Myopia prevention and control system and method based on AI learning
By introducing personnel attribute information and AI learning technology into myopia prevention and control system, personalized myopia prevention and control models are trained, and real-time monitoring information is used for prevention and control, the problem of inaccurate myopia prevention and control caused by the unified correction standards in the existing technology has been solved. , significantly reduces the probability of myopia.
Patent Information
- Application Number
- CN202311704248.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-12-12
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2043-12-12
AI Technical Summary
The existing myopia prevention and control system based on AI learning has unified standards for correcting deviations when dealing with different user attributes, resulting in insufficient accuracy in myopia prevention and control and increasing the probability of myopia.
Introduce personnel attribute information, train personalized myopia prevention and control models based on AI learning technology, and combine real-time monitoring information to prevent and control, so as to improve the accuracy of prevention and control.
Through personalized myopia prevention and control models and real-time monitoring information, the probability of myopia is significantly reduced and the accuracy of prevention and control is improved.
Smart Images

Figure CN117711567B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of AI learning technology, and in particular to a myopia prevention and control system and method based on AI learning. Background Art
[0002] AI learning refers to the process by which an AI system learns and improves through algorithms and data. AI learning can be divided into different methods and techniques such as supervised learning, unsupervised learning, reinforcement learning, and deep learning. Myopia is a common eye health problem, especially in modern society, where many people use their eyes at close range for a long time, such as using electronic devices and reading, which increases the risk of myopia. Myopia prevention and control systems and methods based on AI learning can provide personalized myopia prevention and control advice and support by combining artificial intelligence technology and knowledge in the field of ophthalmology.
[0003] The invention patent with application number CN201910276914.3 discloses an intelligent anti-myopia system based on AI learning, wherein the system includes: an eye behavior monitoring unit, which includes a distance sensor and an acceleration sensor; an MCU central processor is also provided in the device body for receiving and processing the eye data sent back by the sensing unit, the server and the MCU central processor exchange data transmission, the server learns the eye data sent back by a large number of users and continuously optimizes the algorithm, and a mobile terminal is also provided between the path of the wireless signal transceiver and the server, the mobile terminal displays the server's analysis results of the eye data to the user and interacts with the MCU central processor and the server respectively under the control of the user. The above invention can quantitatively detect a variety of variables related to eye habits, and use big data technology and artificial intelligence learning to continuously optimize the algorithm, which is conducive to comprehensive analysis of eye habits and can correct deviations from multiple dimensions, reminding users to gradually correct bad eye habits.
[0004] However, the above-mentioned prior art only describes a comprehensive analysis of the user's eye habits and correction from multiple dimensions. However, when the user attributes (such as eye diseases or eye-related diseases, etc.) are different, the correction standards are also different. If a unified correction standard is used, myopia prevention and control is not accurate enough, which will further increase the user's myopia probability.
[0005] In view of this, there is an urgent need for a myopia prevention and control system and method based on AI learning to at least solve the above-mentioned deficiencies. Summary of the invention
[0006] One of the purposes of the present invention is to provide a myopia prevention and control system based on AI learning, introduce the personal attribute information of the prevention and control personnel who need to carry out myopia prevention and control, train the myopia prevention and control model based on the personnel attribute information based on AI learning technology, determine the prevention and control results based on the myopia prevention and control model and real-time monitoring information, thereby improving the accuracy of prevention and control and further reducing the probability of myopia.
[0007] The myopia prevention and control system based on AI learning provided by the embodiment of the present invention includes:
[0008] The first acquisition subsystem is used to determine the prevention and control personnel who need to carry out myopia prevention and control, and obtain the personnel attribute information of the prevention and control personnel;
[0009] Intelligent learning subsystem, used to train myopia prevention and control models based on AI learning technology and personnel attribute information;
[0010] The second acquisition subsystem is used to obtain real-time monitoring information of prevention and control personnel;
[0011] The prevention and control result determination subsystem is used to conduct myopia prevention and control in real time based on the myopia prevention and control model and real-time monitoring information to obtain prevention and control results.
[0012] Preferably, the first acquisition subsystem includes:
[0013] Identity ID acquisition module, used to obtain the identity ID of the prevention and control personnel;
[0014] The personnel attribute information determination module is used to determine the personnel attribute information according to the identity ID and the preset personnel attribute information database.
[0015] Preferably, the intelligent learning subsystem includes:
[0016] An artificial intelligence learning record acquisition module, used to acquire a first AI learning record;
[0017] A second AI learning record determination module, used to determine the second AI learning record in the first AI learning record according to the personnel attribute information;
[0018] A comprehensive value analysis module is used to conduct a comprehensive analysis of the training of the second AI learning record to obtain a comprehensive value;
[0019] A third AI learning record determination module, configured to use the corresponding second AI learning record as the third AI learning record if the comprehensive value is greater than or equal to a preset comprehensive value threshold;
[0020] A small sample learning module, configured to determine a fourth AI learning record based on the corresponding second AI learning record based on a small sample learning technology if the full value is less than a preset full value threshold;
[0021] The myopia prevention and control model training module is used to train the myopia prevention and control model according to the third AI learning record or the fourth AI learning record.
