Plastic classification method, device and equipment based on spectral information and medium

Abstract

The invention discloses a plastic classification method and device based on spectral information, equipment and a medium, and the method comprises the steps: carrying out the preliminary classification of to-be-classified plastic through a pre-trained plastic classification model, and carrying out the comparison of preliminary classification results through a spectral information divergence method on the basis of the preliminary classification, and obtaining a final classification result of the to-be-classified plastic. According to the plastic classification method and device, on the basis that the to-be-classified plastic is classified through the pre-trained plastic classification model, classification is conducted again, and the accuracy of plastic classification can be improved.

Description

Technical Field

[0001] The present application relates to the field of plastic classification, and in particular to a plastic classification method, device, equipment and medium based on spectral information. Background Art

[0002] Plastic classification plays an important role in many fields such as the recycling industry, quality control, and environmental monitoring. The existing plastic classification mainly uses the PLS-DA (partial least squares discriminant analysis) model. This model is greatly affected by outliers and noise during the modeling process, resulting in low accuracy in plastic classification. Summary of the Invention

[0003] The following is a summary of the subject matter described in detail herein. This summary is not intended to limit the scope of the claims.

[0004] The main purpose of the embodiments of the present disclosure is to propose a plastic classification method, device, equipment and medium based on spectral information, which can improve the accuracy of plastic classification.

[0005] A first aspect of an embodiment of the present application provides a plastic classification method based on spectral information, the method comprising:

[0006] Obtain hyperspectral data of plastics to be classified;

[0007] Inputting the hyperspectral data into a plastic classification model to obtain preliminary classification results;

[0008] The preliminary classification results are compared using a spectral information divergence method to obtain a final classification result of the plastic to be classified.

[0009] The present embodiment provides a spectral information-based plastic classification method. A pre-trained plastic classification model is used to perform a preliminary classification of the plastic to be classified. Based on this preliminary classification, the preliminary classification results are compared using the spectral information divergence method to obtain a final classification result for the plastic to be classified. This shows that the present application further classifies the plastic to be classified based on the classification performed by the pre-trained plastic classification model, thereby improving the accuracy of plastic classification.

[0010] In some embodiments of the present application, before inputting the hyperspectral data of the plastic to be classified into a plastic classification model to obtain a preliminary classification result, the method further includes:

[0011] Obtain spectral data of plastic samples;

[0012] Based on the plastic sample spectral data, a plastic sample standard spectral database is established;

[0013] Based on the plastic sample standard spectrum database, a support vector machine model is trained to obtain the plastic classification model.

[0014] In some embodiments of the present application, establishing a plastic sample standard spectrum database based on the plastic sample spectrum data includes:

[0015] Preprocessing the plastic sample spectral data to obtain standard plastic sample spectral data;

[0016] Based on the plastic sample standard spectrum data, the plastic sample standard spectrum database is established.

[0017] In some embodiments of the present application, the preprocessing of the plastic sample spectral data to obtain standard plastic sample spectral data includes:

[0018] Taking the mean of the spectral data of the plastic sample as a reference spectrum;

[0019] Fitting the spectral data of the plastic sample based on a linear regression method to obtain a linear model of the spectral data of the plastic sample;

[0020] Calculating a scaling factor of the plastic sample spectral data by covariance based on the reference spectrum;

[0021] Based on the linear model and the scaling factor, multivariate scattering correction is performed on the plastic sample spectral data to obtain the plastic sample standard spectral data.

[0022] In some embodiments of the present application, the support vector machine model is trained based on the plastic sample standard spectrum database to obtain the plastic classification model, including:

[0023] Performing dimension expansion on the standard spectral data of the plastic sample based on a Gaussian kernel function;

[0024] The support vector machine model is trained using the expanded standard spectrum data of the plastic samples to obtain the plastic classification model.

