need to be separated. The formula is as follows: .

[0058] .

[0059] Where T is temperature, λ is wavelength, ε is specific emissivity, and c1 and c2 are constants, with c1 = 3.74818 × 10⁻⁶. -4 Wμm 2 c2 = 1.43878 × 104 Kμm. R is the spectral radiance, which can be calculated using the following formula: .

[0060] QCAL is the actual radiometric value of the data, LMINλ is the spectral radiometric value when QCAL=0, LMAXλ is the spectral radiometric value when QCAL=QCALMAX, and QCALMAX is the image radiometric value of the data. The unit of R is W / (m²). 2 ×sr×μm).

[0061] ② Data preprocessing.

[0062] 1. Remove borders.

[0063] like Figure 3 The coordinates (X, Y) of the three bands on the plane (representing the Earth's surface) do not coincide. This mainly reflects the non-overlapping information at the boundaries. Therefore, the property of intersection is used to extract the information from the overlapping middle part. Figure 3 Medium gray area.

[0064] Boundary information refers to the fact that the data acquired for each band during remote sensing data acquisition is different, such as... Figure 4 In part A of the algorithm, if the studied region contains boundary information, this boundary information needs to be removed so that each band contains information. The method used is to determine whether each band contains information; if it does, a value of 1 is assigned; otherwise, a value of 0 is assigned, generating a binary image. Finally, the binary images of each band are multiplied to form a new binary image. Finally, each band is multiplied by the new binary image, thus removing the boundary information. The specific formula is as follows: . Where n refers to the total number of remote sensing image bands used, i=1,…,n, x i and y i These refer to the values ​​of the i-band before and after band removal, respectively. (Next) Figure 4 Part B in the image is the image after removing boundary information.

[0065] 2. Remove interference.

[0066] This invention aims to avoid nine common types of disturbance anomalies, including clouds, water bodies, shaded areas, white mud, ice and snow, vegetation, wetlands, dry riverbeds, and alluvial fans. Disturbance detection employs visual estimation; generally, disturbing features can be identified by their distinct characteristics in TM / ETM 743 color composite images or ASTER 631 color composite images, such as the white color of clouds. Removal methods include ratio method, high-end or low-end cutting method, Q-value method, and spectral angle method.

[0067] (1) High-end or low-end cutting method.

[0068] The main approach utilizes the characteristic high reflectance or strong absorption of interfering features in a specific band of remote sensing images. Specifically, interfering features exhibit high or low values ​​in a particular band. For instance, water bodies show low values ​​in TM / ETM band 7 and are processed using a low-end cropping method, while clouds show high values ​​in TM / ETM band 1 and are processed using a high-end cropping method. Similarly, white clay areas show high values ​​in TM / ETM band 3 and are processed using a high-end cropping method. The formula is as follows: .

[0069] Where i = 0, ..., n, n refers to the total number of remote sensing image bands used, x i and y i These refer to the band values ​​of band i before and after removing interference information, respectively, where b∈[1,…,n], C b It is a constant, x b This is the value corresponding to the original b-band. The purpose of this formula is to retain images that are greater than or less than a certain value, given a constraint, while all others are assigned a value of zero.

[0070] (2) Ratio method.

[0071] The ratio method is commonly used to remove various types of interference, such as shadows, water bodies, ice and snow, and white clay. First, the spectral characteristics of each band of the interfering feature are determined, and a threshold is set for removal. The formula is as follows: .

[0072] Where i = 0, ..., n, n refers to the total number of remote sensing image bands used, x i and y i These refer to the band values ​​of band i before and after removing "sharp" information, respectively, where a∈[1,…,n], C a It is a constant, x a x b These are the original values ​​corresponding to bands a and b. The purpose of this formula is to retain images where the value is greater than or less than a certain constraint, while all others are assigned a value of zero.

[0073] (3) Q-value method.

[0074] This primarily addresses disturbances from snow-covered or lakeside wetlands, dry riverbeds, alluvial areas, and thin clouds. The Q value is defined as follows: .

[0075] Where, x a x b x cFor the bands involved in principal component analysis, k a k b k c For x participating in the principal component changes a x b x c The value of the corresponding eigenvector.

[0076] ③ Mineral mapping.

[0077] Hyperspectral data can be used for mineral mapping. Mineral mapping methods include spectral angle mapping (SAM), binary encoding (BE), spectral information divergence (SID), linear band prediction (LBP), constrained energy minimization (CEM), adaptive coherence estimator (ACE), orthogonal subspace projection (OSP), continuous removal (CR), spectral feature fitting, and multi-range spectral feature fitting (MRSFF). Various methods have readily available software implementations. For example, the algorithm for spectral angle mapping (SAM) is as follows: 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 as follows: .

