Method and system for evaluating and tracing quality of tobacco lamina
By constructing a tobacco leaf quality model and using near-infrared spectroscopy, combined with a two-dimensional graphical model and standard deviation ellipse analysis, the complexity of tobacco leaf quality assessment and the influence of human factors were solved, enabling efficient and accurate assessment and traceability of tobacco leaf quality.
Patent Information
- Application Number
- CN202511670910.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-14
- Publication Date
- 2026-02-06
AI Technical Summary
Existing methods for assessing the quality of tobacco sheets are complex and susceptible to human factors, resulting in inaccurate evaluation results and making it difficult to achieve efficient and accurate quality stability assessment and traceability.
By constructing a tobacco leaf quality model, extracting tobacco leaf characteristics using near-infrared spectroscopy, and combining a two-dimensional graphical model and standard deviation ellipse analysis, the historical and real-time quality projection points of tobacco leaves are determined. Based on the consistency ratio, quality fluctuations are evaluated, and substandard tobacco leaf formulations are screened out.
This paper presents a highly efficient evaluation method that requires no human intervention, enabling accurate assessment and traceability of tobacco sheet quality and formulation quality, and improving the efficiency and accuracy of assessing the stability of tobacco sheet quality.
Smart Images
Figure CN121476104A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of tobacco evaluation technology, and in particular to a method and system for tobacco quality evaluation and traceability. Background Technology
[0002] As an agricultural product, raw tobacco is affected by various factors such as its variety, cultivation methods, and environmental climate, resulting in quality variations even within the same origin and grade, and across different years. Raw tobacco is sorted, threshed, re-dried, and aged, then blended according to a specific formula to obtain sheet tobacco. As the main raw material for cigarettes, the quality stability of sheet tobacco directly affects the taste, consistency, and smoking experience of cigarette products. Therefore, analyzing and evaluating the quality stability of sheet tobacco can provide a theoretical basis for ensuring the continuous stability of product quality.
[0003] However, in current technologies, the evaluation methods for the quality stability of tobacco sheets mainly rely on chemical composition analysis and sensory evaluation. However, these traditional methods have drawbacks such as complex analysis processes, low efficiency, and susceptibility to human factors, resulting in inaccurate evaluation results. Summary of the Invention
[0004] To address the problems existing in the prior art, embodiments of the present invention provide a method and system for evaluating and tracing the quality of tobacco sheets.
[0005] This invention provides a method for evaluating and tracing the quality of tobacco sheets, the method comprising:
[0006] Select a target set of tobacco leaves, extract the tobacco leaf characteristics of different grades in the target set of tobacco leaves, and use the evaluation data as model supervision to construct a tobacco leaf quality model;
[0007] Extract historical tobacco leaf data of the target tobacco leaf, and use the historical projection points of the historical tobacco leaf data in the tobacco leaf quality model to locate the quality and determine the historical quality of the target tobacco leaf.
[0008] Extract real-time tobacco leaf data of the target tobacco leaf, compare the real-time projection points of the real-time tobacco leaf data in the tobacco leaf quality model with the historical projection points, and determine the quality fluctuation evaluation of the target tobacco leaf based on the consistency ratio.
[0009] Obtain the fitted spectrum of the tobacco formula in the target tobacco sheet, compare the fitted spectrum with the positioning range of the real-time projection point, and determine the quality of the target tobacco sheet based on the positioning range;
[0010] When the quality of the tobacco sheet formulation fails to meet the standards, tobacco sheet formulations with substandard quality are screened based on the formulation projection points of the tobacco sheet formulation.
[0011] In one embodiment, the method further includes:
[0012] The near-infrared spectra of sample tobacco leaves in the target tobacco leaf set were extracted using near-infrared spectroscopy, and the near-infrared spectra were preprocessed.
[0013] Using the evaluation data as model supervision, the inter-class and intra-class differences of different grades / scores in the evaluation data are calculated. The feature vectors of different sample tobaccos are determined by the calculation results. The quality grade / score and style grade / score are used as the dimensions of the two-dimensional graph. The quantization range of the feature vectors is adjusted to realize the quantization of the grade / score of the two-dimensional vectors in the two-dimensional graph.
[0014] In one embodiment, the method further includes:
[0015] The historical spectrum corresponding to the historical tobacco leaves is obtained, and the two-dimensional coordinates corresponding to each historical spectrum are calculated in combination with the tobacco leaf quality model to determine the historical projection point.
