Food quality evaluation method and system based on big data

Abstract

The invention relates to the technical field of food quality, in particular to a food quality evaluation method and system based on big data, and the method comprises the following steps: based on food freshness detection data, synchronously collecting food real-time images and environmental sensor data, including temperature, humidity, illumination and gas components, and carrying out data format unification and time synchronization calibration. According to the invention, technical support is provided for real-time monitoring and evaluation of food freshness and environmental conditions by synchronously collecting food real-time images and environmental sensor data, food quality can be accurately quantified through refined data integration and feature analysis, and particularly, the analysis of color, form and surface texture is more accurate. The time difference learning technology is combined with the cloud computing technology, so that the prediction capability of the food quality change trend is effectively enhanced, the food quality management has preventive and accurate future quality prediction, risks in production and supply chains are reduced, and the food safety and consumer satisfaction are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of food quality, and in particular to a food quality assessment method and system based on big data. Background Art

[0002] Food quality technology focuses on ensuring and evaluating the safety, nutritional value and sensory quality of food through various methods. This field includes quality monitoring of the entire food supply chain from raw material collection, processing, packaging to final sales. Food quality assessment involves safety assessment of the use of additives and preservatives in food, as well as laboratory analysis to confirm whether the nutritional content of food meets the standards. With the increasing demand of consumers for food safety and quality, the field of food quality technology is gradually introducing more high-tech means to enhance food quality assurance capabilities.

[0003] Among them, the food quality assessment method of big data involves the use of big data technology to analyze and evaluate the quality of food. Big data is used to collect, process and analyze a large number of data points from food production and supply chains, such as temperature monitoring, humidity records, ingredient analysis results, etc. After processing, the data can predict the shelf life of food under specific conditions, identify potential food safety risks, and optimize production processes to ensure food quality, thereby improving food safety and consumer satisfaction and reducing risks in food production.

[0004] In the existing technology, food quality assessment relies on static laboratory analysis and intermittent data collection, resulting in insufficient real-time monitoring and dynamic assessment capabilities for the entire food quality cycle. This time-delayed data processing method limits the timely identification and response to food safety risks. For example, delayed response to changes in food storage conditions causes its quality to decrease before consumers purchase it. In addition, the existing technology lacks sufficient depth and predictive accuracy when processing and analyzing large-scale data, which limits the improvement of food safety. Inaccurate predictions of shelf life can lead to premature food scrapping or expiration, increasing cost losses. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a food quality assessment method and system based on big data.

[0006] In order to achieve the above object, the present invention adopts the following technical scheme: a food quality assessment method based on big data, comprising the following steps: S1: Based on the food freshness detection data, synchronously collect food real-time images and environmental sensor data, including temperature, humidity, light and gas composition, unify the data format and calibrate the time synchronization, and perform feature analysis on the collected images to identify key morphological features and obtain a standardized data set; S2: Based on the standardized data set, the color, shape and surface texture of the food are quantified, the freshness and quality grade of the food are evaluated according to the quantification results, and the quality of the food is automatically graded to generate a quality grading index; S3: Based on the quality grading index, through time difference learning, collect time series data of food quality, build a dynamic evaluation model, combine cloud computing technology to identify key time points and food quality change trends, and predict food quality changes in future time periods to obtain trend prediction data; S4: Based on the trend prediction data, the dynamic evaluation model is updated, new data is collected and processed, model parameter settings are refined according to user feedback and market dynamics, and real-time evaluation and prediction of food quality are implemented to obtain real-time food evaluation results.

[0007] The present invention is improved in that the steps of data format unification and time synchronization calibration are specifically as follows: S111: Based on the food freshness detection data, the real-time food image and environmental sensor data, including temperature, humidity, light and gas composition, are synchronously collected to obtain the original data set; S112: aligning the timestamps of the original data set using the formula: ; Adjust multiple data points to a unified time baseline to obtain a time-synchronized dataset, where Indicates the timestamp after time synchronization calibration. Represents the timestamp of the original data, represents the reference start time, represents the time deviation correction value; S113: Performing unified formatting processing on the time synchronization data set, and verifying the data type and format consistency, to obtain a data set with a unified format.