[0078] In the formula, (α,β) is the inner product of n-dimensional vectors α and β, defined as (α,β) = α₁β₁ + α₂β₂ + ... + α n β n When α and β are column vectors, (α,β) = α'β = β'α.

[0079] ; .

[0080] .

[0081] Where |α| and |β| are the lengths of vectors α and β, respectively.

[0082] |α|= = = By finding the inner product and length of α and β, we can calculate cos, and then find the included angle from a table.

[0083] ④ Alteration zone analysis.

[0084] Alteration zone analysis, based on mineral mapping, involves comprehensively analyzing the information from altered minerals through superposition. The alteration zones obtained from these superimposed altered minerals are analyzed by combining different altered minerals. These alteration zones reflect the fluid characteristics of hydrothermal deposits, thus guiding mineral exploration. In practice, different alteration zones are determined by combining several altered minerals, and these alteration zones are related to the type of deposit. For example, in porphyry copper deposits, the sericitization of the inner zone and the propylitization of the outer zone can be used for segmentation; gold deposits can be identified by pyrite-sericite alteration.

[0085] In general, first check for zonation. If zonation exists, select k minerals m1, m2, ..., m k As an inner alteration mineral assemblage, h minerals m1, m2, ..., m are selected. h As an altered mineral assemblage in the outer zone; if there is no zoning, select p minerals m1, m2, ..., m p As an alteration mineral assemblage, the formula is as follows: .

[0086] Step 203: Rasterize the mineral survey data to obtain a mineral survey raster image.

[0087] Step 203 specifically includes: Step 203-1: List the coordinates and resource values ​​of the mineral survey data to form a vector layer.

[0088] Step 203-2: Interpolate the vector points in the vector layer to obtain the mineral survey raster image: For vector points in the vector layer with a range greater than the range threshold, use the multifractal method for interpolation. For vector points in the vector layer with a range less than or equal to the range threshold, use the multi-kriging method for interpolation.

[0089] Mineral survey data analysis: ①Gridded mineral survey data.

[0090] Mineral survey data such as Figure 5 The mineral survey data mainly includes mining area data, exploration right data, mining right data, minimum prediction area data, and mineral deposit data. The primary data required are coordinates and resource quantity data, which need to be rasterized to facilitate integration with remote sensing alteration zone data.

[0091] Transform the mineral survey data according to the coordinates, that is, rasterize the data. According to the coordinates, interpolation of the data is required. Note that the interpolation methods include multifractal and Kriging methods: 1) List the coordinates and resource quantity values of the mineral survey data to form a vector layer, which is a point vector .

[0092] is the coordinate value of the i-th row and j-th column of the image, is the calculated temperature value of the i-th row and j-th column of the image. 1 < i < m, 1 < j < n, where m and n are the maximum values of the number of rows and columns of the image.

[0093] 2) Calculate the vector points according to the range.

[0094] If > , use the multifractal method. If < , use the universal Kriging method.

[0095] a. Multifractal.

[0096] Regard each row (or column) of the temperature image as a spatial sequence (assuming m rows and n columns), and find the ranges and standard deviations of each sequence, and establish , data pairs. If the scatter points are near a straight line, the slope of the straight line is the Hurst exponent , which is also the fractal dimension.

[0097] (1) First, calculate the dimension of each row (column), and the steps are as follows: The range of each row (row number i): .

[0098] is the temperature value of the j-th data in the i-th row, and n is the maximum number of columns.

[0099] The standard deviation of each row (row number i): .

[0100] is the temperature value of the j-th data in the i-th row, and n is the maximum number of columns.

[0101] Establish the scatter relationship between the range and the standard deviation: .

[0102] Among them, For row fractal dimension, Let be the temperature value of the j-th data in the i-th row, and n be the maximum column number.

[0103] Similarly, the column fractal dimension can be calculated. .

[0104] (2) Establish a fractal matrix with the following elements: .

[0105] (3) Calculate the total row dimension A: .

[0106] in, .

[0107] .

[0108] Where i is the row number.

[0109] Similarly, the total column dimension B can be calculated. (4) Calculate the fractal vector distance for each unit: .

[0110] For each point Calculate the distance ,use Interpolate point data.

[0111] b. Krieger processing method.

