AI-driven omnibearing posture monitoring method and system, medium and program product

Abstract

The invention provides an AI-driven omni-directional posture monitoring method and system, a medium and a program product, and relates to the technical field of posture monitoring, and the method comprises the steps: firstly, obtaining individual action data from an edge device, and enabling the data to be collected by optical, infrared and acoustic sensors through the fusion of a multi-mode perception technology; then the data are input into a posture determination model, the model is formed by training a posture sample set containing individual actions through a deep neural network and used for determining posture information, and if the posture information is within a standard posture range, monitoring is maintained; if the posture information exceeds the range, the historical posture information of the user is obtained, a personalized correction plan is made according to the historical posture information, and the plan is sent to the mobile terminal to be displayed, so that accurate posture monitoring and correction services can be provided for the user.

Description

Technical Field

[0001] The present application relates to the technical field of posture monitoring, and in particular to an AI-driven all-round posture monitoring method, system, medium and program product. Background Art

[0002] In today's digital age, people are paying more and more attention to their own health and body management. Whether in sports training, rehabilitation treatment, daily office work or daily fitness activities, it is important to accurately understand one's own body condition. Good body posture is not only related to image, but also closely related to physical health. Long-term accumulation of bad body posture may cause various health problems.

[0003] At present, there are two common technical means in the field of posture monitoring. One is to rely on contact sensors for posture monitoring. Such sensors collect data by directly contacting the human body. For example, pressure sensors and acceleration sensors are worn on specific parts of the body. They are used to collect data such as pressure changes and acceleration changes generated when the human body moves, and then analyze the posture information. The other is posture monitoring based on video image recognition technology, which uses a camera to shoot human movement videos, and then uses image analysis algorithms to identify and analyze the human posture in the video to determine the posture.

[0004] However, since contact sensors need to be in direct contact with the human body, this greatly limits the user's freedom of movement. Wearing them for too long will make the user feel uncomfortable, reducing the convenience of use. Image analysis algorithms are limited by complex visual environments and computing resource requirements, so current technology often makes it difficult to achieve non-intrusive, high-precision posture detection and correction in different scenarios. Summary of the invention

[0005] The present application provides an AI-driven all-round posture monitoring method, system, medium and program product for realizing accurate monitoring and correction of user posture.

[0006] In the first aspect, the present application provides an AI-driven all-round posture monitoring method, which is applied to a posture monitoring system, the method comprising: obtaining individual motion data from an edge device, the individual motion data being collected from an optical sensor, an infrared sensor, and an acoustic sensor using a fusion multimodal sensing technology; inputting the individual motion data into a posture determination model to determine the posture information corresponding to the individual motion data, the posture determination model being trained in advance by a deep neural network based on multiple labeled posture sample sets containing individual motions; if the posture information is monitored to be within a preset standard posture range, maintaining the current monitoring state; if the posture information is monitored to be beyond the standard posture range, obtaining multiple historical posture information of the user; after determining a personalized correction plan based on the historical posture information, sending the personalized correction plan to a mobile terminal for display.

[0007] By adopting the above technical solution, the AI-driven all-round posture monitoring method obtains individual motion data collected by fusion multimodal sensing technology from edge devices, and inputs it into the posture determination model to obtain posture information. This multimodal sensing technology can comprehensively and accurately capture individual movements, and the posture determination model trained by deep neural network can accurately analyze the posture. When the posture information exceeds the standard range, the historical posture information is used to formulate a personalized correction plan. This can not only provide users with personalized posture monitoring and correction services, but also make precise adjustments based on individual differences, thereby improving the accuracy and effectiveness of posture monitoring, meeting the diverse needs of users, and helping to improve the user's physical health level.

[0008] In combination with some embodiments of the first aspect, in some embodiments, before the step of obtaining individual motion data from the edge device, it also includes: selecting corresponding monitoring strategies from a pre-set strategy library according to different application scenarios and user needs set by the user, and the strategy library is a set of strategies that configure the accuracy, frequency and feedback method of body monitoring according to different environments and different user groups; adjusting various parameters of the edge device according to the monitoring strategy.

[0009] By adopting the above technical solution, before obtaining individual motion data, the monitoring strategy is selected from the strategy library and the edge device parameters are adjusted according to the application scenarios and needs set by the user. This process takes into account the needs of different environments and user groups, and can configure the accuracy, frequency and feedback method of posture monitoring in a targeted manner. Through the pre-set strategy library and flexible adjustment, the monitoring system can better adapt to different environments and users, ensure the quality and monitoring effect of monitoring data, improve the adaptability and practicality of the system, and provide a good foundation for subsequent posture monitoring.

[0010] In combination with some embodiments of the first aspect, in some embodiments, after the step of obtaining individual motion data from the edge device, it also includes: when receiving an insufficient illumination warning from the optical sensor, automatically adjusting the parameters of the optical camera of the edge device; after obtaining the user's individual motion image through the edge device, enabling the image enhancement algorithm to enhance the brightness of the preset dark area in the individual motion image to obtain the motion outline under dim conditions.

[0011] By adopting the above technical solution, clear individual motion images can still be obtained under dim conditions to ensure the clear presentation of motion contours. This function overcomes the impact of insufficient light on monitoring, enhances the system's adaptability in different lighting environments, and avoids data loss or inaccuracy due to lighting problems.

