Bauxite lithium-rich ore rapid optimization method and system based on imaging hyperspectrum

Abstract

The invention relates to the technical field of mineral resource exploration, in particular to a bauxite lithium-rich ore rapid optimization method and system based on an imaging hyperspectrum. The method comprises the following steps: collecting standard hyperspectral data according to a bauxite sample; using the standard hyperspectral data to construct a multi-dimensional fusion spectral data set; screening optimal characteristic wave band spectral data from the multi-dimensional fusion spectral data set; and constructing a lithium ore rapid optimization model based on the optimal characteristic wave band spectral data, and generating a lithium content spatial distribution thermodynamic diagram through the lithium ore rapid optimization model. According to the method, efficient and accurate recognition of the lithium-rich ore in the bauxite is achieved through multi-dimensional spectrum transformation, characteristic wave band screening and integrated estimation model construction, and the problems that a traditional method is high in cost, low in efficiency and insufficient in optimization precision are solved.

Description

Technical Field

[0001] This invention relates to the field of mineral resource exploration technology, specifically to a rapid optimization method and system for lithium-rich bauxite ore based on imaging hyperspectral imaging. Background Technology

[0002] The refining of clay-type lithium ores mainly relies on acid extraction and roasting methods. These methods increase in cost as impurities in the ore increase, especially due to the generation of more polluting impurities. The key to developing this type of lithium resource lies in extracting more lithium resources at a limited cost while generating less polluting waste. Although remote sensing technology has yielded many successful applications in mineral identification and raw heavy metal estimation, research on how to utilize hyperspectral technology to improve lithium ore grade is scarce.

[0003] Therefore, based on the shortcomings of existing research, this invention develops a rapid optimization technique for lithium-rich bauxite ore based on imaging hyperspectral remote sensing. This technique can quickly and non-destructively optimize lithium-rich ore, reduce beneficiation and smelting costs, and provide technical reference for related industries. Summary of the Invention

[0004] To address the shortcomings of existing methods and the needs of practical applications, and in order to solve the aforementioned problems, this invention provides a rapid optimization method for lithium-rich bauxite ore based on imaging hyperspectral imaging, comprising the following steps: Standard hyperspectral data was collected from bauxite samples; a multi-dimensional fused spectral dataset was constructed using the standard hyperspectral data; the optimal characteristic band spectral data was selected from the multi-dimensional fused spectral dataset; a rapid lithium ore selection model was constructed based on the optimal characteristic band spectral data; and a thermal map of the spatial distribution of lithium content was generated using the rapid lithium ore selection model.

[0005] Optionally, constructing a multi-dimensional fused spectral dataset using the standard hyperspectral data includes the following steps: Logarithmic transformation and continuum removal are performed on the spectral reflectance data for each band; the first derivative is calculated on the logarithmic spectral data, and the original spectrum, logarithmic spectrum, continuum-removed spectrum, and first derivative spectrum are concatenated according to the band dimension to obtain a multi-dimensional fused spectral dataset.

[0006] Optionally, the step of selecting the optimal feature band spectral data from the multi-dimensional fused spectral dataset includes the following steps: A multi-dimensional fused spectral dataset and corresponding lithium content data are combined to form a sample set. Based on the sample set, a spectral-lithium content regression model is constructed using the partial least squares method to obtain the regression coefficients for each band. The weights of the first band are calculated according to the regression coefficients, and the first band set is selected using the weights of the first band and the ratio of the number of retained bands. The weights of the second band are assigned to the first band set using a two-factor adaptive weighting algorithm of variance contribution and information gain. The sample set is divided according to the lithium content to obtain multiple interval sample sets. The optimal feature band spectral data is obtained by combining the weights of the second band and the interval sample sets.

[0007] Optionally, the calculation of the first band weight based on the regression coefficient satisfies: in, Let be the absolute value of the regression coefficient of the j-th band in the i-th sampling. is the absolute value weight of the regression coefficient of the j-th band in the i-th sampling, and m is the number of remaining bands in each sampling.

[0008] Optionally, the first band set is selected using the weight of the first band and the ratio of the number of retained bands, satisfying the following: in, This is the total number of bands. It is the minimum number of bands retained in the Nth sampling, where N is the number of samplings. It is the proportion of the number of bands retained in the i-th sampling. and These represent the initial coefficient and the decay rate coefficient of the decay index, respectively.

