A method and system for extracting copper, lead and zinc mineralization alteration information based on multispectral remote sensing

Abstract

The application belongs to the technical field of remote sensing application, and specifically discloses a copper-lead-zinc mineralization alteration information extraction method and system based on multispectral remote sensing. The method receives multispectral remote sensing image data and generates surface reflectivity data; the aforementioned data is processed based on a spectral feature library to identify target mineral combinations and quantitatively delineate spatial distribution of alteration zoning; vegetation coverage areas are eliminated and background interference is suppressed to generate an alteration intensity grading map; a visual map is generated based on the grading map; a comprehensive report is generated and suggestions for anomaly grade evaluation and ore prospecting target area delineation are proposed; multispectral remote sensing image data, surface reflectivity data and the alteration intensity grading map are archived; and an interface is used to realize interaction with a remote sensing platform and GIS software. The system comprises remote sensing data acquisition, spectral preprocessing, alteration mineral identification, information extraction, result output, data storage and management and system integration modules. The application has the characteristics of high efficiency and precision, clear extracted information and strong practicality.

Description

Technical Field

[0001] This invention belongs to the field of remote sensing application technology, specifically relating to a method and system for extracting copper-lead-zinc mineralization alteration information based on multispectral remote sensing, which is highly efficient, accurate, provides clear information, and is highly practical. Background Technology

[0002] Copper-lead-zinc mineralization and alteration are important geological phenomena in the formation of metallic ore deposits, mainly involving the interaction between the ore body and the surrounding rocks, as well as the characteristics of mineralization zoning. Copper-lead-zinc mineralization and alteration are usually associated with magmatic intrusion. During the ascent of magma, gas and fluid in the magma interact with the surrounding rocks, leading to changes in temperature and composition, and forming alteration zoning. This process can be divided into copper mineralization and lead-zinc mineralization stages, with the mineralization zoning sequentially as follows: copper mineralization zone → copper-lead-zinc mineralization zone → lead-zinc mineralization zone → lead (silver) mineralization zone.

[0003] Methods and systems for extracting alteration information from copper-lead-zinc mineralization mainly rely on traditional geological mapping and geochemical sampling analysis, supplemented by early remote sensing techniques. These methods typically suffer from the following limitations: First, conventional remote sensing alteration information extraction techniques based on principal component analysis (PCA) or band ratio methods struggle to effectively distinguish the spectral characteristics of alterations in different metal mineralizations such as copper, lead, and zinc, leading to information confusion. Second, traditional methods exhibit significantly reduced information extraction capabilities in vegetated areas, Quaternary-covered areas, or areas with complex topography, resulting in severe background interference. Third, existing methods often lack detailed spectral modeling for copper-lead-zinc mineralization alteration mineral assemblages (such as chlorite, epidote, sericite, limonite, and kaolinite) when processing multispectral data, resulting in limited accuracy of extracted alteration information. Finally, insufficient automation in the workflow leads to low information extraction efficiency, making it difficult to meet the needs of large-scale rapid exploration.

[0004] Among existing technologies, multispectral remote sensing is the most commonly used technique for regional-scale alteration extraction because it covers the visible-near-infrared (VNIR) and short-wave infrared (SWIR) spectral bands and balances coverage and spectral resolution. One such technique utilizes the difference in reflectance between the "reflection peak" and "absorption valley" spectral bands of altered minerals, calculating the ratio of "reflection peak spectral band / absorption valley spectral band" to amplify alteration information. This method offers advantages such as requiring no complex parameters, rapidly processing large-scale data, effectively suppressing interference from uniform backgrounds like topographic shadows and atmospheric scattering, and providing clear physical meaning, with results directly correlated to mineral spectral characteristics and easily interpretable. However, because it only uses information from two spectral bands, it is susceptible to contamination by non-altered features and cannot quantify alteration intensity, only determining "presence" and not distinguishing between "weak alteration" and "strong alteration." Therefore, it is only suitable for regional surveys in arid to semi-arid areas. Another approach involves compressing high-dimensional correlated multispectral data into low-dimensional principal components (PCs) through linear transformation. In this approach, "abnormal principal components" concentrate alteration information, while background information is dispersed across other PCs. A threshold segmentation method, Principal Component Analysis (PCA), is then used to extract anomalous areas. This method utilizes all spectral bands and reduces data redundancy, exhibiting higher sensitivity to minor alterations than the ratio method. It can simultaneously extract multiple alteration types, enabling multi-target extraction through different PC combinations. Furthermore, it is well-suited for areas with low to medium vegetation cover when combined with vegetation-suppressed PCs. However, it is susceptible to noise (sensor errors, cloud shadows), and anomalous PCs may contain non-alteration interference. It also cannot distinguish between mineralized and non-mineralized alterations. Therefore, it is only suitable for detailed exploration in areas with medium to high vegetation cover and areas where multiple alteration types coexist. Another approach is spectral angle mapping (SAM), which calculates the angle between the spectral vector of a multispectral pixel and the "standard alteration mineral spectral vector." A smaller angle indicates higher similarity, and pixels with an angle less than a threshold are identified as target alteration pixels. Based on "spectral shape similarity" matching, it is insensitive to changes in absolute reflectance caused by light intensity and terrain slope; it also has strong target targeting, allowing for the extraction of specific minerals; and the results can be quantified for similarity (angle values), facilitating the differentiation of alteration intensity. However, due to the limited number of multispectral bands (<20), it is difficult to capture subtle spectral differences; moreover, the computational load is large, making it unsuitable for ultra-large-scale regional processing. Therefore, it is only applicable to detailed surveys of mining areas with known alteration mineral types (1:50,000 scale). Summary of the Invention

[0005] To address the problems mentioned in the background section, this invention provides a method for extracting copper-lead-zinc mineralization alteration information based on multispectral remote sensing, which is highly efficient, accurate, provides clear information, and is highly practical. It also provides a system for extracting copper-lead-zinc mineralization alteration information based on multispectral remote sensing.

[0006] The method for extracting copper-lead-zinc mineralization alteration information based on multispectral remote sensing in this invention is implemented as follows: it includes remote sensing data acquisition, spectral preprocessing, alteration mineral identification, information extraction, result output, data storage and management, and system integration. The specific steps are as follows: A. Remote sensing data acquisition: Receive raw multispectral remote sensing image data of the target area from satellite and / or airborne sensors; B. Spectral preprocessing: Radiometric calibration, atmospheric correction and noise suppression are performed on the received raw multispectral remote sensing image data to generate surface reflectance data; C. Alteration mineral identification: Based on a pre-built database of copper, lead, and zinc mineralization alteration spectral features, the aforementioned surface reflectance data is processed using a spectral angle mapping algorithm and a hybrid modulation matched filtering algorithm, and the target mineral assemblage in the data is identified and the spatial distribution of alteration zones is quantitatively delineated. D. Information extraction: Eliminate spectral confusion of vegetation cover areas in the data processed in step C, then suppress background interference from complex terrain and loose sediments, and finally extract mineralization alteration anomaly boundaries to generate an alteration intensity grading map. E. Output Results: Based on the alteration intensity grading map, a visual map is generated by integrating alteration zone spatial distribution information, mining area geological map, and mineral point data using a GIS spatial analysis engine. Then, spatial overlay analysis technology and GIS attribute table association function are used to automatically achieve dynamic matching between mineral point data and alteration intensity grading, and a comprehensive report is generated. Based on the comprehensive report, the final mineralization and alteration information product is generated, including an alteration anomaly distribution map, intensity grading report, and mineral point matching results. F. Data storage and management: Distributed archiving of raw multispectral remote sensing image data, surface reflectance data, and alteration intensity classification maps, and implementation of version control and incremental updates; G. System Integration: Utilize system integration interfaces to achieve data interaction with remote sensing platforms and GIS software.