[0025] In some embodiments of the present application, the method of comparing the preliminary classification results using the spectral information divergence method to obtain the final classification result of the plastic to be classified includes:

[0026] Comparing the preliminary classification result with the spectral information divergence of classified samples in the plastic sample standard spectral database using the spectral information divergence method;

[0027] Based on the compared spectral information divergence and the preliminary classification result, a final classification result of the plastic to be classified is obtained.

[0028] In some embodiments of the present application, obtaining hyperspectral data of the plastic to be classified includes:

[0029] The hyperspectral data of the plastics to be classified were collected using a SpecimFX17 hyperspectral camera.

[0030] A second aspect of an embodiment of the present invention provides a plastic classification device based on spectral information, the device comprising:

[0031] A data acquisition module, used to acquire hyperspectral data of plastics to be classified;

[0032] A preliminary classification module, configured to input the hyperspectral data into a plastic classification model to obtain preliminary classification results;

[0033] The final classification module compares the preliminary classification results by using the spectral information divergence method to obtain the final classification results of the plastic to be classified.

[0034] A third aspect of an embodiment of the present invention provides an electronic device, comprising: at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor to enable the at least one control processor to execute the above-mentioned plastic classification method based on spectral information.

[0035] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the above-mentioned plastic classification method based on spectral information.

[0036] It can be understood that the beneficial effects of the second to fourth aspects compared with the relevant technologies are the same as the beneficial effects of the first aspect compared with the relevant technologies. Please refer to the relevant description in the first aspect and no further details will be given here. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0038] Figure 1 This is a schematic diagram of a process for plastic classification based on spectral information provided in an embodiment of the present application;

[0039] Figure 2 This is a structural diagram of a plastic classification device based on spectral information provided in an embodiment of the present application;

[0040] Figure 3It is a structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0041] The following describes in detail embodiments of the present application. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application and are not to be construed as limiting the present application.

[0042] In the description of this application, if there is a description of first, second, etc., it is only for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0043] In the description of this application, it should be understood that descriptions involving orientation, such as the orientation or positional relationship indicated by up, down, etc., are based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on this application.

[0044] In the description of this application, it should be noted that, unless otherwise clearly defined, terms such as setting, installing, and connecting should be understood in a broad sense, and technical personnel in the relevant technical field can reasonably determine the specific meaning of the above terms in this application based on the specific content of the technical solution.

[0045] Existing plastic classification methods mainly use the PLS-DA (Partial Least Squares Discriminant Analysis) algorithm (using the PLS-DA model to establish the association between hyperspectral data and plastic types), using spectral feature bands to classify plastics (using the unique spectral information of each plastic in the near-infrared NIR (900-1700nm) spectral region, and then classifying plastics through a classifier algorithm), and using deep learning algorithms for classification (using convolutional neural networks and recurrent neural networks for training and classification). These methods all have different shortcomings, specifically:

[0046] The PLS-DA (Partial Least Squares Discriminant Analysis) algorithm can be significantly affected by outliers and noise during the modeling process. Spectral data may contain outliers or noise due to factors such as measurement conditions and equipment errors. These values ​​can distort the model's learning process and lead to reduced model performance. Furthermore, the performance of the PLS-DA model depends largely on parameter selection, such as the number of latent variables (n_components). The selection of these parameters typically requires experience, trial and error, or cross-validation, which increases the difficulty and complexity of model construction. Secondly, the PLS-DA algorithm primarily models linear relationships and does not adequately address potential nonlinear relationships in spectral data. This can lead to poor model performance on complex data or data with strong nonlinear relationships. Furthermore, while the PLS-DA algorithm may perform well on the training set, its generalization ability on test sets or new datasets may be limited due to overfitting of the training data or its high sensitivity to changes in the data distribution. This can result in insufficient accuracy in plastic classification for practical applications.

