An UAV pegmatite spectrum acquisition device, a lithium-bearing pegmatite identification method and system

Abstract

The present application discloses a drone pegmatite spectral acquisition device, a lithium-bearing pegmatite identification method and system, which relate to the field of data acquisition and processing. Through reasonable design, a hyperspectral sensor is carried on the drone, which can collect hyperspectral data of pegmatite in real time through the drone, providing a data basis for the identification of lithium-bearing pegmatite. By processing the obtained drone hyperspectral data, reflectance spectral data, etc., a drone hyperspectral spatial vector cloud, a lithium-bearing pegmatite spectral vector cloud and a non-lithium-bearing pegmatite spectral vector cloud are obtained. Using these spectral vector clouds as the input of the constructed network model to obtain the identification result can improve the real-time performance of pegmatite identification. By adopting the spectral angle comparison method based on the identification result and retaining the image containing the information of lithium-bearing pegmatite, the accuracy of lithium-bearing pegmatite identification can be further improved, thereby solving the problem that the prior art cannot carry out the identification of lithium-bearing pegmatite for drone hyperspectral data.

Description

Technical Field

[0001] The present application relates to the field of data acquisition and processing, and particularly to a drone pegmatite spectrum acquisition device, a lithium-bearing pegmatite identification method and system. Background Art

[0002] Based on the electronic transitions of metal cations and the molecular vibrations caused by water, hydroxyl ions, carbonate and sulfate ions, etc., is the infrared spectrum basis for the hyperspectral research of drones (Hunt and Ashley, 1979; Yan Shouxun et al., 2003; Lian Yunzhang et al., 2005). In the identification of lithium minerals, infrared spectroscopy can be used to identify ore-bearing pegmatites and non-ore-bearing pegmatites (Wang Shanshan et al., 2023; Ren Guangli et al., 2022). Because the spectral diagnostic characteristics of clay minerals dominate those of lithium minerals, it is difficult to distinguish the spectra of clay minerals and lithium minerals (Cardoso-Fernandes et al., 2021). A large number of spectral measurement results of lithium-bearing minerals show that it is difficult to distinguish between lepidolite and muscovite in terms of spectra.

[0003] In recent years, satellite remote sensing has gradually been used to study the identification of pegmatite lithium deposits or brine lithium deposits (Cardoso-Fernandes et al., 2020). Satellite hyperspectral and multispectral data are used to extract lithium mineral information. For example, the ZY1E spectrum and GF2 are used to delineate lithium mineralization anomalies (Gao Yihang et al., 2023), neural networks are used to extract pegmatite information (Li Wanyue et al., 2023), and hyperspectral data are used to accurately extract lithium minerals (Booysen et al., 2022). The ASTER data is used to identify remote sensing anomalies of pegmatite lithium deposits and delineate target areas (Jin Moushun et al., 2019; Xu Xingwang et al., 2019; Cardoso-Fernandes et al., 2019). However, satellite multispectral and hyperspectral data with low spatial resolution rarely can directly detect small-sized geological bodies (Eskandari et al., 2023). Satellite high-resolution remote sensing provides high spatial resolution. For example, WorldView-3 has a maximum resolution of 0.3 meters and provides 8 visible and near-infrared (VNIR) bands and 8 short-wave infrared (SWIR) bands, which is a powerful tool for pegmatite identification (Worldview-3 Specifications, 2013; Kruse and Perry, 2012; Jin Moushun et al., 2019). However, it only provides several SWIR bands and cannot accurately identify lithium-bearing pegmatites. Previous studies in the Tashdaban lithium ore field using the WorldView-3 SWIR bands found that the spectral samples of the pegmatite veins and the surrounding wall rocks in the WorldView-3 SWIR images were basically the same, making it impossible to distinguish pegmatites from wall rocks.

[0004] Unmanned aerial vehicle (UAV) systems bridge the observational gap between airborne surveys and ground surveys. They can carry different sensors to obtain high-resolution spatial, spectral, and temporal data, which helps to achieve rapid and efficient breakthroughs in geological prospecting (Beyer et al., 2019; Heincke et al., 2019; Van der Meer et al., 2012). The spectral imaging of UAV systems has been successfully applied to rare earth exploration (Booysen et al., 2019). Jackisch et al. (2020) processed the UAV-borne photography using traditional methods and machine learning methods, combined with ground surveys, and carried out research in Finland. In the northern region of the apatite ore deposit, the ore-bearing rocks are distinguished from the waste rocks. Eskandari et al. (2023) used remote sensing and unmanned aerial vehicles (UAVs) for geological mapping to carry out the exploration of chromite. Martelet et al. (2021) used UAV multi-spectral for geological mapping and mineralization identification, etc. However, there are few studies on pegmatite lithium ore using UAVs. Among them, Bai et al. (2024) used an airborne Hyspex imaging spectrometer to directly map the pegmatite area in Jing'erquan, Xinjiang. However, UAV remote sensing is greatly affected by weather conditions and cannot identify lithium-bearing pegmatites for UAV hyperspectral. Summary of the Invention

[0005] To solve the above problems existing in the prior art, the present application provides a UAV pegmatite spectral acquisition device, a method and a system for identifying lithium-bearing pegmatites.

[0006] To achieve the above object, the present application provides the following solutions:

[0007] In a first aspect, the present application provides a UAV pegmatite spectral acquisition device, which includes: a UAV and a hyperspectral sensor;

[0008] An accommodation cavity is opened on the cabin body of the UAV; the accommodation cavity is used to install the hyperspectral sensor; at the connection between the hyperspectral sensor and the UAV, a shock absorption structure or a calibration structure is provided.

[0009] Optionally, the device further includes landing gears; the landing gears are installed on the UAV; when the UAV lands, the landing gears are used to form a gap between the hyperspectral sensor and the ground.

[0010] Optionally, a battery shielding film is provided in the accommodation cavity.

[0011] Optionally, the UAV is a vertical take-off and fixed-wing UAV; the hyperspectral sensor is a full-spectrum imager in the 400nm - 2500nm band.

