Geriatric rehabilitation monitoring method, yoga mat, device and storage medium

Abstract

The application relates to the technical field of fitness equipment, in particular to an old-age rehabilitation monitoring method, a yoga mat, equipment and a storage medium, the old-age rehabilitation monitoring method first acquires a mode signal, judges a current mode based on the mode signal, acquires a first sensor data set when the current mode is a rehabilitation yoga mode, then carries out feature parameter extraction processing on the first sensor data set to obtain a posture data set, finally acquires a preset rehabilitation standard body position database, carries out rehabilitation effect evaluation on the posture data set based on the rehabilitation standard body position database, obtains a rehabilitation training evaluation report, and aims to accurately adapt to an old-age rehabilitation yoga exclusive training scene, and solves the technical defects that the existing rehabilitation monitoring scheme is insufficient in adaptability and depends on subjective experience in evaluation.

Description

Technical Field

[0001] This invention relates to the field of fitness equipment technology, and in particular to a method for monitoring elderly rehabilitation, a yoga mat, a device, and a storage medium. Background Technology

[0002] The aging population is driving technological innovation in the field of geriatric rehabilitation medicine. As a training method suitable for the recovery of limb function in the elderly, rehabilitation yoga is receiving increasing attention for its scientific validity and safety. Currently, geriatric rehabilitation yoga training largely relies on on-site guidance from professionals, which suffers from resource scarcity and limited guidance coverage. Furthermore, manual assessment depends on experience-based judgment, making it difficult to accurately quantify training postures and rehabilitation effects. While some intelligent monitoring solutions attempt to introduce sensors to collect data, they lack targeted mode adaptation mechanisms and cannot dynamically adjust data collection and assessment logic according to the training scenario.

[0003] At the same time, existing solutions cannot provide personalized training optimization suggestions for elderly users. This often leads to problems such as the inability to correct posture deviations in a timely manner and the difficulty in accurately controlling the rehabilitation process during rehabilitation yoga training for the elderly. This not only affects the rehabilitation effect, but may also cause joint damage and other risks due to improper training. Summary of the Invention

[0004] In order to overcome the shortcomings of the prior art, the present invention aims to provide a method, yoga mat, device and storage medium for elderly rehabilitation monitoring, which is designed to accurately adapt to the special training scenario of elderly rehabilitation yoga, and solve the technical defects of insufficient adaptability and reliance on subjective experience in the existing rehabilitation monitoring program.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A method for monitoring elderly rehabilitation includes: acquiring a pattern signal; determining a current pattern based on the pattern signal; when the current pattern is a rehabilitation yoga pattern, acquiring a first sensor dataset; performing feature parameter extraction processing on the first sensor dataset to obtain a posture dataset; acquiring a preset rehabilitation standard posture database; evaluating the rehabilitation effect of the posture dataset based on the rehabilitation standard posture database to obtain a rehabilitation training evaluation report.

[0006] In the described elderly rehabilitation monitoring method, the step of extracting feature parameters from the first sensor dataset to obtain an attitude dataset includes: using a Kalman filter noise reduction algorithm to perform noise removal processing on the first sensor dataset to obtain a first standard sensor dataset; obtaining preset attitude key parameter screening rules, and performing attitude feature screening processing on the first standard sensor dataset based on the attitude key parameter screening rules to obtain an attitude key feature set; and calling a pre-trained attitude three-dimensional coordinate mapping model to perform coordinate transformation on the attitude key feature set to obtain the attitude dataset.

[0007] In the aforementioned elderly rehabilitation monitoring method, after determining the current mode based on the mode signal, the method further includes: when the current mode is an elderly care monitoring mode, acquiring a second sensor dataset; performing feature parameter extraction processing on the second sensor dataset to obtain a human activity feature dataset; using a fall detection algorithm to judge whether a person has fallen on the human activity feature dataset, and obtaining a judgment result; when the judgment result indicates that a person has fallen, generating an alarm signal.

[0008] In the described elderly rehabilitation monitoring method, the step of extracting feature parameters from the second sensor dataset to obtain a human activity feature dataset includes: using a Kalman filter noise reduction algorithm to perform noise removal processing on the second sensor dataset to obtain a second standard sensor dataset; obtaining preset fall key parameter screening rules, and performing fall feature screening processing on the second standard sensor dataset based on the fall key parameter screening rules to obtain a human activity key feature set; and calling a pre-trained XGBoost (eXtreme Gradient Boosting) feature evaluation model to perform feature sorting and integration processing on the human activity key feature set to obtain the human activity feature dataset.

