Intelligent management method, system and device for dietary nutrient intake

Abstract

The application discloses a kind of intelligent management methods, systems and devices of dietary nutrition intake, belong to dietary management technical field, comprising: obtaining user nutrition demand index, determine the nutritional composition of various food materials;According to nutrition demand index and the nutritional composition of each food material, the intake of each food material is planned and dietary recipe is generated.Determine the nutritional composition of each food material includes: obtaining and determining food material category according to image, determine first overall nutritional composition data and mark area according to food material category, obtain and determine the nutritional composition of the second overall nutritional composition data and mark area of food material according to hyperspectral image, calculate third overall nutritional composition data according to the above nutritional composition, weight is assigned to second and third overall nutritional composition data according to the difference between first and second overall nutritional composition data, and the nutritional composition of food material is calculated;The above scheme can more accurately obtain the nutritional composition of each food material, which helps to realize the accurate management of dietary nutrition and improve the effect of dietary nutrition management.

Description

Technical Field

[0001] This application belongs to the field of dietary management technology and relates to a method, system and device for intelligent management of dietary nutrient intake. Background Technology

[0002] For specific groups such as patients with chronic metabolic diseases, weight managers, post-operative or critically ill patients, athletes, and malnourished patients, it is necessary to control the nutritional composition of their daily diet in order to improve their health and maintain good bodily functions. For example, diabetic patients need to precisely control their carbohydrate and sugar intake to maintain stable blood sugar levels. Post-operative or critically ill patients need to adjust their intake of protein, vitamins, and minerals according to their recovery needs.

[0003] In order to facilitate the statistical analysis and management of the intake of various dietary nutrients, existing technologies have developed image recognition-based dietary nutrient management systems. The basic working principle is as follows: food images are captured by an image acquisition device, and then the nutritional composition of the food (the types and proportions of each nutrient) is determined according to a pre-stored food-nutrient composition table. Finally, the various nutrients and their amounts ingested by the user at different times are statistically analyzed to achieve the purpose of dietary nutrient management.

[0004] While the above methods for managing users' diets are simple and convenient, they present several challenges in practice. These include: For fresh ingredients, their nutritional composition is greatly influenced by their state. For example, bananas at different stages of ripeness have vastly different total sugar content (as ripening increases, the starch in bananas gradually converts into sugar). In fact, for most fruits and vegetables, their nutritional composition (such as sugar, dietary fiber, and vitamins) changes with freshness; simply identifying the type of ingredient is insufficient to accurately determine its nutritional composition. For cooked ingredients, the lack of knowledge about the ingredients' composition and amounts, the proportions of each ingredient, and the changes in their nutritional composition during cooking leads to significant discrepancies between calculated and actual nutritional content. Ultimately, these results are unreliable as a basis for dietary management.

[0005] Clearly, knowing the amount of various nutrients a user needs to consume in their daily meals, the key to improving the effectiveness of dietary nutrient intake management lies in how to more accurately obtain the nutritional composition of various ingredients. Summary of the Invention

[0006] To address the problem that existing technologies struggle to accurately calculate the nutritional composition of various food ingredients, leading to ineffective dietary nutrition management, this application aims to provide an intelligent method for managing dietary nutrient intake. This method can more accurately calculate the nutritional composition of various food ingredients, facilitating the understanding of users' daily intake of various nutrients, thereby achieving precise dietary nutrition management and improving its effectiveness. To implement the aforementioned intelligent management method, this application also aims to provide an intelligent management system for dietary nutrient intake, and a third objective is to provide an intelligent management device for dietary nutrient intake. The specific solutions are as follows:

[0007] A method for intelligent management of dietary nutrient intake includes:

[0008] Obtain the required intake of various nutrients for users and store them as nutritional requirement indicators;

[0009] Analyze and determine the nutritional composition of various food ingredients to be consumed;

[0010] Based on the nutritional requirements and the nutritional composition of each ingredient, the intake of various ingredients is planned and a dietary recipe is generated.

[0011] This includes analyzing and determining the nutritional composition of various food ingredients to be consumed, including:

[0012] Establish and store the first association between food type and marker region and image features, the second association between food type and overall nutritional composition data of food, the third association between various food types and their corresponding marker regions, and the fourth association between the local nutritional composition data of the marker regions on various food types and the overall nutritional composition data of the food.

[0013] Acquire high-resolution images of the ingredients and determine the type of ingredients based on the first association relationship;

[0014] Based on the types of ingredients, the first overall nutritional composition data corresponding to the ingredients is obtained according to the second association relationship, and the marked region corresponding to the ingredients is obtained according to the third association relationship.

[0015] Obtain hyperspectral images of the ingredients, and based on the hyperspectral images and high-resolution images, obtain nutritional composition data of the whole ingredients and the above-mentioned marked areas, and store them as second overall nutritional composition data and local reference data, respectively.

[0016] The third overall nutritional composition data is generated based on the local reference data and according to the fourth correlation relationship;

[0017] Based on the nutritional requirement index, reference weights are assigned to the set types of nutrients. The first overall nutritional composition data and the second overall nutritional composition data are fused based on the above reference weights to generate a first feature value and a second feature value.

[0018] The difference between the first feature value and the second feature value is calculated. Based on the range of this difference, calculation weights are assigned to the second and third overall nutritional composition data. The theoretical nutritional composition data of the food is then generated based on the following formula:

[0019] N ut =ω·N ut2 +(1-ω)N ut3 ,ω∈[0,1];

[0020] Where ω is the calculation weight of the second overall nutrient composition data, and N ut2 N ut3 These are arrays, N, used to characterize the types and proportions of each nutrient in the second and third overall nutritional composition data, respectively. ut N ut2 N ut3 The length of each is n;

[0021] The nutritional composition includes the types of nutrients and their proportions.

[0022] The marked area is defined as: the area on the food that can reflect the overall freshness and / or ripeness of the food.

[0023] The above technical solution determines the nutritional composition of food ingredients not solely based on the type of food, but also by referencing the hyperspectral imaging results. The determination process comprehensively considers the interference of freshness / maturity on nutritional composition, integrating the actual hyperspectral measured nutritional composition data with theoretically predicted data. This results in a more accurate assessment of the nutritional composition of food ingredients, facilitating the understanding of users' daily nutrient intake and enabling precise management of dietary nutrition, thus improving the effectiveness of dietary nutrition management.

[0024] Furthermore, high-resolution and hyperspectral images of the ingredients are acquired, including:

[0025] The ingredients are divided and / or rotated according to set rules, and multiple high-definition images and hyperspectral images are collected;

[0026] Based on the above multiple high-definition images and the first association relationship, multiple food type results are identified and output, and the food type with the most food type results is selected as the food type determination result.

