A navigation method based on privacy protection

By constructing a privacy-preserving navigation system, utilizing blockchain and distributed computing to identify users' health status, and combining Bayesian and A-Star algorithms to recommend meal routes, the problem of users finding it difficult to find suitable meals in the cafeteria has been solved, achieving both privacy protection and health recommendations.

CN115048474BActive Publication Date: 2024-11-05SHENZHEN JILIAN TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210543337.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-05-19
Publication Date
2024-11-05
Estimated Expiration
2042-05-19

AI Technical Summary

Technical Problem

In large canteens, users find it difficult to quickly find meals that meet their tastes and health needs, and the marketing advertisements pushed by the existing canteen monitoring system infringe on user privacy, causing user annoyance.

Method used

A privacy-preserving navigation system is constructed, which identifies users' health and activity status through a blockchain database and distributed computing storage model. It uses a Bayesian classifier and A-Star algorithm to recommend the best meal route and protects user privacy through differential privacy technology to prevent advertising systems from obtaining personal information.

Benefits of technology

It enables the recommendation of suitable meals based on the user's health status, protects user privacy, prevents advertising systems from obtaining personal information, and improves the user's dining experience and health benefits.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115048474B_ABST
    Figure CN115048474B_ABST
Patent Text Reader

Abstract

The present application provides a navigation method based on privacy protection, including: The present application provides a navigation method based on privacy protection, including: constructing a privacy-protected navigation system, specifically including: establishing a blockchain database to protect health privacy; detecting the user's health status and exercise status of the day before dining, the user's health status and exercise status of the day before dining are detected, specifically including: detecting the user's health status at any position, analyzing the uniqueness of the user's health status; analyzing the optimal diet content according to the health status index; navigating indoors to the optimal window according to the optimal diet content; protecting the privacy of the user's status, and obfuscating the recommended data and path.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a navigation method based on privacy protection. Background Art

[0002] In large exhibitions or schools with tens of thousands of people, multiple floors and multiple stalls are often set up in a cafeteria. In super-large cafeterias, users need to try and taste many times to know which window to go to for food, which one is suitable for their taste, which one is healthier, and which place has which kind of dishes. However, how to better recommend suitable dishes and food windows to users so that users can eat the food that suits them at one time is a problem worth solving. On the other hand, general cafeterias have monitoring systems that can push relevant marketing advertising coupons and other things through mobile APP public accounts based on personal eating habits. For example, patients with thyroiditis need to eat high-fiber foods. After the merchant obtains the relevant data, it will promote products to the patient, which exposes personal diet or condition on the one hand, and on the other hand, it is annoying to receive frequent recommended advertisements. Therefore, it is necessary to encrypt the push and confuse the advertising recommendation system. Summary of the invention

[0003] The present invention provides a navigation method based on privacy protection, which mainly includes:

[0004] Construct a privacy-protected navigation system; detect the user's health status and exercise status of the day before dining; analyze the best diet content based on the health status index; navigate to the best window indoors based on the best diet content; protect the privacy of the user's status, and confuse the recommended data and path; the privacy algorithm and the obfuscation algorithm avoid being photographed by the advertising system to see the eating habits and health status; Further optionally, the construction of the privacy-protected navigation system includes: The privacy-protected navigation system includes a user information unit, a user health status detection unit, a user exercise status detection unit of the day, a diet content analysis unit, an indoor navigation unit, and a user data privacy protection unit; the user information query unit includes querying the user's health status, querying the user's exercise status of the day, querying the user's diet content, and the user's facial data information; the user health detection unit includes a function of detecting the user's health status; the diet content analysis unit includes analyzing the best diet content based on the user's health status and exercise status of the day; the indoor navigation unit includes navigating the user to the best window based on the diet content; the user data privacy protection unit protects the user data; including: establishing a blockchain database to protect health privacy;

[0005] The establishment of a blockchain database to protect health privacy specifically includes:

[0006] Based on the user data obtained by the system, including user health information, dietary information, and daily exercise status information, a blockchain database is established. The blockchain database uses a distributed computing storage model to remove the centralized management structure. Each data block uses the Spark distributed computing system to distribute computing tasks to various machines for parallel computing.

