A method and device for exploring clay-type lithium ore

Abstract

The present invention discloses a method and device for exploring clay-type lithium ore, which relates to the technical field of clay-type lithium ore exploration, and solves the problem of low accuracy of clay-type lithium ore exploration in the prior art. It includes: obtaining geological data, mineral data and remote sensing image data of a preset area, analyzing the geological data and mineral data to determine the mineralization theory of clay-type lithium ore, based on a preset orthogonalization algorithm and a preset inverse orthogonalization algorithm, fusing multispectral surface reflectance data and panchromatic band surface reflectance data to obtain target high-resolution multispectral surface reflectance data, and then performing visualization operations and establishing structural interpretation marks, extracting ore-controlling elements from optical surface reflectance data, and determining the target mineralization area based on the mineralization theory, structural interpretation marks and ore-controlling elements of clay-type lithium ore. In this way, the accuracy of clay-type lithium ore exploration can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of clay-type lithium ore exploration, and in particular to a method and device for exploring clay-type lithium ore. Background Art

[0002] Rare metal lithium (Li) is called "new energy metal of the 21st century" and "metal that promotes world progress" and has extremely high economic and strategic value. Currently, the lithium resources available for development and utilization in the world mainly include brine type, pegmatite type and sedimentary type, accounting for 64%, 29% and 7% of the total global lithium resources respectively. Among them, sedimentary lithium ore has important geological and economic significance due to its huge reserves.

[0003] Regarding sedimentary lithium resources, previous researchers have conducted a lot of research on the material sources of lithium-rich clay layers in the study area, the occurrence characteristics and distribution laws of lithium in clay rocks, the enrichment mechanism and mineralization process of lithium, and believed that Li in sedimentary lithium deposits + It may be adsorbed in clay minerals such as montmorillonite and kaolinite, with low solid-liquid ratio, low mineralization, and high Li + concentration, alkaline water medium environment and moderate clay stage are the key factors for Li + Ideal places for enrichment of clay minerals are prone to form high-grade clay-type lithium deposits. The enrichment of sedimentary lithium deposits is closely related to the formation of bauxite. Lithium deposits and bauxite are generally produced together and serve as signs of prospecting for each other.

[0004] Traditional methods for prospecting sedimentary lithium deposits are mainly carried out through field exploration and field geological surveys. This method is labor-intensive, time-consuming and labor-intensive, and cannot meet the requirements for obtaining large-scale mineral anomaly information. Efficient remote sensing technology is an important method for prospecting sedimentary lithium deposits. At present, the use of remote sensing technology to detect sedimentary lithium deposits is mainly used in the detection of lithium-rich pegmatites, and it also focuses on the detection of a single prospecting mark, without a comprehensive analysis of the main prospecting factors.

[0005] Therefore, a new remote sensing prospecting model for sedimentary lithium deposits is needed, and this model should be applied in practical work to identify favorable mineralization areas for lithium deposits, achieve good results, and provide scientific and technological support for maintaining the safety of lithium mineral resources. Summary of the invention

[0006] The object of the present invention is to provide a method and device for exploring clay-type lithium ore, so as to improve the exploration accuracy of clay-type lithium ore.

[0007] In order to achieve the above object, the present invention provides the following technical solutions:

[0008] In a first aspect, the present invention provides a method for exploring clay-type lithium ore, comprising:

[0009] Acquire geological data, mineral data and remote sensing image data of a preset area; the remote sensing image data includes optical surface reflectance data and a discrete mathematical expression model of surface topography; the optical surface reflectance data includes at least multispectral surface reflectance data and panchromatic band surface reflectance data;

[0010] Analyze the geological data and the mineral data to determine the mineralization theory of clay-type lithium ore;

[0011] Based on a preset orthogonalization algorithm and a preset inverse orthogonalization algorithm, the multispectral surface reflectance data and the panchromatic band surface reflectance data are fused to obtain target high-resolution multispectral surface reflectance data;

[0012] Based on the target high-resolution multi-spectral surface reflectance data and the discretized mathematical expression model of the surface topography, a visualization operation is performed and a structural interpretation mark is established; the structural interpretation mark is used to interpret the linear structure, annular structure and fold structure of the preset area in combination with the mineralization theory of the clay-type lithium ore;

[0013] Extracting ore-controlling elements from the optical surface reflectance data; the ore-controlling elements at least include chlorite;

[0014] Based on the mineralization theory of the clay-type lithium ore, the structural interpretation marks and the ore-controlling factors, the target mineralization area is determined.

[0015] Optionally, the multispectral surface reflectance data and the panchromatic band surface reflectance data are fused based on a preset orthogonalization algorithm and a preset inverse orthogonalization algorithm to obtain target high-resolution multispectral surface reflectance data, including:

[0016] Based on the multi-spectral surface reflectance data and a preset panchromatic band spectral response curve, panchromatic simulated surface reflectance data is constructed;

[0017] Based on a preset orthogonalization algorithm, the panchromatic simulated surface reflectance data and the multispectral surface reflectance data are orthogonally transformed to obtain a target orthogonal component space;

[0018] Based on the panchromatic band surface reflectance data, a preset inverse orthogonalization algorithm is used to process the target orthogonal component space to obtain initial high-resolution multispectral surface reflectance data;

[0019] The initial high-resolution multispectral surface reflectance data is feature enhanced to obtain the target high-resolution multispectral surface reflectance data.

[0020] Optionally, the multispectral surface reflectance data includes reflectance data of N bands; N is a positive integer greater than 2;

[0021] The orthogonal transformation of the panchromatic simulated surface reflectance data and the multispectral surface reflectance data based on a preset orthogonalization algorithm to obtain a target orthogonal component space includes:

[0022] Determining the panchromatic simulated surface reflectance data as a first orthogonalized component;

[0023] Substitute the reflectivity data of each band from the reflectivity data of the second band to the reflectivity data of the N bands into the preset orthogonalization formula:

[0024] ;

[0025] Each orthogonalized component of the reflectance data of each band is calculated; where, is the reflectance data of the i-th band in the multispectral surface reflectance data; is the i-th orthogonalized component corresponding to the reflectivity data of the i-th band; i is greater than 1 and i is less than N; is the jth orthogonalized component corresponding to the reflectance data of the jth band; i represents the band currently being processed; j represents the orthogonalized band; express and The inner product between express its own inner product;

[0026] The target orthogonal component space is determined based on the N orthogonalized components.

[0027] Optionally, based on the panchromatic band surface reflectance data, a preset inverse orthogonalization algorithm is used to process the target orthogonal component space to obtain initial high-resolution multispectral surface reflectance data, including:

[0028] Using the formula:

[0029] ;

[0030] The panchromatic band surface reflectance data is determined as the first target orthogonalized component in the target orthogonal component space; wherein, is the first target orthogonalized component, is the panchromatic band surface reflectance data;

[0031] Substitute the reflectance data of each band from the reflectance data of the 2nd band to the reflectance data of the Nth band in the target orthogonal component space into the preset inverse orthogonal formula:

[0032] The initial high-resolution multispectral surface reflectance data is obtained.

