A tea detection, classification and identification method and identification device
By randomly dividing and feature recognition of tea, combined with vertical and horizontal consistency recognition technology, a directional secondary mixed decision tree is constructed, which solves the problem of low subjectivity and accuracy of the existing tea detection and classification methods, and realizes the tea Accurate classification and identification.
Patent Information
- Application Number
- CN202411098553.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-12
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2044-08-12
AI Technical Summary
The existing tea detection classification methods rely on sensory evaluation by professional tea tasters, which are highly subjective and difficult to ensure the objectivity and consistency of the evaluation results, resulting in low accuracy and reliability of the classification results.
By randomly dividing and identifying the target batch of tea, physical and chemical testing is performed after mixing, and vertical and horizontal consistency recognition technology is used to construct a directional secondary mixed decision tree to achieve accurate classification and identification of tea.
It improves the accuracy and reliability of tea classification results, ensures the objectivity and consistency of evaluation results, and meets the precise needs of modern tea production and sales.
Smart Images

Figure CN118629545B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of tea detection, and in particular to a tea detection, classification and identification method and an identification device. Background Art
[0002] In the tea industry, the classification and identification of tea quality is a key task, which is directly related to the market pricing, consumer satisfaction and brand reputation of tea. With the continuous expansion of the tea market and the increasing demand of consumers for tea quality, the traditional tea classification method relies on the taste, smell and other sensory abilities of professional tea tasters to evaluate the quality of tea. This method is highly subjective and easily affected by factors such as the experience, mood and environment of the tea taster. It is difficult to ensure the objectivity and consistency of the evaluation results, and it is difficult to meet the precision needs of modern tea production and sales.
[0003] Among the current related technologies, tea detection, classification and recognition have technical problems such as low accuracy and reliability of classification results. Summary of the invention
[0004] The present application provides a tea detection, classification and identification method and identification device, which adopts technical means such as randomly dividing the target batch of tea, mixing according to appearance characteristics, conducting physical and chemical testing, and performing horizontal and vertical consistency identification, thereby achieving accurate classification and identification of tea and achieving the technical effect of improving the accuracy and reliability of classification results.
[0005] The present application provides a tea detection, classification and identification method, comprising:
[0006] The target batch of tea leaves is randomly divided into N portions of tea leaves, the N portions of tea leaves are traversed for feature recognition, and the N portions of tea leaves are mixed according to the feature recognition results to obtain K portions of mixed tea leaves; K portions of first tea samples and K portions of second tea samples of preset mass are randomly selected from the K portions of mixed tea leaves, and physical and chemical tests are performed on the K portions of first tea samples and the K portions of second tea samples to obtain K first physical and chemical indicator sets and K second physical and chemical indicator sets; the K second physical and chemical indicator sets are used to perform longitudinal consistency recognition on the K first physical and chemical indicator sets to obtain K longitudinal consistency factors; based on the K longitudinal consistency factors The method further comprises the steps of: performing mixed verification of the mixed tea leaves, and when the verification is passed, performing lateral consistency identification between samples on the K first physicochemical indicator sets and the K second physicochemical indicator sets, respectively, to obtain K lateral consistency factor sets; performing directional secondary mixing of the K portions of mixed tea leaves based on the K lateral consistency factor sets, to obtain M portions of divided tea leaves; traversing the M portions of divided tea leaves for physicochemical testing, generating M divided physicochemical testing sets, and obtaining M divided tea leaves grades according to the M divided physicochemical testing sets; and taking the M portions of divided tea leaves and the M divided tea leaves grades as the detection, classification and identification results of the target batch of tea leaves.
[0007] In a possible implementation, the following processing is performed:
[0008] Calling the visual acquisition unit to perform image acquisition on the N portions of tea to obtain N image acquisition results; performing tea feature recognition based on the N image acquisition results to obtain N tea feature sets; performing similarity recognition based on the N tea feature sets to obtain similarity recognition results; mixing the N portions of tea based on the similarity recognition results to generate K portions of mixed tea.
[0009] In a possible implementation, the following processing is performed:
[0010] The K portions of the first tea samples and the K portions of the second tea samples are respectively tested for active ingredients by colorimetry to generate K first active ingredient indices and K second active ingredient indices; the K portions of the first tea samples and the K portions of the second tea samples are tested for pesticide residues by gas chromatography-mass spectrometry to generate K first pesticide residue indices and K second pesticide residue indices; the K portions of the first tea samples and the K portions of the second tea samples are tested for nutritional ingredients by atomic absorption spectroscopy to generate K first nutritional ingredient indices and K second nutritional ingredient indices; the K first active ingredient indices, the K first pesticide residue indices and the K first nutritional ingredient indices are used as K first physical and chemical indicator sets; the K second active ingredient indices, the K second pesticide residue indices and the K second nutritional ingredient indices are used as K second physical and chemical indicator sets.
[0011] In a possible implementation, the following processing is performed:
[0012] A longitudinal consistency identification formula is constructed, wherein the longitudinal consistency identification formula is: ;in, is the vertical consistency factor, n is the number of physical and chemical indicators, n is a positive integer, is the i-th physical and chemical index in the first physical and chemical index set, is the i-th physical and chemical index in the second physical and chemical index set, is the weight of the i-th physical and chemical indicator in the vertical consistency identification; the K first physical and chemical indicator sets and the K second physical and chemical indicator sets are respectively input into the vertical consistency identification formula to obtain the K vertical consistency factors.
[0013] In a possible implementation, the following processing is performed:
[0014] The K first physical and chemical indicator sets are enumerated in pairs to obtain multiple first enumeration combinations, wherein each first enumeration combination includes two first physical and chemical indicator sets; and a horizontal consistency identification formula is constructed, wherein the horizontal consistency identification formula is: ;in, is the horizontal consistency factor, n is the number of physical and chemical indicators, is the jth physical and chemical index of any physical and chemical index set in an enumeration combination, is the jth physical and chemical index of another physical and chemical index set in an enumeration combination, is the weight of the jth physical and chemical indicator in the horizontal consistency identification, and n is a positive integer; based on the horizontal consistency identification formula, the multiple first enumeration combinations are respectively identified for horizontal consistency between samples to obtain multiple first horizontal consistency factors; based on the K first physical and chemical indicator sets, the multiple first horizontal consistency factors are extracted to obtain K first horizontal consistency factor sets; using the horizontal consistency identification formula, the K second physical and chemical indicator sets are respectively identified for horizontal consistency between samples to obtain multiple second horizontal consistency factors, and the K second physical and chemical indicator sets are used as indexes to retrieve the multiple second horizontal consistency factors to obtain K second horizontal consistency factor sets; based on the K first horizontal consistency factor sets and the K second horizontal consistency factor sets, a comprehensive analysis is performed to generate K horizontal consistency factor sets, wherein the K horizontal consistency factor sets correspond one-to-one to the K mixed teas, and the horizontal consistency factor is the degree of similarity in physical and chemical detection between each mixed tea and any one of the K mixed teas.