[0009] Optionally, the second band weights are assigned to the first band set using a two-factor adaptive weighting algorithm based on variance contribution and information gain, satisfying the following: in, This represents the variance-information gain fusion weight for the i-th band, where V is the variance contribution and G is the information gain. It is the variance of the product of variance and information gain.

[0010] Optionally, the variance contribution satisfies in, This represents the variance of the x-th band, used to measure the information richness of that band.

[0011] Optionally, the information gain satisfies: in, It is the entropy of variable Y. It is the conditional entropy of Y given the band x.

[0012] Optionally, the sample set is divided according to lithium content to obtain multiple interval sample sets. The optimal characteristic band spectral data is obtained by combining the second band weight and the interval sample sets, satisfying the following: in, This represents the overall relative root mean square error. This represents the relative root mean square error of the entire sample. This represents the relative root mean square error of samples in the low-content range. This represents the relative root mean square error of samples within the medium content range. This indicates the relative root mean square error of samples in the high-content range. This represents the weighting coefficient.

[0013] This invention collects hyperspectral data from bauxite samples and performs multi-dimensional spectral fusion processing, including logarithmic transformation, continuum removal, and first derivative calculation, to construct a fused dataset containing the original spectra and their transformed features. Based on this, and combined with the lithium content data of the samples, partial least squares regression analysis is used to calculate the regression coefficients of each band, and band weights are calculated accordingly. A preliminary set of feature bands is gradually selected by setting a retention ratio. Furthermore, a two-factor adaptive weighted algorithm based on variance contribution and information gain is introduced to perform secondary weight allocation and optimization screening of candidate bands, ultimately extracting the feature bands most strongly correlated with lithium content. Finally, a lithium content prediction model is established based on these selected bands, and this model is used to perform pixel-by-pixel inversion of hyperspectral images to generate a heatmap that intuitively reflects the spatial distribution of lithium content, thereby achieving rapid, non-contact identification and location of lithium-rich ore in bauxite.

[0014] Secondly, to efficiently execute the rapid selection method for lithium-rich bauxite ore based on imaging hyperspectral imaging provided by this invention, this invention also provides a rapid selection system for lithium-rich bauxite ore based on imaging hyperspectral imaging, including a processor, an input device, an output device, and a memory. The processor, input device, output device, and memory are interconnected. The memory stores a computer program containing program instructions. The processor is configured to call the program instructions to execute the rapid selection method for lithium-rich bauxite ore based on imaging hyperspectral imaging as described in the first aspect of this invention. The rapid selection system for lithium-rich bauxite ore based on imaging hyperspectral imaging of this invention has a compact structure and stable performance, and can stably execute the rapid selection method for lithium-rich bauxite ore based on imaging hyperspectral imaging provided by this invention, further enhancing the overall applicability and practical application capability of this invention. Attached Figure Description

[0015] Figure 1 A flowchart of a rapid optimization method for lithium-rich bauxite ore based on imaging hyperspectral is provided in this embodiment of the invention. Figure 2 This is a framework diagram of a rapid optimization system for lithium-rich bauxite ore based on imaging hyperspectral imaging, provided in an embodiment of the present invention. Detailed Implementation

[0016] Specific embodiments of the present invention will now be described in detail. It should be noted that the embodiments described herein are for illustrative purposes only and are not intended to limit the invention. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be apparent to those skilled in the art that these specific details are not necessary to practice the invention. In other instances, well-known circuits, software, or methods have not been specifically described to avoid obscuring the invention.

[0017] Throughout this specification, references to "an embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in connection with that embodiment or example is included in at least one embodiment of the invention. Therefore, the phrases "in an embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily refer to the same embodiment or example. Furthermore, specific features, structures, or characteristics can be combined in one or more embodiments or examples in any suitable combination and / or sub-combination. Moreover, those skilled in the art will understand that the illustrations provided herein are for illustrative purposes and are not necessarily drawn to scale.

[0018] Please see Figure 1 To address the aforementioned problems, this invention provides a rapid optimization method for lithium-rich bauxite ores based on imaging hyperspectral imaging, such as... Figure 1 As shown, in one embodiment, the method includes the following steps: S1. Standard hyperspectral data were collected based on bauxite samples.

