Food information processing system and method based on AI model

Abstract

The invention discloses a food information processing system and method based on an AI model, relates to the field of food, and solves the problem that a food information processing system is poor in food information processing effect. The product packaging module is used for obtaining a food packaging qualification ratio corresponding to each food production line, obtaining product packaging qualification data, and obtaining a food raw material sampling inspection result coefficient corresponding to each food production line, so as to obtain food raw material sampling inspection data; and the information processing module is used for respectively acquiring a food information monitoring coefficient corresponding to each food production line, analyzing the food information monitoring coefficient to obtain production line food marking data, and performing traceable marking on the food according to the production line food marking data. The food information processing system can improve objectivity and transparency of a food information processing process.

Description

Technical Field

[0001] The present invention belongs to the field of food and relates to AI model technology, specifically a food information processing system and method based on AI model. Background Art

[0002] The existing food information processing system has the following specific defects when processing food information:

[0003] 1. Existing food information is often processed directly based on the data that has been recorded, which makes the food data lack objectivity and authenticity, making it difficult to provide accurate commodity information and making international food trade lack transparency;

[0004] 2. The existing food information processing methods in the field of international trade often only focus on the qualification testing of the final food products, thus ignoring the testing of the production lines and raw materials, and failing to classify and mark the food produced by different production lines, which increases the uncertainty of food export trade.

[0005] To this end, we propose a food information processing system and method based on AI model. Summary of the invention

[0006] In view of the deficiencies in the prior art, the present invention aims to provide a food information processing system and method based on an AI model. The present invention is based on respectively obtaining multiple food production lines, performing raw material sampling inspections on the multiple food production lines, and obtaining the food raw material sampling inspection result coefficient corresponding to each food production line, obtaining food raw material sampling inspection data, performing finished product packaging monitoring on each food production line in the food production area, and respectively obtaining the food packaging qualified ratio corresponding to each food production line, obtaining product packaging qualified data, respectively obtaining the food information monitoring coefficient corresponding to each food production line according to the product packaging qualified data and the food raw material sampling inspection data, obtaining multiple food information monitoring coefficients, obtaining the food information monitoring coefficient threshold, and performing numerical comparison with the multiple food information monitoring coefficients, obtaining the assembly line food marking data according to the numerical comparison, and performing traceability marking on the food through the assembly line food marking data;

[0007] In order to achieve the above purpose, the present invention adopts the following technical solution: a food information processing system based on AI model, the specific working process of each module is as follows:

[0008] Raw material sampling module: used to obtain multiple food production lines respectively, conduct raw material sampling inspections on multiple food production lines, and obtain the food raw material sampling inspection result coefficient corresponding to each food production line to obtain food raw material sampling inspection data;

[0009] Finished product packaging module: used to monitor the finished product packaging of each food production line in the food production area, and obtain the food packaging qualified ratio corresponding to each food production line, and obtain the product packaging qualified data;

[0010] Information processing module: used to obtain the food information monitoring coefficient corresponding to each food production line according to the product packaging qualification data and the food raw material sampling data, obtain multiple food information monitoring coefficients, obtain the food information monitoring coefficient threshold, and compare it with the multiple food information monitoring coefficients numerically, obtain the assembly line food marking data according to the numerical comparison, and use the assembly line food marking data to mark the food for traceability.

[0011] A food information processing method based on an AI model comprises the following specific steps:

[0012] Step S1: obtaining multiple food production lines respectively, performing raw material sampling inspections on the multiple food production lines, and obtaining the food raw material sampling inspection result coefficient corresponding to each food production line to obtain food raw material sampling inspection data;

[0013] Step S2: Perform finished product packaging monitoring on each food production line in the food production area, and obtain the food packaging qualified ratio corresponding to each food production line to obtain product packaging qualified data;

[0014] Step S3: According to the qualified product packaging data and the food raw material sampling data, the food information monitoring coefficient corresponding to each food production line is obtained respectively, and multiple food information monitoring coefficients are obtained. The food information monitoring coefficient threshold is obtained, and it is numerically compared with the multiple food information monitoring coefficients. The assembly line food marking data is obtained according to the numerical comparison, and the food is traceably marked through the assembly line food marking data.