[0015] In a possible implementation, the following processing is performed:
[0016] Determine whether there is a longitudinal consistency factor less than a preset longitudinal consistency factor threshold among the K longitudinal consistency factors, if not, the verification is passed; if so, obtain Q portions of longitudinal abnormal mixed tea, and perform overall mixing and feature identification on the Q portions of longitudinal abnormal mixed tea to obtain Q portions of updated mixed tea; perform sample physicochemical testing on the Q portions of updated mixed tea respectively to obtain Q first updated physicochemical indicator sets and Q second updated physicochemical indicator sets; perform longitudinal consistency identification and mixing verification based on the Q first updated physicochemical indicator sets and the Q second updated physicochemical indicator sets, and when the verification is passed, update the K portions of mixed tea based on the Q portions of updated mixed tea; update the K first physicochemical indicator sets and the K second physicochemical indicator sets based on the Q first updated physicochemical indicator sets and the Q second updated physicochemical indicator sets.
[0017] In a possible implementation, the following processing is performed:
[0018] Randomly select M portions of mixed tea from the K portions of mixed tea as M decision nodes, and construct a directional secondary mixed decision tree, wherein the lateral consistency factor of any two portions of mixed tea among the M portions of mixed tea is less than or equal to a preset lateral consistency factor; input the K portions of mixed tea into the directional secondary mixed decision tree, analyze the lateral consistency factor of each portion of mixed tea with the M decision nodes in turn, and when the lateral consistency factor with the decision node is the maximum value in the set of lateral consistency factors of the mixed tea, add the mixed tea into the internal node of the corresponding decision node to obtain M internal nodes; mix the mixed tea in the M internal nodes to obtain the M portions of divided tea.
[0019] The present application also provides a tea detection, classification and identification device, comprising:
[0020] A feature recognition module, the feature recognition module is used to randomly divide the target batch of tea into N portions of tea, traverse the N portions of tea to perform feature recognition, and mix the N portions of tea according to the feature recognition results to obtain K portions of mixed tea; a physical and chemical detection module, the physical and chemical detection module is used to randomly extract K portions of first tea samples and K portions of second tea samples of preset mass from the K portions of mixed tea, and perform physical and chemical detection on the K portions of first tea samples and the K portions of second tea samples to obtain K first physical and chemical indicator sets and K second physical and chemical indicator sets; a longitudinal consistency recognition module, the longitudinal consistency recognition module is used to perform longitudinal consistency recognition on the K first physical and chemical indicator sets using the K second physical and chemical indicator sets to obtain K longitudinal consistency factors; a transverse consistency recognition module is used to perform transverse consistency recognition on the K portions of first physical and chemical indicator sets based on the K portions of second physical and chemical indicator sets. A longitudinal consistency factor is used to verify the mixing of the mixed tea leaves. When the verification is passed, the K first physical and chemical indicator sets and the K second physical and chemical indicator sets are respectively identified for inter-sample horizontal consistency to obtain K horizontal consistency factor sets; a secondary mixing module is used to perform directional secondary mixing of the K portions of mixed tea leaves based on the K horizontal consistency factor sets to obtain M portions of divided tea leaves; a tea grade acquisition module is used to traverse the M portions of divided tea leaves for physical and chemical testing, generate M divided physical and chemical test sets, and obtain M divided tea grades according to the M divided physical and chemical test sets; a detection classification recognition result acquisition module is used to use the M portions of divided tea leaves and the M divided tea grades as the detection classification recognition results of the target batch of tea leaves.
[0021] The present application proposes a tea detection, classification and identification method and identification device. First, the target batch of tea is randomly divided into N portions of tea, and the N portions of tea are traversed for feature identification. According to the feature identification results, the N portions of tea are mixed to obtain K portions of mixed tea. Then, K portions of first tea samples and K portions of second tea samples of preset quality are randomly extracted from the K portions of mixed tea, and physical and chemical tests are performed on the K portions of first tea samples and the K portions of second tea samples to obtain K first physical and chemical indicator sets and K second physical and chemical indicator sets. Then, the K second physical and chemical indicator sets are used to perform longitudinal consistency identification on the K first physical and chemical indicator sets to obtain K longitudinal consistency factors. Then, based on the K The vertical consistency factor is used to perform mixed verification of the mixed tea. When the verification is passed, the K first physical and chemical indicator sets and the K second physical and chemical indicator sets are respectively identified for inter-sample horizontal consistency to obtain K horizontal consistency factor sets. Then, based on the K horizontal consistency factor sets, the K portions of mixed tea are directionally secondary mixed to obtain M portions of divided tea. Then, the M portions of divided tea are traversed for physical and chemical testing to generate M divided physical and chemical test sets. According to the M divided physical and chemical test sets, M divided tea grades are obtained. Finally, the M portions of divided tea and the M divided tea grades are used as the detection, classification and identification results of the target batch of tea, thereby achieving the technical effect of improving the accuracy and reliability of the classification results. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the device according to the embodiment of the present application. It should be understood that the previous or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.
[0023] Figure 1 A flowchart of a tea detection, classification and identification method provided in an embodiment of the present application.
[0024] Figure 2 A schematic diagram of the structure of a tea detection, classification and identification device provided in an embodiment of the present application.
[0025] Explanation of the reference numerals: feature recognition module 10 , physical and chemical detection module 20 , longitudinal consistency recognition module 30 , transverse consistency recognition module 40 , secondary mixing module 50 , tea grade acquisition module 60 , detection classification recognition result acquisition module 70 . DETAILED DESCRIPTION
[0026] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.
[0027] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.
[0028] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, device, product, or server 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 modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.
[0029] The present application embodiment provides a method for detecting, classifying and identifying tea leaves. Figure 1 As shown, the method includes:
[0030] Step S100, randomly divide the target batch of tea into N portions of tea, traverse the N portions of tea for feature recognition, mix the N portions of tea according to the feature recognition results, and obtain K portions of mixed tea. Specifically, the target batch of tea is randomly and evenly divided into N portions (N is an integer greater than 1), and the appearance features of each portion of tea are recognized, such as color, shape, texture, etc., using image recognition or other non-destructive detection technology. According to the feature recognition results, the portions of tea with similar features are merged to finally obtain K portions of mixed tea (K≤N).