[0022] Preferably, the comprehensive value analysis module includes:
[0023] a marked record collection submodule, used to collect the fifth AI learning record; wherein each fifth AI learning record is pre-marked with a comprehensive label;
[0024] A first feature vector determination submodule, configured to determine a first data feature based on a preset data feature extraction template and according to a fifth AI learning record, and to represent the first data feature as a first feature vector;
[0025] A deep learning model acquisition submodule is used to acquire a preset deep learning model;
[0026] A comprehensiveness analysis model determination submodule, used to determine a comprehensiveness analysis model based on a deep learning model according to the first feature vector and the corresponding comprehensiveness label;
[0027] A second feature vector determination submodule, configured to extract a second data feature of the second AI learning record based on the data feature extraction template, and represent the second data feature as a second feature vector;
[0028] The comprehensive value output submodule is used to input the second eigenvector into the comprehensiveness analysis model to obtain the comprehensive value.
[0029] Preferably, the prevention and control result determination subsystem includes:
[0030] An information sorter acquisition module is used to acquire an information sorter preset at the front end of the myopia prevention and control model;
[0031] A comparison result determination module is used to input the real-time monitoring information into the information sorter and determine the comparison result of the preset comparator corresponding to the sorting unit of the information sorter;
[0032] A feature information determination module, for obtaining, if there is a mismatch in the comparison result, the feature information of the sorting unit whose output comparison result is a mismatch;
[0033] The myopia prevention and control module is used to prevent and control myopia based on feature information.
[0034] Preferably, the myopia prevention and control module includes:
[0035] A feature information parsing submodule, used to parse the feature information and obtain the first prevention and control target and the first prevention and control type;
[0036] A prevention and control strategy determination submodule, used to determine a prevention and control strategy according to the first prevention and control type;
[0037] The strategy type acquisition submodule is used to obtain the strategy type of the prevention and control strategy; the strategy types include: immediate strategy and subsequent strategy;
[0038] The real-time prevention and control submodule is used to determine the target position of the first prevention and control target when the strategy type is the real-time strategy, and to dispatch the prevention and control robot to perform corresponding prevention and control according to the target position;
[0039] The subsequent prevention and control submodule is used to carry out corresponding prevention and control of the first prevention and control target based on the questionnaire reminder rules when the strategy type is a subsequent strategy.
[0040] Preferably, the real-time prevention and control submodule includes:
[0041] A target map acquisition unit, used to acquire a target map where the first prevention and control target is located;
[0042] A point determination unit, used to determine that the target position corresponds to a first point on the target map, and at the same time, determine that the idle control robot corresponds to a second point on the target map;
[0043] The first control target traversal unit is used to traverse the first control target in sequence, take the first control target currently being traversed as the second control target, determine the first point corresponding to the second control target, and use it as the third point;
[0044] An associated prevention and control robot determining unit, used to determine an associated prevention and control robot corresponding to the second prevention and control target according to the second point, the third point and the target map;
[0045] A fourth point determination unit, used to determine a fourth point of the associated control robot;
[0046] A second prevention and control type determination unit is used to control the associated prevention and control robot to go to a target position of a second prevention and control target corresponding to the associated prevention and control machine based on a target route according to the third point, the fourth point and the target map, and when the associated prevention and control robot arrives at the target position of the corresponding second prevention and control target, obtain the first prevention and control type of the second prevention and control target and use it as the second prevention and control type;
[0047] The prevention and control reminder unit is used to obtain the preset prevention and control reminder rules corresponding to the second prevention and control type, and remind the corresponding second prevention and control target according to the prevention and control reminder rules.
[0048] Preferably, the associated prevention and control robot determination unit includes:
[0049] A map route acquisition subunit, used for acquiring a map route between the second point and the third point according to the target map;
[0050] A map route feature acquisition subunit, used to acquire map route features;
[0051] The consumption time determination subunit is used to determine the consumption time corresponding to the map route feature according to the map route feature and a preset route feature-consumption time library;
[0052] The associated prevention and control robot determination subunit is used to sum up the consumed time to obtain the time cost, take the second point corresponding to the map route with the smallest time cost as the first target point, take the third point corresponding to the map route where the first target point is located as the second target point, and take the prevention and control robot corresponding to the first target point as the associated prevention and control robot of the second prevention and control target corresponding to the second target point.