[0015] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0016] 1. The present invention can effectively improve the transparency and objectivity of food information processing by acquiring food raw material sampling data and product packaging qualified data to detect the food production line;

[0017] 2. The present invention marks foods produced by different production process lines, which can improve the certainty of food export trade and reduce the trade risks of import and export products. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0019] Figure 1 is the overall system block diagram of the present invention;

[0020] Figure 2 It is a diagram of the implementation steps of the present invention;

[0021] Figure 3 It is a schematic diagram of the product traceability process in the present invention. DETAILED DESCRIPTION

[0022] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] Embodiment 1

[0024] See also Figure 1 , the present invention provides a technical solution: a food information processing system based on an AI model, comprising a raw material sampling module, a production process module, a finished product packaging module, an information processing module and a server, wherein the raw material sampling module, the production process module, the finished product packaging module and the information processing module are respectively connected to the server, and the server controls the raw material sampling module, the production process module, the finished product packaging module and the information processing module respectively;

[0025] The raw material sampling module obtains multiple food production lines respectively, performs raw material sampling on the multiple food production lines, and obtains the food raw material sampling result coefficient corresponding to each food production line to obtain food raw material sampling data;

[0026] Acquire working data of food production lines, mark food production lines in production according to the working data of food production lines, obtain multiple food production lines, and randomly select one of the marked multiple food production lines as a sample food production line;

[0027] Conduct random inspections on raw materials of sample food production lines to obtain the food raw material random inspection result coefficients corresponding to the sample food production lines;

[0028] The details are as follows:

[0029] In the process of sampling the sample food production line, several different types of raw materials for sampling are set, and the set raw materials for sampling are named as the first type of raw materials for sampling to the ath type of raw materials for sampling;

[0030] It should be noted here that:

[0031] In this application, a is the number of types corresponding to the sampled raw materials, and a is an integer greater than 0;

[0032] In this application, assuming that the food produced is instant noodles, the first type of raw materials for random inspection may be flour, the second type of raw materials for random inspection may be edible oil, the third type of raw materials for random inspection may be salt, and the fourth type of raw materials for random inspection may be dried vegetables;

[0033] From the production raw materials currently used in the sample food production line, the first type of raw materials to the ath type of raw materials are respectively selected by unit mass to obtain the first raw material samples to the ath raw material samples;

[0034] Performing a sample quality test on the first raw material sample and obtaining a quality deviation of the first raw material;

[0035] The details are as follows:

[0036] Obtain an ingredient list corresponding to the first raw material sample, obtain material ingredients corresponding to the first raw material sample according to the ingredient list, obtain multiple different material ingredients, and name the obtained multiple material ingredients as the first material ingredient to the bth material ingredient respectively;

[0037] It should be noted here that:

[0038] In the present application, b is the material ingredient type and quantity value corresponding to the first raw material sample, and b is an integer greater than 0;

[0039] In the present application, assuming that the first raw material sample is specifically flour, the corresponding first material ingredient may be glutenin, the second material ingredient may be gluten, the third material ingredient may be starch, and the fourth material ingredient may be impurities;

[0040] The impurities involved here are specifically substances that are not marked in the ingredient list;

[0041] Obtaining a mass value of the first raw material sample to obtain a total mass value of the sample;

[0042] Respectively obtain the mass values ​​corresponding to the first material ingredient to the bth material ingredient in the first raw material sample, and obtain the mass values ​​of the first ingredient to the bth material ingredient;

[0043] Calculate the ratios of the mass values ​​of the first ingredient to the mass values ​​of the bth ingredient to the total mass of the sample to obtain the mass proportion of the first ingredient to the mass proportion of the bth ingredient;

[0044] According to the ingredient list, respectively obtain the reference mass proportions of the first material ingredient to the bth material ingredient in the first raw material sample to obtain the first reference mass proportion to the bth reference mass proportion;

[0045] The mass deviation of the first raw material is obtained by calculating the mass ratio of the first ingredient to the mass ratio of the bth ingredient and the mass ratio of the first reference mass to the bth reference mass;

[0046] The quality deviation of the first raw material is calculated using the following formula:

[0047]

[0048] Among them, Czp1 is the mass deviation of the first raw material, Pz1 to Pzb are the mass proportion of the first ingredient to the mass proportion of the bth ingredient respectively, Pj1 to Pjb are the mass proportion of the first reference mass to the bth reference mass, and b is the quantity value of the material ingredient type corresponding to the first raw material sample;

[0049] Repeat the process of obtaining the quality deviation of the first raw material, and obtain the quality deviations of the raw materials corresponding to the second type of sampled raw materials to the ath type of sampled raw materials, respectively, to obtain the quality deviation of the second raw material to the ath raw material;