[0007] Furthermore, in step B, the radiation calibration uses the official calibration coefficients of the sensor, the atmospheric correction is based on the radiative transfer model or the empirical linear method to eliminate the effects of scattering and absorption, and the noise suppression is achieved through an adaptive filtering algorithm.

[0008] Furthermore, in step C, the pre-set spectral feature library of copper, lead, and zinc mineralization alteration is integrated from laboratory measured spectra and standard ground cover spectra of typical mining areas; the spectral angle mapping algorithm is used to calculate the corresponding spectral angle θ in the aforementioned surface reflectance data, and compare it with the set spectral angle θ threshold to identify target mineral combinations including chlorite, epidote, and sericite; the hybrid modulation matched filtering algorithm is used to optimize the endmember abundance estimation based on the linear spectral mixing model, and quantitatively delineate the spatial distribution of alteration zonation including limonite and kaolinite zones in the surface reflectance data.

[0009] Furthermore, the formula for calculating the spectral angle θ in the spectral angle mapping algorithm is as follows:

[0010] In the formula: t i r is the value of the reference spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library. i Let θ be the value of the image spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library, and n be the number of bands in the spectrum; where a smaller spectral angle θ indicates a higher spectral matching degree. The hybrid modulation matched filtering algorithm is based on a linear spectral mixing model, which calculates the matched filter response and noise adjustment factor using the following formula: , , In the formula: c i Let be the endmember spectral coefficients of the image spectrum in the i-th band. g i Let be the image spectral vector in the i-th band. s i For the image spectrum at the i-th observation or signal, m i Let be the mean or expected value of the i-th variable in the image spectrum. Let be the variance of the image spectrum in the i-th variable; set dual thresholds based on the matched filter response value and noise adjustment factor to delineate the spatial distribution of alteration zonation, including limonite and kaolinite zones.

[0011] Furthermore, in step D, multi-scale segmentation and object-oriented analysis techniques are used to eliminate spectral obfuscation of vegetation cover areas in the data processed in step C; then, a terrain shadow correction model and a Quaternary cover masking algorithm are used sequentially to suppress background interference from complex terrain and loose sediments in the data after spectral obfuscation is eliminated; finally, the abnormal boundaries of mineralization alteration are automatically extracted from the data with suppressed background interference through adaptive threshold segmentation to generate an alteration intensity grading map. The steps for automatically extracting abnormal boundaries of mineralization alteration and generating an alteration intensity grading map are as follows: First, based on data that has suppressed background interference, the mineralization alteration anomaly index of each pixel is calculated; Secondly, an adaptive threshold segmentation algorithm is used to dynamically adjust the threshold based on the statistical characteristics within the local window, thereby segmenting out abnormal boundaries; the formula is: In the formula, m It is a local mean. k For local standard deviation, s This is an empirical coefficient; Finally, the boundary continuity is optimized through morphological post-processing to generate a high-precision alteration intensity grading map.

[0012] Furthermore, in step E, the visualized map includes mineral assemblage types, alteration intensity grading, and spatial distribution of alteration zones; in step F, a distributed architecture is used to structurally store the original multispectral remote sensing image data, surface reflectance data, and alteration intensity grading map; in step G, the OGC standard service and API gateway of the system integration interface are used to achieve seamless data exchange and function calls with remote sensing platforms and GIS software including ENVI, ArcGIS, and QGIS.

[0013] The copper-lead-zinc mineralization alteration information extraction system based on multispectral remote sensing of the present invention is implemented as follows: it includes a remote sensing data acquisition module, a spectral preprocessing module, an alteration mineral identification module, an information extraction module, a result output module, a data storage and management module, and a system integration module. The remote sensing data acquisition module is connected to the spectral preprocessing module and is used to receive raw multispectral remote sensing image data of the target area from satellite and / or airborne sensors. The spectral preprocessing module, connected to the alteration mineral identification module, is used to perform radiometric calibration, atmospheric correction, and noise suppression on the received raw multispectral remote sensing image data to generate high-quality surface reflectance data. The alteration mineral identification module is connected to the information extraction module. It is used to process the aforementioned surface reflectance data based on a pre-set copper, lead, and zinc mineralization alteration spectral feature library, using a spectral angle mapping algorithm and a hybrid modulation matched filtering algorithm, and to identify the target mineral combination in the data and quantitatively delineate the spatial distribution of alteration zones. The information extraction module is connected to the result output module. It is used to eliminate spectral confusion of vegetation cover area in the data output by the alteration mineral identification module, then suppress background interference from complex terrain and loose sediments, and finally extract mineralization alteration anomaly boundaries to generate an alteration intensity grading map. The result output module is connected to the data storage and management module. It is used to generate a visual map based on the alteration intensity grading map, by integrating alteration zone spatial distribution information, mining area geological map and mineral point data based on the GIS spatial analysis engine. Then, it uses spatial overlay analysis technology and GIS attribute table association function to automatically realize dynamic matching between mineral point data and alteration intensity grading, and generate a comprehensive report. Based on the comprehensive report, it generates the final mineralization and alteration information product, which includes an alteration anomaly distribution map, intensity grading report and mineral point matching results. The data storage and management module is connected to the system integration module and is used to perform distributed archiving of raw multispectral remote sensing image data, surface reflectance data and alteration intensity grading maps, and to perform version control and incremental updates. The system integration module is used to achieve data interaction with the remote sensing platform and GIS software through the system integration interface.