[0008] The present invention is improved in that the step of obtaining the standardized data set is specifically as follows: S121: Extract key visual indicators from food images, including color, shape and surface texture, and obtain a feature extraction result set; S122: Based on the feature extraction result set and in combination with the environmental sensor data, a multi-layer perceptron algorithm is used to process the data, the information of the different data sources is integrated, and key morphological features are identified to obtain a standardized data set.

[0009] The present invention is improved in that the steps of evaluating the freshness and quality grade of the food are specifically as follows: S211: quantifying the color, morphology and surface texture of the food based on the standardized data set, analyzing the color depth, edge clarity and texture complexity of the food image, and obtaining visual feature data; S212: Perform freshness analysis on the visual feature data using the formula: ; Get food freshness rating ,in, Indicates The weight of the visual features, Indicates The quantized value of a visual feature, Represents the total number of visual features; S213: Compare the food freshness score with the market quality standard, evaluate the food quality grade, including excellent, good and medium, and obtain the quality grade result.

[0010] The present invention is improved in that the steps of obtaining the quality grading index are specifically as follows: S221: extracting scores based on the freshness and quality grade of the food, and performing preliminary grouping according to the grading logic to obtain the food grade of the preliminary grouping; S222: Based on the food grade of the initial grouping, the formula is used: ; Automatically determine the quality grade of food , and obtain the quality grading index, where Represents the food freshness score, is the weight parameter, is a bias term used to calibrate the classification threshold.

[0011] The present invention is improved in that the steps of collecting the time series data of food quality are specifically as follows: S311: Based on the quality grading index, automatically identify and collect time series data of each rating stage through time difference learning, including the quality changes of food in multiple stages, to obtain a time series data set; S312: Based on the time series data set, perform real-time synchronization and upload, and regularly update to the cloud database to obtain a synchronously updated cloud database; S313: Based on the synchronously updated cloud database, the formula is adopted: ; Process the time series data in the cloud database to obtain a time-differenced data set, where: Indicates the quality level at the current time point, Indicates the quality level at the previous time point, Represents the difference in quality level between two time points, is the smoothing coefficient.

[0012] The present invention is improved in that the step of acquiring the trend prediction data is specifically as follows: S321: Using the cloud computing technology, marking key time points at which food quality changes occur, and obtaining marked time nodes; S322: Based on the marked time nodes, apply STL decomposition, analyze the data around the key time points, extract the key patterns of food quality changes, and use the formula: ; Calculate the factors affecting future quality changes and obtain the predicted quality level ,in, is the current quality level, is the adjusted value obtained based on seasonal decomposition, It is a random fluctuation based on the current data analysis; S323: Based on the predicted quality level, predict the food quality trend in a future time period to obtain trend prediction data.

[0013] The present invention is improved in that the steps of obtaining the real-time evaluation results of the food are specifically as follows: S411: Based on the trend prediction data, new food quality data and user feedback are collected to obtain a real-time updated data set; S412: Analyze the real-time updated data set to identify quality trends and potential problem points, and obtain an updated dynamic evaluation model; S413: Based on the updated dynamic evaluation model, the quality rating of the current food batch is calculated using the formula: ; Get real-time food evaluation results, including: represents the food quality results of real-time assessment, is a newly collected dataset, and is the linear relationship coefficient derived from historical data.