[0012] In combination with some embodiments of the first aspect, in some embodiments, after the step of inputting the individual motion data into the posture determination model to determine the posture information corresponding to the individual motion data, it also includes: determining the user's posture deviation degree based on the posture information and the standard posture range; determining a corresponding reminder strategy based on the posture deviation degree, the reminder strategy including when the posture deviation degree is within a preset mild deviation range, generating a preset mild vibration reminder through the smart wearable device; when the posture deviation degree is within a preset moderate deviation range, generating a preset medium intensity vibration reminder through the smart wearable device; when the posture deviation degree is within a preset severe deviation range, generating a preset strong vibration reminder through the smart wearable device, and simultaneously sending a warning message to the mobile terminal.

[0013] By adopting the above technical solution, through the vibration of smart wearable devices and the warning of mobile terminals, users can timely perceive their own posture problems, which can effectively draw users' attention to posture problems, guide users to adjust their posture in time, and avoid health risks that may be caused by posture deviations.

[0014] In combination with some embodiments of the first aspect, in some embodiments, after the step of determining the corresponding reminder strategy based on the degree of posture deviation, it also includes: establishing a user behavior record based on the user's response to the reminder strategy, the user behavior record including the posture adjustment time after the user receives the reminder, the improvement effect of the user after receiving the reminder, and evaluating the user's posture improvement trend, the posture adjustment time is determined based on the time interval required for the user to receive the reminder and the posture data to return to the standard posture range, the improvement effect is determined based on the difference between the posture parameters of the user after receiving the reminder strategy and the standard posture parameters, and the posture improvement trend is determined based on the statistical analysis results of the user's posture parameters within a continuous set time; dynamically adjust the reminder strategy based on the user behavior record; after receiving the posture information viewing instruction sent by the mobile terminal, generate a user posture improvement report based on the user behavior record and send it to the mobile terminal for display.

[0015] By adopting the above technical solution, by recording the user's posture adjustment time, improvement effect and improvement trend, the system can accurately grasp the user's feedback on the reminder and the posture improvement situation. This can not only provide the user with detailed feedback on the posture improvement process, but also dynamically adjust the reminder strategy, so that the user can obtain guidance that is more suitable for themselves, thereby improving the user's posture adjustment efficiency and the final effect.

[0016] In combination with some embodiments of the first aspect, in some embodiments, after the step of establishing a user behavior record based on the user's response to the reminder strategy, it also includes: evaluating the user's adaptation degree to the personalized correction plan based on the user behavior record; and dynamically adjusting the personalized correction plan based on the adaptation degree.

[0017] By adopting the above technical solutions, this dynamic adjustment can continuously optimize the correction plan, avoid poor user experience or poor correction effect due to plan inadaptability, and improve the success rate of personalized correction plans and user acceptance.

[0018] In combination with some embodiments of the first aspect, in some embodiments, before the step of obtaining individual motion data from the edge device, it also includes: performing multiple identity authentications on the target user; when the multiple identity authentications are passed, determining that the target user is an object to be monitored.

[0019] By adopting the above technical solution and through multiple identity authentication, it can be ensured that only the target users who have passed the authentication will be confirmed as the objects to be monitored, thus ensuring the legality and security of the use of the body monitoring system, preventing unauthorized users from using it, and avoiding the risk of data leakage.

[0020] In a second aspect, the present application provides a posture monitoring system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the posture monitoring system to perform the method described in the first aspect and any possible implementation method of the first aspect.

[0021] In a third aspect, the present application provides a computer-readable storage medium comprising instructions, which, when executed on a posture monitoring system, enables the posture monitoring system to execute the method described in the first aspect and any possible implementation of the first aspect.

[0022] In a fourth aspect, the present application provides a computer program product. When the computer program product is run on a body posture monitoring system, the body posture monitoring system executes the method described in the first aspect and any possible implementation manner of the first aspect.

[0023] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. Due to the use of fusion multimodal sensing technology to collect individual motion data, the use of deep neural networks to train posture determination models, and the technical means of formulating personalized correction plans based on historical posture information, the problem that existing posture monitoring technologies are difficult to provide personalized monitoring and correction services and the monitoring accuracy is not high is effectively solved. It then realizes the technical effect of providing users with accurate and personalized posture monitoring and correction services, improving the accuracy and effectiveness of posture monitoring, and helping to improve the user's physical health level.

[0024] 2. Since the technical means of determining the degree of posture deviation according to posture information and standard posture range and setting vibration reminders of different intensities and reminder strategies for sending warning information to mobile terminals are adopted, the problem that the existing posture monitoring technology cannot provide differentiated reminders according to the degree of deviation, resulting in users being insensitive to posture problems, is effectively solved. In addition, the technical effect of timely feedback of posture problems to users based on different degrees of deviation is achieved, guiding users to adjust their posture, improving the timeliness and effectiveness of posture adjustment, and reducing health risks caused by posture deviation is achieved.

[0025] 3. The invention adopts the technical means of establishing user behavior records according to the user's response to the reminder strategy, and dynamically adjusting the reminder strategy and generating the posture improvement report accordingly. Therefore, the problem that the existing posture monitoring technology is difficult to adjust the monitoring strategy according to the user's behavior feedback, resulting in poor effect of user posture adjustment is effectively solved. Then, the invention realizes the technical effect of dynamically optimizing the reminder strategy through detailed recording and feedback of the user's posture adjustment process, and improving the efficiency and final effect of the user's posture adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of the AI-driven all-round posture monitoring method in an embodiment of the present application; Figure 2 This is another flowchart of the AI-driven all-round posture monitoring method in an embodiment of the present application; Figure 3 It is a schematic diagram of the structure of a physical device of the posture monitoring system in the embodiment of the present application. DETAILED DESCRIPTION

[0027] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification and appended claims of the present application, the singular expressions "one", "a kind of", "said", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to and includes any or all possible combinations of one or more listed items.

[0028] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more.