[0009] The present invention also provides a yoga mat for elderly rehabilitation monitoring, wherein the yoga mat employs any of the elderly rehabilitation monitoring methods described above for work control; the yoga mat includes a control unit, a sensor unit, a data acquisition unit, and a mode switching unit, wherein the data acquisition unit is connected to the control unit, and the sensor unit and the mode switching unit are respectively connected to the data acquisition unit; the data acquisition unit is used to acquire sensor data sets from the sensor unit and mode signals from the mode switching unit, and to send the sensor data sets and the mode signals to the control unit; the control unit is used to determine the current mode based on the mode signals, and when the current mode is rehabilitation yoga mode, to evaluate the rehabilitation effect based on the sensor data to obtain a rehabilitation training evaluation report.

[0010] The elderly rehabilitation monitoring yoga mat also includes an alarm unit, which is connected to the control unit. The control unit is also used to determine whether a person has fallen based on the sensor data when the current mode is the elderly monitoring mode, so as to obtain a judgment result. The alarm unit is used to issue an audible and visual alarm when the judgment result indicates that a person has fallen.

[0011] In the aforementioned yoga mat for elderly rehabilitation monitoring, the sensor unit includes a hand detection sensor group, a waist detection sensor group, a knee joint detection sensor group, and a foot detection sensor group. These sensor groups are connected to the data acquisition unit. The hand detection sensor group is used to collect hand support pressure and balance parameters; the waist detection sensor group is used to collect waist fit and bending curvature; the knee joint detection sensor group is used to collect knee joint load-bearing pressure and bending angle; and the foot detection sensor group is used to collect foot pressure and center of gravity shift.

[0012] In the aforementioned yoga mat for elderly rehabilitation monitoring, the sensor unit further includes a fall detection sensor group, which is connected to the data acquisition unit. The fall detection sensor group includes a flexible pressure array sensor, a three-axis accelerometer sensor, and a gyroscope sensor. The flexible pressure array sensor is used to collect the contact area between the human body and the yoga mat, the three-axis accelerometer sensor is used to collect the abrupt change value of the three-axis acceleration, and the gyroscope sensor is used to collect the tilt angle of the human body posture.

[0013] A third aspect of the present invention provides an elderly rehabilitation monitoring device, the elderly rehabilitation monitoring device comprising: a memory and at least one processor, the memory storing instructions; at least one processor calling the instructions in the memory to cause the elderly rehabilitation monitoring device to perform the various steps of the elderly rehabilitation monitoring method described in any of the preceding claims.

[0014] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions which, when executed by a processor, implement the steps of any of the above-described methods for monitoring and caring for the elderly.

[0015] In the technical solution of this invention, the elderly rehabilitation monitoring method first acquires a pattern signal, determines the current mode based on the pattern signal, and when the current mode is rehabilitation yoga mode, acquires a first sensor dataset, then performs feature parameter extraction processing on the first sensor dataset to obtain a posture dataset, and finally acquires a preset rehabilitation standard posture database, evaluates the rehabilitation effect of the posture dataset based on the rehabilitation standard posture database, and obtains a rehabilitation training evaluation report. The aim is to achieve accurate quantitative evaluation of training posture and rehabilitation effect through targeted sensor data acquisition and feature extraction, output scientific rehabilitation training evaluation reports for elderly users, and provide personalized training optimization guidance based on a standardized rehabilitation posture reference system, thereby improving the standardization and safety of rehabilitation training, making up for the shortcomings of existing solutions in accurately controlling the rehabilitation process and having limited guidance coverage, and helping the elderly to obtain an efficient and safe rehabilitation training experience. Attached Figure Description

[0016] Figure 1 A logic flowchart of the elderly rehabilitation monitoring method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of the yoga mat for elderly rehabilitation monitoring provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of the elderly rehabilitation monitoring device provided in an embodiment of the present invention. Detailed Implementation

