A method for identifying the origin of coix seed, a terminal device, and a storage medium

Abstract

The application discloses a method for identifying the producing area of coix seed, a terminal device and a storage medium, and comprises the following steps: S1: collecting a plurality of coix seed samples of the producing area to be identified, and obtaining a three-dimensional data array of the excitation emission matrix (EEM) fluorescence spectrum of the coix seed; S2: converting the three-dimensional data array of the EEM fluorescence spectrum into a two-dimensional data array of the sample set; S3: obtaining a training set and a test set by using the two-dimensional data array of the sample set; S4: training a model constructed by using an improved random forest algorithm by using the training set, and obtaining a producing area identification model; and S5: evaluating the performance of the model by using the test set and a prediction set. The method for identifying the coix seed from different producing areas based on the fluorescence spectrum technology realizes the classification of the coix seed from different producing areas by using the model constructed by combining the improved random forest algorithm and the fluorescence spectrum.

Description

TECHNICAL FIELD

[0001] The present application relates to a technology for provenance identification using machine learning, in particular to a method for identifying the provenance of coix seed, a terminal device and a storage medium. BACKGROUND

[0002] Coix seed is the dried mature seed of the annual herb coix, which contains 8 essential amino acids, various mineral elements, and is rich in various unsaturated fatty acids and polysaccharide substances. Coix seed has high medicinal value and has the effects of anti-tumor, anti-inflammatory and analgesic, heat-clearing and dampness-removing, blood circulation and stasis removal, blood glucose improvement, and immune system enhancement. In addition, coix seed contains rich vitamin E and other substances, and has beauty effects. Coix seed has a wide range of provenance, and the quality of coix seed from different provenances is significantly different, so it is necessary to develop a technology that can quickly identify the provenance of coix seed.

[0003] Existing coix seed provenance identification technologies, many scholars use various detection technologies such as gas chromatography-mass spectrometry (GC-MS), high performance liquid chromatography (HPLC), infrared spectroscopy (IR), and other chemical metrology methods to identify the provenance of coix seed. These methods can accurately distinguish coix seed from other provenances, but they have the disadvantages of long sample testing time, the need for large and expensive analytical instruments, and the consumption of more harmful organic reagents. Therefore, it is of great significance to establish a faster, more accurate, and more economical coix seed quality evaluation method to improve the coix seed quality evaluation system, protect consumer rights and interests, and promote the healthy and rapid development of the coix seed industry.

[0004] CN115436517A provides a linear discriminant method for provenance tracing of cloud production of wild beans, which uses GC-MS method, the experiment is relatively complex, and the linear discriminant model used has certain limitations, which has the disadvantages of long sample testing time, the need for large and expensive analytical instruments, and the consumption of more harmful organic reagents. SUMMARY

[0005] The purpose of the present application is to provide a method for identifying the provenance of coix seed, a terminal device and a storage medium to solve the above technical problems.

[0006] The present application adopts the following specific technical solutions:

[0007] A method for identifying the provenance of coix seed, comprising the following steps:

[0008] S1: Collecting a plurality of coix seed samples from the provenance to be identified, obtaining a three-dimensional data array of the excitation emission matrix (EEM) fluorescence spectrum of coix seed;

[0009] S2: preprocessing the three-dimensional data array of the EEM fluorescence spectrum, and converting the three-dimensional data array of the preprocessed EEM fluorescence spectrum into a two-dimensional data array of a sample set;

[0010] S3: obtaining a training set and a test set by using the two-dimensional data array of the sample set;

[0011] S4: obtaining a provenance identification model by training an improved random forest algorithm combined with a model constructed by a fluorescence spectrum by using the training set;

[0012] S5: evaluating the performance of the model by using the test set and the prediction set.

[0013] The EEM fluorescence spectrum of the coix seed extract is obtained, the three-dimensional data array of the EEM fluorescence spectrum is converted into a two-dimensional data array of a sample set, a training set is obtained based on the data set, and then a provenance identification model is obtained by training, which combines the EEM fluorescence spectrum method and the machine learning method, avoids the complex sample pretreatment process, and improves the efficiency and accuracy of coix seed provenance identification.