[0016] Based on the mean and standard deviation of the historical projection points, a historical double standard deviation ellipse is plotted. Based on the coverage of the historical double standard deviation ellipse over the historical projection points, the historical quality of the target tobacco sheet is determined.
[0017] In one embodiment, the method further includes:
[0018] The real-time spectrum corresponding to the real-time tobacco leaf is obtained, and the two-dimensional coordinates corresponding to each real-time spectrum are calculated in combination with the tobacco leaf quality model to determine the real-time projection point.
[0019] The consistency of the target tobacco sheet is determined by comparing the coverage of the historical double standard deviation ellipse to the real-time projection points.
[0020] In one embodiment, the method further includes:
[0021] By combining the fitted spectrum with the tobacco quality model, the two-dimensional coordinates corresponding to each fitted spectrum are calculated, thereby determining the fitted projection point;
[0022] Based on the mean and standard deviation of the real-time projection points, a real-time double standard deviation ellipse is plotted. Based on the coverage of the real-time double standard deviation ellipse over the fitted projection points, the quality of the target tobacco sheet is determined.
[0023] In one embodiment, the method further includes:
[0024] Obtain the tobacco leaf spectrum corresponding to the historical data of each tobacco leaf formula, and determine the formula projection point of the tobacco leaf formula by combining it with the tobacco leaf quality model;
[0025] Based on the mean and standard deviation of the formula projection points, a double standard deviation ellipse of the formula is plotted. Based on the coverage of the double standard deviation ellipse of the formula projection points, the formula conformity of the corresponding formula is determined.
[0026] Output the substandard tobacco formula whose conformity to the formula is less than the conformity standard.
[0027] This invention provides a tobacco leaf quality evaluation and traceability system, the system comprising:
[0028] The quality model module is used to select a target tobacco leaf set, extract tobacco leaf characteristics of different grades in the target tobacco leaf set, and use the evaluation data as model supervision to construct a tobacco leaf quality model.
[0029] The historical data module is used to extract historical tobacco leaf data of the target tobacco leaf, and to perform quality positioning by using the historical projection points of the historical tobacco leaf data in the tobacco leaf quality model to determine the historical quality of the target tobacco leaf.
[0030] The real-time data module is used to extract real-time tobacco leaf data of the target tobacco leaf, compare the real-time projection points of the real-time tobacco leaf data in the tobacco leaf quality model with the historical projection points, and determine the quality fluctuation evaluation of the target tobacco leaf based on the consistency ratio.
[0031] The fitting module is used to obtain the fitted spectrum of the tobacco formula in the target tobacco sheet, compare the fitted spectrum with the positioning range of the real-time projection point, and determine the quality of the target tobacco sheet based on the positioning range;
[0032] The formulation module is used to screen for substandard tobacco formulations based on the formulation projection points of the tobacco formulation when the quality of the tobacco formulation fails to meet the standards.
[0033] In one embodiment, the system further includes:
[0034] The spectral module is used to extract the near-infrared spectra of sample tobacco leaves in the target tobacco leaf set using near-infrared spectroscopy technology, and to preprocess the near-infrared spectra.
[0035] The model calculation module is used to use the evaluation data as model supervision to calculate the inter-class and intra-class differences of different grades / scores in the evaluation data, determine the feature vectors of different sample tobaccos through the calculation results, and adjust the quantization range of the feature vectors using the quality grade / score and style grade / score as dimensions of the two-dimensional graph to realize the quantization of the grade / score of the two-dimensional vector in the two-dimensional graph.
[0036] This invention provides an electronic device, including a processor and a memory;
[0037] The processor is connected to the memory;
[0038] The memory is used to store executable program code;
[0039] The processor runs a program corresponding to the executable program code stored in the memory to perform the methods described in one or more embodiments.
[0040] This invention provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described method for evaluating and tracing the quality of tobacco sheets.
[0041] In view of the above, in one or more embodiments of this specification, a target tobacco leaf set is selected, and tobacco leaf characteristics of different grades in the target tobacco leaf set are extracted. Evaluation data is used as model supervision to construct a tobacco leaf quality model. Historical tobacco leaf data of the target tobacco leaf is extracted, and the historical projection points of the historical tobacco leaf data in the tobacco leaf quality model are used for quality positioning to determine the historical quality of the target tobacco leaf. Real-time tobacco leaf data of the target tobacco leaf is extracted, and the consistency between the real-time projection points of the real-time tobacco leaf data in the tobacco leaf quality model and the historical projection points is compared. Based on the consistency ratio, the quality fluctuation evaluation of the target tobacco leaf is determined. The fitted spectrum of the tobacco leaf formula in the target tobacco leaf is obtained, and the positioning range of the fitted spectrum and the real-time projection points is compared. Based on the positioning range, the quality of the target tobacco leaf is determined. When the quality of the tobacco leaf formula fails to meet the standard, tobacco leaf formulas with substandard quality are screened based on the formula projection points. This provides an efficient evaluation method without manual intervention, providing accurate assessment and corresponding traceability methods for the quality, quality stability, and formula quality of the target tobacco leaf. Attached Figure Description
[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0043] Figure 1 This is a flowchart of a method for evaluating and tracing the quality of tobacco sheets, provided in one embodiment of this specification.