[0033] Optionally, the optical surface reflectance data further includes hyperspectral surface reflectance data;

[0034] The step of extracting ore-controlling elements from the optical surface reflectivity data comprises:

[0035] The first mineral information and the second mineral information are extracted from the multispectral surface reflectance data, and the mineral anomaly information is extracted from the hyperspectral surface reflectance data; the first mineral information includes at least hydroxyl, iron dye and carbonate; the second mineral information includes at least Fe-rich silicate minerals and iron-rich magnesium minerals; the mineral anomaly information includes at least lithium chlorite, montmorillonite, illite, the chlorite, kaolinite and dolomite.

[0036] Optionally, extracting mineral anomaly information from the hyperspectral surface reflectance data includes:

[0037] Extracting image pixel spectral vectors from the hyperspectral surface reflectance data;

[0038] Based on the spectral characteristics of the mineral anomaly information, a reference sample spectrum corresponding to the image pixel spectrum vector is selected from a preset reference spectrum library, and a reference sample spectrum vector is extracted from the reference sample spectrum;

[0039] Substitute the image pixel spectral vector and the reference sample spectral vector into the formula:

[0040] ;

[0041] The similarity value between the image pixel spectral vector and the reference sample spectral vector is calculated; wherein, is the spectral vector of the i-th image pixel; is the spectral vector of the i-th reference sample; The numerical value representing the similarity, Representation vector and vector The inner product of Representation vector Model; Representation vector Model;

[0042] If the value of the similarity is less than or equal to a preset angle threshold, the mineral anomaly information is extracted.

[0043] Optionally, extracting the first mineral information from the multispectral surface reflectance data includes:

[0044] Preprocessing the multispectral surface reflectance data;

[0045] Analyzing and obtaining a plurality of characteristic bands corresponding to the first mineral information from the preprocessed multispectral surface reflectance data;

[0046] Based on the multiple characteristic bands, performing data dimension reduction on the reflectivity data corresponding to the multiple characteristic bands to obtain surface reflectivity data after dimension reduction;

[0047] For the surface reflectivity data after dimensionality reduction, a preset Spectrum-Area model is used to remove background information in the surface reflectivity data after dimensionality reduction to obtain reflectivity data to be separated;

[0048] Density segmentation is performed on the reflectivity data to be separated to obtain the first mineral information.

[0049] Optionally, extracting the second mineral information from the multispectral surface reflectance data includes:

[0050] Information extraction is performed on the Fe-rich silicate mineral and the Fe-rich magnesia mineral based on the band ratio method.

[0051] Optionally, the determining of the target mineralization area based on the mineralization theory of the clay-type lithium ore, the structural interpretation mark and the ore-controlling elements includes:

[0052] According to the mineralization theory of the clay-type lithium ore, the structural interpretation marks and the ore-controlling elements, combined with geological characteristics and geochemical information, the initial mineralization area of ​​the clay-type lithium ore is determined; the geological characteristics and the geochemical information are obtained based on the analysis of the geological data;

[0053] obtaining field samples collected from the initial mineralization area;

[0054] Conducting mineral identification and geochemical analysis on the field samples to determine a remote sensing prospecting model for clay-type lithium deposits;

[0055] Based on the clay-type lithium ore remote sensing prospecting model, the clay-type lithium ore mineralization theory, the geological characteristics, the geochemical information and the ore-controlling factors, GIS technology is used to perform multivariate information superposition and analysis to determine the target mineralization area.

[0056] Compared with the prior art, the present invention provides a method for prospecting clay-type lithium ore. After obtaining relevant information, remote sensing image data and determining the mineralization theory of clay-type lithium ore, the multispectral surface reflectance data and the panchromatic band surface reflectance data are fused to obtain the target high-resolution multispectral surface reflectance, providing high-quality data for the subsequent establishment of structural interpretation marks, reflecting the characteristics of clay-type geology in the preset area, and conducting in-depth and precise structural interpretation of the preset area. In combination with the unique ore-controlling elements (represented by chlorite) in clay-type lithium ore, the target mineralization area is finally determined, and a more accurate lithium mining area is obtained to improve the accuracy of lithium ore exploration.

[0057] In a second aspect, the present invention further provides an exploration device for clay-type lithium ore, comprising:

[0058] An acquisition module is used to acquire geological data, mineral data and remote sensing image data of a preset area; the remote sensing image data includes optical surface reflectance data and a discretized mathematical expression model of surface topography; the optical surface reflectance data includes at least multispectral surface reflectance data and panchromatic band surface reflectance data;

[0059] An analysis module, used for analyzing the geological data and the mineral data to determine the mineralization theory of clay-type lithium ore;

[0060] A fusion module, for fusing the multispectral surface reflectance data and the panchromatic band surface reflectance data based on a preset orthogonalization algorithm and a preset inverse orthogonalization algorithm to obtain target high-resolution multispectral surface reflectance data;

[0061] Establishing a module for performing visualization operations and establishing structural interpretation marks based on the target high-resolution multispectral surface reflectance data and the discretized mathematical expression model of the surface topography; the structural interpretation marks are used to interpret the linear structure, annular structure and fold structure of the preset area in combination with the mineralization theory of the clay-type lithium ore;

[0062] An extraction module, used to extract ore-controlling elements from the optical surface reflectance data; the ore-controlling elements at least include chlorite;

[0063] A determination module is used to determine the target mineralization area based on the mineralization theory of the clay-type lithium ore, the structural interpretation marks and the ore-controlling factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0065] Figure 1 A schematic flow chart of a method for exploring clay-type lithium ore provided in accordance with an embodiment of the present invention;

[0066] Figure 2 A schematic diagram of the structure of an exploration device for clay-type lithium ore provided in one embodiment of the present invention. DETAILED DESCRIPTION

[0067] In order to clearly describe the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, words such as "first" and "second" are used to distinguish the same items or similar items with basically the same functions and effects. For example, the first threshold and the second threshold are only used to distinguish different thresholds, and their order is not limited. Those skilled in the art can understand that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit them to be different.

[0068] It should be noted that, in the present invention, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0069] In the present invention, "at least one" means one or more, "more than one" means two or more. "And / or" describes the association relationship of the associated objects, indicating that three types of relationships may exist.

[0070] like Figure 1 As shown, an embodiment of the present invention provides a method for exploring clay-type lithium ore, which may include:

[0071] Step 110: Acquire geological data, mineral data and remote sensing image data of a preset area; the remote sensing image data includes optical surface reflectance data and a discrete mathematical expression model of surface topography; the optical surface reflectance data includes at least multispectral surface reflectance data and hyperspectral surface reflectance data;

[0072] In specific implementation, different remote sensors on different satellites are used to collect images of the ground to obtain various remote sensing image data.

[0073] For example, remote sensing image data may include ALOS 12.5-meter DEM (Digital Elevation Model) data, Landsat-8 data, ASTER, and ZY1-02D ​​data.

[0074] The ALOS (Advanced Land Observing Satellite) satellite was launched on January 24, 2006. The 12.5-meter DEM data was collected by the PALSAR sensor of ALOS. The horizontal and vertical accuracy of the ALOS 12.5-meter DEM data can reach 12 meters. The embodiment of the present invention obtains 17 scenes of ALOS 12.5-meter DEM data, for example, and mosaicks and cuts the DEM data according to the range of the preset area.