[0019] First, bauxite rock samples were collected, and indoor hyperspectral imaging data acquisition, lithium content determination, and data standardization were completed simultaneously to provide a high-quality data source for subsequent analysis. The specific steps include the following: S11. Sample pretreatment and powder preparation: Remove the surface weathering layer and impurities from the rock sample, and perform gradient grinding using an RS200 disc grinder. First, grind at 100 mesh for 30 seconds, and then finely grind at 200 mesh for 60 seconds to obtain a powder sample with uniform particle size. Anhydrous ethanol is used for cooling during the grinding process to avoid sample oxidation.

[0020] S12. Precise Acquisition of Hyperspectral Images: Building upon step S11, a constant temperature and humidity (temperature 20±1℃, relative humidity 35±5%) darkroom environment was established in the laboratory for acquisition. An AisaFENIX imaging spectrometer was used to acquire hyperspectral images of the rocks. During acquisition, a 6×35W halogen tungsten lamp was used as the light source, and the vertical distance between the spectrometer lens and the sample was 100cm to ensure uniform illumination. Spectral data included the 0.38-2.5μm range, with spectral sampling widths of 1.7nm and 5.45nm for the 0.38-0.97μm and 0.97-2.5μm bands, respectively, totaling 629 bands. S13. Accurate Lithium Content Determination and Data Calibration: 0.1g of powder sample was digested using a mixed acid system of hydrofluoric acid, nitric acid, and perchloric acid via microwave digestion. After digestion, the volume was adjusted to 50mL. The lithium content was determined using an Agilent 7900 inductively coupled plasma mass spectrometer (ICP-MS) at 21℃ and 40% relative humidity. One standard reference material (GBW07107) was inserted for every 10 samples for quality control, ensuring that the measurement error was less than 3%. Environmental parameters during hyperspectral acquisition and chemical analysis were recorded simultaneously, and a data-environmental parameter correlation table was established for subsequent outlier removal.

[0021] S2. Using the standard hyperspectral data, construct a multi-dimensional fused spectral dataset.

[0022] In this embodiment, constructing a multi-dimensional fused spectral dataset using the standard hyperspectral data includes the following steps: S21. Perform logarithmic transformation and continuum removal on the spectral reflectance data for each band.

[0023] The spectral reflectance data for each band are logarithmically transformed, and the calculation formula is as follows: In the formula, This represents the value of band i. Let be the logarithmic spectrum of band i.

[0024] Furthermore, the original spectral data undergoes continuum removal to eliminate the effects of spectral baseline drift; the logarithmic spectral data is then subjected to first derivative calculation (using a 5-point sliding window for smoothing) to enhance the resolution of spectral absorption peaks.

[0025] S22. Calculate the first derivative of the logarithmic spectral data, and continuum-removed spectrum and first derivative spectrum are concatenated according to band dimension to obtain a multi-dimensional fused spectral dataset.

[0026] Specifically, the original spectrum, logarithmic spectrum, continuum-removed spectrum, and first derivative spectrum are spliced ​​together according to the band dimension to construct a 4×288-dimensional multi-dimensional fused spectral data.

[0027] S3. Select the optimal feature band spectral data from the multi-dimensional fused spectral dataset.

[0028] In this embodiment, the step of selecting the optimal feature band spectral data from the multi-dimensional fused spectral dataset includes the following steps: S31. Combine the multi-dimensional fused spectral dataset with the corresponding lithium content data to form a sample set.

[0029] A sample set was formed by combining multi-dimensional fused spectral data with corresponding lithium content data. The Grubbs criterion (dynamic significance level) was used to remove outliers. Specifically, the first round of detection was performed with an initial significance level of 0.05. If the proportion of outliers exceeded 5%, the significance level was adaptively lowered to 0.01 for re-detection to avoid excessive removal of valid samples. The remaining samples were randomly divided into training and test sets in a 2:1 ratio. During the division process, stratified sampling combined with K-means pre-clustering was used to ensure that the lithium content distribution of the two sets of samples was consistent.

[0030] S32. Based on the sample set, construct a spectral-lithium content regression model using the partial least squares method to obtain the regression coefficients for each band.