[0044] Figure 2 This is a two-dimensional image model diagram of a target tobacco leaf quality provided in one embodiment of this specification.
[0045] Figure 3This is a schematic diagram of a double standard deviation ellipse of the first three years of tobacco product A origin and grade X, provided in one embodiment of this specification.
[0046] Figure 4 This is a schematic diagram of an elliptical projection of a real-time projection point of grade X of tobacco product A over the previous three years, based on two standard deviations, provided in one embodiment of this specification.
[0047] Figure 5 This is a schematic diagram of a fitted projection point in a real-time elliptical distribution of two standard deviations, provided by one embodiment of this specification.
[0048] Figure 6 This is a schematic diagram of an elliptical projection of the real-time projection points of the x grade of raw tobacco (origin A) over the previous three years at twice the standard deviation, provided by one embodiment of this specification.
[0049] Figure 7 This is a schematic diagram of an elliptical projection of the real-time projection points of the raw tobacco b origin y grade over the previous three years, provided by one embodiment of this specification, using twice the standard deviation.
[0050] Figure 8 This is a schematic diagram of a tobacco quality evaluation and traceability system provided in one embodiment of this specification.
[0051] Figure 9 This is a schematic diagram of the structure of an electronic device provided in one embodiment of this specification. Detailed Implementation
[0052] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed merely to enable those skilled in the art to better understand and implement the subject matter described herein, and are not intended to limit the scope, applicability, or examples set forth in the claims. The function and arrangement of the elements discussed may be changed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the various examples. For example, the described methods may be performed in a different order than described, and steps may be added, omitted, or combined. Furthermore, features described in some examples may be combined in other examples.
[0053] As used herein, the term "comprising" and its variations are open terms meaning "including but not limited to". The term "based on" means "at least partially based on". The terms "one embodiment" and "an embodiment" mean "at least one embodiment". The term "another embodiment" means "at least one other embodiment". The terms "first", "second", etc., may refer to different or the same objects. Other definitions, whether explicit or implicit, may be included below. Unless explicitly indicated by the context, the definition of a term shall remain consistent throughout the specification.
[0054] like Figure 1 As shown, this embodiment of the invention provides a method for evaluating and tracing the quality of sheet tobacco, including:
[0055] Step S102: Select a target tobacco leaf set, extract tobacco leaf features of different grades in the target tobacco leaf set, and use the evaluation data as model supervision to construct a tobacco leaf quality model.
[0056] Specifically, taking tobacco leaves A from origin X as the target tobacco leaf as an example, a set of tobacco leaves A is selected. When selecting sample tobacco leaves from the target tobacco leaf set, it is advisable to select tobacco leaves of different qualities and styles, i.e., tobacco leaves with certain differences, to make the sample as comprehensive as possible. For example, covering 3-5 core production areas, selecting high / medium / low quality tobacco leaves from each area, with a total of ≥500 samples. Near-infrared spectroscopy is used on the sample tobacco leaves to obtain the corresponding near-infrared spectra of the sample tobacco leaves in the target tobacco leaf set. The spectra are preprocessed, and then tobacco leaf features are extracted. The tobacco leaf feature extraction steps and the subsequent steps for constructing the tobacco leaf quality model can be as follows:
[0057] 1. Preprocessing steps: including moving window smoothing and derivative processing.
[0058] Moving window smoothing involves smoothing the spectral data using a moving window method. Through data fitting and differentiation, a 17-point moving window can be used to traverse the spectral data, with each window fitted using a third-order polynomial. The aim is to eliminate high-frequency noise (such as instrument fluctuations) while preserving the characteristic absorption peak shapes of key compounds such as sugars and nicotine.
[0059] Derivative processing involves differentiating the smoothed polynomial (usually the first derivative). Its purpose is to eliminate baseline drift (such as the effect of scattering from the sample surface), enhance the separation of overlapping peaks, and make characteristic peaks such as nicotine (1700 nm) and total sugar (2100 nm) more prominent.
