Multi-mode dairy acidity detection method and system

Abstract

The invention provides a multi-mode dairy acidity detection method and system, and is suitable for the technical field of data processing, and the method comprises the steps: generating a plurality of pieces of to-be-detected dairy association information according to to-be-detected dairy spectral information, to-be-detected dairy electrochemical information, to-be-detected dairy physical parameter information and a preset to-be-detected dairy information association vector; and based on preset dairy acidity grade quantity information, performing classification processing according to the multiple pieces of to-be-detected dairy association information, and generating multiple pieces of to-be-detected dairy acidity detection information. The method and the device are used for realizing multi-dimensional and all-directional detection of the acidity of the dairy product, the accuracy and the reliability of detection of the acidity of the dairy product are remarkably improved, different acidity grade division requirements are met, and the application scene of rapid and accurate detection of batch samples in dairy product production and quality inspection is met.

Description

Technical Field

[0001] This application belongs to the field of data processing technology, and in particular relates to a multimodal method and system for detecting the acidity of dairy products. Background Technology

[0002] Acidity testing of dairy products is a crucial step in ensuring their quality and safety. Currently, the industry's testing technologies have evolved from traditional methods to modern technologies. Traditional testing primarily uses titratable acidity methods and pH measurement methods. Titratable acidity methods offer higher accuracy but are complex and time-consuming to operate, while pH measurement methods are convenient and suitable for rapid testing on production lines, but their accuracy is slightly lower.

[0003] However, in existing technologies, detection schemes that rely solely on single-modal data are susceptible to environmental interference or the influence of other components in the sample itself, resulting in limited detection accuracy and insufficient stability of the detection results. This leads to low efficiency in batch sample detection scenarios and makes it difficult to meet the high-throughput and high-precision quality inspection requirements of modern dairy production. Summary of the Invention

[0004] In view of this, embodiments of this application provide a multimodal dairy acidity detection method and system, aiming to solve the problems of insufficient single-modal data detection and low batch detection efficiency in the prior art.

[0005] The first aspect of this application provides a multimodal method for detecting the acidity of dairy products, including:

[0006] Acquire spectral information, electrochemical information, and physical parameter information of multiple dairy products to be tested;

[0007] Based on the spectral information, electrochemical information, physical parameter information, and preset association vectors of the dairy products to be tested, multiple association information of the dairy products to be tested is generated.

[0008] Based on the preset number of dairy acidity levels, the multiple dairy products to be tested are classified and processed according to their associated information to generate multiple dairy acidity detection information.

[0009] A second aspect of this application provides a multimodal dairy acidity detection system, comprising:

[0010] The dairy product information acquisition module is used to acquire spectral information, electrochemical information, and physical parameter information of multiple dairy products to be tested.

[0011] The module for generating association information of dairy products to be tested is used to generate multiple association information of dairy products to be tested based on the multiple spectral information of dairy products to be tested, multiple electrochemical information of dairy products to be tested, multiple physical parameter information of dairy products to be tested, and multiple preset association vectors of information of dairy products to be tested.

[0012] The dairy product acidity detection information generation module is used to classify and process multiple dairy products based on preset dairy product acidity level quantity information and the associated information of the multiple dairy products to be tested, and generate multiple dairy product acidity detection information.

[0013] A third aspect of this application provides a terminal device, the terminal device including a memory and a processor, the memory storing a computer program executable on the processor, the processor executing the computer program to implement the steps of the multimodal dairy acidity detection method described in the first aspect above.

[0014] A fourth aspect of this application provides a computer-readable storage medium, comprising: storing a computer program, wherein when executed by a processor, the computer program implements the steps of the multimodal dairy acidity detection method described in the first aspect above.

[0015] The beneficial effects of this application embodiment compared with the prior art are: this application effectively integrates key information related to dairy acidity that is difficult to cover with a single data dimension, avoids the detection deviation caused by the limitation of data dimension in the single detection method in the prior art, and is used to realize multi-dimensional and all-round detection of dairy acidity, thereby significantly improving the accuracy and reliability of dairy acidity detection, and adapting to the needs of different acidity level classification, meeting the application scenario of rapid and accurate detection of batch samples in dairy production quality inspection. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the implementation process of the multimodal dairy acidity detection method provided in Embodiment 1 of this application;

[0018] Figure 2 This is a schematic diagram of the implementation process of the multimodal dairy acidity detection method provided in Embodiment 2 of this application;

[0019] Figure 3 This is a schematic diagram of the implementation process of the multimodal dairy acidity detection method provided in Embodiment 3 of this application;

[0020] Figure 4 This is a schematic diagram of the implementation process of the multimodal dairy acidity detection method provided in Embodiment 4 of this application;

[0021] Figure 5 This is a schematic diagram of the implementation process of the multimodal dairy acidity detection method provided in Embodiment 5 of this application;

[0022] Figure 6 This is a schematic diagram of the implementation process of the multimodal dairy acidity detection method provided in Embodiment Six of this application;

[0023] Figure 7 This is a schematic diagram of the implementation process of the multimodal dairy acidity detection method provided in Embodiment 7 of this application;

[0024] Figure 8 This is a schematic diagram of the structure of the multimodal dairy acidity detection system provided in the embodiments of this application;

[0025] Figure 9 This is a schematic diagram of the terminal device provided in the embodiments of this application. Detailed Implementation

[0026] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0027] To illustrate the technical solution described in this application, specific embodiments are provided below.

[0028] Figure 1 A flowchart illustrating the implementation of the multimodal dairy acidity detection method provided in Embodiment 1 of this application is shown, and detailed below:

[0029] Step S101: Obtain spectral information, electrochemical information, and physical parameter information of multiple dairy products to be tested.

[0030] In this embodiment, the spectral information of the dairy products to be tested can refer to the absorbance or reflectance data sequence of each sample at different wavelengths, collected after non-contact scanning of multiple dairy product samples by a near-infrared spectrometer. This data reflects the differences in absorption and reflection of infrared light of specific wavelengths by dairy products with different acidities. The electrochemical information of the dairy products to be tested can refer to the digital signal data obtained by immersing electrochemical sensors such as pH electrodes and ion-selective electrodes into multiple dairy product samples, utilizing the electrochemical reaction between the electrodes and hydrogen ions and organic acid anions in the samples to convert the ion concentration change into a measurable potential or current signal, which is then recorded and converted by a data acquisition instrument. The physical parameter information of the dairy products to be tested can refer to the density, viscosity, and refractive index data obtained by measuring multiple dairy product samples separately using professional testing instruments, as well as the ambient temperature data of each sample collected in real time by a temperature sensor. This data can be accurately measured by specialized equipment such as near-infrared spectrometers, electrochemical sensors, densitometers, viscometers, refractometers, and temperature sensors.