[0031] The number of Monte Carlo samplings was set to N=500. Each time, 80% of the samples were randomly selected from the training set. The partial least squares (PLS) method was used to construct a spectrum-lithium content regression model, and the regression coefficients of each band were recorded during each modeling process.

[0032] S33. Calculate the weight of the first band based on the regression coefficient, and use the weight of the first band and the ratio of the number of retained bands to filter the first band set.

[0033] In this embodiment, the average absolute value of the regression coefficients for each band in 500 samples is calculated as the comprehensive weight of that band. The weight of the first band is calculated based on the regression coefficients, satisfying the following: in, Let be the absolute value of the regression coefficient of the j-th band in the i-th sampling. is the absolute value weight of the regression coefficient of the j-th band in the i-th sampling, and m is the number of remaining bands in each sampling.

[0034] Furthermore, bands with smaller absolute values ​​are removed using the attenuation index method. The first band set is then selected using the weight of the first band and the ratio of the number of retained bands, satisfying the following: in, This is the total number of bands. It is the minimum number of bands retained in the Nth sampling, where N is the number of samplings. It is the proportion of the number of bands retained in the i-th sampling. and These represent the initial coefficient and the decay rate coefficient of the decay index, respectively.

[0035] By combining the weight of the first band and the ratio of the number of retained bands, the bands are filtered by setting the retention ratio r to be greater than 0.8 and the weight to be greater than 1.5 times the comprehensive weight, and the first band set is obtained.

[0036] S34. The second band weights are assigned to the first band set using a two-factor adaptive weighting algorithm of variance contribution and information gain.

[0037] Specifically, the second band weights are assigned to the first band set using a two-factor adaptive weighting algorithm based on variance contribution and information gain, satisfying the following: in, This represents the variance-information gain fusion weight for the i-th band, where V is the variance contribution and G is the information gain. It is the variance of the product of variance and information gain.

[0038] The variance contribution satisfies in, This represents the variance of the x-th band, used to measure the information richness of that band.

[0039] The information gain satisfies: in, It is the entropy of variable Y. It is the conditional entropy of Y given the band x.

[0040] S35. Divide the sample set according to the lithium content to obtain multiple interval sample sets, and combine the second band weight and the interval sample sets to obtain the optimal characteristic band spectral data.

[0041] Based on the sample set, the lithium content was divided into three intervals: low (<1394ppm), medium (1394-3716ppm), and high (>3716ppm), and the corresponding interval sample sets were obtained.

[0042] Furthermore, the feature values ​​of each band for each sample are weighted according to the variance-information gain fusion weight, and the calculation formula is as follows: In the formula, It is the weighted value of the i-th band of the j-th sample. It is the spectral value of the i-th band of the j-th sample. It is the variance-information gain fusion weight of the i-th band.

[0043] Based on this, the weighted samples are further divided into training and test sets, a partial least squares regression model is constructed, and the lithium content of the test set is predicted; then the relative root mean square error of the whole sample set and the relative root mean square error of the samples in the low, medium and high intervals are calculated.

[0044] Record the band set and corresponding relative root mean square error after each screening, and combine them according to weight to obtain the comprehensive relative root mean square error. Select the band set with the smallest comprehensive relative root mean square error as the initial feature bands.

[0045] The Pearson correlation coefficient between each pair of bands in the initial band set is calculated, and highly redundant bands with a correlation coefficient greater than 0.85 are removed to obtain the optimal feature band set.

[0046] In this embodiment, the sample set is divided according to lithium content to obtain multiple interval sample sets. The optimal characteristic band spectral data is obtained by combining the second band weight and the interval sample sets, satisfying the following: in, This represents the overall relative root mean square error. This represents the relative root mean square error of the entire sample. This represents the relative root mean square error of samples in the low-content range. This represents the relative root mean square error of samples within the medium content range. This indicates the relative root mean square error of samples in the high-content range. In this example, the weighting coefficients are represented. .

[0047] Relative root mean square error ,satisfy: In the formula, , and These are the true value, predicted value, and average value of sample i, respectively.

[0048] S4. Construct a rapid optimization model for lithium ore based on the optimal characteristic band spectral data, and generate a thermal map of the spatial distribution of lithium content through the rapid optimization model for lithium ore.