[0017] This invention provides a method for monitoring rehabilitation in the elderly, a yoga mat, a device, and a storage medium. In this invention, the terms "first," "second," "third," "fourth," etc. (if present)," in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0018] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 One embodiment of the elderly rehabilitation monitoring method of the present invention includes: 101. Acquire the mode signal and determine the current mode based on the mode signal; 102. When the current mode is rehabilitation yoga mode, acquire the first sensor dataset; In this embodiment, the pattern signal acquisition and current pattern determination process is based on a preset rehabilitation yoga mode and elderly care monitoring mode system and switching mechanism. The system first loads the pre-configured core parameters of the rehabilitation yoga mode and elderly care monitoring mode. The generation of pattern signals originates from two trigger paths: one is the manual switching operation of the terminal App, where a clear pattern signal is formed through the user's active input of the mode selection command; the other is the system's built-in scene perception-driven automatic switching mechanism. Through real-time monitoring of the operating status data, when it detects that the elderly user is in a prolonged static state, it automatically generates an elderly care monitoring mode signal; when it identifies that the user is exhibiting a regular movement trajectory, it automatically generates a rehabilitation yoga mode signal. Subsequently, the system analyzes and matches the pattern signals generated by the above two paths to accurately determine the current pattern. After completing the pattern determination, if the current pattern is determined to be the rehabilitation yoga mode, the system will acquire the first sensor dataset. The acquisition range and parameter configuration of this dataset are optimized for the posture monitoring needs of rehabilitation yoga training, ensuring that the acquired data can accurately reflect the limb posture characteristics of the user during training, providing effective data support for subsequent feature parameter extraction and rehabilitation effect evaluation. By deeply coupling the preset dual-mode and flexible switching mechanism with the mode judgment and data acquisition process, the system achieves accurate mode recognition and scene adaptability, effectively avoiding conflicts between monitoring and training functions in a single mode. The dual triggering path of manual switching and automatic recognition not only meets the autonomous needs of the elderly to actively control the mode, but also realizes intelligent mode switching without manual intervention through scene perception, adapting to the needs of the elderly with different operating abilities and different usage scenarios. Based on the mode judgment results, the system can obtain the first sensor dataset in a targeted manner, enabling targeted screening of data collection, reducing the system resource consumption caused by invalid data collection, and ensuring a high degree of adaptability of the collected data to the feature extraction and effect evaluation functions of the subsequent rehabilitation yoga mode, thereby improving the overall operating efficiency and evaluation accuracy of the rehabilitation monitoring system.

[0019] 103. Perform feature parameter extraction processing on the first sensor dataset to obtain the attitude dataset; 104. Obtain a preset rehabilitation standard posture database, evaluate the rehabilitation effect of the posture dataset based on the rehabilitation standard posture database, and obtain a rehabilitation training evaluation report.

[0020] In this embodiment, feature parameter extraction processing is performed on the first sensor dataset to obtain a posture dataset. This process involves systematically analyzing multi-dimensional data closely related to elderly rehabilitation yoga training, such as pressure, posture, and heart rate collected by the first sensor. Through precise screening, quantification, conversion, and integration, a posture dataset that objectively reflects the limb posture and physical load of elderly users during yoga training is extracted. A pre-set rehabilitation standard posture database is obtained. This database is a standardized posture information set pre-constructed based on the physiological characteristics of the elderly population, clinical research results in rehabilitation medicine, and professional standards for yoga rehabilitation training. It covers core parameters such as standard limb joint angles, body center of gravity distribution, and limb pressure transmission paths corresponding to different rehabilitation stages. Each standard posture is associated with a clear rehabilitation goal and a suitable training population, serving as a core reference for rehabilitation effect evaluation. This provides an objective and scientific benchmark for rehabilitation effect evaluation, avoiding subjectivity and experience-based biases in the evaluation work and significantly improving the authority and credibility of the evaluation results. The rehabilitation effect is evaluated based on the rehabilitation standard posture database, resulting in a rehabilitation training evaluation report. The core of this report relies on a posture assessment algorithm for in-depth analysis. This algorithm comprehensively compares various indicators in the posture dataset with preset elderly rehabilitation standard posture parameters in the rehabilitation standard posture database, accurately identifying deviations between the user's current training posture and the standard posture. Simultaneously, it combines heart rate data to determine the body's load during training. Based on the comparison and analysis results, it generates practical posture correction suggestions, clearly indicating the limb parts and movement details that need adjustment. A scientifically reasonable training intensity adjustment plan is simultaneously developed to suit the elderly user's physical tolerance. Finally, the evaluation conclusions, posture correction suggestions, and training intensity adjustment plan are systematically integrated to form a complete rehabilitation training evaluation report. The comprehensive information contained in this report provides clear guidance for elderly users' self-training and detailed data support for medical staff or rehabilitation therapists in developing subsequent rehabilitation plans, significantly improving the standardization and effectiveness of elderly rehabilitation yoga training.

[0021] In this embodiment of the invention, the step of extracting feature parameters from the first sensor dataset to obtain an attitude dataset includes: performing noise reduction processing on the first sensor dataset using a Kalman filter noise reduction algorithm to obtain a first standard sensor dataset; obtaining preset attitude key parameter screening rules, performing attitude feature screening processing on the first standard sensor dataset based on the attitude key parameter screening rules to obtain an attitude key feature set; and calling a pre-trained attitude 3D coordinate mapping model to perform coordinate transformation on the attitude key feature set to obtain the attitude dataset.