[0031] Step S102: Generate multiple dairy product association information based on the multiple spectral information of the dairy products to be tested, the multiple electrochemical information of the dairy products to be tested, the multiple physical parameter information of the dairy products to be tested, and the multiple preset association vectors of the dairy products to be tested.

[0032] In this embodiment, the multiple preset association vectors for dairy product information to be tested can be manually set. First, the spectral information, electrochemical information, and physical parameter information of the dairy products to be tested can be standardized and preprocessed to unify data dimensions and units, eliminating systematic errors caused by different testing equipment. Then, the multiple preset association vectors for dairy product information to be tested are matched with the three standardized data types at the feature level to uncover the correspondence between each data type and the acidity-related features in the association vectors. Next, the three matched data types are aligned in terms of feature dimensions to ensure consistency of different modal data in the feature space. Then, differentiated weights are assigned based on the degree of matching between each modal data and the association vector, highlighting the key features for detecting dairy product acidity and weakening the influence of irrelevant data. Finally, a weighted fusion algorithm is used to integrate the weighted three types of data with the multiple preset association vectors for dairy product information to be tested, generating multiple association information for dairy products to be tested that comprehensively reflects the intrinsic correlation of multimodal data and the acidity characteristics of dairy products.

[0033] Step S103: Based on the preset number of dairy acidity levels, classify and process the multiple dairy products to be tested according to their association information to generate multiple dairy product acidity detection information.

[0034] In this embodiment, the preset number of dairy acidity levels can be manually set. First, the number of classification categories can be determined based on the preset number of dairy acidity levels, defining a clear range for the classification of multiple dairy product association information to be tested. Then, a corresponding number of initial category centers are selected from the multiple dairy product association information to be tested as a benchmark for classification. Next, the similarity between each dairy product association information to be tested and all initial category centers is calculated. Based on the similarity results, each dairy product association information to be tested is assigned to the most suitable category, completing the first round of classification. Then, the category center of the corresponding category is updated based on the feature mean of all dairy product association information to be tested in each category, making the category center more closely match the core features of the category. This iterative process of similarity calculation, category assignment, and category center updating is repeated until the category centers tend to stabilize. Finally, based on the stable category to which each dairy product association information to be tested belongs, combined with the specific acidity level label corresponding to the preset number of dairy acidity levels, multiple dairy product acidity detection information containing dairy acidity levels and related feature descriptions are generated.

[0035] The multimodal dairy acidity detection method provided in this application effectively integrates key information related to dairy acidity that is difficult to cover with a single data dimension. It avoids the detection bias caused by the limitation of data dimension in the single detection method in the prior art. It is used to realize multi-dimensional and all-round detection of dairy acidity, thereby significantly improving the accuracy and reliability of dairy acidity detection. It is also adapted to the different acidity level classification requirements and meets the application scenarios of rapid and accurate detection of batch samples in dairy production quality inspection.

[0036] Figure 2 The flowchart illustrating the implementation of the multimodal dairy acidity detection method provided in Embodiment 2 of this application is shown. The difference between this method and Embodiment 1 described above is that:

[0037] Multiple preset information association vectors for dairy products to be detected include preset acidity feature localization vectors for dairy products to be detected, preset acidity feature matching vectors for dairy products to be detected, and preset acidity feature quantization mapping vectors for dairy products to be detected.

[0038] Step S102 specifically includes:

[0039] Step S201: Based on the multiple spectral information, multiple electrochemical information, and multiple physical parameter information of the dairy products to be tested, the data are spliced ​​and format converted to generate multiple information vectors of the dairy products to be tested.

[0040] In this embodiment, the splicing process integrates multiple spectral information, electrochemical information, and physical parameter information of the dairy products to be tested into a unified data sequence in a preset order. The format conversion process converts the integrated multi-source data into a unified data format that meets the requirements of subsequent calculations, thereby eliminating the differences in storage structure between different types of data. Through the coordinated operation of splicing and format conversion, multiple information vectors of the dairy products to be tested that can fully carry the original multimodal information are generated.

[0041] Step S202: Multiply the multiple dairy product information vectors to be detected and the preset dairy product acidity feature positioning vectors to be detected to generate multiple dairy product acidity feature positioning information.

[0042] In this embodiment, the preset acidity feature localization vector of the dairy product to be detected can be set manually. It can be used to clearly point to the core feature dimension related to the acidity of the dairy product. By multiplying multiple dairy product information vectors to be detected with this localization vector, the feature components that are highly related to acidity in the multiple dairy product information vectors to be detected are highlighted, thereby suppressing the interference of irrelevant features. Then, the feature positions that play a key role in acidity detection in each dairy product information vector to be detected are accurately locked, thereby generating multiple dairy product acidity feature localization information that can identify the location of key acidity features.

[0043] Step S203: Multiply the multiple dairy product information vectors to be detected with the preset acidity feature matching vectors of the dairy products to be detected to generate multiple acidity feature matching information of the dairy products to be detected.

[0044] In this embodiment, the preset acidity feature matching vector of the dairy product to be tested can be manually set and can be used to carry known acidity feature pattern information. Multiple dairy product information vectors to be tested can be multiplied with this matching vector to quantify the degree of fit between each feature in the multiple dairy product information vectors and the known acidity feature pattern, thereby obtaining a matching score for each feature dimension. This score then reflects the correlation strength between different dairy product information vectors to be tested and the acidity feature pattern, thus generating multiple dairy product acidity feature matching information that can reflect the degree of feature matching.

[0045] Step S204: Multiply the multiple acidity feature location information and multiple acidity feature matching information of the dairy products to be detected to generate multiple acidity feature location matching information of the dairy products to be detected.

[0046] In this embodiment, multiple acidity feature location information of dairy products to be detected can be multiplied with multiple acidity feature matching information of dairy products to be detected, and then the location identifier of the key feature and the matching degree of the feature can be associated and fused. Then, the location of the key feature is attached with the corresponding matching strength information. Then, the core feature combination with clear location and high matching degree is selected, thereby generating multiple acidity feature location matching information of dairy products to be detected that can accurately reflect the location and matching strength of the key acidity feature.