[0060] 2. Model Building
[0061] Establish a two-dimensional graph model, such as Figure 2As shown, Fisher (FDA) is used to introduce category labels (good / medium / poor grades, where the introduction of scores / grades is supervised by assessment data, such as professional smoking data from an expert panel), maximizing inter-class differences and minimizing intra-class differences in the dimensionality-reduced coordinate axes. Inter-class deviation matrix B (reflecting differences between grades) and intra-class deviation matrix A (reflecting fluctuations within the same grade) are calculated. In the two-dimensional graph, the discriminant scores of the first and second dimensions are selected as the quality-style values. The generated two-dimensional graph displays the class centers of each category, with the class center coordinates calculated as (mean of discriminant scores for all samples in that class) × (non-standardized discriminant function coefficient matrix). Rotation transformations can be applied based on the actual good / medium / poor grades. The magnitude of the quality value represents the quality of the tobacco, while the magnitude of the style value only indicates the degree of style manifestation, not necessarily quality.
[0062] 3. Model score quantization
[0063] To more intuitively reflect the differences in tobacco leaf quality, the score on the horizontal axis (-1) is numerically quantified. The first dimension score data of the modeling data is mapped to the interval [0, 100]. The maximum value of the first dimension of the data projection is divided into 0 points, and the minimum value is divided into 100 points. A6 has the best quality and the highest score. The minimum value (min) and maximum value (max) of the first dimension (quality axis) score of all samples are taken, and the linear transformation parameters are calculated to obtain the corresponding scores within the [0, 100] score scale. For example, the highest original score (optimal grade A6) is mapped to 100 points, and the lowest score is mapped to 0 points. This constructs a tobacco leaf quality model, and the target tobacco leaf quality model is scored.
[0064] Step S104: Extract historical tobacco leaf data of the target tobacco leaf, use the historical projection points of the historical tobacco leaf data in the tobacco leaf quality model to locate the quality, and determine the historical quality of the target tobacco leaf.
[0065] Specifically, historical tobacco leaf data for the target tobacco leaf is extracted. For example, data on tobacco leaf A from origin X over the past three years can be obtained from historical data aggregation. The amount of data for each year should be greater than or equal to 30 batches to avoid statistical bias due to small samples. After extracting the near-infrared spectra corresponding to the historical tobacco leaf data, the preprocessing parameters (17-point smoothing / 3rd-order polynomial / first derivative) from the previous step are reused to ensure consistency of model input.
[0066] After preprocessing, each historical spectrum is input into the FDA model from the previous step to calculate its coordinates in two-dimensional space, thus determining the two-dimensional data corresponding to each historical spectrum. Then, the mean and standard deviation of all projection points (historical projection points) are calculated. An ellipse with the mean as the center point and twice the standard deviation (including the X-axis and Y-axis standard deviations) as the major and minor axes is plotted. The number of historical projection points within the ellipse is then compared; this is the ellipse coverage rate. The proportion of projection points within the ellipse serves as the historical quality evaluation standard for the target tobacco sheet. The two-standard-deviation method, compared to the one-standard-deviation method, can detect small fluctuations, and compared to the three-standard-deviation method, it is more realistic and does not mask true deviations, resulting in the most balanced data. For example, as illustrated... Figure 3 As shown, the projection points of all samples and the ellipse of twice the standard deviation are as follows: Figure 3 As can be seen, the projected data are slightly scattered, but 95% of the data are distributed within the range of the double standard deviation ellipse, indicating that the target tobacco in the historical data is relatively stable.
[0067] Step S106: Extract real-time tobacco leaf data of the target tobacco leaf, compare the real-time projection points of the real-time tobacco leaf data in the tobacco leaf quality model with the historical projection points, and determine the quality fluctuation evaluation of the target tobacco leaf based on the consistency ratio.
[0068] Specifically, real-time tobacco leaf data for the target tobacco leaf of the current year is extracted, i.e., data on tobacco leaf A from origin X in the current year is obtained. The data volume should be greater than or equal to 30 batches to avoid statistical bias due to small samples. Referring to the previous step, near-infrared spectral extraction is performed on the real-time tobacco leaf data. After preprocessing, the projection points (real-time projection points) of each spectrum are determined in conjunction with the tobacco leaf quality model. Then, the projection points are compared with historical projection points. Based on the comparison results, the stability of the target tobacco leaf between years is determined. The comparison process includes comparing all real-time projection points with the double standard deviation ellipse drawn from the historical projection points to determine the proportion of real-time projection points covered by the double standard deviation ellipse.