[0053] Preferably, the subsequent prevention and control submodule includes:
[0054] A third prevention and control target determination unit, configured to, when the strategy type is a subsequent strategy, use the corresponding first prevention and control target as the third prevention and control target;
[0055] A survey time acquisition unit, used to acquire the survey time of the subsequent survey;
[0056] A first eye-use habit questionnaire obtaining unit, used to obtain a first eye-use habit questionnaire of a third prevention and control target for a preset time length before the survey time;
[0057] An eye habit improvement rate determination unit, used for obtaining an eye habit improvement rate according to a first eye habit questionnaire;
[0058] A current eye habit determination unit, used to obtain a second eye habit questionnaire of the third prevention and control target before the survey time and the most recent survey time, to determine the current eye habit;
[0059] A difference value acquisition unit, used to acquire a difference value between the current eye usage habit and the recommended eye usage habit;
[0060] A first questionnaire cycle determining unit, used for determining a first questionnaire cycle according to the difference value;
[0061] A second questionnaire cycle determining unit, used for determining a second questionnaire cycle according to the first questionnaire cycle and the eye habit improvement rate;
[0062] The questionnaire reminder determination unit is used to provide a questionnaire reminder for the third prevention and control target according to the second questionnaire cycle.
[0063] The myopia prevention and control method based on AI learning provided by the embodiment of the present invention includes:
[0064] Step 1: Determine the prevention and control personnel who need to carry out myopia prevention and control, and obtain the personnel attribute information of the prevention and control personnel;
[0065] Step 2: Based on AI learning technology, train the myopia prevention and control model according to personnel attribute information;
[0066] Step 3: Obtain real-time monitoring information of prevention and control personnel;
[0067] Step 4: Based on the myopia prevention and control model, conduct myopia prevention and control in real time according to real-time monitoring information to obtain prevention and control results.
[0068] The beneficial effects of the present invention are:
[0069] The present invention introduces the personnel attribute information of the prevention and control personnel who need to carry out myopia prevention and control, and based on AI learning technology, trains the myopia prevention and control model according to the personnel attribute information. According to the myopia prevention and control model and real-time monitoring information, the prevention and control results are determined, thereby improving the accuracy of prevention and control and further reducing the probability of myopia.
[0070] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.
[0071] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0072] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:
[0073] Figure 1 Schematic diagram of a myopia prevention and control system based on AI learning in an embodiment of the present invention;
[0074] Figure 2 Schematic diagram of a myopia prevention and control method based on AI learning in an embodiment of the present invention. DETAILED DESCRIPTION
[0075] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.
[0076] The embodiment of the present invention provides a myopia prevention and control system based on AI learning, such as Figure 1 As shown, including:
[0077] The first acquisition subsystem 1 is used to determine the prevention and control personnel who need to carry out myopia prevention and control, and obtain the personnel attribute information of the prevention and control personnel; wherein the prevention and control personnel who need to carry out myopia prevention and control are pre-entered by the myopia prevention and control management personnel; the personnel attribute information is: eye diseases (such as myopia, refractive error, fundus lesions, etc.) or eye-related diseases, etc.;
[0078] Intelligent learning subsystem 2, used to train a myopia prevention and control model based on AI learning technology and personnel attribute information; wherein the myopia prevention and control model is: an AI model used to replace manual myopia prevention and control based on personnel attribute information adaptability;
[0079] The second acquisition subsystem 3 is used to obtain real-time monitoring information of the prevention and control personnel; wherein the real-time monitoring information is obtained based on eye tracking technology or sensors such as cameras;
[0080] The prevention and control result determination subsystem 4 is used to perform myopia prevention and control in real time based on the myopia prevention and control model and real-time monitoring information to obtain the prevention and control results. Among them, myopia prevention and control is to determine whether the eye behavior of the prevention and control personnel is standardized. If not, the corresponding prevention and control personnel are managed; the prevention and control result is the management result after the prevention and control personnel are managed for myopia, and the management result is the eye behavior of the prevention and control personnel after the myopia management.
[0081] The working principle and beneficial effects of the above technical solution are:
[0082] This application introduces the personnel attribute information of the prevention and control personnel who need to carry out myopia prevention and control, and trains a myopia prevention and control model based on the personnel attribute information based on AI learning technology. The myopia prevention and control model can adaptively carry out myopia prevention and control based on the personnel attribute information, and determine the prevention and control results based on the myopia prevention and control model and real-time monitoring information, thereby improving the accuracy of prevention and control and further reducing the probability of myopia.
[0083] In specific applications, the personnel attribute information of the prevention and control personnel who need to carry out myopia prevention and control is determined according to the input of the personnel identity information, and the myopia prevention and control model is trained based on the personnel attribute information. Then, the real-time monitoring information of the prevention and control personnel is obtained based on eye tracking technology, cameras or other sensors to obtain the prevention and control results.
[0084] In one embodiment, the first acquisition subsystem includes:
[0085] An identity ID acquisition module is used to obtain the identity ID of the prevention and control personnel; wherein the identity ID is: face ID and / or pupil ID;
[0086] The personnel attribute information determination module is used to determine the personnel attribute information according to the identity ID and the preset personnel attribute information database, wherein the preset personnel attribute information database is pre-entered with a plurality of one-to-one corresponding identity IDs and personnel attribute information.
[0087] The working principle and beneficial effects of the above technical solution are:
[0088] This application introduces the identity ID of the prevention and control personnel, determines the personnel attribute information based on the identity ID and the personnel attribute information database, and improves the accuracy of the personnel attribute information.