[0007] Further optionally, the detecting the user's health status and exercise status of the day before dining includes:

[0008] First, the system obtains the camera information of the cafeteria, which includes the image information of the cafeteria; based on the obtained cafeteria image information, MediaPipe is used to identify and detect faces. Mideapipe takes face information as input and detection and recognition results as output; among them, data packets are the most basic data units of Mediapipe, and a data packet is the information data of a specific time node; based on the face data of the detection and recognition results of Mediapipe, the face data of the detection and recognition results are classified with the face data information in the database through the Bayesian classic classifier, and the classification result is the user recognition result; the user information is determined according to the classification result of the Bayesian classic classifier, and the user's height information is calculated according to the posture information detected by MediaPipe; the exercise status information of the day is obtained, and the exercise status information includes the exercise location, exercise content, and exercise duration, and the diet content is adjusted according to the exercise status information; including: detecting the user's health status at any position; analyzing the uniqueness of the user's health status;

[0009] The detecting of the user's health status at any position specifically includes:

[0010] First, the system uses the facial data information detected and recognized by Mediapipe to mark the indoor map location of the unmeasured user. The indoor map location of the unmeasured user is sent to the mobile weight measuring instrument terminal. The mobile weight measuring instrument randomly selects the indoor map location of the unmeasured user to measure the weight and body fat of different users. A health status index is established, and the health status index = weight - (height - 105) + body fat. A health status index less than 18 or a health status index greater than 24 indicates that the user is in poor health. A health status index greater than 18 and a health status index less than 24 indicate that the user is in good health.

[0011] The analysis of the uniqueness of the user's health status specifically includes:

[0012] According to the user's health status index, a mathematical model is established to analyze the uniqueness of the user's health status. The health status index of each user is analyzed, and the frequency of all different health status indexes is calculated, that is, the number of people with a certain user's health status index / the number of all users. The result is the uniqueness of the health status of a certain user. If a user's health status index is lower than 10%, it means that the user's health status is unique. Adjust the navigation parameters according to the uniqueness of the user's health status. Among the paths recommended by the path navigation, the parameter path with fewer billboards in the navigation parameters is given priority. If the number of billboards on the paths is the same, the path with less traffic is preferred. Change the navigation parameters to protect the privacy and security of people with relatively unique health status, and avoid the heterogeneous characteristics of the population being more easily collected when the advertising recommendation system collects information.

[0013] Further optionally, analyzing the optimal diet content according to the health status index includes:

[0014] A dietary content dataset is established, including a vegetable dietary content dataset, a meat dietary content dataset, and a beverage dietary content dataset, wherein the corresponding calories of each item in the dataset are annotated as the dataset label, and the unit of the dataset label is calories / 100g; based on the analyzed health status index, a Bayesian model is used to train the relationship between various food attributes and health status scores to predict the optimal dietary content; different weights are assigned to different types of dishes from small to large according to the health status index being greater than 24, less than 24 and greater than 18, and less than 18; the prediction results are obtained according to the basic mathematical formula of the Bayesian model (Word document output); based on the prediction results of the Bayesian model, the optimal dietary content for the user is obtained.

[0015] Further optionally, the indoor navigation to the best window according to the best diet content includes:

[0016] Image recognition technology is used according to the images corresponding to the cameras in the cafeteria. The Mediapipe framework is used to identify the flow of people at different locations in the cafeteria, and a path planning algorithm is constructed. The input is the position of the window corresponding to the best food content and the user's location, and the output is the best path for indoor navigation. The A-Star algorithm is used to construct a path planning algorithm. The A-Star algorithm combines the Dijkstra algorithm with the breadth-first search algorithm (BFS). The A-Star algorithm adds the initial point position to the open list openlist; adds the node to be processed to the closed list closelist; for each of the 8 adjacent nodes of the current node, if it is unreachable or already in the closed list closelist, the A-Star algorithm ignores the node; traverses the adjacent nodes until the best window position is added to the open list openlist, indicating that the path has been found; the path planning combines the size of the flow of people in different paths with the flow of people, and recommends the best navigation path for users.

[0017] Further optionally, the privacy protection of the user status and the obfuscation of the recommended data and path include:

[0018] The database is established using blockchain technology, and a distributed computing storage model is used to remove the centralized management structure and encrypt private data, which includes the user's health information, diet content, and exercise status of the day. Differential privacy encryption technology is used to add uncertainty to the query recommendation data and user path results. Among them, differential privacy, through the Laplacian mechanism, for adjacent data sets, the corresponding mapping function is given to add Laplacian noise to the results of the user query, generating a final result that is slightly different from the original result. The Laplacian noise formula is: (Word document output).

[0019] Further optionally, the privacy algorithm and the obfuscation algorithm to avoid the eating habits and health status being captured by the advertising system include: first, obtaining the image corresponding to the camera in the cafeteria; according to the obtained image, using the obfuscation algorithm to add Laplacian noise to blur the image, so that the advertising system cannot obtain clear image data from the camera; the image data corresponding to the camera in the cafeteria is stored in the database using a privacy algorithm, wherein the database uses blockchain technology to encrypt the data to protect the security of user privacy, and the advertising system cannot modify the data in the blockchain database; the user's health status data is stored in the database using a privacy algorithm, wherein the database uses blockchain technology to encrypt the data to protect the security of user privacy, and the advertising system cannot modify the data in the blockchain database; including: adjusting navigation parameters according to the uniqueness of the user's health status to avoid advertising; adjusting navigation parameters according to the user's exercise status to avoid advertising recommendations; adjusting navigation parameters based on the user's position to avoid advertising recommendations;