[0022] In this embodiment, a Kalman filter denoising algorithm is used to eliminate noise in the first sensor dataset. Essentially, it relies on the state prediction and observation update iteration mechanism of the Kalman filter to suppress and remove random noise introduced during sensor acquisition due to environmental interference, equipment errors, and other factors. This preserves and strengthens the effective signals in the original data, resulting in a first standard sensor dataset with higher purity and reliability, laying the foundation for accurate feature extraction in subsequent steps. After obtaining the preset posture key parameter screening rules, posture feature screening is performed on the first standard sensor dataset based on these rules. These posture key parameter screening rules are based on the core needs of elderly rehabilitation yoga training, and a parameter screening system preset according to rehabilitation medicine theory and yoga posture training standards. The system focuses on retaining core parameters related to limb posture, joint range of motion, body center of gravity distribution, and limb movement trajectory, while simultaneously removing redundant data unrelated to posture assessment. The posture key feature set obtained through this screening process simplifies data dimensions and focuses core information, significantly improving the efficiency and relevance of subsequent data processing. Based on this, a pre-trained 3D coordinate mapping model of posture is invoked to perform coordinate transformation on the key feature set of posture. This model has been pre-trained on a large number of elderly rehabilitation yoga posture samples and has a mature mapping relationship between feature parameters and 3D spatial coordinates. It can transform the filtered two-dimensional discrete feature parameters or single-dimensional data into a set of quantifiable and visualized posture coordinates in three-dimensional space, that is, the final posture dataset. This dataset can intuitively represent the spatial position and posture of elderly users' limbs during training, and provide a concrete and measurable core basis for subsequent comparison and evaluation with the rehabilitation standard posture database.

[0023] In this embodiment of the invention, after determining the current mode based on the mode signal, the method further includes: when the current mode is an elderly care monitoring mode, acquiring a second sensor dataset; performing feature parameter extraction processing on the second sensor dataset to obtain a human activity feature dataset; using a fall detection algorithm to judge whether a person has fallen on the human activity feature dataset to obtain a judgment result; and generating an alarm signal when the judgment result indicates that a person has fallen.

[0024] In this embodiment, after entering the elderly monitoring mode, the system will selectively acquire a second sensor dataset. This dataset mainly comes from real-time data collection of the daily activities of elderly users by sensors such as flexible pressure array sensors, triaxial accelerometers, and gyroscopes. It covers multi-dimensional data closely related to the safety of daily activities, including body contact pressure distribution, human motion acceleration, limb rotation angle, movement speed, and environmental spatial location. The acquisition frequency and parameter configuration are optimized for the activity characteristics of elderly users to ensure comprehensive capture of behavioral signals related to potential safety risks. Based on this, feature parameter extraction processing is performed on the second sensor dataset to obtain a human activity feature dataset. The core of this process is to extract core information directly related to fall risk from massive amounts of raw data. By eliminating irrelevant data such as environmental interference and equipment redundancy, the system focuses on retaining key features such as human tilt rate, vertical fall acceleration, amplitude of sudden changes in body position, and continuity of activity trajectory, forming a standardized feature set that can accurately characterize the safety status of human activities, providing high-value data support for subsequent fall detection. Subsequently, a fall detection algorithm is used to judge whether a person has fallen based on the human activity feature dataset. This algorithm relies on a pre-set fall behavior feature model, matching and comparing extracted human activity features with characteristic thresholds and variation patterns of typical fall behaviors to accurately identify the differences between normal activities and fall behaviors, ultimately outputting a clear judgment result as to whether a fall has occurred. When the judgment result indicates a fall has been detected, the system immediately triggers an alarm response mechanism to generate an alarm signal. This signal can be simultaneously pushed to the monitoring personnel's terminal, the community elderly care platform, and on-site audible and visual alarm equipment, achieving multi-dimensional rapid risk warning. This ensures that monitoring personnel and relevant management platforms can promptly obtain risk information and quickly initiate rescue and disposal procedures, significantly improving the safety level of elderly users' daily activities.

[0025] In this embodiment of the invention, the step of extracting feature parameters from the second sensor dataset to obtain a human activity feature dataset includes: using a Kalman filter noise reduction algorithm to perform noise removal processing on the second sensor dataset to obtain a second standard sensor dataset; obtaining preset fall key parameter screening rules, and performing fall feature screening processing on the second standard sensor dataset based on the fall key parameter screening rules to obtain a human activity key feature set; and calling a pre-trained XGBoost feature evaluation model to perform feature sorting and integration processing on the human activity key feature set to obtain the human activity feature dataset.