[0049] In this embodiment, based on the optimal characteristic band spectral data, a random forest ensemble model integrating Bayesian optimization and leave-one-out cross-validation is constructed to improve the accuracy of lithium content estimation and model stability, including the following steps: S41. Feature dataset partitioning: Combine the optimal feature band spectral data obtained in step S3 with the corresponding lithium content data to form a feature sample set, and divide it into a training set and a test set in a 2:1 ratio. S42. Bayesian Optimization-Random Forest Model Construction: The training set samples are used to construct a random forest ensemble estimation model by combining leave-one-out cross-validation with the Bayesian optimization algorithm. Furthermore, let the number of training set samples be m, create m sub-training sets, each sub-training set contains m-1 samples, and use the remaining 1 sample as the validation set to ensure that each sample is used as the validation set once; Then, with R 2 Maximizing the objective function, the Bayesian optimization algorithm is used to optimize the key hyperparameters of the random forest (number of decision trees, maximum tree depth, minimum number of sample splits) to obtain the optimal combination of hyperparameters; Based on the optimal hyperparameters, the RandomForestRegressor class under the ensemble module in the Scikit-learn machine learning library is used to construct an independent random forest sub-model for each sub-training set and predict the lithium content value of the corresponding validation set.

[0050] Calculate the validation set R for all sub-models 2 The RMSE (Resolution Rate) is calculated by integrating the sub-models using a weighted average method to obtain a preliminary integrated model. This preliminary integrated model is then applied to the test set to calculate the coefficient of determination (R²) for the test set. 2The model is evaluated using the relative root mean square error (RRMSE) and relative analysis error (RPD). If RPD ≥ 2.5, the model is considered qualified; otherwise, the model is re-optimized by adjusting the hyperparameter range to obtain the best prediction model as the rapid selection model for lithium ore.

[0051] Furthermore, the rapid lithium ore selection model is applied to hyperspectral images of bauxite cores or blocks, and combined with spatial texture features to achieve graded selection and localization of lithium-rich ores, including the following steps: Acquire hyperspectral images of bauxite cores or blocks, perform dark current correction and radiometric calibration according to standards, and extract spectral data of corresponding bands from the images based on the optimal feature bands to form an image feature dataset. The image feature dataset is input into a rapid lithium ore optimization model to obtain the predicted lithium content value for each pixel in the image, generating a heatmap of the spatial distribution of lithium content. Simultaneously, the gray-level co-occurrence matrix texture features (contrast, entropy, correlation) of the image are extracted, and a correlation model between texture features and lithium content prediction error is established to correct the prediction results and reduce interference from foreign matter of the same spectrum. Based on economic indicators for bauxite lithium resource development and combined with extensive measured data, grading thresholds were established: ore with a lithium content less than 1394 ppm was classified as low-grade, ore with a lithium content between 1394 ppm and 3716 ppm as medium-grade, and ore with a lithium content greater than 3716 ppm as high-grade. Based on the corrected lithium content prediction results, the boundaries of different grade ores were marked on the heat map, and the spatial coordinates and area of ​​high-grade ore were output, providing accurate data for subsequent mining.

[0052] Please see Figure 2 In this embodiment, to efficiently execute the rapid selection method for lithium-rich bauxite ore based on imaging hyperspectral imaging provided by this invention, the present invention also provides a rapid selection system for lithium-rich bauxite ore based on imaging hyperspectral imaging, comprising: an input device 1, an output device 2, a processor 3, and a memory 4, wherein the input device 1, output device 2, processor 3, and memory 4 are interconnected, and the memory 4 contains program instructions for executing the steps of the rapid selection method for lithium-rich bauxite ore based on imaging hyperspectral imaging. The rapid selection system for lithium-rich bauxite ore based on imaging hyperspectral imaging of this invention has a compact structure and stable performance, and can stably execute the rapid selection method for lithium-rich bauxite ore based on imaging hyperspectral imaging of this invention, further enhancing the overall applicability and practical application capability of this invention.

[0053] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the present invention.