[0031] In a possible implementation, step S100 further includes step S110, calling the visual acquisition unit to perform image acquisition on the N portions of tea leaves to obtain N image acquisition results. Specifically, activate or start the device or system responsible for image acquisition, such as a high-definition camera, a scanner or a dedicated image acquisition device, set acquisition parameters such as resolution, exposure time, focal length, etc. according to the characteristics of the tea leaves and the detection requirements, place the N portions of tea leaves one by one in the field of view of the visual acquisition unit, perform image acquisition operations, record the appearance characteristics of each portion of tea leaves, save the collected image data as a digital file, and form N image acquisition results. Step S120, perform tea feature recognition based on the N image acquisition results to obtain N tea feature sets. Specifically, preprocess the collected images, including denoising, contrast enhancement, cropping, etc., to improve image quality, extract feature information of tea leaves, such as color, shape, texture, edge, etc., from the preprocessed images using image processing algorithms or machine learning models, organize the extracted feature information into a structured data set, and form N tea feature sets, each set containing multiple feature values corresponding to the tea leaves. Step S130, similarity recognition is performed based on the N tea feature sets to obtain similarity recognition results. Specifically, a similarity measurement method (such as Euclidean distance, cosine similarity, Manhattan distance, etc.) is used to compare the N tea feature sets in pairs to calculate the similarity between them. According to the set similarity threshold, it is determined which tea feature sets have a high enough similarity and can be regarded as similar or the same category. The results of the similarity judgment are organized into a form that is easy to understand and process, such as a similarity matrix, a grouping list, etc., as a basis for the mixing operation. Step S140, based on the similarity recognition result, the N portions of tea are mixed to generate K portions of mixed tea. Specifically, according to the similarity recognition result, a tea mixing strategy is formulated, that is, which tea leaves should be mixed together to form a new mixed tea. According to the mixing strategy, similar tea leaves are merged together to form K portions of mixed tea. This implementation method uses machine vision and image processing algorithms to quickly and accurately identify the feature information of tea leaves, and perform similarity comparison and classification, thereby achieving the technical effect of improving the efficiency and accuracy of feature recognition.
[0032] Step S200, randomly extracting K first tea samples and K second tea samples of preset weight from K mixed teas, and performing physical and chemical tests on the K first tea samples and K second tea samples to obtain K first physical and chemical index sets and K second physical and chemical index sets. Specifically, two samples of preset tea weight are randomly extracted from each mixed tea, namely, K first tea samples and K second tea samples. Performing physical and chemical tests on the extracted samples, i.e., detecting the composition, properties or content of the samples by physical or chemical methods, including but not limited to moisture content, tea polyphenols content, caffeine content, heavy metal content, etc., to obtain K first physical and chemical index sets and K second physical and chemical index sets.
[0033] In a possible implementation, step S200 further includes step S210, using colorimetry to detect active ingredients in the K first tea samples and the K second tea samples, respectively, to generate K first active ingredient indices and K second active ingredient indices. Specifically, a series of standard solutions of active ingredients (caffeine, polyphenols, amino acids, etc.) with known concentrations are prepared, and these solutions are used to draw a standard curve. The K first tea samples and the second tea samples are subjected to pretreatment steps such as grinding and extraction to obtain a test solution containing active ingredients. The test solution is mixed with a specific color developer, so that the active ingredient reacts chemically with the color developer to generate a colored product, and the absorbance of the post-reaction solution is measured at a specific wavelength using a spectrophotometer. According to the standard curve and the measured absorbance, the content of the active ingredient in each sample is calculated to generate K first active ingredient indices and K second active ingredient indices. Step S220, using gas chromatography-mass spectrometry to detect pesticide residues on the K first tea samples and the K second tea samples, and generate K first pesticide residue indices and K second pesticide residue indices. Specifically, the tea samples are subjected to pre-treatment steps such as grinding, extraction, and purification to remove impurities and enrich pesticide residues. The treated samples are injected into a gas chromatograph, and the difference in the distribution coefficients of different substances between the gas phase and the stationary phase is used for separation. The separated compounds enter the mass spectrometer, and an ion beam is formed through ionization, acceleration, focusing and other processes, and then separated into ions with different mass-to-charge ratios through a mass analyzer, and qualitative and quantitative analysis is performed based on the intensity and mass-to-charge ratio of the ions. According to the comparison of the mass spectrum and the database, the type and content of pesticides in the sample are determined, and K first pesticide residue indices and K second pesticide residue indices are generated. Step S230, using atomic absorption spectroscopy to detect the nutritional components of the K first tea samples and the K second tea samples, and K first nutritional component indices and K second nutritional component indices are generated. Specifically, the tea sample is subjected to acid digestion and other treatments to convert the nutrients (minerals, vitamins, proteins, etc.) into soluble ion forms, and the digested sample is converted into gaseous atoms or ions by means of flame or graphite furnace, and the atomized sample is irradiated with light of a specific wavelength to measure the intensity of light absorbed by the specific element atoms. According to the relationship between absorbance and element concentration (Lambert-Beer law), the content of the nutrients in the sample is calculated to generate K first nutrient component indicators and K second nutrient component indicators. Step S240, the K first active component indicators, K first pesticide residue indicators and the K first nutrient component indicators are used as K first physical and chemical indicator sets; step S250, the K second active component indicators, K second pesticide residue indicators and the K second nutrient component indicators are used as K second physical and chemical indicator sets.This implementation method uses a combination of multiple analytical techniques (colorimetry, gas chromatography-mass spectrometry, atomic absorption spectrometry) to comprehensively and accurately detect the active ingredients, pesticide residues and nutrients in tea samples. Each technology has its unique advantages and scope of application, covering different aspects of tea quality assessment. By integrating these test results into a set of physical and chemical indicators, the technical effect of providing a scientific basis for tea quality evaluation is achieved.
[0034] Step S300, using the K second physical and chemical indicator sets, the K first physical and chemical indicator sets are identified for longitudinal consistency, and K longitudinal consistency factors are obtained. Specifically, the longitudinal consistency refers to the consistency of the physical and chemical properties between different samples within the same mixed tea. Specifically, the physical and chemical indicators of the two samples (the first and second samples) of each mixed tea are compared, and the difference in physical and chemical indicators of each comparison group (i.e., the two samples of each mixed tea) is calculated to obtain K longitudinal consistency factors, which reflect the degree of consistency within each mixed tea.