[0020] The method of adjusting navigation parameters according to the uniqueness of the user's health status to avoid advertisements specifically includes:

[0021] First, the uniqueness of the user's health status is obtained. According to the uniqueness of the user's health status, the weight of the crowd flow analysis data parameters in the navigation algorithm on the navigation result output is adjusted. The uniqueness of the user's health status is high, and the weight of the crowd flow data in the navigation algorithm is reduced, and the path with fewer billboards is preferred. The adjustment of navigation parameters according to the user's motion status to avoid advertising recommendations specifically includes:

[0022] First, obtain the user's movement status. According to the user's movement status, adjust the weight of the crowd flow analysis data parameters in the navigation algorithm for the navigation result output. If the user's movement status is high intensity, reduce the weight of the crowd flow data in the navigation algorithm, and give priority to paths with fewer billboards to avoid users who are tired and unable to rationally choose the advertising content on the billboards.

[0023] The step of adjusting navigation parameters based on user positions to avoid advertisement recommendations specifically includes:

[0024] First, obtain the user's position information and the salary level corresponding to the position. Set a salary level threshold, and adjust the weight of the traffic analysis data parameters in the navigation algorithm to the navigation result output according to the threshold. If the user's salary level is greater than the threshold, reduce the weight of the traffic data in the navigation algorithm, give priority to paths with fewer billboards, and avoid frequent exposure to advertisements for users with strong purchasing power. The technical solution provided by the embodiment of the present invention may include the following beneficial effects:

[0025] The present invention identifies the user's identity, determines the user's health information status, guides the user to dine according to the user's health status, and protects the user's health data privacy, preventing businesses from using marketing advertisements and other means to control the user's privacy to coerce the user's consumption. The present invention recommends exercise routes based on the user's dining health, and adjusts the user's next dining navigation according to the exercise route, so that the user can combine the exercise route with the dining health to obtain higher health benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 The present invention is a flowchart of a navigation method based on privacy protection.

[0027] Figure 2 A schematic diagram of a navigation method based on privacy protection of the present invention. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is described in detail below with reference to the accompanying drawings and specific embodiments. Figure 1 FIG. 1 is a flow chart of a navigation method based on privacy protection according to the present invention. Figure 1 As shown, a navigation method based on privacy protection in this embodiment may specifically include:

[0029] Step 101, construct a privacy-protected navigation system. The privacy-protected navigation system includes a user information unit, a user health status detection unit, a user's current day exercise status detection unit, a diet content analysis unit, an indoor navigation unit, and a user data privacy protection unit. The user information query unit includes querying the user's health status, querying the user's current day exercise status, querying the user's diet content, and the user's face data information. The user health detection unit includes a function for detecting the user's health status. The diet content analysis unit includes analyzing the best diet content based on the user's health status and the current day exercise status. The indoor navigation unit includes navigating the user to the best window based on the diet content. The user data privacy protection unit protects the user data. A blockchain database is established to protect health privacy. According to the user data obtained by the system, the user data includes user health information, diet content information, and current day exercise status information, and a blockchain database is established. The blockchain database uses a distributed computing storage mode to remove the centralized management structure. Each data block distributes the computing tasks to each machine by using the Spark distributed computing system for parallel computing.