[0075] DEM (Digital Elevation Model) is a digital model that represents ground elevation. In the embodiment of the present invention, the discretized mathematical expression model of the surface topography is the digital elevation model DEM. The digital elevation model DEM expresses the surface topography through a discretization method, that is, converting continuous surface elevation data into a gridded numerical representation.

[0076] For example, the digital elevation model DEM can be expressed as formula (1) as follows:

[0077] (1)

[0078] and is the plane coordinate, for( ) corresponding elevation.

[0079] Landsat-8 is the eighth satellite of the Landsat program, which carries the Land Imager (OLI) and Thermal Infrared Sensor (TIRS), including 8 multispectral bands with a spatial resolution of 30m, 1 panchromatic band with a spatial resolution of 15m, and 2 thermal infrared bands with a spatial resolution of 100m.

[0080] The Landsat-8 data preprocessing process includes radiometric calibration, atmospheric correction, image mosaicking and cropping.

[0081] ASTER is a multispectral sensor that integrates near-infrared, mid-infrared and thermal infrared on board the Terra satellite. It has a total of 14 spectral bands (Table 1) and a spatial resolution ranging from 15m to 90m.

[0082] The preprocessing of ASTER multispectral data mainly includes radiation calibration, atmospheric correction, band combination, image mosaicking and cropping.

[0083] The ZY1-02D ​​satellite, also known as the 5m optical satellite, operates in a sun-synchronous orbit and can obtain, for example, 60km wide hyperspectral data of 166 spectral bands. ZY1-02D ​​hyperspectral data has the advantages of high spatial spectral resolution and a large number of bands, which is more conducive to mineral extraction.

[0084] Step 120: Analyze geological data and mineral data to determine the mineralization theory of clay-type lithium ore;

[0085] In step 120, after obtaining the geological data and mineral data, the geological data and mineral data are analyzed to obtain important information such as the ore-controlling factors, geological characteristics, geochemical information, etc. of clay-type lithium deposits, and the mineralization conditions are summarized to facilitate the establishment of a mineralization theory for clay-type lithium deposits.

[0086] Specifically, step 120 further includes: analyzing the material source of the lithium-rich clay layer of the clay-type lithium ore, the occurrence characteristics and distribution law of lithium in clay rocks, the enrichment mechanism and mineralization process of lithium to determine the clay minerals (i.e., ore-controlling factors) that the lithium element in the clay-type lithium ore may be adsorbed on, geological characteristics, and geochemical information.

[0087] For example, the ore-controlling elements include at least minerals such as montmorillonite, lithium chlorite, chlorite, kaolinite, illite and dolomite, as well as Fe (iron)-rich iron silicate minerals and iron-rich magnesium minerals.

[0088] Step 130: Based on a preset orthogonalization algorithm and a preset inverse orthogonalization algorithm, the multispectral surface reflectance data and the panchromatic band surface reflectance data are fused to obtain target high-resolution multispectral surface reflectance data;

[0089] Assume that the multispectral surface reflectance data is , where N is the number of bands of the multispectral surface reflectance data. That is, the multispectral surface reflectance data includes reflectance data of N bands; is the reflectance data of the i-th band (such as red, green, blue, etc.) in the multispectral surface reflectance data.

[0090] Before performing step 130, the multispectral surface reflectance data may be preprocessed, and the preprocessing process at least includes radiation calibration, atmospheric correction, image mosaicking and cropping, etc.

[0091] Step 130 may specifically include:

[0092] 1) Based on the multispectral surface reflectance data and the preset panchromatic band spectral response curve, the panchromatic simulated surface reflectance data is constructed;

[0093] Step 1) may specifically include:

[0094] Selecting surface reflectance data of multiple target bands from multispectral surface reflectance data;

[0095] Determine the weight coefficient of each target band from the preset panchromatic band spectral response curve;

[0096] Based on the weight coefficient of each target band, the formula is used:

[0097] (2)

[0098] The surface reflectance data of multiple target bands are weighted and summed, and the summation result is determined as the full-color simulated surface reflectance data. Simulates the surface reflectance data for full color.

[0099] For example, the panchromatic band (band 8) of Landsat8 usually covers 0.50-0.68um, containing mixed information in the visible and near-infrared regions. By linearly combining multispectral bands (such as R, G, B, and NIR) using the above formula (2), an image with a spectral response similar to that of the PAN band can be generated, which is called a panchromatic simulated image (i.e., panchromatic simulated surface reflectance data). , ; is the i-th band (such as red, green, blue, etc.) of the multispectral image (i.e., multispectral surface reflectance data); is the weight coefficient, which is usually set according to the spectral response curve of the PAN band (for example, the weights of Landsat8 PAN corresponding to R, G, B, and NIR are approximately 0.3, 0.3, 0.2, and 0.2.

[0100] 2) Based on the preset orthogonalization algorithm, the panchromatic simulated surface reflectance data and the multispectral surface reflectance data are orthogonally transformed to obtain the target orthogonal component space;

[0101] First, the panchromatic simulation of surface reflectance data as the first orthogonal component.

[0102] Secondly, for each MS i (i=2,…,n) Apply the orthogonalization formula (3), that is, substitute the reflectance data of each band from the second band to the Nth band in the reflectance data of the N bands into the preset orthogonalization formula:

[0103] (3)

[0104] Each orthogonalized component of the reflectance data of each band is calculated; where, is the reflectance data of the i-th band in the multispectral surface reflectance data; is the i-th orthogonalized component corresponding to the reflectivity data of the i-th band; i is greater than 1 and i is less than N; is the jth orthogonalized component corresponding to the reflectance data of the jth band; i represents the band currently being processed; j represents the orthogonalized band; express and The inner product between express The square of the modulus.

[0105] Through this process, ensure that each newly generated band With all previous They are all orthogonal, that is, the bands are uncorrelated with each other, thus removing redundant information.

[0106] A target orthogonal component space is determined based on the N orthogonalized components.

[0107] 3) Based on the panchromatic band surface reflectance data, the preset inverse orthogonal algorithm is used to process the target orthogonal component space to obtain the initial high-resolution multispectral surface reflectance data;

[0108] Step 3) may specifically include:

[0109] The first step is to use the formula:

[0110] (4)

[0111] The panchromatic band surface reflectance data is determined as the first target orthogonalized component in the target orthogonal component space; wherein, is the first objective orthogonalized component, It is the surface reflectance data of the panchromatic band;

[0112] The first step is to understand that the target orthogonal component space includes N multispectral bands. (normalized and matched to adjust its range to Replace the first orthogonal component of the above target orthogonal component space It becomes the new first orthogonal component (that is, the first target orthogonalized component), and then the remaining multispectral bands in the target orthogonal component space (starting from the second band) are gradually orthogonalized to make them mutually orthogonal (no correlation) with the previous components. For example, the following formula (4) is used for the i-th band.