[0047] Step S205: Generate multiple dairy product association information based on the multiple dairy product acidity feature localization and matching information, multiple dairy product information vectors, and the preset dairy product acidity feature quantization mapping vector.

[0048] In this embodiment, the preset quantization mapping vector of the acidity features of the dairy products to be tested can be manually set and can be used to establish the correspondence rules between feature associations and quantified values. First, the acidity feature location matching information of multiple dairy products to be tested can be collaboratively associated with multiple dairy product information vectors to be tested. Then, core features with accurate location, high matching degree, and carrying original multimodal information can be extracted. Next, the preset quantization mapping vector of the acidity features of the dairy products to be tested is combined to perform calculations, transforming the feature associations into standardized quantified data. Finally, the quantification results from various dimensions are integrated to generate multiple dairy product association information that comprehensively reflects the multimodal acidity feature associations and quantified attributes.

[0049] The multimodal dairy acidity detection method provided in this application enables accurate processing of multimodal data, enhances the ability to locate, match, and quantify dairy acidity-related features, effectively improves the relevance and reliability of information related to multiple dairy products to be tested, thereby significantly optimizing the accuracy and stability of dairy acidity detection. At the same time, it can adapt to different acidity level classification requirements, efficiently meet the application scenarios of rapid batch sample detection in dairy production quality inspection, and provide more targeted technical support for dairy quality control.

[0050] Figure 3 The flowchart illustrating the implementation of the multimodal dairy acidity detection method provided in Embodiment 3 of this application is shown. The difference between this method and Embodiment 2 is that step S205 specifically includes:

[0051] Step S301: Multiply the multiple dairy product information vectors to be detected and the preset acidity feature quantization mapping vector of the dairy product to be detected to generate multiple acidity feature quantization mapping information of the dairy product to be detected.

[0052] In this embodiment, the preset quantification mapping vector for the acidity features of the dairy products to be tested can be manually set and can be used to carry the mapping rules for converting multimodal raw features into acidity-related quantitative indicators. By multiplying multiple dairy product information vectors to be tested with this quantification mapping vector, the multi-dimensional raw features contained in the multiple dairy product information vectors, such as spectral, electrochemical, and physical parameters, can be converted into quantitative data directly related to acidity detection. This eliminates the interference caused by differences in the units of the raw features, and allows features of different modalities to reflect their correlation with acidity using a unified quantification standard. This generates multiple quantification mapping information for the acidity features of the dairy products to be tested that accurately reflects the acidity correlation attributes of multimodal features.

[0053] Step S302: Multiply the multiple acidity feature localization matching information of the multiple dairy products to be detected and the multiple acidity feature quantization mapping information of the multiple dairy products to be detected to generate multiple acidity feature mapping vectors of the dairy products to be detected.

[0054] In this embodiment, the location matching information of multiple acidity features of the dairy products to be detected clarifies the position and matching strength of the key acidity features, and the quantitative mapping information of multiple acidity features of the dairy products to be detected provides acidity correlation quantitative data of multimodal features. The two can be multiplied to achieve deep fusion of the location matching attribute and the quantitative attribute of the key features, so that the quantitative data focuses on the core feature region with accurate location and high matching degree, and filters out the quantitative data corresponding to features with ambiguous location or low matching degree, thereby generating multiple acidity feature mapping vectors of the dairy products to be detected that can condense the quantitative information of the core acidity features.

[0055] Step S303: The multiple acidity feature mapping vectors of the dairy products to be tested and the multiple information vectors of the dairy products to be tested are added together to generate multiple association information of the dairy products to be tested.

[0056] In this embodiment, the multiple acidity feature mapping vectors of the dairy products to be tested are the concentrated and quantified results of the core acidity features, while the multiple information vectors of the dairy products to be tested completely retain the original multimodal information. The two can be added together to calculate the result. This way, while preserving the integrity of the original multimodal information, the enhanced quantified information of the core acidity features is superimposed. Consequently, the generated information contains comprehensive original data support and highlights the key role of the core acidity features. This results in a final result that is both complete and targeted, thereby generating multiple dairy product association information that can integrate the original information and the core features.

[0057] The multimodal dairy acidity detection method provided in this application realizes hierarchical processing and precise fusion of original multimodal information and core acidity features. This ensures the accuracy of feature quantification and strengthens the prominent role of core features, effectively improving the information quality and accuracy of multiple dairy products to be detected. This makes subsequent classification processing more reliable and significantly optimizes the overall performance of dairy acidity detection. At the same time, it maintains adaptability to different acidity level classification requirements, which can efficiently meet the actual application scenarios of rapid and accurate detection of batch samples in dairy production quality inspection, and provide more comprehensive technical support for dairy quality control.

[0058] Figure 4 The flowchart illustrating the implementation of the multimodal dairy acidity detection method provided in Embodiment 4 of this application is shown. Its difference from Embodiment 1 described above lies in:

[0059] Multiple preset association vectors for dairy product information to be tested include preset anchor vectors for acidity features of dairy products to be tested, preset calibration vectors for acidity features of dairy products to be tested, and preset association weight vectors for acidity features of dairy products to be tested.

[0060] Step S102 specifically includes:

[0061] Step S401: Based on the multiple spectral information, multiple electrochemical information, and multiple physical parameter information of the dairy products to be tested, generate multiple spectral splicing vectors, multiple electrochemical splicing vectors, and multiple physical parameter splicing vectors of the dairy products to be tested.

[0062] In this embodiment, for multiple spectral information of dairy products to be tested, the absorbance or reflectance data sequences of each sample can be continuously spliced ​​according to the wavelength dimension to eliminate the feature fragmentation problem caused by data discreteness, thereby generating multiple spliced ​​vectors of spectral information of dairy products to be tested; for multiple electrochemical information of dairy products to be tested, the potential or current digital signal data of each sample can be sequentially spliced ​​according to the time dimension of signal acquisition, thereby integrating the temporal characteristics of electrochemical signals to generate multiple spliced ​​vectors of electrochemical information of dairy products to be tested; for multiple physical parameter information of dairy products to be tested, the density, viscosity, refractive index and temperature data of each sample can be spliced ​​according to the preset parameter category order, and then the data presentation format of the physical parameters can be unified to generate multiple spliced ​​vectors of physical parameters of dairy products to be tested.