[0089] In one embodiment, the intelligent learning subsystem includes:
[0090] An artificial intelligence learning record acquisition module is used to acquire a first AI learning record; wherein the first AI learning record is: a record of manual myopia prevention and control acquired by big data;
[0091] A second AI learning record determination module is used to determine a second AI learning record in the first AI learning record according to the personnel attribute information; wherein the second AI learning record is: the first AI learning record having similar personnel attribute information of the prevention and control personnel who performs myopia prevention and control in the first AI learning record and the personnel attribute information;
[0092] A comprehensive value analysis module is used to perform a comprehensive training analysis on the second AI learning record to obtain a comprehensive value; wherein the comprehensive training analysis is to determine whether the training data is comprehensive;
[0093] A third AI learning record determination module, configured to use the corresponding second AI learning record as the third AI learning record if the comprehensive value is greater than or equal to a preset comprehensive value threshold; wherein the comprehensive value threshold is manually preset;
[0094] A small sample learning module, for determining a fourth AI learning record based on a small sample learning technique and a corresponding second AI learning record if the overall value is less than a preset overall value threshold; wherein the small sample learning belongs to the scope of the prior art and will not be described in detail;
[0095] The myopia prevention and control model training module is used to train the myopia prevention and control model according to the third AI learning record or the fourth AI learning record.
[0096] The working principle and beneficial effects of the above technical solution are:
[0097] Generally, although big data can obtain enough data, it cannot guarantee that all data is available, especially in situations with more restrictions. For example, people with eye diseases use their eyes for less time than healthy people, but not all patients with eye diseases can find corresponding prevention and control data. Therefore, this application introduces personnel attribute information, determines the prevention and control personnel attribute information of the personnel who perform myopia prevention and control in the first AI learning record and the second AI learning record with similar personnel attribute information, and conducts a comprehensive training analysis on the second AI learning record, thereby avoiding the problem of inaccurate myopia prevention and control model caused by incomplete training data.
[0098] In one embodiment, the full value analysis module includes:
[0099] a marked record collection submodule, for collecting fifth AI learning records; wherein each fifth AI learning record is pre-marked with a comprehensiveness label; wherein the comprehensiveness label is a comprehensive value of the record manually marked in advance;
[0100] A first feature vector determination submodule is used to determine the first data feature based on a preset data feature extraction template and the fifth AI learning record, and to express the first data feature as a first feature vector; wherein the data feature extraction template constrains only the extraction of data features, and the data features are, for example, text features, semantic features, and structural features; and the construction rules of the first feature vector are manually preset;
[0101] A deep learning model acquisition submodule is used to acquire a preset deep learning model;
[0102] A comprehensiveness analysis model determination submodule, used to determine a comprehensiveness analysis model based on a deep learning model according to the first feature vector and the corresponding comprehensiveness label;
[0103] A second feature vector determination submodule is used to extract a second data feature of the second AI learning record based on the data feature extraction template, and express the second data feature as a second feature vector; the second data feature is, for example, a text feature, a semantic feature, and a structural feature; the construction rule of the second feature vector is the same as the first feature vector;
[0104] The comprehensive value output submodule is used to input the second eigenvector into the comprehensiveness analysis model to obtain the comprehensive value.
[0105] The working principle and beneficial effects of the above technical solution are:
[0106] This application introduces a fifth AI learning record that is pre-labeled with a comprehensiveness label, introduces a data feature extraction template to determine the first data feature and construct a first feature vector, introduces a deep learning model to train a comprehensiveness analysis model, obtains the second feature vector and inputs it into the comprehensiveness analysis model to obtain a comprehensive value, and the determination of the comprehensive value is more appropriate.
[0107] In one embodiment, the prevention and control result determination subsystem includes:
[0108] An information sorter acquisition module is used to acquire an information sorter preset in the front end of the myopia prevention and control model; wherein the front end of the myopia prevention and control model is: the first model level of the myopia prevention and control model; the information sorter is: an intelligent model for classifying real-time monitoring information based on the type of monitoring information, and the type of monitoring information is, for example: related monitoring information for determining the duration of eye use, and for example: related monitoring information for determining the distance of eye use;
[0109] The comparison result determination module is used to input the real-time monitoring information into the information sorter, and determine that the sorting unit of the information sorter corresponds to the comparison result of the preset comparator; wherein the sorting unit is: a storage space preset for storing real-time monitoring information of a certain type of monitoring information; wherein the comparison result is: matching and non-matching;
[0110] The characteristic information determination module is used for obtaining the characteristic information of the sorting unit whose comparison result is not matched if the comparison result does not match; wherein the characteristic information is: the identity information of the monitoring personnel corresponding to the sorting unit and the real-time monitoring information corresponding to the sorting unit that does not match;
[0111] The myopia prevention and control module is used to prevent and control myopia based on feature information.
[0112] The working principle and beneficial effects of the above technical solution are:
[0113] This application introduces an information sorter to sort the real-time monitoring information, obtains the comparison results of the sorting units of the comparator, and when there is a match mismatch in the comparison results, outputs the characteristic information of the corresponding sorting unit, and performs myopia prevention and control based on the characteristic information, thereby further improving the accuracy of myopia prevention and control.