[0027] Multiple nutrient compositions are generated based on the analysis of multiple hyperspectral images. The proportions of various nutrients in the above nutrient compositions are averaged to obtain the second overall nutrient composition data.

[0028] The above technical solutions can identify ingredients from multiple angles and in all directions, ensuring the accuracy of ingredient type determination. By flipping or dividing the ingredients, high-spectral measurement technology can more accurately determine the content of each nutrient contained in the ingredients. After averaging, the nutritional composition data of the second whole ingredient can be more accurate.

[0029] Furthermore, based on the aforementioned nutritional requirements and the nutritional composition of each ingredient, the intake of various ingredients is planned and a dietary recipe is generated, including:

[0030] Establish and store the fifth correlation between each dietary recipe and the types and weight ratios of ingredients;

[0031] Establish and store the sixth correlation between changes in the nutritional composition of various ingredients and cooking methods;

[0032] Obtain the user's nutritional needs indicators and generate multiple sets of weight ratio data for various ingredients based on the nutritional composition of each ingredient;

[0033] Based on the above weight ratio data set and the fifth correlation, alternative recipes are generated;

[0034] Based on the candidate recipes and the sixth correlation, the changes in the nutritional composition of the ingredients are statistically analyzed, and the weight ratio of various ingredients in the candidate recipes is adjusted to match the dietary nutrition with the nutritional requirements, thereby generating the dietary recipes.

[0035] The above technical solution can automatically generate dietary recipes for users to choose from based on the weight ratio of various ingredients required for intake. The process fully considers the changes in the nutritional composition of ingredients during cooking and adjusts the weight ratio of each type of ingredient to ensure that the final cooked food meets the user's nutritional needs. Users can choose the above process according to their personal preferences, which facilitates dietary management.

[0036] Further analysis and determination of the nutritional composition of various edible ingredients also includes:

[0037] Acquire food ingredient label images and automatically identify the nutritional composition of the food ingredients based on the food ingredient label images; and / or

[0038] Establish a seventh correlation between user descriptions of ingredients and their nutritional composition;

[0039] Obtain the user's description of the ingredients and generate the nutritional composition of the ingredients based on the seventh association relationship;

[0040] Among them, obtaining users' descriptions of ingredients includes: collecting and recognizing users' voice data / text descriptions or click operations on the user's interactive interface;

[0041] The descriptive information includes the type of ingredients and one or more of the following: color, plumpness, smell, texture, juice viscosity, flavor, source of ingredients, pre-processing procedures, and cooking methods.

[0042] The above technical solutions can be used to directly obtain the nutritional composition of packaged food through image recognition, or to automatically determine the nutritional composition of food based on the user's description of the food's condition, which is convenient and fast.

[0043] Furthermore, the management method also includes:

[0044] Establish an eighth correlation between user-defined physiological characteristic parameters and nutritional requirement indicators;

[0045] Set the target trend of change for the physiological characteristic parameters data of the above settings;

[0046] Acquire physiological characteristic parameter data of a user within a set time period after the user consumes food prepared according to the dietary recipe, and analyze the actual change trend of the above physiological characteristic parameter data;

[0047] Analyze the difference between the actual trend and the target trend, and adjust the nutritional requirement index according to the eighth correlation.

[0048] The physiological characteristic parameters set include: weight, body fat percentage, blood pressure, blood lipid content, blood glucose content, muscle mass, systemic inflammation level, and intestinal flora structure.

[0049] The above technical solutions can be used to continuously optimize nutritional needs indicators based on feedback from users' own physiological characteristic parameters, thereby improving the effectiveness of dietary nutrition management.

[0050] A smart management system for dietary nutrient intake, comprising:

[0051] The nutrition requirement acquisition unit is configured to acquire the intake of various nutrients required by the user and store them as nutrition requirement indicators.

[0052] The food nutrition determination unit is configured to analyze and determine the nutritional composition of various food ingredients to be consumed.

[0053] The recipe generation unit is configured to plan the intake of various ingredients and generate a dietary recipe based on the nutritional requirement indicators and the nutritional composition of each ingredient.

[0054] The food nutrition determination unit includes:

[0055] The relational model storage subunit is configured to store a first relational model reflecting the relationship between food types and marker regions and image features, a second relational model reflecting the relationship between food types and overall nutritional composition data of food, a third relational model reflecting the relationship between various types of food and their corresponding marker regions, and a fourth relational model reflecting the relationship between the local nutritional composition data contained in the marker regions of various types of food and the overall nutritional composition data of food.

[0056] A high-definition image acquisition subunit is configured to acquire high-definition image data of the food to be eaten;

[0057] The hyperspectral image acquisition subunit is configured to acquire hyperspectral image data of the food to be consumed.

[0058] The ingredient type determination subunit is configured to identify and output the ingredient type based on high-definition image data and the first relationship model.

[0059] The first nutritional data generation subunit is configured to generate first overall nutritional composition data reflecting the overall nutritional composition of the current food based on the type of food and the second relationship model.

[0060] The second nutrition data generation subunit is configured to generate second overall nutrition composition data reflecting the overall nutritional composition of the current food based on the hyperspectral image data;

[0061] The third nutritional data generation subunit is configured to determine the location of the marker region based on the high-definition image data and the first relation model, obtain the nutritional composition data in the marker region by combining the hyperspectral image data, and generate third overall nutritional composition data that reflects the overall nutritional composition of the current food by combining the nutritional composition data in the marker region with the fourth correlation relationship.

[0062] The first data fusion subunit is configured to assign reference weights to a set type of nutrient according to the nutritional requirement index, and to perform fusion processing on the first overall nutritional composition data and the second overall nutritional composition data based on the reference weights to generate a first feature value and a second feature value.

[0063] The second data fusion subunit is configured to calculate the difference between the first feature value and the second feature value, assign calculation weights to the second overall nutritional composition data and the third overall nutritional composition data according to the range of the difference, and generate theoretical nutritional composition data of the ingredients based on the following calculation formula:

[0064] N ut =ω·N ut2 +(1-ω)N ut3,ω∈[0,1];

[0065] Where ω is the calculation weight of the second overall nutrient composition data, and N ut2 N ut3 These are arrays, N, used to characterize the types and proportions of each nutrient in the second and third overall nutritional composition data, respectively. ut N ut2 N ut3 The length of each is n;

[0066] The nutritional composition includes the types of nutrients and their proportions.