[0030] Step 102, before dining, the user's health status and exercise status of the day are detected. First, the system obtains the camera information of the cafeteria, and the camera information includes the cafeteria image information. According to the obtained cafeteria image information, MediaPipe is used to identify and detect faces. Mideapipe uses face information as input and detection and recognition results as output. Among them, the data packet is the most basic data unit of Mediapipe, and a data packet is the information data of a specific time node. According to the face data of the detection and recognition results of Mediapipe, the face data of the detection and recognition results and the face data information in the database are classified by the Bayesian classic classifier, and the classification result is the user identification result. Determine the user information according to the classification result of the Bayesian classic classifier, and calculate the user's height information according to the posture information detected by MediaPipe. Obtain the exercise status information of the day, which includes the exercise location, exercise content, and exercise duration, and adjust the diet content according to the exercise status information. For example, the user can access the user information unit to query his or her health status, exercise status, and diet content. The user uses the health detection unit to detect his or her own health status. After the health detection unit completes the detection, the system analyzes the best diet content according to the user's health status and the day's exercise status. After the system analyzes the best diet content, it navigates the user to the best window according to the diet content. The system's data passes through the user data privacy protection unit to protect and process user data. Detect the user's health status at any location. First, the system marks the indoor map location of the unmeasured user with the face data information detected and recognized by Mediapipe. The indoor map location of the unmeasured user is sent to the mobile weight measuring instrument terminal. The mobile weight measuring instrument randomly selects the indoor map location of the unmeasured user to measure the weight and body fat of different users. Establish a health status index, health status index = weight - (height - 105) + body fat. If the health status index is less than 18 or the health status index is greater than 24, it indicates that the user is in poor health. If the health status index is greater than 18 and the health status index is less than 24, it indicates that the user is in good health. For example, after obtaining the user's health information, diet content information, and exercise status information of the day, the system uses the Spark distributed computing system to distribute the computing tasks to each machine for parallel computing for each data block, thereby improving the system's computing efficiency. Analyze the uniqueness of the user's health status. According to the user's health status index, a mathematical model is established to analyze the uniqueness of the user's health status. Analyze the health status index of each user and calculate the frequency of all different health status indexes, that is, the number of people with a certain user's health status index / the number of all users. The result is the uniqueness of the health status of a certain user. If a user's health status index is less than 10%, it means that the user's health status is unique.Adjust navigation parameters according to the uniqueness of the user's health status. Among the paths recommended by the path navigation, give priority to the path with fewer billboards in the navigation parameters. If the number of billboards on the paths is the same, give priority to the path with less traffic. Change the navigation parameters to protect the privacy and security of people with relatively unique health status, and avoid the advertising recommendation system from collecting heterogeneous characteristics of the crowd more easily when collecting information. For example, after a user enters the cafeteria, the system obtains the cafeteria image information from the camera information. Mediapipe detects and recognizes faces and classifies the face data of the detection and recognition results with the face data information in the database through the Bayesian classic classifier. Based on the classification results, the system obtains the user's height information and exercise status information for the day.

[0031] Step 103, analyze the best diet content according to the health status index. Establish a diet content data set, including a vegetable diet content data set, a meat diet content data set, and a beverage diet content data set, wherein the corresponding calories of each content in the data set are marked as a data set label, and the unit of the data set label is calories / 100g. According to the analyzed health status index, use the Bayesian model to train the relationship between various food attributes and health status scores to predict the best diet content. According to the health status index greater than 24, less than 24 and greater than 18, less than 18, different weights are assigned to different types of dishes from small to large. According to the basic mathematical formula of the Bayesian model (Word document output), the prediction result is obtained. According to the prediction result of the Bayesian model, the best diet content for the user is obtained. For example, when there are multiple people in the cafeteria, the system detects a face and marks the indoor map location of the unmeasured user. The marked indoor map location of the unmeasured user is assigned to a movable weight measuring instrument. The movable weight measuring instrument then goes to the corresponding user location and randomly selects the indoor map location of the unmeasured user to measure the weight and body fat of different users. If the health status index of a user is 19, it indicates that the user is in good health.

[0032] Step 104, navigate to the best window indoors according to the best food content. Use image recognition technology based on the image corresponding to the camera in the cafeteria, use the Mediapipe framework to identify the flow of people at different locations in the cafeteria, build a path planning algorithm, input the position of the window corresponding to the best food content and the user's location, and output the best path for indoor navigation. Use the A-Star algorithm to build a path planning algorithm, which combines the Dijkstra algorithm with the breadth-first search algorithm (BFS). The A-Star algorithm adds the initial point position to the open list openlist. Add the node to be processed to the closed list closelist. For each of the 8 adjacent nodes of the current node, if it is unreachable or already in the closed list closelist, the A-Star algorithm ignores the node. Traverse the adjacent nodes until the best window position is added to the open list openlist, indicating that the path has been found. Path planning combines the size of the flow of people in different paths with the flow of people to recommend the best navigation path for users. For example, a user's health status index is 19, there are 30 users with a health status index of 19, and there are 1,000 system users. The uniqueness of the user's health status is 3%, which is less than 10%, indicating that the user's status is unique. Adjust navigation parameters based on the uniqueness of the user's health status to avoid advertisements.

[0033] Step 105, privacy protection of user status, obfuscation of recommended data and paths. Use blockchain technology to establish a database, and use a distributed computing storage model to remove the centralized management structure and encrypt private data. The private data includes the user's health information, diet content, and exercise status of the day. Use differential privacy encryption technology to add uncertainty to the query recommendation data and user path results. Among them, differential privacy, through the Laplacian mechanism, for adjacent data sets, given the corresponding mapping function, adds Laplacian noise to the results of the user query, and generates a final result that is slightly different from the original result. Among them, the Laplacian noise formula is: (Word document output). For example, a user's health status index is 19. Among vegetables, kelp has 12 calories / 100g, lettuce has 14 calories / 100g, bean sprouts have 18 calories / 100g, among meats, pig ears have 190 calories / 100g, chicken has 200 calories / 100g, ham has 320 calories / 100g, and among drinks, soy milk has 14 calories / 100g, sugarcane juice has 64 calories / 100g, and cola has 95 calories / 100g. According to the user's health status index, the weights of different dishes are assigned. Among them, the weights of vegetables, meat, and drinks are divided from small to large according to calories, which are 0.25, 0.5, and 0.25 respectively. According to the Bayesian prediction model analysis, the best diet content is lettuce, chicken, and sugarcane juice. The best diet content for the user is kelp, ham, and sugarcane juice.