[0112] The Kriging method is a way to make unbiased optimal estimates of regionalized variables over a finite region.

[0113] Based on the temperature value graph T calculated earlier, calculate the temperature value at a certain point. The n nearest temperature values ​​are represented as .

[0114] (1) Calculate the expectation and covariance.

[0115] Mathematical expectation vector : .

[0116] Calculate the covariance c(x): .

[0117] in, for The transpose of .

[0118] (2) Calculate unbiasedness: .

[0119] That is to say: .

[0120] in, The value to be determined.

[0121] (3) Calculate the optimality.

[0122] Under the condition of unbiasedness, the variance is: .

[0123] Construct the Lagrange equation: .

[0124] in, and For the value to be found, calculate with respect to F. and Take the derivative of the equation and set it to 0. This gives us the Kriging equations. (4) Calculate the Kriging equation: .

[0125] (5) Perform interpolation.

[0126] Kriging interpolation methods in the field of computation: .

[0127] For each point's domain Interpolation is performed on all of them to obtain a raster image of the mineral survey.

[0128] Step 204: Register and fuse the alteration zone information and the mineral survey raster image to obtain fused data.

[0129] Step 204 specifically includes: Step 204-1: Based on the alteration zone information, determine the mineral assemblage that each alteration zone passes through.

[0130] Step 204-2: Determine any alteration zone as the current alteration zone.

[0131] Step 204-3: After performing minimum noise separation transform on the images of multiple bands corresponding to the current alteration zone, a multidimensional image of the current alteration zone is obtained. The multidimensional image includes multiple single-channel images. Each channel of the multidimensional image corresponds one-to-one with the mineral assemblages traversed by the alteration zone. Each band corresponds one-to-one with the mineral assemblages traversed by the alteration zone.

[0132] Step 204-4: Determine any channel as the current channel.

[0133] Step 204-5: Determine the structural similarity coefficient between the mineral survey raster image and the single-channel image of the current channel.

[0134] Step 204-6: If the structural similarity coefficient is greater than the structural similarity coefficient threshold, replace the single-channel image of the current channel in the mineral survey raster image with the mineral survey raster image. The replacement involves replacing the entire mineral survey raster image with the single-channel image. For example, if there are 5 channels and channel 2 is the mineral survey channel, then that channel will be replaced.

[0135] Step 204-7: Update the current channel and return to step 204-5 until all channels are traversed to obtain the fused data of the current alteration zone.

[0136] Step 204-8: Update the current alteration zone and return to step 204-3 to obtain the fused data for each alteration zone.

[0137] Data fusion: ① Data registration and analysis.

[0138] The raster image generated from mineral survey data processing is matched with the remote sensing alteration zone raster (matrix). The main focus here is on comparing structural similarity; therefore, the image's brightness, contrast, and structure must be considered. Let's assume the alteration zone after p mineral assemblages is... Similarly, the rasterized image of a mineral survey is X. Then, the structural similarity coefficient... .

[0139] The SSIM value is [0, 1], and the larger the value, the higher the image similarity.

[0140] In the formula, and It is image X and image The mean, and It is image X and image variance It is image X and image The covariance.

[0141] .

[0142] .

[0143] .

[0144] .

[0145] Where m and n are the number of rows and columns of the image. It is a constant, usually taken as k1 can take the value 0.01, K2 can take the value 0.03, and L can take the value 255.

[0146] ②Merge mineral survey data with alteration zone data.

[0147] An image similar to the rasterized national conditions data X was obtained using a matching device. We use similarity to... Perform the substitution operation with X. Suppose that the multidimensional image Y with p channels obtained after undergoing MNF (Minimum NoiseFraction Rotation) transformation on p-band images is represented by the following formula: .

[0148] Using X to replace the image The replaced multidimensional image is: .

[0149] Step 205: Generate a vector cloud space by mapping the metallogenic belt and fused data according to coordinates, and perform metallogenic belt constraint analysis based on the vector cloud space.

[0150] Step 205 specifically includes: Step 205-1: Generate corresponding data based on coordinates from the ore-forming belt and the fused data to form a vector cloud space.

[0151] Step 205-2: Construct a vector co-occurrence matrix based on the vector cloud space: Take a vector in the vector cloud space and another vector offset along a certain direction and by an offset distance Forming a matrix If the calculation involves movement within a spatial vector cloud, multiple matrices are obtained. Statistical matrix The number of opportunities is the first opportunity number. Dividing the first opportunity number by the total number of opportunities yields the first opportunity probability density, forming a vector co-occurrence matrix. The total number of opportunities refers to the number of moves.