[0067] The marked area is defined as: the area on the food that can reflect the overall freshness and / or ripeness of the food.

[0068] Furthermore, the food nutrition determination unit also includes:

[0069] The data acquisition prompting subunit is configured to prompt the user to divide and / or flip the food according to the set rules, and to acquire multiple high-definition images and hyperspectral images;

[0070] The first data reliability optimization subunit is configured to count the number of each food type in the identification results and select the food type with the largest number as the food type determination result.

[0071] The second data reliability optimization subunit is configured to generate multiple nutrient compositions based on the analysis of multiple hyperspectral images, and to average the proportions of various nutrients in the above nutrient compositions to obtain the second overall nutrient composition data.

[0072] The above technical solutions can improve the reliability of directly acquired high-definition image data and hyperspectral image data.

[0073] Furthermore, the relation model storage subunit also stores a fifth relation model to reflect the relationship between each dietary recipe and the types and weight ratios of ingredients, and a sixth relation model to reflect the relationship between changes in the nutritional composition of various ingredients and cooking methods.

[0074] The recipe generation unit includes:

[0075] The optional ingredient weight ratio subunit can be configured to obtain the user's nutritional needs indicators and generate multiple sets of weight ratio data for various ingredients based on the nutritional composition of each ingredient.

[0076] The alternative recipe generation subunit is configured to generate alternative recipes based on the weight ratio data set and according to the fifth relationship model;

[0077] The dietary recipe generation subunit is configured to, based on the candidate recipes and the sixth relationship model, statistically analyze the changes in the nutritional composition of the ingredients, adjust the weight ratio of various ingredients in the candidate recipes to make the dietary nutrition match the nutritional requirements, and generate the dietary recipes.

[0078] Furthermore, the management system also includes a nutritional indicator optimization unit, comprising:

[0079] A physiological characteristic parameter-nutritional index relationship model is stored to reflect the correlation between user-defined physiological characteristic parameter data and nutritional requirement indicators;

[0080] The target acquisition subunit is optimized and configured to set the target change trend of the physiological characteristic parameter data of the above-mentioned settings.

[0081] The physiological data acquisition subunit is configured to acquire set physiological characteristic parameter data of a user within a set time period after the user consumes food prepared according to the dietary recipe.

[0082] The trend analysis subunit is configured to analyze the actual changing trend of the physiological characteristic parameter data of the above-mentioned settings, analyze the difference between the actual changing trend and the target changing trend, and adjust the user's nutritional requirement index according to the physiological characteristic parameter-nutritional index relationship model.

[0083] The physiological characteristic parameters set include: weight, body fat percentage, blood pressure, blood lipid content, blood glucose content, muscle mass, systemic inflammation level, and intestinal flora structure.

[0084] The above technical solutions can optimize users' nutritional needs indicators and improve the effectiveness of dietary management.

[0085] A smart device for managing dietary nutrient intake, comprising:

[0086] An image acquisition component is configured to acquire and output high-resolution image data of the food to be eaten;

[0087] A hyperspectral camera, configured to acquire and output hyperspectral image data of food ingredients to be consumed;

[0088] Physiological characteristic parameter data acquisition component, configured to collect physiological characteristic parameter data set by the user;

[0089] Interactive components, including a voice input module and a touch screen, are configured to receive user control commands and nutritional requirement indicators, and output the nutritional composition of ingredients and dietary recipes; and

[0090] The control component has a built-in intelligent management system for dietary nutrient intake as described above. It is configured to acquire the nutritional requirement indicators input by the user through the interactive component, as well as high-definition and hyperspectral images of food ingredients captured by the image acquisition component and hyperspectral camera, and output the nutritional composition of various food ingredients and dietary recipes.

[0091] This application includes at least one of the following beneficial effects:

[0092] (1) When determining the nutritional composition of ingredients, the type of ingredients is not the sole basis. Instead, the hyperspectral imaging results of the ingredients are used as a reference to determine the nutritional composition of the ingredients. The interference of the freshness / maturity of the ingredients on the nutritional composition is comprehensively considered during the determination process. The nutritional composition data actually measured by hyperspectral imaging is combined with the nutritional composition data predicted by theory, so that the determination results of the nutritional composition of ingredients are more accurate. This makes it easier to know the intake of various nutrients in the user's daily meals, and can achieve precise management of dietary nutrition and improve the effect of dietary nutrition management.

[0093] (2) During the process of generating dietary recipes, continuously optimize the user's nutritional needs indicators based on the user's physiological characteristic parameters, so that dietary management is more in line with the user's needs and improves the effectiveness of dietary management.

[0094] (3) The entire system can be built on the user's smartphone or tablet, making it easy for ordinary household users to use. Attached Figure Description

[0095] Figure 1 This is a schematic diagram of the intelligent management method of this application;

[0096] Figure 2 A schematic diagram illustrating the method for analyzing and determining the nutritional composition of various food ingredients to be consumed;

[0097] Figure 3 A schematic diagram illustrating a method for planning the intake of various ingredients and generating dietary recipes;

[0098] Figure 4 A schematic diagram illustrating the steps for optimizing nutritional requirement indicators;

[0099] Figure 5 This is a schematic diagram of the functional modules of the intelligent management system of this application;

[0100] Figure 6 A schematic diagram of the functional modules of the food nutrition determination unit;

[0101] Figure 7 A schematic diagram of the functional modules of the nutritional indicator optimization unit;

[0102] Figure 8 This is a schematic diagram of the functional modules of the intelligent management device of this application.

[0103] Figure reference numerals: 100, Nutritional Requirements Acquisition Unit; 200, Food Nutritional Determination Unit; 210, Relationship Model Storage Subunit; 220, High-Resolution Image Acquisition Subunit; 230, Hyperspectral Image Acquisition Subunit; 240, Food Type Determination Subunit; 250, First Nutritional Data Generation Subunit; 260, Second Nutritional Data Generation Subunit; 270, Third Nutritional Data Generation Subunit; 280, First Data Fusion Subunit; 290, Second Data Fusion Subunit; 300, Recipe Generation Unit 310. Optional ingredient weight ratio subunit; 320. Alternative recipe generation subunit; 330. Dietary recipe generation subunit; 400. Nutritional index optimization unit; 410. Physiological characteristic parameter-nutritional index relationship model; 420. Optimization target acquisition subunit; 430. Physiological data acquisition subunit; 440. Trend analysis subunit; 500. Image acquisition component; 600. Hyperspectral camera; 700. Physiological characteristic parameter data acquisition component; 800. Interaction component; 900. Control component. Detailed Implementation

[0104] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.