[0063] Step S402: Based on the multiple spectral splicing vectors of the dairy products to be tested, the multiple electrochemical splicing vectors of the dairy products to be tested, the multiple physical parameter splicing vectors of the dairy products to be tested, and the preset acidity feature anchoring vectors of the dairy products to be tested, generate multiple spectral feature anchoring vectors, multiple electrochemical feature anchoring vectors of the dairy products to be tested, and multiple physical parameter feature anchoring vectors of the dairy products to be tested.

[0064] In this embodiment, the preset acidity feature anchoring vector of the dairy product to be tested can be manually set and can be used to anchor the core feature dimensions of spectral, electrochemical, and physical parameters that are strongly correlated with the acidity of the dairy product. Multiple spectral splicing vectors of the dairy product to be tested can be processed with this anchoring vector to filter out the feature components in the spectral splicing vectors that are highly correlated with acidity, generating multiple spectral feature anchoring vectors of the dairy product to be tested; multiple electrochemical splicing vectors of the dairy product to be tested can be processed with this anchoring vector to lock in the key features reflecting acidity changes in the electrochemical splicing vectors, generating multiple electrochemical feature anchoring vectors of the dairy product to be tested; multiple physical parameter splicing vectors of the dairy product to be tested can be processed with this anchoring vector, and then the core parameter features related to acidity in the physical parameter splicing vectors are extracted to generate multiple physical parameter feature anchoring vectors of the dairy product to be tested.

[0065] Step S403: Based on the multiple spectral splicing vectors of the dairy products to be tested, the multiple electrochemical splicing vectors of the dairy products to be tested, the multiple physical parameter splicing vectors of the dairy products to be tested, and the preset acidity feature calibration vectors of the dairy products to be tested, generate multiple spectral feature calibration vectors, multiple electrochemical feature calibration vectors, and multiple physical parameter feature calibration vectors of the dairy products to be tested.

[0066] In this embodiment, the preset acidity characteristic calibration vector of the dairy product to be tested can be manually set and can be used to calibrate the characteristic deviations caused by equipment errors and environmental interference in the three modal data. Multiple spectral splicing vectors of the dairy product to be tested can be calculated with this calibration vector to correct abnormal features such as baseline drift and noise interference in the spectral splicing vectors, generating multiple spectral characteristic calibration vectors of the dairy product to be tested. Multiple electrochemical splicing vectors of the dairy product to be tested can be calculated with this calibration vector to compensate for signal deviations caused by sensor accuracy issues in the electrochemical splicing vectors, generating multiple electrochemical characteristic calibration vectors of the dairy product to be tested. Multiple physical parameter splicing vectors of the dairy product to be tested can be calculated with this calibration vector, and then the parameter errors caused by changes in the measurement environment in the physical parameter splicing vectors can be adjusted to generate multiple physical parameter characteristic calibration vectors of the dairy product to be tested, thereby achieving accurate calibration and optimization of the characteristics of the three modal data.

[0067] Step S404: Generate multiple sets of related information for the dairy products to be tested based on the multiple anchor vectors of spectral features of the dairy products to be tested, the multiple anchor vectors of electrochemical features of the dairy products to be tested, the multiple anchor vectors of physical parameters of the dairy products to be tested, the multiple calibration vectors of spectral features of the dairy products to be tested, the multiple calibration vectors of electrochemical features of the dairy products to be tested, the multiple calibration vectors of physical parameters of the dairy products to be tested, and the preset correlation weight vector of acidity features of the dairy products to be tested.

[0068] In this embodiment, the preset acidity feature association weight vector of the dairy product to be tested can be manually set and can be used to allocate the contribution weights of anchoring features and calibration features of the three modalities in acidity detection. First, multiple anchoring vectors of spectral features of the dairy product to be tested can be fused with multiple calibration vectors of spectral features of the dairy product to be tested; multiple anchoring vectors of electrochemical features of the dairy product to be tested can be fused with multiple calibration vectors of electrochemical features of the dairy product to be tested; and multiple anchoring vectors of physical parameter features of the dairy product to be tested can be fused with multiple calibration vectors of physical parameter features of the dairy product to be tested, thereby obtaining fused feature vectors for each of the three modalities. Then, these fused feature vectors are weighted and calculated with the preset acidity feature association weight vector of the dairy product to be tested, thereby strengthening the features that play a major role in acidity detection and weakening the influence of secondary features according to the weight allocation. Finally, the weighted three types of fused feature vectors are integrated to generate multiple dairy product association information that comprehensively reflects the anchoring, calibration, and weight allocation features of multimodal data.

[0069] The multimodal dairy acidity detection method provided in this application performs dimensional anchoring, calibration, and fusion processing on three types of modal data: spectral, electrochemical, and physical parameters. This enables precise identification of the core features of acidity in each modality, corrects various biases in the data, and highlights the role of key features based on weight allocation. This effectively improves the accuracy and reliability of the correlation information of multiple dairy products to be tested, providing a better data foundation for subsequent classification processing. It significantly optimizes the overall performance of dairy acidity detection while maintaining adaptability to different acidity level classification requirements. This method can more efficiently meet the practical application scenarios of rapid and accurate detection of batch samples in dairy production quality inspection, providing more refined technical support for dairy quality control.

[0070] Figure 5 The flowchart illustrating the implementation of the multimodal dairy acidity detection method provided in Embodiment 5 of this application is shown. The difference between this method and Embodiment 4 above is that step S404 specifically includes:

[0071] Step S501: The multiple spectral feature anchoring vectors, multiple electrochemical feature anchoring vectors, multiple physical parameter feature anchoring vectors of the dairy products to be tested, and the preset acidity feature association weighting vector of the dairy products to be tested are weighted and summed to obtain the multiple acidity feature anchoring weighted information of the dairy products to be tested.

[0072] In this embodiment, the preset acidity feature association weight vector of the dairy product to be tested can be manually set and can include the weight allocation ratio of three types of anchor vectors: spectral, electrochemical, and physical parameters, in acidity detection. First, the weight values ​​of multiple spectral feature anchor quantities corresponding to the acidity feature association weight vector of the dairy product to be tested can be extracted and weighted with multiple spectral feature anchor vectors. Then, the weight values ​​of multiple electrochemical feature anchor quantities corresponding to the dairy product to be tested can be extracted and weighted with multiple electrochemical feature anchor vectors. Next, the weight values ​​of multiple physical parameter feature anchor quantities corresponding to the dairy product to be tested can be extracted and weighted with multiple physical parameter feature anchor vectors. Finally, the weighted results of the three types of anchor quantities are summed to obtain multiple acidity feature anchor weight information of the dairy product to be tested that reflects the contribution of different anchor features.