[0152] Step 205-3: Calculate the co-occurrence constraints based on the vector co-occurrence matrix. The co-occurrence constraints include: entropy, inertia, and energy.

[0153] Step 205-4: Construct a vector difference matrix based on the vector cloud space: Take a vector in the vector cloud space and another vector offset along a certain direction and by an offset distance The spatial angles are calculated. If the calculation involves movement within a spatial vector cloud, multiple spatial angles are obtained. The number of outflows of these spatial angles is counted as the second outflow count. The second outflow count is divided by the total number of outflows to obtain the second outflow probability density, forming a vector difference matrix. The number of rows in the vector difference matrix is ​​equal to the number of rows in the vector difference matrix. The number of columns in the vector difference matrix is ​​equal to the number of columns in the vector difference matrix.

[0154] Step 205-5: Calculate the difference constraint indices based on the vector difference matrix. The difference constraint indices include: curvature, tensile strength, and enthalpy.

[0155] Step 205-6: Conduct metallogenic belt constraint analysis based on symbiotic constraint index and differential constraint index.

[0156] Wherein, the entropy is: .

[0157] in, .

[0158] In the formula, Entropy. This represents the total number of rows in the vector co-occurrence matrix. This represents the total number of columns in the vector co-occurrence matrix. Let be the element in the i-th row and j-th column of the vector co-occurrence matrix. Let be the i-th vector in the vector cloud space. Let be the j-th vector in the vector cloud space. For matrix The first time he made a name for himself. This represents the total number of times.

[0159] Inertia is: .

[0160] In the formula, It is due to inertia.

[0161] Energy is: .

[0162] In the formula, It is energy.

[0163] The curvature is: .

[0164] .

[0165] In the formula, For curvature. Let be the element in the i-th row and j-th column of the vector difference matrix. Vector before and after movement The spatial angle. Vector before and after movement The spatial angle. This is the second time.

[0166] Zhang Duwei: .

[0167] In the formula, For Zhang Du.

[0168] Enthalpy is: .

[0169] In the formula, It is enthalpy.

[0170] The higher the symbiotic index, the better; the lower the differential index, the better. The formula for calculating the total index of metallogenic belt constraint division is as follows: .

[0171] Among them, ZZ is the overall indicator. Here, H is the calculation factor, I is the entropy, E is the energy, Q is the curvature, Z is the tension, and S is the enthalpy.

[0172] Metallogenic belt constraints: ① The symbiotic matrix of metallogenic belts and fused data.

[0173] The ore-forming belt and the fused data are used to generate corresponding data according to coordinates, forming a vector cloud space. A vector from the vector cloud space is then selected. and another vector offset along a certain direction and by an offset distance Obtain the matrix If calculating movement within a spatial vector cloud, various matrices are calculated. Statistical matrix The number of successful occurrences divided by the total number of occurrences yields the success probability density, forming a vector co-occurrence matrix. .

[0174] ② Calculation of constraint indicators.

[0175] Entropy, inertia, and energy are calculated by computed co-occurrence matrices. A vector in the vector cloud space is taken. and another vector offset along a certain direction and by an offset distance Calculate the included angle in space. If calculating movement within a spatial vector cloud, calculate various included angles and statistically analyze the included angles. The probability density function of a success is obtained by dividing the number of successes by the total number of successes, forming a vector difference matrix. Curvature, tensile strength, and enthalpy are calculated by using a difference matrix. ③ Analysis of constraints on metallogenic belts By analyzing entropy, inertia, energy, curvature, tension, and enthalpy, it is generally believed that smaller entropy and enthalpy, larger inertia and energy, and smaller curvature and larger tension indicate stronger constraints on ore-forming zones.

[0176] Step 206: Based on the results of the metallogenic belt constraint analysis, determine the potential mineral resource prediction area.

[0177] Step 206 specifically includes: Step 206-1: When both the symbiotic constraint index and the differential constraint index are within the preset range, the ore-forming zone is determined to be a forced ore-forming zone.

[0178] Step 206-2: Obtain all alteration zones in the forced ore zone as undetermined alteration zones.

[0179] Step 206-3: Delete known alteration zones from the undetermined alteration zones to obtain multiple prediction areas. The known alteration zones are from mineral survey data.

[0180] Step 206-4: Define a random variable as the dependent variable and construct a multiple linear regression model with multiple prediction regions as independent variables.