[0026] In this embodiment, a Kalman filter denoising algorithm is used to eliminate noise in the second sensor dataset. Its core relies on the bidirectional iterative mechanism of state prediction and observation update in Kalman filtering to specifically suppress and remove random noise introduced into the original data due to factors such as sensor accuracy deviations, environmental electromagnetic interference, and non-target motion disturbances from human activity. After obtaining the preset key fall parameter screening rules, fall feature screening is performed on the second standard sensor dataset based on these rules. These key fall parameter screening rules combine the physiological characteristics of the elderly population with the mechanical characteristics of fall behavior, and are based on a parameter screening system preset from clinical fall prevention research results. The focus is on retaining core parameters directly related to fall risk. These parameters are selected around three core dimensions: body posture imbalance characteristics, mechanical impact characteristics, and contact state abrupt change characteristics during a fall in the elderly population. For example, body posture imbalance characteristics may include real-time pressure values ​​of different areas of a yoga mat and pressure acquisition frequency data. These parameters can accurately reflect changes in the body's support state before and after a fall. For instance, during normal activity, the plantar pressure distribution shows a stable and gradual trend, while before a fall, there is a rapid shift in the pressure center of gravity and a sudden increase in local pressure peaks. During a fall, the pressure distribution changes dramatically. Features include a sharp increase or decrease in the contact area; mechanical impact characteristics can include changes in the acceleration of the human body on the yoga mat plane (such as acceleration values ​​in the X, Y, and Z axes). Clinical studies have shown that when elderly people fall, their bodies will exhibit significant free fall or lateral sliding movements due to imbalance, resulting in a negative acceleration peak in the vertical direction far exceeding that of daily activities, and a sudden change in acceleration in the horizontal direction. These parameters are core mechanical indicators for distinguishing between normal activity and falls. Features of abrupt changes in contact state can include parameters such as the body tilt angle and tilt angular velocity. Before a fall, the body usually exhibits a tilt angle that rapidly exceeds the physiological balance threshold and a sharp increase in tilt angular velocity, while during normal activity, changes in body angle are gradual and controllable. These parameters can intuitively characterize the degree and trend of imbalance in body posture. Simultaneously, redundant data irrelevant to fall assessment is systematically removed. The key feature set of human activity obtained through this screening process simplifies data dimensions and focuses core risk information, significantly improving the efficiency and targeting of subsequent data processing. Based on this, the pre-trained XGBoost feature evaluation model is invoked to perform feature ranking and integration processing on the key feature set of human activity. This model has been trained and optimized with a large number of samples of normal activities of the elderly and simulated fall scenarios, and has built a mature feature importance evaluation system. It can assign different weights based on the correlation strength between features and fall behavior, prioritize and integrate key features, remove feature variables with information redundancy or interference, and finally output the human activity feature dataset.

[0027] Please see Figure 2The present invention also provides a yoga mat for elderly rehabilitation monitoring, wherein the yoga mat for elderly rehabilitation monitoring employs any of the elderly rehabilitation monitoring methods described above for operation control; the yoga mat for elderly rehabilitation monitoring includes a control unit, a sensor unit, a data acquisition unit, and a mode switching unit, wherein the data acquisition unit is connected to the control unit, and the sensor unit and the mode switching unit are respectively connected to the data acquisition unit; the data acquisition unit is used to acquire sensor data sets from the sensor unit and mode signals from the mode switching unit, and to send the sensor data sets and the mode signals to the control unit; the control unit is used to determine the current mode based on the mode signals, and when the current mode is rehabilitation yoga mode, to evaluate the rehabilitation effect based on the sensor data to obtain a rehabilitation training evaluation report.