[0035] In a possible implementation, step S300 further includes step S310, constructing a longitudinal consistency identification formula, wherein the longitudinal consistency identification formula is:
[0036] ;
[0037] in, is the vertical consistency factor, n is the number of physical and chemical indicators, n is a positive integer, is the i-th physical and chemical index in the first physical and chemical index set, is the i-th physical and chemical index in the second physical and chemical index set, is the weight of the ith physical and chemical index in the vertical consistency identification. Specifically, the total number of physical and chemical indexes involved in the vertical consistency identification is n, and the weight reflects the importance of different physical and chemical indexes in evaluating the vertical consistency of tea samples. The weight distribution is determined based on expert knowledge, historical data or statistical analysis. Step S320, respectively input the K first physical and chemical index sets and the K second physical and chemical index sets into the vertical consistency identification formula to obtain the K vertical consistency factors. Specifically, for each physical and chemical index set pair (i.e., a set in the first set and a corresponding set in the second set), the vertical consistency identification formula is used for calculation, each physical and chemical index is traversed, its difference is calculated, and multiplied by the corresponding weight, and then all weighted differences are summed to output a vertical consistency factor, which is used as a quantitative indicator for evaluating the degree of vertical consistency between the two sets. This implementation method systematically evaluates the stability or consistency of the test results between different samples of the same mixed tea by constructing a vertical consistency identification formula and inputting a set of physical and chemical indicators to calculate the vertical consistency factor, thereby achieving the technical effect of improving the accuracy of vertical consistency identification.
[0038] Step S400, perform mixed verification of the mixed tea leaves based on the K vertical consistency factors. When the verification is passed, perform horizontal consistency identification between samples on the K first physical and chemical indicator sets and the K second physical and chemical indicator sets, respectively, to obtain K horizontal consistency factor sets. Specifically, based on the results of the vertical consistency factors, verify the uniformity of the mixed tea leaves. If all mixed tea leaves pass the verification (that is, the consistency factor is within an acceptable range), perform horizontal consistency identification, that is, compare the physical and chemical indicators of each portion of mixed tea leaves with the remaining K-1 portions of mixed tea leaves horizontally, analyze the similarity of the physical and chemical properties between different mixed tea leaves, and calculate K horizontal consistency factor sets, which reflect the similarity between different mixed tea leaves.
[0039] In a possible implementation, step S400 further includes step S410, enumerating the K first physical and chemical indicator sets in pairs to obtain multiple first enumeration combinations, wherein each first enumeration combination includes two first physical and chemical indicator sets. Specifically, initialize an empty list for storing the first enumeration combinations. Use two layers of loops to traverse the K first physical and chemical indicator sets, the outer loop selects a set as the first element of the combination, and the inner loop selects a set different from the outer loop as the second element of the combination. The two sets selected in each loop are added to the list as a combination. After the loop ends, all possible combinations of the two sets are stored in the list. Step S420, construct a horizontal consistency identification formula, wherein the horizontal consistency identification formula is:
[0040] ;
[0041] in, is the horizontal consistency factor, n is the number of physical and chemical indicators, is the jth physical and chemical index of any physical and chemical index set in an enumeration combination, is the jth physical and chemical index of another physical and chemical index set in an enumeration combination, is the weight of the jth physical and chemical indicator in the horizontal consistency identification, and n is a positive integer. Specifically, the horizontal consistency identification formula is used to quantify the degree of consistency between different physical and chemical indicator sets on the same or similar physical and chemical indicators. The horizontal consistency factor is a quantitative indicator that represents the degree of consistency between two physical and chemical indicator sets, and is used to evaluate whether the two sets present similar characteristics or trends in the physical and chemical indicators they contain. The higher the horizontal consistency factor, the higher the consistency between the two physical and chemical indicator sets. Step S430, based on the horizontal consistency identification formula, the multiple first enumeration combinations are respectively subjected to the horizontal consistency identification between samples to obtain multiple first horizontal consistency factors. Specifically, for each first enumeration combination, the consistency of the corresponding indicators in the two first physical and chemical indicator sets is calculated using the horizontal consistency identification formula, and a horizontal consistency factor is generated according to the calculation result. The factor reflects the degree of consistency between the two first physical and chemical indicator sets on the physical and chemical indicators, and the horizontal consistency factors of all enumeration combinations are stored. Step S440, the multiple first horizontal consistency factors are extracted based on the K first physical and chemical indicator sets to obtain K first horizontal consistency factor sets. Specifically, for any first physicochemical indicator set, extract its first horizontal consistency factor with all other sets from the multiple first horizontal consistency factors to obtain a first horizontal consistency factor set, repeat the above steps for K first physicochemical indicator sets, and finally obtain K first horizontal consistency factor sets. Step S450, use the horizontal consistency identification formula to perform horizontal consistency identification between samples on the K second physicochemical indicator sets respectively, obtain multiple second horizontal consistency factors, and use the K second physicochemical indicator sets as indexes to retrieve the multiple second horizontal consistency factors to obtain K second horizontal consistency factor sets. Specifically, it is similar to steps S410-S440, but applied to the second physicochemical indicator set. Step S460, based on the K first horizontal consistency factor sets and the K second horizontal consistency factor sets, a comprehensive analysis is performed to generate K horizontal consistency factor sets, wherein the K horizontal consistency factor sets correspond one-to-one to K portions of mixed tea, and the horizontal consistency factor is the degree of similarity between the physical and chemical test of each portion of mixed tea and any portion of mixed tea among the K portions of mixed tea.Specifically, each of the K mixed teas corresponds to a first horizontal consistency factor set and a second horizontal consistency factor set. The first and second horizontal consistency factor sets are merged into a comprehensive horizontal consistency factor set using a fusion method (such as weighted average, principal component analysis, etc.). When merging, the weights are adjusted according to the actual situation or other strategies are used to balance the influence of the two sets to obtain a comprehensive horizontal consistency factor set. The same operation is performed on each mixed tea to finally obtain K comprehensive horizontal consistency factor sets, each of which reflects the degree of consistency between the corresponding physical and chemical indicators of each mixed tea and other mixed teas in the K mixed teas. This implementation method comprehensively evaluates the consistency of the physical and chemical properties between different mixed teas through pairwise enumeration consistency identification and comprehensive analysis of the first horizontal consistency factor set and the second horizontal consistency factor set, achieving the technical effect of improving the comprehensiveness and accuracy of horizontal consistency identification.