[0114] In one embodiment, the myopia prevention and control module includes:
[0115] The feature information parsing submodule is used to parse the feature information and obtain the first prevention and control target and the first prevention and control type; wherein the first prevention and control target is: personnel with irregular eye behavior among the prevention and control personnel; the first prevention and control type is: eye use time control, for example: eye use distance control, and the first prevention and control type is determined by the comparison result matching the non-compliant monitoring information type;
[0116] The prevention and control strategy determination submodule is used to determine the prevention and control strategy according to the first prevention and control type; wherein the prevention and control strategy is: how to prevent and control myopia for the first prevention and control target;
[0117] The strategy type acquisition submodule is used to obtain the strategy type of the prevention and control strategy; the strategy types include: immediate strategy and subsequent strategy;
[0118] The real-time control submodule is used to determine the target position of the first control target when the strategy type is the real-time strategy, and to dispatch the control robot to perform corresponding control according to the target position; wherein the target position is: the position of the first control target, for example: the relative position of the first control target in a behavior correction classroom;
[0119] The subsequent prevention and control submodule is used to carry out corresponding prevention and control of the first prevention and control target based on the questionnaire reminder rule when the strategy type is the subsequent strategy. Among them, the questionnaire reminder rule is: what kind of questionnaire cycle is used to determine the eye habits of the prevention and control personnel to urge the first prevention and control target to improve their eye habits.
[0120] The working principle and beneficial effects of the above technical solution are:
[0121] This application introduces the first prevention and control type, determines the prevention and control strategy, and conducts myopia prevention and control according to the strategy type of the prevention and control strategy, thereby improving the comprehensiveness of prevention and control.
[0122] In one embodiment, the real-time prevention and control submodule includes:
[0123] A target map acquisition unit is used to acquire a target map where the first prevention and control target is located; wherein the target map is: a schematic diagram of the first prevention and control target in a behavior correction classroom;
[0124] A point determination unit, used to determine that the target position corresponds to a first point on the target map, and at the same time, determine that the idle control robot corresponds to a second point on the target map;
[0125] The first control target traversal unit is used to traverse the first control target in sequence, take the first control target currently being traversed as the second control target, determine the first point corresponding to the second control target, and use it as the third point;
[0126] The associated prevention and control robot determining unit is used to determine the associated prevention and control robot corresponding to the second prevention and control target according to the second point, the third point and the target map; wherein, when determining the associated prevention and control robot, if the prevention and control robot can reach the vicinity of the second prevention and control target the fastest, the corresponding prevention and control robot is the associated prevention and control robot corresponding to the second prevention and control target;
[0127] A fourth point determination unit, used to determine a fourth point of the associated control robot;
[0128] A second prevention and control type determination unit is used to control the associated prevention and control robot to go to a target position of a second prevention and control target corresponding to the associated prevention and control machine based on a target route according to the third point, the fourth point and the target map, and when the associated prevention and control robot arrives at the target position of the corresponding second prevention and control target, obtain the first prevention and control type of the second prevention and control target and use it as the second prevention and control type;
[0129] The prevention and control reminder unit is used to obtain the preset prevention and control reminder rules corresponding to the second prevention and control type, and remind the corresponding second prevention and control target according to the prevention and control reminder rules. Among them, the prevention and control reminder rules are: how to prevent and control reminders first, and then how to prevent and control reminders, for example: first display a reminder slogan on the preset display device on the associated prevention and control robot, and then use the robot arm to perform physical intervention, and the physical intervention is, for example: straighten the sitting posture.
[0130] The working principle and beneficial effects of the above technical solution are:
[0131] When the prevention and control robot is used for myopia prevention and control intelligently, it is necessary to consider reasonable scheduling based on the relative positions of the target prevention and control personnel and the prevention and control robot in the vision correction classroom. Therefore, this application introduces a target map, determines the first point of the target position and the second point of the prevention and control robot on the target map, determines the second prevention and control target and the associated prevention and control robot corresponding to the second prevention and control target based on the first point and the second point, controls the associated prevention and control robot to the target position of the second prevention and control target, and then reminds the second prevention and control target based on the prevention and control reminder rules and the second prevention and control type, and promptly determines the associated prevention and control robot that can arrive the fastest, thereby improving the rationality of scheduling and making prevention and control more timely.
[0132] In one embodiment, the associated prevention and control robot determination unit includes:
[0133] A map route acquisition subunit, used for acquiring a map route between the second point and the third point according to the target map;
[0134] The map route feature acquisition subunit is used to acquire map route features; wherein the map route features are: what type of map route is what length;
[0135] The consumption time determination subunit is used to determine the consumption time corresponding to the map route feature according to the map route feature and the preset route feature-consumption time library; wherein the consumption time is the time required for the prevention and control robot to travel in one or more sections of the map route corresponding to the map route feature;
[0136] The associated control robot determination subunit is used to sum up the consumed time to obtain the time cost, and use the second point corresponding to the map route with the smallest time cost as the first target point, the third point corresponding to the map route where the first target point is located as the second target point, and the control robot corresponding to the first target point as the associated control robot of the second control target corresponding to the second target point. The time cost is, for example, 6s.