[0105] In the description of this specification, the references to "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples" refer to specific features, structures, materials, or characteristics described in connection with the described embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0106] A smart management method for dietary nutrient intake, such as Figure 1 As shown, it includes:

[0107] S100: Obtain the intake of various nutrients required by the user and store them as nutritional requirement indicators;

[0108] S200 analyzes and determines the nutritional composition of various food ingredients to be consumed;

[0109] S300: Based on the nutritional requirements and the nutritional composition of each ingredient, plan the intake of various ingredients and generate a dietary recipe.

[0110] In step S100 above, nutrients include substances such as protein, fat, carbohydrates, vitamins, minerals, and dietary fiber, which are essential for maintaining life. Users obtain the appropriate daily nutrient intake based on the advice of a nutritionist or doctor.

[0111] In step S200, the nutritional composition includes the types of nutrients and their proportions. For example, 100g of carrots contains 8.1g of carbohydrates, 0.2g of fat, 1g of protein, 100ug of carotene, 120.7mg of sodium, and 119mg of potassium.

[0112] In step S200, the nutritional composition of various food ingredients to be consumed is analyzed and determined, such as... Figure 2 As shown, it includes:

[0113] S210, Data Relationship Construction:

[0114] S2101, Establish and store the first association between food type, marker region, and image features. This facilitates subsequent steps in determining the food type and corresponding marker region based on one or more image features in the high-resolution image of the food. The extraction of image features from the high-resolution image is accomplished by a pre-defined image recognition algorithm. The marker region is defined as: an area on the food that reflects the overall freshness and / or ripeness of the food, such as the edge of a vegetable leaf or the stem area of ​​a fruit.

[0115] S2102, Establish and store a second association between food type and overall nutritional composition data. This second association can be obtained from existing food nutrition databases, such as the nutritional composition of various food ingredients from the "Chinese Food Composition Table".

[0116] S2103, Establish and store a third association between various food ingredients and their corresponding marker regions. This third association is used to determine the marker regions corresponding to the food ingredients after the ingredient types have been determined, facilitating the acquisition of the locations of these marker regions and their corresponding hyperspectral image data in subsequent steps.

[0117] S2104, Establish and store the fourth association between the local nutritional composition data of the marked regions on various food ingredients and the overall nutritional composition data of the food ingredients. In practical applications, the above-mentioned fourth association is obtained by: classifying and labeling various regions on the food ingredients according to the difficulty of image recognition, for example, leafy green vegetables can be labeled with the outer edge of the leaf, the inner edge of the leaf, the edge of the petiole, the center of the petiole, the root and stem region, etc.; acquiring multiple hyperspectral image data of different individuals of the same type of food ingredients, statistically analyzing the nutritional composition in each marked region, and storing them in association with the marked regions to form a training dataset; based on the above-mentioned training dataset, obtaining marked regions that are closely related to the nutritional composition data of their own nutritional composition and the overall nutritional composition data of the food ingredients through neural network training, selecting them as marked regions, and simultaneously generating the association between the local nutritional composition data contained in the above-mentioned marked regions and the overall nutritional composition data of the food ingredients.

[0118] S220: Obtain high-definition images of the ingredients and determine the types of ingredients, such as the names of vegetables, fruits, meat, eggs, and poultry, based on the first association relationship.

[0119] S230, based on the type of food ingredient, obtain the first overall nutritional composition data corresponding to the food ingredient according to the second association relationship, and obtain the marked region corresponding to the food ingredient according to the third association relationship. The aforementioned first overall nutritional composition data comes from an existing food nutrition database, and the marked region is mainly for food ingredients with uneven distribution of nutrients, such as fruits and vegetables.

[0120] S240, acquire a hyperspectral image of the food ingredient, and acquire nutritional composition data of the whole food ingredient and the above-mentioned marked area based on the hyperspectral image and the high-definition image, and store them as second overall nutritional composition data and local reference data respectively.

[0121] S250, based on the local reference data and according to the fourth correlation, generate third overall nutritional composition data.

[0122] S260, assign reference weights to the specified nutrient types based on the nutritional requirement indicators, and fuse the first overall nutritional composition data and the second overall nutritional composition data based on the aforementioned reference weights to generate a first feature value and a second feature value. The significance of this data fusion is to simplify the representation of the nutritional composition of food ingredients, discard unnecessary interfering data items, and facilitate the subsequent generation of dietary recipes. For example, if a user needs to supplement 600mg of potassium per day, it is recommended to eat 200g of cauliflower (each 100g of cauliflower contains approximately 299mg of potassium). When fusing the nutritional composition data of cauliflower, potassium can be given a higher weight, while the weight of other nutrients, such as vitamin C and dietary fiber, can be reduced. The specific weight allocation method needs to be determined based on the nutritional requirement indicators.

[0123] S270, calculate the difference between the first feature value and the second feature value, assign calculation weights to the second overall nutritional composition data and the third overall nutritional composition data according to the range of the difference, and generate theoretical nutritional composition data of the ingredients based on the following calculation formula:

[0124] N ut =ω·N ut2 +(1-ω)N ut3 ,ω∈[0,1];

[0125] Where ω is the calculation weight of the second overall nutrient composition data, and N ut2 N ut3 These are arrays, N, used to characterize the types and proportions of each nutrient in the second and third overall nutritional composition data, respectively. ut N ut2 N ut3 The length of each array is n, where n is the number of selected nutrients. For example, an array with n = 5, where the 5 elements represent carbohydrates, fats, proteins, sodium, and potassium, can be represented as: N ut2 =[8.1, 0.2, 1, 0.12, 0.12], the default unit is g / 100g.

[0126] In this application, the proportion of each type of nutrient is a weight percentage, and in a specific embodiment, it can be defined as the proportion of energy supplied in the total calories.

[0127] In practical applications, when the difference between the first and second feature values ​​is too large, it indicates that the overall detection error of the current hyperspectral image data may be too large. In this case, the calculation weight of the second overall nutritional composition data can be reduced, while the calculation weight of the third overall nutritional composition data can be increased. Reasons for a large overall detection error in food products include factors such as the surface shape of the food itself and its placement during image capture. For example, the placement of a banana can significantly affect the results of hyperspectral measurements; in this case, the accuracy of hyperspectral measurement data for a local area of ​​the food is much higher than the accuracy of the overall hyperspectral measurement data.

[0128] The significance of step S270 is that when the accuracy of the overall hyperspectral measurement results of the food is low, the calculation weight of the overall nutritional composition data is reduced, and the calculation weight of the local nutritional composition data is increased, so that the final theoretical nutritional composition data is more accurate.