[0042] In a possible implementation, step S400 further includes step S470, determining whether there is a longitudinal consistency factor less than a preset longitudinal consistency factor threshold among the K longitudinal consistency factors, if not, the verification is passed. Specifically, traverse the K longitudinal consistency factors, compare each factor with the preset longitudinal consistency factor threshold, and the longitudinal consistency factor threshold is used to measure the uniformity of tea mixing. If the K longitudinal consistency factors are all greater than or equal to the preset longitudinal consistency factor threshold, the mixing verification is passed. Step S480, if yes, obtain Q portions of longitudinal abnormal mixed tea, and perform overall mixing and feature identification on the Q portions of longitudinal abnormal mixed tea to obtain Q portions of updated mixed tea. Specifically, if there is a longitudinal consistency factor less than the preset longitudinal consistency factor threshold among the K longitudinal consistency factors, then identify those teas whose longitudinal consistency factors are lower than the threshold from the K portions of mixed tea, i.e., Q portions of longitudinal abnormal mixed tea. Mix these Q portions of tea as a whole in the hope of improving uniformity, perform feature identification on the mixed tea, such as color, shape, aroma, etc., to ensure mixing uniformity, and re-divide the mixed tea into Q portions of updated mixed tea. Step S490, respectively perform sample physicochemical tests on the Q updated mixed teas to obtain Q first updated physicochemical index sets and Q second updated physicochemical index sets. Specifically, two samples of preset tea weight are randomly selected from each updated mixed tea, namely Q first tea samples and Q second tea samples, and the selected samples are subjected to physicochemical tests, including but not limited to moisture content, tea polyphenol content, caffeine content, etc., to obtain the first updated physicochemical index set and the second updated physicochemical index set. Step S4100, perform longitudinal consistency identification and mixed verification according to the Q first updated physicochemical index sets and the Q second updated physicochemical index sets, and when the verification passes, update the K mixed teas according to the Q updated mixed teas. Specifically, use the updated physicochemical index set to perform longitudinal consistency identification, calculate the longitudinal consistency factor of the updated mixed teas, and determine whether the longitudinal consistency factors of all updated mixed teas are not less than the preset longitudinal consistency factor threshold. If all pass, replace the corresponding mixed teas in the original K mixed teas according to the physicochemical properties and quality of the Q updated mixed teas. Step S4110, the K first physical and chemical indicator sets and the K second physical and chemical indicator sets are updated according to the Q first updated physical and chemical indicator sets and the Q second updated physical and chemical indicator sets. Specifically, for each of the original K portions of mixed tea, the mixed tea corresponding to the Q portions of updated mixed tea is found, the physical and chemical indicator set of the updated mixed tea is replaced with the original physical and chemical indicator set, and the database or record file is updated to save the latest physical and chemical indicator information.If the longitudinal consistency factor of the mixed tea is lower than the threshold, it means that there is unevenness inside the tea. This implementation method improves the uniformity of the tea by remixing and re-dividing these abnormal mixed teas, thereby achieving the technical effect of ensuring the uniformity of the mixed tea itself.
[0043] Step S500, based on the K sets of horizontal consistency factors, directional secondary mixing of the K portions of mixed tea leaves is performed to obtain M portions of divided tea leaves. Specifically, according to the horizontal consistency factors, the mixed tea leaves with high similarity are classified into the same group, that is, the K portions of mixed tea leaves are mixed according to the consistency of physical and chemical properties, and the mixed tea leaves with the same or similar physical and chemical properties are mixed to finally obtain M portions of divided tea leaves (M≤K). Among them, step S100 is to mix tea leaves with similar appearance characteristics in N portions of tea leaves according to the similarity of the appearance of the tea leaves, so as to obtain K portions of mixed tea leaves; step S500 further mixes tea leaves with similar physical and chemical characteristics in K portions of mixed tea leaves according to the similarity of the physical and chemical properties of the tea leaves, so as to obtain M portions of divided tea leaves.
[0044] In a possible implementation, step S500 further includes step S510, randomly selecting M portions of mixed tea from the K portions of mixed tea as M decision nodes, and constructing a directional secondary mixed decision tree, wherein the horizontal consistency factor of any two portions of mixed tea from the M portions of mixed tea is less than or equal to the preset horizontal consistency factor. Specifically, from the K portions of mixed tea, randomly but non-repeatedly select M portions of tea as the root node or decision node of the decision tree, and the selection here should ensure that the horizontal consistency factor (an index indicating the similarity of physical and chemical properties) between any two selected mixed teas is less than or equal to the preset threshold value to ensure the representativeness and difference of the node. With these M portions of mixed tea as the starting point, a directional secondary mixed decision tree is constructed. Step S520, input the K portions of mixed tea into the directional secondary mixed decision tree, analyze the horizontal consistency factor of each portion of mixed tea with the M decision nodes in turn, and when the horizontal consistency factor with the decision node is the maximum value in the horizontal consistency factor set of the mixed tea, add the mixed tea to the internal node of the corresponding decision node to obtain M internal nodes. Specifically, K portions of mixed tea are input one by one into the constructed directional quadratic mixed decision tree. For each portion of mixed tea, its horizontal consistency factor with all decision nodes is analyzed, and each portion of mixed tea is assigned to the internal node corresponding to the decision node with the highest horizontal consistency factor, that is, mixed teas with similar physical and chemical properties are grouped together, that is, the internal nodes represent a set of mixed teas with similar specific physical and chemical properties. Step S530, the mixed teas in the M internal nodes are mixed to obtain the M portions of divided tea. Specifically, all the mixed teas in each internal node are mixed to form a new portion of divided tea with higher internal physical and chemical property consistency, and finally M portions of divided tea are obtained, and the physical and chemical properties of each portion of tea are consistent. This implementation method realizes the orderly management and optimized mixing of mixed teas by constructing a decision tree and assigning mixed teas to internal nodes, reduces unnecessary mixing steps and costs, and achieves the technical effect of optimizing the mixing process.
[0045] Step S600, traverse the M divided tea leaves for physical and chemical testing, generate M divided physical and chemical testing sets, and obtain M divided tea leaves grades according to the M divided physical and chemical testing sets. Specifically, perform physical and chemical testing on the M divided tea leaves, generate M divided physical and chemical testing sets, and grade each divided tea leaf according to the results of the physical and chemical testing sets and the preset rating standard, that is, classify or rate the tea leaves quality according to the physical and chemical testing data.
[0046] Step S700, the M divided tea leaves and the M divided tea leaves grades are used as the detection, classification and identification results of the target batch of tea leaves. Specifically, the M divided tea leaves and their corresponding grades are output as the detection, classification and identification results of the target batch of tea leaves, that is, the detection, classification and identification results refer to the tea classification and grade information obtained after a series of detection, identification and classification. The embodiment of the present application adopts technical means such as randomly dividing the target batch of tea leaves, mixing them according to appearance characteristics, conducting physical and chemical tests, and conducting horizontal and vertical consistency identification, so as to realize the accurate classification and identification of tea leaves, and achieve the technical effect of improving the accuracy and reliability of the classification results.
[0047] In the above, refer to Figure 1 A method for detecting, classifying and identifying tea leaves according to an embodiment of the present invention is described in detail. Figure 2 A tea leaf detection, classification and identification device according to an embodiment of the present invention is described.