[0137] The working principle and beneficial effects of the above technical solution are:
[0138] This application introduces map route features and line features-consumption time library, determines the consumption time, sums and calculates the time cost of the map line, determines the associated prevention and control robot with the smallest time cost, improves the timeliness of the scheduling of the associated prevention and control robots, and improves the prevention and control efficiency.
[0139] In one embodiment, the subsequent prevention and control submodule includes:
[0140] A third prevention and control target determination unit, configured to, when the strategy type is a subsequent strategy, use the corresponding first prevention and control target as the third prevention and control target;
[0141] A survey time acquisition unit is used to acquire the survey time of a subsequent survey; wherein the subsequent survey is: a subsequent survey of the myopia habit tracking of the first prevention and control target; the survey time is: the time of the subsequent survey of the eye habits of the third prevention and control target;
[0142] A first eye-use habit questionnaire obtaining unit is used to obtain a first eye-use habit questionnaire of a third prevention and control target for a preset time length before the survey time; wherein the preset time length is, for example, one month;
[0143] An eye habit improvement rate determination unit is configured to obtain an eye habit improvement rate according to a first eye habit questionnaire; wherein, when obtaining the eye habit improvement rate according to the first eye habit questionnaire, the eye habit of the third prevention and control target is evaluated according to the first eye habit questionnaire to obtain an evaluation score, and the change in the evaluation score each time is divided by the time interval between each scoring to obtain the eye habit improvement rate;
[0144] A current eye habit determination unit, used to obtain a second eye habit questionnaire of the third prevention and control target before the survey time and the most recent survey time, to determine the current eye habit;
[0145] A difference value acquisition unit, used to acquire a difference value between the current eye usage habit and the recommended eye usage habit; wherein the difference value is the degree of dissimilarity between the current eye usage habit and the recommended eye usage habit;
[0146] A first questionnaire cycle determination unit is used to determine the first questionnaire cycle according to the difference value; wherein the first questionnaire cycle can be determined by matching data based on a preset difference value-questionnaire cycle library and the difference value, and the larger the difference value, the shorter the first questionnaire cycle;
[0147] A second questionnaire cycle determination unit is used to determine a second questionnaire cycle according to the first questionnaire cycle and the eye habit improvement rate; wherein each eye habit improvement rate corresponds to a fusion coefficient, and the corresponding relationship between the eye habit improvement rate and the fusion coefficient is manually preset; the greater the eye habit improvement rate, the higher the fusion coefficient, and the first questionnaire cycle and the fusion coefficient are correspondingly multiplied to obtain the second questionnaire cycle;
[0148] The questionnaire reminder determination unit is used to provide a questionnaire reminder for the third prevention and control target according to the second questionnaire cycle.
[0149] The working principle and beneficial effects of the above technical solution are:
[0150] This application determines the rate of improvement of eye habits based on the first eye habits questionnaire obtained for the length of time before the survey time of the subsequent survey of the third prevention and control target. The current eye habits are determined based on the second eye habits questionnaire, and the difference between the current eye habits and the recommended eye habits is obtained. The first questionnaire cycle is determined based on the difference value, and the second questionnaire cycle is determined based on the first questionnaire cycle and the rate of improvement of eye habits. In the subsequent myopia prevention and control tasks, questionnaire reminders are conducted for the third prevention and control target based on different performances, and the reminder process is more humane and more appropriate.
[0151] The embodiment of the present invention provides a myopia prevention and control method based on AI learning, which is characterized by comprising:
[0152] Step 1: Determine the prevention and control personnel who need to carry out myopia prevention and control, and obtain the personnel attribute information of the prevention and control personnel;
[0153] Step 2: Based on AI learning technology, train the myopia prevention and control model according to personnel attribute information;
[0154] Step 3: Obtain real-time monitoring information of prevention and control personnel;
[0155] Step 4: Based on the myopia prevention and control model, conduct myopia prevention and control in real time according to real-time monitoring information to obtain prevention and control results.