[0181] Step 206-5: Construct a set of equations for a multiple linear regression model using multiple sets of observation data, and solve the set of equations for the multiple linear regression model using the least squares estimation method to obtain the coefficients of the set of equations for the multiple linear regression model.

[0182] Step 206-6: Determine any one of the independent variables as the current independent variable.

[0183] Step 206-7: Use analysis of deviations to perform hypothesis testing on the multiple linear regression model before and after deleting the current independent variable.

[0184] Step 206-8: Determine the discriminant by the absolute value of the difference between the coefficients of determination before and after removing the current independent variable from the multiple linear regression model. The coefficients of determination are: .in, is the coefficient of determination for the multiple linear regression model. SS is the sum of squares of the regressions for the multiple linear regression model. represents the total variance of the multiple linear regression model. MS represents the sum of squared residuals.

[0185] Step 206-9: When the discriminant is less than the discriminant threshold, delete the prediction region corresponding to the current independent variable.

[0186] Step 206-10: Update the current independent variable and return to step 206-7 until all independent variables have been traversed to obtain the potential mineral resource prediction area.

[0187] Delineate the prediction area: ① Delineate the prediction area.

[0188] Based on the above analysis, the prediction area is delineated by alteration zones. The same metallogenic belt has the same alteration zone, and the alteration zone of the known deposit is consistent with the metallogenic belt. Therefore, the alteration zone without known deposits is delineated as the prediction area.

[0189] ② Optimize the analysis and prediction area.

[0190] Define y as a random variable (unknown alteration zone data X), x1, x2, ..., x n For n independent variables (corresponding to the changed data) A total of m observations were conducted. First, it was assumed that there is a linear relationship between y and the n independent variables: .

[0191] In the formula, a0, a1, a2, …, a n The regression coefficient is a constant, representing the regression coefficient x when other independent variables remain unchanged. j (j=1,2,…n) represents the average change when y increases or decreases by one unit, and ε is the random error after removing the influence of n independent variables on y. The above formula is called a multiple linear regression model.

[0192] First use To estimate the mean E(y) of y, assume ε follows a mean of 0 and a variance of σ. 2 The normal distribution is ε ~ N(0, σ). 2 Then y follows a pattern with mean E(y) and variance σ. 2 The normal distribution, i.e., y ~ N[E(y), σ 2 Then, for m groups of sample observation data: .

[0193] In the formula, x ij x represents j The i-th observation has the following formula: .

[0194] The above equation is the mathematical model for n-variable linear regression, where a0, a1, a2, …, a n Let there be n+1 undetermined parameters, and ε1, ε2, …, εm be m independent random variables that follow the same normal distribution. For simplicity, matrix form is used: .

[0195] , , .

[0196] The mathematical model for n-variable linear regression is: .

[0197] To perform least squares estimation using the formula, first assume b0, b1, b2, ..., b n Let a0, a1, a2, …, a be the n+1 regression coefficients. n The least squares estimate of the observed values ​​is given as follows: .

[0198] e j For the error ε j The estimated value is called the residual, assuming for The estimated value, then .

[0199] .

[0200] In the above formula, j = 1, 2, …, m. The residual e j Represents actual value Compared with the estimated value Degree of deviation. To make the estimated value... Compared with actual value For the best fit, the sum of squared residuals must be minimized. .

[0201] To reach the minimum, according to the principles of advanced mathematics, the extreme value is where the differential of the function is 0. Therefore, the equation is established. .

[0202] From the above formula, we obtain the normal equation: .

[0203] Given matrix X, the equations on both sides of the coefficients are expressed in terms of C and D. Then, .

[0204] .

[0205] .

[0206] Therefore, the matrix form of the normal equation is: .

[0207] Where B is an unknown vector, if the matrix coefficient C is of full rank, its inverse matrix exists, and the unknown vector B can be solved in reverse: .

[0208] Where vector B is the optimization parameter. ③ Precision controller.

[0209] Hypothesis testing and evaluation of regression equations generally employ analysis of deviations. Total variation is defined as follows: .

[0210] Where SS is the regression sum of squares, and is the regression value. with the mean The sum of squares of the differences reflects the fluctuations in Y caused by changes in the independent variable X, representing the degrees of freedom. (n is the number of independent variables). MS is the sum of squared residuals, which is the measured value. With regression value The sum of squares of the differences is caused by experimental error and other factors, and the degrees of freedom The total degrees of freedom of variation are m-1.