[0069] Furthermore, the standard for evaluating the quality fluctuation of the target tobacco sheet can be as follows: when the consistency is greater than or equal to 80% (greater than or equal to 80% of the real-time projection points are covered by an ellipse of twice the standard deviation), it indicates stable quality, and further quality traceability can be performed in subsequent steps. Conversely, when it is less than 80%, it indicates insufficient stability, and further quality traceability is required in subsequent steps. Figure 4 For example, the proportion of projection points within the range of the ellipse with two standard deviations in the previous three years is calculated to be 68.5% consistent. Since the consistency is less than 80%, the inter-year stability of grade X from origin A is poor this year, and further analysis is needed to determine the reasons for the quality fluctuations.
[0070] Step S108: Obtain the fitted spectrum of the tobacco formula in the target tobacco sheet, compare the fitted spectrum with the positioning range of the real-time projection point, and determine the quality of the target tobacco sheet based on the positioning range.
[0071] Specifically, further quality traceability of target tobacco leaves includes tracing the tobacco leaf formulation, i.e., obtaining the fitted spectrum of the tobacco leaf formulation. For example, the quality traceability of the fitted spectrum can be illustrated by the following: Tobacco leaf of origin A, grade X, is obtained from 51% grade X tobacco from origin A and 49% grade Y tobacco from origin B through processes such as tobacco blending, leaf and stem separation, and re-drying. The average of the spectra of grade X tobacco from origin A and grade Y tobacco from origin B is then weighted at a ratio of 51:49 to obtain the fitted spectrum of the tobacco leaf. This fitted spectrum is then input into the tobacco leaf quality model to obtain the projection points of the fitted spectrum.
[0072] Further, based on the real-time projection points from the previous step, draw the corresponding real-time double standard deviation ellipse. The drawing process can refer to the procedure in step S104 for drawing the double standard deviation ellipse of historical projection points. By calculating the mean and standard deviation of all projection points, draw the double standard deviation ellipse with the mean as the center point and the double standard deviation as the major and minor semi-axes. Then, compare the positions of the projection points of the fitted spectrum with the real-time double standard deviation ellipse, such as... Figure 5 As shown, if the projection point of the fitted spectrum is located within the real-time double standard deviation ellipse, the quality is stable after processes such as raw tobacco blending, leaf and stem separation, and re-drying from raw tobacco to sheet tobacco, meaning there is no problem with the production process or formula.
[0073] Step S110: When the quality of the tobacco sheet formulation fails to meet the standard, tobacco sheet formulations with substandard quality are screened based on the formulation projection points of the tobacco sheet formulation.
[0074] Specifically, when the quality of the tobacco sheet formulation fails to meet standards, or when other instructions require testing of each formulation within the tobacco sheet formulation, each formulation within the tobacco sheet is tested individually. For example, based on the initial formulation of tobacco sheet of grade X from origin A, the consistency between the corresponding raw tobacco of grade X from origin a and raw tobacco of grade Y from origin b is analyzed.
[0075] Formula consistency testing can assess the stability of a formula. This involves acquiring historical data for different formulas, such as the tobacco leaf spectra corresponding to tobacco grade A from its origin and grade X over the past three years. These data are then combined with a tobacco quality model to determine projection points, and the corresponding double-standard-deviation ellipse is plotted using these projection points. Next, real-time data for the formula is acquired, such as this year's tobacco leaf spectra for tobacco grade A from its origin and grade X. Again, these projection points are determined using the tobacco quality model, and the double-standard-deviation ellipse plotted using historical data is used to assess the consistency of formula A from its origin and grade X. For example,... Figure 6As shown, the standard for conformity can be set at 80%, for example, if the conformity of origin A and grade X reaches 89.0%, the quality stability between years is good.
[0076] Similarly, the testing of the grade y of the raw tobacco b from another formulation can be conducted in the same manner as the testing of the grade x of the raw tobacco a from another formulation. The final result can be as follows: Figure 7 As shown, in Figure 7 In the data, the consistency between origin B and grade Y was 71.2%, and the coverage between today's tobacco leaf spectrum and the ellipse plotted with two standard deviations of historical data was only 71.2%, indicating poor inter-year stability. The slightly poor inter-year stability of raw tobacco from origin B and grade Y can be attributed to blending raw tobacco with grade X from origin A, followed by leaf and stem separation and re-drying processes. This can lead to poor inter-year stability of the tobacco leaves. Research can be conducted by adjusting the initial formula or using other technical methods to achieve overall stability in the quality of tobacco leaves from the same origin and grade.