[0014] A food quality assessment system based on big data, the system comprising: The data calibration module collects real-time food images and environmental sensor data based on food freshness detection data, unifies the data format and calibrates the time synchronization, and performs feature analysis on the collected images to identify key morphological features and obtain a standardized data set; The quality analysis module quantifies the color, shape and surface texture of the food based on the standardized data set, evaluates the freshness and quality grade of the food, and automatically grades the quality of the food to generate a quality grading index; The trend prediction module collects time series data of food quality through time difference learning based on the quality grading index, identifies key time points and food quality change trends in combination with cloud computing technology, and predicts food quality changes in future time periods to obtain trend prediction data; The model optimization module updates the dynamic evaluation model based on the trend prediction data, collects new user feedback and market data, and refines the parameter settings of the model to generate adjusted model parameters; The real-time monitoring module performs real-time evaluation and prediction of food quality based on the adjusted model parameters, and dynamically generates real-time food evaluation results according to current quality standards.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by synchronously collecting real-time images of food and environmental sensor data, technical support is provided for real-time monitoring and evaluation of food freshness and environmental conditions. Through refined data integration and feature analysis, food quality can be accurately quantified, especially in the analysis of color, morphology and surface texture. The temporal difference learning technology is combined with cloud computing technology to effectively enhance the ability to predict the trend of food quality changes, making food quality management more preventive. Accurate future quality prediction reduces risks in production and supply chains, and improves food safety and consumer satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of a food quality assessment method based on big data is proposed for the present invention; Figure 2 A flow chart of data format unification and time synchronization calibration in the present invention; Figure 3 A flowchart for obtaining a standardized data set in the present invention; Figure 4 A flow chart for evaluating the freshness and quality level of food in the present invention; Figure 5 This is a flow chart for obtaining quality grading indicators in the present invention; Figure 6 It is a flow chart for collecting time series data of food quality in the present invention; Figure 7 A flowchart for obtaining trend prediction data in the present invention; Figure 8 The figure is a flow chart for obtaining the real-time evaluation results of food in the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0018] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating positions or positional relationships, are based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0019] Example: See Figure 1 The present invention provides a technical solution: a food quality assessment method based on big data, comprising the following steps: S1: Based on the food freshness detection data, synchronously collect food real-time images and environmental sensor data, including temperature, humidity, light and gas composition, unify the data format and calibrate the time synchronization, and perform feature analysis on the collected images to identify key morphological features and obtain a standardized data set; Temperature: Appropriate low temperature can significantly prolong the freshness of food and prevent the growth of microorganisms. Humidity will affect the dryness of food and microbial activity, which will directly affect the quality and safety of food. For example, a high humidity environment may cause some foods to mold. Light: Light can have a negative impact on certain foods (such as meat and dairy products), leading to fat oxidation and pigment changes. Gas composition: The ratio of gas components will affect the color and taste of food. For example, by reducing oxygen and increasing carbon dioxide and nitrogen, the shelf life of meat products can be significantly extended while maintaining their color and taste. Cooked food and pastries can control the ratio of oxygen and carbon dioxide to slow down microbial growth and maintain the flavor and texture of food. The adjustment of gas composition can be carried out according to modified atmosphere packaging (MAP) technology.

[0020] S2: Based on a standardized data set, the color, morphology and surface texture of food are quantified, the freshness and quality grade of food are evaluated based on the quantification results, and the food quality is automatically graded to generate quality grading indicators; S3: Based on the quality grading index, through time difference learning, collect the time series data of food quality, build a dynamic evaluation model, combine cloud computing technology to identify key time points and food quality change trends, and predict the food quality changes in the future time period to obtain trend prediction data; S4: Update the dynamic evaluation model based on trend prediction data, collect and process new data, refine model parameter settings according to user feedback and market dynamics, and implement real-time evaluation and prediction of food quality to obtain real-time food evaluation results.

[0021] The standardized data set includes image feature data, synchronized time data, and environmental monitoring data. The quality grading indicators include freshness level, color score, and morphological integrity score. The trend prediction data includes key time markers, quality change information, and future quality prediction results. The real-time food evaluation results include model optimization records and feedback response analysis results.