[0129] In order to identify the types of ingredients from multiple angles and in a comprehensive manner, and to ensure the accuracy of ingredient identification, in step S220, high-resolution images of the ingredients are acquired, including:

[0130] S221, Divide and / or flip the ingredients according to the set rules, and collect multiple high-definition images;

[0131] S222, based on the above multiple high-definition images and the first association relationship, identify and output multiple food type results, and select the food type with the most food type results as the food type determination result.

[0132] In step S221 above, the food ingredients are segmented according to the set rules, including cutting and separating the food ingredients. The purpose is to highlight the image features that can represent the type of food ingredients in the high-definition image.

[0133] In step S240, a hyperspectral image of the food ingredient is acquired, including:

[0134] S241, Divide and / or flip the ingredients according to the set rules, and collect multiple hyperspectral images;

[0135] S242, generate multiple nutrient compositions based on the analysis of multiple hyperspectral images, and average the proportions of various nutrients in the above nutrient compositions to obtain the second overall nutrient composition data.

[0136] The segmentation operation in step S241 above includes slicing the ingredients, such as removing the skin of fruits and vegetables and only counting the edible flesh, separating the outer layer of vegetables from the main body, and estimating the overall nutritional composition of vegetables by measuring specific nutrients in the heart of the vegetable.

[0137] The above method involves flipping or dividing the ingredients, and then using hyperspectral imaging technology to more accurately determine the content of each nutrient in the ingredients. After averaging, the nutritional composition data of the second whole ingredient becomes more accurate.

[0138] In the optimized step S200, the analysis and determination of the nutritional composition of various edible ingredients also includes:

[0139] S280, direct acquisition steps:

[0140] The nutritional composition of the food can be obtained by directly acquiring the food label image and automatically identifying it based on the food label image. For example, the characters in the image can be identified directly through an OCR recognition algorithm to obtain the percentage of the corresponding nutrients.

[0141] S290, manual acquisition steps:

[0142] S291, establish the seventh association between user descriptions of ingredients and their nutritional composition. The aforementioned descriptions include the type of ingredient and one or more of its characteristics, such as color, plumpness, aroma, texture, juice viscosity, flavor, origin, pre-processing procedures, and cooking methods. Obtaining user descriptions of ingredients includes: collecting and recognizing user voice data / text descriptions or clicks on the user interface.

[0143] S292, Obtain the user's description information of the ingredients, and generate the nutritional composition of the ingredients based on the seventh association relationship.

[0144] Unlike image recognition or direct hyperspectral measurement, the above-mentioned manual acquisition step determines the nutritional composition of food by directly collecting the user's description information. It has good applicability for some food products whose types are difficult to determine through images (such as beef and mutton products).

[0145] In step S300, based on the nutritional requirement indicators and the nutritional composition of each ingredient, the intake of various ingredients is planned and a dietary recipe is generated, such as... Figure 3 As shown, it includes:

[0146] S310, Steps for constructing a relational model:

[0147] S3101, Establish and store the fifth association relationship between each dietary recipe and the types and weight ratios of ingredients. The weight ratios of various ingredients in each dietary recipe can be obtained from the existing recipe database.

[0148] S3102, Establish and store the sixth correlation between changes in the nutritional composition of various ingredients and cooking methods. The aforementioned sixth correlation can be obtained from existing culinary nutrition databases.

[0149] S320 obtains the user's nutritional needs indicators and generates multiple sets of weight ratio data for various ingredients based on their nutritional composition. For example, when 200ug of folic acid needs to be supplemented, one can choose to eat about 200g of broccoli (broccoli contains about 100ug / 100g of folic acid) or 1000g of carrots (carrots contain about 20.7ug / 100g of folic acid). Obviously, it is also possible to generate a weight ratio of 160g of broccoli plus 200g of carrots, which is easier to form alternative recipes.

[0150] S330, Based on the above weight ratio data set and the fifth correlation, generate alternative recipes.

[0151] S340, based on the candidate recipes and the sixth correlation, statistically analyze the changes in the nutritional composition of the ingredients, adjust the weight ratio of various ingredients in the candidate recipes to make the dietary nutrition match the nutritional requirements, and generate the dietary recipes.

[0152] The above technical solution can automatically generate a dietary recipe for users to choose from based on the weight ratio of various ingredients to be ingested. The process fully considers the changes in the nutritional composition of ingredients during cooking and adjusts the weight ratio of each type of ingredient to ensure that the final cooked food meets the user's nutritional needs. Users can choose the above process according to their personal preferences, which is convenient for dietary management.

[0153] In order to continuously optimize nutritional requirements based on feedback from users' own physiological characteristic parameters and improve the effectiveness of dietary nutrition management, such as Figure 4 As shown, the management method described in the embodiments of this application further includes:

[0154] S400, steps to optimize nutritional requirements indicators:

[0155] S410, establish an eighth correlation between user-defined physiological characteristic parameters and nutritional requirement indicators. These physiological characteristic parameters include, but are not limited to, one or more of the following: weight, body fat percentage, blood pressure, blood lipid levels, blood glucose levels, muscle mass, systemic inflammation levels, and gut microbiota structure. This eighth correlation can be obtained from existing professional medical databases, such as the *Chinese Dietary Reference Intakes* and *Nutrition and Food Hygiene*, or it can be customized by a professional nutritionist.

[0156] S420, set the target change trend of the physiological characteristic parameter data of the above settings, such as controlling the rise rate of the user's postprandial blood glucose within the set range.

[0157] S430: Obtain physiological characteristic parameter data of the user within a set time period after the user consumes food prepared according to the dietary recipe, and analyze the actual change trend of the physiological characteristic parameter data.

[0158] S440, Analyze the difference between the actual trend and the target trend, and adjust the nutritional requirement index according to the eighth correlation.

[0159] To implement the above-mentioned intelligent management method for dietary nutrient intake, this application also discloses an intelligent management system for dietary nutrient intake, combined with... Figure 5 As shown, it mainly includes: a nutritional requirement acquisition unit 100, an ingredient nutrition determination unit 200, and a recipe generation unit 300.

[0160] The nutrition requirement acquisition unit 100 is configured to acquire the required intake of various nutrients for the user and store them as nutrition requirement indicators. The food nutrition determination unit 200 is configured to analyze and determine the nutritional composition of various foods to be consumed. The recipe generation unit 300 is configured to plan the intake of various foods and generate a dietary recipe based on the nutrition requirement indicators and the nutritional composition of each food.