[0077] In this embodiment, the method for restoring cloud-polluted EO images has the following advantages: high visibility, achieving high-visibility image restoration while meeting fidelity requirements; high fidelity, through bidirectional generation technology and adaptive adjustment strategies, the generated regional edge mapping features have clear physical concepts and uninterrupted edges, and the parameters are controllable and adjustable, ensuring the fidelity of the restored image; high computational efficiency, unlike similar methods that require dual deep learning models, only one deep learning model is needed, simplifying the system complexity and improving computational efficiency by about 50%; and wide applicability, ensuring data availability, improving image quality, enhancing ground feature identification capabilities, supporting continuous monitoring and time series analysis, and reducing data acquisition costs.
[0078] This invention provides a method for evaluating and tracing the quality of tobacco flakes. The method involves selecting a target tobacco flake set, extracting tobacco leaf characteristics of different grades within the set, and using evaluation data as model supervision to construct a tobacco leaf quality model. Historical tobacco leaf data for the target tobacco flakes is extracted, and the historical projection points of this data in the tobacco leaf quality model are used for quality localization to determine the historical quality of the target tobacco flakes. Real-time tobacco leaf data for the target tobacco flakes is extracted, and the consistency between the real-time projection points and historical projection points in the model is compared to determine the quality fluctuation of the target tobacco flakes based on the consistency ratio. The method also involves obtaining the fitted spectrum of the tobacco flake formula within the target tobacco flakes, comparing the fitted spectrum with the localization range of the real-time projection points, and determining the quality of the target tobacco flakes based on this localization range. Furthermore, when the quality of the tobacco flake formula fails to meet standards, formulas with substandard quality are screened based on their formula projection points. This provides a highly efficient evaluation method without manual intervention, offering accurate assessment and traceability methods for the quality, quality stability, and formula quality of target tobacco flakes.
[0079] Please see Figure 8 , Figure 8 This is a schematic diagram of a tobacco quality evaluation and traceability system provided in an embodiment of this application. Figure 8 As shown, the system includes:
[0080] The quality model module S802 is used to select a target tobacco leaf set, extract tobacco leaf characteristics of different grades in the target tobacco leaf set, and use the evaluation data as model supervision to construct a tobacco leaf quality model.
[0081] The historical data module S804 is used to extract historical tobacco leaf data of the target tobacco leaf, and to perform quality positioning by using the historical projection points of the historical tobacco leaf data in the tobacco leaf quality model to determine the historical quality of the target tobacco leaf.
[0082] The real-time data module S806 is used to extract real-time tobacco leaf data of the target tobacco leaf, compare the real-time projection points of the real-time tobacco leaf data in the tobacco leaf quality model with the historical projection points, and determine the quality fluctuation evaluation of the target tobacco leaf based on the consistency ratio.
[0083] The fitting module S808 is used to acquire the fitted spectrum of the tobacco formula in the target tobacco sheet, compare the fitted spectrum with the positioning range of the real-time projection point, and determine the quality of the target tobacco sheet based on the positioning range;
[0084] The formulation module S810 is used to screen for tobacco formulations that do not meet quality standards based on the formulation projection points of the tobacco formulation when the quality of the tobacco formulation does not meet the standards.
[0085] In another embodiment, a tobacco leaf quality evaluation and traceability system further includes:
[0086] The spectral module is used to extract the near-infrared spectra of sample tobacco leaves in the target tobacco leaf set using near-infrared spectroscopy technology, and to preprocess the near-infrared spectra.
[0087] The model calculation module is used to use the evaluation data as model supervision to calculate the inter-class and intra-class differences of different grades / scores in the evaluation data, determine the feature vectors of different sample tobaccos through the calculation results, and adjust the quantization range of the feature vectors using the quality grade / score and style grade / score as dimensions of the two-dimensional graph to realize the quantization of the grade / score of the two-dimensional vector in the two-dimensional graph.
[0088] Those skilled in the art will clearly understand that the technical solutions of the embodiments of this application can be implemented by means of software and / or hardware. In this specification, "unit" and "module" refer to software and / or hardware that can independently complete or cooperate with other components to complete a specific function, wherein the hardware may be, for example, a field-programmable gate array (FPGA), an integrated circuit (IC), etc.
[0089] Each processing unit and / or module in the embodiments of this application can be implemented by an analog circuit that implements the functions described in the embodiments of this application, or by software that executes the functions described in the embodiments of this application.