[0028] In this embodiment, the sensor unit provides basic sensing data support for the system, accurately capturing multi-dimensional data related to rehabilitation training and safety monitoring, such as body contact pressure distribution, limb movement acceleration, and body rotation angle, during use by elderly users. Considering the core needs of elderly users for thickness and comfortable fit when using yoga mats, the sensing units in the sensor unit all adopt a sheet-like structure. These sheet-like sensors are both ultra-thin and flexible, allowing for seamless integration with the yoga mat body. This ensures that the thickness of the sensor itself does not alter the flatness and tactile experience of the mat, and maintains a good fit when the user's limbs are in contact with the mat, avoiding interference from rigid or thick sensors on training movements, while also ensuring the stability and accuracy of sensor data acquisition. The mode switching unit is responsible for generating and outputting mode trigger signals. It can generate corresponding mode signals through manual operation or automatic scene recognition, providing a basis for system mode judgment. The data acquisition unit is used to collect multi-dimensional sensor datasets output by the sensor unit in real time, and to synchronously collect mode signals generated by the mode switching unit. After preliminary data normalization, both types of information are stably transmitted to the control unit to ensure the real-time and completeness of data transmission, providing reliable data input for subsequent mode judgment and rehabilitation assessment. The control unit accurately determines the current operating mode and then executes targeted processing logic based on the mode type. When the current mode is determined to be rehabilitation yoga mode, the control unit calls a preset feature extraction algorithm, a three-dimensional posture coordinate mapping model, and a rehabilitation standard posture database to systematically analyze and process the sensor dataset. Through a series of calculations such as posture feature extraction and comparison with standard posture parameters, it completes the rehabilitation effect assessment and finally generates a rehabilitation training assessment report containing core content such as posture correction suggestions and training intensity adjustment plans. This achieves fully automated processing from data acquisition to assessment output, enabling accurate assessment and personalized guidance of rehabilitation effects, fully adapting to the physiological characteristics and rehabilitation training needs of the elderly, improving the standardization and effectiveness of rehabilitation training, and promoting the intelligent upgrading of rehabilitation training scenarios. Further, please refer to Figure 2 It also includes an alarm unit, which is connected to the control unit. The control unit is also used to determine whether a person has fallen based on the sensor data when the current mode is the elderly care monitoring mode, so as to obtain a judgment result. The alarm unit is used to issue an audible and visual alarm when the judgment result is that a person has fallen.

[0029] In this embodiment, after determining that the current mode is elderly care monitoring, the control unit receives multi-dimensional sensor data transmitted by the data acquisition unit, determines whether a person has fallen, and outputs a clear result. When the determination result indicates that a fall has been detected, the control unit immediately sends a trigger signal to the connected alarm unit. Upon receiving the signal, the alarm unit quickly activates the audible and visual alarm mechanism. A dual warning is formed through an audio warning signal that can penetrate environmental interference and a highly recognizable visual warning signal, providing immediate alerts to those in the vicinity. This dual audible and visual alarm mode is well-suited to the environmental characteristics of elderly care scenarios. The audio signal can cover a certain spatial range to alert distant monitoring personnel, while the visual signal can quickly transmit warning information in dimly lit environments or complex scenes. This dual warning dimension significantly improves the efficiency and coverage of alarm information transmission. Furthermore, the integrated design of this functional module allows the yoga mat to retain its basic rehabilitation training attributes while further enhancing its core value in elderly care safety monitoring. It eliminates the need for additional independent alarm equipment, simplifies the system deployment process, and adapts to the usage needs of various scenarios such as home-based elderly care and community-based elderly care.

[0030] In this embodiment of the invention, the sensor unit includes a hand detection sensor group, a waist detection sensor group, a knee joint detection sensor group, and a foot detection sensor group. The hand detection sensor group, the waist detection sensor group, the knee joint detection sensor group, and the foot detection sensor group are respectively connected to the data acquisition unit. The hand detection sensor group is used to collect hand support pressure and balance parameters; the waist detection sensor group is used to collect waist fit and bending curvature; the knee joint detection sensor group is used to collect knee joint load-bearing pressure and bending angle; and the foot detection sensor group is used to collect foot pressure and center of gravity offset.

[0031] In this embodiment, the sensor unit adopts a modular layout of hand detection sensor group, waist detection sensor group, knee joint detection sensor group and foot detection sensor group, combined with the targeted sensor unit configuration of each area, to achieve accurate collection of key part status parameters during the rehabilitation training of elderly users. Specifically, the hand detection sensor group is located on both sides of the front end of the yoga mat, equipped with two sets of plate-type pressure sensing units. By sensing the pressure changes when the hand contacts the mat, it accurately collects hand support pressure parameters. Simultaneously, based on the difference and balance analysis of the pressure data from both sides, it simultaneously acquires balance parameters during hand support, providing data support for assessing the standardization and stability of upper limb support movements. The waist detection sensor group integrates one plate-type flexible posture sensor and two sets of plate-type pressure sensing units, positioned to fit the waist activity area of ​​elderly users. The flexible posture sensor can capture the shape changes during waist bending in real time, converting them into quantifiable bending curvature parameters. The two sets of pressure sensors determine the waist fit by sensing the contact pressure distribution between the waist and the mat, and indirectly characterize the core muscle exertion state by combining pressure change patterns, achieving comprehensive monitoring of key waist training parameters. Orientation data acquisition: The knee joint detection sensor group is equipped with two sets of sheet-type pressure sensors and one set of sheet-type infrared distance sensors, which are deployed corresponding to the knee joint support area. The pressure sensors are responsible for sensing the pressure changes during the knee joint's weight-bearing process and accurately collecting the knee joint's weight-bearing pressure parameters. The infrared distance sensor converts the distance changes during the knee joint's movement into knee joint flexion angle data, thereby achieving synchronous monitoring of the knee joint's force and activity status. The foot detection sensor group adopts an array layout of four sets of sheet-type high-precision pressure sensors, corresponding to four key force-bearing areas: the medial side of the forefoot, the lateral side of the forefoot, the medial side of the heel, and the lateral side of the heel. Through real-time data acquisition from the pressure sensors in each area, the plantar pressure distribution characteristics can be accurately obtained. At the same time, based on the pressure difference of the four sets of sensors and the analysis of the center of gravity migration trajectory, the center of gravity offset parameter is calculated, providing core data for assessing the body's balance status. The modular sensor layout and targeted sensor unit configuration enable precise and focused acquisition of parameters for key parts of rehabilitation training. This overcomes the limitations of single-sensor full-domain acquisition and significantly improves the accuracy and relevance of parameter acquisition for each part, ensuring that the acquired data can truly reflect the training status of key parts.