[0073] Step S502: The multiple spectral feature calibration vectors, electrochemical feature calibration vectors, physical parameter feature calibration vectors of the dairy products to be tested, and the preset acidity feature correlation weight vector of the dairy products to be tested are weighted and summed to obtain the acidity feature calibration weight information of the dairy products to be tested.

[0074] In this embodiment, the preset acidity feature association weight vector of the dairy product to be tested can be manually set, and corresponding acidity detection weight ratios can be assigned to the three types of calibration vectors: spectral, electrochemical, and physical parameters. First, the weight values ​​corresponding to multiple spectral feature calibration vectors of the dairy product to be tested can be extracted from the preset acidity feature association weight vector, and weighted together with multiple spectral feature calibration vectors of the dairy product to be tested. Then, the weight values ​​corresponding to multiple electrochemical feature calibration vectors of the dairy product to be tested can be extracted and weighted together with multiple electrochemical feature calibration vectors of the dairy product to be tested. Next, the weight values ​​corresponding to multiple physical parameter feature calibration vectors of the dairy product to be tested can be extracted and weighted together with multiple physical parameter feature calibration vectors of the dairy product to be tested. Finally, the weighted results of the three types of calibration vectors are summed to obtain multiple acidity feature calibration weighted information of the dairy product to be tested that reflects the contribution of different calibration features.

[0075] Step S503: The multiple acidity feature anchoring weighted information and multiple acidity feature calibration weighted information of the dairy products to be tested are added together to generate multiple dairy product association information.

[0076] In this embodiment, the anchoring weighted information of multiple acidity features of the dairy products to be tested focuses on the core anchoring features that are strongly correlated with acidity in the multimodal data, and reflects the contribution differences of different anchoring features. The calibration weighted information of multiple acidity features of the dairy products to be tested focuses on the corrected multimodal calibration features, and also reflects the contribution differences of different calibration features. The two can be added together to deeply fuse the weighted information of the core anchoring features and the weighted information of the corrected calibration features. This allows the fused information to retain the key value of the core acidity anchoring features. Then, the advantages of the two types of weighted information are integrated through the addition operation, thereby generating multiple dairy product association information that can comprehensively anchor and calibrate the weighted features.

[0077] The multimodal dairy acidity detection method provided in this application implements hierarchical weighted processing of anchoring features and calibration features, which makes the contribution of anchoring and calibration features of different modalities in acidity detection more accurate. This effectively improves the refinement and accuracy of the correlation information of multiple dairy products to be tested, provides data support that is more in line with the actual acidity detection needs for subsequent classification processing, significantly optimizes the overall accuracy and stability of dairy acidity detection, and maintains adaptability to the needs of different acidity level classification. It can more efficiently meet the actual application scenarios of rapid and accurate detection of batch samples in dairy production quality inspection, and provide detailed and reliable technical support for dairy quality control.

[0078] Figure 6 The flowchart illustrating the implementation of the multimodal dairy acidity detection method provided in Embodiment Six of this application is shown. The difference between this method and Embodiment One is that step S103 specifically includes:

[0079] Step S601: Based on the preset number of dairy acidity levels, randomly select the association information of the multiple dairy products to be tested to obtain multiple selected association information of dairy products to be tested.

[0080] In this embodiment, the preset number of dairy acidity levels can be set manually. The random sampling process uses this number as a benchmark to determine the number of samples to be drawn. That is, the number of multiple dairy product association information samples to be tested is consistent with the number of categories corresponding to the preset number of dairy acidity levels. By randomly selecting samples from multiple dairy product association information samples to be tested, the sample bias caused by manual selection is avoided, thereby ensuring that the sample samples can reflect the feature distribution of the overall association information in a balanced way. Then, through this random sampling operation, multiple dairy product association information samples to be tested that can be used as the initial classification benchmark are obtained.

[0081] Step S602: Based on the multiple dairy product association information to be tested and the multiple extracted dairy product association information to be tested, multiple remaining dairy product association information to be tested are obtained; the remaining dairy product association information to be tested corresponds one-to-one with the dairy product association information to be tested and the extracted dairy product association information to be tested.

[0082] In this embodiment, the multiple dairy product association information to be tested is a complete data set for classification processing. The multiple extracted dairy product association information to be tested is a partial sample selected from the complete set. By comparing and filtering the multiple dairy product association information to be tested with the multiple extracted dairy product association information to be tested, the extracted parts of the multiple dairy product association information to be tested are removed, and the remaining data that has not been extracted is obtained. Since each extracted dairy product association information to be tested comes from the multiple dairy product association information to be tested, the multiple remaining dairy product association information to be tested obtained after filtering forms a one-to-one correspondence with the original multiple dairy product association information to be tested and the extracted multiple extracted dairy product association information to be tested.

[0083] Step S603: Calculate the Euclidean distance between the multiple extracted dairy product association information and the multiple remaining dairy product association information to obtain the multiple dairy product association feature distance information.

[0084] In this embodiment, Euclidean distance can be used to quantify the similarity between multiple extracted dairy product association information and multiple remaining dairy product association information in the feature space. During the calculation, each remaining dairy product association information is used as a benchmark, and Euclidean distance is calculated with each of the multiple extracted dairy product association information to obtain the feature distance between the remaining dairy product association information and each extracted dairy product association information. Then, for each remaining dairy product association information, a set of corresponding distance data is generated. All these distance data are then organized to obtain multiple dairy product association feature distance information that can reflect the feature similarity of each dairy product association information.

[0085] Step S604: Extract the minimum value of the distance information of the multiple dairy product association features corresponding to the multiple extracted dairy product association information to be detected, and obtain the maximum and minimum value information of the distance of the multiple dairy product association features to be detected.

[0086] In this embodiment, for each set of multiple dairy product association feature distance information corresponding to the remaining dairy product association information to be detected, the numerical values ​​are compared and filtered to find the element with the smallest value in the set of distance information. This minimum value represents that the feature similarity between the remaining dairy product association information and one of the multiple extracted dairy product association information is the highest. Then, the minimum distance values ​​corresponding to each remaining dairy product association information are summarized to form a dataset that can identify the distance between each remaining association information and the most similar benchmark sample, thereby obtaining the maximum and minimum distance information of multiple dairy product association feature distances.

[0087] Step S605: Generate multiple acidity detection information of dairy products to be tested based on the remaining dairy product association information and the dairy product association information to be tested and the extracted dairy product association information corresponding to the maximum and minimum distance information of the multiple dairy product association features to be tested.