[0014] Furthermore, the pre-set copper, lead, and zinc mineralization alteration spectral feature library in the alteration mineral identification module is integrated from laboratory measured spectra and standard ground cover spectra of typical mining areas; the spectral angle mapping algorithm is used to calculate the corresponding spectral angle θ in the aforementioned surface reflectance data, and compare it with the set spectral angle θ threshold to identify target mineral combinations including chlorite, epidote, and sericite; the hybrid modulation matched filtering algorithm is used to optimize the endmember abundance estimation based on the linear spectral mixing model, and quantitatively delineate the spatial distribution of alteration zonation including limonite and kaolinite zones in the surface reflectance data; The formula for calculating the spectral angle θ in the spectral angle mapping algorithm is as follows:

[0015] In the formula: t i r is the value of the reference spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library. i Let θ be the value of the image spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library, and n be the number of bands in the spectrum; where a smaller spectral angle θ indicates a higher spectral matching degree. The hybrid modulation matched filtering algorithm is based on a linear spectral mixing model, which calculates the matched filter response and noise adjustment factor using the following formula: , , In the formula: c i Let be the endmember spectral coefficients of the image spectrum in the i-th band. g i Let be the image spectral vector in the i-th band. s i For the image spectrum at the i-th observation or signal, m i Let be the mean or expected value of the i-th variable in the image spectrum. The variance of the image spectrum in the i-th variable is used; based on the matched filter response value and noise adjustment factor, a dual threshold is set to delineate the spatial distribution of alteration zonation, including the limonite and kaolinite zones. Further, the information extraction module employs multi-scale segmentation and object-oriented analysis techniques to eliminate spectral obfuscation of vegetation cover areas in the data processed by the alteration mineral identification module; then, a terrain shadow correction model and a Quaternary cover layer masking algorithm are sequentially used to suppress background interference from complex terrain and loose sediments in the data after spectral obfuscation is eliminated; finally, the abnormal boundaries of mineralization alteration are automatically extracted from the data with suppressed background interference through adaptive threshold segmentation to generate an alteration intensity grading map. The step of automatically extracting the abnormal boundaries of mineralization alteration is as follows: First, based on data that has suppressed background interference, the mineralization alteration anomaly index of each pixel is calculated; Secondly, an adaptive threshold segmentation algorithm is used to dynamically adjust the threshold based on the statistical characteristics within the local window, thereby segmenting out abnormal boundaries; the formula is: ; In the formula, m It is a local mean. k For local standard deviation, s This is an empirical coefficient; Finally, the boundary continuity is optimized through morphological post-processing to generate a high-precision alteration intensity grading map.

[0016] Furthermore, the visualization map in the result output module includes mineral assemblage types, alteration intensity grading, and spatial distribution of alteration zones; the data storage and management module employs a distributed architecture to structurally store the original multispectral remote sensing image data, surface reflectance data, and alteration intensity grading map; the system integration module utilizes the OGC standard service and API gateway of the system integration interface to achieve seamless data exchange and function calls with remote sensing platforms and GIS software including ENVI, ArcGIS, and QGIS.

[0017] The present invention has the following beneficial effects: 1. This invention addresses the shortcomings of existing methods (such as band ratio method and principal component analysis) due to their coarse spectral feature extraction, which makes it difficult to distinguish subtle differences in mineralization alteration of copper, lead, and zinc. It utilizes a pre-built spectral feature library specifically for copper, lead, and zinc mineralization (integrating laboratory-measured spectra and standard ground cover spectra from typical mining areas). Combined with a spectral angle mapping algorithm (SAM) to calculate the angle (θ) between the image spectrum and the reference spectrum, it accurately matches target mineral assemblages (such as chlorite, epidote, sericite, etc.), thereby avoiding interference from non-target ground cover and preventing spectral confusion between different metal mineralization alterations from the source. Furthermore, it innovatively introduces a hybrid modulation matched filtering algorithm (MTMF), which calculates the matched filter response value and noise adjustment factor based on a linear spectral mixing model. This quantitatively delineates the spatial distribution of alteration zones such as limonite and kaolinite zones. Moreover, by generating an alteration intensity grading map through adaptive threshold segmentation, it clearly distinguishes between "weak alteration, medium alteration, and strong alteration," providing quantitative data support for mineral resource reserve estimation and mineralization potential assessment, thus overcoming the "qualitative but not quantitative" deficiency of traditional techniques. Ultimately, it achieves the dual goals of "qualitative identification + quantitative zoning," not only breaking through the spectral confusion bottleneck of traditional technologies but also significantly improving the accuracy of mineralization alteration identification.

[0018] 2. This invention addresses the shortcomings of existing technologies in vegetated areas, where misjudgment can easily occur due to the overlap of vegetation and alteration mineral spectra. By employing multi-scale segmentation and object-oriented analysis techniques, it upgrades "pixel-level spectral analysis" to "object-level feature extraction." This allows for precise separation of the spatial-spectral correlation features between vegetation and alteration features, effectively eliminating spectral confusion in vegetated areas and expanding the technology's applicability from arid regions to highly vegetated copper-lead-zinc mining areas. Furthermore, addressing the limitations of existing methods that are sensitive to topographic shadows and loose sediments (Quaternary cover), easily misjudging shadows as weak alteration and sediments as mineralization zones, this invention utilizes a topographic shadow correction model to eliminate reflectivity deviations caused by slope / aspect. Combined with a Quaternary cover masking algorithm, it actively shields areas with loose sediments, significantly reducing background interference from complex terrain and cover layers, and improving the extraction reliability in complex geomorphological areas such as mountains and valleys. By employing the above technologies, this invention strengthens its anti-interference capability against complex backgrounds and overcomes the extraction limitations of special areas, effectively expanding its applicability and improving the reliability of information extraction.

[0019] 3. This invention addresses the shortcomings of existing technologies, which require manual selection of ratio bands, analysis of PCA load vectors, and manual segmentation of anomaly regions, resulting in cumbersome and inefficient processes. It constructs a fully automated workflow of "data acquisition-preprocessing-identification-extraction-output": the spectral preprocessing stage standardizes radiometric calibration (official coefficients), atmospheric correction (FLAASH / 6S model), and noise suppression (adaptive filtering), reducing manual parameter adjustments; the alteration identification stage automatically matches spectral libraries and algorithm parameters; and the results output stage automatically integrates GIS data to generate visual maps and comprehensive reports, meeting the timeliness requirements of rapid mineral resource exploration, ultimately achieving full automation and significantly improving extraction efficiency. Furthermore, it addresses the shortcomings of existing technologies that rely on single software (such as ENVI) and struggle to integrate data with GIS platforms. This invention achieves seamless data exchange and function calls with mainstream remote sensing / GIS software such as ENVI, ArcGIS, and QGIS through OGC standard services and API gateways. Simultaneously, it employs a distributed architecture to enable structured archiving and incremental updates of raw, preprocessed, and result data, supporting collaborative work among multiple teams. This solves the problems of "data silos" and "software compatibility," lowering the technical application threshold for grassroots exploration units. Through these measures, this invention significantly improves the level of process automation and integration, meeting the needs of large-scale, rapid exploration.

[0020] 4. This invention addresses the shortcomings of existing technologies, which often output only single alteration anomaly maps and lack correlation with geological attributes. Based on a GIS spatial analysis engine, it deeply integrates alteration zoning, alteration intensity classification, and geological maps and mineral point data of mining areas to generate a visualized map containing "mineral assemblage type - alteration intensity - spatial distribution of alteration zoning." This visually presents the zoning pattern of copper-lead-zinc mineralization as "copper mineralization zone → copper-lead-zinc mineralization zone → lead-zinc mineralization zone," enabling geologists to quickly identify favorable mineralization areas. Furthermore, in the output stage, it can automatically associate geological attributes (such as stratigraphy and structural lines) to generate a comprehensive report, including anomaly level assessment and mineral exploration target area delineation suggestions. This directly provides precise guidance for field exploration in terms of "target area coordinates - alteration intensity - mineralization probability," avoiding the limitations of existing technologies that "only output data without application suggestions," and achieving a closed loop from "information extraction" to "mineral exploration guidance."