[0050] For the finished food products produced by the sample food assembly line, the raw material proportions corresponding to the first type of sampled raw materials to the ath type of sampled raw materials are obtained respectively, and the first raw material proportion to the ath raw material proportion are obtained;

[0051] It should be noted here that:

[0052] In this application, the raw material ratios involved here are the raw material mass ratios in a single food product produced;

[0053] The food raw material sampling result coefficient corresponding to the sample food production line is obtained by calculating the quality deviation of the first raw material to the quality deviation of the ath raw material and the proportion of the first raw material to the ath raw material;

[0054] The food raw material sampling result coefficient corresponding to the sample food production line is calculated. The specific formula is as follows:

[0055] Jcj=Czp1×Ycl1+Czp2×Ycl2++Czpa×Ycla;

[0056] Among them, Jcj is the food raw material sampling result coefficient corresponding to the sample food production line, Czp1 to Czpa are the quality deviation of the first raw material to the quality deviation of the ath raw material, and Yc l 1 to Yc la are the proportion of the first raw material to the proportion of the ath raw material;

[0057] Repeat the process of obtaining the food raw material sampling result coefficient corresponding to the sample food production line, respectively obtain the food raw material sampling result coefficient corresponding to each food production line, and obtain multiple food raw material sampling result coefficients;

[0058] The coefficients of multiple food raw material sampling results are defined as food raw material sampling data;

[0059] The raw material sampling module acquires the food raw material sampling data and transmits it to the information processing module;

[0060] The finished product packaging module monitors the finished product packaging of each food production line in the food production area, and obtains the food packaging qualified ratio corresponding to each food production line to obtain the product packaging qualified data;

[0061] Acquire multiple food production lines in the food production area, and randomly select one of the marked multiple food production lines as a characteristic food production line;

[0062] Establish a food packaging inspection model;

[0063] The details are as follows:

[0064] Acquire historical finished product packaging data of food, and acquire multiple food outer packaging images according to the historical finished product packaging data of food;

[0065] manually labeling a plurality of food outer packaging images into first-type food packaging images and second-type food packaging images to obtain food packaging image labeling data;

[0066] It should be noted here that:

[0067] In the present application, the first type of food packaging image referred to herein is a qualified food packaging image, and the second type of food packaging image referred to herein is an unqualified food packaging image;

[0068] Dividing the food packaging image annotation data into an image test set and an image training set according to a preset training-test ratio;

[0069] It should be noted here that:

[0070] In the present application, the preset training-test ratio involved here is specifically 7:3, that is, the ratio of the number of image test sets to the number of image training sets is 7:3;

[0071] Create an image recognition model through the existing AI model platform, and train the image recognition model using the image training set until each food packaging image in the image training set is trained at least once;

[0072] The image recognition model is tested using the image test set, and the accuracy of the test results is obtained to obtain the model recognition accuracy, the target model recognition accuracy is obtained, and the model recognition accuracy is numerically compared with the target model recognition accuracy. When the model recognition accuracy is greater than or equal to the target model recognition accuracy, the image recognition model training is completed to obtain the food packaging detection model. When the model recognition accuracy is less than the target model recognition accuracy, the image training set is continued to be used to train the image recognition model until the model recognition accuracy is greater than or equal to the target model recognition accuracy.

[0073] Use the food packaging inspection model to conduct packaging qualification inspection on the characteristic food production line, and obtain the food packaging qualification ratio corresponding to the characteristic food production line;

[0074] The details are as follows:

[0075] In the process of performing packaging qualified line inspection on the characteristic food production line, the time value corresponding to the current moment is marked as the first packaging characteristic time point, the time value corresponding to the characteristic monitoring time before the first packaging characteristic time point is marked as the second packaging characteristic time point, and the period between the first packaging characteristic time point and the second packaging characteristic time point is marked as the food packaging inspection cycle;

[0076] During the food packaging detection cycle, the food packaging detection model is used to perform image detection on the outer packaging of each finished food product, and the food outer packaging image is marked as a first type of food packaging image and a second type of food packaging image according to the detection result;

[0077] When one of the multiple images corresponding to the outer packaging of the finished food product is a second type of food packaging image, the corresponding finished food product is classified as an unqualified packaged finished product; when the multiple images corresponding to the outer packaging of the finished food product are all first type of food packaging images, the corresponding finished food product is classified as a qualified packaged finished product;

[0078] Count the number of qualified packaged finished products to obtain the number value of qualified packaged finished products;

[0079] Obtain the cumulative number of finished food products detected by the food packaging detection model during the food packaging detection cycle to obtain the cumulative number of finished food products detected;

