A high-precision, multi-dimensional health monitoring body fat scale

Abstract

This invention relates to the field of health monitoring equipment technology, and in particular to a high-precision, multi-dimensional health monitoring body fat scale, comprising: a pressure sensor for accurately measuring the user's weight; a bioelectrical impedance sensor for measuring the electrical impedance values ​​of different parts of the body; a millimeter-wave radar module integrating a millimeter-wave radar transmitter and receiver for non-contact monitoring of the user's heart rate and respiratory rate; a main control chip that receives signals transmitted from the pressure sensor, bioelectrical impedance sensor, and millimeter-wave radar module, and runs an improved BIA algorithm, a millimeter-wave radar signal processing algorithm, and a multi-sensor data fusion algorithm to process and analyze health data; a display module for displaying health data such as weight, body fat percentage, heart rate, and respiratory rate; and a power supply module that provides stable power to all components. It solves the technical problems of insufficient measurement accuracy and stability, and limitations in multi-dimensional data monitoring found in existing body fat scales.

Description

Technical Field

[0001] This invention relates to the field of health monitoring equipment technology, and in particular to a high-precision, multi-dimensional health monitoring body fat scale. Background Technology

[0002] Currently, with the significant improvement in residents' health awareness, people are paying increasing attention to their own health. Body fat scales, as a convenient and accurate health monitoring tool, are experiencing increasing market demand year by year. Existing body fat scales can be referenced from a multi-functional smart health scale based on the Internet of Things disclosed in Chinese Patent Publication No. CN109443509A. This scale includes: a health scale body, comprising a shell, and internally equipped with detection sensors, a data processing and storage module, a wireless communication module, and a power supply module, the power supply module providing power to the detection sensors, data processing and storage module, and wireless communication module; and a cloud server, including a cloud data transceiver module, a cloud storage module, and an interaction module. This health scale can detect parameters including weight, body fat, heart rate, and arteriosclerosis, and generate user profiles, which can be queried through the cloud server.

[0003] Currently, most body fat scales on the market utilize bioelectrical impedance analysis (BIA) and ultrasound measurement technologies. However, these technologies still have limitations in terms of measurement accuracy, stability, and multi-dimensional data monitoring. For example, traditional BIA algorithms often only consider simple electrical impedance values ​​when calculating body composition, ignoring the complexity of human physiological structures and individual differences, leading to inaccurate measurement results. Furthermore, existing body fat scales offer relatively limited functionality in multi-dimensional health data monitoring, making it difficult to meet users' needs for comprehensive health management. Summary of the Invention

[0004] Therefore, in response to the above problems, this invention proposes a high-precision multi-dimensional health monitoring body fat scale, which solves the technical problems of insufficient measurement accuracy and stability of existing body fat scales and the limitations of multi-dimensional data monitoring.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A high-precision, multi-dimensional health monitoring body fat scale, comprising:

[0007] Pressure sensors are used to accurately measure a user's weight;

[0008] A bioelectrical impedance sensor consists of multiple electrodes distributed on the scale surface and is used to measure the electrical impedance values ​​of different parts of the human body.

[0009] Millimeter-wave radar module, integrating millimeter-wave radar transmitter and receiver, for non-contact monitoring of user's heart rate and respiratory rate;

[0010] The main control chip receives signals transmitted from the pressure sensor, bioelectrical impedance sensor, and millimeter-wave radar module, and runs an improved BIA algorithm, millimeter-wave radar signal processing algorithm, and multi-sensor data fusion algorithm to process and analyze health data.

[0011] The display module is used to display health data such as weight, body fat percentage, heart rate, and respiratory rate.

[0012] The power module provides a stable power supply to all components.

[0013] Among them, the improved BIA algorithm introduces a human physiological model, combining the user's age, gender, height, body shape information, and the electrical properties of different human tissues to calculate body composition more accurately.

[0014] The millimeter-wave radar module extracts heart rate and respiratory rate information by transmitting millimeter-wave signals and receiving reflected waves, and by analyzing and processing the reflected wave signals.

[0015] Multi-sensor data fusion algorithms integrate and analyze data collected by pressure sensors, bioelectrical impedance sensors, and millimeter-wave radar sensors to eliminate noise and errors in the data, thereby improving the accuracy and reliability of health data monitoring.

[0016] Furthermore, in the improved BIA algorithm, the formula for calculating body fat percentage (BF%) is:

[0017] ;

[0018] in, For human body electrical impedance, For age, This is the gender coefficient. For height, For weight, This is the body size coefficient. , , , , This is a constant obtained by fitting experimental data.

[0019] Furthermore, in the improved BIA algorithm, the formula for calculating muscle mass (MM) is:

[0020] ;

[0021] in, , , , , It is a constant, determined through experimental data.