[0156] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. The myopia prevention and control system based on AI learning is characterized by: include: The first acquisition subsystem is used to determine the prevention and control personnel who need to carry out myopia prevention and control, and obtain the personnel attribute information of the prevention and control personnel; Intelligent learning subsystem, used to train myopia prevention and control models based on AI learning technology and personnel attribute information; The second acquisition subsystem is used to obtain real-time monitoring information of prevention and control personnel; Real-time monitoring information is obtained based on eye tracking technology or cameras; The prevention and control result determination subsystem is used to carry out myopia prevention and control in real time based on the myopia prevention and control model and real-time monitoring information to obtain the prevention and control results; The prevention and control result determination subsystem includes: An information sorter acquisition module is used to acquire an information sorter preset at the front end of the myopia prevention and control model; A comparison result determination module is used to input the real-time monitoring information into the information sorter and determine the comparison result of the preset comparator corresponding to the sorting unit of the information sorter; A feature information determination module, for obtaining, if there is a mismatch in the comparison result, the feature information of the sorting unit whose output comparison result is a mismatch; Myopia prevention and control module, used to prevent and control myopia based on feature information; Myopia prevention and control module, including: A feature information parsing submodule, used to parse the feature information and obtain the first prevention and control target and the first prevention and control type; A prevention and control strategy determination submodule, used to determine a prevention and control strategy according to the first prevention and control type; The strategy type acquisition submodule is used to obtain the strategy type of the prevention and control strategy; the strategy types include: immediate strategy and subsequent strategy; The real-time prevention and control submodule is used to determine the target position of the first prevention and control target when the strategy type is the real-time strategy, and to dispatch the prevention and control robot to perform corresponding prevention and control according to the target position; The subsequent prevention and control submodule is used to carry out corresponding prevention and control of the first prevention and control target based on the questionnaire reminder rules when the strategy type is a subsequent strategy; Real-time prevention and control submodule, including: A target map acquisition unit, used to acquire a target map where the first prevention and control target is located; A point determination unit, used to determine that the target position corresponds to a first point on the target map, and at the same time, determine that the idle control robot corresponds to a second point on the target map; The first control target traversal unit is used to traverse the first control target in sequence, take the first control target currently being traversed as the second control target, determine the first point corresponding to the second control target, and use it as the third point; An associated prevention and control robot determining unit, used to determine an associated prevention and control robot corresponding to the second prevention and control target according to the second point, the third point and the target map; A fourth point determination unit, used to determine a fourth point of the associated control robot; A second prevention and control type determination unit is used to control the associated prevention and control robot to go to a target position of a second prevention and control target corresponding to the associated prevention and control machine based on a target route according to the third point, the fourth point and the target map, and when the associated prevention and control robot arrives at the target position of the corresponding second prevention and control target, obtain the first prevention and control type of the second prevention and control target and use it as the second prevention and control type; A prevention and control reminder unit, used to obtain a preset prevention and control reminder rule corresponding to the second prevention and control type, and remind the corresponding second prevention and control target according to the prevention and control reminder rule; The associated prevention and control robot determination unit includes: A map route acquisition subunit, used for acquiring a map route between the second point and the third point according to the target map; A map route feature acquisition subunit, used to acquire map route features; The consumption time determination subunit is used to determine the consumption time corresponding to the map route feature according to the map route feature and a preset route feature-consumption time library; The associated prevention and control robot determination subunit is used to sum up the consumed time to obtain the time cost, take the second point corresponding to the map route with the smallest time cost as the first target point, take the third point corresponding to the map route where the first target point is located as the second target point, and take the prevention and control robot corresponding to the first target point as the associated prevention and control robot of the second prevention and control target corresponding to the second target point; Subsequent prevention and control submodules include: A third prevention and control target determination unit, configured to, when the strategy type is a subsequent strategy, use the corresponding first prevention and control target as the third prevention and control target; A survey time acquisition unit, used to acquire the survey time of the subsequent survey; A first eye-use habit questionnaire obtaining unit, used to obtain a first eye-use habit questionnaire of a third prevention and control target for a preset time length before the survey time; An eye habit improvement rate determination unit, used for obtaining an eye habit improvement rate according to a first eye habit questionnaire; A current eye habit determination unit, used to obtain a second eye habit questionnaire of the third prevention and control target before the survey time and the most recent survey time, to determine the current eye habit; A difference value acquisition unit, used to acquire a difference value between the current eye usage habit and the recommended eye usage habit; A first questionnaire cycle determining unit, used for determining a first questionnaire cycle according to the difference value; A second questionnaire cycle determining unit, used for determining a second questionnaire cycle according to the first questionnaire cycle and the eye habit improvement rate; The questionnaire reminder determination unit is used to provide a questionnaire reminder for the third prevention and control target according to the second questionnaire cycle.
2. The myopia prevention and control system based on AI learning as claimed in claim 1, characterized in that: The first acquisition subsystem includes: Identity ID acquisition module, used to obtain the identity ID of the prevention and control personnel; The personnel attribute information determination module is used to determine the personnel attribute information according to the identity ID and the preset personnel attribute information database.
3. The myopia prevention and control system based on AI learning as claimed in claim 1, characterized in that: Intelligent learning subsystem, including: An artificial intelligence learning record acquisition module, used to acquire a first AI learning record; A second AI learning record determination module, used to determine the second AI learning record in the first AI learning record according to the personnel attribute information; A comprehensive value analysis module is used to conduct a comprehensive analysis of the training of the second AI learning record to obtain a comprehensive value; A third AI learning record determination module, configured to use the corresponding second AI learning record as the third AI learning record if the comprehensive value is greater than or equal to a preset comprehensive value threshold; A small sample learning module, configured to determine a fourth AI learning record based on the corresponding second AI learning record based on a small sample learning technology if the full value is less than a preset full value threshold; The myopia prevention and control model training module is used to train the myopia prevention and control model according to the third AI learning record or the fourth AI learning record.