[0080] Calculate the ratio of the number of qualified packaged finished products to the cumulative number of tested finished products to obtain the qualified ratio of food packaging corresponding to the characteristic food production line;

[0081] Repeat the process of obtaining the qualified ratio of food packaging corresponding to the characteristic food production line, perform packaging qualification test on each food production line respectively, and obtain multiple qualified ratios of food packaging;

[0082] A plurality of food packaging qualified ratios are defined as product packaging qualified data;

[0083] The finished product packaging module acquires the qualified data of product packaging and transmits it to the information processing module;

[0084] The information processing module obtains the food information monitoring coefficient corresponding to each food production line according to the qualified product packaging data and the food raw material sampling data, obtains multiple food information monitoring coefficients, obtains the food information monitoring coefficient threshold, and compares it with the multiple food information monitoring coefficients, obtains the line food marking data according to the numerical comparison, and uses the line food marking data to mark the food for traceability;

[0085] Obtain food raw material sampling data, and obtain the food raw material sampling result coefficient corresponding to each food production line based on the food raw material sampling data;

[0086] Obtain product packaging qualified data, and obtain the food packaging qualified ratio corresponding to each food production line according to the product packaging qualified data;

[0087] The food information monitoring coefficient is obtained by calculating the food raw material sampling result coefficient and the food packaging qualified ratio corresponding to the same food production line;

[0088] The food information monitoring coefficient is calculated as follows:

[0089]

[0090] Among them, Sjx is the food information monitoring coefficient, Jcj is the food raw material sampling result coefficient, and Hgb is the food packaging qualified ratio;

[0091] Obtain food information monitoring coefficient threshold;

[0092] The details are as follows:

[0093] Obtain the food raw material sampling result coefficient threshold and food packaging qualified ratio threshold respectively;

[0094] It should be noted here that:

[0095] In this application, the food raw material sampling result coefficient threshold involved here is the maximum food raw material sampling result coefficient corresponding to the qualified food production line tested, and the food packaging qualified ratio threshold involved here is the minimum food packaging qualified ratio corresponding to the qualified food production line tested;

[0096] The food information monitoring coefficient threshold is obtained by calculating the food raw material sampling result coefficient threshold and the food packaging qualified ratio threshold;

[0097] The food information monitoring coefficient threshold is calculated using the following formula:

[0098]

[0099] Among them, Sjxy is the food information monitoring coefficient threshold, Jcjy is the food raw material sampling result coefficient threshold, and Hgby is the food packaging qualified ratio threshold;

[0100] Compare the food packaging qualified ratio threshold with the food information monitoring coefficient, mark the food produced by the assembly line according to the numerical comparison, and obtain assembly line food marking data;

[0101] The details are as follows:

[0102] When the food information monitoring coefficient is greater than or equal to the food information monitoring coefficient threshold, the corresponding food production line is judged to be qualified for food production inspection, and the food produced by it is marked as a first type of product;

[0103] When the food information monitoring coefficient is less than the food information monitoring coefficient threshold, the corresponding food production line is judged to be unqualified in food production inspection, and the food produced by it is marked as a second type of product;

[0104] It should be noted here that:

[0105] In this application, the first type of products referred to herein are qualified products, and the second type of products are unqualified products;

[0106] The food labeling data of the assembly line is marked on the outer packaging of the food through a QR code for traceability, and the product type traceability query can be carried out through the existing product sales channels;

[0107] It should be noted here that:

[0108] The product sales channels involved here include but are not limited to official websites, shopping platforms, and import and export trading platforms;

[0109] In this application, if corresponding calculation formulas appear, the above calculation formulas are all dimensionless and take their numerical calculations. The weight coefficients, proportional coefficients and other coefficients in the formulas are set to a result value obtained by quantifying each parameter. The size of the weight coefficient and the proportional coefficient can be determined as long as it does not affect the proportional relationship between the parameter and the result value.