[0113] In the second step, the reflectance data of each band from the reflectance data of the second band to the reflectance data of the Nth band in the target orthogonal component space is substituted into the preset inverse orthogonalization formula (4):

[0114] (5)

[0115] Obtain initial high-resolution multispectral surface reflectance data.

[0116] From the above content, we can see that formula (5) restores the orthogonal component (GS') back to the original multispectral band space through inverse transformation, and uses the linear combination of the orthogonal components to approximately restore the original image. The first orthogonal component The remaining components provide the spectral information of the multispectral image. Through the weighted linear combination of these components, the fused multispectral bands, i.e. the initial high-resolution multispectral surface reflectance data, can be reconstructed.

[0117] 4) Feature enhancement is performed on the initial high-resolution multispectral surface reflectance data to obtain the target high-resolution multispectral surface reflectance data.

[0118] For example, the data feature and information enhancement method is used to enhance the features of the second surface reflectivity data.

[0119] From the above content, it can be seen that in the embodiment of the present invention, the preset orthogonalization algorithm and the preset inverse orthogonalization algorithm are used to fuse the reflectance data of the multispectral band and the reflectance data of the panchromatic band of the remote sensing image, and the feature enhancement is performed after the fusion. Such a processing process has the following technical effects: 1. The spectral characteristics of the multispectral image are effectively retained to ensure that the spectral information is not lost while the spatial resolution is improved; 2. Through the orthogonalization process, the correlation between the bands is reduced and the independence between the bands is improved, so that the fusion is more effective and the aliasing effect is reduced; 3. The orthogonalization method is relatively simple to calculate and the implementation is relatively efficient, which is suitable for large-scale data processing; 4. The main component to be replaced can be selected as needed, so as to flexibly control the spectral and spatial characteristics of the fusion result; 5. The fused image usually has better visual effects and richer details, which is suitable for visual analysis and decision support; 6. Since the main features of the multispectral data are retained, the image distortion can be effectively reduced and more realistic surface information can be provided; In general, the orthogonalization method has significant advantages in improving image quality; It provides a multi-source source of signs for the establishment of subsequent structural interpretation signs, which can improve the accuracy of the interpretation signs.

[0120] Step 140: Based on the target high-resolution multispectral surface reflectance data and the discretized mathematical expression model of the surface topography, a visualization operation is performed and a structural interpretation mark is established; the structural interpretation mark is used to interpret the linear structure, annular structure and fold structure of the preset area in combination with the mineralization theory of clay-type lithium deposits;

[0121] For example, ArcGIS Pro (ArcGIS Pro is a powerful desktop geographic information system software developed by Esri Environmental Systems Research Institute) can be used to perform three-dimensional visualization operations based on the target high-resolution multispectral surface reflectance data and the discrete mathematical expression model of the surface topography and landforms to establish structural interpretation signs.

[0122] It should be noted that linear structure refers to the geological structure pattern of a preset area, and becomes the fracture structure information of ore guidance, ore control, mineralization and ore-bearing space. In remote sensing images, linear structure is usually indirectly displayed by controlling local lithology, lithofacies, landforms and river characteristics, showing different tones, forms and landforms. Therefore, the establishment of structural interpretation marks is mainly achieved from the aspects of tones, forms and landform development.

[0123] Annular structures are mainly divided into tectonic rings related to tectonic genesis, hydrothermal rings related to hydrothermal activity, and magmatic rings related to magmatic hydrothermal intrusion.

[0124] The fold structure is manifested in the remote sensing image as the fold form of layered stratum strips or stripes. The color tone, terrain, and shadow strips on the image are closed or semi-closed circles, ellipses, arcs, strips, and have obvious symmetrical figures; the rock triangular faces, single-sided mountains, pig back ridges and other structural landforms are symmetrically repeated along a certain interface; the rock strips at the turning end are curved, and the rock triangular faces are regularly deflected into arc-shaped geometric shapes.

[0125] For example, the linear structure, annular structure and fold structure of the preset area can be interpreted based on the established structural interpretation symbols and the geological map of the preset area.

[0126] Step 150: extracting ore-controlling elements from the optical surface reflectance data; the ore-controlling elements at least include chlorite; wherein the optical surface reflectance data also includes hyperspectral surface reflectance data;

[0127] Chlorite is a representative ore-controlling factor that adsorbs lithium in clay-type lithium ores.

[0128] Step 150 may specifically include: extracting first mineral information and second mineral information from multispectral surface reflectance data, and extracting mineral anomaly information from hyperspectral surface reflectance data; wherein the first mineral information includes at least hydroxyl, iron dye and carbonate; the second mineral information includes at least Fe-rich silicate minerals and iron-rich magnesium minerals; the mineral anomaly information includes at least lithium chlorite, montmorillonite, illite, chlorite, kaolinite and dolomite.

[0129] It can be understood that mineral anomaly information refers to information reflected by the situation where the concentration of certain elements or minerals is significantly higher than the background value.

[0130] The first mineral information is extracted from the multispectral surface reflectance data, which may specifically include:

[0131] Step 1: Preprocess the multispectral surface reflectance data;

[0132] In the first step, a relatively interference-free area of ​​multispectral surface reflectance data such as vegetation cover, shadow, cloud and snow is selected. For characteristic minerals of different alteration types, according to the characteristics of preset mineral reflectance spectrum curves, the preset mineral reflectance spectrum is resampled to the multispectral surface reflectance data spectrum and a mineral reflectance spectrum curve based on the multispectral surface reflectance data is generated.

[0133] Step 2: Analyze and obtain multiple characteristic bands corresponding to the first mineral information from the pre-processed multispectral surface reflectance data;

[0134] Specifically, the preprocessed multispectral surface reflectance data is a mineral reflectance spectrum curve based on the multispectral surface reflectance data.

[0135] From the mineral reflection spectrum curve, multiple characteristic bands corresponding to each mineral information in hydroxyl, iron-stained and carbonated minerals and characteristic spectra corresponding to multiple characteristic bands are analyzed;

[0136] For example, taking the extraction of hydroxyl groups from ASTER data as an example: in ASTER bands 1 to 4, except for halloysite, the reflectance of other minerals increases with the increase of wavelength and reaches a peak in band 4; the reflectance of halloysite reaches a peak in band 3 and decreases in band 4; in band 6, most OH - The minerals initially show an absorption valley; therefore, it is believed that most OH can be extracted from bands 1, 3, 4, and 6. - Group mineral information. OH - The characteristics of the mineral anomaly component should be: in the ASTER data, the contributions of bands 1, 3, and 4 are opposite to those of band 6, and band 6 has a high loading.

[0137] Step 3: Based on multiple characteristic bands, the reflectivity data corresponding to the multiple characteristic bands are subjected to data dimension reduction to obtain the surface reflectivity data after dimension reduction;

[0138] The purpose of dimensionality reduction of reflectance data corresponding to multiple characteristic bands is to recombine the information of each band so that the mineral anomaly information exists only in one of the components.

[0139] The dimensionality reduction process in the third step refers to the following multiple sub-steps:

[0140] Suppose a data set (original matrix) X contains n samples, each sample contains p features. Therefore, the data set can be represented as an n×p matrix X.