[0211] If the observed values ​​are given and the total variance is determined, the regression effect can be measured using SS and MS. The larger the SS, the more significant the regression effect; the larger the MS, the worse the regression effect.

[0212] To test the overall regression effect, a dimensionless index—the coefficient of determination R—was defined. 2 To indicate: R 2 This reflects the proportion of the regression deviation to the total variation. R = R0 1 / 2 This is called the multiple correlation coefficient, which reflects the degree of correlation between all independent variables and the dependent variable. R 2 The larger the R value, the better the regression effect.

[0213] The above-mentioned regression analysis cannot explain the effect of each independent variable x1, x2, ..., x. n While all independent variables are important to the dependent variable y, some may have no effect on the dependent variable, or their effect may be replaced by other independent variables. These latter independent variables need to be removed from the regression equation. It is recommended that each independent variable x... i Whether it is significant, assuming H0: a i =0, i=1,2, …n.

[0214] (1) F-value test.

[0215] In H0:a i Under the assumption of =0, .

[0216] For a given confidence level α, find the critical value F corresponding to β from the F-value distribution table. β If |F i |〉F βIf we reject the hypothesis H0, we assume that the overall regression effect is significant when n independent variables are not present; otherwise, the overall regression effect is not significant.

[0217] (2) t test.

[0218] In H0:a i Under the assumption of 0, the t-test formula is as follows: .

[0219] For a given significance level β, find the critical value t corresponding to β from the t-value distribution table. β If |t i |〉t β Reject hypothesis H0, and believe that a i Values ​​that are significantly different from 0 should not be removed; conversely, values ​​that are significantly different from 0 should be removed.

[0220] (3) p-value test.

[0221] Assume H0: a i =0, and follows a p-distribution statistic with 1 and mn-1 degrees of freedom: .

[0222] For a given significance level β, the critical value p can be found in the p-value distribution table. β (1, mn-1), if p i >p β (1, mn-1), reject the hypothesis H0, and believe that x i If it plays an important role in the y-value, it should not be removed; conversely, if it does not play an important role, it should be removed.

[0223] Step 207: Select the optimal band combination using the optimal index method; based on the optimal band combination, use coordinate layering to overlay the raster and vector to generate a schematic diagram of the potential mineral resource prediction area: the base map uses a false-color map of the band combination with the highest information entropy, and the vectors are represented by points, lines and surfaces with the same projection.

[0224] Base map generation: Selecting wave bands using the optimal index method: .

[0225] In the formula, Let be the standard deviation of the i-th band. Let be the correlation coefficients of bands i and j. A higher OIF indicates a greater amount of information contained; therefore, the band combination with the highest OIF is the optimal band combination, and the image synthesized using the optimal band combination is used as the base map. Raster and vector image synthesis: A false-color image with the highest information entropy band combination is used as the base map, and vectors are represented by points, lines, and surfaces with the same projection. The raster and vector images are then overlaid using coordinate layering to create an image suitable for human vision.

[0226] vector x and y are the corresponding coordinates, and z is the eigenvalue. Vector values, raster , and For the corresponding coordinates, Let be the raster grayscale value. Thus achieving a grid Grayscale values ​​and vectors The superposition of.

[0227] Image Output: Outputs a superimposed image suitable for human visual observation. The software outputs the final image in JPG or TIF format, such as... Figure 6 .

[0228] 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.

[0229] 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 rapidly delineating potential mineral resource prediction areas based on remote sensing alteration zones, characterized in that, include: Acquire remote sensing data and mineral survey data for the target area; Based on the remote sensing data, alteration zone information was extracted using atmospheric correction and mineral mapping. The mineral survey data is rasterized to obtain a mineral survey raster image. The alteration zone information and the mineral survey raster image are registered, analyzed, and fused to obtain fused data. The metallogenic belt and fused data are used to generate a vector cloud space according to coordinate correspondence, and the metallogenic belt constraint analysis is performed based on the vector cloud space; Based on the results of the metallogenic belt constraint analysis, potential mineral resource prediction areas were identified.