[0029] For ease of understanding, the following is a description of the process of the method provided by this implementation. Figure 1 , which is a flow chart of the AI-driven all-round posture monitoring method in an embodiment of the present application.

[0030] S101, obtaining individual motion data from an edge device, where the individual motion data is collected from an optical sensor, an infrared sensor, and an acoustic sensor using a fusion multimodal sensing technology; The posture monitoring system first establishes a stable communication connection with various edge devices. These edge devices are equipped with optical sensors, infrared sensors and acoustic sensors, which are distributed in the monitoring area to form a multi-sensor collaborative collection network.

[0031] In the data collection process, the fusion of multimodal sensing technology plays a key role. The optical sensor is responsible for collecting individual visual image data and recording information such as the body's posture, limb position, and movement trajectory. In order to improve the quality and efficiency of image acquisition, the system uses high-speed image acquisition technology, which can capture the moment of human movement at a high frame rate to ensure that key movement details are not missed. At the same time, image anti-shake and autofocus technology are used to overcome the problem of image blur caused by equipment shaking or individual movement. Infrared sensors obtain thermal imaging data of the human body by detecting infrared rays emitted by the human body, thereby sensing the outline, position, and tiny changes in movement of the human body. The posture monitoring system uses the thermal imaging characteristics of infrared sensors, combined with advanced image processing algorithms, to accurately identify human targets in complex background environments and track their movements.

[0032] In some embodiments, during posture monitoring, lighting conditions will have a significant impact on the quality of individual motion data obtained from edge devices. Therefore, the embodiments of the present application adopt a series of technical means to ensure that clear and usable image data can be obtained even in dim environments, as follows: When the optical sensor detects insufficient illumination and issues a warning, the posture monitoring system will respond quickly and automatically adjust the optical camera parameters of the edge device. The aperture size of the optical camera determines the amount of light entering the lens. When the illumination is insufficient, the system will automatically increase the aperture to allow more light to enter and increase the overall brightness of the image. At the same time, the system will adjust the shutter speed and appropriately extend the exposure time to give the sensor more time to capture light. After obtaining the user's individual motion image through the edge device, the posture monitoring system will enable advanced image enhancement algorithms to enhance the brightness of preset dark areas in the image. The image enhancement algorithm based on deep learning will first perform feature extraction on the input individual motion image to identify the location and features of the dark area. Then, using the principle of Generative Adversarial Network (GAN), the generator generates an enhanced image, and the discriminator determines the difference between the generated image and the real clear image. Through adversarial training, the generator is continuously optimized so that the generated image can maintain the details and authenticity of the image while improving the brightness of the dark area. In this way, even in dim conditions, the action contours of the user's individual action images can be clearly obtained.

[0033] Acoustic sensors are used to collect sound signals in the environment, such as footsteps, joint movement sounds, etc. These sound information can help determine the individual's movement status and rhythm. In order to accurately capture these sound signals, the system uses high-sensitivity, wide-bandwidth microphones in the selection of acoustic sensors, and uses noise reduction technology and sound enhancement algorithms to remove environmental noise interference and highlight the sound characteristics related to human movements. The posture monitoring system uses time synchronization technology to ensure that the data collected from different sensors are consistent in time. Through a precise clock synchronization mechanism, each sensor data is accurately timestamped, so that subsequent data fusion and analysis can be performed on a unified time basis.

[0034] In some embodiments, in the posture monitoring system, considering the different application scenarios and needs of different users, as well as the impact of different environments on monitoring, more accurate and efficient posture monitoring is achieved by selecting monitoring strategies from the strategy library and adjusting the edge device parameters. Specifically, a strategy library is set in advance. The strategy library is a set of strategies that configure the accuracy, frequency and feedback mode of posture monitoring according to different environments and user groups. Different environmental factors, such as indoor and outdoor scenes, lighting conditions, space size, etc., will affect the performance of the monitoring device and the data collection effect; different user groups, such as athletes, office workers, and the elderly, have different needs for posture monitoring. For athletes, they need high-precision and high-frequency posture monitoring during training, so as to correct their movements in time, improve training effects, and reduce the risk of injury. High-precision sensor parameters will be configured in the strategy library, such as using more accurate optical sensors to capture subtle changes in movements, and the monitoring frequency may be set to collect data multiple times per second. The feedback method focuses more on real-time feedback, such as real-time vibration reminders through smart wearable devices, and displaying detailed posture data and analysis results on the training equipment.

[0035] When using the posture monitoring system, users will set it according to their own application scenarios and needs. If the user is doing high-intensity sports training, such as professional athletes doing special training, they will choose a strategy with high precision, high frequency monitoring and real-time feedback. After receiving the user's settings, the system will quickly match and select the corresponding monitoring strategy from the strategy library. The system will search and filter in the strategy library based on the scene keywords entered by the user (such as "sports training", "office", etc.), user type (such as "athlete", "office worker", etc.), and specific requirements for accuracy, frequency and feedback methods. : After selecting the monitoring strategy, the system will adjust the various parameters of the edge device according to the strategy. This can enhance the adaptability of the posture monitoring system, improve data quality and monitoring effects, optimize user experience, and achieve accurate posture monitoring and personalized services.