[0047] Plastic classification using spectral signature bands is a method that requires large amounts of data, increasing the complexity and computational cost of data processing because spectral data from plastics of different materials and compositions may vary within the same spectral range. Furthermore, spectral data is significantly affected by various environmental factors, such as lighting and atmosphere. These factors can cause fluctuations and distortions in spectral data, affecting the accuracy of classification results. Furthermore, the diversity and complexity of spectral data make it challenging to build a robust classification model. Different application scenarios may require different classification models and parameter settings, increasing the difficulty of model development and debugging.

[0048] Using deep learning algorithms to classify plastics typically requires a large amount of data to train the model. In hyperspectral image classification, the limited availability and uneven quality of labeled samples can lead to inaccurate estimation of model parameters, which in turn affects classification accuracy. Furthermore, the high dimensionality and data redundancy of hyperspectral images increase the demand for training data. Furthermore, deep learning algorithms lack the ability to explain the model's internal operating mechanisms, making it difficult to interpret the model's predictions. Furthermore, due to the complexity and nonlinear relationships of hyperspectral image data, optimizing deep learning models can be challenging.

[0049] For this purpose, refer to Figure 1 The embodiment of the present application provides a plastic classification method based on spectral information. The method is applied to a controller, which may be a server, an electronic device, or a mobile terminal, etc., and is not specifically limited here. The method includes the following steps S101 to S103:

[0050] Step S101: Acquire hyperspectral data of plastics to be classified.

[0051] In this step, the Specim FX17 hyperspectral camera covers the visible light and near-infrared bands from 400 nm to 1000 nm, can provide detailed spectral information, and provides high spectral resolution, which can distinguish subtle spectral differences between different substances. In this example, the Specim FX17 hyperspectral camera is used to collect hyperspectral data of the plastics to be classified.

[0052] Step S102: Input the hyperspectral data into the plastic classification model to obtain preliminary classification results.

[0053] In this step, the hyperspectral data is preprocessed and a plastic classification model is pre-trained. The pre-processed hyperspectral data is classified using the pre-trained plastic classification model to obtain preliminary classification results for the plastics to be classified. The pre-trained plastic classification model is a support vector machine model, which is highly robust to noise and outliers, and has stronger generalization capabilities when dealing with small samples and nonlinear problems.

[0054] Step S103: Compare the preliminary classification results using the spectral information divergence method to obtain the final classification results of the plastics to be classified.

[0055] In this step, the spectral information divergence method (SID) quantifies the similarity between two spectral curves by calculating the divergence between them, providing a clear and comparable standard, which helps to more accurately analyze and compare spectral data. In the case of a small plastic sample database and unknown sources, it can make plastic classification more accurate; therefore, after classifying the plastics to be classified, this embodiment performs a secondary classification using the spectral information divergence method to further improve the accuracy of plastic classification.

[0056] Unlike the prior art, the present embodiment uses a pre-trained plastic classification model to perform preliminary classification of the plastics to be classified. Based on this preliminary classification, the spectral information divergence method is used to compare the preliminary classification results to obtain the final classification results of the plastics to be classified. This shows that the present embodiment further classifies the plastics to be classified based on the classification of the plastics to be classified using the pre-trained plastic classification model, thereby improving the accuracy of plastic classification.

[0057] The specific implementation methods of the above steps are introduced below.

[0058] Before the hyperspectral data of the plastic to be classified in step S102 is input into the plastic classification model to obtain a preliminary classification result, the method further includes the following steps S201 to S203:

[0059] Step S201, obtaining spectral data of a plastic sample;

[0060] Step S202: establishing a plastic sample standard spectrum database based on the plastic sample spectrum data;

[0061] Step S203: Based on the plastic sample standard spectrum database, the support vector machine model is trained to obtain a plastic classification model.

[0062] In this embodiment, the hyperspectral data of the plastics to be classified is collected using a Specim FX17 hyperspectral camera. During collection, a halogen lamp provides a constant light source for the plastics. A uniform reflectance whiteboard is used for calibration. A laboratory bench is used to ensure high uniformity from the plastic to the camera, thereby ensuring the quality of the collected spectral data of the plastic samples.