[0088] In this embodiment, the extracted dairy product association information corresponding to the maximum and minimum distance information of multiple dairy product association features can be used as a benchmark sample under this category and assigned a corresponding acidity level label. The remaining dairy product association information with the smallest distance to the benchmark sample is classified into the same acidity level because of its highest feature similarity. Then, the original feature data of multiple dairy product association information is combined to supplement the feature description of each classified association information. The acidity level label and the corresponding feature description are then integrated and associated with the original multiple dairy product association information, thereby generating multiple dairy product acidity detection information containing acidity level, feature description and the correspondence of the original association information.

[0089] The multimodal dairy acidity detection method provided in this application enables accurate classification of multiple dairy products related to each other, improves the objectivity and reliability of the classification results, significantly optimizes the classification accuracy and reliability of dairy acidity detection, and is adaptable to different acidity level classification requirements. It can efficiently meet the practical application scenarios of rapid and accurate detection of batch samples in dairy production quality inspection, and provides more objective technical support for dairy quality control.

[0090] Figure 7 The flowchart illustrating the implementation of the multimodal dairy acidity detection method provided in Embodiment 7 of this application is shown. The difference between this method and Embodiment 6 above is that step S605 specifically includes:

[0091] Step S701: Generate multiple dairy product association classification information based on the remaining dairy product association information, the dairy product association information to be detected, and the extracted dairy product association information corresponding to the maximum and minimum distance information of the multiple dairy product association features to be detected.

[0092] In this embodiment, the minimum and maximum distance information of multiple dairy product association features is used to identify the most similar extracted dairy product association information corresponding to each remaining dairy product association information. Multiple remaining dairy product association information with the smallest distance to the same extracted dairy product association information can be grouped into one category, and then together with the extracted dairy product association information, they form a classification set. Then, the original feature data of the corresponding multiple dairy product association information is added to each classification set to clarify the source and feature details of each data in the classification set. Finally, a temporary category identifier is assigned to each classification set, thereby generating multiple dairy product association classification information containing classification sets, original feature data, and temporary category identifiers.

[0093] Step S702: Calculate the median vector of the multiple dairy product association classification information to obtain the median information of the multiple dairy product association classification.

[0094] In this embodiment, each category set in the multiple dairy product association classification information to be detected contains multiple association information in the form of feature vectors. The median vector calculation requires calculating the median value of each feature vector in each category set according to the feature dimension, and then combining the median values ​​of each dimension in the corresponding dimension order to form the median vector of the category set. Then, this calculation process is performed on all the category sets corresponding to the multiple dairy product association classification information to be detected. Finally, the median vectors of each category set are sorted to obtain the median information of multiple dairy product association classifications to be detected that can reflect the core features of each category set.

[0095] Step S703: Determine whether the information in the multiple categories of dairy products to be tested is the same as the information in the multiple extracted dairy products to be tested; if yes, proceed to step S704; if no, proceed to step S705.

[0096] In this embodiment, the numerical information of each of the multiple dairy product association categories to be detected can be compared with the corresponding multiple extracted dairy product association information in each dimension to determine whether the values ​​of the two are completely consistent in all feature dimensions. If the numerical information of the multiple dairy product association categories to be detected and the multiple extracted dairy product association information to be detected are completely matched for all categories, the classification result is determined to be stable; if the values ​​of the two are not completely matched for any category, the classification benchmark is determined to need to be optimized.

[0097] Step S704: Generate multiple acidity detection information for the multiple dairy products to be tested based on the multiple classification information associated with the dairy products to be tested.

[0098] In this embodiment, a correspondence can be established between the specific acidity level labels corresponding to the preset number of dairy product acidity levels and the temporary category identifiers of each classification set. Then, a clear acidity level can be assigned to the associated classification information of multiple dairy products to be tested within each classification set. Subsequently, the core features of the associated information within each classification set are extracted. Combined with the original spectral information, electrochemical information, and physical parameter information of multiple dairy products to be tested, the feature sources and detailed descriptions of the tests are supplemented. Then, the acidity level, core feature descriptions, and original test information are integrated to generate complete and clearly labeled acidity test information for multiple dairy products to be tested.

[0099] Step S705: The information in the multiple categories of dairy products to be tested is used as multiple extracted information of dairy products to be tested, and the process is returned to step S602.

[0100] In this embodiment, when the numerical information of multiple dairy product association categories to be detected is inconsistent with the multiple extracted dairy product association information, it indicates that the initially extracted benchmark samples have failed to fully reflect the core features of the classification set. The numerical information of multiple dairy product association categories to be detected can replace the original multiple extracted dairy product association information as new classification benchmark samples, thereby ensuring that the new benchmark samples can accurately reflect the core features of each classification set. Then, the classification benchmark is continuously optimized through iterative updates, thereby improving the accuracy of the final classification result.

[0101] The multimodal dairy acidity detection method provided in this application introduces the median vector as a criterion for judging classification stability, realizing dynamic iterative optimization of the classification benchmark. This effectively solves the benchmark bias problem that may exist in the initial random sampling, making the classification results more consistent with the true characteristic distribution of each acidity level. It significantly improves the accuracy and stability of the classification of multiple dairy products to be tested, making the generated acidity detection information of multiple dairy products to be tested more reliable and valuable for reference. It ensures adaptability to the needs of different acidity level classifications, efficiently meets the actual application scenario of rapid and accurate detection of batch samples in dairy production quality inspection, and provides more rigorous and efficient technical support for dairy quality control.

[0102] Corresponding to the method in the above embodiments, Figure 8 The diagram shows a structural block diagram of the multimodal dairy acidity detection system provided in the embodiments of this application. For ease of explanation, only the parts related to the embodiments of this application are shown. Figure 8 The example multimodal dairy acidity detection system can be the execution subject of the multimodal dairy acidity detection method provided in the aforementioned embodiment 1.

[0103] Reference Figure 8 The multimodal dairy acidity detection system includes:

[0104] The dairy product information acquisition module 810 is used to acquire multiple spectral information, multiple electrochemical information, and multiple physical parameter information of the dairy products to be tested.

[0105] The dairy product association information generation module 820 is used to generate multiple dairy product association information based on the multiple dairy product spectral information, multiple dairy product electrochemical information, multiple dairy product physical parameter information and multiple preset dairy product information association vectors.