[0012] In a second aspect, the present application provides a method for identifying lithium-bearing pegmatites, which includes:

[0013] Obtain reflectance spectral data, UAV hyperspectral data, target spectra, dark current data, attitude data, and ground station data; the UAV hyperspectral data is a spectral image collected by the UAV pegmatite spectral acquisition device provided above in this application; the dark current data is the current data collected before and after the UAV takes off and lands; the reflectance spectral data is the emissivity data of lithium-bearing pegmatite and the emissivity data of non-lithium-bearing pegmatite measured by an infrared spectrometer; the target spectra are obtained based on targets with different reflectivities; the targets with different reflectivities are set within a set range centered on the UAV takeoff position; the attitude data includes one or more of the yaw angle, pitch angle, and roll angle of the hyperspectral sensor;

[0014] Fit the UAV hyperspectral data based on the reflectance spectral data to obtain the fitted infrared spectrum;

[0015] Calibrate the key parameters based on the fitted infrared spectrum, and construct a lithium-bearing pegmatite spectral vector cloud and a non-lithium-bearing pegmatite spectral vector cloud based on the key parameters and the reflectance spectral data;

[0016] Correct the UAV hyperspectral data using the target spectra and the dark current data, and match the ground station data and the attitude data to obtain the corrected UAV hyperspectral data;

[0017] Based on the corrected UAV hyperspectral data, construct a UAV hyperspectral space vector cloud according to the key parameters;

[0018] Input the UAV hyperspectral space vector cloud, the lithium-bearing pegmatite spectral vector cloud, and the non-lithium-bearing pegmatite spectral vector cloud into the constructed network model to obtain an identification result; the identification result is one or more of an image containing lithium-bearing pegmatite information, an image containing non-lithium-bearing pegmatite information, and an image containing both lithium-bearing pegmatite information and non-lithium-bearing pegmatite information;

[0019] Based on the identification result using the spectral angle comparison method, retain the image containing lithium-bearing pegmatite information;

[0020] Process the image containing lithium-bearing pegmatite information using a filtering technique to obtain a binary image, and vectorize the binary image to form a vector file of the lithium-bearing pegmatite distribution;

[0021] Generate a color rendering map containing lithium-bearing pegmatite information based on the vector file of the lithium-bearing pegmatite distribution and the UAV hyperspectral data.

[0022] Optionally, based on the calibrated key parameters of the infrared spectrum after fitting, and based on the key parameters and the reflectance spectrum data, construct the spectral vector cloud of lithium-bearing pegmatite and the spectral vector cloud of lithium-free pegmatite, specifically including:

[0023] Based on the characteristics of lithium-bearing pegmatite and lithium-free pegmatite calibrated from the infrared spectrum after fitting, and use the characteristics of lithium-bearing pegmatite and lithium-free pegmatite as key parameters;

[0024] Based on the characteristics of lithium-bearing pegmatite and the reflectance spectrum data, construct the spectral vector cloud of lithium-bearing pegmatite;

[0025] Based on the characteristics of lithium-free pegmatite and the reflectance spectrum data, construct the spectral vector cloud of lithium-free pegmatite.

[0026] Optionally, use the target spectrum and the dark current data to correct the UAV hyperspectral data, and match the ground station data and the attitude data to obtain the corrected UAV hyperspectral data, specifically including:

[0027] According to the acquisition duration, divide the UAV hyperspectral data into the first part of UAV hyperspectral data and the second part of UAV hyperspectral data;

[0028] Use the current data collected before the UAV takes off to perform radiometric correction on the first part of UAV hyperspectral data, and use the current data collected after the UAV lands to perform radiometric correction on the second part of UAV hyperspectral data to obtain the radiometrically corrected UAV hyperspectral data;

[0029] Take the target spectrum as the standard spectrum, obtain the spectral data corresponding to the target in the radiometrically corrected UAV hyperspectral data, and use the obtained spectral data corresponding to the target as white balance, and use the white balance to perform reflectance correction on the spectral data in the UAV hyperspectral data except for the obtained spectral data corresponding to the target to obtain the reflectance-corrected UAV hyperspectral data;

[0030] Use the DEM model and the satellite attitude to perform orthorectification on the reflectance-corrected UAV hyperspectral data, and in this rectification process, match the ground station data and the attitude data to obtain the corrected UAV hyperspectral data.

[0031] Optionally, based on the recognition result, use the spectral angle comparison method to retain the image containing lithium-bearing pegmatite information, specifically including:

[0032] When the recognition result is an image that contains both lithium-bearing pegmatite information and lithium-free pegmatite information, calculate the spectral angle between the spectrum in the UAV hyperspectral data and the spectrum in the emissivity data of lithium-bearing pegmatite to obtain the first spectral angle result;

[0033] Calculate the spectral angle between the spectrum in the UAV hyperspectral data and the spectrum in the emissivity data of the lithium-free pegmatite to obtain the second spectral angle result;

[0034] When the first spectral angle result is less than the second spectral angle result, retain the image that contains both lithium-bearing pegmatite information and lithium-free pegmatite information;

[0035] When the first spectral angle result is greater than or equal to the second spectral angle result, do not retain the image that contains both lithium-bearing pegmatite information and lithium-free pegmatite information.

[0036] Optionally, generate a color rendering map containing lithium-bearing pegmatite information based on the vector file of the lithium-bearing pegmatite distribution and the UAV hyperspectral data, specifically including:

[0037] Overlay the vector file of the lithium-bearing pegmatite distribution and the UAV hyperspectral data by coordinate stratification to generate a color rendering map containing lithium-bearing pegmatite information.

[0038] In a third aspect, the present application provides a lithium-bearing pegmatite identification system, which includes:

[0039] The above-provided UAV pegmatite spectral acquisition device, which is used to collect UAV hyperspectral data, target spectra, and dark current data; the dark current data is the current data collected before and after the UAV takes off; the target spectra are obtained based on targets with different reflectivities; the targets with different reflectivities are set within a set range centered on the UAV take-off position;

[0040] A UAV control system, carried on the UAV, which is used to control the flight of the UAV based on flight instructions; the flight instructions include one or more of flight routes, flight speeds, flight distances, and flight altitudes;

[0041] A hyperspectral sensor control system, carried on the UAV, which is used to control the working data of the hyperspectral sensor in the UAV pegmatite spectral acquisition device; the working data of the hyperspectral sensor includes one or more of acquisition duration and spectral bands;

[0042] A ground infrared spectrometer, which is used to collect reflectivity spectral data; the reflectivity spectral data includes the emissivity data of lithium-bearing pegmatite and the emissivity data of lithium-free pegmatite;

[0043] A ground station, which is used to obtain ground station data and attitude data; the attitude data includes one or more of the yaw angle, pitch angle, and roll angle of the hyperspectral sensor;

[0044] A service system that conducts data interaction with the drone pegmatite spectral acquisition device, the drone control system, the hyperspectral sensor control system, the ground infrared spectrometer, and the ground station, and is used to implement the provided lithium-bearing pegmatite identification method.