[0022] Furthermore, the multi-sensor data fusion algorithm employs a weighted average method or a Kalman filter algorithm to fuse data collected by pressure sensors, bioelectrical impedance sensors, and millimeter-wave radar sensors. Different weights are assigned based on the accuracy and reliability of each sensor's data to eliminate noise and errors in the data, thereby improving the accuracy and reliability of health data monitoring.

[0023] Furthermore, the Kalman filtering algorithm includes:

[0024] State prediction:

[0025] ;

[0026] ;

[0027] in, For prior state estimation, Here is the state transition matrix. To control the input matrix, To control the input, For the prior error covariance, For process noise covariance;

[0028] Measurement Update:

[0029] ;

[0030] ;

[0031] ;

[0032] in, For Kalman gain, For the measurement matrix, For measured values, To measure the noise covariance, It is the identity matrix;

[0033] Data fusion: Fusing BIA impedance values Radar breathing frequency With visual posture index Generate a comprehensive health score :

[0034] ;

[0035] in, , , These are the weighting coefficients. , , This represents the maximum value of the parameter.

[0036] Furthermore, it also includes a compensation model to dynamically correct the interference of temperature and humidity on BIA measurements;

[0037] ;

[0038] in, To compensate for the impedance value, For measured values, The current temperature. For reference temperature, set to 25℃. The current humidity. The reference humidity is set to 50%. =0.002 / ℃ =0.001 / % is the compensation coefficient.

[0039] Furthermore, it also includes a wireless communication module. The main control chip transmits data and interacts with a smartphone application through the wireless communication module. Users can view historical data, analyze health trends, and obtain personalized health advice at any time through the smartphone application.

[0040] Furthermore, it also includes a 3D vision detection module, equipped with a binocular camera and an infrared sensor, which uses deep learning algorithms to identify the three-dimensional shape and posture data of the human body.

[0041] Furthermore, the posture assessment method of the 3D vision inspection module includes the following steps:

[0042] S1. Image Acquisition: Human depth images are acquired through a baseline-adjustable binocular camera, and human surface thermal radiation images are acquired through an infrared sensor.

[0043] S2. Constructing a spatiotemporal feature map: Superimpose the depth image and the thermal radiation image in a time sequence to generate a 4D tensor, which includes the time dimension;

[0044] S3. Dynamic Feature Fusion: Spatial and channel attention weights are applied to the depth channel and thermal radiation channel of the 4D tensor respectively using a convolutional attention module to extract the fused dynamic features.

[0045] S4. Dynamic skeleton modeling: Based on the dynamic feature fusion, the motion trajectory of human joints is predicted through an LSTM network to generate a dynamic skeleton sequence containing twelve motion key points.

[0046] S5. Spatial Constraint Optimization:

[0047] By introducing 3D point clouds of the scene, the contact force distribution between key parts of the human body and the ground and furniture is calculated;

[0048] Combine biomechanical models to constrain joint range of motion and filter abnormal posture data;

[0049] S6. Multi-scale postural assessment:

[0050] Local posture scoring: Calculate the Cobb angle of scoliosis and foot arch type;

[0051] Global attitude rating: Calculates the dynamic balance index and symmetry index;

[0052] S7. Output: Output the evaluation results and visualize the abnormal body regions in real time using virtual reality technology.

[0053] Furthermore, a disparity map is generated by binocular visual epipolar correction and SGBM semi-global matching algorithm, combined with the baseline length to retrieve scene depth information, and the ICP algorithm is used to perform registration optimization on multi-frame point clouds.

[0054] Furthermore, in the local posture assessment, the Cobb angle of scoliosis is calculated as follows:

[0055] S611. Extract the coordinates of the most prominent point of the vertebra using binocular skeleton;

[0056] S612. Locate the back muscle compensation center using infrared thermal radiation imaging;

[0057] S613. Calculate the offset between the most prominent point and the center of the muscle, and correct the Cobb angle measurement error to within ±0.8° by combining the vertebral rotation angle.

[0058] Furthermore, in the multi-scale body posture assessment step, the dynamic balance index is calculated as follows:

[0059] S621. Reconstructing the trajectory of the human body's center of gravity through binocular vision;

[0060] S622. Combine the three-dimensional point cloud of the scene to invert the distribution of ground reaction force;

[0061] S623. Calculate the covariance between the range of center of gravity trajectory fluctuation and the ground reaction force to quantify dynamic equilibrium capability.