[0022] See also Figure 2 ,The specific steps for data format unification and time synchronization calibration are: S111: Based on the food freshness detection data, the real-time food image and environmental sensor data, including temperature, humidity, light and gas composition, are synchronously collected to obtain the original data set; Based on the demand for food freshness detection data, the temperature, humidity, light and gas composition data are monitored and recorded in real time. The data is accurately time-stamped to ensure the time consistency of subsequent analysis. During the collection process, the data of each sensor must be corrected through the built-in time synchronization module to ensure that the timestamps of all data are completely consistent. This ensures that the real-time status of the food can be accurately reflected when multiple environmental variables change. In addition, all collected data are stored in a unified data format to facilitate subsequent data processing and analysis, and obtain the original data set containing multiple environmental parameters.

[0023] S112: Align the timestamps of the original data set using the formula: ; Adjust multiple data points to a unified time baseline to obtain a time-synchronized dataset, where Indicates the timestamp after time synchronization calibration. Represents the timestamp of the original data, Represents the reference start time, used as the base point for all data timestamp corrections. represents the time deviation correction value; The collected data are as follows: (UNIX timestamp, representing the time of a specific sampling point), (The earliest sampling point timestamp of the day), the time deviation obtained through data collection and monitoring Seconds, substitute into the formula to calculate the timestamp after time synchronization calibration: ; This result shows the difference in seconds between the calibrated timestamp of a sampling point and the earliest sampling point of the day, providing a clear time baseline for the alignment of all data points.

[0024] S113: Performing unified formatting processing on the time synchronization data set, and verifying the data type and format consistency, to obtain a data set with unified format; By uniformly formatting the time-synchronized data sets, including encoding, conversion and verification of different types of data, we ensure that all data meet the preset quality standards. The data validation stage checks the data set to eliminate any errors or format inconsistencies. This stage is an important part of data processing, ensuring that subsequent analysis can be based on accurate and unified data. The result of the formatting process is a clean and consistent data set. The data set will directly support subsequent data analysis tasks, such as pattern recognition and trend prediction, to ensure accurate data analysis.

[0025] See also Figure 3 , the specific steps for obtaining the standardized data set are: S121: Extract key visual indicators from food images, including color, shape and surface texture, and obtain a feature extraction result set; Extracting key visual indicators of food images relies on image capture devices to ensure that the image quality is sufficient for subsequent analysis. Then, the color, morphology and surface texture in the image are quantitatively analyzed to determine the brightness and fading of the food. Morphological analysis uses edge detection technology to determine the shape and size of the food. Surface texture analysis uses texture recognition technology to evaluate the smoothness or roughness of the food surface. Together, they constitute the feature extraction process, and the resulting feature extraction result set contains quantitative data for various visual indicators.

[0026] S122: Based on the feature extraction result set, combined with the environmental sensor data, a multi-layer perceptron algorithm is used to process the data, the information of the different data sources is integrated, and key morphological features are identified to obtain a standardized data set; Based on the feature extraction result set and combined with environmental sensor data, it involves integrating visual data and environmental data, and using a multi-layer perceptron algorithm to process and fuse information. The multi-layer perceptron is a feedforward artificial neural network that inputs, processes and outputs data through multiple levels. The nodes between layers interact with each other through weighted connections. The adjustment of weights is automatically completed through the learning process so that the model can better adapt to the data characteristics. During the fusion process, the color, morphology and surface texture features in the image data are combined with environmental data such as temperature and humidity. Through algorithm optimization, key food freshness indicators are identified and enhanced to obtain a standardized data set, providing a quantitative standard for comprehensively evaluating food freshness.