[0032] In this embodiment of the invention, the sensor unit further includes a fall detection sensor group, which is connected to the data acquisition unit. The fall detection sensor group includes a flexible pressure array sensor, a three-axis accelerometer sensor, and a gyroscope sensor. The flexible pressure array sensor is used to collect the contact area between the human body and the yoga mat, the three-axis accelerometer sensor is used to collect the abrupt change value of the three-axis acceleration, and the gyroscope sensor is used to collect the tilt angle of the human body posture.

[0033] In this embodiment, the flexible pressure array sensor, the three-axis accelerometer, and the gyroscope sensor all employ a sheet-type sensor structure. The flexible pressure array sensor is evenly distributed across the yoga mat, accurately collecting contact area data by sensing changes in pressure distribution at the interface between the human body and the mat. Sudden imbalances in the human posture during a fall cause a rapid change in the contact area with the mat; this data serves as a crucial basis for determining whether the human body has lost its normal support state. The three-axis accelerometer captures the mechanical characteristics of human movement, collecting acceleration abrupt changes in the three axes. Rapid falls and impacts accompanying a fall produce significant acceleration peak abrupt changes; this parameter is a core mechanical indicator distinguishing normal activity from a fall. The gyroscope sensor collects the tilt angle of the human posture in real time. By tracking the angular change from a stable posture to an unbalanced tilt, the degree and rate of posture imbalance can be accurately characterized, providing core data support in the posture dimension for determining the occurrence of a fall. The multi-dimensional data collected by the three sensors is transmitted to the control unit in real time through the data acquisition unit, forming a fall risk perception data system covering contact state, mechanical characteristics, and posture changes. This provides comprehensive and high-value data input for the accurate calculation of the fall judgment algorithm. The collaborative acquisition architecture of multiple types of sensor units breaks through the limitations of a single sensor in perceiving fall behavior. Through complementary verification of multi-dimensional data such as contact area, acceleration change, and tilt angle, the integrity and accuracy of fall behavior feature capture are greatly improved, effectively reducing the risk of misjudgment or missed judgment due to single data deviation.

[0034] Figure 3 This is a schematic diagram of the structure of an elderly rehabilitation monitoring device 300 provided in an embodiment of the present invention. The elderly rehabilitation monitoring device 300 can vary significantly due to differences in configuration or performance. It may include one or more central processing units (CPUs) 310 (e.g., one or more processors) and a memory 320, and one or more storage media 330 (e.g., one or more mass storage devices) for storing application programs 333 or data 332. The memory 320 and storage media 330 can be temporary or persistent storage. The program stored in the storage media 330 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the elderly rehabilitation monitoring device 300. Furthermore, the processor 310 may be configured to communicate with the storage media 330 and execute the series of instruction operations in the storage media 330 on the elderly rehabilitation monitoring device 300 to implement the steps of the elderly rehabilitation monitoring method provided in the above-described method embodiments.

[0035] The elderly rehabilitation monitoring device 300 may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 331, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 3 The structure of the elderly rehabilitation monitoring equipment shown does not constitute a limitation on the elderly rehabilitation monitoring equipment. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0036] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of the elderly rehabilitation monitoring method.

[0037] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the system, device, or unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0038] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0039] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for monitoring and caring for elderly patients during rehabilitation, characterized in that, include: Acquire a pattern signal and determine the current pattern based on the pattern signal; When the current mode is rehabilitation yoga mode, acquire the first sensor dataset; The first sensor dataset is processed to extract feature parameters to obtain an attitude dataset. Obtain a preset standard rehabilitation posture database, evaluate the rehabilitation effect of the posture dataset based on the standard rehabilitation posture database, and obtain a rehabilitation training evaluation report.