[0034] Step 106, the privacy algorithm and the obfuscation algorithm avoid being photographed by the advertising system for dietary rules and health status. First, obtain the image corresponding to the camera in the cafeteria. According to the obtained image, the obfuscation algorithm is used to add Laplacian noise to blur the image, and the advertising system cannot obtain clear image data from the camera. The image data corresponding to the camera in the cafeteria is stored in the database using a privacy algorithm, wherein the database uses blockchain technology to encrypt the data to protect the security of user privacy, and the advertising system cannot modify the data in the blockchain database. The user's health status data is stored in the database using a privacy algorithm, wherein the database uses blockchain technology to encrypt the data to protect the security of user privacy, and the advertising system cannot modify the data in the blockchain database. For example, the Mediapipe algorithm obtains the cafeteria image, analyzes the flow of people at different locations in the cafeteria, finds the flow of people on the corresponding path according to the A* algorithm, and recommends the path with the least flow of people to the user. Adjust the navigation parameters according to the uniqueness of the user's health status to avoid advertising. First, obtain the uniqueness of the user's health status. According to the uniqueness of the user's health status, adjust the weight of the traffic analysis data parameters in the navigation algorithm to the output of the navigation result. The uniqueness of the user's health status is high, so the weight of the traffic data in the navigation algorithm is reduced, and the path with fewer billboards is given priority. For example, a user suffers from diabetes. The user's health information, diet content, and exercise status of the day are stored in the established blockchain database. The disease requires a large amount of black fungus, and black fungus is in Area 1 of the cafeteria. The Laplacian mechanism is used to confuse the diet content and path in the user's recommended data, so that the results are slightly different, and other users cannot query the correct recommended data and path. Adjust the navigation parameters according to the user's exercise status to avoid advertising recommendations. First, obtain the user's exercise status. According to the user's exercise status, adjust the weight of the traffic analysis data parameters in the navigation algorithm to the output of the navigation results. The user's exercise status is high in exercise intensity, so the weight of the traffic data in the navigation algorithm is reduced, and the path with fewer billboards is given priority to avoid the user being unable to rationally choose the advertising content on the billboard when he is tired. For example, a cafeteria advertising system wants to obtain images in the cafeteria to analyze the user's dietary habits. The obfuscation algorithm uses Laplacian noise to blur the image. The advertising system cannot obtain valid image data from the image. The image data is stored in the blockchain database, and the advertising system cannot modify the data in the blockchain database. A cafeteria advertising system hopes to obtain the user's health status to gain benefits. It uses a privacy algorithm to store the user's health status data in the database. The advertising system cannot modify the data in the blockchain database. Adjust navigation parameters based on the user's position to avoid advertising recommendations. First, obtain the user's position information and the salary level corresponding to the position. Set the salary level threshold and adjust the weight of the traffic analysis data parameters in the navigation algorithm on the navigation result output according to the threshold.When the user's salary level is greater than the threshold, the weight of the traffic data in the navigation algorithm is reduced, and the path with fewer billboards is preferred to avoid frequent exposure to advertisements by users with strong purchasing power. For example, the uniqueness of the user's health status is high, and the weight of the traffic data in the navigation algorithm is reduced. If there is a path with fewer billboards, the path with fewer billboards is preferred, otherwise the path with fewer people is preferred. The above description is only an embodiment of the present invention, and does not limit the patent scope of the present invention. All equivalent structures or equivalent process transformations made by using the contents of the present invention specification and drawings, or directly or indirectly used in other related technical fields, are also included in the patent protection scope of the present invention. The program used to implement the information control of the present invention can be written in one or more programming languages ​​or a combination thereof to write computer program codes for performing the operations of the present invention, and the programming language includes object-oriented programming languages-such as Java, python, C++, and also includes conventional procedural programming languages-such as C language or similar programming languages. The program code can be executed completely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet). In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the present embodiment. In addition, the functional units in the various embodiments of the present invention can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of hardware plus software functional units. The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium, and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to perform some steps of the method described in each embodiment of the present invention.The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