[0014] Further, in the present application, in order to obtain suitable fluorescence intensity and relatively complete spectrum in a short time, the sample in step S1 is processed as follows:

[0015] The collected sample is dissolved in a 70% ethanol solution, then ultrasonic extraction and centrifugal separation are performed, the supernatant after standing is collected, the supernatant is mixed with a 70% ethanol solution, and a coix seed powder extract is prepared for detecting the EEM fluorescence spectrum thereof; when detecting the EEM fluorescence spectrum of the sample, the excitation wavelength range of the fluorescence spectrometer is set to 200-450 nm, the emission wavelength range is set to 250-750 nm, the wavelength interval is 5 nm, the slit width is 5 nm, the coix seed powder extract is placed in a fluorescence dish of the fluorescence spectrometer for scanning, and a three-dimensional data array of the EEM fluorescence spectrum is obtained.

[0016] Further, when the collected sample is subjected to ultrasonic extraction and centrifugal separation, the ultrasonic time is 30 min, and centrifugal separation is performed at 5000 rpm for 10 min.

[0017] Further, in the present application, in order to further improve the efficiency and accuracy of coix seed provenance identification, the three-dimensional data array of the EEM fluorescence spectrum is preprocessed, and the content of preprocessing the three-dimensional data array of the EEM fluorescence spectrum in step S2 includes: selecting a set excitation wavelength and emission wavelength range, removing background data, and obtaining a three-dimensional data set of the EEM fluorescence spectrum.

[0018] Further, in the step S2, when converting the three-dimensional data array of the pretreated EEM fluorescence spectrum into a two-dimensional data array of the sample set, for each sample, the three-dimensional data array thereof is read row by row, and the three-dimensional data array is unfolded in a manner that the head and tail of each row are connected, so as to obtain a row of multi-column data corresponding to each sample, and a two-dimensional data array of the sample set is obtained.

[0019] Further, before the step S3 of obtaining the training set and the test set by using the two-dimensional data array of the sample set, the two-dimensional data array of the sample set is subjected to standard deviation normalization processing, so as to obtain a pretreated two-dimensional data array of the sample set.

[0020] Further, the step S4 of model includes PCA dimension reduction, feature extraction, random forest construction, and model optimization. The step of building the model in the step S4 is as follows: the pretreated two-dimensional data array of the sample set is split into a training set and a test set, the test set is subjected to PCA dimension reduction processing, feature extraction is realized, the number of retained PCR principal components is obtained, then the extracted features are used to construct a random forest (Gini coefficient), and finally, the number of retained PCR principal components and the parameters (the number of decision trees, the maximum depth of the decision tree, and the minimum number of leaf nodes) of the random forest are adjusted by using the test set, so as to obtain an optimal model.

[0021] Further, the specific step of evaluating the performance of the model by using the test set and the prediction set in the step S5 is as follows: the training set and the newly built test set are loaded into the model constructed by using the improved random forest algorithm combined with the fluorescence spectrum, and the accuracy, specificity, and sensitivity of the classification of the coix seed samples are calculated.

[0022] As an inventive concept, the present application also provides a terminal device, which comprises a memory, a processor, and a computer program stored in the memory; the processor executes the computer program to realize the steps of the method of the present application.

[0023] As an inventive concept, the present application also provides a computer readable storage medium, which stores a computer program / instruction; the computer program / instruction is executed by a processor to realize the steps of the method of the present application.

[0024] Compared with the prior art, the method for identifying the production area of coix seed based on the EEM fluorescence spectrum technology has the beneficial effects that the model constructed based on the improved random forest algorithm can well identify the coix seed from different production areas, the classification accuracy of the training set, the test set and the prediction set reaches 100%, and the sensitivity and specificity of the production area also reach 100%, fully proving that the model constructed based on the improved random forest algorithm combined with the fluorescence spectrum can quickly and reliably identify the production area of coix seed, provides a new idea for the production area tracing of traditional Chinese medicinal materials, provides a basis for accurately, quickly and reliably detecting the medicinal value and quality of coix seed, and promotes the healthy and rapid development of the coix seed industry. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only illustrate the embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of the provided drawings.