[0045] According to the specific embodiments provided in this application, the following technical effects are disclosed in this application:

[0046] A drone pegmatite spectral acquisition device, a lithium-bearing pegmatite identification method, and a system provided in this application, through reasonable design, mount a hyperspectral sensor on a drone, and can collect hyperspectral data of pegmatite in real time through the drone, providing a data basis for the identification of lithium-bearing pegmatite. By processing the obtained drone hyperspectral data, reflectance spectral data, etc., a drone hyperspectral spatial vector cloud, a lithium-bearing pegmatite spectral vector cloud, and a non-lithium-bearing pegmatite spectral vector cloud are obtained. Using these spectral vector clouds as the input of the constructed network model to obtain the identification result can improve the real-time performance of pegmatite identification. By adopting the spectral angle comparison method based on the identification result and retaining the image containing information of lithium-bearing pegmatite, the accuracy of lithium-bearing pegmatite identification can be further improved, thereby solving the problem that the prior art cannot carry out the identification of lithium-bearing pegmatite for drone hyperspectral.

[0047] In addition, in this application, a shock absorption structure or a calibration structure is provided at the connection between the hyperspectral sensor and the drone, which can effectively prevent the hyperspectral sensor from being damaged when the drone lands, thereby improving the service life of the hyperspectral sensor. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of this application. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0049] Figure 1 Schematic structural diagram of the drone pegmatite spectral acquisition device provided in the embodiment of this application;

[0050] Figure 2 Schematic structural diagram of the connection between the hyperspectral sensor and the drone provided in the embodiment of this application;

[0051] Figure 3 Flowchart of the lithium-bearing pegmatite identification method provided in the embodiment of this application;

[0052] Figure 4 Schematic diagram of the reflectance of lithium-bearing pegmatite provided in the embodiment of this application;

[0053] Figure 5 Schematic diagram of the reflectivity of the lithium-free pegmatite provided by the embodiment of the present application;

[0054] Figure 6 Schematic diagram of the construction of the infrared spectral vector cloud provided by the embodiment of the present application;

[0055] Figure 7 Schematic diagram of the drone hyperspectral data provided by the embodiment of the present application;

[0056] Figure 8 Schematic diagram of the drone hyperspectral spatial vector cloud provided by the embodiment of the present application;

[0057] Figure 9 Schematic diagram of the input device configuration provided by the embodiment of the present application;

[0058] Figure 10 Schematic diagram of the test results of the ground infrared spectrum of the wood-cut lithium-bearing pegmatite provided by the embodiment of the present application;

[0059] Figure 11 Schematic diagram of the test results of the ground infrared spectrum of the wood-cut lithium-free pegmatite provided by the embodiment of the present application;

[0060] Figure 12 Schematic diagram of the structure of the typical convolutional neural network provided by the embodiment of the present application;

[0061] Figure 13 Schematic diagram of the screening of the recognition results provided by the embodiment of the present application;

[0062] Figure 14 Field landscape map of lithium-bearing pegmatite provided by the embodiment of the present application;

[0063] Figure 15 Based on the embodiment of the present application Figure 14 Obtained drone hyperspectral lithium-bearing pegmatite recognition result map;

[0064] Figure 16 Based on the embodiment of the present application Figure 14 Generated color rendering map containing lithium-bearing pegmatite information;

[0065] Figure 17 Schematic diagram of the flight altitude value provided by the embodiment of the present application;

[0066] Figure 18 Technical flow chart of realizing the recognition of lithium-bearing pegmatite by using the lithium-bearing pegmatite recognition system provided by the embodiment of the present application.

[0067] Explanation of the reference numerals:

[0068] 1 UAV, 1-1 left wing, 1-2 right wing, 1-3 left rotor wing, 1-4 right rotor wing, 1-5 tail wing, 2 hyperspectral sensor, 3 landing gear, 4 shock absorption structure or calibration structure. Detailed implementation manners

[0069] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0070] The purpose of the present application is to provide a spodumene spectral acquisition device for UAVs, a lithium-bearing spodumene identification method and system, aiming to solve the problem that the prior art cannot identify lithium-bearing spodumene for UAV hyperspectral.

[0071] To make the above objects, features and advantages of the present application more obvious and understandable, the present application will be further described in detail below with reference to the accompanying drawings and specific implementation manners.

[0072] This embodiment provides a spodumene spectral acquisition device for UAVs, as Figure 1 shown. The device includes: a UAV 1 and a hyperspectral sensor 2.

[0073] A receiving cavity is opened on the cabin body of the UAV 1. For example, the aircraft cabin is cut and modified according to the size of the hyperspectral sensor 2 to obtain the receiving cavity. The receiving cavity is used to install the hyperspectral sensor. As Figure 2 shown, at the connection between the hyperspectral sensor 2 and the UAV 1, a shock absorption structure or calibration structure 4 is provided. For example, spherical shock absorption or a gimbal can be set for shock absorption or calibration.

[0074] In actual application, the UAV can be a vertical takeoff fixed-wing UAV, as Figure 1 shown. It includes: a left wing 1-1, a right wing 1-2, a left rotor wing 1-3, a right rotor wing 1-4, a tail wing 1-5 and a cabin body. The hyperspectral sensor can be a full-spectrum imager in the 400nm - 2500nm band.

[0075] Based on the spodumene spectral acquisition device for UAVs provided above in the present application, the problem that the current rotor UAVs are limited by flight time and speed and cannot carry out large-area operations can be solved, and it is suitable for large-area scanning of lithium-bearing spodumene in exposed areas.

[0076] In an exemplary embodiment, in order to prevent the hyperspectral sensor from being damaged when landing on the ground, the present application may further be provided with a landing gear 3. The landing gear 3 is installed on the UAV 1. For example, the landing gear 3 is welded at two areas at the head and tail of the UAV 1, so as to ensure the balance when landing. When the UAV 1 lands, the landing gear 3 can ensure that there is a gap between the hyperspectral sensor 2 and the ground, thereby increasing the service life of the hyperspectral sensor 2.

[0077] In an exemplary embodiment, in order to avoid the influence of the electromagnetic field of the UAV on the hyperspectral sensor 2, electromagnetic shielding is required. Based on this, this embodiment uses an electromagnetic shielding film as the electromagnetic shielding material, and reasonably arranges the accommodation cavity where the hyperspectral sensor 2 is located to shield the electromagnetic influence. Hyperspectral sensors without electromagnetic shielding and those installed on the UAV are used to test the same ground object under the same conditions, and it is found that the test results of the shielding effect are exactly the same.

[0078] In an exemplary embodiment, in order to avoid the problem that the front and back or left and right weights of the UAV are inconsistent due to the installation of the hyperspectral sensor, the weight matching of the front and back and left and right needs to be considered when installing the hyperspectral sensor. After installing the hyperspectral sensor, it is necessary to verify the consistency of the weights. The verification method adopted in this embodiment is to verify whether the center of gravity remains unchanged before and after installing the hyperspectral sensor to ensure the safety of the UAV flight.