Claims

1. A navigation method based on privacy protection, It is characterized in that The method comprises: Establish a blockchain database to protect health privacy; detect the user's health status and exercise status of the day before dining, which specifically includes: detecting the user's health status at any position and analyzing the uniqueness of the user's health status; analyzing the best diet content according to the health status index; navigating to the best window indoors according to the best diet content; protecting the privacy of the user's status, and obfuscating the recommended data and paths; using privacy algorithms and obfuscation algorithms to avoid being photographed by the advertising system to capture the diet rules and health status, which specifically includes: adjusting navigation parameters according to the uniqueness of the user's health status to avoid advertising, adjusting navigation parameters according to the user's exercise status to avoid advertising recommendations, and adjusting navigation parameters based on the user's position to avoid advertising recommendations; The dining process first detects the user's health status and exercise status for the day, including: First, the system obtains the camera information of the cafeteria, which includes the image information of the cafeteria; based on the obtained cafeteria image information, MediaPipe is used to identify and detect faces. Mideapipe takes face information as input and detection and recognition results as output; among them, data packets are the most basic data units of Mediapipe, and a data packet is the information data of a specific time node; based on the face data of the detection and recognition results of Mediapipe, the face data of the detection and recognition results are classified with the face data information in the database through the Bayesian classic classifier, and the classification result is the user recognition result; the user information is determined according to the classification result of the Bayesian classic classifier, and the user's height information is calculated according to the posture information detected by MediaPipe; the exercise status information of the day is obtained, and the exercise status information includes the exercise location, exercise content, and exercise duration, and the diet content is adjusted according to the exercise status information; including: detecting the user's health status at any position; analyzing the uniqueness of the user's health status; The detecting of the user's health status at any position specifically includes: First, the system uses the facial data information detected and recognized by Mediapipe to mark the indoor map location of the unmeasured user; the indoor map location of the unmeasured user is sent to the mobile weight measuring instrument terminal; the mobile weight measuring instrument randomly selects to go to the indoor map location of the unmeasured user to measure the weight and body fat of different users; a health status index is established, where the health status index = weight - (height - 105) + body fat; a health status index less than 18 or a health status index greater than 24 indicates that the user is in poor health; a health status index greater than 18 and a health status index less than 24 indicates that the user is in good health; The analysis of the uniqueness of the user's health status specifically includes: According to the user's health status index, a mathematical model is established to analyze the uniqueness of the user's health status; the health status index of each user is analyzed, and the frequency of all different health status indexes is calculated, that is, the number of people with a certain user's health status index / the number of all users, and the result is the uniqueness of the health status of a certain user; if the health status index of a certain user is lower than 10%, it means that the health status of the user is unique; according to the uniqueness of the user's health status, the navigation parameters are adjusted, and among the paths recommended by the path navigation, the parameter path with fewer billboards in the navigation parameters is given priority. If the number of billboards on the paths is the same, the path with less traffic is given priority; the navigation parameters are changed to protect the privacy and security of people with relatively unique health status, so as to avoid the heterogeneous characteristics of the population being more easily collected when the advertising recommendation system collects information; Said indoor navigation to the best window according to the best diet content includes: According to the images corresponding to the cameras in the cafeteria, image recognition technology is used, and the Mediapipe framework is used to identify the flow of people at different locations in the cafeteria, and a path planning algorithm is constructed. The input is the position of the window corresponding to the best food content and the user's location, and the output is the best path for indoor navigation; the path planning algorithm is constructed using the A-Star algorithm, which combines the Dijkstra algorithm with the breadth-first search algorithm (BFS); the A-Star algorithm adds the initial point position to the open list openlist; adds the node to be processed to the closed list closelist; for each of the 8 adjacent nodes of the current node, if it is unreachable or already in the closed list closelist, the A-Star algorithm ignores the node; traverses the adjacent nodes until the best window position is added to the open list openlist, indicating that the path has been found; the path planning combines the size of the flow of people in different paths with the flow of people, and recommends the best navigation path for users; The privacy protection of the user status, recommended data and path obfuscation, includes: The database is established using blockchain technology, and a distributed computing storage model is used to remove the centralized management structure and encrypt private data, including the user's health information, diet content, and exercise status of the day. Differential privacy encryption technology is used to add uncertainty to the query recommendation data and user path results. Among them, differential privacy, through the Laplacian mechanism, for adjacent data sets, the corresponding mapping function is given to add Laplacian noise to the result of the user query, generating a final result that is slightly different from the original result. Among them, the Laplacian noise formula is: P(A i B i C i )=P(A i )*P(B i )*P(C i ), where A i A type of vegetable diet, B i A type of meat diet, C i Get prediction results for a certain category of beverage and food content; The privacy algorithm and obfuscation algorithm prevent the eating habits and health status from being captured by the advertising system, including: First, obtain the image corresponding to the camera in the cafeteria; based on the obtained image, use the obfuscation algorithm to add Laplacian noise to blur the image, so that the advertising system cannot obtain clear image data from the camera; the image data corresponding to the camera in the cafeteria is stored in the database using a privacy algorithm, where the database uses blockchain technology to encrypt the data to protect the security of user privacy, and the advertising system cannot modify the data in the blockchain database; the user's health status data is stored in the database using a privacy algorithm, where the database uses blockchain technology to encrypt the data to protect the security of user privacy, and the advertising system cannot modify the data in the blockchain database; including: adjusting navigation parameters according to the uniqueness of the user's health status to avoid advertising; adjusting navigation parameters according to the user's exercise status to avoid advertising recommendations; adjusting navigation parameters based on the user's position to avoid advertising recommendations.