[0048] A tea detection, classification and identification device according to an embodiment of the present invention is used to solve the technical problem of low accuracy and reliability of classification results in existing tea detection, classification and identification, and achieve the technical effect of improving the accuracy and reliability of classification results. A tea detection, classification and identification device includes: a feature recognition module 10, a physical and chemical detection module 20, a longitudinal consistency recognition module 30, a transverse consistency recognition module 40, a secondary mixing module 50, a tea grade acquisition module 60, and a detection, classification and identification result acquisition module 70.
[0049] The feature recognition module 10 is used to randomly divide the target batch of tea leaves into N portions of tea leaves, traverse the N portions of tea leaves for feature recognition, and mix the N portions of tea leaves according to the feature recognition results to obtain K portions of mixed tea leaves; the physical and chemical detection module 20 is used to randomly extract K portions of first tea samples and K portions of second tea samples of preset mass from the K portions of mixed tea leaves, respectively, and perform physical and chemical detection on the K portions of first tea samples and the K portions of second tea samples to obtain K first physical and chemical indicator sets and K second physical and chemical indicator sets; the longitudinal consistency recognition module 30 is used to use the K second physical and chemical indicator sets to perform longitudinal consistency recognition on the K first physical and chemical indicator sets to obtain K longitudinal consistency factors; the transverse consistency recognition module 40 is used to perform longitudinal consistency recognition on the K first physical and chemical indicator sets based on the K longitudinal consistency factors. The mixed tea is verified based on the consistency factor. When the verification is passed, the K first physical and chemical indicator sets and the K second physical and chemical indicator sets are respectively identified for cross-sample consistency to obtain K cross-sample consistency factor sets. The secondary mixing module 50 is used to perform directional secondary mixing of the K portions of mixed tea based on the K cross-sample consistency factor sets to obtain M portions of divided tea. The tea grade acquisition module 60 is used to traverse the M portions of divided tea for physical and chemical testing, generate M divided physical and chemical test sets, and obtain M divided tea grades according to the M divided physical and chemical test sets. The detection classification recognition result acquisition module 70 is used to use the M portions of divided tea and the M divided tea grades as the detection classification recognition results of the target batch of tea.
[0050] The specific configuration of the feature recognition module 10 will be described in detail below. As described above, the feature recognition module 10 may further include: an image acquisition unit for calling the visual acquisition unit to perform image acquisition on the N portions of tea leaves to obtain N image acquisition results; a tea feature recognition unit for performing tea feature recognition based on the N image acquisition results to obtain N tea feature sets; a similarity recognition unit for performing similarity recognition based on the N tea feature sets to obtain similarity recognition results; and a tea mixing unit for mixing the N portions of tea leaves based on the similarity recognition results to generate K portions of mixed tea leaves.
[0051] The specific configuration of the physical and chemical detection module 20 will be described in detail below. As described above, the physical and chemical detection module 20 may further include: an active ingredient detection unit for performing active ingredient detection on the K first tea samples and the K second tea samples respectively by colorimetry, generating K first active ingredient indices and K second active ingredient indices; a pesticide residue detection unit for performing pesticide residue detection on the K first tea samples and the K second tea samples by gas chromatography-mass spectrometry, generating K first pesticide residue indices and K second pesticide residue indices; a nutrient component detection unit for performing nutrient component detection on the K first tea samples and the K second tea samples by atomic absorption spectroscopy, generating K first nutrient component indices and K second nutrient component indices; a physicochemical indicator set generation unit for using the K first active ingredient indices, the K first pesticide residue indices and the K first nutrient component indices as K first physicochemical indicator sets, and using the K second active ingredient indices, the K second pesticide residue indices and the K second nutrient component indices as K second physicochemical indicator sets.
[0052] The specific configuration of the longitudinal consistency identification module 30 will be described in detail below. As described above, the longitudinal consistency identification module 30 may further include: a longitudinal consistency identification formula construction unit for constructing a longitudinal consistency identification formula, wherein the longitudinal consistency identification formula is: ,in, is the vertical consistency factor, n is the number of physical and chemical indicators, n is a positive integer, is the i-th physical and chemical index in the first physical and chemical index set, is the i-th physical and chemical index in the second physical and chemical index set, is the weight of the i-th physical and chemical indicator in the vertical consistency identification; the vertical consistency factor acquisition unit is used to input the K first physical and chemical indicator sets and the K second physical and chemical indicator sets into the vertical consistency identification formula respectively to obtain the K vertical consistency factors.
[0053] The specific configuration of the transverse consistency identification module 40 will be described in detail below. As described above, the transverse consistency identification module 40 may further include: a pairwise enumeration unit for enumerating the K first physical and chemical indicator sets in pairs to obtain a plurality of first enumeration combinations, wherein each first enumeration combination includes two first physical and chemical indicator sets; and a transverse consistency identification formula construction unit for constructing a transverse consistency identification formula, wherein the transverse consistency identification formula is: ;in, is the horizontal consistency factor, n is the number of physical and chemical indicators, is the jth physical and chemical index of any physical and chemical index set in an enumeration combination, is the jth physical and chemical index of another physical and chemical index set in an enumeration combination, is the weight of the jth physical and chemical indicator in the horizontal consistency identification, and n is a positive integer; the inter-sample horizontal consistency identification unit is used to perform inter-sample horizontal consistency identification on the multiple first enumeration combinations based on the horizontal consistency identification formula to obtain multiple first horizontal consistency factors; the horizontal consistency factor set acquisition unit is used to extract the multiple first horizontal consistency factors based on the K first physical and chemical indicator sets to obtain K first horizontal consistency factor sets, use the horizontal consistency identification formula to perform inter-sample horizontal consistency identification on the K second physical and chemical indicator sets to obtain multiple second horizontal consistency factors, and use the K second physical and chemical indicator sets as indexes to retrieve the multiple second horizontal consistency factors to obtain K second horizontal consistency factor sets, and perform comprehensive analysis based on the K first horizontal consistency factor sets and the K second horizontal consistency factor sets to generate K horizontal consistency factor sets, wherein the K horizontal consistency factor sets correspond one-to-one to the K mixed teas, and the horizontal consistency factor is the degree of similarity in physical and chemical detection between each mixed tea and any one of the K mixed teas.