[0036] In some embodiments, multiple identity authentication can be performed on the target user in the posture monitoring system. The posture monitoring system involves the user's personal health data, which has high privacy and sensitivity. Performing multiple identity authentication can effectively prevent unauthorized access, avoid user data leakage, and protect user privacy and security. Common multiple identity authentication methods include password authentication, SMS verification code authentication, fingerprint recognition authentication, facial recognition authentication and other combinations. When the target user passes the multiple identity authentication, the system will determine it as the object to be monitored. The system will create a corresponding record for the user in the database, which includes the user's basic information, historical posture data (if any), etc. At the same time, the system will configure the corresponding monitoring parameters for the user based on the user's identity information, such as age, gender, occupation, etc., and the previously set monitoring strategy (if set). For an older retired user, the system may select a lower intensity monitoring frequency for it, and adjust the feedback method to a larger font display and voice prompt that is easier to understand; for young fitness enthusiasts, the system will configure higher-precision monitoring parameters and real-time feedback functions according to their fitness needs, so that users can adjust their exercise postures in time. In this way, the system can provide personalized, safe and reliable posture monitoring services for different users.

[0037] S102, inputting the individual motion data into a posture determination model to determine posture information corresponding to the individual motion data, wherein the posture determination model is trained in advance by a deep neural network based on a plurality of labeled posture sample sets containing individual motions; Before inputting individual motion data into the posture determination model, the posture monitoring system will pre-process the data, including standardizing the data and converting the data into a format suitable for model input. For image data, normalization will be performed to map pixel values ​​to a specific interval to eliminate the impact of data magnitude differences on model training and prediction; for numerical data collected by sensors, feature scaling will be performed to make them have the same scale.

[0038] The posture determination model uses a deep neural network architecture that can automatically learn complex features and patterns in the data. During the training phase, the system uses a large number of labeled posture sample sets containing individual movements to train the model. These sample sets cover a variety of posture types, action scenes, and individual differences, ensuring that the model can learn comprehensive posture characteristics. During the training process, the posture monitoring system uses the back propagation algorithm to calculate the error between the model prediction results and the labeled data, and adjusts the model parameters according to the error, so that the model's prediction results continue to approach the real posture information. When the individual action data is input into the trained posture determination model, the model will process the data layer by layer. For image data, the posture determination model extracts local features in the image through the convolution kernel, such as the position of the joints of the human body, the angle of the limbs, etc.; the pooling layer compresses the features, reducing the amount of data while retaining the key features. Finally, the model outputs the posture information corresponding to the individual action data, including detailed information such as the posture, action type, and action amplitude of the human body.

[0039] In some embodiments, the posture monitoring system can determine the degree of posture deviation and formulate a reminder strategy accordingly, so that users can understand their own posture problems in a timely and effective manner, thereby guiding users to adjust their posture and prevent health problems caused by long-term poor posture. The specific implementation process is as follows: After the posture monitoring system obtains the posture information output by the posture determination model, it will compare and analyze it with the pre-set standard posture range. The standard posture range is obtained by combining multidisciplinary knowledge such as ergonomics, kinematics and medicine, combined with a large amount of statistical analysis of human motion data, covering the normal posture parameters of the human body in various common activity scenarios. The system uses a specific calculation method to measure the degree of difference between the posture information and the standard posture. For posture angle information, the deviation value between the actual angle and the standard angle is calculated; for position information, the distance or coordinate difference between the actual position and the standard position is calculated. In order to comprehensively evaluate the degree of conformity between the posture and the standard posture, the system will assign corresponding weights to each posture parameter according to its importance to the overall posture health. Through weighted calculation, a comprehensive posture deviation value is obtained which can accurately reflect the difference between body posture and standard posture. The degree of posture deviation of the user is determined based on this value, which is generally divided into mild, moderate and severe deviation ranges.

[0040] When the degree of posture deviation is within the preset mild deviation range, the system determines that the user's posture problem is relatively mild at this time, but it still needs attention. The preset slight vibration is generated by the smart wearable device to remind the user. This reminder method is relatively gentle and will not cause much interference to the user. If the degree of posture deviation reaches the preset moderate deviation range, it means that the user's posture problem is more obvious and requires a stronger reminder to prompt adjustment. At this time, the smart wearable device will generate a preset medium-intensity vibration to remind the user. When the degree of posture deviation is within the preset severe deviation range, it means that the user's posture problem is more serious and may cause great harm to physical health. The system not only generates a preset strong vibration to remind the user through the smart wearable device, but also sends a warning message to the mobile terminal simultaneously. In this way, whether the user is wearing a smart wearable device or viewing mobile terminal information, he can promptly discover and pay attention to posture problems and take corresponding corrective measures.

[0041] S103, if the body posture information is monitored to be within the preset standard posture range, the current monitoring state is maintained; The posture monitoring system will pre-set a comprehensive and scientific standard posture range, which covers the normal posture parameters of the human body in different activity scenarios. For standing posture, the standard posture range will specify the vertical axis position of the human body, the ideal angle and position relationship of the head, shoulders, spine, hips and legs; for walking posture, it will clarify the parameter range of the size, frequency, and transfer trajectory of the body's center of gravity; for sitting posture, it will define the degree of back straightness, the placement angle of arms and legs, etc. These standard posture ranges are based on multidisciplinary knowledge such as ergonomics, kinematics, and medicine, and are combined with a large amount of statistical analysis results of human motion data to ensure their scientificity and rationality.

[0042] After the posture determination model outputs posture information, the posture monitoring system will compare this information with the preset standard posture range. During the comparison process, the system uses precise calculation methods to measure the degree of difference between the posture information and the standard posture. For posture angle information, the system calculates the deviation value between the actual angle and the standard angle; for position information, the distance or coordinate difference between the actual position and the standard position is calculated. In order to comprehensively evaluate the degree of conformity between the posture and the standard posture, the system assigns corresponding weights to different posture parameters according to their importance to the overall posture health. For example, the posture of the spine has a greater impact on physical health and will be given a higher weight; while some relatively minor limb position parameters have relatively lower weights. Through weighted calculation, a comprehensive posture deviation value is obtained to accurately determine whether the posture is within the standard posture range.