4. The myopia prevention and control system based on AI learning as claimed in claim 3, characterized in that: Comprehensive value analysis module, including: a marked record collection submodule, used to collect the fifth AI learning record; wherein each fifth AI learning record is pre-marked with a comprehensive label; A first feature vector determination submodule, configured to determine a first data feature based on a preset data feature extraction template and according to a fifth AI learning record, and to represent the first data feature as a first feature vector; A deep learning model acquisition submodule is used to acquire a preset deep learning model; A comprehensiveness analysis model determination submodule, used to determine a comprehensiveness analysis model based on a deep learning model according to the first feature vector and the corresponding comprehensiveness label; A second feature vector determination submodule, configured to extract a second data feature of the second AI learning record based on the data feature extraction template, and represent the second data feature as a second feature vector; The comprehensive value output submodule is used to input the second eigenvector into the comprehensiveness analysis model to obtain the comprehensive value.
5. A myopia prevention and control method based on AI learning, characterized in that: include: Step 1: Determine the prevention and control personnel who need to carry out myopia prevention and control, and obtain the personnel attribute information of the prevention and control personnel; Step 2: Based on AI learning technology, train the myopia prevention and control model according to personnel attribute information; Step 3: Obtain real-time monitoring information of prevention and control personnel; real-time monitoring information is obtained based on eye tracking technology or cameras; Step 4: Based on the myopia prevention and control model, carry out myopia prevention and control in real time according to real-time monitoring information to obtain prevention and control results; Step 4: Based on the myopia prevention and control model, carry out myopia prevention and control in real time according to real-time monitoring information, and obtain prevention and control results, including: Obtain an information sorter preset in the front end of the myopia prevention and control model; Input the real-time monitoring information into the information sorter, and determine the comparison result of the preset comparator corresponding to the sorting unit of the information sorter; If there is a mismatch in the comparison result, the comparison result obtained as an output is the feature information of the mismatching sorting unit; Carry out myopia prevention and control based on characteristic information; Based on the characteristic information, myopia prevention and control is carried out, including: Analyze the characteristic information to obtain the first prevention and control target and the first prevention and control type; Determine the prevention and control strategy based on the first prevention and control type; Obtain the strategy type of the prevention and control strategy; the strategy type includes: immediate strategy and subsequent strategy; When the strategy type is an immediate strategy, the target position of the first control target is determined, and the control robot is dispatched immediately to perform corresponding control according to the target position; When the strategy type is a follow-up strategy, the first prevention and control target is controlled accordingly based on the questionnaire reminder rules; When the strategy type is an immediate strategy, the target position of the first control target is determined, and the control robot is dispatched immediately to perform corresponding control according to the target position, including: Obtain the target map where the first prevention and control target is located; Determine that the target position corresponds to a first point on the target map, and at the same time, determine that the idle control robot corresponds to a second point on the target map; Traverse the first control target in sequence, take the first control target currently being traversed as the second control target, determine the first point corresponding to the second control target, and use it as the third point; Determine the associated prevention and control robot corresponding to the second prevention and control target according to the second point, the third point and the target map; Determine the fourth point of the associated prevention and control robot; According to the third point, the fourth point and the target map, the associated prevention and control robot is controlled to go to the target position of the second prevention and control target corresponding to the associated prevention and control machine based on the target route. When the associated prevention and control robot reaches the target position of the corresponding second prevention and control target, the first prevention and control type of the second prevention and control target is obtained and used as the second prevention and control type. Obtain the preset prevention and control reminder rules corresponding to the second prevention and control type, and remind the corresponding second prevention and control target according to the prevention and control reminder rules; According to the second point, the third point and the target map, the associated control robot corresponding to the second control target is determined, including: According to the target map, a map route between the second point and the third point is obtained; Get map route features; Determine the consumption time corresponding to the map route feature according to the map route feature and the preset route feature-consumption time library; The consumed time is summed up to obtain the time cost, and the second point corresponding to the map route with the smallest time cost is used as the first target point, the third point corresponding to the map route where the first target point is located is used as the second target point, and the prevention and control robot corresponding to the first target point is used as the associated prevention and control robot of the second prevention and control target corresponding to the second target point; When the strategy type is a follow-up strategy, based on the questionnaire reminder rules, corresponding prevention and control measures are taken for the first prevention and control target, including: When the strategy type is a follow-up strategy, the corresponding first prevention and control target will be used as the third prevention and control target; Get survey moments for follow-up surveys; Obtaining the first eye-use habit questionnaire of the third prevention and control target for a preset time period before the survey time; According to the first eye habit questionnaire, obtain the rate of improvement of eye habits; Obtain the second eye-use habit questionnaire of the third prevention and control target before the survey time and the most recent survey time to determine the current eye-use habit; Get the difference between current eye habits and recommended eye habits; According to the difference value, the first questionnaire cycle is determined; Determine the second questionnaire cycle based on the first questionnaire cycle and the rate of improvement of eye habits; Based on the second questionnaire cycle, questionnaire reminders are conducted for the third prevention and control target.