[0079] In an exemplary embodiment, a method for identifying lithium pegmatite is provided, as Figure 3 shown, the method includes:

[0080] Step 100: Obtain reflectance spectral data, UAV hyperspectral data, target spectrum, dark current data, attitude data, and ground station data.

[0081] In the actual application process, the UAV hyperspectral data is a spectral image collected by the UAV pegmatite spectral acquisition device provided above in the present application. For example, according to the survey area range, the flight route can be planned, and the flight route can be arranged in two ways: north-south and east-west. Whiteboards and targets with multiple different reflectivities (for example, reflectivities of 11%, 33%, and 66%) are arranged near the take-off point as the targets for hyperspectral sensor calibration. At the same time, the relationships between the exposure time, frame period, flight altitude, flight speed, and frame period are configured. Generally, the frame period is about 80% of the DN value (Digital Number, that is, the pixel brightness value of the remote sensing image) reflected by the target. The flight speed is matched with the exposure time, frame period, and flight altitude to make the flight speed of the hyperspectral sensor consistent with the flight speed of the UAV. Set the excitation area and height, and gradually obtain the UAV hyperspectral data according to the flight route. At the same time, collect the dark current data before take-off and after landing.

[0082] The reflectance spectral data are the emissivity data of lithium-bearing pegmatite and non-lithium-bearing pegmatite measured by an infrared spectrometer. For example, the infrared spectrometer is configured as a portable spectrometer covering the wavelength range of 400 - 2500 nm. A white board with a reflectance of 1 is used as the calibration standard. Before measurement, the electrical signal is corrected by the white board to convert it into reflectance, and then the reflectance spectral curve is obtained. The target spectra are obtained based on targets with different reflectances. The targets with different reflectances are set within a set range centered on the take-off position of the drone. The attitude data includes one or more of the yaw angle, pitch angle, and roll angle of the hyperspectral sensor.

[0083] When configuring the ground station and the inertial navigation system, the ground station is configured to use RTK with post-differential processing (PPS) as the ground station, and the sampling time interval of the ground station is consistent with the timestamp.

[0084] The acquisition of ground station data is configured in a fixed-point mode. The sampling time interval is consistent with the timestamp of the drone hyperspectral data acquisition. Sampling is carried out to obtain fixed-point data, that is, ground station data.

[0085] Step 101: Fit the drone hyperspectral data based on the reflectance spectral data to obtain the fitted infrared spectrum.

[0086] In the actual application process, an infrared spectrometer can be used to test lithium-bearing pegmatite and non-lithium-bearing pegmatite to obtain their reflectance data. Based on the reflectance data, the drone hyperspectral data is fitted to obtain the fitted infrared spectrum.

[0087] Step 102: Calibrate the key parameters based on the fitted infrared spectrum, and construct the spectral vector cloud of lithium-bearing pegmatite and the spectral vector cloud of non-lithium-bearing pegmatite based on the key parameters and the reflectance spectral data.

[0088] In the actual application process, after obtaining the fitted infrared spectrum, the characteristics of lithium-bearing pegmatite and non-lithium-bearing pegmatite are calibrated as the key parameters for constructing the infrared spectral vector cloud. The obtained key parameters are mainly two characteristics, emissivity and absorption peak. For example, there is a high reflectance at 400 nm and an obvious absorption peak at 1000 nm. Among them, the reflectance of lithium-bearing pegmatite and the reflectance of non-lithium-bearing pegmatite are as Figure 4 and Figure 5 shown.

[0089] For example, according to the measured lithium-bearing pegmatite and non-lithium-bearing pegmatite, the main parameters of lithium-bearing pegmatite are determined as the POS position: n 1 ,n 2 ,n 3 ,n 4 ,……, and the main parameters of non-lithium-bearing pegmatite are determined as the POS position: m 1 ,m2 , m 3 , m 4 , …….

[0090] Construct an infrared spectral vector cloud based on the parameters of lithium-bearing pegmatite and lithium-free pegmatite. The vectors in the infrared spectral vector cloud include vector length, angle, and starting position. As Figure 6 shown, the coordinate a j (x j , y j ) and a m (x m , y m ) where the x-coordinate is the POS position and the y-coordinate is the reflectance (%), O j is the vector angle, and l j is the vector length.

[0091]

[0092] In the formula, j is the sequence number of the constructed vector, is the vector from j to m, l io is the complex form of the vector, i is the imaginary unit of the complex number, o is the direction angle of the vector, and l is the amplitude.

[0093] Step 103: Calibrate the UAV hyperspectral data using the target spectrum and dark current data, and match the ground station data and attitude data to obtain the calibrated UAV hyperspectral data.

[0094] In the actual application process, the implementation process of this step can be:

[0095] (1) Divide the UAV hyperspectral data into the first part of the UAV hyperspectral data and the second part of the UAV hyperspectral data according to the acquisition duration.

[0096] (2) Since the UAV data acquisition time is relatively long, generally, the data collected in the first half is corrected using the dark current before takeoff, and the data collected in the second half is corrected using the dark current after landing. Based on this, radiometric calibration is performed on the first part of the UAV hyperspectral data using the current data collected before the UAV takes off, and radiometric calibration is performed on the second part of the UAV hyperspectral data using the current data collected after the UAV lands to obtain the radiometrically calibrated UAV hyperspectral data.

[0097] Among them, DN 0～l / 2 = DN o1 - DN BC1 .

[0098] DN l / 2～l = DN o2 - DNB C2 .

[0099] In the formula, DN 0-l / 2 is the DN value after radiometric correction in the time period from 0 to l / 2, and DN o1 is the original DN value in the time period from 0 to l / 2, and DN BC1 is the dark current value collected in the time period from 0 to l / 2, and DN l / 2-l is the DN value after radiometric correction in the time period from l / 2 to l, and DN o2 is the original DN value in the time period from l / 2 to l, and DN BC2 is the dark current value collected in the time period from l / 2 to l, and l is the data acquisition time.

[0100] (3) Using the target spectrum as the standard spectrum, the spectral data corresponding to the target is obtained from the radiometrically corrected UAV hyperspectral data, and the obtained spectral data corresponding to the target is used as white balance. The spectral data other than the obtained spectral data corresponding to the target in the UAV hyperspectral data is corrected by white balance to obtain the radiometrically corrected UAV hyperspectral data.

[0101] For example, in the radiometrically corrected UAV hyperspectral data, several modes of the target (11%, 33%, 66%, 99.99%) are circled as white balance, and other data are corrected according to the white balance.