[0043] If the posture information is detected to be within the preset standard posture range, the posture monitoring system will maintain the current monitoring state. During the monitoring state, the system will continue to monitor the working status of the sensor to ensure its normal operation, and will also regularly perform self-inspections on the sensor to check whether the sensor's acquisition frequency, data transmission stability and other indicators are normal. Once a sensor failure or abnormality is found, the system will promptly issue an alarm and attempt to automatically repair it or switch to a backup sensor to ensure the continuity of data collection.

[0044] S104, if the posture information is detected to be beyond the standard posture range, obtaining a plurality of historical posture information of the user; Once the posture monitoring system determines that the current posture information exceeds the preset standard posture range, it will quickly start the historical posture information acquisition program. The system will first locate the storage location of the historical posture data related to the user in the local database or cloud storage based on the user's identity. To ensure the efficiency of data acquisition, the system uses indexing technology to establish an index based on key information such as user ID and timestamp to quickly retrieve the corresponding historical data file.

[0045] When obtaining historical posture information, the system will consider the integrity and timeliness of the data. Since the user's posture may change over time, recent data can better reflect the user's current physical condition and habits, so the system will give priority to obtaining historical posture information within a recent period of time, such as data within the past week or month. However, for some users with long-term posture problems, earlier data is also of reference value. The system will obtain data in chronological order, from recent to far, until the preset data volume or time span requirements are reached. After obtaining historical posture information, the system will preliminarily organize and analyze the data, and sort the posture data at different time points in chronological order to facilitate subsequent analysis and comparison.

[0046] S105: After determining a personalized correction plan according to the historical body posture information, the personalized correction plan is sent to a mobile terminal for display.

[0047] After obtaining and organizing the user's historical posture information, the posture monitoring system begins to formulate a personalized correction plan. The system first conducts an in-depth analysis of the historical posture information, and uses data mining and machine learning algorithms to extract the rules and trends of the user's posture changes. Through cluster analysis, the system can find the types of bad postures that users often have, such as hunched backs, uneven shoulders, etc., and analyze the frequency, duration, and association of these bad postures with the user's daily activities. Combining the user's historical posture data, current posture deviations, and the user's personal information (such as age, gender, physical condition, athletic ability, etc.), the system generates a personalized correction strategy. For older users with poor body flexibility, the correction plan will focus on gentle stretching and posture adjustment exercises to avoid physical injuries caused by excessive exercise; for young users with strong athletic ability, some slightly more intense exercises can be added to speed up the correction of posture.

[0048] The system will develop a detailed correction plan based on the correction strategy. The correction plan includes a series of specific corrective movements, training frequency and time schedule. The design of the corrective movements is based on the principles of ergonomics and sports medicine to ensure the scientificity and effectiveness of the movements. In order to improve the hunched posture, the plan may include back stretching, core muscle training and other movements. The system will generate detailed instructions and demonstration videos for each corrective movement, and through animation or real-person demonstrations, let users clearly understand the key points and correct practices of the movements.

[0049] In terms of training frequency and time arrangement, the system will make reasonable plans based on the user's daily activity habits and physical recovery ability. For users with busy work, the system will design some short-term, high-intensity training plans to facilitate users to exercise in fragmented time; for users with more time, a more systematic and comprehensive training plan can be arranged. The system will also take into account the user's physical recovery time to avoid fatigue or injury caused by excessive training.

[0050] After formulating a personalized correction plan, the posture monitoring system sends the plan to the mobile terminal for display. The system transmits the correction plan to the user's mobile terminal in a secure and encrypted manner through network communication technology. To ensure the stability and security of data transmission, the system uses a secure and reliable communication protocol, such as the HTTPS protocol, to encrypt the transmitted data to prevent the data from being stolen or tampered with during transmission.

[0051] In terms of mobile terminal display, the system will provide users with an intuitive and friendly interface. The correction plan will be presented in the form of a list or calendar, and users can clearly see the daily training tasks, action instructions and estimated training time. Users can click on a specific training task to view detailed action demonstration videos and text instructions. The mobile terminal will also set a reminder function to remind users to train on time according to the time schedule of the correction plan, helping users develop good training habits.

[0052] In the embodiment of the present application, due to the use of fusion multimodal sensing technology to collect individual motion data from optical, infrared and acoustic sensors, the use of deep neural networks to train posture determination models to accurately analyze posture information, differentiated processing is performed based on the comparison results of posture information with the standard posture range, and historical information is obtained when the posture is abnormal to formulate a personalized correction plan and push it to the mobile terminal. Therefore, the technical means effectively solve the problems of existing posture monitoring technologies such as limited monitoring methods, poor accuracy, inability to provide personalized services, and difficulty in adapting to different user needs and complex environments, thereby achieving non-intrusive high-precision posture detection in different scenarios, providing users with accurate and personalized posture monitoring and correction services, improving the accuracy and effectiveness of posture monitoring, meeting the diverse needs of users, and improving the technical effect of users' physical health level.

[0053] After combining the above content, the following is a more detailed description of the process of the method provided by this implementation. Figure 2 , which is another flow chart of the AI-driven all-round posture monitoring method in an embodiment of the present application.