[0161] Detailed, such as Figure 6As shown, the food nutrition determination unit 200 includes: a relational model storage subunit 210, a high-definition image acquisition subunit 220, a hyperspectral image acquisition subunit 230, a food type determination subunit 240, a first nutrition data generation subunit 250, a second nutrition data generation subunit 260, a third nutrition data generation subunit 270, a first data fusion subunit 280, and a second data fusion subunit 290.

[0162] The relational model storage subunit 210 includes a data storage device configured to store a first relational model reflecting the relationship between food types and marker regions and image features, a second relational model reflecting the relationship between food types and overall nutritional composition data of food, a third relational model reflecting the relationship between various types of food and their corresponding marker regions, and a fourth relational model reflecting the relationship between the local nutritional composition data contained in the marker regions of various types of food and the overall nutritional composition data of the food.

[0163] The high-definition image acquisition subunit 220 includes an image data interface module that is connected to an image acquisition device, such as a smartphone camera or a dedicated high-definition camera, and is configured to acquire high-definition image data of the food to be consumed. The hyperspectral image acquisition subunit 230 includes a hyperspectral measurement data interface module that is connected to a hyperspectral camera 600 and is configured to acquire hyperspectral image data of the food to be consumed.

[0164] The ingredient type determination subunit 240 includes an image recognition algorithm module, configured to connect with the high-definition image acquisition subunit 220 to acquire and identify the ingredient type based on the high-definition image data and the first relationship model.

[0165] The first nutritional data generation subunit 250 includes a data mapping module configured to be data-connected to the food ingredient type determination subunit 240 and the relational model storage subunit 210. This module generates first overall nutritional composition data reflecting the overall nutritional composition of the current food ingredient based on the received food ingredient type and a second relational model. The second nutritional data generation subunit 260 includes a data integration module configured to be data-connected to the hyperspectral image acquisition subunit 230 and the relational model storage subunit 210. This module generates second overall nutritional composition data reflecting the overall nutritional composition of the current food ingredient based on the received hyperspectral image data.

[0166] The third nutrition data generation subunit 270 is configured to be data-connected with the high-definition image acquisition subunit 220 and the relational model storage subunit 210. It is used to determine the location of the marker area based on the acquired high-definition image data and the first relational model, obtain the nutritional composition data in the marker area in combination with the hyperspectral image data, and generate the third overall nutritional composition data to reflect the overall nutritional composition of the current food based on the nutritional composition data in the marker area and the fourth correlation relationship.

[0167] Both the first data fusion subunit 280 and the second data fusion subunit 290 are built into the data processor and are data-connected to the nutritional requirement acquisition unit 100, the first nutritional data generation subunit 250, and the second nutritional data generation subunit 260. The first data fusion subunit 280 is configured to assign reference weights to a set type of nutrient based on nutritional requirement indicators, and to fuse the first overall nutritional composition data and the second overall nutritional composition data based on the aforementioned reference weights to generate a first feature value and a second feature value. The second data fusion subunit 290 is configured to calculate the difference between the first feature value and the second feature value, and to assign calculation weights to the second overall nutritional composition data and the third overall nutritional composition data based on the difference range, generating theoretical nutritional composition data of the ingredients based on the following calculation formula:

[0168] N ut =ω·N ut2 +(1-ω)N ut3 ,ω∈[0,1];

[0169] Where ω is the calculation weight of the second overall nutrient composition data, and N ut2 N ut3 These are arrays, N, used to characterize the types and proportions of each nutrient in the second and third overall nutritional composition data, respectively. ut N ut2 N ut3 The length of each is n.

[0170] The above nutritional composition includes the types of nutrients and their proportions, and the marked area is defined as the area on the food that can reflect the overall freshness and / or maturity of the food.

[0171] The optimized food nutrition determination unit 200 also includes: a data acquisition prompt subunit, a first data reliability optimization subunit, and a second data reliability optimization subunit.

[0172] The data acquisition prompt subunit is configured to output prompts to guide the user to segment and / or flip the food according to set rules, and to acquire multiple high-resolution and hyperspectral images. The first data reliability optimization subunit is configured to count the number of each food type in the identification results and select the food type with the highest number as the food type determination result. The second data reliability optimization subunit is configured to analyze multiple hyperspectral images to generate multiple nutritional compositions, and to average the proportion of each nutrient in the above nutritional compositions to obtain the second overall nutritional composition data.

[0173] In this embodiment of the application, the relational model storage subunit 210 also stores a fifth relational model used to reflect the relationship between each dietary recipe and the types and weight ratios of ingredients, and a sixth relational model used to reflect the relationship between changes in the nutritional composition of various ingredients and cooking methods.

[0174] Based on the above configuration, the aforementioned recipe generation unit 300 includes: an ingredient optional weight ratio subunit 310, an alternative recipe generation subunit 320, and a dietary recipe generation subunit 330.

[0175] The ingredient selection weight ratio subunit 310 is configured to acquire the user's nutritional requirements and generate multiple sets of weight ratio data for various ingredients based on their nutritional composition. The alternative recipe generation subunit 320 is configured to connect to the ingredient selection weight ratio subunit 310 and the relational model storage subunit 210, receive and generate alternative recipes based on the weight ratio data sets and the fifth relational model. The dietary recipe generation subunit 330 is configured to receive and, based on the alternative recipes, the sixth relational model, and the user-selected cooking method, statistically analyze changes in the nutritional composition of the ingredients, adjust the weight ratios of various ingredients in the alternative recipes to match dietary nutrition with nutritional requirements, and generate the dietary recipe.

[0176] In this embodiment of the application, the management system further includes a nutritional index optimization unit 400, such as... Figure 7 As shown, it includes: a physiological characteristic parameter-nutritional index relationship model 410, an optimization target acquisition subunit 420, a physiological data acquisition subunit 430, and a trend analysis subunit 440.

[0177] The physiological characteristic parameter-nutritional index relationship model 410 is stored to reflect the correlation between user-defined physiological characteristic parameter data and nutritional requirement indicators. The optimization target acquisition subunit 420 is configured to set a target change trend for the aforementioned physiological characteristic parameter data based on user input instructions. The physiological data acquisition subunit 430 is configured to connect to external detection devices such as smart bracelets to acquire set physiological characteristic parameter data for a set period after the user consumes food prepared according to the dietary recipe. The trend analysis subunit 440 is configured to analyze the actual change trend of the acquired set physiological characteristic parameter data, analyze the difference between the actual change trend and the target change trend, and adjust the user's nutritional requirement indicators according to the physiological characteristic parameter-nutritional index relationship model 410.