[0102]

[0103] In the formula, ρ target is the ground object reflectance, DN target is the DN value after radiometric correction, DN panel is the DN value of the target after radiometric correction, and ρ panel is the reflectance of the target.

[0104] (4) The orthorectification of the reflectance-corrected UAV hyperspectral data is performed using the DEM model and satellite attitude, and during this rectification process, the ground station data and attitude data are matched to obtain the rectified UAV hyperspectral data.

[0105] For example, the original attitude of the hyperspectral sensor is (Y o , P o , R o ), which represents the initial yaw, pitch, and roll angles of the hyperspectral sensor. The attitude of each point (timestamp) is Y q , P q , R q , representing the yaw, pitch, and roll angles of the sensor at timestamp q. First, the original attitude of the instrument is corrected.

[0106] Y c = Y q - Y o , Pc = P q -P O , R c = R q -R o are the yaw, pitch, and roll angles of the corrected timestamp i, respectively.

[0107] The coordinates of each corrected timestamp q are:

[0108]

[0109] In the formula, m represents the cartographic coordinate system, u represents the image coordinate system, x c , y c , z c are the coordinates of the projection center of timestamp q in the cartographic coordinate system, x q , y q , z q are the coordinates of the ground point in the cartographic coordinate system of timestamp q, x, y, -g are the spatial coordinates of the pixel point of timestamp q in the image coordinate system, k is the scale factor k = h / g, h is the flight altitude, T is the transformation matrix, and is represented by the following formula:

[0110]

[0111] In the formula, W is the WGS84 coordinate system, g is the local horizontal coordinate system, b is the IMU coordinate system, c is the imaging spectrometer coordinate system, j is the image coordinate system, is the transformation matrix from the IMU coordinate system to the local horizontal coordinate system, Y c , P c , R c are the yaw, pitch, and roll angles of the corrected timestamp q, is the transformation matrix from the local horizontal coordinate system to the WGS84 coordinate system, is the transformation matrix from the WGS84 coordinate system to the cartographic coordinate system, is the transformation matrix from the image coordinate system to the imaging spectrometer coordinate system, is the transformation matrix from the imaging spectrometer coordinate system to the IMU coordinate system.

[0112] Step 104: Based on the corrected UAV hyperspectral data, construct a UAV hyperspectral spatial vector cloud according to the key parameters.

[0113] Since the hyperspectral data of drones obtained by hyperspectral sensors has certain noise and often requires smoothing processing, which takes a lot of computing time, constructing a hyperspectral vector cloud can directly participate in the calculation and save a large amount of time. The construction of the hyperspectral vector cloud of drones needs to be consistent with the construction of the infrared spectral vector cloud to ensure mutual consistency. Input the corrected hyperspectral data of drones, and construct the hyperspectral spatial vector cloud of drones according to the key parameters for constructing the infrared spectral vector cloud. Among them, the hyperspectral data of drones is as Figure 7 shown, and the constructed hyperspectral spatial vector cloud of drones is as Figure 8 shown.

[0114] For example, according to the infrared spectral calibration parameters and the infrared spectral vector construction formula Take the value band a j , a m as the parameters for constructing the hyperspectral vector cloud of drones, then any point on the image (i.e., the vectorized data) can be expressed as:

[0115]

[0116] Among them, j and m are the POS positions of band j and band m, k and l are the number of rows and columns of the drone image matrix, i is the imaginary unit of the complex number, 1…k, 1…t are the matrix from row 1 to row k and from column 1 to column t. In this way, each point of the image can form a new multi-dimensional data set, and each point in the multi-dimensional data layer has an amplitude and a phase, and can be compared through the two parameters of amplitude and phase, so as to form the amplitude and phase vector space data cloud.

[0117] It is represented by a simple example. For example, the spectral data of lithium-bearing pegmatite is:

[0118] Wavelength nm Reflectance %

[0119]

[0120] V 400,2100 ={0.2 0.8}=1.8 i13.25 .

[0121] At the same time, the band representation of point a in the hyperspectral image of the drone is:

[0122] Wavelength nm Reflectance %

[0123]

[0124] The vector cloud constructed according to the spectral parameters of lithium-bearing pegmatite is:

[0125] a 400,2100 ={0.5 0.7]=1.7i6.71 。

[0126] Then the spectrum at location a is inconsistent with that of the lithium-bearing pegmatite.

[0127] Step 105: Input the UAV hyperspectral spatial vector cloud, the lithium-bearing pegmatite spectral vector cloud, and the non-lithium-bearing pegmatite spectral vector cloud into the constructed network model to obtain the recognition result. The recognition result is one or more of an image containing lithium-bearing pegmatite information, an image containing non-lithium-bearing pegmatite information, and an image containing both lithium-bearing pegmatite information and non-lithium-bearing pegmatite information.

[0128] For example, configure data such as the UAV hyperspectral spatial vector cloud, the lithium-bearing pegmatite spectral vector cloud, and the non-lithium-bearing pegmatite spectral vector cloud as the input device shown in Figure 9 the figure.

[0129] For example, for data with obvious spectral features, the spectral analysis software uses TSG (Spectral Geologist) to analyze altered minerals. Tests show that although the spectra of lepidolite and muscovite are not very different, the spectra of spodumene and lepidolite are significantly different from that of quartz. The common spectral features of ore-bearing pegmatite and non-ore-bearing pegmatite are: both have a spectral absorption at 2200 nm, a low reflectance between 2400 - 2500 nm, and a high reflectance between 350 - 750 nm (the reason for being visible white to the naked eye). The differences are that the ore-bearing pegmatite has an obvious absorption feature at 2200 nm ( Figure 10 ), and also has two small absorption valleys at 2360 nm and 2450 nm ( Figure 11 ). The spectrometer actually has a spectral resolution lower than 10 nm and is not affected by the atmosphere. The spatial resolution can reach 2 cm, and the electrical signal is stable. The comparison of its spectral features is shown in Table 1. The obvious differences in the spectra of spodumene, lepidolite, and quartz in the short-wave infrared range indicate that the hyperspectral method can be used to distinguish lithium-bearing pegmatite from non-lithium-bearing pegmatite. Therefore, UAV hyperspectral data can be used for the research on the identification of lithium pegmatite.

[0130] Table 1 Comparison table of spectral characteristics between lithium-bearing pegmatite and non-lithium-bearing pegmatite in Tamuoche

[0131]

[0132] Furthermore, in 1986, Rumelhart et al. proposed the backpropagation algorithm for artificial neural networks, which is superior to shallow machine learning models such as support vector machines, BOOSTING, and LOGISTIC regression methods based on statistical learning theory. In 2006, Hinton et al. proposed that deep learning artificial neural networks with multiple hidden layers have excellent feature learning capabilities. They can overcome the training difficulty through unsupervised training layer by layer, utilize the non-linear operators of the deep structure, obtain higher-level features, and improve the recognition rate and accuracy. Based on this, the network model adopted in this application is a typical convolutional neural network, such as Figure 12 shown, which mainly includes an input layer, a convolutional layer, a sampling layer, a connection layer, and an output layer.