Claims

1. A method for rapid beneficiation of lithium-rich bauxite ores based on imaging hyperspectral, characterized in that, The method comprises the following steps: Collecting standard hyperspectral data according to bauxite sample collection standards; Using the standard hyperspectral data to construct a multi-dimensional fusion spectral data set; Screening optimal characteristic band spectral data from the multi-dimensional fusion spectral data set; Constructing a lithium ore rapid optimization model based on the optimal characteristic band spectral data, and generating a lithium content spatial distribution thermal map through the lithium ore rapid optimization model.

2. The method for rapid beneficiation of lithium-rich bauxite ores based on imaging hyperspectral according to claim 1, characterized in that, The step of using the standard hyperspectral data to construct a multi-dimensional fusion spectral data set comprises the following steps: Performing logarithmic transformation and continuum removal processing on spectral reflectance data of each band; Performing first-order derivative calculation on the logarithmic spectral data, and splicing the original spectrum, the logarithmic spectrum, the continuum removal spectrum and the first-order derivative spectrum according to the band dimension to obtain a multi-dimensional fusion spectral data set.

3. The method according to claim 1, wherein the method is characterized by, The step of screening optimal characteristic band spectral data from the multi-dimensional fusion spectral data set comprises the following steps: Combining the multi-dimensional fusion spectral data set and corresponding lithium content data to form a sample set; Constructing a spectrum-lithium content regression model through a partial least squares method based on the sample set to obtain regression coefficients of each band; Calculating first band weights according to the regression coefficients, and screening a first band set using the first band weights and a reserved band number ratio; Assigning second band weights to the first band set through a variance contribution-information gain double-factor adaptive weighting algorithm; Dividing the sample set according to lithium content to obtain multiple interval sample sets, and combining the second band weights and the interval sample sets to obtain optimal characteristic band spectral data.

4. The method according to claim 3, wherein the method is characterized by, The step of calculating the first band weights according to the regression coefficients satisfies: wherein, is the absolute value of the regression coefficient of the jth band in the ith sample, is the absolute value of the regression coefficient of the jth band in the ith sample, and m is the number of remaining bands in each sample.

5. The method for rapid beneficiation of lithium-rich bauxite ores based on imaging hyperspectral according to claim 3, characterized in that, The step of screening the first band set using the first band weights and the reserved band number ratio satisfies: wherein, is the total number of wavebands, is the minimum number of wavebands reserved for the Nth sampling, N being the number of sampling, is the ratio of the number of wavebands reserved for the i-th sampling, and respectively represent the initial coefficient of the decay exponent and the decay rate coefficient.

6. The method for rapid beneficiation of lithium-rich bauxite ores based on imaging hyperspectral according to claim 3, characterized in that, The step of assigning the second band weights to the first band set through the variance contribution-information gain double-factor adaptive weighting algorithm satisfies: wherein, is the variance-information gain fusion weight for the i-th band, V is the variance contribution, G is the information gain, is the variance of the variance and information gain product value.

7. The method for rapid beneficiation of lithium-rich bauxite ores based on imaging hyperspectral according to claim 6, characterized in that, The variance contribution satisfies wherein, represents the variance of the xth waveband, which measures the information richness of the waveband itself.

8. The method for rapid beneficiation of lithium-rich bauxite ores based on imaging hyperspectral according to claim 6, characterized in that, The information gain satisfies: wherein is the entropy of the variable Y, is the conditional entropy of Y given the waveband x.

9. The method for rapid beneficiation of lithium-rich bauxite ores based on imaging hyperspectral according to claim 3, characterized in that, The step of dividing the sample set according to lithium content to obtain multiple interval sample sets, and combining the second band weights and the interval sample sets to obtain optimal characteristic band spectral data satisfies: wherein, denotes the overall relative root mean square error, denotes the full-sample relative root mean square error, denotes the low-content interval sample relative root mean square error, denotes the medium-content interval sample relative root mean square error, denotes the high-content interval sample relative root mean square error, denotes the weight coefficient.

10. An imaging hyperspectral based bauxite lithium-rich ore rapid beneficiation system characterized in that, The bauxite lithium-rich ore rapid optimization system based on imaging hyperspectrum comprises an input device, an output device, a processor and a memory, the input device, the output device, the processor and the memory are connected to each other, the memory comprises program instructions, and the program instructions are used to execute the bauxite lithium-rich ore rapid optimization method based on imaging hyperspectrum in any one of claims 1 to 9.

Images