[0063] In this embodiment, a standard spectrum database of plastic samples is constructed based on the collected spectrum data of plastic samples, wherein the standard spectrum database of plastic samples contains spectrum characteristics of different types of plastics and corresponding category labels.

[0064] In some embodiments, the standard spectral database of plastic samples is divided into a training set and a test set, and a support vector machine model (SVM) is trained using the training set. During the training process, the SVM attempts to find an optimal hyperplane so that plastic samples of different categories are separated as much as possible in the spectral feature space, thereby achieving the purpose of classification. After the training is completed, the trained SVM is tested and verified using the test set to further improve the accuracy of the plastic classification model.

[0065] The step S202 of establishing a standard spectrum database of plastic samples based on the spectrum data of the plastic samples includes the following steps S301 to S302:

[0066] Step S301: pre-processing the spectral data of the plastic sample to obtain standard spectral data of the plastic sample;

[0067] Step S302: establishing a plastic sample standard spectrum database based on the plastic sample standard spectrum data.

[0068] In this embodiment, the spectral data of the plastic sample is preprocessed by a multivariate scattering correction method to eliminate the spectral baseline drift and amplitude variation caused by the scattering effect and particle size difference, so as to improve the quality of the spectral data and ensure the accuracy of subsequent modeling and analysis.

[0069] The preprocessing of the plastic sample spectral data in step S301 to obtain the standard spectral data of the plastic sample includes the following steps S401 to S404:

[0070] Step S401: taking the mean value of the spectral data of the plastic sample as a reference spectrum;

[0071] Step S402: fitting the spectral data of the plastic sample based on a linear regression method to obtain a linear model of the spectral data of the plastic sample;

[0072] Step S403: Calculate the scaling factor of the plastic sample spectrum data through covariance based on the reference spectrum;

[0073] Step S404: Based on the linear model and the scaling factor, perform multivariate scattering correction on the spectral data of the plastic sample to obtain standard spectral data of the plastic sample.

[0074] In this embodiment, the spectral data of all plastic samples are used as the reference spectrum, wherein the reference spectrum can be expressed by formula (1):

[0075]

[0076] Among them, X ref is the reference spectrum, n is the number of plastic sample spectra, X i is the i-th spectrum data (a vector).

[0077] Linear regression was performed on each plastic sample spectrum, that is, the linear model corresponding to each plastic sample spectrum was fitted using the least squares method through formula (2):

[0078]

[0079] Among them, a i is the scaling factor, b i is the regression coefficient, a i It can be calculated using formula (3), b i It can be calculated using formula (4):

[0080]

[0081] The strength of the linear relationship between two variables is shown by covariance, which can be expressed by formula (5):

[0082]

[0083] Among them, X ij is the value of the i-th spectrum data at the j-th wavelength, X ref,j is the value of the reference spectrum at the jth wavelength, is the mean of the ith spectral data, is the mean of the reference spectrum, and m represents the number of bands.

[0084] Secondly, the variance of the reference spectrum can be expressed by formula (6):

[0085]

[0086] Among them, Var(X ref ) represents the reference spectrum X ref The variance of X ref [k] is the value of the reference spectrum at the kth sample, is the mean of the reference spectrum, and n is the number of data points (i.e., the number of samples) of the reference spectrum.

[0087] In this embodiment, the plastic sample spectrum X is measured by covariance i and reference spectra A high covariance value indicates that the two spectra have similar variation trends at the same wavelength.

[0088] In this embodiment, the covariance is used to calculate the scaling factor a of each plastic sample spectrum. i The measured spectrum is adjusted to the same scale as the reference spectrum by the scaling factor; and by adjusting the scaling factor, the scattering effect in the spectral data can be eliminated, so that the measured spectrum is better aligned with the reference spectrum.