[0133] For example, convolution filtering is performed on each channel of the input image, and the features of the image training area are taken to obtain feature maps with the same number as the convolution kernels. After these features pass through the resampling layer, feature maps with a small size but a large number are obtained. The feature maps are unfolded in a certain way and spliced into a one-dimensional vector connection layer, and then output through several connection layers to achieve the recognition task.

[0134] The convolutional layer can be represented by the following formula. The j-th feature matrix can be obtained by convolution weighting of several feature maps in the previous layer, and there is:

[0135]

[0136] In the formula, f is the neuron activation function, N j represents the combination of input feature maps, represents the convolution operation, convolution core matrix, is the bias matrix, represents the feature matrix of the previous layer.

[0137] Common neuron activation functions include the Sigmoid function, Tanh function, ReLU function, etc.

[0138] The sampling layer is also called the "pooling layer". Its function is to perform pooling sampling based on the principle of local correlation, reduce the data volume, and retain useful information. There is:

[0139] represents the feature matrix of the previous layer.

[0140] In the formula, d represents the sampling function, including maximum sampling, mean sampling, etc.

[0141] After passing through the convolutional layer and the sampling layer, it is linked to a connection layer. Each neuron is connected to each neuron in the next layer. The feature vector of the l-th connection layer can be expressed as x l :

[0142] x l = f(w l x l-1 + b l )。

[0143] In the formula, w l is the weight matrix, b l is the bias vector, and x l-1 is the feature vector of the connection layer of the (l - 1)-th layer.

[0144] When the last output layer of the model is a logistic regression layer, each node output by the convolutional neural network represents the probability P(Y = t|x, w, b) of a certain category t of the input data, and there is:

[0145]

[0146] In the formula, w is the weight parameter of the last layer, b is the corresponding bias parameter, x is the input data, and w t is the weight parameter of the last layer for category t, s t () is the probability of category t, and w j is the weight parameter of the last layer for category j.

[0147] Step 106: Based on the recognition result, use the spectral angle comparison method to retain the image containing the information of lithium-bearing pegmatite.

[0148] In the actual application process, the implementation process of this step 106 can be:

[0149] (1) When the recognition result is an image that contains both the information of lithium-bearing pegmatite and the information of non-lithium-bearing pegmatite, calculate the spectral angle between the spectrum in the UAV hyperspectral data and the spectrum in the emissivity data of lithium-bearing pegmatite to obtain the first spectral angle result.

[0150] (2) Calculate the spectral angle between the spectrum in the UAV hyperspectral data and the spectrum in the emissivity data of non-lithium-bearing pegmatite to obtain the second spectral angle result.

[0151] (3) When the first spectral angle result is less than the second spectral angle result, retain the image that contains both the information of lithium-bearing pegmatite and the information of non-lithium-bearing pegmatite.

[0152] (4) When the first spectral angle result is greater than or equal to the second spectral angle result, do not retain the image that contains both the information of lithium-bearing pegmatite and the information of non-lithium-bearing pegmatite.

[0153] For example, the output results of the above data processing of lithium-bearing pegmatite and lithium-free pegmatite can be used as the input layer of the judgment device. If it is lithium-bearing pegmatite information, it is retained; if it is lithium-free pegmatite information, it is deleted; if it contains both lithium-bearing pegmatite information and lithium-free pegmatite information, then the spectra of the hyperspectral images collected by the UAV here are respectively compared with the spectra of lithium-bearing pegmatite and lithium-free pegmatite obtained from the ground infrared spectra for spectral angle, as Figure 13 shown. Judge the size of the spectral angle. If the spectral angle of the lithium-bearing pegmatite is less than that of the lithium-free pegmatite, it is retained; otherwise, it is deleted.

[0154] The spectral angle method represents each multi-dimensional space point with its spatial vector and compares the similarity of the spatial vector angles. It is a supervised classification method that requires a known reference spectrum for each category. This reference spectrum can be measured on the ground and stored in the reference spectrum library, or it can be statistically analyzed from the regions of interest of the graphic units with known conditions and stored in the reference spectrum library. The formula is as follows:

[0155]

[0156] In the formula, (α,β) is the inner product of the n-dimensional vectors α and β. According to the definition of the inner product, there is:

[0157] (α,β) = α 1 β 1 +α 2 β 2 +…+α n β n .

[0158] When α and β are column vectors, (α,β) = α’β = β’α, there is:

[0159]

[0160] In the formula, |α| and |β| are the lengths of the vectors α and β,

[0161] By calculating the inner product and length of α and β, cos can be obtained, and its included angle can be found by looking up the table.

[0162] Step 107: Process the image containing lithium-bearing pegmatite information using filtering technology to obtain a binary image, and vectorize the binary image to form a vector file of the lithium-bearing pegmatite distribution.

[0163] In the actual application process, median filtering can be used to process the two-dimensional image of lithium-bearing pegmatite. For example, the output two-dimensional matrix is X ij , and the median filtering is:

[0164]

[0165] Wherein, F ij is the median filtering result, A is the filtering window, and the window can be selected as 3×3 or 5×5, etc., until the filtering effect is satisfactory.

[0166] After filtering, a two-dimensional {0, 1} binary image is obtained. By vectorizing the two-dimensional binary image, a vector file of the distribution of lithium-bearing pegmatite can be formed.

[0167] Step 108: Generate a color rendering map containing lithium-bearing pegmatite information based on the vector file of the distribution of lithium-bearing pegmatite and the UAV hyperspectral data.

[0168] In the actual application process, input the optimized lithium-bearing pegmatite vector, R (red): 669.43nm, G (green): 538.96nm, B (blue): 479.25nm in the UAV hyperspectral bands, as the input items for the RGB (red, green, blue) mode color composite image, synthesize the color image, and the vector is represented by points, lines, and surfaces with the same projection. Use coordinate layering to overlay the raster and the vector.

[0169] Among them, the vector is f(x, y, z), x and y are the corresponding coordinates, z is the eigenvalue, f(x, y, z) is the vector value, the raster is g(x′, y′), x′ and y′ are the corresponding coordinates, and g(x′, y′) is the raster gray value

[0170] Let x = x′ and y = y′, so as to realize the overlay of the raster gray value g(x′, y′) and the vector g(x′, y′).