[0027] See also Figure 4 , the specific steps for evaluating the freshness and quality grade of food are: S211: Quantify the color, morphology and surface texture of food based on standardized data sets, analyze the color depth, edge clarity and texture complexity of food images, and obtain visual feature data; Based on standardized data sets, image enhancement technology is applied to improve image clarity, including adjusting brightness and contrast to better identify colors and edges, using Sobel or Canny algorithms to determine the shape and outline of food, and identifying edges by calculating the gradient between image pixels. Color analysis uses color space conversion, such as from RGB to HSV, to more effectively analyze color saturation and hue, making color analysis more in line with the human eye's sensitivity to color. Finally, texture analysis calculates the grayscale co-occurrence matrix of a specific area in the image or performs a Fourier transform to identify and quantify the roughness or complexity of the surface texture, completing the conversion from the original image to visual feature data.

[0028] S212: Perform freshness analysis on visual feature data using the formula: ; Get food freshness rating ,in, Indicates The weight of the visual features, Indicates The quantized value of a visual feature, Represents the total number of visual features; Consider three visual features, namely color depth , edge clarity , texture complexity , the weights determined by industry standards are , , , and the visual feature value is , , , substitute into the formula to calculate: ; The results show that according to the weighted scoring model, the freshness score of the food is 79, and the score reflects the comparison with the freshness of the food according to the market quality standards, which can help further determine the quality grade of the food.

[0029] S213: Compare the food freshness score with the market quality standard to evaluate the food quality level, including excellent, good and fair, and obtain the quality level result; Standardizing food freshness scores and converting raw scores into standard scores is achieved by mapping the scores to a fixed range (such as 0 to 100) to facilitate comparison with quality standards from different sources. According to the industry's quality grade standards, the standards are divided into three main categories: excellent, good, and medium. Each category corresponds to a specific score range. Finally, the scores are matched with the categories, and the specific grades are determined according to their numerical range. For example, scores above 90 are excellent, 80 to 90 are good, and 70 to 80 are medium, ensuring the objectivity and standardization of the evaluation.

[0030] See also Figure 5 , the specific steps for obtaining the quality grading index are: S221: extracting scores based on the freshness and quality grade of the food, performing preliminary grouping according to the grading logic, and obtaining the food grade of the preliminary grouping; Specific scores are extracted from the obtained freshness and quality grade data. The scores reflect the basic quality characteristics of the food. The scores are preliminarily classified through grading logic, and the foods are grouped according to their freshness scores. For example, foods with high scores are classified into the high-quality group, and foods with medium scores are classified into the medium-quality group. This ensures a systematic preliminary rating of foods based on objective data. The grouping results will provide a basis for further detailed quality analysis.

[0031] S222: Food grade based on initial grouping, using the formula: ; Automatically determine the quality grade of food , and obtain the quality grading index, where Represents the food freshness score, is a weight parameter used to adjust the influence of freshness score in quality rating. is a bias term used to calibrate the classification threshold. For example, if After weight adjustment, scores above 90 are classified as excellent, scores between 80 and 90 are good, and scores between 70 and 80 are medium. The following data was collected, there is a food freshness score is 85, weight is 1.1, bias If it is -10, the food quality level The calculation is as follows: ; ; The result shows that the quality grade score of this food is 83.5, which is classified as good according to the set classification criteria, indicating that the food performs well in freshness and overall quality.

[0032] See also Figure 6 ,The specific steps for collecting time series data of food quality are: S311: Based on the quality grading index, through time difference learning, the time series data of each rating stage is automatically identified and collected, including the quality changes of food in multiple stages, to obtain a time series data set; Based on the existing quality grading indicators, the time difference learning method is used to automatically identify and collect food quality changes at key stages. The process involves comprehensive monitoring of the production process and collects food quality data at various stages including before, during and after production. The data is continuously recorded by sensors at each rating stage to ensure the accuracy and timeliness of the data. The collected data covers the entire process from raw material inspection to finished product packaging, forming a time series data set.