2. The method according to claim 1, in, The establishment of a blockchain database to protect health privacy specifically includes: Based on the user data obtained by the system, which includes user health information, dietary content information, and daily exercise status information, a blockchain database is established; the blockchain database uses a distributed computing storage model to remove the centralized management structure; each data block distributes the computing tasks to each machine by using the Spark distributed computing system for parallel computing.

3. The method according to claim 1, in, The optimal diet content according to the health status index analysis includes: Establish a dietary content data set, including a vegetable dietary content data set, a meat dietary content data set, and a beverage dietary content data set. The corresponding calories of each content in the data set are marked as data set labels, and the unit of the data set label is calories / 100g. According to the analyzed health status index, the Bayesian model is used to train the relationship between various food attributes and health status scores to predict the optimal dietary content. According to the health status index greater than 24, less than 24 and greater than 18, and less than 18, different types of dishes are assigned different weights from small to large. According to the basic mathematical formula of the Bayesian model, P(A i B i C i )=P(A i )*P(B i )*P(C i ), where A i A type of vegetable diet, B i A type of meat diet, C i Get the prediction result for a certain type of beverage and food content; get the best food content for the user based on the prediction result of the Bayesian model.

4. The method according to claim 1, in, The method of adjusting navigation parameters according to the uniqueness of the user's health status to avoid advertisements specifically includes: First, the uniqueness of the user's health status is obtained; according to the uniqueness of the user's health status, the weight of the crowd flow analysis data parameters in the navigation algorithm on the navigation result output is adjusted; when the uniqueness of the user's health status is high, the weight of the crowd flow data in the navigation algorithm is reduced, and the path with fewer billboards is given priority; The step of adjusting navigation parameters according to the user's motion state to avoid advertisement recommendations specifically includes: First, the user's movement status is obtained; according to the user's movement status, the weight of the crowd flow analysis data parameters in the navigation algorithm on the navigation result output is adjusted; if the user's movement status is high intensity, the weight of the crowd flow data in the navigation algorithm is reduced, and the path with fewer billboards is given priority to avoid the user being unable to rationally choose the advertising content on the billboard when he is tired; The step of adjusting navigation parameters based on user positions to avoid advertisement recommendations specifically includes: First, obtain the user's position information and the salary level corresponding to the position; set a salary level threshold, and adjust the weight of the traffic analysis data parameters in the navigation algorithm for the navigation result output according to the threshold; if the user's salary level is greater than the threshold, reduce the weight of the traffic data in the navigation algorithm, and give priority to paths with fewer billboards to avoid frequent exposure to advertisements for users with strong purchasing power.