[0021] In summary, this invention, through its innovative approach of "dedicated spectral library + dual-algorithm fusion + anti-interference processing + automated integration," effectively solves the problems of "spectral confusion, weak anti-interference, low efficiency, and poor practicality" in existing technologies. It not only improves the accuracy and efficiency of extracting copper-lead-zinc mineralization alteration information but also expands the applicability of the technology in complex geological backgrounds (high vegetation, complex terrain). It provides efficient and reliable technical support for rapid regional exploration and precise delineation of prospecting targets for copper-lead-zinc mineral resources, possessing significant geological application value and promising prospects for wider application. Attached Figure Description

[0022] Figure 1 This is a flowchart of the copper-lead-zinc mineralization alteration information extraction method based on multispectral remote sensing of the present invention. Detailed Implementation

[0023] The present invention will be further described below with reference to embodiments, but this is not intended to limit the present invention in any way. Any changes or improvements made based on the teachings of the present invention shall fall within the protection scope of the present invention.

[0024] This invention relates to a method for extracting copper-lead-zinc mineralization alteration information based on multispectral remote sensing, comprising the steps of remote sensing data acquisition, spectral preprocessing, alteration mineral identification, information extraction, result output, data storage and management, and system integration. The specific steps are as follows: A. Remote sensing data acquisition: Receive raw multispectral remote sensing image data of the target area from satellite and / or airborne sensors; B. Spectral preprocessing: Radiometric calibration, atmospheric correction and noise suppression are performed on the received raw multispectral remote sensing image data to generate surface reflectance data; C. Alteration mineral identification: Based on a pre-built database of copper, lead, and zinc mineralization alteration spectral features, the aforementioned surface reflectance data is processed using a spectral angle mapping algorithm and a hybrid modulation matched filtering algorithm, and the target mineral assemblage in the data is identified and the spatial distribution of alteration zones is quantitatively delineated. D. Information extraction: Eliminate spectral confusion of vegetation cover areas in the data processed in step C, then suppress background interference from complex terrain and loose sediments, and finally extract mineralization alteration anomaly boundaries to generate an alteration intensity grading map. E. Output Results: Based on the alteration intensity grading map, a visual map is generated by integrating alteration zone spatial distribution information, mining area geological map, and mineral point data using a GIS spatial analysis engine. Then, spatial overlay analysis technology and GIS attribute table association function are used to automatically achieve dynamic matching between mineral point data and alteration intensity grading, and a comprehensive report is generated. Based on the comprehensive report, the final mineralization and alteration information product is generated, including an alteration anomaly distribution map, intensity grading report, and mineral point matching results. F. Data storage and management: Distributed archiving of raw multispectral remote sensing image data, surface reflectance data, and alteration intensity classification maps, and implementation of version control and incremental updates; G. System Integration: Utilize system integration interfaces to achieve data interaction with remote sensing platforms and GIS software.

[0025] In step B, the radiometric calibration uses the official calibration coefficients of the sensor, the atmospheric correction is based on the radiative transfer model (such as FLAASH, 6S) or the empirical linear method to eliminate the effects of scattering and absorption, and the noise suppression is achieved through an adaptive filtering algorithm.

[0026] The radiative transfer model simulates the relationship between satellite observations and atmospheric conditions (temperature, pressure, gas concentration, etc.) and surface information (emissivity) by solving the radiative transfer equation, generating a reference spectrum. This model is used for sensitivity analysis and channel selection during the inversion process. By perturbing atmospheric variables and comparing spectral differences, sensitive channels for key parameters are determined, and the surface reflectivity is corrected to the true radiative value, thereby improving the accuracy of remote sensing data.

[0027] The empirical linear method establishes a linear relationship between surface reflectance and remote sensing images based on statistical characteristics. It identifies "unchanging targets" with stable reflectance in the images and uses the reflectance variation patterns of these targets at different times for correction.

[0028] The noise suppression is achieved through an adaptive filtering algorithm, which dynamically adjusts the weighting coefficients based on the error between the input signal and the desired signal (such as the original signal) to gradually reduce the impact of noise.

[0029] In step C, the pre-set spectral feature library of copper, lead, and zinc mineralization alteration is integrated from laboratory measured spectra and standard ground cover spectra of typical mining areas; the spectral angle mapping algorithm is used to calculate the corresponding spectral angle θ in the aforementioned surface reflectance data, and compare it with the set spectral angle θ threshold to identify target mineral combinations including chlorite, epidote, and sericite; the hybrid modulation matched filtering algorithm is used to optimize the endmember abundance estimation based on the linear spectral mixing model, and quantitatively delineate the spatial distribution of alteration zonation including limonite and kaolinite zones in the surface reflectance data.

[0030] Furthermore, the formula for calculating the spectral angle θ in the spectral angle mapping algorithm is as follows:

[0031] In the formula: t i r is the value of the reference spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library. i Let θ be the value of the image spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library, and n be the number of bands in the spectrum; where a smaller spectral angle θ indicates a higher spectral matching degree. The hybrid modulation matched filtering algorithm is based on a linear spectral mixing model, which calculates the matched filter response and noise adjustment factor using the following formula: , , In the formula: c i Let be the endmember spectral coefficients of the image spectrum in the i-th band. g i Let be the image spectral vector in the i-th band. si For the image spectrum at the i-th observation or signal, m i Let be the mean or expected value of the i-th variable in the image spectrum. Let be the variance of the image spectrum in the i-th variable; set dual thresholds based on the matched filter response value and noise adjustment factor to delineate the spatial distribution of alteration zonation, including limonite and kaolinite zones.

[0032] The linear spectral mixing model is a core method in hyperspectral remote sensing for resolving mixed pixel problems. Its core idea is to reconstruct observation data by linearly combining known endmember spectra. It is a mature existing technology, and the details will not be elaborated here. The endmember spectral coefficient refers to the proportion of spectral contribution of each endmember in a mixed pixel in a specific band, which is used to describe the spectral characteristics of ground cover components.

[0033] In step D, multi-scale segmentation and object-oriented analysis techniques are used to eliminate spectral obfuscation of vegetation cover areas in the data processed in step C. Then, the terrain shadow correction model and Quaternary cover mask algorithm are used in sequence to suppress background interference from complex terrain and loose sediments in the data after eliminating spectral obfuscation. Finally, the abnormal boundaries of mineralization alteration are automatically extracted from the data with suppressed background interference through adaptive threshold segmentation to generate an alteration intensity grading map. The steps for automatically extracting abnormal boundaries of mineralization alteration and generating an alteration intensity grading map are as follows: First, based on data that has suppressed background interference, the mineralization alteration anomaly index of each pixel is calculated; Secondly, an adaptive threshold segmentation algorithm is used to dynamically adjust the threshold based on statistical characteristics (such as mean and standard deviation) within the local window to segment out abnormal boundaries; the formula is: In the formula, m It is a local mean. k For local standard deviation, s This is an empirical coefficient (usually taken as 1.5~2.5); Finally, boundary continuity is optimized through morphological post-processing (such as dilation and corrosion) to generate a high-precision alteration intensity grading map.