[0033] S312: Based on the time series data set, real-time synchronization and uploading are performed, and the cloud database is updated regularly to obtain a synchronously updated cloud database; Real-time synchronization and upload are implemented on the collected time series data sets to ensure that the data is uploaded to the cloud database in a timely manner, including data formatting, error detection and correction, and verification process to ensure data integrity. During the upload process, key time points are automatically identified and marked, such as temperature changes, humidity adjustments or other key production environment changes. The marked time points are processed through the cloud computing platform to optimize data storage and access efficiency, forming a synchronously updated cloud database, which provides real-time and accurate data support for food quality monitoring and future predictions.

[0034] S313: Based on the synchronously updated cloud database, the formula is: ; Process the time series data in the cloud database to obtain a time-differenced data set, where: Indicates the quality level at the current time point, Indicates the quality level at the previous time point, Represents the difference in quality level between two time points, is the smoothing coefficient; The following data were collected: initial food quality level is 80, the current time point changes is 2, the smoothing coefficient is 0.8, and the calculation process is as follows: ; ; The results show that the food quality level at the current time point is 64.4, indicating that the food quality has declined compared to the previous time point, which helps to adjust the production strategy in time to control quality.

[0035] See also Figure 7 ,The specific steps for obtaining trend forecast data are as follows: S321: Using cloud computing technology, mark the key time points when food quality changes, and obtain the marked time nodes; Through cloud computing technology, time series data is extracted from the cloud database to perform change point detection and analysis operations. First, all time points in the data are checked one by one, and the food quality change rate at each time point is calculated. Then, a threshold for the change rate is set, and time points greater than the threshold are marked as key time points. Then, the context data range of the marked time points is further analyzed to ensure that the marking is accurate and serves as the basis for subsequent trend predictions.

[0036] S322: Based on the marked time nodes, apply STL decomposition to analyze the data around the key time points and extract the key patterns of food quality changes using the formula: ; Calculate the factors affecting future quality changes and obtain the predicted quality level ,in, is the current quality level, is the adjusted value obtained based on seasonal decomposition, It is a random fluctuation based on the current data analysis; The following data was collected, the current quality level 75, seasonally adjusted 3, random fluctuation = -1, and the predicted future quality level is calculated according to the formula : ; The results show that after taking into account seasonal changes and recent fluctuations, the predicted future quality grade is 77, which is an increase from the current level of 75, indicating a potential direction for improvement in production quality.

[0037] S323: Based on the predicted quality level, predict the food quality trend in a future time period to obtain trend prediction data; Based on the predicted quality grade, combined with historical trends and real-time analysis results, the predicted quality grades are arranged in chronological order, and the trend line is fitted. Then, the fluctuation range of food quality in the future period is calculated. The predicted value is corrected based on the cyclical change model of historical data to generate trend data containing changes in food quality in the future period, providing data basis for subsequent production decisions and quality optimization.

[0038] See also Figure 8 , the specific steps for obtaining the real-time food evaluation results are as follows: S411: Based on trend prediction data, collect new food quality data and user feedback, including data points extracted from real-time production lines and market dynamics, to obtain a real-time updated data set; Through cloud computing technology, data from real-time production lines is regularly collected and synchronized, down to food quality data points extracted from specific product batches, as well as user evaluation data from market feedback. This process is executed through data acquisition, processing large amounts of real-time data streams, and formatting the data into a unified standard to accommodate subsequent analysis and processing. Key data points are obtained from a variety of data sources. The data points reflect the quality change trends of food and consumer satisfaction, and a real-time updated data set containing the latest market and production information is obtained.

[0039] S412: Analyze the real-time updated data set to identify quality trends and potential problem points, and obtain an updated dynamic evaluation model; Based on the collected real-time updated data sets, data analysis technology is applied to conduct in-depth trend analysis. Through statistical methods and machine learning models, patterns in the data are identified. The patterns mine the fluctuation trends and quality problem points of food quality. Through analysis, adjustment and optimization of the parameters of the evaluation model, in order to more accurately predict and respond to future quality problems, an updated and improved dynamic evaluation model is obtained that can respond to market and production changes in real time.