[0141] First sub-step: data standardization.

[0142] In order to eliminate the dimensional differences between different feature dimensions, the data needs to be standardized. A common method is to return the mean of each feature to zero and normalize the standard deviation. Standardization can eliminate dimensional differences and allow different features to be compared at the same level.

[0143] For each feature j, calculate the mean and standard deviation , and then normalize it using formula (6):

[0144] (6)

[0145] Among them, after standardization, the matrix Z is obtained, and the mean of each column is 0 and the variance is 1. i represents the index of the i-th sample, and j represents the index of the j-th feature. Therefore, Z ij Represents the standardized value of the i-th sample in the matrix Z on the j-th feature; is the value of the data set (that is, the original matrix); is the mean of the jth feature; is the standard deviation of the jth feature.

[0146] Second sub-step: Calculate the covariance matrix

[0147] The covariance matrix represents the linear correlation between the various features of the data; for the standardized data, the covariance matrix C is calculated, and its formula is:

[0148] (7)

[0149] in, is the transpose of Z. The covariance matrix C is a p×p symmetric matrix, where the element C in the i-th row and j-th column is ij Represents the covariance between the i-th feature and the j-th feature;

[0150] The third sub-step: Calculate the eigenvalues ​​and eigenvectors of the covariance matrix.

[0151] The covariance matrix is ​​symmetric and can be decomposed into a set of eigenvalues ​​and corresponding eigenvectors; the eigenvalues ​​represent the variance of the data in the direction of the corresponding eigenvectors, and the eigenvectors represent the new orthogonal basis directions;

[0152] The covariance matrix C is a symmetric matrix, so it can be decomposed into eigenvalues, namely:

[0153] (8)

[0154] Among them, λ is the eigenvalue, and v is the eigenvector corresponding to the eigenvalue. The eigenvector v represents the direction of the principal component of the data, and the eigenvalue λ represents the variance of the data in this direction. Usually, the first few principal components with larger eigenvalues ​​are selected to retain the data information to the maximum extent;

[0155] Fourth sub-step: Selecting the main components

[0156] According to the order of the eigenvalues, the eigenvectors corresponding to the characteristic bands in the second sub-step are selected as the main components;

[0157] Fifth sub-step: Representation of surface reflectivity data after dimensionality reduction

[0158] The original data X is projected onto the selected principal component direction to obtain the reduced-dimensional surface reflectance data Y. The projection formula is:

[0159] (9)

[0160] Among them, Y is an n×k matrix, that is, the data after dimensionality reduction, each row represents the coordinates of a sample in the new k-dimensional space; W is the preset dimensionality reduction matrix.

[0161] Step 4: For the surface reflectivity data after dimensionality reduction, the preset Spectrum-Area model is used to remove the background information in the surface reflectivity data after dimensionality reduction to obtain the reflectivity data to be separated; and the reflectivity data to be separated is density segmented to obtain the first mineral information.

[0162] It should be noted that in the frequency domain, geophysical and geochemical anomalies have self-similarity different from the background. This self-similarity is presented as a power law relationship, and the calculation formula is as follows:

[0163] (10)

[0164] In the formula, S represents the energy spectrum density in the Fourier energy spectrum space, A(≥S) represents the area in the Fourier energy spectrum space where the energy spectrum density is greater than S. The index representing the fractal model;

[0165] Then, the fourth step can specifically include:

[0166] Perform Fourier transform on the reduced-dimensional surface reflectivity data, so that the reduced-dimensional surface reflectivity data can be converted from spatial domain data to frequency domain spatial data; the frequency is regarded as the energy spectrum density S, and the corresponding number of pixels is regarded as the area A, and log-log transform is performed to design a suitable filter (threshold), and then the mineral anomaly information is separated from the complex background information through inverse Fourier transform;

[0167] Among them, the DFT (Discrete Fourier Transform) forward transform formula is as follows: Given a discrete sequence x[n] of length N, its DFT formula is:

[0168] (11)

[0169] IDFT (Inverse Discrete Fourier Transform) formula:

[0170] (12)

[0171] X[k] is the representation of the signal in the frequency domain, x[n] is the representation of the signal in the time domain, N is the number of sampling points of the signal, and j is the imaginary unit.

[0172] Finally, the spatial data of separated geophysical anomalies, geochemical anomalies and background information are obtained through density segmentation method to obtain remote sensing information such as hydroxyl, iron staining, and carbonate alteration.

[0173] The above is the process of extracting the first mineral information such as hydroxyl, iron staining, and carbonation.

[0174] Next, the extraction of the second mineral information (Fe-rich silicate mineral and ferromagnesian mineral) will be described.

[0175] Extracting the second mineral information from the multispectral surface reflectance data may specifically include: extracting information on Fe-rich silicate minerals and Fe-rich magnesia minerals based on a band ratio method.

[0176] It should be noted that the principle of the band ratio method is as follows:

[0177] The band ratio method is based on the principle of algebraic operation. When the difference between bands is similar but the slopes are different, the ratio of the reflection band to the absorption band is used to enhance the spectral difference between various rock types, suppress the influence of terrain, and display the dynamic range.

[0178] Analyze the spectral curve of the mineral to find the interval with the largest slope change and the reflection peaks and absorption valleys in the curve, determine the spectral range, perform ratio enhancement processing, and form an image that highlights the mineral information.

[0179] The mineral spectrum sensitivity analysis is as follows:

[0180] For example, there are three bands in the ASTER data in the visible-near infrared (VNIR) wavelength range. The absorption in these bands is mainly caused by electronic transition processes such as color centers, conduction bands, lattice field effects or valence bond migration.

[0181] Fe2+ ​​ions in the silicate lattice framework often cause low reflectivity in the spectral range of 0.6-1.2μm. The ASTER band ratio band1 / band2 is sensitive to ferrous absorption, and the ratio band5 / band3 represents the steep slope caused by ferrous absorption in the VNIR wavelength region and the general lack of Al-OH absorption in the band5 part of the short infrared SWIR band. Therefore, the ASTER band ratio (band1 / band2+band5 / band3) can be mainly used to interpret the rocks of Fe-rich silicate minerals exposed on the surface. band refers to the band.

[0182] In addition, practical applications show that the band ratio (band1 / band2+band5 / band3) is particularly effective in highlighting Mg-OH-containing iron silicate minerals, such as chlorite and epidote.

[0183] For rocks mainly composed of silicate minerals, as the SiO2 content gradually decreases, the reflectivity valley will gradually migrate toward the long-wave direction, which will cause the emissivity of band 12 to gradually increase relative to the emissivity of band 13; therefore, the band ratio band12 / band13 shows an inverse trend with the SiO2 content to a certain extent, and is thus used to measure the content of iron-rich magnesium minerals in surface rocks.

[0184] Next, the extraction of mineral anomaly information (such as hectorite, montmorillonite, illite, chlorite, kaolinite, and dolomite) is described.

[0185] Extract mineral anomaly information from hyperspectral surface reflectance data, including:

[0186] Image pixel spectral vector extraction steps: extract pixel spectral vector from hyperspectral surface reflectance data.