[0034] In step E, based on the GIS spatial analysis engine, the spatial distribution information of alteration zones, the geological map of the mining area, and the data of mineral deposits are integrated. This involves using an overlay analysis tool to overlay the geological map of the mining area with the alteration zone data, identifying the intersection area between the geological structure and the alteration zone, creating a buffer zone for the mineral deposits through buffer analysis, and then overlaying it with the alteration zone and the geological map to analyze the relationship between the distribution of mineral deposits and the geological structure.

[0035] In step E, spatial overlay analysis technology and GIS attribute table association function are used to automatically achieve dynamic matching between mineral point data and alteration intensity classification, and generate a comprehensive report. Based on the comprehensive report, a final mineralization and alteration information product is generated, including an alteration anomaly distribution map, intensity classification report, and mineral point matching results. This involves spatial overlay analysis of mineral point location data and alteration intensity classification map. Using the overlay analysis function of the GIS spatial analysis engine, the alteration intensity level and its corresponding alteration zone of each mineral point are accurately calculated. Through the GIS attribute table association function, a dynamic link relationship library is established between mineral point attributes (such as mineral type, scale, and genetic type) and corresponding alteration anomaly area characteristics (such as alteration mineral assemblage, zoning type, and intensity level). Based on the spatial association results, a comprehensive report is automatically generated, including the spatial distribution of mineralization and alteration anomaly areas, alteration intensity classification statistics, and mineral point and alteration anomaly matching degree analysis.

[0036] In step E, the visualization map includes mineral assemblage type, alteration intensity classification, and spatial distribution of alteration zoning.

[0037] In step F, a distributed architecture is used to perform structured storage of the original multispectral remote sensing image data, surface reflectance data, and alteration intensity grading map.

[0038] In step G, the OGC standard service and API gateway of the system integration interface are used to achieve seamless data exchange and function calls with remote sensing platforms and GIS software including ENVI, ArcGIS, and QGIS.

[0039] This invention relates to a multispectral remote sensing-based copper-lead-zinc mineralization alteration information extraction system, comprising a remote sensing data acquisition module, a spectral preprocessing module, an alteration mineral identification module, an information extraction module, a result output module, a data storage and management module, and a system integration module. The remote sensing data acquisition module is connected to the spectral preprocessing module and is used to receive raw multispectral remote sensing image data of the target area from satellite and / or airborne sensors. The spectral preprocessing module, connected to the alteration mineral identification module, is used to perform radiometric calibration, atmospheric correction, and noise suppression on the received raw multispectral remote sensing image data to generate high-quality surface reflectance data. The alteration mineral identification module is connected to the information extraction module. It is used to process the aforementioned surface reflectance data based on a pre-set copper, lead, and zinc mineralization alteration spectral feature library, using a spectral angle mapping algorithm and a hybrid modulation matched filtering algorithm, and to identify the target mineral combination in the data and quantitatively delineate the spatial distribution of alteration zones. The information extraction module is connected to the result output module. It is used to eliminate spectral confusion of vegetation cover area in the data output by the alteration mineral identification module, then suppress background interference from complex terrain and loose sediments, and finally extract mineralization alteration anomaly boundaries to generate an alteration intensity grading map. The result output module is connected to the data storage and management module. It is used to generate a visual map based on the alteration intensity grading map, by integrating alteration zone spatial distribution information, mining area geological map and mineral point data based on the GIS spatial analysis engine. Then, it uses spatial overlay analysis technology and GIS attribute table association function to automatically realize dynamic matching between mineral point data and alteration intensity grading, and generate a comprehensive report. Based on the comprehensive report, it generates the final mineralization and alteration information product, which includes an alteration anomaly distribution map, intensity grading report and mineral point matching results. The data storage and management module is connected to the system integration module and is used to perform distributed archiving of raw multispectral remote sensing image data, surface reflectance data and alteration intensity grading maps, and to perform version control and incremental updates. The system integration module is used to achieve data interaction with the remote sensing platform and GIS software through the system integration interface.

[0040] The alteration mineral identification module contains a pre-set spectral feature library of copper, lead, and zinc mineralization alteration, which is integrated from laboratory measured spectra and standard ground cover spectra of typical mining areas. The spectral angle mapping algorithm is used to calculate the corresponding spectral angle θ in the aforementioned surface reflectance data and compare it with the set spectral angle θ threshold to identify target mineral combinations including chlorite, epidote, and sericite. The hybrid modulation matched filtering algorithm is used to optimize the endmember abundance estimation based on the linear spectral mixing model to quantitatively delineate the spatial distribution of alteration zonation, including limonite and kaolinite zones, in the surface reflectance data. The formula for calculating the spectral angle θ in the spectral angle mapping algorithm is as follows:

[0041] In the formula: t i r is the value of the reference spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library. i Let θ be the value of the image spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library, and n be the number of bands in the spectrum; where a smaller spectral angle θ indicates a higher spectral matching degree. The hybrid modulation matched filtering algorithm is based on a linear spectral mixing model, which calculates the matched filter response and noise adjustment factor using the following formula: , , In the formula: c iLet be the endmember spectral coefficients of the image spectrum in the i-th band. g i Let be the image spectral vector in the i-th band. s i For the image spectrum at the i-th observation or signal, m i Let be the mean or expected value of the i-th variable in the image spectrum. Let be the variance of the image spectrum in the i-th variable; set dual thresholds based on the matched filter response value and noise adjustment factor to delineate the spatial distribution of alteration zonation, including limonite and kaolinite zones.

[0042] The information extraction module employs multi-scale segmentation and object-oriented analysis techniques to eliminate spectral confusion in vegetation-covered areas from the data processed by the alteration mineral identification module. Then, it sequentially uses a terrain shadow correction model and a Quaternary overburden masking algorithm to suppress background interference from complex terrain and loose sediments in the data after spectral confusion has been eliminated. Finally, it automatically extracts abnormal boundaries of mineralization alteration from the data with suppressed background interference through adaptive threshold segmentation to generate an alteration intensity grading map. The step of automatically extracting the abnormal boundaries of mineralization alteration is as follows: First, based on data that has suppressed background interference, the mineralization alteration anomaly index of each pixel is calculated; Secondly, an adaptive threshold segmentation algorithm is used to dynamically adjust the threshold based on statistical characteristics (such as mean and standard deviation) within the local window to segment out abnormal boundaries; the formula is: In the formula, m It is a local mean. k For local standard deviation, s This is an empirical coefficient (usually taken as 1.5~2.5); Finally, boundary continuity is optimized through morphological post-processing (such as dilation and corrosion) to generate a high-precision alteration intensity grading map.