2. The method for rapid delineation of potential mineral resource prediction areas based on remote sensing alteration zones according to claim 1, characterized in that, Based on the aforementioned remote sensing data, alteration zone information is extracted using atmospheric correction and mineral mapping, specifically including: The remote sensing data is subjected to quality assessment to obtain remote sensing data that meets the quality assessment requirements. The indicators used in the quality assessment include: spectral angle, root mean square error, peak signal-to-noise ratio, structural similarity, average structural similarity, and image homogenization error. Atmospheric correction processing is performed on remote sensing data that meets the quality assessment requirements to obtain corrected remote sensing data; The corrected remote sensing data is subjected to border removal processing to obtain border-removed remote sensing data; The borderless remote sensing data is processed using interference removal methods to obtain interference-free remote sensing data; the interference removal methods include: ratio method, segmentation method, Q-value method and spectral angle method; Mineral mapping is carried out based on the aforementioned de-interference remote sensing data; Based on the comprehensive overlay analysis of the mineral mapping information, the alteration zone information is determined by the combination of alteration minerals.

3. The method for rapid delineation of potential mineral resource prediction areas based on remote sensing alteration zones according to claim 2, characterized in that, The corrected remote sensing data is subjected to border removal processing to obtain border-removed remote sensing data, specifically including: Determine any band as the current band; Determine the binary image corresponding to the current band: For each point in the target area, assign a value of 1 if there is corrected remote sensing data for the current band, and assign a value of 0 if there is no corrected remote sensing data for the current band, thus obtaining the binary image corresponding to the current band. Update the current band and return to the step "determine the binary image corresponding to the current band" until all bands are traversed to obtain the binary image corresponding to each band; Multiply the binary images corresponding to all bands to obtain a debounded binary image; The corrected remote sensing data is processed by removing the borders using a de-bordered binary image to obtain de-bordered remote sensing data.

4. The method for rapid delineation of potential mineral resource prediction areas based on remote sensing alteration zones according to claim 2, characterized in that, Based on the comprehensive overlay analysis of alteration mineral information using the aforementioned mineral mapping, alteration zone information is determined through alteration mineral assemblages, specifically including: Based on the mineral mapping, determine whether there are zonations; When there are zones, select k minerals as the inner zone alteration mineral assemblage and select h minerals as the outer zone alteration mineral assemblage; When there is no zoning, p minerals are selected as the alteration mineral assemblage; where the values ​​of k, h, and p are determined according to the deposit type of a typical deposit. Alteration mineral assemblage is used as information about alteration zones.

5. The method for rapid delineation of potential mineral resource prediction areas based on remote sensing alteration zones according to claim 2, characterized in that, The mineral survey data is rasterized to obtain a mineral survey raster image, specifically including: List the coordinates and resource values ​​of the mineral survey data to form a vector layer; Interpolation processing is performed on the vector points in the vector layer to obtain a mineral survey raster image: for vector points in the vector layer with a range greater than the range threshold, the multifractal method is used for interpolation processing; for vector points in the vector layer with a range less than or equal to the range threshold, the multi-kriging method is used for interpolation processing.

6. The method for rapid delineation of potential mineral resource prediction areas based on remote sensing alteration zones according to claim 2, characterized in that, The alteration zone information and the mineral survey raster image are registered, analyzed, and fused to obtain fused data, which specifically includes: Based on the alteration zone information, the mineral assemblage traversed by each alteration zone is determined; Identify any alteration zone as the current alteration zone; After performing minimum noise separation transformation on the images of multiple bands corresponding to the current alteration zone, a multidimensional image of the current alteration zone is obtained; the multidimensional image includes multiple single-channel images; the channels of the multidimensional image correspond one-to-one with the mineral combinations traversed by the alteration zone; and the multiple bands correspond one-to-one with the mineral combinations traversed by the alteration zone. Select any channel as the current channel; Determine the structural similarity coefficient between the mineral survey raster image and the single-channel image of the current channel; If the structural similarity coefficient is greater than the structural similarity coefficient threshold, the single-channel image of the current channel in the mineral survey raster image is replaced with the mineral survey raster image. Update the current channel and return to the step "Determine the structural similarity coefficient between the mineral survey raster image and the single-channel image of the current channel" until all channels are traversed to obtain the fused data of the current alteration zone; Update the current alteration zone and return to step "After performing minimum noise separation transformation on the images of multiple bands corresponding to the current alteration zone, obtain the multidimensional image of the current alteration zone", to obtain the fused data of each alteration zone.