[0171] Use the corresponding coordinates to overlay the raster and the vector, and the output is a GIS vector and raster overlay image, thus forming an image suitable for the human eye habit (i.e., a color rendering map).

[0172] That is, input the UAV hyperspectral base map and the lithium-bearing pegmatite vector, and output an overlay image (GIS) suitable for human eye observation, which can be output as a final image in JPG or TIF format through software. Among them, taking the Figure 14 field landscape of lithium-bearing pegmatite shown as an example, the recognition result of the UAV hyperspectral lithium-bearing pegmatite is as Figure 15 shown, and the finally generated image suitable for the human eye habit is as Figure 16 shown.

[0173] In an exemplary embodiment, before acquiring the hyperspectral data of the drone, it is necessary to match the timestamps and flight attitudes. For example, each frame exposure has a timestamp, and each timestamp records the attitude of the hyperspectral sensor. Generally, the attitude of the hyperspectral sensor includes one or more of data such as position (GPS), velocity in the X direction, velocity in the Y direction, velocity in the Z direction, altitude, yaw, roll, and pitch.

[0174] In an exemplary embodiment, a lithium-bearing pegmatite identification system is provided. The system includes: the drone pegmatite spectral acquisition device provided above, a drone control system, a hyperspectral sensor control system, a ground infrared spectrometer, a ground station, and a service system.

[0175] The drone pegmatite spectral acquisition device provided above is used to acquire drone hyperspectral data, target spectra, and dark current data. The dark current data is the current data acquired before and after the drone takes off and lands. The target spectra are obtained based on targets with different reflectivities. The targets with different reflectivities are set within a set range centered on the takeoff position of the drone.

[0176] The drone control system is carried on the drone and is used to control the flight of the drone based on flight instructions. The flight instructions include one or more of flight routes, flight speeds, flight distances, and flight altitudes.

[0177] The hyperspectral sensor control system is carried on the drone and is used to control the working data of the hyperspectral sensor in the drone pegmatite spectral acquisition device. The working data of the hyperspectral sensor includes one or more of acquisition duration and spectral bands.

[0178] The ground infrared spectrometer is used to acquire reflectance spectral data. The reflectance spectral data includes emissivity data of lithium-bearing pegmatite and emissivity data of non-lithium-bearing pegmatite.

[0179] The ground station is used to acquire ground station data and attitude data. The attitude data includes one or more of the yaw angle, pitch angle, and roll angle of the hyperspectral sensor.

[0180] The service system conducts data interaction with the drone pegmatite spectral acquisition device, the drone control system, the hyperspectral sensor control system, the ground infrared spectrometer, and the ground station, and is mainly used to implement the lithium-bearing pegmatite identification method provided above.

[0181] In an exemplary embodiment, in order to ensure the mobility of the platform, the drone control system that controls the drone and the hyperspectral sensor control system that controls the hyperspectral sensor are two different systems to match the relationships of flight altitude, flight distance, and flight speed. Generally, as Figure 17The flight altitude value shown is the difference between the altitude of the aircraft in the survey area and the average altitude of the survey area. The speed measurement of the hyperspectral sensor system should be consistent with the flight speed of the UAV. The flight speed is generally the airspeed. The flight distance is related to the viewing angle of the sensor. The mutual relationship is shown in the following formula:

[0182] The flight distance ll = htan(a / 2).

[0183] Flight altitude h, viewing angle a, flight distance ll.

[0184] In summary, the overall process of the lithium-bearing pegmatite identification method implemented by the lithium-bearing pegmatite identification system of the present application is as Figure 18 shown. By distinguishing the infrared spectra of lithium-bearing pegmatites and non-lithium-bearing pegmatites, it calibrates the characteristic spectral bands for lithium-bearing pegmatite identification as input items, and uses the vector cloud space neural network method to identify lithium-bearing pegmatites in the hyperspectral data of the UAV, forming a system for lithium-bearing pegmatite identification. This system mainly makes up for the inability of existing satellite hyperspectral remote sensing with a resolution of 30 meters to identify pegmatite veins at the meter and centimeter levels. Although satellite high-resolution remote sensing has a panchromatic band with a resolution of 0.3 meters, it does not have high-resolution spectra and cannot distinguish lithium-bearing pegmatite veins. At the same time, it solves the problem that current rotor UAVs are limited by flight time and speed and cannot carry out large-area operations. This device and system are very suitable for large-area scanning and identifying lithium-bearing pegmatites in exposed areas.

[0185] In an exemplary embodiment, the present application provides a computer device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor. The processor executes the computer program to implement the lithium-bearing pegmatite identification method provided above.

[0186] In an exemplary embodiment, the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the lithium-bearing pegmatite identification method provided above.

[0187] In an exemplary embodiment, the present application provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the lithium-bearing pegmatite identification method provided above.

[0188] In an exemplary embodiment, the present application provides a computer device, which may be a database. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store transactions to be processed. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, it can implement the provided lithium-bearing pegmatite identification method.

[0189] It should be noted that the object information (including but not limited to object device information, object personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the object or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions.

[0190] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0191] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in this specification.

[0192] In this article, specific examples are used to elaborate on the principles and implementation manners of the present application. The descriptions of the above embodiments are only used to help understand the methods and core ideas of the present application. The same or similar parts among the various embodiments can be referred to each other; at the same time, for those of ordinary skill in the art, according to the ideas of the present application, there will be changes in the specific implementation manners and application scopes. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A method for identifying lithium-containing pegmatite, characterized in that: The lithium-bearing pegmatite identification method comprises: Acquire reflectivity spectrum data, UAV hyperspectral data, target spectrum, dark current data, attitude data and ground station data; the UAV hyperspectral data is a spectrum image acquired by a UAV pegmatite spectrum acquisition device; the dark current data is current data acquired before and after the UAV takes off and lands; the reflectivity spectrum data is the emissivity data of lithium-bearing pegmatite and the emissivity data of lithium-free pegmatite measured by an infrared spectrometer; the target spectrum is obtained based on targets with different reflectivities; the targets with different reflectivities are set within a set range centered on the take-off position of the UAV; the attitude data includes one or more of the yaw angle, pitch angle and roll angle of the hyperspectral sensor; Fitting the UAV hyperspectral data based on the reflectivity spectral data to obtain a fitted infrared spectrum; Calibrate key parameters based on the fitted infrared spectrum, and construct a spectrum vector cloud of lithium-bearing pegmatite and a spectrum vector cloud of lithium-free pegmatite based on the key parameters and the reflectivity spectrum data; The target spectrum and the dark current data are used to correct the UAV hyperspectral data, and the ground station data and the attitude data are matched to obtain the corrected UAV hyperspectral data; Based on the corrected UAV hyperspectral data, constructing a UAV hyperspectral spatial vector cloud according to the key parameters; Inputting the drone hyperspectral spatial vector cloud, the lithium-bearing pegmatite spectral vector cloud, and the lithium-free pegmatite spectral vector cloud into the constructed network model, obtaining a recognition result; the recognition result is one or more of an image containing lithium-bearing pegmatite information, an image containing lithium-free pegmatite information, and an image containing both lithium-bearing pegmatite information and lithium-free pegmatite information; retaining images containing information of lithium-bearing pegmatite based on the identification results by using a spectral angle comparison method; Processing the image containing lithium-bearing pegmatite information by using filtering technology to obtain a binary image, and vectorizing the binary image to form a vector file of lithium-bearing pegmatite distribution; A color rendering containing lithium-bearing pegmatite information is generated based on the vector file of the lithium-bearing pegmatite distribution and the drone hyperspectral data.