[0054] S201. Establish a user behavior record according to the user's response to the reminder strategy. The user behavior record includes the posture adjustment time after the user receives the reminder, the improvement effect of the user after receiving the reminder, and the posture improvement trend of the user. The posture adjustment time is determined according to the time interval required for the posture data of the user to return to the standard posture range after receiving the reminder. The improvement effect is determined according to the difference between the posture parameters of the user after receiving the reminder strategy and the standard posture parameters. The posture improvement trend is determined according to the statistical analysis results of the user's posture parameters within a continuous set time. The posture monitoring system will first continuously monitor the user's response information to the reminder strategy fed back by the smart wearable device and mobile terminal. When receiving the relevant feedback signal triggered by the reminder, the system starts the user behavior recording program. In order to accurately determine the posture adjustment time, the system uses high-precision timestamp technology. At the time of reminder sending, the system records the precise timestamp T1. When the user's posture data is monitored to return to the standard posture range, the timestamp T2 is recorded again. By calculating the difference between T2 and T1, the posture adjustment time is obtained.

[0055] In terms of calculating the improvement effect, the system will obtain the real-time posture parameters of the user after receiving the reminder strategy from the posture determination model. These parameters cover multi-dimensional data such as the posture angle and position information of various parts of the body. The system compares these real-time posture parameters with the pre-set standard posture parameters. In order to accurately calculate the difference change, the system uses vector calculation and spatial geometry algorithms. The posture parameters are regarded as multi-dimensional vectors, and the degree of difference between the posture parameters and the standard posture parameters is quantified by calculating indicators such as the Euclidean distance and the cosine value of the angle between the vectors. For the angle parameters of the body joints, the absolute value or relative ratio of the difference between the actual angle and the standard angle is calculated; for the position parameters of the body parts, the distance deviation in the spatial coordinate system is calculated. Through the comprehensive calculation of the differences in multiple posture parameters, a quantitative value that fully reflects the effect of posture improvement is obtained.

[0056] For the evaluation of posture improvement trends, the system will collect the user's posture parameter sequence within a set continuous time. Time series analysis methods, such as autoregressive moving average model (ARIMA), Kalman filtering and other technologies, are used to model and predict posture parameter sequences. By analyzing the changing trend of posture parameters in the time dimension, the system can determine whether the user's posture is gradually improving, remaining stable or deteriorating. The system will also use machine learning algorithms, such as support vector machines (SVMs) and decision trees, to classify and cluster posture parameters and explore potential patterns and laws of posture changes. Combined with these analysis results, the system can more accurately evaluate the user's posture improvement trend and provide a strong basis for subsequent reminder strategy adjustments and personalized correction plan optimization.

[0057] S202, dynamically adjusting the reminder strategy according to the user behavior record; After obtaining the user's behavior records, the posture monitoring system will conduct an in-depth analysis to dynamically adjust the reminder strategy. The system will judge the effectiveness of the reminder based on the posture adjustment time. If the user can adjust the posture back to the standard range in a short time after receiving the reminder, it means that the current reminder strategy is more effective. The system will appropriately reduce the frequency of reminders to avoid excessive reminders causing trouble to the user. Adopt an adaptive frequency adjustment algorithm to adjust the reminder frequency according to a certain proportional coefficient based on the length of the posture adjustment time. If the posture adjustment time is less than the preset minimum time threshold, the system will reduce the reminder frequency by a certain percentage; conversely, if the posture adjustment time is too long and exceeds the preset maximum time threshold, the system will increase the reminder frequency to ensure that users can pay attention to posture problems in a timely manner.

[0058] The system will analyze the improvement effect based on the calculated quantitative values. If the improvement effect is obvious, that is, the difference between the body parameters and the standard body parameters is large and close to the standard body, the system will adjust the reminder intensity to make it relatively mild.

[0059] When evaluating the trend of posture improvement, if the system determines through time series analysis and machine learning algorithms that the user's posture is in a state of continuous improvement, the system will further optimize the reminder strategy to make it more personalized and targeted. Adjust the timing and method of reminders based on the user's posture improvement patterns and habits. If it is found that the user is more likely to accept reminders and adjust their posture during a certain period of time each day, the system will increase the frequency and accuracy of reminders during these time periods; if the user is more sensitive to a certain type of reminder (such as vibration reminder), the system will appropriately increase the proportion of that type of reminder.

[0060] S203, after receiving the posture information viewing instruction sent by the mobile terminal, generating a user posture improvement report according to the user behavior record and sending it to the mobile terminal for display; After receiving the posture information viewing instruction from the mobile terminal, the posture monitoring system will quickly start the user posture improvement report generation program. The system will first conduct a comprehensive review and in-depth analysis of the existing user behavior records. Starting with the posture adjustment time data, the system will count the speed differences of posture adjustment after users receive reminders in different time periods.

[0061] For the improvement effect data, the system will further subdivide the changes in posture parameters. Taking spinal posture as an example, the system will analyze in detail the changing trends of parameters such as the curvature angle of the spine and the distance from the central axis under the reminder strategy. By comparing the improvement effects at different stages, the system can determine which corrective movements or training are more effective in improving spinal posture. At the same time, combined with the posture improvement trend data, time series prediction models, such as long short-term memory networks (LSTM), are used to predict the user's posture development trend in the future. If it is predicted that the user's hunchback problem will be significantly improved through the current training plan within the next week, then this positive trend will be clearly shown in the report; conversely, if there is a potential risk of deterioration in posture, it will also be warned in the report in a timely manner.

[0062] During the report generation process, the system uses visualization technology to convert complex data into intuitive and easy-to-understand charts and graphs. In order to show the changes in posture adjustment time, the system will generate a line graph with time as the horizontal axis and posture adjustment time as the vertical axis, clearly showing the increase or decrease in the time required for users to adjust their posture over time. For the improvement effect, a bar graph will be used to compare the degree of improvement of different posture parameters, allowing users to see at a glance the optimization of their posture in various parts of their body. The posture improvement trend is displayed through a trend graph to help users better understand the direction of their posture changes.