[0054] Among them, the horizontal consistency identification module 40 may further include: a verification unit is used to determine whether there is a longitudinal consistency factor less than a preset longitudinal consistency factor threshold among the K longitudinal consistency factors, and if not, the verification is passed; an updated mixed tea acquisition unit is used to, if yes, obtain Q portions of longitudinal abnormal mixed tea, and perform overall mixing and feature identification on the Q portions of longitudinal abnormal mixed tea to obtain Q portions of updated mixed tea; a sample physicochemical detection unit is used to perform sample physicochemical detection on the Q portions of updated mixed tea, respectively, to obtain Q first updated physicochemical indicator sets and Q second updated physicochemical indicator sets; the update unit performs longitudinal consistency identification and mixing verification based on the Q first updated physicochemical indicator sets and the Q second updated physicochemical indicator sets, and when the verification is passed, the K portions of mixed tea are updated according to the Q portions of updated mixed tea, and the K first physicochemical indicator sets and the K second physicochemical indicator sets are updated according to the Q first updated physicochemical indicator sets and the Q second updated physicochemical indicator sets.
[0055] The specific configuration of the secondary mixing module 50 will be described in detail below. As described above, the secondary mixing module 50 may further include: a directional secondary mixing decision tree construction unit for randomly selecting M portions of mixed tea from the K portions of mixed tea as M decision nodes to construct a directional secondary mixing decision tree, wherein the horizontal consistency factor of any two portions of mixed tea in the M portions of mixed tea is less than or equal to a preset horizontal consistency factor; an internal node acquisition unit for inputting the K portions of mixed tea into the directional secondary mixing decision tree, sequentially analyzing the horizontal consistency factor of each portion of mixed tea with the M decision nodes, and when the horizontal consistency factor with the decision node is the maximum value in the set of horizontal consistency factors of the mixed tea, adding the mixed tea into the internal node of the corresponding decision node to obtain M internal nodes; and a divided tea acquisition unit for mixing the mixed tea in the M internal nodes to obtain the M portions of divided tea.
[0056] A tea detection, classification and identification device provided in an embodiment of the present invention can execute a tea detection, classification and identification method provided in any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.
[0057] Although the present application makes various references to certain modules in the device according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.
[0058] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
Claims
1. A method for detecting, classifying and identifying tea leaves, characterized in that: The method comprises: Randomly divide the target batch of tea into N portions of tea, traverse the N portions of tea to perform feature recognition, and mix the N portions of tea according to the feature recognition results to obtain K portions of mixed tea; Randomly extracting K first tea samples and K second tea samples of preset mass from K mixed teas respectively, and performing physical and chemical tests on the K first tea samples and the K second tea samples to obtain K first physical and chemical index sets and K second physical and chemical index sets; Using the K second physical and chemical indicator sets, performing longitudinal consistency identification on the K first physical and chemical indicator sets to obtain K longitudinal consistency factors; Performing mixed verification of the mixed tea leaves based on the K vertical consistency factors, and when the verification is passed, performing horizontal consistency identification between samples on the K first physical and chemical indicator sets and the K second physical and chemical indicator sets, respectively, to obtain K horizontal consistency factor sets; Performing directional secondary mixing of the K portions of mixed tea leaves based on the K sets of lateral consistency factors to obtain M portions of divided tea leaves; Traversing the M divided tea leaves for physical and chemical testing, generating M divided physical and chemical testing sets, and obtaining M divided tea leaves grades according to the M divided physical and chemical testing sets; The M divided tea leaves and the M divided tea leaf grades are used as the detection, classification and identification results of the target batch of tea leaves; A longitudinal consistency identification formula is constructed, wherein the longitudinal consistency identification formula is: ; in, is the vertical consistency factor, n is the number of physical and chemical indicators, n is a positive integer, is the i-th physical and chemical index in the first physical and chemical index set, is the i-th physical and chemical index in the second physical and chemical index set, is the weight of the ith physical and chemical index in the vertical consistency identification; Inputting the K first physical and chemical indicator sets and the K second physical and chemical indicator sets into a longitudinal consistency identification formula respectively to obtain the K longitudinal consistency factors; Enumerate the K first physical and chemical indicator sets in pairs to obtain a plurality of first enumeration combinations, wherein each first enumeration combination includes two first physical and chemical indicator sets; Construct a horizontal consistency identification formula, wherein the horizontal consistency identification formula is: ; in, is the horizontal consistency factor, n is the number of physical and chemical indicators, is the jth physical and chemical index of any physical and chemical index set in an enumeration combination, is the jth physical and chemical index of another physical and chemical index set in an enumeration combination, is the weight of the jth physical and chemical index in horizontal consistency identification, and n is a positive integer; Based on the horizontal consistency identification formula, the multiple first enumeration combinations are respectively subjected to horizontal consistency identification between samples to obtain multiple first horizontal consistency factors; Extracting the plurality of first transverse consistency factors based on the K first physical and chemical indicator sets to obtain K first transverse consistency factor sets; Using the horizontal consistency identification formula, respectively, the K second physical and chemical indicator sets are subjected to horizontal consistency identification between samples to obtain a plurality of second horizontal consistency factors, and the K second physical and chemical indicator sets are used as indexes to retrieve the plurality of second horizontal consistency factors to obtain K second horizontal consistency factor sets; Based on the K first horizontal consistency factor sets and the K second horizontal consistency factor sets, a comprehensive analysis is performed to generate K horizontal consistency factor sets, wherein the K horizontal consistency factor sets correspond to the K portions of mixed tea one by one, and the horizontal consistency factor is the degree of similarity between each portion of mixed tea and any portion of the K portions of mixed tea in physical and chemical tests; Randomly selecting M portions of mixed tea from the K portions of mixed tea as M decision nodes, and constructing a directional secondary mixed decision tree, wherein the lateral consistency factor of any two portions of mixed tea from the M portions of mixed tea is less than or equal to a preset lateral consistency factor; Input the K portions of mixed tea into the directional secondary mixed decision tree, analyze the horizontal consistency factor of each portion of mixed tea with the M decision nodes in turn, and when the horizontal consistency factor with the decision node is the maximum value in the set of horizontal consistency factors of mixed tea, add the mixed tea into the internal node of the corresponding decision node to obtain M internal nodes; The mixed tea leaves in the M internal nodes are mixed to obtain the M portions of divided tea leaves.
2. A tea detection, classification and identification method as claimed in claim 1, characterized in that: include: Calling a visual acquisition unit to acquire images of the N portions of tea leaves to obtain N image acquisition results; Perform tea feature recognition based on the N image acquisition results to obtain N tea feature sets; Performing similarity recognition based on the N tea feature sets to obtain a similarity recognition result; The N portions of tea are mixed based on the similarity recognition results to generate K portions of mixed tea.