7. The method for rapid delineation of potential mineral resource prediction areas based on remote sensing alteration zones according to claim 2, characterized in that, The metallogenic belt and fused data are mapped to a vector cloud space, and a metallogenic belt constraint analysis is performed based on the vector cloud space, specifically including: The metallogenic belt and fused data are used to generate corresponding data according to coordinates, forming a vector cloud space; Based on the aforementioned vector cloud space, construct a vector co-occurrence matrix: take a vector in the vector cloud space. and another vector offset along a certain direction and by an offset distance Forming a matrix If the calculation involves movement within a spatial vector cloud, then multiple matrices are obtained. Statistical matrix The number of times an opportunity arises is the first opportunity number. Dividing the first opportunity number by the total number of opportunities yields the first opportunity probability density, forming a vector co-occurrence matrix. Based on the vector co-occurrence matrix, co-occurrence constraint indices are calculated; the co-occurrence constraint indices include: entropy, inertia, and energy; Based on the aforementioned vector cloud space, construct a vector difference matrix: take a vector in the vector cloud space. and another vector offset along a certain direction and by an offset distance The spatial angles are calculated. If the calculation involves movement within a spatial vector cloud, the spatial angles of multiple vectors are obtained. The number of outflows of the spatial angles is counted as the second outflow count. The second outflow count is divided by the total number of outflows to obtain the second outflow probability density, forming a vector difference matrix. The number of rows in the vector difference matrix is ​​equal to the number of rows in the vector difference matrix. The number of columns in the vector difference matrix is ​​equal to the number of columns in the vector difference matrix. Based on the vector difference matrix, the difference constraint index is calculated; the difference constraint index includes: curvature, tensile strength, and enthalpy; Based on the symbiotic constraint index and the differential constraint index, an analysis of the constraints on metallogenic belts is conducted.

8. The method for rapid delineation of potential mineral resource prediction areas based on remote sensing alteration zones according to claim 7, characterized in that, The entropy is: ; in, ; In the formula, Entropy; This represents the total number of rows in the vector co-occurrence matrix; This represents the total number of columns in the vector co-occurrence matrix; Let be the element in the i-th row and j-th column of the vector co-occurrence matrix; Let i be the i-th vector in the vector cloud space; Let j be the j-th vector in the vector cloud space; For matrix The first number of times one can achieve success; Total number of times; The inertia is: ; In the formula, For inertia; The energy is: ; In the formula, For energy; The curvature is: ; ； In the formula, Curvature; Let be the element in the i-th row and j-th column of the vector difference matrix; Vector before and after movement The spatial angle; Vector before and after movement The spatial angle; This is the second highest number of times a student can earn a living. The tensor is: ; In the formula, For the degree of tension; The enthalpy is: ; In the formula, It is enthalpy.

9. The method for rapid delineation of potential mineral resource prediction areas based on remote sensing alteration zones according to claim 7, characterized in that, Based on the results of the metallogenic belt constraint analysis, potential mineral resource prediction areas were identified, specifically including: When both the symbiotic constraint index and the differential constraint index are within a preset range, the ore-forming zone is determined to be a forced ore-forming zone; All alteration zones within the forced ore zone are identified as undetermined alteration zones. By deleting known alteration zones from the undetermined alteration zones, multiple prediction areas are obtained; the known alteration zones are based on mineral survey data. A multiple linear regression model is constructed by defining a random variable as the dependent variable and multiple prediction regions as independent variables. A set of equations for a multiple linear regression model is constructed using multiple sets of observation data, and the coefficients of the multiple linear regression model are obtained by solving the set of equations using the least squares estimation method. Determine any one of the independent variables as the current independent variable; Analysis of deviations was used to test the hypotheses in a multiple linear regression model before and after removing the current independent variable. The absolute value of the difference between the coefficients of determination before and after removing the current independent variable in a multiple linear regression model is used as the discriminant; the coefficients of determination are: ;in, is the coefficient of determination for the multiple linear regression model; SS is the sum of squares of the regressions for the multiple linear regression model. The total variance of the multiple linear regression model is represented by MS, which is the sum of squared residuals. When the discriminant is less than the discriminant threshold, the prediction region corresponding to the current independent variable is deleted; Update the current independent variable and return to the step "Use the analysis of deviations to perform hypothesis testing on the multiple linear regression model before and after deleting the current independent variable" until all independent variables are traversed to obtain the potential mineral resource prediction area.

10. The method for rapid delineation of potential mineral resource prediction areas based on remote sensing alteration zones according to claim 1, characterized in that, After determining the potential mineral resource prediction area based on the results of the metallogenic belt constraint analysis, the following steps are also included: Selecting the optimal band combination using the optimal index method; Based on the optimal band combination, the grid and vector are superimposed using coordinate layering to generate a schematic diagram of the potential mineral resource prediction area: the base map is a false color map with the band combination with the highest information entropy, and the vectors are represented by points, lines and surfaces with the same projection.

Images