[0040] S413: Calculate the quality rating of the current food batch based on the updated dynamic assessment model, using the formula: ; Get real-time food evaluation results, including: represents the food quality results of real-time assessment, is a newly collected dataset, and is the linear relationship coefficient derived from historical data; If you have the following data, the newly collected data set The value is 85, and the coefficient obtained by historical data analysis is 1.2, offset is -5. Then the calculation process is: ; ; The results indicate that the current food batch has a real-time rating of 97, meaning that based on the latest data collected and processed, the quality rating of this food batch is high, which bodes well for market feedback and consumer satisfaction.

[0041] A food quality assessment system based on big data, the system comprising: The data calibration module collects real-time food images and environmental sensor data based on food freshness detection data, unifies the data format and calibrates the time synchronization, and performs feature analysis on the collected images to identify key morphological features and obtain a standardized data set; The quality analysis module quantifies the color, shape and surface texture of food based on standardized data sets, evaluates the freshness and quality grade of food, and automatically grades the quality of food to generate quality grading indicators; The trend prediction module collects time series data of food quality through time difference learning based on quality grading indicators, combines cloud computing technology to identify key time points and food quality change trends, and predicts food quality changes in the future time period to obtain trend prediction data; The model optimization module updates the dynamic evaluation model based on trend forecast data, collects new user feedback and market data, and refines the model parameter settings to generate adjusted model parameters; The real-time monitoring module performs real-time evaluation and prediction of food quality based on the adjusted model parameters, and dynamically generates real-time food evaluation results according to current quality standards.

[0042] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A food quality assessment method based on big data, characterized in that: The following steps are involved: Based on the food freshness detection data, real-time food images and environmental sensor data, including temperature, humidity, light and gas composition, are collected synchronously to unify the data format and calibrate the time synchronization. The collected images are analyzed for features, key morphological features are identified, and standardized data sets are obtained. Based on the standardized data set, the color, shape and surface texture of the food are quantified, the freshness and quality grade of the food are evaluated according to the quantification results, and the quality of the food is automatically graded to generate a quality grading index; Based on the quality grading index, through time difference learning, collect time series data of food quality, build a dynamic evaluation model, combine cloud computing technology to identify key time points and food quality change trends, and predict food quality changes in future time periods to obtain trend prediction data; Based on the trend prediction data, the dynamic evaluation model is updated, new data is collected and processed, model parameter settings are refined according to user feedback and market dynamics, and real-time evaluation and prediction of food quality are implemented to obtain real-time food evaluation results.

2. The food quality assessment method based on big data according to claim 1, characterized in that: The steps of data format unification and time synchronization calibration are specifically as follows: Based on the food freshness detection data, real-time food images and environmental sensor data, including temperature, humidity, light and gas composition, are collected synchronously to obtain the original data set; The timestamps of the original data set are aligned using the formula: ； Adjust multiple data points to a unified time baseline to obtain a time-synchronized dataset, where Indicates the timestamp after time synchronization calibration. Represents the timestamp of the original data, represents the reference start time, represents the time deviation correction value; The time synchronization data set is uniformly formatted, and the data type and format consistency are verified to obtain a data set with a uniform format.

3. The food quality assessment method based on big data according to claim 1, characterized in that: The steps for obtaining the standardized data set are specifically as follows: Extract key visual indicators from food images, including color, shape and surface texture, and obtain a feature extraction result set; Based on the feature extraction result set and combined with environmental sensor data, a multi-layer perceptron algorithm is used to process the data, integrate information from different data sources, and identify key morphological features to obtain a standardized data set.