3. A tea detection, classification and identification method as claimed in claim 1, characterized in that: include: Using a colorimetric method, respectively, the K portions of the first tea samples and the K portions of the second tea samples are tested for active ingredients to generate K first active ingredient indices and K second active ingredient indices; Using a gas chromatography-mass spectrometry method to perform pesticide residue detection on the K first tea samples and the K second tea samples to generate K first pesticide residue indices and K second pesticide residue indices; Using atomic absorption spectroscopy to detect the nutritional components of the K first tea samples and the K second tea samples, and generate K first nutritional component indices and K second nutritional component indices; The K first active ingredient indices, the K first pesticide residue indices, and the K first nutrient component indices are used as K first physical and chemical indicator sets; The K second active ingredient indices, the K second pesticide residue indices and the K second nutrient component indices are taken as K second physical and chemical indicator sets.
4. A tea detection, classification and identification method as claimed in claim 1, characterized in that: include: Determine whether there is a longitudinal consistency factor of the K longitudinal consistency factors that is less than a preset longitudinal consistency factor threshold, if not, the verification is passed; If yes, obtain Q portions of abnormal longitudinal mixed tea leaves, and perform overall mixing and feature recognition on the Q portions of abnormal longitudinal mixed tea leaves to obtain Q portions of updated mixed tea leaves; Performing physical and chemical tests on the Q updated mixed tea samples respectively to obtain Q first updated physical and chemical index sets and Q second updated physical and chemical index sets; Performing longitudinal consistency identification and mixed verification according to the Q first updated physical and chemical indicator sets and the Q second updated physical and chemical indicator sets, and when the verification passes, updating the K portions of mixed tea according to the Q portions of updated mixed tea; The K first physicochemical indicator sets and the K second physicochemical indicator sets are updated according to the Q first updated physicochemical indicator sets and the Q second updated physicochemical indicator sets.
5. A tea detection, classification and identification device, characterized in that: The device is used to implement a tea detection, classification and identification method according to any one of claims 1 to 4, and the device comprises: A feature recognition module, wherein the feature recognition module is used to randomly divide a target batch of tea leaves into N portions of tea leaves, traverse the N portions of tea leaves for feature recognition, and mix the N portions of tea leaves according to the feature recognition results to obtain K portions of mixed tea leaves; A physical and chemical testing module, the physical and chemical testing module is used to randomly extract K first tea samples and K second tea samples of preset mass from K mixed teas, and perform physical and chemical testing on the K first tea samples and the K second tea samples to obtain K first physical and chemical index sets and K second physical and chemical index sets; A longitudinal consistency identification module, the longitudinal consistency identification module is used to use the K second physical and chemical indicator sets to perform longitudinal consistency identification on the K first physical and chemical indicator sets to obtain K longitudinal consistency factors; A transverse consistency identification module, wherein the transverse consistency identification module is used to perform mixed verification of mixed tea leaves based on the K longitudinal consistency factors. When the verification is passed, the K first physical and chemical index sets and the K second physical and chemical index sets are respectively subjected to transverse consistency identification between samples to obtain K transverse consistency factor sets; A secondary mixing module, the secondary mixing module is used to perform directional secondary mixing of the K portions of mixed tea leaves based on the K sets of lateral consistency factors to obtain M portions of divided tea leaves; A tea grade acquisition module, the tea grade acquisition module is used to traverse the M divided tea leaves for physical and chemical testing, generate M divided physical and chemical testing sets, and obtain M divided tea grades according to the M divided physical and chemical testing sets; A detection classification recognition result acquisition module, wherein the detection classification recognition result acquisition module is used to use the M divided tea leaves and the M divided tea leaf grades as the detection classification recognition result of the target batch of tea leaves; The longitudinal consistency identification module further includes: a longitudinal consistency identification formula construction unit for constructing a longitudinal consistency identification formula, wherein the longitudinal consistency identification formula is: ,in, is the vertical consistency factor, n is the number of physical and chemical indicators, n is a positive integer, is the i-th physical and chemical index in the first physical and chemical index set, is the i-th physical and chemical index in the second physical and chemical index set, is the weight of the i-th physical and chemical indicator in the vertical consistency identification; the vertical consistency factor acquisition unit is used to input the K first physical and chemical indicator sets and the K second physical and chemical indicator sets into the vertical consistency identification formula respectively to obtain the K vertical consistency factors; The transverse consistency identification module further includes: a pairwise enumeration unit for enumerating the K first physical and chemical indicator sets in pairs to obtain multiple first enumeration combinations, wherein each first enumeration combination includes two first physical and chemical indicator sets; a transverse consistency identification formula construction unit for constructing a transverse consistency identification formula, wherein the transverse consistency identification formula is: ,in, is the horizontal consistency factor, n is the number of physical and chemical indicators, is the jth physical and chemical index of any physical and chemical index set in an enumeration combination, is the jth physical and chemical index of another physical and chemical index set in an enumeration combination, is the weight of the jth physical and chemical indicator in the horizontal consistency identification, and n is a positive integer; the inter-sample horizontal consistency identification unit is used to perform inter-sample horizontal consistency identification on the multiple first enumeration combinations based on the horizontal consistency identification formula to obtain multiple first horizontal consistency factors; the horizontal consistency factor set acquisition unit is used to extract the multiple first horizontal consistency factors based on the K first physical and chemical indicator sets to obtain K first horizontal consistency factor sets, use the horizontal consistency identification formula to perform inter-sample horizontal consistency identification on the K second physical and chemical indicator sets to obtain multiple second horizontal consistency factors, and use the K second physical and chemical indicator sets as indexes to retrieve the multiple second horizontal consistency factors to obtain K second horizontal consistency factor sets, and perform comprehensive analysis based on the K first horizontal consistency factor sets and the K second horizontal consistency factor sets to generate K horizontal consistency factor sets, wherein the K horizontal consistency factor sets correspond one-to-one to the K portions of mixed tea, and the horizontal consistency factor is the degree of similarity between each portion of mixed tea and any portion of mixed tea in the K portions of mixed tea in physical and chemical testing; The secondary mixing module further includes: a directional secondary mixing decision tree construction unit for randomly selecting M portions of mixed tea from the K portions of mixed tea as M decision nodes to construct a directional secondary mixing decision tree, wherein the lateral consistency factor of any two portions of mixed tea among the M portions of mixed tea is less than or equal to a preset lateral consistency factor; an internal node acquisition unit for inputting the K portions of mixed tea into the directional secondary mixing decision tree, and analyzing the lateral consistency factor of each portion of mixed tea with the M decision nodes in turn, and when the lateral consistency factor with the decision node is the maximum value in the set of lateral consistency factors of the mixed tea, adding the mixed tea into the internal node of the corresponding decision node to obtain M internal nodes; and a divided tea acquisition unit for mixing the mixed tea in the M internal nodes to obtain the M portions of divided tea.