[0043] The visualization map in the results output module includes mineral assemblage types, alteration intensity grading, and spatial distribution of alteration zones; the data storage and management module uses a distributed architecture to structure and store the original multispectral remote sensing image data, surface reflectance data, and alteration intensity grading map; the system integration module utilizes the OGC standard service and API gateway of the system integration interface to achieve seamless data exchange and function calls with remote sensing platforms and GIS software including ENVI, ArcGIS, and QGIS.

[0044] Example

[0045] like Figure 1 As shown, the extraction of copper-lead-zinc mineralization alteration information based on multispectral remote sensing is carried out in the following specific process: S100: The remote sensing data acquisition module automatically receives raw multispectral remote sensing image data of the target area from satellite platforms (such as Landsat-8 OLI, Sentinel-2) and airborne sensors, and acquires raw multispectral image data of the target area covering the visible light, near-infrared and short-wave infrared bands.

[0046] The remote sensing data acquisition module has a built-in standardized data interface that supports receiving raw DN value data or surface reflectance products that have undergone radiometric calibration and atmospheric correction preprocessing. It automatically performs data format unification and metadata extraction (specifically as follows: First, the module parses the input data file format (e.g., GeoTIFF, HDF, or NetCDF), identifies the encoding structure of the raw data, and applies standard conversion algorithms (e.g., the GDAL library) to unify heterogeneous formats into an internally common raster data model; second, it extracts metadata information, including image acquisition time, spatial resolution (e.g., 10 meters or 30 meters), band center wavelength, sensor type, projection coordinate system, and data quality identifier, and writes this information into a structured metadata database using an XML or JSON parser; next, it performs data integrity verification, comparing the metadata with the number of image bands, file size, and checksum value to ensure no missing or corrupted data; if an anomaly is detected, it triggers an error log and notifies the system administrator; finally, it encapsulates the unified data into a standard output product (e.g., COG format) and generates a metadata summary report for direct use by subsequent modules).

[0047] S200: The spectral preprocessing module performs radiometric calibration (using the official calibration coefficients of the sensor) on the acquired multispectral raw image data, converting the raw DN values ​​into physically meaningful apparent radiance; it uses atmospheric correction algorithms based on radiative transfer models (such as FLAASH, 6S) or empirical linear methods to eliminate the effects of atmospheric scattering and absorption, obtaining high-precision surface reflectance data; it uses adaptive filters (such as mean filters, median filters) to dynamically adjust the weighting coefficients according to the error between the input signal and the desired signal (such as the raw signal), gradually reducing the impact of noise and ensuring the fidelity of the spectral data; finally, it outputs the preprocessed standardized reflectance data to the alteration mineral identification module.

[0048] S300: The alteration mineral identification module is based on a pre-built spectral feature library of copper, lead, and zinc mineralization alteration (integrated from laboratory measured spectra and standard ground cover spectra of typical mining areas). It uses the spectral angle mapping algorithm (SAM) to calculate the corresponding spectral angle θ in the preprocessed standardized reflectance data and compares it with the set spectral angle θ threshold. A machine learning classifier (such as support vector machine SVM or random forest) is introduced to identify target mineral combinations including chlorite, epidote, and sericite. At the same time, the hybrid modulation matched filtering algorithm (MTMF) is used to optimize the endmember abundance estimation based on the linear spectral mixture model to quantitatively delineate the spatial distribution of alteration zonation in the surface reflectance data, including limonite and kaolinite zones. Finally, the identified target mineral combinations and the quantitatively delineated spatial distribution of alteration zonation are output to the information extraction module.

[0049] The formula for calculating the spectral angle θ in the spectral angle mapping algorithm is as follows:

[0050] In the formula: t i r is the value of the reference spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library. i denoted as the value of the image spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library, where n is the number of bands in the spectrum; and the smaller the spectral angle θ, the higher the spectral matching degree.

[0051] The hybrid modulation matched filtering algorithm is based on a linear spectral mixing model, which calculates the matched filter response and noise adjustment factor using the following formula: , , In the formula: c i Let be the endmember spectral coefficients of the image spectrum in the i-th band. g i Let be the image spectral vector in the i-th band. s i For the image spectrum at the i-th observation or signal, m i Let be the mean or expected value of the i-th variable in the image spectrum. Let be the variance of the image spectrum in the i-th variable; set dual thresholds based on the matched filter response value and noise adjustment factor to delineate the spatial distribution of alteration zonation, including limonite and kaolinite zones.

[0052] S400: The information extraction module integrates multi-source geographic information data (such as Digital Elevation Model (DEM) and National Distance Index (NDVI)). First, it employs multi-scale segmentation and object-oriented analysis techniques to eliminate spectral obfuscation of vegetation cover areas in the data processed by the alteration mineral identification module. Then, it uses a terrain shadow correction model and a Quaternary cover mask algorithm to suppress background interference from complex terrain and loose sediments in the data after spectral obfuscation elimination. Finally, for the data with suppressed background interference, morphological filtering (such as opening and closing operations) and region growing techniques are used to enhance the spatial continuity of alteration zoning. The module also fuses the distribution map output by the alteration mineral identification module with the confidence assessment results (the fusion method is as follows: a weighted Bayesian probabilistic fusion algorithm is used, with the confidence assessment results as weighting factors, to spatially weight and overlay the alteration mineral distribution map to generate a comprehensive mineralization and alteration intensity map; simultaneously, combined with adaptive threshold segmentation results, decision-level fusion technology is applied to optimize the alteration zoning boundaries, eliminate local inconsistencies, and ensure high accuracy and interpretability of the information). This system automatically generates high-precision mineralization alteration zoning raster maps and vector boundary files. The automatic generation process is as follows: based on a comprehensive mineralization alteration intensity map, an adaptive raster segmentation algorithm (such as watershed segmentation or region growing) is applied to accurately divide alteration zoning units; secondly, edge extraction and polygonization are used to convert raster boundaries into vector format, ensuring topological integrity and minimizing geometric errors; simultaneously, an automatic metadata filling mechanism is embedded, including coordinate system information, resolution parameters, and confidence scores, to generate standardized GeoTIFF raster files and Shapefile vector files; finally, post-processing optimization is performed, applying the Douglas-Peucker algorithm to simplify redundant points on vector boundaries, and integrating DEM data to correct terrain distortion, outputting high-fidelity results that can be directly used for mineral exploration. Furthermore, statistical methods (such as Kriging interpolation) are used to quantify uncertainty and intelligently extract abnormal boundaries of mineralization alteration, generating an alteration intensity grading map to provide reliable spatial decision support for mineral resource exploration.