[0089] In step S203, the support vector machine model is trained based on the plastic sample standard spectrum database to obtain a plastic classification model, which includes the following steps S501 to S502:

[0090] Step S501: performing dimension expansion on the standard spectrum data of the plastic sample based on a Gaussian kernel function;

[0091] Step S502: Use the expanded standard spectrum data of the plastic sample to train the support vector machine model to obtain the plastic classification model.

[0092] In this embodiment, the kernel function of the support vector machine model adopts the Gaussian kernel function, which can be expressed by formula (7):

[0093]

[0094] Here, x1 and x2 are two points in feature space, ||x1-x2|| is the Euclidean distance between the two points, and γ is a parameter in the Gaussian kernel function that determines the kernel width. It controls the size of the range of influence of a data point. Smaller values ​​of γ result in a wider Gaussian function, while larger values ​​of γ make the Gaussian function narrower, meaning that each data point has a smaller local influence.

[0095] The exponential function part in the Gaussian kernel function can be expanded by Taylor, and then the Gaussian kernel function can be extended to infinite dimensions, as shown in formula (8):

[0096]

[0097] Among them, e is a natural constant (approximately 2.71828...), R n is the remainder of the nth-order Taylor polynomial.

[0098] Expanding equation (8) using Taylor's formula yields equation (9):

[0099]

[0100] Here, x1 and x2 are two points in the feature space, and σ is the width parameter of the Gaussian kernel function, which controls the "width" or range of influence of the kernel function. A larger σ value means a wider peak of the kernel function and a faster decay; a smaller σ value produces a sharper peak and a slower decay rate. φ(x) is a feature mapping function that maps the input point x1 or x2 to a point in high-dimensional space. This mapping allows the distance between the two input points to be measured by their inner product in the high-dimensional space, thus achieving a nonlinear similarity measure.

[0101] In this embodiment, the Gaussian kernel function can be represented as an inner product in an infinite-dimensional feature space through Taylor expansion. This enables linear operations in high-dimensional space, effectively applying the "kernel technique." The kernel technique allows complex nonlinear transformations to be performed using the kernel function in the original feature space without explicitly calculating the features in the high-dimensional space, improving the accuracy of spectral data processing. Finally, different plastics can be classified in the kernel space using a hyperplane.

[0102] Comparing the preliminary classification results by the spectral information divergence method in step S103 to obtain the final classification results of the plastics to be classified includes steps S601 to S602:

[0103] Step S601: using the spectral information divergence method, comparing the preliminary classification result with the spectral information divergence of classified samples in the plastic sample standard spectral database;

[0104] Step S602: Obtain a final classification result of the plastic to be classified based on the compared spectral information divergence and the preliminary classification result.

[0105] In this implementation, spectral data can be affected by the environment, the number of database samples, and the quality of the training samples. Therefore, for more accurate classification, the spectral vectors are treated as random variables and the similarity between two random vectors is analyzed based on probability and statistics theory. Specifically, the smaller the spectral information divergence, the more similar the two sets of spectra are.

[0106] Specifically, it can be expressed by formula (10), assuming that x and y represent the target spectrum and the test spectrum respectively, and the band is n:

[0107] SID(x,y)=D(x||y)+D(y||x) (10)

[0108] in:

[0109]

[0110]

[0111]

[0112] I i (x)=-lga i (15)

[0113] I i (x) = -lgb i (16)

[0114] In this embodiment, the preliminary classification results of the plastic to be classified and the spectral information divergence of the classified samples in the standard spectral database of plastic samples are calculated by the above formulas (10) to (16) to solve the problem of possible classification errors during the preliminary classification and further improve the accuracy of the classification of the plastic to be classified.

[0115] In this embodiment, after obtaining the plastic sample spectrum, the plastic sample spectrum is preprocessed by a multivariate scattering correction method to eliminate the influence of uneven light source and surface scattering in the sample, so that the measured spectrum is more accurate and reliable; a plastic sample standard spectrum database is constructed based on the preprocessed plastic sample spectrum, and the support vector machine model is trained by the training set in the plastic sample standard spectrum database. During the training process, the Gaussian kernel function is used to make linear operations in high-dimensional space possible, and the plastic is effectively classified to obtain a trained plastic classification model. The plastic to be classified is input into the plastic classification model to obtain a preliminary classification result, and then the plastic is further classified by the spectral information scattering method to make the classification result more accurate.