[0090] See Figure 9 It shows a schematic diagram of the structure of an electronic device according to an embodiment of this application, which can be used to implement... Figure 1 The method in the illustrated embodiment. (As shown) Figure 9 As shown, the electronic device 900 may include: at least one processor 901, at least one network interface 904, a user interface 903, a memory 905, and at least one communication bus 902.
[0091] The communication bus 902 is used to enable communication between these components.
[0092] The user interface 903 may include a display screen and a camera. Optionally, the user interface 903 may also include a standard wired interface and a wireless interface.
[0093] The network interface 904 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).
[0094] The processor 901 may include one or more processing cores. The processor 901 connects to various parts within the electronic device 900 using various interfaces and lines, and performs various functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in the memory 905, and by calling data stored in the memory 905. Optionally, the processor 901 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 901 may integrate one or a combination of several of the following: a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), and a modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content required for display; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 901 and may be implemented as a separate chip.
[0095] The memory 905 may include random access memory (RAM) or read-only memory. Optionally, the memory 905 may include a non-transitory computer-readable storage medium. The memory 905 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 905 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory 905 may also be at least one storage device located remotely from the aforementioned processor 901. Figure 9 As shown, the memory 905, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and program instructions.
[0096] exist Figure 9In the illustrated electronic device 900, the user interface 903 is mainly used to provide an input interface for the user and acquire user input data; while the processor 901 can be used to call the image-based interactive application stored in the memory 905 and specifically perform the following operations: select a target tobacco leaf set, extract tobacco leaf features of different grades in the target tobacco leaf set, use evaluation data as model supervision to construct a tobacco leaf quality model; extract historical tobacco leaf data of the target tobacco leaf, use the historical projection points of the historical tobacco leaf data in the tobacco leaf quality model to locate the quality and determine the historical quality of the target tobacco leaf; extract real-time tobacco leaf data of the target tobacco leaf, compare the real-time projection points of the real-time tobacco leaf data in the tobacco leaf quality model with the historical projection points, and determine the quality fluctuation evaluation of the target tobacco leaf based on the consistency ratio; acquire the fitted spectrum of the tobacco leaf formula in the target tobacco leaf, compare the fitted spectrum with the location range of the real-time projection points, and determine the quality of the target tobacco leaf based on the location range; and when the quality of the tobacco leaf formula does not meet the standard, screen for tobacco leaf formulas with substandard quality based on the formula projection points of the tobacco leaf formula.
[0097] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method. The computer-readable storage medium may include, but is not limited to, any type of disk, including floppy disks, optical disks, DVDs, CD-ROMs, microdrives, as well as magneto-optical disks, ROMs, RAMs, EPROMs, EEPROMs, DRAMs, VRAMs, flash memory devices, magnetic cards or optical cards, nanosystems (including molecular memory ICs), or any type of medium or device suitable for storing instructions and / or data.
[0098] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.
[0099] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.
[0100] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some service interface; the indirect coupling or communication connection between devices or units may be electrical or other forms.
[0101] The units described as separate components may or may not be physically separate. The 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 the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0102] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.
[0103] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, 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. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.
[0104] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc.
[0105] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.
Claims
1. A method for evaluating and tracing the quality of tobacco flakes, the method comprising: Select a target set of tobacco leaves, extract the tobacco leaf characteristics of different grades in the target set of tobacco leaves, and use the evaluation data as model supervision to construct a tobacco leaf quality model; Extract historical tobacco leaf data of the target tobacco leaf, and use the historical projection points of the historical tobacco leaf data in the tobacco leaf quality model to locate the quality and determine the historical quality of the target tobacco leaf. Extract real-time tobacco leaf data of the target tobacco leaf, compare the real-time projection points of the real-time tobacco leaf data in the tobacco leaf quality model with the historical projection points, and determine the quality fluctuation evaluation of the target tobacco leaf based on the consistency ratio. Obtain the fitted spectrum of the tobacco formula in the target tobacco sheet, compare the fitted spectrum with the positioning range of the real-time projection point, and determine the quality of the target tobacco sheet based on the positioning range; When the quality of the tobacco sheet formulation fails to meet the standards, tobacco sheet formulations with substandard quality are screened based on the formulation projection points of the tobacco sheet formulation.
2. The method according to claim 1, characterized in that, The method further includes: The near-infrared spectra of sample tobacco leaves in the target tobacco leaf set were extracted using near-infrared spectroscopy, and the near-infrared spectra were preprocessed. Using the evaluation data as model supervision, the inter-class and intra-class differences of different grades / scores in the evaluation data are calculated. The feature vectors of different sample tobaccos are determined by the calculation results. The quality grade / score and style grade / score are used as the dimensions of the two-dimensional graph. The quantization range of the feature vectors is adjusted to realize the quantization of the grade / score of the two-dimensional vectors in the two-dimensional graph.