[0187] Optionally, when extracting the pixel spectral vector, the hyperspectral surface reflectance data may be preprocessed in a certain process, and then the pixel spectral vector may be extracted from the preprocessed hyperspectral surface reflectance data.

[0188] For example, the hyperspectral surface reflectance data may be ZY1-02D ​​data.

[0189] Selection step: based on the spectral characteristics of the mineral anomaly information, a reference sample spectrum corresponding to the image pixel spectrum vector is selected from a preset reference spectrum library, and a reference sample spectrum vector is extracted from the reference sample spectrum;

[0190] Calculation steps: Substitute the image pixel spectral vector and the reference sample spectral vector into the formula:

[0191] (13)

[0192] The similarity value between the image pixel spectral vector and the reference sample spectral vector is calculated; where, is the spectral vector of the i-th image pixel; is the spectral vector of the i-th reference sample; The numerical value representing the similarity, Representation vector and vector The inner product of Representation vector Model; Representation vector Model.

[0193] Mineral anomaly information extraction step: If the similarity value is less than or equal to the preset angle threshold, the mineral anomaly information is extracted.

[0194] It is understandable that the smaller the angle between the two, the higher the similarity between the characterizing pixel spectrum and the reference sample spectrum. Therefore, when the angle difference is less than or equal to the preset angle threshold, the similarity between the characterizing pixel spectrum and the reference sample spectrum meets the set requirement.

[0195] From the above content, it can be seen that based on the reference sample spectrum, if the similarity between the image pixel spectrum and the reference sample spectrum meets the set requirements, the mineral anomaly information corresponding to the reference sample spectrum can be finally extracted from the hyperspectral surface reflectance data.

[0196] In the process of extracting mineral anomaly information, the above extraction method regards the spectrum of a single image element as a vector in n-dimensional space. The spectrum of N bands is used to represent the vector in N-dimensional space. The similarity between the pixel spectrum and the reference sample spectrum is determined by calculating the vector angle between the two. The smaller the angle between the two, the higher the similarity. This improves the extraction accuracy of mineral anomaly information.

[0197] It is understandable that all the above mineral information extraction processes can be extracted using different data sources. For example, the results of ASTER data and ZY1-02D ​​data extraction can be verified with each other, and then verified with geological map lithology information to ensure the accuracy of mineral information extraction.

[0198] Step 160: Determine the target mineralization area based on the mineralization theory, structural interpretation signs and ore-controlling factors of clay-type lithium deposits.

[0199] Step 160 may specifically include:

[0200] According to the mineralization theory, structural interpretation signs and ore-controlling factors of clay-type lithium deposits, combined with geological characteristics and geochemical information, the initial mineralization area of ​​clay-type lithium deposits is determined; both geological characteristics and geochemical information can be obtained based on the analysis of geological data;

[0201] Obtain field samples from the initial mineralization area;

[0202] Conduct mineral identification and geochemical analysis on field samples to determine the remote sensing prospecting model for clay-type lithium deposits;

[0203] Based on the remote sensing prospecting model of clay-type lithium deposits, the mineralization theory, geological characteristics, geochemical information and ore-controlling factors of clay-type lithium deposits, GIS technology is used to superimpose and analyze multivariate information to determine the target mineralization area.

[0204] It is understandable that determining the target mineralization area is actually to delineate the favorable areas for clay-type lithium deposits. There are some delineation principles that can be used as reference when delineating favorable areas for clay-type lithium deposits, such as:

[0205] ① Strata of the xx (xx represents a place name) group and the yy (yy represents a place name) group in the preset area;

[0206] ② The arc structure and anticline structure area are developed;

[0207] ③ Mainly located in the high-value extraction areas of Fe-rich ferrosilicate minerals and Fe-rich magnesia minerals, and the high-value areas of Al-OH and Mg-OH also have an indicative effect;

[0208] ④ The mineral information of montmorillonite, lithium chlorite, chlorite, kaolinite, illite and dolomite all have an indicative effect on lithium ore, and the mineral information of montmorillonite and lithium chlorite has a stronger indicative effect;

[0209] ⑤Located in the high-value lithium geochemical area.

[0210] In a specific embodiment, multi-source remote sensing image data, such as ALOS12.5mDEM data and Landsat8 remote sensing images, are combined to perform linear structure interpretation. For example, Fe-rich silicate minerals, iron-rich magnesium minerals, lithium chlorite, montmorillonite, illite, chlorite, kaolinite and dolomite are successfully extracted using ASTER multispectral data and ZY-02D hyperspectral data.

[0211] Combined with the characteristic analysis of geological data and mineral data, 10 prospecting target areas were finally determined. The research results show that the clay-type lithium deposits in Wenshan area are controlled by the EW arc structure between the Wenshan-Malipo Fault and the Guangnan-Funing Fault, and the main mineral-rich strata include the Wujiaping Formation (P3w) and the Longtan Formation (P3l). The high-value areas rich in Fe silicate minerals and ferromagnesian minerals extracted by ASTER data may indicate the enrichment of clay-type lithium deposits, while montmorillonite, lithium chlorite, chlorite, kaolinite, illite and dolomite have important indicative effects on the existence of clay-type lithium deposits, especially montmorillonite and lithium chlorite show strong indicative characteristics.

[0212] From the above content, it can be seen that in the exploration method for clay-type lithium ore provided in the embodiment of the present invention, after obtaining the geological data, mineral data and remote sensing image data of the preset area, first, the preset orthogonalization algorithm and the preset inverse orthogonalization algorithm are used to fuse the reflectance data of the multispectral band of the remote sensing image and the reflectance data of the panchromatic band, and the features are enhanced after fusion. Such a processing process not only retains the texture features and spectral features of the remote sensing image, but also improves the spatial resolution, greatly enhances the accuracy of the data, and provides a multi-source source of marks for the establishment of subsequent structural interpretation marks, which can improve the accuracy of the interpretation marks; second, when extracting the first mineral information such as hydroxyl, iron staining, and carbonate minerals, the spectra of the multiple characteristic bands are reduced in dimension according to the multiple characteristic bands of each mineral, and the preset Spectrum-Area model is used to remove the dimensionality after the dimensionality reduction. The background information in the surface reflectivity data after dimensionality reduction is used to obtain the reflectivity data to be separated; and the reflectivity data to be separated is density segmented to obtain the first mineral information. This extraction method, because it extracts according to the characteristics of each mineral and adopts the Spectrum-Area model, the results proposed can be more accurate; thirdly, when extracting abnormal information of minerals such as lithium chlorite, montmorillonite, illite, chlorite, kaolinite and dolomite, the pixel spectral vector is extracted from the hyperspectral surface reflectivity data, based on the spectral characteristics of the mineral abnormality information, the corresponding reference sample spectrum is selected from the preset reference spectral library and the vector of the reference sample spectrum is extracted, and then the vector of the pixel spectrum and the vector of the reference sample spectrum are substituted into the designed formula to calculate the similarity between the two, which can improve the similarity of data extraction, thereby improving the accuracy of extracting mineral abnormality information. In general, through more accurate extraction of various data and mineral information, a more accurate remote sensing prospecting model for clay-type lithium deposits can be obtained, which can make the final delineation of favorable areas for clay-type lithium deposits more accurate and improve the accuracy of exploration of clay-type lithium deposits.