4. The food quality assessment method based on big data according to claim 1, characterized in that: The steps for evaluating the freshness and quality grade of the food are specifically as follows: Based on the standardized data set, quantify the color, morphology and surface texture of the food, analyze the color depth, edge clarity and texture complexity of the food image, and obtain visual feature data; The freshness analysis of the visual feature data is performed using the formula: ； Get food freshness rating ,in, Indicates The weight of the visual features, Indicates The quantized value of a visual feature, Represents the total number of visual features; The food freshness score is compared with the market quality standard to evaluate the food quality grade, including excellent, good and medium, to obtain the quality grade result.

5. The food quality assessment method based on big data according to claim 1, characterized in that: The steps for obtaining the quality grading index are specifically as follows: Based on the freshness and quality grade of the food, extract the score, perform preliminary grouping according to the grading logic, and obtain the food grade of the preliminary grouping; Based on the food grade of the initial grouping, the formula is used: ； Automatically determine the quality grade of food , and obtain the quality grading index, where Represents the food freshness score, is the weight parameter, is a bias term used to calibrate the classification threshold.

6. The food quality assessment method based on big data according to claim 1, characterized in that: The steps for collecting the time series data of food quality are specifically as follows: Based on the quality grading index, the time series data of each rating stage is automatically identified and collected through time difference learning, including the quality changes of food in multiple stages, to obtain a time series data set; Based on the time series data set, real-time synchronization and uploading are performed, and the cloud database is updated regularly to obtain a synchronously updated cloud database; Based on the synchronously updated cloud database, the formula is adopted: ； Process the time series data in the cloud database to obtain a time-differenced data set, where: Indicates the quality level at the current time point, Indicates the quality level at the previous time point, Represents the difference in quality level between two time points, is the smoothing coefficient.

7. The food quality assessment method based on big data according to claim 1, characterized in that: The steps for obtaining the trend prediction data are specifically as follows: Using the cloud computing technology, key time points at which food quality changes occur are marked to obtain marked time nodes; Based on the marked time nodes, STL decomposition is applied to analyze the data around the key time points and extract the key patterns of food quality changes using the formula: ； Calculate the factors affecting future quality changes and obtain the predicted quality level ,in, is the current quality level, is the adjusted value obtained based on seasonal decomposition, It is a random fluctuation based on the current data analysis; Based on the predicted quality level, the food quality trend in a future time period is predicted to obtain trend prediction data.

8. The food quality assessment method based on big data according to claim 1, characterized in that: The steps for obtaining the real-time food evaluation results are specifically as follows: Based on the trend prediction data, new food quality data and user feedback are collected to obtain a real-time updated data set; Analyze the real-time updated data set to identify quality trends and potential problem points, and obtain an updated dynamic evaluation model; Based on the updated dynamic evaluation model, the quality rating of the current food batch is calculated using the formula: ； Get real-time food evaluation results, including: represents the food quality results of real-time assessment, is a newly collected dataset, and is the linear relationship coefficient derived from historical data.

9. A food quality assessment system based on big data, characterized in that: According to any one of claims 1 to 8, the method for evaluating food quality based on big data is implemented, and the system comprises: The data calibration module collects real-time food images and environmental sensor data based on food freshness detection data, unifies the data format and calibrates the time synchronization, and performs feature analysis on the collected images to identify key morphological features and obtain a standardized data set; The quality analysis module quantifies the color, shape and surface texture of the food based on the standardized data set, evaluates the freshness and quality grade of the food, and automatically grades the quality of the food to generate a quality grading index; The trend prediction module collects time series data of food quality through time difference learning based on the quality grading index, identifies key time points and food quality change trends in combination with cloud computing technology, and predicts food quality changes in future time periods to obtain trend prediction data; The model optimization module updates the dynamic evaluation model based on the trend prediction data, collects new user feedback and market data, and refines the parameter settings of the model to generate adjusted model parameters; The real-time monitoring module performs real-time evaluation and prediction of food quality based on the adjusted model parameters, and dynamically generates real-time food evaluation results according to current quality standards.

Images