2. The method for monitoring elderly rehabilitation according to claim 1, characterized in that, The step of extracting feature parameters from the first sensor dataset to obtain an attitude dataset includes: The first sensor dataset is noise-reducing processed by the Kalman filter noise reduction algorithm to obtain the first standard sensor dataset. Obtain preset attitude key parameter filtering rules, and perform attitude feature filtering processing on the first standard sensor dataset based on the attitude key parameter filtering rules to obtain attitude key feature set; The pose key feature set is transformed by calling a pre-trained pose 3D coordinate mapping model to obtain the pose dataset.

3. The method for monitoring elderly rehabilitation according to claim 1, characterized in that, After determining the current mode based on the mode signal, the process further includes: When the current mode is elderly care monitoring mode, acquire the second sensor dataset; The second sensor dataset is processed to extract feature parameters to obtain a human activity feature dataset; A fall detection algorithm is used to judge whether a person has fallen on the human activity feature dataset, and the judgment result is obtained. An alarm signal is generated when the judgment result indicates that a person has fallen.

4. The method for monitoring elderly rehabilitation according to claim 3, characterized in that, The step of extracting feature parameters from the second sensor dataset to obtain a human activity feature dataset includes: The second sensor dataset is noise-reducing processed using the Kalman filter noise reduction algorithm to obtain the second standard sensor dataset. Obtain preset key fall parameter filtering rules, and perform fall feature filtering processing on the second standard sensor dataset based on the key fall parameter filtering rules to obtain a key feature set of human activity; The pre-trained XGBoost feature evaluation model is invoked to perform feature sorting and integration processing on the key feature set of human activities, thereby obtaining the human activity feature dataset.

5. A yoga mat for elderly rehabilitation monitoring, wherein the yoga mat for elderly rehabilitation monitoring employs the elderly rehabilitation monitoring method as described in any one of claims 1-4 for work control; the yoga mat for elderly rehabilitation monitoring includes a control unit, a sensor unit, a data acquisition unit, and a mode switching unit, wherein the data acquisition unit is connected to the control unit, and the sensor unit and the mode switching unit are respectively connected to the data acquisition unit; the data acquisition unit is used to acquire sensor data sets from the sensor unit and acquire mode signals from the mode switching unit, and send the sensor data sets and the mode signals to the control unit; the control unit is used to determine the current mode based on the mode signals, and when the current mode is a rehabilitation yoga mode, to perform a rehabilitation effect evaluation based on the sensor data to obtain a rehabilitation training evaluation report.

6. The elderly rehabilitation monitoring yoga mat according to claim 5, characterized in that, It also includes an alarm unit, which is connected to the control unit. The control unit is also used to determine whether a person has fallen based on the sensor data when the current mode is the elderly care monitoring mode, so as to obtain a judgment result. The alarm unit is used to issue an audible and visual alarm when the judgment result is that a person has fallen.

7. The elderly rehabilitation monitoring yoga mat according to claim 5, characterized in that, The sensor unit includes a hand detection sensor group, a waist detection sensor group, a knee joint detection sensor group, and a foot detection sensor group. The hand detection sensor group, the waist detection sensor group, the knee joint detection sensor group, and the foot detection sensor group are respectively connected to the data acquisition unit. The hand detection sensor group is used to collect hand support pressure and balance parameters; the waist detection sensor group is used to collect waist fit and bending curvature; the knee joint detection sensor group is used to collect knee joint load-bearing pressure and bending angle; and the foot detection sensor group is used to collect foot pressure and center of gravity offset.

8. The elderly rehabilitation monitoring yoga mat according to claim 7, characterized in that, The sensor unit also includes a fall detection sensor group, which is connected to the data acquisition unit. The fall detection sensor group includes a flexible pressure array sensor, a three-axis accelerometer sensor, and a gyroscope sensor. The flexible pressure array sensor is used to collect the contact area between the human body and the yoga mat, the three-axis accelerometer sensor is used to collect the abrupt change value of the three-axis acceleration, and the gyroscope sensor is used to collect the tilt angle of the human body posture.

9. A geriatric rehabilitation monitoring device, characterized in that, The elderly rehabilitation monitoring device includes: a memory and at least one processor, wherein the memory stores instructions; At least one of the processors invokes the instructions in the memory to cause the geriatric rehabilitation monitoring device to perform the steps of the geriatric rehabilitation monitoring method as described in any one of claims 1-4.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement the various steps of the elderly rehabilitation monitoring method as described in any one of claims 1-4.

Images