[0110] Embodiment 2

[0111] See also Figure 2 Based on another concept of the same invention, a food information processing method based on an AI model is now proposed, comprising the following steps:

[0112] Step S1: obtaining multiple food production lines respectively, performing raw material sampling inspections on the multiple food production lines, and obtaining the food raw material sampling inspection result coefficient corresponding to each food production line to obtain food raw material sampling inspection data;

[0113] Step S11: obtaining the working data of the food production line, marking the food production lines in production state according to the working data of the food production line, obtaining multiple food production lines, and randomly selecting one of the marked multiple food production lines as a sample food production line;

[0114] Step S12: performing random inspection on the raw materials of the sample food production line to obtain the food raw material random inspection result coefficient corresponding to the sample food production line;

[0115] The details are as follows:

[0116] Step S121: in the process of sampling the sample food production line, a plurality of different types of raw materials for sampling are set respectively, and the set raw materials for sampling are named as the first type of raw materials for sampling to the ath type of raw materials for sampling;

[0117] Step S122: extracting the first type of sampled raw materials to the ath type of sampled raw materials per unit mass from the production raw materials currently used in the sample food production line, and obtaining the first raw material samples to the ath raw material samples;

[0118] Step S123: performing sample quality detection on the first raw material sample and obtaining the first raw material quality deviation;

[0119] The details are as follows:

[0120] Step S1231: obtaining an ingredient list corresponding to the first raw material sample, obtaining material ingredients corresponding to the first raw material sample according to the ingredient list, obtaining a plurality of different material ingredients, and naming the obtained plurality of material ingredients as the first material ingredient to the bth material ingredient respectively;

[0121] Step S1232: obtaining a mass value of the first raw material sample to obtain a total mass value of the sample;

[0122] Step S1233: respectively obtaining the mass values ​​corresponding to the first material ingredient to the b-th material ingredient in the first raw material sample, and obtaining the mass values ​​of the first ingredient to the b-th material ingredient;

[0123] Step S1234: Calculate the ratios of the mass values ​​of the first ingredient to the mass values ​​of the bth ingredient to the total mass value of the sample to obtain the mass proportion of the first ingredient to the mass proportion of the bth ingredient;

[0124] Step S1235: Obtain the reference mass proportions of the first material ingredient to the bth material ingredient in the first raw material sample according to the ingredient list, and obtain the first reference mass proportion to the bth reference mass proportion;

[0125] Step S1236: Calculate the mass ratio of the first ingredient to the mass ratio of the bth ingredient and the mass ratio of the first reference mass to the bth reference mass to obtain the mass deviation of the first raw material;

[0126] The quality deviation of the first raw material is calculated using the following formula:

[0127]

[0128] Among them, Czp1 is the mass deviation of the first raw material, Pz1 to Pzb are the mass proportion of the first ingredient to the mass proportion of the bth ingredient respectively, Pj1 to Pjb are the mass proportion of the first reference mass to the bth reference mass, and b is the quantity value of the material ingredient type corresponding to the first raw material sample;

[0129] Step S124: respectively obtaining the raw material quality deviations corresponding to the second type of sampled raw materials to the ath type of sampled raw materials, and obtaining the second raw material quality deviation to the ath raw material quality deviation;

[0130] Step S125: For the finished food products produced by the sample food assembly line, respectively obtain the raw material proportions corresponding to the first type of sampled raw materials to the ath type of sampled raw materials, and obtain the first raw material proportion to the ath raw material proportion;

[0131] Step S126: Calculate the food raw material sampling result coefficient corresponding to the sample food production line by dividing the first raw material quality deviation to the ath raw material quality deviation and the first raw material proportion to the ath raw material proportion;

[0132] The food raw material sampling result coefficient corresponding to the sample food production line is calculated. The specific formula is as follows:

[0133] Jcj=Czp1×Ycl1+Czp2×Ycl2++Czpa×Ycla;

[0134] Among them, Jcj is the food raw material sampling result coefficient corresponding to the sample food production line, Czp1 to Czpa are the quality deviation of the first raw material to the quality deviation of the ath raw material, and Yc l 1 to Yc la are the proportion of the first raw material to the proportion of the ath raw material;

[0135] Step S13: acquiring the food raw material sampling result coefficients corresponding to each food production line respectively, to obtain a plurality of food raw material sampling result coefficients;

[0136] Step S14: defining multiple food raw material sampling result coefficients as food raw material sampling data;

[0137] Step S2: Perform finished product packaging monitoring on each food production line in the food production area, and obtain the food packaging qualified ratio corresponding to each food production line to obtain product packaging qualified data;

[0138] Step S21: acquiring multiple food production lines in the food production area, and randomly selecting one of the marked multiple food production lines as a characteristic food production line;

[0139] Step S22: establishing a food packaging detection model;

[0140] The details are as follows:

[0141] Step S221: Acquire historical finished product packaging data of food, and acquire multiple food outer packaging images according to the historical finished product packaging data of food;