[0053] S500: The results output module, based on the alteration intensity grading map, integrates alteration zoning spatial distribution information, mining area geological maps, and mineral point data through a Geographic Information System (GIS) platform interface (such as ArcGIS or QGIS). It renders the alteration information distribution map into a high-resolution color thematic map (including mineral assemblage type, alteration intensity grading, and alteration zoning spatial distribution), and integrates statistical analysis functions to automatically correlate geological attributes and generate a comprehensive report (specifically, based on the mineral distribution map and confidence assessment results output by the alteration mineral identification module, automatically...). Key statistical indicators of mineralized alteration areas are extracted, including total alteration area, abundance proportions of various minerals (such as chlorite, epidote, sericite, limonite, and kaolinite), average alteration intensity, and confidence level coefficient of variation. Secondly, multi-source geographic information data (such as DEM and NDVI) are integrated, and statistical analysis algorithms (such as principal component analysis (PCA) and spatial autocorrelation analysis) are applied to calculate the correlation between alteration zonation and topographic relief and vegetation cover, generating quantitative report tables and visualization charts (such as histograms and scatter plots). Finally, through an interface, the rendered high-resolution data is... The report document embeds thematic maps (including alteration zoning raster maps and vector boundaries) and automatically annotates metadata (such as coordinate system, resolution, and confidence score). Finally, it automatically generates a structured geological report document (PDF or Word format) using standard templates, including an executive summary, detailed description of the alteration area, statistical results, charts, and appendices (such as uncertainty quantification reports). Users can export or print the report with one click, facilitating resource assessment and exploration decisions. The report content includes the abundance distribution of alteration minerals, spatial variability characteristics, confidence assessment results, and uncertainty quantification indicators, and proposes anomaly level assessments and mineral target area delineation suggestions. It also supports multiple output formats (such as GeoTIFF raster files, Shapefile vector boundaries, and PDF documents), allowing users to directly import the data into mineral exploration software or print it. Furthermore, it provides a web-based interactive visualization platform (such as using Leaflet or OpenLayers frameworks) to enable online dynamic display, spatial querying, and multi-scale zooming of alteration information in the mining area, improving user decision-making efficiency and data sharing capabilities.

[0054] S600: The data storage and management module constructs an elastic storage architecture based on a distributed database (such as Hadoop HDFS or cloud storage services) to achieve efficient archiving and long-term preservation of raw multispectral remote sensing image data, surface reflectance data, and alteration intensity grading maps. It uses a spatial database engine (such as PostGIS or GeoMesa) to organize multi-source data, establish metadata and spatial indexes, and support fast queries, fuzzy searches, and dynamic updates based on geographic coordinates, timestamps, and attribute conditions. It integrates a data version control mechanism and a regular backup strategy to ensure data integrity and recoverability. At the same time, it provides a graphical management interface and API interface, allowing users to perform batch data import and export, permission allocation, and data lifecycle monitoring, realizing intelligent management of multi-source data and providing unified and reliable data support for all modules of the system.

[0055] S700: The system integration module provides standardized application programming interfaces (APIs) and middleware support through the system integration interface, enabling seamless integration with mainstream remote sensing platforms (such as ENVI and ERDAS IMAGINE) and GIS software (such as ArcGIS and QGIS). This interface is built based on Open Geospatial Consortium (OGC) standards (such as WMS and WFS), supports automated data flow and status monitoring between modules, and enables real-time data interaction and command invocation with external systems (such as mineral databases and exploration management systems) through RESTful or SOAP protocols. At the same time, it integrates security authentication mechanisms (such as OAuth 2.0) and encrypted data transmission to ensure the security of cross-platform collaboration and improve the overall scalability and interoperability of the system.

[0056] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for extracting copper-lead-zinc mineralization alteration information based on multispectral remote sensing, characterized in that: The process includes remote sensing data acquisition, spectral preprocessing, alteration mineral identification, information extraction, result output, data storage and management, and system integration. The specific steps are as follows: A. Remote sensing data acquisition: Receive raw multispectral remote sensing image data of the target area from satellite and / or airborne sensors; B. Spectral preprocessing: Radiometric calibration, atmospheric correction and noise suppression are performed on the received raw multispectral remote sensing image data to generate surface reflectance data; C. Alteration mineral identification: Based on a pre-built database of copper, lead, and zinc mineralization alteration spectral features, the aforementioned surface reflectance data is processed using a spectral angle mapping algorithm and a hybrid modulation matched filtering algorithm, and the target mineral assemblage in the data is identified and the spatial distribution of alteration zones is quantitatively delineated. D. Information extraction: Eliminate spectral confusion of vegetation cover areas in the data processed in step C, then suppress background interference from complex terrain and loose sediments, and finally extract mineralization alteration anomaly boundaries to generate an alteration intensity grading map. E. Output Results: Based on the alteration intensity grading map, a visual map is generated by integrating alteration zone spatial distribution information, mining area geological map, and mineral point data using a GIS spatial analysis engine. Then, spatial overlay analysis technology and GIS attribute table association function are used to automatically achieve dynamic matching between mineral point data and alteration intensity grading, and a comprehensive report is generated. Based on the comprehensive report, the final mineralization and alteration information product is generated, including an alteration anomaly distribution map, intensity grading report, and mineral point matching results. F. Data storage and management: Distributed archiving of raw multispectral remote sensing image data, surface reflectance data, and alteration intensity classification maps, and implementation of version control and incremental updates; G. System Integration: Utilize system integration interfaces to achieve data interaction with remote sensing platforms and GIS software.

2. The method for extracting copper-lead-zinc mineralization alteration information based on multispectral remote sensing according to claim 1, characterized in that: In step B, the radiometric calibration uses the official calibration coefficients of the sensor, the atmospheric correction is based on the radiative transfer model or the empirical linear method to eliminate the effects of scattering and absorption, and the noise suppression is achieved through an adaptive filtering algorithm.

3. The method for extracting copper-lead-zinc mineralization alteration information based on multispectral remote sensing according to claim 1, characterized in that: In step C, the pre-set spectral feature library of copper, lead, and zinc mineralization alteration is integrated from laboratory measured spectra and standard ground cover spectra of typical mining areas; the spectral angle mapping algorithm is used to calculate the corresponding spectral angle θ in the aforementioned surface reflectance data, and compare it with the set spectral angle θ threshold to identify target mineral combinations including chlorite, epidote, and sericite; the hybrid modulation matched filtering algorithm is used to optimize the endmember abundance estimation based on the linear spectral mixing model, and quantitatively delineate the spatial distribution of alteration zonation including limonite and kaolinite zones in the surface reflectance data.

4. The method for extracting copper-lead-zinc mineralization alteration information based on multispectral remote sensing according to claim 3, characterized in that: The formula for calculating the spectral angle θ in the spectral angle mapping algorithm is as follows: In the formula: t i r is the value of the reference spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library. i Let be the value of the image spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library, and n be the number of bands in the spectrum; where a smaller spectral angle θ indicates a higher spectral matching degree. The hybrid modulation matched filtering algorithm is based on a linear spectral mixing model, which calculates the matched filter response and noise adjustment factor using the following formula: ， ， In the formula: c i Let be the endmember spectral coefficients of the image spectrum in the i-th band. g i Let be the image spectral vector in the i-th band. s i For the image spectrum at the i-th observation or signal, μ i Let be the mean or expected value of the i-th variable in the image spectrum. Let be the variance of the image spectrum in the i-th variable; set dual thresholds based on the matched filter response value and noise adjustment factor to delineate the spatial distribution of alteration zonation, including limonite and kaolinite zones.