[0213] The exploration device for clay-type lithium ore provided by the present invention is described below. The exploration device for clay-type lithium ore described below and the exploration method for clay-type lithium ore described above can be referred to each other.

[0214] like Figure 2 As shown, an embodiment of the present invention further provides an exploration device for clay-type lithium ore, which is used to implement the exploration method for clay-type lithium ore in any of the above embodiments. The exploration device for clay-type lithium ore may include:

[0215] The acquisition module 210 is used to acquire geological data, mineral data and remote sensing image data of a preset area; the remote sensing image data includes optical surface reflectance data and a discrete mathematical expression model of surface topography; the optical surface reflectance data includes at least multispectral surface reflectance data and panchromatic band surface reflectance data;

[0216] Analysis module 220, used to analyze geological data and mineral data to determine the mineralization theory of clay-type lithium ore;

[0217] A fusion module 230 is used to fuse the multispectral surface reflectance data and the panchromatic band surface reflectance data based on a preset orthogonalization algorithm and a preset inverse orthogonalization algorithm to obtain target high-resolution multispectral surface reflectance data;

[0218] Establishing module 240, for performing visualization operations and establishing structural interpretation marks based on target high-resolution multi-spectral surface reflectance data and a discrete mathematical expression model of surface topography; the structural interpretation marks are used to interpret linear structures, annular structures and fold structures in a preset area in combination with the mineralization theory of clay-type lithium deposits;

[0219] An extraction module 250 is used to extract ore-controlling elements from the optical surface reflectance data; the ore-controlling elements at least include chlorite;

[0220] The determination module 260 is used to determine the target mineralization area based on the mineralization theory, structural interpretation signs and ore-controlling factors of clay-type lithium deposits.

[0221] On the other hand, the present invention also provides a non-transitory computer-readable storage medium, in which instructions are stored. When the instructions are executed, the method for exploring clay-type lithium ore in any of the above embodiments is implemented.

[0222] Although the present invention is described herein in conjunction with various embodiments, in the process of implementing the claimed invention, those skilled in the art may understand and implement other variations of the disclosed embodiments by viewing the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other components or steps, and "one" or "an" does not exclude multiple situations. A single processor or other unit may implement several functions listed in the claims. Certain measures are recorded in different dependent claims, but this does not mean that these measures cannot be combined to produce good results.

[0223] Although the present invention has been described in conjunction with specific features and embodiments thereof, it is apparent that various modifications and combinations may be made thereto without departing from the spirit and scope of the present invention. Accordingly, this specification and the accompanying drawings are merely exemplary illustrations of the present invention as defined by the appended claims and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present invention. Obviously, those skilled in the art may make various modifications and variations to the present invention without departing from the spirit and scope of the present invention. Thus, the present invention is intended to include such modifications and variations if they fall within the scope of the claims of the present invention and their equivalents.

Claims

1. A method for exploring clay-type lithium ore, characterized in that: include: Acquire geological data, mineral data and remote sensing image data of a preset area; the remote sensing image data includes optical surface reflectance data and a discrete mathematical expression model of surface topography; the optical surface reflectance data includes at least multispectral surface reflectance data and panchromatic band surface reflectance data; Analyze the geological data and the mineral data to determine the mineralization theory of clay-type lithium ore; Based on a preset orthogonalization algorithm and a preset inverse orthogonalization algorithm, the multispectral surface reflectance data and the panchromatic band surface reflectance data are fused to obtain target high-resolution multispectral surface reflectance data; Based on the target high-resolution multi-spectral surface reflectance data and the discretized mathematical expression model of the surface topography, a visualization operation is performed and a structural interpretation mark is established; the structural interpretation mark is used to interpret the linear structure, annular structure and fold structure of the preset area in combination with the mineralization theory of the clay-type lithium ore; Extracting ore-controlling elements from the optical surface reflectance data; the ore-controlling elements at least include chlorite; Based on the mineralization theory of the clay-type lithium ore, the structural interpretation marks and the ore-controlling factors, determine the target mineralization area; The method of fusing the multispectral surface reflectance data and the panchromatic band surface reflectance data based on a preset orthogonalization algorithm and a preset inverse orthogonalization algorithm to obtain target high-resolution multispectral surface reflectance data includes: constructing panchromatic simulated surface reflectance data based on the multispectral surface reflectance data and a preset panchromatic band spectral response curve; orthogonally transforming the panchromatic simulated surface reflectance data and the multispectral surface reflectance data based on a preset orthogonalization algorithm to obtain a target orthogonal component space; processing the target orthogonal component space based on the panchromatic band surface reflectance data using a preset inverse orthogonalization algorithm to obtain initial high-resolution multispectral surface reflectance data; and performing feature enhancement on the initial high-resolution multispectral surface reflectance data to obtain the target high-resolution multispectral surface reflectance data; The multispectral surface reflectance data includes reflectance data of N bands; N is a positive integer greater than 2; the orthogonal transformation of the panchromatic simulated surface reflectance data and the multispectral surface reflectance data based on a preset orthogonalization algorithm is performed to obtain a target orthogonal component space, including: Determining the panchromatic simulated surface reflectance data as a first orthogonalized component; Substitute the reflectivity data of each band from the reflectivity data of the second band to the reflectivity data of the N bands into the preset orthogonalization formula: ； Each orthogonalized component of the reflectance data of each band is calculated; where, is the reflectance data of the i-th band in the multispectral surface reflectance data; is the i-th orthogonalized component corresponding to the reflectivity data of the i-th band; i is greater than 1 and i is less than N; is the jth orthogonalized component corresponding to the reflectance data of the jth band; i represents the band currently being processed; j represents the orthogonalized band; express and The inner product between express Its own inner product; determining the target orthogonal component space based on the N orthogonalized components; The method of processing the target orthogonal component space based on the panchromatic band surface reflectance data using a preset inverse orthogonalization algorithm to obtain initial high-resolution multispectral surface reflectance data includes: Using the formula: ； The panchromatic band surface reflectance data is determined as the first target orthogonalized component in the target orthogonal component space; wherein, is the first target orthogonalized component, is the panchromatic band surface reflectance data; Substitute the reflectance data of each band from the reflectance data of the 2nd band to the reflectance data of the Nth band in the target orthogonal component space into the preset inverse orthogonal formula: ； The initial high-resolution multispectral surface reflectance data is obtained; wherein, is the initial high-resolution multispectral surface reflectance data obtained after the reflectance data of the i-th band is processed by inverse intersection. is the i-th inverse orthogonalized component of the reflectance data of the i-th band in the target orthogonal component space, is the jth inverse orthogonalized component of the reflectance data of the jth band in the target orthogonal component space.