[0106] The dairy product acidity detection information generation module 830 is used to generate multiple dairy product acidity detection information by classifying and processing the multiple dairy products to be tested based on the preset dairy product acidity level quantity information and the associated information of the multiple dairy products to be tested.

[0107] The process by which each module in the multimodal dairy acidity detection system provided in this application implements its respective function can be found in the foregoing. Figure 1 The description of Embodiment 1 shown will not be repeated here.

[0108] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0109] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0110] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0111] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0112] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance. It should also be understood that although the terms "first," "second," etc., are used in the text to describe various elements in some embodiments of this application, these elements should not be limited by these terms. These terms are merely used to distinguish one element from another. For example, a first table may be named a second table, and similarly, a second table may be named a first table, without departing from the scope of the various described embodiments. Both the first table and the second table are tables, but they are not the same table.

[0113] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0114] The multimodal dairy acidity detection method provided in this application can be applied to terminal devices such as mobile phones, tablets, wearable devices, and vehicle-mounted devices. This application does not impose any restrictions on the specific type of terminal device.

[0115] For example, the terminal device may be a station in a WLAN, a cellular phone, a cordless phone, a handheld device with wireless communication capabilities, a computing device or other processing device connected to a wireless modem, an in-vehicle device, a vehicle networking terminal, a computer, a laptop computer, a handheld communication device, a handheld computing device, a satellite wireless device, a wireless modem card and / or other devices for communication over a wireless system, as well as next-generation communication systems, such as mobile terminals in 5G networks or mobile terminals in future evolved public terrestrial mobile networks.

[0116] Figure 9 This is a schematic diagram of the structure of a terminal device provided in an embodiment of this application. For example... Figure 9 As shown, the terminal device 9 of this embodiment includes: at least one processor 90 ( Figure 9(Only one is shown in the image), memory 91, which stores a computer program 92 that can run on the processor 90. When the processor 90 executes the computer program 92, it implements the steps in the various embodiments of the multimodal dairy acidity detection method described above, for example... Figure 1 Steps S101 to S103 are shown. Alternatively, when the processor 90 executes the computer program 92, it implements the functions of each module / unit in the above system embodiments, for example... Figure 8 The functions of modules 810 to 830 are shown.

[0117] The terminal device 9 can be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor 90 and a memory 91. Those skilled in the art will understand that... Figure 9 This is merely an example of terminal device 9 and does not constitute a limitation on terminal device 9. It may include more or fewer components than shown, or combine certain components, or different components. For example, the terminal device may also include input transmission devices, network access devices, buses, etc.

[0118] The processor 90 can be a central processing unit, or it can be other general-purpose processors, digital signal processors, or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0119] In some embodiments, the memory 91 may be an internal storage unit of the terminal device 9, such as a hard disk or memory of the terminal device 9. The memory 91 may also be an external storage device of the terminal device 9, such as a plug-in hard disk or smart memory card equipped on the terminal device 9. Furthermore, the memory 91 may include both internal and external storage units of the terminal device 9. The memory 91 is used to store operating systems, applications, data, and other programs, such as the program code of the computer program. The memory 91 can also be used to temporarily store data that has been sent or will be sent.

[0120] 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.

[0121] This application also provides a terminal device, which includes at least one memory, at least one processor, and a computer program stored in the at least one memory and executable on the at least one processor. When the processor executes the computer program, it causes the terminal device to implement the steps in any of the above method embodiments.

[0122] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps described in the various method embodiments above.

[0123] This application provides a computer program product that, when run on a terminal device, enables the terminal device to implement the steps described in the various method embodiments above.

[0124] If the integrated module / 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 medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0125] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0126] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0127] 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.

[0128] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A multimodal method for detecting the acidity of dairy products, characterized in that, include: Acquire spectral information, electrochemical information, and physical parameter information of multiple dairy products to be tested; Based on the spectral information, electrochemical information, physical parameter information, and preset association vectors of the dairy products to be tested, multiple association information of the dairy products to be tested is generated. Based on the preset number of dairy acidity levels, the multiple dairy products to be tested are classified and processed according to their associated information to generate multiple dairy acidity detection information.

2. The multimodal dairy acidity detection method as described in claim 1, characterized in that, Multiple preset information association vectors for dairy products to be detected include preset acidity feature localization vectors for dairy products to be detected, preset acidity feature matching vectors for dairy products to be detected, and preset acidity feature quantization mapping vectors for dairy products to be detected. The step of generating multiple sets of association information for dairy products to be tested based on the multiple spectral information, multiple electrochemical information, multiple physical parameter information, and multiple preset association vectors of information for dairy products to be tested specifically includes: The multiple spectral information, electrochemical information, and physical parameter information of the dairy products to be tested are spliced ​​and converted to generate multiple information vectors of the dairy products to be tested. The multiple dairy product information vectors to be detected and the preset acidity feature location vectors of the dairy products to be detected are multiplied together to generate multiple acidity feature location information of the dairy products to be detected. The multiple dairy product information vectors to be detected and the preset acidity feature matching vectors to be detected are multiplied together to generate multiple acidity feature matching information of dairy products to be detected. The multiple acidity feature location information and multiple acidity feature matching information of the dairy products to be tested are multiplied together to generate multiple acidity feature location matching information of the dairy products to be tested. Based on the location and matching information of multiple acidity features of the dairy products to be tested, multiple information vectors of the dairy products to be tested, and the preset quantization mapping vector of acidity features of the dairy products to be tested, multiple association information of the dairy products to be tested is generated.

3. The multimodal dairy acidity detection method as described in claim 2, characterized in that, The step of generating association information for multiple dairy products to be tested based on the multiple acidity feature localization and matching information of the multiple dairy products to be tested, the multiple information vectors of the dairy products to be tested, and the preset quantization mapping vector of the acidity features of the dairy products to be tested, specifically includes: The multiple dairy product information vectors to be detected and the preset acidity feature quantization mapping vectors to be detected are multiplied together to generate multiple acidity feature quantization mapping information of dairy products to be detected. The acidity feature localization and matching information of the multiple dairy products to be detected and the quantization mapping information of the multiple dairy products to be detected are multiplied together to generate multiple acidity feature mapping vectors of the dairy products to be detected. Multiple acidity feature mapping vectors and multiple information vectors of the dairy products to be tested are added together to generate multiple association information of the dairy products to be tested.