5. The method for extracting copper-lead-zinc mineralization alteration information based on multispectral remote sensing according to claim 1, characterized in that: In step D, multi-scale segmentation and object-oriented analysis techniques are used to eliminate spectral obfuscation of vegetation cover areas in the data processed in step C. Then, the terrain shadow correction model and Quaternary cover mask algorithm are used in sequence to suppress background interference from complex terrain and loose sediments in the data after eliminating spectral obfuscation. Finally, the abnormal boundaries of mineralization alteration are automatically extracted from the data with suppressed background interference through adaptive threshold segmentation to generate an alteration intensity grading map. The steps for automatically extracting abnormal boundaries of mineralization alteration and generating an alteration intensity grading map are as follows: First, based on data that has suppressed background interference, the mineralization alteration anomaly index of each pixel is calculated; Secondly, an adaptive threshold segmentation algorithm is used to dynamically adjust the threshold based on the statistical characteristics within the local window, thereby segmenting out abnormal boundaries; the formula is: In the formula, μ It is a local mean. k For local standard deviation, σ This is an empirical coefficient; Finally, the boundary continuity is optimized through morphological post-processing to generate a high-precision alteration intensity grading map.

6. The method for extracting copper-lead-zinc mineralization alteration information based on multispectral remote sensing according to any one of claims 1 to 5, characterized in that: In step E, the visualized map includes mineral assemblage types, alteration intensity grading, and spatial distribution of alteration zones; in step F, a distributed architecture is used to structurally store the original multispectral remote sensing image data, surface reflectance data, and alteration intensity grading map; in step G, the OGC standard service and API gateway of the system integration interface are used to achieve seamless data exchange and function calls with remote sensing platforms and GIS software including ENVI, ArcGIS, and QGIS.

7. A copper-lead-zinc mineralization alteration information extraction system based on multispectral remote sensing, characterized in that: It includes modules for remote sensing data acquisition, spectral preprocessing, alteration mineral identification, information extraction, result output, data storage and management, and system integration. The remote sensing data acquisition module is connected to the spectral preprocessing module and is used to receive raw multispectral remote sensing image data of the target area from satellite and / or airborne sensors. The spectral preprocessing module, connected to the alteration mineral identification module, is used to perform radiometric calibration, atmospheric correction, and noise suppression on the received raw multispectral remote sensing image data to generate high-quality surface reflectance data. The alteration mineral identification module is connected to the information extraction module. It is used to process the aforementioned surface reflectance data based on a pre-set copper, lead, and zinc mineralization alteration spectral feature library, using a spectral angle mapping algorithm and a hybrid modulation matched filtering algorithm, and to identify the target mineral combination in the data and quantitatively delineate the spatial distribution of alteration zones. The information extraction module is connected to the result output module. It is used to eliminate spectral confusion of vegetation cover area in the data output by the alteration mineral identification module, then suppress background interference from complex terrain and loose sediments, and finally extract mineralization alteration anomaly boundaries to generate an alteration intensity grading map. The result output module is connected to the data storage and management module. It is used to generate a visual map based on the alteration intensity grading map, by integrating alteration zone spatial distribution information, mining area geological map and mineral point data based on the GIS spatial analysis engine. Then, it uses spatial overlay analysis technology and GIS attribute table association function to automatically realize dynamic matching between mineral point data and alteration intensity grading, and generate a comprehensive report. Based on the comprehensive report, it generates the final mineralization and alteration information product, which includes an alteration anomaly distribution map, intensity grading report and mineral point matching results. The data storage and management module is connected to the system integration module and is used to perform distributed archiving of raw multispectral remote sensing image data, surface reflectance data and alteration intensity grading maps, and to perform version control and incremental updates. The system integration module is used to achieve data interaction with the remote sensing platform and GIS software through the system integration interface.

8. The copper-lead-zinc mineralization alteration information extraction system based on multispectral remote sensing according to claim 7, characterized in that: The alteration mineral identification module contains a pre-set spectral feature library of copper, lead, and zinc mineralization alteration, which is integrated from laboratory measured spectra and standard ground cover spectra of typical mining areas. The spectral angle mapping algorithm is used to calculate the corresponding spectral angle θ in the aforementioned surface reflectance data and compare it with the set spectral angle θ threshold to identify target mineral combinations including chlorite, epidote, and sericite. The hybrid modulation matched filtering algorithm is used to optimize the endmember abundance estimation based on the linear spectral mixing model to quantitatively delineate the spatial distribution of alteration zonation, including limonite and kaolinite zones, in the surface reflectance data. The formula for calculating the spectral angle θ in the spectral angle mapping algorithm is as follows: In the formula: t i r is the value of the reference spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library. i Let be the value of the image spectrum in the i-th band from the aforementioned mineralization alteration spectral feature library, and n be the number of bands in the spectrum; where a smaller spectral angle θ indicates a higher spectral matching degree. The hybrid modulation matched filtering algorithm is based on a linear spectral mixing model, which calculates the matched filter response and noise adjustment factor using the following formula: ， ， In the formula: c i Let be the endmember spectral coefficients of the image spectrum in the i-th band. g i Let be the image spectral vector in the i-th band. s i For the image spectrum at the i-th observation or signal, μ i Let be the mean or expected value of the i-th variable in the image spectrum. Let be the variance of the image spectrum in the i-th variable; set dual thresholds based on the matched filter response value and noise adjustment factor to delineate the spatial distribution of alteration zonation, including limonite and kaolinite zones.

9. The copper-lead-zinc mineralization alteration information extraction system based on multispectral remote sensing according to claim 7, characterized in that: The information extraction module employs multi-scale segmentation and object-oriented analysis techniques to eliminate spectral confusion in vegetation-covered areas from the data processed by the alteration mineral identification module. Then, it sequentially uses a terrain shadow correction model and a Quaternary overburden masking algorithm to suppress background interference from complex terrain and loose sediments in the data after spectral confusion has been eliminated. Finally, it automatically extracts abnormal boundaries of mineralization alteration from the data with suppressed background interference through adaptive threshold segmentation to generate an alteration intensity grading map. The step of automatically extracting the abnormal boundaries of mineralization alteration is as follows: First, based on data that has suppressed background interference, the mineralization alteration anomaly index of each pixel is calculated; Secondly, an adaptive threshold segmentation algorithm is used to dynamically adjust the threshold based on the statistical characteristics within the local window, thereby segmenting out abnormal boundaries; the formula is: In the formula, μ It is a local mean. k For local standard deviation, σ This is an empirical coefficient; Finally, the boundary continuity is optimized through morphological post-processing to generate a high-precision alteration intensity grading map.

10. The copper-lead-zinc mineralization alteration information extraction system based on multispectral remote sensing according to claim 7, 8, or 9, characterized in that: The visualization map in the results output module includes mineral assemblage types, alteration intensity grading, and spatial distribution of alteration zones; the data storage and management module uses a distributed architecture to structure and store the original multispectral remote sensing image data, surface reflectance data, and alteration intensity grading map; the system integration module utilizes the OGC standard service and API gateway of the system integration interface to achieve seamless data exchange and function calls with remote sensing platforms and GIS software including ENVI, ArcGIS, and QGIS.

Images