[0063] In addition, the report will also incorporate professional analysis and suggestions. The system will give personalized health suggestions based on the user's age, gender, occupation and other personal information combined with body shape data. After the report is generated, the body shape monitoring system will send the user's body shape improvement report to the mobile terminal through secure and stable network communication technology.

[0064] S204, evaluating the user's adaptability to the personalized correction plan based on the user behavior record; When the posture monitoring system evaluates the user's adaptability to the personalized correction plan based on the user's behavior records, it will make comprehensive considerations from multiple dimensions. First, the system will analyze the user's completion of the training movements in the correction plan. By monitoring the data collected by smart wearable devices and edge devices, the system can accurately determine whether the user is exercising in accordance with the movement specifications, training frequency and duration required by the plan. If more than 90% of the user's training movements meet the standard specifications within a week, and the training frequency and duration are in line with the plan, then it can be preliminarily judged that the user has a high degree of adaptability to the training movements; conversely, if the movement deformation and insufficient training times often occur, it means that the user may have problems with movement adaptation.

[0065] The system will also pay attention to the user's physiological feedback data during training, and collect data such as heart rate changes and muscle fatigue during training through heart rate sensors and pressure sensors integrated in smart wearable devices. When a user performs a corrective exercise, if the heart rate is within a reasonable range and muscle fatigue can be recovered quickly after the training, it indicates that the user's body has adapted well to the training intensity; if the heart rate is too high or the user is in a fatigued state for a long time, it may mean that the training intensity is too high and the user's body is difficult to adapt. The posture monitoring system will set up a feedback portal on the mobile terminal to encourage users to regularly evaluate the correction plan, such as whether they feel that the training movements are too difficult and whether the training time arrangement is reasonable. At the same time, the system will use natural language processing technology to analyze the user's feedback content, extract key information, and judge the user's satisfaction and needs.

[0066] S205: Dynamically adjust the personalized correction plan according to the degree of adaptation.

[0067] After obtaining the evaluation results of the user's adaptability to the personalized correction plan, the posture monitoring system will dynamically adjust the correction plan in a targeted manner. If the evaluation results show that the user has adapted well to the current correction plan and the posture improvement effect is obvious, the system will appropriately increase the difficulty and intensity of the training to further promote the optimization of the posture. For users who have been doing posture correction training for a period of time and have a high degree of adaptability, the system can add some challenging balance training movements on the basis of the original simple stretching movements, such as single-leg standing stretch, to improve the body's balance ability and better exercise the core muscles, thereby improving the overall posture.

[0068] When the assessment finds that the user has partially adapted to the correction plan and there are some difficulties in execution or the effect is not obvious, the system will optimize and adjust the training movements. If the user is unable to complete a specific corrective movement correctly due to the difficulty of the movement, the system will use movement decomposition and simplification technology to split the complex movement into several simple sub-movements, allowing the user to gradually master it. For twisting movements that require greater body coordination, the system can first let the user perform simple side flexion exercises, and then gradually increase the complexity of the movement after the user adapts. In addition, the system will adjust the training intensity based on the user's physiological feedback data. If the user has a high heart rate or excessive muscle fatigue during training, the system will appropriately reduce the training intensity, such as reducing the number of repetitions of the movement or shortening the training time, to ensure the safety and effectiveness of the training.

[0069] For users with low adaptability, the system will re-formulate a correction plan. First, the system will conduct an in-depth analysis of the reasons for the user's lack of adaptability, whether it is because the training movements are too difficult, the training time is too long, or it does not match the user's personal physical condition. If it is a problem of movement difficulty, the system will re-select some movements that are more in line with the user's current physical ability and sports foundation. For users with poor body flexibility, some high-intensity stretching movements are abandoned and replaced with simple joint movement exercises, such as wrist and ankle rotation, slight neck rotation, etc., to help users gradually improve their body flexibility and mobility.

[0070] In the embodiment of the present application, due to the use of a technical means of establishing a user behavior record according to the user's response to the reminder strategy, dynamically adjusting the reminder strategy, generating a posture improvement report, and evaluating the user's adaptability to the personalized correction plan and dynamically adjusting the correction plan, it is possible to accurately grasp the actual situation and demand changes in the user's posture adjustment process. This not only effectively solves the problem that the existing posture monitoring technology cannot flexibly adjust the monitoring and correction strategy according to the user's real-time feedback, resulting in poor user posture improvement effects, lack of pertinence and continuity in monitoring and correction, but also realizes the provision of users with more accurate, personalized and continuously optimized posture monitoring and correction services, significantly improving the user's posture adjustment efficiency and final effect, and effectively meeting the user's diversified and dynamic needs in posture management.

[0071] The following describes the posture monitoring system in the embodiment of the present invention from the perspective of hardware processing. Figure 3 , is a schematic diagram of a physical device structure of a body posture monitoring system in an embodiment of the present application.

[0072] It should be noted that Figure 3The structure of the body posture monitoring system shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0073] like Figure 3 As shown, the body posture monitoring system includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 302 or the program loaded from the storage part 308 to the random access memory (RAM) 303, such as executing the method described in the above embodiment. In the RAM 303, various programs and data required for system operation are also stored. The CPU 301, the ROM 302 and the RAM 303 are connected to each other through the bus 304. The input / output (I / O) interface 305 is also connected to the bus 304.

[0074] The following components are connected to the I / O interface 305: an input section 306 including an audio input device, a button switch, etc.; an output section 307 including a liquid crystal display (LCD) and an audio output device, an indicator light, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to the I / O interface 305 as needed. A removable medium 311, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 310 as needed so that a computer program read therefrom is installed into the storage section 308 as needed.