3. The method according to claim 2, characterized in that, The step of determining the historical quality of the target tobacco leaf by using historical projection points of historical tobacco leaf data in the tobacco leaf quality model for quality localization includes: The historical spectrum corresponding to the historical tobacco leaves is obtained, and the two-dimensional coordinates corresponding to each historical spectrum are calculated in combination with the tobacco leaf quality model to determine the historical projection point. Based on the mean and standard deviation of the historical projection points, a historical double standard deviation ellipse is plotted. Based on the coverage of the historical double standard deviation ellipse over the historical projection points, the historical quality of the target tobacco sheet is determined.
4. The method according to claim 3, characterized in that, The comparison of the consistency between the real-time projection points of the real-time tobacco leaf data in the tobacco leaf quality model and the historical projection points includes: The real-time spectrum corresponding to the real-time tobacco leaf is obtained, and the two-dimensional coordinates corresponding to each real-time spectrum are calculated in combination with the tobacco leaf quality model to determine the real-time projection point. The consistency of the target tobacco sheet is determined by comparing the coverage of the historical double standard deviation ellipse to the real-time projection points.
5. The method according to claim 4, characterized in that, The comparison of the fitted spectrum with the positioning range of the real-time projection point, and the determination of the quality of the target tobacco sheet based on the positioning range, includes: By combining the fitted spectrum with the tobacco quality model, the two-dimensional coordinates corresponding to each fitted spectrum are calculated, thereby determining the fitted projection point; Based on the mean and standard deviation of the real-time projection points, a real-time double standard deviation ellipse is plotted. Based on the coverage of the real-time double standard deviation ellipse over the fitted projection points, the quality of the target tobacco sheet is determined.
6. The method according to claim 5, characterized in that, The screening of tobacco leaf formulas with substandard quality based on historical projection points of the formula includes: Obtain the tobacco leaf spectrum corresponding to the historical data of each tobacco leaf formula, and determine the formula projection point of the tobacco leaf formula by combining it with the tobacco leaf quality model; Based on the mean and standard deviation of the formula projection points, a double standard deviation ellipse of the formula is plotted. Based on the coverage of the double standard deviation ellipse of the formula projection points, the formula conformity of the corresponding formula is determined. Output the substandard tobacco formula whose conformity to the formula is less than the conformity standard.
7. A tobacco leaf quality evaluation and traceability system, characterized in that, The system includes; The quality model module is used to select a target tobacco leaf set, extract tobacco leaf characteristics of different grades in the target tobacco leaf set, and use the evaluation data as model supervision to construct a tobacco leaf quality model. The historical data module is used to extract historical tobacco leaf data of the target tobacco leaf, and to perform quality positioning by using the historical projection points of the historical tobacco leaf data in the tobacco leaf quality model to determine the historical quality of the target tobacco leaf. The real-time data module is used to extract real-time tobacco leaf data of the target tobacco leaf, compare the real-time projection points of the real-time tobacco leaf data in the tobacco leaf quality model with the historical projection points, and determine the quality fluctuation evaluation of the target tobacco leaf based on the consistency ratio. The fitting module is used to obtain the fitted spectrum of the tobacco formula in the target tobacco sheet, compare the fitted spectrum with the positioning range of the real-time projection point, and determine the quality of the target tobacco sheet based on the positioning range; The formulation module is used to screen for substandard tobacco formulations based on the formulation projection points of the tobacco formulation when the quality of the tobacco formulation fails to meet the standards.
8. The system according to claim 7, characterized in that, The system also includes: The spectral module is used to extract the near-infrared spectra of sample tobacco leaves in the target tobacco leaf set using near-infrared spectroscopy technology, and to preprocess the near-infrared spectra. The model calculation module is used to use the evaluation data as model supervision to calculate the inter-class and intra-class differences of different grades / scores in the evaluation data, determine the feature vectors of different sample tobaccos through the calculation results, and adjust the quantization range of the feature vectors using the quality grade / score and style grade / score as dimensions of the two-dimensional graph to realize the quantization of the grade / score of the two-dimensional vector in the two-dimensional graph.
9. An electronic device, comprising a processor and a memory; The processor is connected to the memory; The memory is used to store executable program code; The processor runs a program corresponding to the executable program code stored in the memory to perform the method as described in any one of claims 1-6.
10. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method as described in any one of claims 1-6.