[0116] like Figure 2 As shown, some embodiments of the present application provide a plastic classification device based on spectral information, the device including a data acquisition module 1100, a preliminary classification module 1200, and a final classification module 1300. Specifically:

[0117] A data acquisition module, used to acquire hyperspectral data of plastics to be classified;

[0118] A preliminary classification module, configured to input the hyperspectral data into a plastic classification model to obtain preliminary classification results;

[0119] The final classification module compares the preliminary classification results by using the spectral information divergence method to obtain the final classification results of the plastic to be classified.

[0120] It should be noted that the plastic classification device based on spectral information provided in this embodiment and the above-mentioned plastic classification method based on spectral information are based on the same inventive concept. Therefore, the relevant content of the above-mentioned plastic classification method based on spectral information is also applicable to the content of the plastic classification device based on spectral information. Therefore, it will not be repeated here.

[0121] like Figure 3 , an embodiment of the present application further provides an electronic device, the electronic device comprising:

[0122] At least one hydrogen fuel cell;

[0123] at least one memory;

[0124] at least one processor;

[0125] at least one program;

[0126] The programs are stored in the memory, and the processor executes at least one program to implement the above-mentioned plastic classification method based on spectral information in the present disclosure.

[0127] The electronic device may be any intelligent terminal including a mobile phone, a tablet computer, a personal digital assistant (PDA), a car computer, etc.

[0128] The electronic device according to the embodiment of the present application is described in detail below.

[0129] The processor 1600 may be implemented as a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is configured to execute relevant programs to implement the technical solutions provided by the embodiments of the present disclosure.

[0130] Memory 1700 can be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). Memory 1700 can store an operating system and other application programs. When implementing the technical solutions provided in the embodiments of this specification through software or firmware, the relevant program code is stored in memory 1700 and is called by processor 1600 to execute the plastic classification method based on spectral information in the embodiments of this disclosure.

[0131] Input / output interface 1800, used for information input and output;

[0132] Communication interface 1900, used to implement communication interaction between this device and other devices, which can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WiFi, Bluetooth, etc.);

[0133] bus 2000 , which transmits information between various components of the device (e.g., processor 1600 , memory 1700 , input / output interface 1800 , and communication interface 1900 );

[0134] The processor 1600 , the memory 1700 , the input / output interface 1800 , and the communication interface 1900 are connected to each other in communication within the device via the bus 2000 .

[0135] An embodiment of the present disclosure also provides a storage medium, which is a computer-readable storage medium. The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the above-mentioned plastic classification method based on spectral information.

[0136] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0137] The embodiments described in the embodiments of the present disclosure are intended to more clearly illustrate the technical solutions of the embodiments of the present disclosure and do not constitute a limitation on the technical solutions provided by the embodiments of the present disclosure. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of the present disclosure are also applicable to similar technical problems.

[0138] Those skilled in the art will understand that the technical solutions shown in the drawings do not constitute a limitation on the embodiments of the present disclosure, and may include more or fewer steps than shown in the drawings, or a combination of certain steps, or different steps.

[0139] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, i.e., they may be located in one place or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of this embodiment.

[0140] Those skilled in the art will appreciate that all or some of the steps in the methods, systems, and functional modules / units in the devices disclosed above may be implemented as software, firmware, hardware, or appropriate combinations thereof.

[0141] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0142] It should be understood that in this application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.

[0143] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0144] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0145] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0146] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes multiple instructions for enabling an electronic device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes various media that can store programs, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0147] The above is a specific description of the preferred implementation of the embodiments of the present application, but the embodiments of the present application are not limited to the above-mentioned implementation methods. Technical personnel familiar with the art can also make various equivalent modifications or substitutions without violating the spirit of the embodiments of the present application. These equivalent modifications or substitutions are all included in the scope defined by the claims of the embodiments of the present application.