[0142] Step S222: annotating the plurality of food outer packaging images into first-type food packaging images and second-type food packaging images by manual annotation to obtain food packaging image annotation data;

[0143] Step S223: dividing the food packaging image annotation data into an image test set and an image training set according to a preset training test ratio;

[0144] Step S224: creating an image recognition model through an existing AI model platform, and training the image recognition model using the image training set until each food packaging image in the image training set is trained at least once;

[0145] Step S225: Use the image test set to test the image recognition model, and obtain the accuracy of the test results to obtain the model recognition accuracy, obtain the target model recognition accuracy, and compare the model recognition accuracy with the target model recognition accuracy. When the model recognition accuracy is greater than or equal to the target model recognition accuracy, the image recognition model training is completed to obtain the food packaging detection model. When the model recognition accuracy is less than the target model recognition accuracy, continue to use the image training set to train the image recognition model until the model recognition accuracy is greater than or equal to the target model recognition accuracy.

[0146] Step S23: using the food packaging detection model to perform packaging qualification detection on the characteristic food production line to obtain the food packaging qualification ratio corresponding to the characteristic food production line;

[0147] The details are as follows:

[0148] Step S231: in the process of performing packaging qualified line inspection on the characteristic food production line, the time value corresponding to the current moment is marked as the first packaging characteristic time point, the time value corresponding to the characteristic monitoring time before the first packaging characteristic time point is marked as the second packaging characteristic time point, and the period between the first packaging characteristic time point and the second packaging characteristic time point is marked as the food packaging inspection cycle;

[0149] Step S232: During the food packaging detection cycle, use the food packaging detection model to perform image detection on the outer packaging of each finished food product, and mark the food outer packaging image as a first type of food packaging image and a second type of food packaging image according to the detection result;

[0150] Step S233: when one of the multiple images corresponding to the outer packaging of the finished food product is a second type of food packaging image, the corresponding finished food product is classified as an unqualified packaged finished product; when the multiple images corresponding to the outer packaging of the finished food product are all first type of food packaging images, the corresponding finished food product is classified as a qualified packaged finished product;

[0151] Step S234: Count the number of qualified packaged finished products to obtain the number value of qualified packaged finished products;

[0152] Step S235: obtaining the cumulative number of finished foods detected by the food packaging detection model during the food packaging detection cycle, and obtaining the cumulative number of finished foods detected;

[0153] Step S236: Calculate the ratio of the number of qualified packaged finished products to the cumulative number of tested finished products to obtain the qualified ratio of food packaging corresponding to the characteristic food production line;

[0154] Step S24: performing packaging qualification test on each food production line respectively to obtain multiple food packaging qualification ratios;

[0155] Step S25: defining a plurality of food packaging qualified ratios as product packaging qualified data;

[0156] Step S3: obtaining the food information monitoring coefficient corresponding to each food production line according to the qualified product packaging data and the food raw material sampling data, obtaining multiple food information monitoring coefficients, obtaining the food information monitoring coefficient threshold, and numerically comparing it with the multiple food information monitoring coefficients, obtaining the line food marking data according to the numerical comparison, and marking the food traceably through the line food marking data;

[0157] Step S31: obtaining food raw material sampling data, and obtaining the food raw material sampling result coefficient corresponding to each food production line according to the food raw material sampling data;

[0158] Step S32: obtaining product packaging qualified data, and obtaining the food packaging qualified ratio corresponding to each food production line according to the product packaging qualified data;

[0159] Step S33: Calculating the food information monitoring coefficient by taking the food raw material sampling result coefficient and the food packaging qualified ratio corresponding to the same food production line;

[0160] The food information monitoring coefficient is calculated as follows:

[0161]

[0162] Among them, Sjx is the food information monitoring coefficient, Jcj is the food raw material sampling result coefficient, and Hgb is the food packaging qualified ratio;

[0163] Step S34: obtaining a food information monitoring coefficient threshold;

[0164] The details are as follows:

[0165] Step S341: respectively obtaining a food raw material sampling result coefficient threshold and a food packaging qualified ratio threshold;

[0166] Step S342: Calculating the food information monitoring coefficient threshold by using the food raw material sampling result coefficient threshold and the food packaging qualified ratio threshold;

[0167] Step S35: numerically comparing the food packaging qualified ratio threshold with the food information monitoring coefficient, marking the food produced by the assembly line according to the numerical comparison, and obtaining assembly line food marking data;

[0168] The details are as follows:

[0169] Step S351: when the food information monitoring coefficient is greater than or equal to the food information monitoring coefficient threshold, the corresponding food production line is judged to be qualified for food production inspection, and the food produced by the food production line is marked as a first type of product;

[0170] Step S352: when the food information monitoring coefficient is less than the food information monitoring coefficient threshold, the corresponding food production line is judged to be unqualified in food production inspection, and the food produced by the food production line is marked as a second type of product;

[0171] Step S36: Please refer to Figure 3 , the assembly line food labeling data can be marked on the food packaging through a QR code, and product traceability can be queried through existing product sales channels.