2. The method for exploring clay-type lithium ore according to claim 1, characterized in that: The optical surface reflectance data also includes hyperspectral surface reflectance data; The step of extracting ore-controlling elements from the optical surface reflectivity data comprises: The first mineral information and the second mineral information are extracted from the multispectral surface reflectance data, and the mineral anomaly information is extracted from the hyperspectral surface reflectance data; the first mineral information includes at least hydroxyl, iron dye and carbonate; the second mineral information includes at least Fe-rich silicate minerals and iron-rich magnesium minerals; the mineral anomaly information includes at least lithium chlorite, montmorillonite, illite, the chlorite, kaolinite and dolomite.

3. The method for exploring clay-type lithium ore according to claim 2, characterized in that: The extracting of mineral anomaly information from the hyperspectral surface reflectance data comprises: Extracting image pixel spectral vectors from the hyperspectral surface reflectance data; Based on the spectral characteristics of the mineral anomaly information, a reference sample spectrum corresponding to the image pixel spectrum vector is selected from a preset reference spectrum library, and a reference sample spectrum vector is extracted from the reference sample spectrum; Substitute the image pixel spectral vector and the reference sample spectral vector into the formula: ； The similarity value between the image pixel spectral vector and the reference sample spectral vector is calculated; wherein, is the spectral vector of the i-th image pixel; is the spectral vector of the i-th reference sample; The numerical value representing the similarity, Representation vector and vector The inner product of Representation vector Model; Representation vector Model; If the value of the similarity is less than or equal to a preset angle threshold, the mineral anomaly information is extracted.

4. The method for exploring clay-type lithium ore according to claim 2, characterized in that: The step of extracting first mineral information from the multispectral surface reflectance data comprises: Preprocessing the multispectral surface reflectance data; Analyzing and obtaining a plurality of characteristic bands corresponding to the first mineral information from the preprocessed multispectral surface reflectance data; Based on the multiple characteristic bands, performing data dimension reduction on the reflectivity data corresponding to the multiple characteristic bands to obtain surface reflectivity data after dimension reduction; For the surface reflectivity data after dimensionality reduction, a preset Spectrum-Area model is used to remove background information in the surface reflectivity data after dimensionality reduction to obtain reflectivity data to be separated; Density segmentation is performed on the reflectivity data to be separated to obtain the first mineral information.

5. The method for exploring clay-type lithium ore according to claim 2, characterized in that: The step of extracting second mineral information from the multispectral surface reflectance data comprises: Information extraction is performed on the Fe-rich silicate mineral and the Fe-rich magnesium mineral based on the band ratio method.

6. The method for exploring clay-type lithium ore according to claim 1, characterized in that: The target mineralization area is determined based on the mineralization theory of the clay-type lithium ore, the structural interpretation marks and the ore-controlling factors, including: According to the mineralization theory of the clay-type lithium ore, the structural interpretation marks and the ore-controlling elements, combined with geological characteristics and geochemical information, the initial mineralization area of ​​the clay-type lithium ore is determined; the geological characteristics and the geochemical information are obtained based on the analysis of the geological data; obtaining field samples collected from the initial mineralization area; Conducting mineral identification and geochemical analysis on the field samples to determine a remote sensing prospecting model for clay-type lithium deposits; Based on the clay-type lithium ore remote sensing prospecting model, the clay-type lithium ore mineralization theory, the geological characteristics, the geochemical information and the ore-controlling factors, GIS technology is used to perform multivariate information superposition and analysis to determine the target mineralization area.

7. An exploration device for clay-type lithium ore, characterized in that: include: An acquisition module is used to acquire geological data, mineral data and remote sensing image data of a preset area; the remote sensing image data includes optical surface reflectance data and a discretized mathematical expression model of surface topography; the optical surface reflectance data includes at least multispectral surface reflectance data and panchromatic band surface reflectance data; An analysis module, used for analyzing the geological data and the mineral data to determine the mineralization theory of clay-type lithium ore; A fusion module, for fusing the multispectral surface reflectance data and the panchromatic band surface reflectance data based on a preset orthogonalization algorithm and a preset inverse orthogonalization algorithm to obtain target high-resolution multispectral surface reflectance data; Establishing a module for performing visualization operations and establishing structural interpretation marks based on the target high-resolution multispectral surface reflectance data and the discretized mathematical expression model of the surface topography; the structural interpretation marks are used to interpret the linear structure, annular structure and fold structure of the preset area in combination with the mineralization theory of the clay-type lithium ore; An extraction module, used to extract ore-controlling elements from the optical surface reflectance data; the ore-controlling elements at least include chlorite; A determination module, for determining a target mineralization area based on the mineralization theory of the clay-type lithium ore, the structural interpretation mark and the ore-controlling elements; The method of fusing the multispectral surface reflectance data and the panchromatic band surface reflectance data based on a preset orthogonalization algorithm and a preset inverse orthogonalization algorithm to obtain target high-resolution multispectral surface reflectance data includes: constructing panchromatic simulated surface reflectance data based on the multispectral surface reflectance data and a preset panchromatic band spectral response curve; orthogonally transforming the panchromatic simulated surface reflectance data and the multispectral surface reflectance data based on a preset orthogonalization algorithm to obtain a target orthogonal component space; processing the target orthogonal component space based on the panchromatic band surface reflectance data using a preset inverse orthogonalization algorithm to obtain initial high-resolution multispectral surface reflectance data; and performing feature enhancement on the initial high-resolution multispectral surface reflectance data to obtain the target high-resolution multispectral surface reflectance data; The multispectral surface reflectance data includes reflectance data of N bands; N is a positive integer greater than 2; the orthogonal transformation of the panchromatic simulated surface reflectance data and the multispectral surface reflectance data based on a preset orthogonalization algorithm is performed to obtain a target orthogonal component space, including: Determining the panchromatic simulated surface reflectance data as a first orthogonalized component; Substitute the reflectivity data of each band from the reflectivity data of the second band to the reflectivity data of the N bands into the preset orthogonalization formula: ； Each orthogonalized component of the reflectance data of each band is calculated; where, is the reflectance data of the i-th band in the multispectral surface reflectance data; is the i-th orthogonalized component corresponding to the reflectivity data of the i-th band; i is greater than 1 and i is less than N; is the jth orthogonalized component corresponding to the reflectance data of the jth band; i represents the band currently being processed; j represents the orthogonalized band; express and The inner product between express Its own inner product; determining the target orthogonal component space based on the N orthogonalized components; The method of processing the target orthogonal component space based on the panchromatic band surface reflectance data using a preset inverse orthogonalization algorithm to obtain initial high-resolution multispectral surface reflectance data includes: Using the formula: ； The panchromatic band surface reflectance data is determined as the first target orthogonalized component in the target orthogonal component space; wherein, is the first target orthogonalized component, is the panchromatic band surface reflectance data; Substitute the reflectance data of each band from the reflectance data of the 2nd band to the reflectance data of the Nth band in the target orthogonal component space into the preset inverse orthogonal formula: ； The initial high-resolution multispectral surface reflectance data is obtained; wherein, is the initial high-resolution multispectral surface reflectance data obtained after the reflectance data of the i-th band is processed by inverse intersection. is the i-th inverse orthogonalized component of the reflectance data of the i-th band in the target orthogonal component space, is the jth inverse orthogonalized component of the reflectance data of the jth band in the target orthogonal component space.

Images