[0026] Figure 1 Flow chart for identifying the production area of coix seed based on EEM fluorescence spectrum technology combined with machine learning;

[0027] Figure 2 Spectrum diagram of EEM fluorescence spectrum of coix seed from nine production areas (Anhui, Fujian, Hebei, Heilongjiang, Jilin, Liaoning, Inner Mongolia, Shandong and Shaanxi);

[0028] Figure 3 Structure diagram of the model for identifying the production area of coix seed. DETAILED DESCRIPTION

[0029] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific implementation manner of the present application will be described in detail in the following with reference to the embodiments of the present application.

[0030] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, therefore the present application is not limited by the specific embodiments disclosed below. Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation manner of the present application. The "in one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0031] Embodiment 1

[0032] The EEM fluorescence spectrum technology based method for identifying the production area of coix seed provided by the present embodiment identifies coix seed from different production areas through the EEM fluorescence spectrum technology combined with the improved random forest algorithm, and the process includes:

[0033] The exemplary implementation methods of the present application will be described in more detail below, and these implementation methods are provided for the purpose of enabling a more thorough understanding of the present application and fully conveying the scope of the present application to those skilled in the art.

[0034] As shown in Figure 1 The EEM fluorescence spectrum technology based method for identifying the production area of coix seed provided by the present embodiment identifies coix seed from different production areas through the EEM fluorescence spectrum technology combined with the improved random forest algorithm, and the process includes:

[0035] S1: Collecting a plurality of coix seed samples to be identified for production area, obtaining a three-dimensional data array of the excitation emission matrix (EEM) fluorescence spectrum of the coix seed;

[0036] S2: Preprocessing the three-dimensional data array of the EEM fluorescence spectrum, and converting the three-dimensional data array of the preprocessed EEM fluorescence spectrum into a two-dimensional data array of the sample set;

[0037] After preprocessing the EEM fluorescence spectrum, the three-dimensional data array of each sample is read row by row, and the data is unfolded by connecting the head and tail of each row, obtaining a row of multi-column data corresponding to each sample, and a two-dimensional data array of the sample set is obtained, and then the two-dimensional data array is subjected to standard deviation standardization processing.

[0038] S3: Obtaining a training set and a test set using the two-dimensional data array of the sample set, and the specific operation is to collect the processed two-dimensional data array of the sample set along the dimension of the sample, obtaining the training set and the test set;

[0039] S4: Training the model constructed by the improved random forest algorithm using the training set, obtaining a production area identification model;

[0040] S5: Evaluating the performance of the model using the test set and the prediction set.

[0041] The present embodiment designs a model for identifying the production area of coix seed based on the improved random forest algorithm, which realizes the identification of coix seed samples from different production areas, and the specific operation is as follows:

[0042] The model for identifying the origin of coix seed is a model based on an improved random forest algorithm, including PCA dimension reduction, feature extraction, random forest construction, and model optimization. The final model for identifying the origin of coix seed is obtained. The specific model building steps in step S4 are as follows: the two-dimensional data array of the pretreated sample set is split into a training set and a test set, the test set is subjected to PCA dimension reduction processing, feature extraction is realized, the number of retained PCR principal components is obtained, then the extracted features are used to construct a random forest (Gini coefficient), and finally the number of retained PCR principal components and the parameters (number of decision trees, maximum depth of decision tree, and minimum number of leaf nodes) of the random forest are adjusted to obtain the optimal model.

[0043] After multiple model optimization, it is determined that the number of retained PCR principal components is 16, the number of decision trees is 100, the maximum depth of decision tree is 3, and the minimum number of leaf nodes is 1, the accuracy of the test set and the prediction set reaches 100%, and the sensitivity and specificity reach 100%, which fully proves that the model based on the EEM fluorescence spectrum technology combined with the improved random forest algorithm can quickly and reliably identify the origin of coix seed samples.

[0044] The following is a detailed description of the process of the model based on the EEM fluorescence spectrum technology combined with the improved random forest algorithm for identifying the origin of coix seed samples according to the specific embodiment.

[0045] 1. Experimental instruments and materials

[0046] Instruments: F-7000 fluorescence spectrometer; data analysis and programs used are run in python3.10.2 environment.