2. The method for identifying lithium-bearing pegmatite according to claim 1, characterized in that: Calibrate key parameters based on the fitted infrared spectrum, and construct a spectrum vector cloud of lithium-bearing pegmatite and a spectrum vector cloud of lithium-free pegmatite based on the key parameters and the reflectivity spectrum data, specifically including: Calibrate the characteristics of lithium-bearing pegmatite and the characteristics of lithium-free pegmatite based on the fitted infrared spectrum, and use the characteristics of lithium-bearing pegmatite and the characteristics of lithium-free pegmatite as key parameters; Constructing the spectral vector cloud of the lithium-bearing pegmatite based on the characteristics of the lithium-bearing pegmatite and the reflectivity spectral data; The spectral vector cloud of the lithium-free pegmatite is constructed based on the characteristics of the lithium-free pegmatite and the reflectance spectrum data.

3. The lithium-bearing pegmatite identification method according to claim 1, characterized in that: The target spectrum and the dark current data are used to correct the UAV hyperspectral data, and the ground station data and the attitude data are matched to obtain the corrected UAV hyperspectral data, specifically including: According to the acquisition time, the UAV hyperspectral data is divided into the first part of UAV hyperspectral data and the second part of UAV hyperspectral data; The first part of the UAV hyperspectral data is radiometrically corrected using the current data collected before the UAV takes off, and the second part of the UAV hyperspectral data is radiometrically corrected using the current data collected after the UAV lands, to obtain the radiometrically corrected UAV hyperspectral data; The target spectrum is used as a standard spectrum, and the spectrum data corresponding to the target is obtained from the UAV hyperspectral data after radiation correction, and the obtained spectrum data corresponding to the target is used as white balance. The spectrum data in the UAV hyperspectral data other than the obtained spectrum data corresponding to the target is subjected to reflection correction by using the white balance to obtain the UAV hyperspectral data after reflection correction; The DEM model and satellite attitude are used to perform orthorectification on the reflectance-corrected UAV hyperspectral data, and in this correction process, the ground station data and the attitude data are matched to obtain the corrected UAV hyperspectral data.

4. The method for identifying lithium-bearing pegmatite according to claim 1, characterized in that: The spectral angle comparison method is used to retain images containing lithium-bearing pegmatite information based on the identification results, specifically including: When the recognition result is an image containing both lithium-bearing pegmatite information and non-lithium-bearing pegmatite information, a spectral angle calculation is performed on the spectrum in the drone hyperspectral data and the spectrum in the emissivity data of the lithium-bearing pegmatite to obtain a first spectral angle result; The spectrum angle of the spectrum in the UAV hyperspectral data and the spectrum in the emissivity data of the lithium-free pegmatite are calculated to obtain the second spectrum angle result; When the first spectral angle result is less than the second spectral angle result, retaining the image containing both lithium-bearing pegmatite information and non-lithium-bearing pegmatite information; When the first spectral angle result is greater than or equal to the second spectral angle result, the image containing both lithium-bearing pegmatite information and non-lithium-bearing pegmatite information is not retained.

5. The method for identifying lithium-bearing pegmatite according to claim 1, characterized in that: A color rendering containing lithium-bearing pegmatite information is generated based on the vector file of the lithium-bearing pegmatite distribution and the drone hyperspectral data, specifically including: The vector file of the lithium-bearing pegmatite distribution and the drone hyperspectral data are superimposed using coordinate layering to generate a color rendering containing lithium-bearing pegmatite information.

6. A lithium-containing pegmatite identification system, characterized in that: The system comprises: The UAV pegmatite spectrum acquisition device is used to collect UAV hyperspectral data, target spectrum and dark current data; the dark current data is the current data collected before the UAV takes off and after landing; the target spectrum is obtained based on targets with different reflectivities; the targets with different reflectivities are set within a set range centered on the take-off position of the UAV; A UAV control system is mounted on the UAV and is used to control the flight of the UAV based on flight instructions; the flight instructions include one or more of route, speed, distance and altitude; A hyperspectral sensor control system, carried on the UAV, is used to control the working data of the hyperspectral sensor in the UAV pegmatite spectrum acquisition device; the working data of the hyperspectral sensor includes one or more of acquisition time and spectral band; A ground-based infrared spectrometer for collecting reflectance spectrum data; the reflectance spectrum data includes the emissivity data of lithium-bearing pegmatite and the emissivity data of lithium-free pegmatite; A ground station, used to obtain ground station data and attitude data; the attitude data includes one or more of a yaw angle, a pitch angle, and a roll angle of the hyperspectral sensor; The service system interacts with the UAV pegmatite spectrum acquisition device, the UAV control system, the hyperspectral sensor control system, the ground infrared spectrometer and the ground station for implementing the lithium-bearing pegmatite identification method as described in any one of claims 1 to 5.

7. The lithium-bearing pegmatite identification system according to claim 6, characterized in that: The UAV pegmatite spectrum collection device comprises: a UAV and a hyperspectral sensor; A housing is provided on the body of the drone; the housing is used to install the hyperspectral sensor; a shock absorbing structure or a correction structure is provided at the connection between the hyperspectral sensor and the drone.

8. The lithium-bearing pegmatite identification system according to claim 6, characterized in that: The device also includes a landing gear; the landing gear is installed on the UAV; when the UAV lands, the landing gear is used to form a gap between the hyperspectral sensor and the ground.

9. The lithium-bearing pegmatite identification system according to claim 7, characterized in that: A battery shielding film is arranged in the accommodating cavity.

10. The lithium-bearing pegmatite identification system according to claim 6, characterized in that: The UAV is a vertical take-off fixed-wing UAV; the hyperspectral sensor is a full-spectrum imager with a wavelength range of 400nm-2500nm.

Images