[0075] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a computer program for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through the communication part 309, and / or installed from a removable medium 311. When the computer program is executed by the central processing unit (CPU) 301, various functions defined in the present invention are performed.

[0076] It should be noted that specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device.

[0077] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present invention. Each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings.

[0078] Specifically, the posture monitoring system of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, the AI-driven omnidirectional posture monitoring method provided in the above embodiment is implemented.

[0079] As another aspect, the present invention further provides a computer-readable storage medium, which may be included in the posture monitoring system described in the above embodiment; or may exist independently without being assembled into the posture monitoring system. The above storage medium carries one or more computer programs, and when the above one or more computer programs are executed by a processor of the posture monitoring system, the posture monitoring system implements the AI-driven all-round posture monitoring method provided in the above embodiment.

[0080] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

[0081] As used in the above embodiments, the term "when..." may be interpreted to mean "if..." or "after..." or "in response to determining..." or "in response to detecting...", depending on the context. Similarly, the phrases "upon determining..." or "if (the stated condition or event) is detected" may be interpreted to mean "if determining..." or "in response to determining..." or "upon detecting (the stated condition or event)" or "in response to detecting (the stated condition or event)", depending on the context.

[0082] Those skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by computer programs to instruct related hardware, and the programs can be stored in computer-readable storage media. When the programs are executed, they can include the processes of the above-mentioned method embodiments. The aforementioned storage media include: ROM or random access memory RAM, magnetic disk or optical disk and other media that can store program codes.

Claims

1. An AI-driven all-round posture monitoring method, applied to a posture monitoring system, characterized in that: The method comprises: Acquire individual motion data from edge devices, where the individual motion data is collected from optical sensors, infrared sensors, and acoustic sensors using fusion multimodal sensing technology; Inputting the individual motion data into a posture determination model to determine posture information corresponding to the individual motion data, wherein the posture determination model is trained in advance by a deep neural network based on a plurality of labeled posture sample sets containing individual motions; If the body posture information is detected to be within the preset standard posture range, the current monitoring state is maintained; If the posture information is detected to be beyond the standard posture range, multiple historical posture information of the user is obtained; After determining a personalized correction plan based on the historical body posture information, the personalized correction plan is sent to the mobile terminal for display.

2. The method according to claim 1, characterized in that Before the step of obtaining individual motion data from edge devices, it also includes: According to different application scenarios and user needs set by the user, the corresponding monitoring strategy is selected from the pre-set strategy library, which is a strategy set that configures the accuracy, frequency and feedback method of body posture monitoring according to different environments and different user groups; Adjust various parameters of edge devices according to the monitoring strategy.

3. The method according to claim 1, characterized in that After the step of obtaining individual motion data from the edge device, it also includes: Automatically adjust the parameters of the optical camera of the edge device when receiving a warning of insufficient illumination from the optical sensor; After acquiring the user's individual action image through the edge device, the image enhancement algorithm is enabled to enhance the brightness of the preset dark area in the individual action image to obtain the action outline under dim conditions.

4. The method according to claim 1, characterized in that: After the step of inputting the individual motion data into the posture determination model to determine the posture information corresponding to the individual motion data, the method further includes: Determining the degree of posture deviation of the user according to the posture information and the standard posture range; Determining a corresponding reminder strategy based on the degree of posture deviation, the reminder strategy comprising generating a preset slight vibration reminder through the smart wearable device when the degree of posture deviation is within a preset mild deviation range; When the degree of posture deviation is within a preset moderate deviation range, a preset moderate intensity vibration reminder is generated through the smart wearable device; When the degree of posture deviation is within a preset severe deviation range, a preset strong vibration reminder is generated through the smart wearable device, and a warning message is sent to the mobile terminal simultaneously.

5. The method according to claim 1, characterized in that After the step of determining a corresponding reminder strategy based on the degree of posture deviation, the method further includes: According to the user's response to the reminder strategy, a user behavior record is established, and the user behavior record includes the posture adjustment time after the user receives the reminder, the improvement effect of the user after receiving the reminder, and the evaluation of the posture improvement trend of the user. The posture adjustment time is determined according to the time interval required for the posture data of the user to return to the standard posture range after receiving the reminder, the improvement effect is determined according to the difference between the posture parameters of the user after receiving the reminder strategy and the standard posture parameters, and the posture improvement trend is determined according to the statistical analysis results of the user's posture parameters within a continuous set time; Dynamically adjust the reminder strategy according to the user behavior record; After receiving the posture information viewing instruction sent by the mobile terminal, a user posture improvement report is generated according to the user behavior record and sent to the mobile terminal for display.

6. The method according to claim 5, characterized in that After the step of establishing a user behavior record according to the user's response to the reminder strategy, the method further includes: Assessing the user's adaptability to the personalized correction plan based on the user behavior record; The personalized correction plan is dynamically adjusted according to the degree of adaptation.

7. The method according to claim 1, characterized in that Before the step of obtaining individual motion data from edge devices, it also includes: Perform multi-factor authentication on target users; When multiple identity authentications are passed, the target user is determined to be an object to be monitored.

8. A body posture monitoring system, characterized in that: The posture monitoring system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to enable the posture monitoring system to execute the method described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that: When the instructions are executed on a body posture monitoring system, the body posture monitoring system is caused to execute the method according to any one of claims 1 to 7.

10. A computer program product, characterized in that When the computer program product is executed on a body posture monitoring system, the body posture monitoring system is enabled to execute the method according to any one of claims 1 to 7.

Images