[0178] The physiological characteristics parameters set above include, but are not limited to: weight, body fat percentage, blood pressure, blood lipid content, blood glucose content, muscle mass, systemic inflammation level, and gut microbiota structure.

[0179] Finally, this application also discloses an intelligent dietary nutrient intake management device, combined with Figure 8 As shown, it includes an image acquisition component 500, a hyperspectral camera 600, a physiological characteristic parameter data acquisition component 700, an interaction component 800, and a control component 900.

[0180] Image acquisition component 500 includes a smartphone camera or a dedicated camera for capturing and outputting high-resolution image data of the food to be consumed. Hyperspectral camera 600 is used to acquire and output hyperspectral image data of the food to be consumed. Physiological characteristic parameter data acquisition component 700 is configured to acquire user-defined physiological characteristic parameter data, including but not limited to devices such as smart bracelets, electronic blood glucose meters, and blood pressure monitors with relevant acquisition functions.

[0181] The interactive component 800 includes a voice input module and a touch screen. In this embodiment, it is preferably a tablet computer or a smartphone, configured to receive user control commands and nutritional requirement indicators, and output the nutritional composition of ingredients and dietary recipes.

[0182] The control component 900 includes a control APP and related modules installed on the aforementioned tablet or smartphone. It has a built-in intelligent dietary nutrient intake management system, as described above, configured to acquire nutritional requirement indicators input by the user via the interaction component 800, as well as high-resolution and hyperspectral images of food ingredients acquired by the image acquisition component 500 and the hyperspectral camera 600. It then outputs the nutritional composition of various ingredients and dietary recipes. In practical applications, the control component 900 connects to the image acquisition component 500 and the hyperspectral camera 600 via Bluetooth or Wi-Fi communication modules to acquire relevant data. Simultaneously, it outputs prompts via the interaction component 800 to remind the user to pay attention to their daily nutritional intake and guides the user in optimizing their dietary recipes.

[0183] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for intelligent management of dietary nutrient intake, characterized in that, include: Obtain the required intake of various nutrients for users and store them as nutritional requirement indicators; Analyze and determine the nutritional composition of various food ingredients to be consumed; Based on the nutritional requirements and the nutritional composition of each ingredient, the intake of various ingredients is planned and a dietary recipe is generated. This includes analyzing and determining the nutritional composition of various food ingredients to be consumed, including: Establish and store the first association between food type and marker region and image features, the second association between food type and overall nutritional composition data of food, the third association between various food types and their corresponding marker regions, and the fourth association between the local nutritional composition data of the marker regions on various food types and the overall nutritional composition data of the food. Acquire high-resolution images of the ingredients and determine the type of ingredients based on the first association relationship; Based on the types of ingredients, the first overall nutritional composition data corresponding to the ingredients is obtained according to the second association relationship, and the marked region corresponding to the ingredients is obtained according to the third association relationship. Obtain hyperspectral images of the ingredients, and based on the hyperspectral images and high-resolution images, obtain nutritional composition data of the whole ingredients and the above-mentioned marked areas, and store them as second overall nutritional composition data and local reference data, respectively. The third overall nutritional composition data is generated based on the local reference data and according to the fourth correlation relationship; Based on the nutritional requirement index, reference weights are assigned to the set types of nutrients. The first overall nutritional composition data and the second overall nutritional composition data are fused based on the above reference weights to generate a first feature value and a second feature value. The difference between the first feature value and the second feature value is calculated. Based on the range of this difference, calculation weights are assigned to the second and third overall nutritional composition data. The theoretical nutritional composition data of the food is then generated based on the following formula: N ut =ω·N ut2 +(1-ω)N ut3 ,ω∈[0,1]; Where ω is the calculation weight of the second overall nutrient composition data, and N ut2 N ut3 These are arrays, N, used to characterize the types and proportions of each nutrient in the second and third overall nutritional composition data, respectively. ut N ut2 N ut3 The length of each is n; The nutritional composition includes the types of nutrients and their proportions. The marked area is defined as: the area on the food that can reflect the overall freshness and / or ripeness of the food.

2. The intelligent management method for dietary nutrient intake according to claim 1, characterized in that, Acquire high-resolution and hyperspectral images of the ingredients, including: The ingredients are divided and / or rotated according to set rules, and multiple high-definition images and hyperspectral images are collected; Based on the above multiple high-definition images and the first association relationship, multiple food type results are identified and output, and the food type with the most food type results is selected as the food type determination result. Multiple nutrient compositions are generated based on the analysis of multiple hyperspectral images. The proportions of various nutrients in the above nutrient compositions are averaged to obtain the second overall nutrient composition data.

3. The intelligent management method for dietary nutrient intake according to claim 2, characterized in that, Based on the aforementioned nutritional requirements and the nutritional composition of each ingredient, the intake of various ingredients is planned and a dietary menu is generated, including: Establish and store the fifth correlation between each dietary recipe and the types and weight ratios of ingredients; Establish and store the sixth correlation between changes in the nutritional composition of various ingredients and cooking methods; Obtain the user's nutritional needs indicators and generate multiple sets of weight ratio data for various ingredients based on the nutritional composition of each ingredient; Based on the above weight ratio data set and the fifth correlation, alternative recipes are generated; Based on the candidate recipes and the sixth correlation, the changes in the nutritional composition of the ingredients are statistically analyzed, and the weight ratio of various ingredients in the candidate recipes is adjusted to match the dietary nutrition with the nutritional requirements, thereby generating the dietary recipes.

4. The intelligent management method for dietary nutrient intake according to claim 1, characterized in that, Analyzing and determining the nutritional composition of various food ingredients to be consumed also includes: Acquire food ingredient label images and automatically identify the nutritional composition of the food ingredients based on the food ingredient label images; and / or Establish a seventh correlation between user descriptions of ingredients and their nutritional composition; Obtain the user's description of the ingredients and generate the nutritional composition of the ingredients based on the seventh association relationship; Among them, obtaining users' descriptions of ingredients includes: collecting and recognizing users' voice data / text descriptions or click operations on the user's interactive interface; The descriptive information includes the type of ingredients and one or more of the following: color, plumpness, smell, texture, juice viscosity, flavor, source of ingredients, pre-processing procedures, and cooking methods.

5. The intelligent management method for dietary nutrient intake according to claim 1, characterized in that, The management method also includes: Establish an eighth correlation between user-defined physiological characteristic parameters and nutritional requirement indicators; Set the target trend of change for the physiological characteristic parameters data of the above settings; Acquire physiological characteristic parameter data of a user within a set time period after the user consumes food prepared according to the dietary recipe, and analyze the actual change trend of the above physiological characteristic parameter data; Analyze the difference between the actual trend and the target trend, and adjust the nutritional requirement index according to the eighth correlation. The physiological characteristic parameters set include: weight, body fat percentage, blood pressure, blood lipid content, blood glucose content, muscle mass, systemic inflammation level, and intestinal flora structure.