[0172] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A food information processing system based on AI model, characterized in that: include: Raw material sampling module: used to obtain multiple food production lines respectively, conduct raw material sampling on multiple food production lines, and obtain the food raw material sampling result coefficient corresponding to each food production line to obtain food raw material sampling data; Finished product packaging module: used to monitor the finished product packaging of each food production line in the food production area, and obtain the food packaging qualified ratio corresponding to each food production line, and obtain the product packaging qualified data; Information processing module: used to obtain the food information monitoring coefficient corresponding to each food production line according to the product packaging qualification data and the food raw material sampling data, obtain multiple food information monitoring coefficients, obtain the food information monitoring coefficient threshold, and compare it with the multiple food information monitoring coefficients numerically, obtain the assembly line food marking data according to the numerical comparison, and use the assembly line food marking data to mark the food for traceability.

2. A food information processing system based on an AI model according to claim 1, characterized in that: The raw material sampling module acquires the food raw material sampling data, as follows: Acquire working data of food production lines, mark food production lines in production according to the working data of food production lines, obtain multiple food production lines, and randomly select one of the marked multiple food production lines as a sample food production line; Conduct random inspections on raw materials of sample food production lines to obtain the food raw material random inspection result coefficients corresponding to the sample food production lines; Obtaining the food raw material sampling result coefficients corresponding to each food production line respectively, and obtaining multiple food raw material sampling result coefficients; The coefficients of multiple food raw material sampling results are defined as food raw material sampling data; The raw material sampling module obtains the food raw material sampling result coefficients corresponding to the sample food production line, as follows: In the process of sampling the sample food production line, several different types of raw materials for sampling are set, and the set raw materials for sampling are named as the first type of raw materials for sampling to the ath type of raw materials for sampling; From the production raw materials currently used in the sample food production line, the first type of raw materials to the ath type of raw materials are respectively selected by unit mass to obtain the first raw material samples to the ath raw material samples; A sample quality test is performed on the first raw material sample, and a quality deviation of the first raw material is obtained.

3. A food information processing system based on an AI model according to claim 2, characterized in that: The raw material sampling module obtains the food raw material sampling result coefficient corresponding to the sample food assembly line, specifically as follows: respectively obtain the raw material quality deviations corresponding to the second type of sampled raw materials to the ath type of sampled raw materials, and obtain the second raw material quality deviation to the ath raw material quality deviation; For the finished food products produced by the sample food assembly line, the raw material proportions corresponding to the first type of sampled raw materials to the ath type of sampled raw materials are obtained respectively, and the first raw material proportion to the ath raw material proportion are obtained; The food raw material sampling result coefficient corresponding to the sample food production line is obtained by calculating the quality deviation of the first raw material to the quality deviation of the ath raw material and the proportion of the first raw material to the ath raw material; Calculate the coefficient of food raw material sampling results corresponding to the sample food production line.

4. A food information processing system based on an AI model according to claim 3, characterized in that: The raw material sampling module obtains the quality deviation of the first raw material, as follows: Obtain an ingredient list corresponding to the first raw material sample, obtain material ingredients corresponding to the first raw material sample according to the ingredient list, obtain multiple different material ingredients, and name the obtained multiple material ingredients as the first material ingredient to the bth material ingredient respectively; Obtaining a mass value of the first raw material sample to obtain a total mass value of the sample; The mass values ​​corresponding to the first material ingredient to the bth material ingredient in the first raw material sample are respectively obtained to obtain the mass values ​​of the first ingredient to the bth material ingredient.

5. A food information processing system based on an AI model according to claim 4, characterized in that: The raw material sampling module obtains the quality deviation of the first raw material, as follows: Calculate the ratios of the mass values ​​of the first ingredient to the mass values ​​of the bth ingredient to the total mass of the sample to obtain the mass proportion of the first ingredient to the mass proportion of the bth ingredient; According to the ingredient list, respectively obtain the reference mass proportions of the first material ingredient to the bth material ingredient in the first raw material sample to obtain the first reference mass proportion to the bth reference mass proportion; The mass deviation of the first raw material is obtained by calculating the mass ratio of the first ingredient to the mass ratio of the bth ingredient and the mass ratio of the first reference mass to the bth reference mass; The first raw material quality deviation is calculated.