[0148] The embodiments of the present application are described in detail above in conjunction with the accompanying drawings, but the present application is not limited to the above embodiments. Various changes can be made within the scope of knowledge possessed by ordinary technicians in the relevant technical field without departing from the purpose of the present application.

Claims

1. A plastic classification method based on spectral information, characterized in that: The method comprises: Obtain hyperspectral data of plastics to be classified; Inputting the hyperspectral data into a plastic classification model to obtain a preliminary classification result; The preliminary classification results are compared by using a spectral information divergence method to obtain a final classification result of the plastic to be classified.

2. The plastic classification method based on spectral information according to claim 1, characterized in that: Before inputting the hyperspectral data of the plastic to be classified into a plastic classification model to obtain a preliminary classification result, the method further includes: Obtain spectral data of plastic samples; Based on the plastic sample spectral data, a plastic sample standard spectral database is established; Based on the plastic sample standard spectrum database, a support vector machine model is trained to obtain the plastic classification model.

3. The plastic classification method based on spectral information according to claim 2, characterized in that: The step of establishing a standard spectrum database of plastic samples based on the spectrum data of the plastic samples comprises: Preprocessing the spectral data of the plastic sample to obtain standard spectral data of the plastic sample; Based on the plastic sample standard spectrum data, the plastic sample standard spectrum database is established.

4. The plastic classification method based on spectral information according to claim 3, characterized in that: The preprocessing of the plastic sample spectral data to obtain the plastic sample standard spectral data includes: Taking the mean of the spectral data of the plastic sample as a reference spectrum; Fitting the spectral data of the plastic sample based on a linear regression method to obtain a linear model of the spectral data of the plastic sample; Based on the reference spectrum, calculating the scaling factor of the plastic sample spectral data by covariance; Based on the linear model and the scaling factor, multivariate scattering correction is performed on the plastic sample spectral data to obtain the plastic sample standard spectral data.

5. The plastic classification method based on spectral information according to claim 2, characterized in that: The support vector machine model is trained based on the plastic sample standard spectrum database to obtain the plastic classification model, including: Performing dimension expansion on the standard spectral data of the plastic sample based on a Gaussian kernel function; The support vector machine model is trained using the expanded standard spectrum data of the plastic samples to obtain the plastic classification model.

6. The plastic classification method based on spectral information according to claim 2, characterized in that: The method of comparing the preliminary classification results by using the spectral information divergence method to obtain the final classification results of the plastics to be classified includes: By using the spectral information divergence method, the preliminary classification result is compared with the spectral information divergence of the classified samples in the plastic sample standard spectral database; Based on the compared spectral information divergence and the preliminary classification result, a final classification result of the plastic to be classified is obtained.

7. The plastic classification method based on spectral information according to claim 1, characterized in that: The method of obtaining the hyperspectral data of the plastic to be classified comprises: The hyperspectral data of the plastics to be classified are collected by a SpecimFX17 hyperspectral camera.

8. A plastic classification device based on spectral information, characterized in that: The device comprises: A data acquisition module, used for acquiring hyperspectral data of plastics to be classified; A preliminary classification module, used for inputting the hyperspectral data into a plastic classification model to obtain a preliminary classification result; The final classification module compares the preliminary classification results by using the spectral information divergence method to obtain the final classification results of the plastic to be classified.

9. An electronic device, characterized in that: It includes at least one control processor and a memory for communicating with the at least one control processor; the memory stores instructions that can be executed by the at least one control processor, and the instructions are executed by the at least one control processor so that the at least one control processor can execute a plastic classification method based on spectral information as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and the computer-executable instructions are used to enable a computer to execute the plastic classification method based on spectral information as described in any one of claims 1 to 7.

Images