6. A smart management system for dietary nutrient intake, characterized in that, include: Nutritional requirement acquisition unit (100) is configured to acquire the intake of various nutrients required by the user and store them as nutritional requirement indicators; The food nutrition determination unit (200) is configured to analyze and determine the nutritional composition of various food ingredients to be consumed. The recipe generation unit (300) is configured to plan the intake of various ingredients and generate a dietary recipe based on the nutritional requirements indicators and the nutritional composition of each ingredient. The food nutrition determination unit (200) includes: The relational model storage subunit (210) is configured to store a first relational model reflecting the relationship between food types and marker regions and image features, a second relational model reflecting the relationship between food types and overall nutritional composition data of food, a third relational model reflecting the relationship between various food types and their corresponding marker regions, and a fourth relational model reflecting the relationship between the local nutritional composition data contained in the marker regions of various food types and the overall nutritional composition data of food. High-definition image acquisition subunit (220) is configured to acquire high-definition image data of the food to be eaten; The hyperspectral image acquisition subunit (230) is configured to acquire hyperspectral image data of the food to be consumed; The ingredient type determination subunit (240) is configured to identify and output the ingredient type based on high-definition image data and the first relationship model; The first nutritional data generation subunit (250) is configured to generate first overall nutritional composition data reflecting the overall nutritional composition of the current food based on the type of food and the second relationship model. The second nutrition data generation subunit (260) is configured to generate second overall nutrition composition data reflecting the overall nutritional composition of the current food based on the hyperspectral image data; The third nutritional data generation subunit (270) is configured to determine the location of the marker region based on the high-definition image data and the first relation model, obtain the nutritional composition data in the marker region by combining the hyperspectral image data, and generate third overall nutritional composition data to reflect the overall nutritional composition of the current food by combining the nutritional composition data in the marker region with the fourth correlation relationship. The first data fusion subunit (280) is configured to assign reference weights to a set type of nutrient according to the nutritional requirement index, and to perform fusion processing on the first overall nutritional composition data and the second overall nutritional composition data based on the reference weights to generate a first feature value and a second feature value. The second data fusion subunit (290) is configured to calculate the difference between the first feature value and the second feature value, assign calculation weights to the second overall nutritional composition data and the third overall nutritional composition data according to the range of the difference, and generate theoretical nutritional composition data of the ingredients based on the following calculation formula: N ut =ω·N ut2 +(1-ω)N ut3 ,ω∈[0,1]; Where ω is the calculation weight of the second overall nutrient composition data, and N ut2 N ut3 These are arrays, N, used to characterize the types and proportions of each nutrient in the second and third overall nutritional composition data, respectively. ut N ut2 N ut3 The length of each is n; The nutritional composition includes the types of nutrients and their proportions. The marked area is defined as: the area on the food that can reflect the overall freshness and / or ripeness of the food.

7. The intelligent management method for dietary nutrient intake according to claim 6, characterized in that, The food nutrition determination unit (200) also includes: The data acquisition prompting subunit is configured to prompt the user to divide and / or flip the food according to the set rules, and to acquire multiple high-definition images and hyperspectral images; The first data reliability optimization subunit is configured to count the number of each food type in the identification results and select the food type with the largest number as the food type determination result. The second data reliability optimization subunit is configured to generate multiple nutrient compositions based on the analysis of multiple hyperspectral images, and to average the proportions of various nutrients in the above nutrient compositions to obtain the second overall nutrient composition data.

8. The intelligent management system for dietary nutrient intake according to claim 7, characterized in that, The relation model storage subunit (210) also stores a fifth relation model to reflect the relationship between each dietary recipe and the types and weight ratios of ingredients, and a sixth relation model to reflect the relationship between changes in the nutritional composition of various ingredients and cooking methods. The recipe generation unit (300) includes: The optional ingredient weight ratio subunit (310) is configured to obtain the user's nutritional needs indicators and generate multiple sets of weight ratio data for various ingredients based on the nutritional composition of each ingredient. The alternative recipe generation subunit (320) is configured to generate alternative recipes based on the weight ratio data set and according to the fifth relationship model; The dietary recipe generation subunit (330) is configured to generate the dietary recipe by statistically analyzing the changes in the nutritional composition of the ingredients based on the candidate recipes and the sixth relationship model, adjusting the weight ratio of various ingredients in the candidate recipes to make the dietary nutrition match the nutritional requirements indicators.

9. The intelligent management system for dietary nutrient intake according to claim 7, characterized in that, The management system also includes a nutrition index optimization unit (400), comprising: Physiological characteristic parameter-nutritional index relationship model (410) is stored to reflect the correlation between user-defined physiological characteristic parameter data and nutritional requirement indicators; The target acquisition subunit (420) is optimized and configured to set the target change trend of the physiological characteristic parameter data of the above-mentioned setting items; The physiological data acquisition subunit (430) is configured to acquire set physiological characteristic parameter data of a user within a set time period after the user consumes food prepared according to the dietary recipe; The trend analysis subunit (440) is configured to analyze the actual change trend of the physiological characteristic parameter data of the above-mentioned settings, analyze the difference between the actual change trend and the target change trend, and adjust the user's nutritional needs index according to the physiological characteristic parameter-nutritional index relationship model (410). The physiological characteristic parameters set include: weight, body fat percentage, blood pressure, blood lipid content, blood glucose content, muscle mass, systemic inflammation level, and intestinal flora structure.

10. A smart device for managing dietary nutrient intake, characterized in that, include: An image acquisition component (500) is configured to acquire and output high-resolution image data of the food to be eaten; A hyperspectral camera (600) is configured to acquire and output hyperspectral image data of the food to be consumed; Physiological characteristic parameter data acquisition component (700), configured to acquire user-defined physiological characteristic parameter data; Interactive component (800), including voice input module and touch screen, is configured to receive user control commands and nutritional requirement indicators, and output the nutritional composition of ingredients and dietary recipes; as well as The control component (900) has a built-in intelligent management system for dietary nutrient intake as described above. It is configured to acquire the nutritional requirement indicators input by the user through the interaction component (800), as well as high-definition images and hyperspectral images of food ingredients acquired by the image acquisition component (500) and the hyperspectral camera (600), and output the nutritional composition of various food ingredients and dietary recipes.

Images