6. A food information processing system based on an AI model according to claim 1, characterized in that: The finished product packaging module obtains the qualified data of product packaging, as follows: Acquire multiple food production lines in the food production area, and randomly select one of the marked multiple food production lines as a characteristic food production line; Establish a food packaging inspection model; Use the food packaging inspection model to conduct packaging qualification inspection on the characteristic food production line, and obtain the food packaging qualification ratio corresponding to the characteristic food production line; Conduct packaging qualification tests on each food production line to obtain multiple food packaging qualification ratios; A plurality of food packaging qualified ratios are defined as product packaging qualified data.

7. A food information processing system based on an AI model according to claim 6, characterized in that: The finished product packaging module creates a food packaging detection model, as follows: Acquire historical finished product packaging data of food, and acquire multiple food outer packaging images according to the historical finished product packaging data of food; manually labeling a plurality of food outer packaging images into first-type food packaging images and second-type food packaging images to obtain food packaging image labeling data; Dividing the food packaging image annotation data into an image test set and an image training set according to a preset training-test ratio; Create an image recognition model through the existing AI model platform, and train the image recognition model using the image training set until every food packaging image in the image training set has been trained at least once.

8. A food information processing system based on an AI model according to claim 7, characterized in that: The finished product packaging module creates a food packaging detection model, specifically as follows: use an image test set to test the image recognition model, and obtain the accuracy of the test results to obtain the model recognition accuracy, obtain the target model recognition accuracy, and numerically compare the model recognition accuracy with the target model recognition accuracy. When the model recognition accuracy is greater than or equal to the target model recognition accuracy, the image recognition model training is completed to obtain the food packaging detection model. When the model recognition accuracy is less than the target model recognition accuracy, continue to use the image training set to train the image recognition model until the model recognition accuracy is greater than or equal to the target model recognition accuracy.

9. A food information processing system based on an AI model according to claim 8, characterized in that: The finished product packaging module obtains the qualified ratio of food packaging corresponding to the characteristic food production line, as follows: In the process of performing packaging qualified line inspection on the characteristic food production line, the time value corresponding to the current moment is marked as the first packaging characteristic time point, the time value corresponding to the characteristic monitoring time before the first packaging characteristic time point is marked as the second packaging characteristic time point, and the period between the first packaging characteristic time point and the second packaging characteristic time point is marked as the food packaging inspection cycle; During the food packaging detection cycle, the food packaging detection model is used to perform image detection on the outer packaging of each finished food product, and the food outer packaging image is marked as a first type of food packaging image and a second type of food packaging image according to the detection result; When one of the multiple images corresponding to the outer packaging of the finished food product is a second type of food packaging image, the corresponding finished food product is classified as an unqualified packaged finished product; when the multiple images corresponding to the outer packaging of the finished food product are all first type of food packaging images, the corresponding finished food product is classified as a qualified packaged finished product; Count the number of qualified packaged finished products to obtain the number value of qualified packaged finished products; Obtain the cumulative number of finished food products detected by the food packaging detection model during the food packaging detection cycle to obtain the cumulative number of finished food products detected; Calculate the ratio of the number of qualified packaged finished products to the cumulative number of tested finished products to obtain the qualified ratio of food packaging corresponding to the characteristic food production line.

10. A food information processing method based on an AI model, applicable to a food information processing system based on an AI model as claimed in any one of claims 1 to 9, characterized in that: The food information processing method comprises the following specific steps: Step S1: obtaining multiple food production lines respectively, performing raw material sampling inspections on the multiple food production lines, and obtaining the food raw material sampling inspection result coefficient corresponding to each food production line to obtain food raw material sampling inspection data; Step S2: Perform finished product packaging monitoring on each food production line in the food production area, and obtain the food packaging qualified ratio corresponding to each food production line to obtain product packaging qualified data; Step S3: According to the qualified product packaging data and the food raw material sampling data, the food information monitoring coefficient corresponding to each food production line is obtained respectively, and multiple food information monitoring coefficients are obtained. The food information monitoring coefficient threshold is obtained, and it is numerically compared with the multiple food information monitoring coefficients. The assembly line food marking data is obtained according to the numerical comparison, and the food is traceably marked through the assembly line food marking data.

Images