[0047] Materials: The coix seed sample powder used in the experiment comes from 9 origins of Anhui, Fujian, Hebei, Heilongjiang, Jilin, Liaoning, Inner Mongolia, Shandong, and Shaanxi. First, 24 samples from each origin were collected, a total of 216 samples were used for model construction, then 6 new samples from each origin were collected, a total of 54 samples were used for prediction set construction, and the specific information is shown in Table 1, Table 1 is a table of coix seed samples from different origins.

[0048] Table 1

[0049]

[0050] 2. Experimental method

[0051] 2.1 Sample processing and spectrum detection

[0052] Each sample powder was weighed 15 mg, dissolved in 1 mL 70% ethanol solution, ultrasonic for 30 min, then centrifuged at 5000 rpm for 10 min, after standing for 1 h, 150 μL supernatant was mixed with 1350 μL 70% ethanol solution, and then the EEM fluorescence spectrum was detected.

[0053] The fluorescence spectrometer was set as follows: excitation wavelength range 200-450 nm, emission wavelength range 250-750 nm, wavelength interval 5 nm, slit width 5 nm, voltage 700 V, and scanning speed 30000 nm·min-1. The fluorescence dish containing the extract of coix seed powder was placed in the support of the fluorescence spectrometer for scanning to obtain the EEM fluorescence spectrum.

[0054] 2.2 Data processing

[0055] 1) Obtaining EEM fluorescence spectrum data set

[0056] Each coix seed sample obtained an EEM fluorescence spectrum matrix with a size of 101×51 (number of emission wavelengths × number of excitation wavelengths) after fluorescence scanning. The EEM fluorescence spectrum array composed of the sample set of the samples for constructing the model was a three-dimensional data array with a size of 101×51×216 (number of emission wavelengths × number of excitation wavelengths × number of samples). The background data in the set spectral range were removed, and then the data were subjected to standard deviation standardization. Finally, the data were expanded into a row vector with a length of 5151 by connecting the head and tail of the row. The data of the sample set of 216 coix seed samples became a two-dimensional array with a size of 216×5151. The EEM fluorescence spectrum of the coix seed from nine producing areas (Anhui, Fujian, Hebei, Heilongjiang, Jilin, Liaoning, Inner Mongolia, Shandong, and Shaanxi) after removing the background is shown in Fig. 1. Figure 2

[0057] 2) Obtaining training set and test set

[0058] The sample set data array was collected in layers along the sample dimension to obtain the training set and the test set, and the ratio was 3:1. Table 2 shows the data set of coix seed from different producing areas.

[0059] Table 2

[0060]

[0061] 2.3 Establishment of coix seed producing area identification model

[0062] ​In machine learning, a random forest is a classifier containing multiple decision trees, and its output category is determined by the mode of the categories output by individual trees, and its prediction accuracy is higher than that of most single algorithms. However, the dimensionality of EEM fluorescence spectrum data is high (usually hundreds or thousands of data), and the calculation cost of directly using the random forest algorithm to construct a discriminant model based on the original EEM fluorescence spectrum data is too high, so the data dimension reduction method is combined with the random forest to reduce the number of unimportant features, thereby improving the training speed of the random forest discriminant model, and to a certain extent, the overfitting situation can be avoided. Embodiment 1 of the present application builds a model constructed by an improved random forest algorithm, including dimension conversion, standardization processing, PCA dimension reduction, feature extraction, random forest construction, model optimization. According to the classification accuracy results of the training set, the model parameters are optimized, the number of retained PCR principal components is set to 16, the number of decision trees is 100, the maximum depth of the decision tree is 3, and the minimum sample number of the leaf node is 1, and the specific structure is as shown in Figure 3

[0063] 2.4 Model evaluation

[0064] Accuracy, sensitivity and specificity are often used to evaluate the classification performance of the model. Among them, the accuracy is the proportion of the number of correctly classified samples to the total number of samples, which is used to evaluate the overall performance of the model; the sensitivity is the ability of the model to correctly identify a certain class of samples as that class; the specificity is the ability of the model to reject other class samples for a certain class of samples. The closer the accuracy, sensitivity and specificity are to 100%, the better the classification performance of the model, and the calculation formulas of the accuracy, sensitivity and specificity are as follows:

[0065]

[0066]

[0067]

[0068] Among them, ngg represents the number of samples belonging to g class and correctly identified as g class; G represents the number of categories; N represents the total number of samples; TP represents true positive; TN represents true negative; FN represents false negative; and FP represents false positive.