4. The multimodal dairy acidity detection method as described in claim 1, characterized in that, Multiple preset association vectors for dairy product information to be tested include preset anchor vectors for acidity features of dairy products to be tested, preset calibration vectors for acidity features of dairy products to be tested, and preset association weight vectors for acidity features of dairy products to be tested. The step of generating multiple sets of association information for dairy products to be tested based on the multiple spectral information, multiple electrochemical information, multiple physical parameter information, and multiple preset association vectors of information for dairy products to be tested specifically includes: Based on the spectral information, electrochemical information, and physical parameter information of the multiple dairy products to be tested, multiple spectral splicing vectors, multiple electrochemical splicing vectors, and multiple physical parameter splicing vectors of the dairy products to be tested are generated. Based on the multiple spectral splicing vectors of the dairy products to be tested, the multiple electrochemical splicing vectors of the dairy products to be tested, the multiple physical parameter splicing vectors of the dairy products to be tested, and the preset acidity feature anchoring vectors of the dairy products to be tested, multiple spectral feature anchoring vectors of the dairy products to be tested, multiple electrochemical feature anchoring vectors of the dairy products to be tested, and multiple physical parameter feature anchoring vectors of the dairy products to be tested are generated. Based on the multiple spectral splicing vectors of the dairy products to be tested, the multiple electrochemical splicing vectors of the dairy products to be tested, the multiple physical parameter splicing vectors of the dairy products to be tested, and the preset acidity feature calibration vectors of the dairy products to be tested, multiple spectral feature calibration vectors of the dairy products to be tested, multiple electrochemical feature calibration vectors of the dairy products to be tested, and multiple physical parameter feature calibration vectors of the dairy products to be tested are generated. Based on the multiple spectral feature anchoring vectors, multiple electrochemical feature anchoring vectors, multiple physical parameter feature anchoring vectors, multiple spectral feature calibration vectors, multiple electrochemical feature calibration vectors, multiple physical parameter feature calibration vectors, and a preset acidity feature association weight vector for the dairy products to be tested, multiple association information for the dairy products to be tested is generated.

5. The multimodal dairy acidity detection method as described in claim 4, characterized in that, The step of generating association information for multiple dairy products to be tested based on the multiple spectral feature anchoring vectors, multiple electrochemical feature anchoring vectors, multiple physical parameter feature anchoring vectors, multiple spectral feature calibration vectors, multiple electrochemical feature calibration vectors, multiple physical parameter feature calibration vectors, and a preset acidity feature association weight vector of the dairy products to be tested, specifically includes: The multiple spectral feature anchoring vectors, multiple electrochemical feature anchoring vectors, multiple physical parameter feature anchoring vectors of the dairy products to be tested, and the preset acidity feature association weighting vector of the dairy products to be tested are weighted and summed to obtain the multiple acidity feature anchoring weighting information of the dairy products to be tested. The multiple spectral feature calibration vectors, electrochemical feature calibration vectors, physical parameter feature calibration vectors, and preset acidity feature correlation weight vectors of the dairy products to be tested are weighted and summed to obtain the acidity feature calibration weight information of the dairy products to be tested. The multiple acidity feature anchoring weighted information and multiple acidity feature calibration weighted information of the dairy products to be tested are added together to generate multiple association information of the dairy products to be tested.

6. The multimodal dairy acidity detection method as described in claim 1, characterized in that, The step of classifying and processing multiple dairy acidity levels based on preset information and the associated information of multiple dairy products to be tested to generate multiple acidity detection information for dairy products to be tested specifically includes: Based on the preset number of dairy acidity levels, the association information of the multiple dairy products to be tested is randomly selected to obtain multiple association information of the selected dairy products to be tested. Based on the multiple association information of the dairy products to be tested and the multiple association information of the extracted dairy products to be tested, multiple association information of the remaining dairy products to be tested is obtained; the association information of the remaining dairy products to be tested corresponds one-to-one with the association information of the dairy products to be tested and the association information of the extracted dairy products to be tested. Calculate the Euclidean distance between the multiple extracted dairy product association information and the multiple remaining dairy product association information to obtain the multiple dairy product association feature distance information; Extract the minimum value of the distance information of multiple dairy product association features corresponding to the multiple extracted dairy product association information to obtain the maximum and minimum value information of the distance of multiple dairy product association features to be detected; Based on the maximum and minimum distance information of the multiple dairy products to be tested, the remaining dairy products to be tested association information, the dairy products to be tested association information, and the extracted dairy products to be tested association information, multiple dairy products to be tested acidity detection information are generated.

7. The multimodal dairy acidity detection method as described in claim 6, characterized in that, The step of generating multiple acidity detection information for dairy products to be tested based on the maximum and minimum distance information of the multiple dairy product association features to be tested, the remaining dairy product association information to be tested, the dairy product association information to be tested, and the extracted dairy product association information to be tested, specifically includes: Based on the maximum and minimum distance information of the multiple dairy products to be detected, the remaining dairy products to be detected association information, the dairy products to be detected association information, and the extracted dairy products to be detected association information, multiple dairy products to be detected association classification information are generated. Calculate the median vector of the multiple dairy product association classification information to obtain the median information of the multiple dairy product association classifications to be detected; Determine whether the numerical information of the multiple dairy products to be detected in the associated classification is the same as the associated information of the multiple extracted dairy products to be detected; If so, then based on the multiple classification information associated with the multiple dairy products to be tested, multiple acidity detection information of the dairy products to be tested will be generated; If not, the information in the multiple categories of dairy products to be tested is used as multiple extracted dairy product association information, and the process is returned to the step of obtaining multiple remaining dairy product association information based on the multiple dairy product association information and the multiple extracted dairy product association information; the remaining dairy product association information corresponds one-to-one with the dairy product association information to be tested and the extracted dairy product association information.

8. A multimodal dairy acidity detection system, characterized in that, include: The dairy product information acquisition module is used to acquire spectral information, electrochemical information, and physical parameter information of multiple dairy products to be tested. The module for generating association information of dairy products to be tested is used to generate multiple association information of dairy products to be tested based on the multiple spectral information of dairy products to be tested, multiple electrochemical information of dairy products to be tested, multiple physical parameter information of dairy products to be tested, and multiple preset association vectors of information of dairy products to be tested. The dairy product acidity detection information generation module is used to classify and process multiple dairy products based on preset dairy product acidity level quantity information and the associated information of the multiple dairy products to be tested, and generate multiple dairy product acidity detection information.

9. A terminal device, characterized in that, The terminal device includes a memory and a processor. The memory stores a computer program that can run on the processor. When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7.

Images