5. A navigation system based on privacy protection, It is characterized in that The system includes: a user information unit, a user health status detection unit, a user's current day exercise status detection unit, a diet content analysis unit, an indoor navigation unit, and a user data privacy protection unit; the user information query unit includes querying the user's health status, querying the user's current day exercise status, querying the user's diet content, and the user's face data information; the user health detection unit includes a detection function for the user's health status, which is specifically used to detect the user's health status before dining, and the detection of the user's health status before dining specifically includes: detecting the user's health status at any position and analyzing the uniqueness of the user's health status; the user's current day exercise status detection unit is used to detect the user's current day exercise status before dining; the diet content analysis unit includes, according to the user's health status and the current day exercise status, The optimal diet content of the state analysis is specifically used to analyze the optimal diet content according to the health state index; the indoor navigation unit includes navigating the user to the optimal window according to the diet content; the user data privacy protection unit protects the user data; including: establishing a blockchain database to protect health privacy; privacy protection of user status, and obfuscating recommended data and paths; privacy algorithm and obfuscation algorithm to avoid being photographed by the advertising system to see the diet rules and health status, and the privacy algorithm and obfuscation algorithm to avoid being photographed by the advertising system to see the diet rules and health status, specifically including: adjusting navigation parameters according to the uniqueness of the user's health status to avoid advertising, adjusting navigation parameters according to the user's exercise status to avoid advertising recommendations, and adjusting navigation parameters based on the user's position to avoid advertising recommendations; Before dining, the user's health status is detected, including: first, the system obtains the camera information of the cafeteria, and the camera information includes the cafeteria image information; based on the obtained cafeteria image information, MediaPipe is used to identify and detect faces, and Mideapipe uses face information as input and detection and recognition results as output; among them, data packets are the most basic data units of Mediapipe, and a data packet is information data at a specific time node; based on the face data of the Mediapipe detection and recognition results, the face data of the detection and recognition results are classified with the face data information in the database through the Bayesian classic classifier, and the classification result is the user identification result; the user information is determined according to the classification result of the Bayesian classic classifier, and the user's height information is calculated according to the posture information detected by MediaPipe; before dining, the exercise status of the day is detected, including: obtaining the exercise status information of the day, the exercise status information includes exercise location, exercise content, exercise duration, and adjusting the diet content according to the exercise status information; The arbitrary position detection of user health status specifically includes: first, the system marks the indoor map position of the unmeasured user with the face data information detected and recognized by Mediapipe; the indoor map position of the unmeasured user is sent to the mobile weight measuring instrument terminal; the mobile weight measuring instrument randomly selects to go to the indoor map position of the unmeasured user to measure the weight and body fat of different users; establishes a health status index, where the health status index = weight - (height - 105) + body fat; if the health status index is less than 18 or the health status index is greater than 24, it indicates that the user is in poor health; if the health status index is greater than 18 and the health status index is less than 24, it indicates that the user is in good health; The analysis of the uniqueness of the user's health status specifically includes: establishing a mathematical model based on the user's health status index to analyze the uniqueness of the user's health status; analyzing the health status index of each user, calculating the frequency of all different health status indexes, that is, the number of people with the health status index of a certain user / the number of all users, and the result obtained is the uniqueness of the health status of a certain user; if the health status index of a certain user is lower than 10%, it means that the health status of the user is unique; adjusting the navigation parameters according to the uniqueness of the user's health status, among the paths recommended by the path navigation, giving priority to the parameter path with fewer billboards in the navigation parameters, if the number of billboards of the paths is the same, giving priority to the path with less traffic; changing the navigation parameters to protect the privacy and security of people with relatively unique health status, and avoid the heterogeneous characteristics of the population being more easily collected when the advertising recommendation system collects information; The method of indoor navigation to the best window according to the best dietary content comprises: using image recognition technology according to the image corresponding to the camera in the cafeteria, using the Mediapipe framework to identify the flow of people at different positions in the cafeteria, constructing a path planning algorithm, inputting the position of the window corresponding to the best dietary content and the position of the user, and outputting the best path for indoor navigation; constructing a path planning algorithm using the A-Star algorithm, which combines the Dijkstra algorithm with the breadth-first search algorithm (BFS); the A-Star algorithm adds the initial point position to the open list openlist; adds the node to be processed to the closed list closelist; for each of the eight adjacent nodes of the current node, if it is unreachable or already in the closed list closelist, the A-Star algorithm ignores the node; traverses the adjacent nodes until the best window position is added to the open list openlist, indicating that the path has been found; the path planning combines the size of the flow of people in different paths with the flow of people, and recommends the best navigation path for the user; The privacy protection of the user status and the obfuscation of the recommended data and paths include: using blockchain technology to establish a database and using a distributed computing storage mode to remove the centralized management structure and encrypt the private data, which includes the user's health information, diet content, and exercise status of the day; using differential privacy encryption technology to add uncertainty to the query recommendation data and user path results; wherein, differential privacy, through the Laplacian mechanism, for adjacent data sets, gives a corresponding mapping function to add Laplacian noise to the result of the user query, and generates a final result that is slightly different from the original result; wherein, the Laplacian noise formula is: P(A i B i C i )=P(A i )*P(B i )*P(C i ), where A i A type of vegetable diet, B i A type of meat diet, C i Get prediction results for a certain category of beverage and food content; The privacy algorithm and the obfuscation algorithm avoid the eating habits and health status being captured by the advertising system, including: first, obtaining the image corresponding to the camera in the cafeteria; according to the obtained image, using the obfuscation algorithm to add Laplacian noise to blur the image, so that the advertising system cannot obtain clear image data from the camera; the image data corresponding to the camera in the cafeteria is stored in the database using the privacy algorithm, wherein the database uses the blockchain technology to encrypt the data to protect the security of the user's privacy, and the advertising system cannot modify the data in the blockchain database; the user's health status data is stored in the database using the privacy algorithm, wherein the database uses the blockchain technology to encrypt the data to protect the security of the user's privacy, and the advertising system cannot modify the data in the blockchain database; including: adjusting the navigation parameters according to the uniqueness of the user's health status to avoid advertising; adjusting the navigation parameters according to the user's exercise status to avoid advertising recommendations; adjusting the navigation parameters based on the user's position to avoid advertising recommendations.