[0069] 3. Experimental results

[0070] ​Although the EEM fluorescence spectra of coix seed from different producing areas have certain differences, the differences are small enough to be distinguished by naked eyes, and still need more objective mathematical statistics for analysis. The coix seed producing area identification model proposed in Embodiment 1 of the present application can realize accurate and rapid identification of the coix seed producing area, the model is established by using the training set, and then the classification performance of the model is verified by using the test set and the prediction set, so as to realize identification of coix seed samples from different producing areas. The accurate classification rates of the corresponding data sets and the sensitivity and specificity of each sample are shown in Table 3, and Table 3 is the classification results of the coix seed producing area identification model. As can be seen from Table 3, the accuracy of the test set and the prediction set reaches 100%, and shows excellent sensitivity and specificity, especially the sensitivity and specificity of the test set and the prediction set are both 100%. In addition, the confusion matrix of the test set and the prediction set is shown in Table 4, and Table 4 is the confusion matrix of the test set and the prediction set obtained by the coix seed producing area identification model, and from the table it can be seen that all coix seed samples are correctly determined as the corresponding producing area by the coix seed producing area identification model.

[0071] Table 3

[0072]

[0073]

[0074] Table 4

[0075]

[0076]

[0077] In summary, Embodiment 1 of the present application proposes a method for identifying the producing area of coix seed based on EEM fluorescence spectrum technology, and the model constructed based on the improved random forest algorithm well realizes identification of coix seed from different producing areas, wherein the classification accuracy of the test set and the prediction set reaches 100%, and in addition, the sensitivity and specificity of each producing area also reaches 100%, which can fully prove that the model constructed based on the EEM fluorescence spectrum technology combined with the improved random forest algorithm can quickly and reliably identify the producing area of coix seed, and provides a new idea for identification of the producing area of traditional Chinese medicinal materials.

[0078] Embodiment 2

[0079] Embodiment 2 of the present application provides a terminal device corresponding to the above-mentioned Embodiment 1, which can be a processing device for a client, such as a mobile phone, a notebook computer, a tablet computer, a desktop computer, etc., to execute the method of the above-mentioned Embodiment.

[0080] The terminal device of the embodiment comprises a memory, a processor and a computer program stored in the memory; the processor executes the computer program stored in the memory to realize the method steps of the above-mentioned embodiment 1.

[0081] Further, the memory can be a high-speed random access memory (RAM), and can also comprise a non-volatile memory, for example at least one disk memory.

[0082] Further, the processor can be a central processing unit (CPU), a digital signal processor (DSP) or various types of general-purpose processors, which are not limited herein.

[0083] Embodiment 3

[0084] The embodiment 3 of the present application provides a computer readable storage medium corresponding to the above-mentioned embodiment 1, which stores a computer program / instruction. The computer program / instruction is executed by the processor to realize the method steps of the above-mentioned embodiment 1.

[0085] The computer readable storage medium can be a tangible device that maintains and stores instructions for use by an instruction execution device. The computer readable storage medium can be, for example but not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any combination thereof.

[0086] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROMs, optical storage devices, etc.) containing computer usable program code. The solutions in the embodiments of the present application can be implemented in various computer languages, for example, object-oriented programming languages such as Java and interpreted scripting languages such as JavaScript.

[0087] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems) and computer program products of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices produce a device that implements the functions described in the flowcharts and / or block diagrams. Figure 1 one flow or multiple flows and / or blocks Figure 1means for performing the function specified by the block or blocks.

[0088] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the operations specified in the flowchart block or blocks. Figure 1 one or more flowcharts and / or blocks Figure 1 steps of a function specified in one or more blocks.

[0089] Obviously, the above-described embodiments of the application are only examples for clearly illustrating the application, but not to limit the implementation of the application. Based on the above description, other different forms of changes or variations can be made by those skilled in the art. Here, all the implementations are not required to be exhausted. Any modification, equivalent replacement and improvement within the spirit and principle of the application should be included in the protection scope of the claims of the application.

Claims

1. A method for identifying the origin of seeds of Coix lacryma-jobi, characterized by, The method comprises the following steps: S1: Collecting a plurality of seed samples of coix seed from different producing areas to obtain a three-dimensional data array of the EEM fluorescence spectrum of the coix seed; S2: Preprocessing the three-dimensional data array of the EEM fluorescence spectrum, and converting the three-dimensional data array of the preprocessed EEM fluorescence spectrum into a two-dimensional data array of the sample set; S3: Obtaining a training set and a test set by using the two-dimensional data array of the sample set; S4: Training an improved random forest algorithm combined with a model constructed by the fluorescence spectrum by using the training set to obtain a producing area identification model; S5: Evaluating the performance of the model by using the test set and a prediction set, wherein the prediction set is composed of new samples obtained by secondary collection of coix seed from the producing areas; The model of step S4 comprises PCA dimension reduction, feature extraction, random forest construction, model optimization, and the specific steps of building the model are as follows: the preprocessed two-dimensional data array of the sample set is split into a training set and a test set, the test set is subjected to PCA dimension reduction processing, feature extraction is performed to obtain the number of retained PCR principal components, the extracted features are used to construct a random forest, and the number of retained PCR principal components and the parameters of the random forest are adjusted by the test set to obtain an optimal model.

2. The method of claim 1, wherein the method is characterized by, In step S1, the sample is treated as follows: The collected sample is dissolved in a 70% ethanol solution, then subjected to ultrasonic extraction and centrifugal separation, the supernatant after standing is collected, the supernatant is mixed with a 70% ethanol solution to prepare a coix seed powder extract for detecting the EEM fluorescence spectrum thereof, when detecting the EEM fluorescence spectrum, the excitation wavelength range of the fluorescence spectrometer is set to 200-450 nm, the emission wavelength range is set to 250-750 nm, the wavelength interval is 5 nm, the slit width is 5 nm, the coix seed powder extract is placed in a fluorescence dish of the fluorescence spectrometer for scanning to obtain a three-dimensional data array of the EEM fluorescence spectrum.

3. The method of claim 2, wherein the seeds are from a plant of the species Coix lacryma-jobi L. var. ma-yuen (Romer) Stapf. When the collected sample is subjected to ultrasonic extraction and centrifugal separation, the ultrasonic time is 30 min, and the centrifugal separation is performed at 5000 rpm for 10 min.

4. The method of claim 1, wherein the method is characterized by, In step S2, the preprocessing of the three-dimensional data array of the EEM fluorescence spectrum comprises the following steps: selecting a set of excitation wavelength and emission wavelength ranges, and removing background data to obtain a three-dimensional data set of the EEM fluorescence spectrum.

5. The method of claim 1, wherein the method is characterized by, In step S2, when the three-dimensional data array of the preprocessed EEM fluorescence spectrum is converted into a two-dimensional data array of the sample set, the three-dimensional data array of each sample is read row by row, and the row is expanded to obtain a row of multi-column data corresponding to each sample, and a two-dimensional data array of the sample set is obtained.

6. The method of claim 1, wherein the method is characterized by, Before step S3, the two-dimensional data array of the sample set is subjected to standard deviation standardization processing to obtain a preprocessed two-dimensional data array of the sample set.

7. The method of claim 1, wherein the method is characterized by, In step S5, the specific steps of evaluating the performance of the model by using the test set and the prediction set are as follows: loading the training set and the newly established test set into the model constructed by the improved random forest algorithm combined with the fluorescence spectrum, and calculating the accuracy, specificity and sensitivity of the classification of the coix seed samples.

8. A terminal device comprising a memory, a processor, and a computer program stored on the memory; characterized in that, The processor executes the computer program to implement the method of any one of claims 1-7.

9. A storage medium, characterized by The storage medium stores computer programs / instructions, and the computer programs / instructions are executed by a processor to implement the method for identifying the production area of coix seed according to any one of claims 1-7.

Images