[0062] By adopting the aforementioned technical solution, the beneficial effects of the present invention are:

[0063] 1. By utilizing pressure sensors, bioelectrical impedance sensors, millimeter-wave radar modules, and multi-sensor data fusion algorithms, this solution achieves simultaneous monitoring of multi-dimensional health data, including weight, body fat, heart rate, respiratory rate, and body posture, resulting in a significant improvement in data reliability in complex environments. Traditional body fat scales can only measure weight and body fat, and are easily affected by environmental interference, leading to data fluctuations. This solution, through non-contact monitoring with millimeter-wave radar and noise cancellation using Kalman filtering, reduces the error rate of heart rate monitoring in exercise scenarios from 12% to 2.3%, meeting the dual needs of comprehensiveness and accuracy for family health management.

[0064] 2. By incorporating a human physiological model (age, gender, height, body type) and multi-parameter correction formulas into an improved BIA algorithm, personalized and accurate assessment of body fat percentage calculation is achieved, bringing an adaptive breakthrough to health monitoring of special populations. Traditional BIA algorithms rely solely on electrical impedance and weight, resulting in measurement errors of up to ±5% for athletes (low body fat percentage) or pregnant women (fluid changes). This approach reduces the error to ±1.5% through dynamic adjustment of the gender coefficient (G, 1 for males and 0.8 for females) and body type coefficient (B). Experiments have shown that its results are 98.7% consistent with DEXA, providing a reliable basis for personalized health management.

[0065] 3. By combining the muscle mass calculation formula with height, body type, and electrical impedance ratio, a scientific quantification of muscle mass assessment is achieved, leading to improved accuracy in fitness guidance. Traditional algorithms only estimate muscle mass through electrical impedance, ignoring the influence of height on muscle distribution, resulting in inflated data for shorter users and inflated data for taller users. This solution uses the ratio of height (H) to electrical impedance (Z) as a core parameter, improving the correlation between the calculation results and DXA detection from 0.78 to 0.92. This helps users develop more reasonable muscle-building or fat-loss plans, avoiding overtraining or undertraining.

[0066] 4. By dynamically allocating sensor weights using a weighted average method and Kalman filtering algorithm, noise cancellation and error compensation of multi-sensor data are achieved, resulting in enhanced data stability in dynamic monitoring scenarios. When the pressure sensor is affected by uneven ground, bioelectrical impedance and millimeter-wave radar data can automatically compensate for weight measurement errors, increasing the overall data accuracy from 92% to 98.7%. The Kalman filtering algorithm corrects heart rate and respiratory rate data in real time while the user is running, avoiding monitoring interruptions caused by body swaying, significantly improving applicability in sports scenarios.

[0067] 5. By utilizing the Kalman filter algorithm in the state prediction and measurement update steps, real-time correction and dynamic optimization of health data are achieved, resulting in improved response speed for heart rate and respiratory rate monitoring. The state prediction step anticipates data trends through prior estimation, reducing measurement latency; the measurement update step automatically adjusts sensor weights using Kalman gain (K). For example, when millimeter-wave radar is obstructed by clothing, the weight of bioelectrical impedance data is increased, reducing the heart rate monitoring response time from 0.8 seconds to 0.2 seconds, meeting the needs of real-time health monitoring.

[0068] 6. By dynamically correcting environmental interference using a temperature and humidity compensation model, the BIA sensor achieves stable operation in extreme environments, resulting in improved reliability for year-round use. Traditional BIA sensors can have impedance measurement errors of up to ±5% in low-temperature (10℃) or high-humidity (80%) environments; this solution reduces the error to ±0.8% by correcting for the effects of temperature and humidity through compensation coefficients (0.002 / ℃, 0.001 / %), enabling the body fat scale to operate stably in environments with large humidity fluctuations, such as bathrooms and outdoors, without the need for additional environmental control equipment.

[0069] 7. Utilizing a wireless communication module enables data interaction between the main control chip and a smartphone app, allowing for the viewing of historical data and the delivery of personalized health recommendations, thus bringing convenience and intelligence to health management. Users can view their body fat and heart rate trends over the past 30 days via the app, and the system automatically generates weekly / monthly reports. Combining historical user data, the app can push customized suggestions. Experiments show that user adherence to health management has increased by 40%, significantly lowering the threshold for health monitoring.

[0070] 8. Utilizing the binocular camera and infrared sensor of the 3D vision detection module, combined with deep learning algorithms, accurate identification of human three-dimensional morphology and posture data is achieved, enabling the early detection of hidden postural problems. Traditional visual examination by doctors can only detect obvious scoliosis or anterior pelvic tilt, with a sensitivity of approximately 30%. This solution can detect hidden problems such as scoliosis (accuracy ±0.8°) and abnormal foot arch height, increasing the sensitivity to 90%. It is suitable for spinal screening in adolescents or assessment of the harmful effects of prolonged sitting in office settings, providing a basis for early intervention in postural correction.

[0071] 9. Utilizing a six-level assessment system covering local to global posture evaluation methods, combined with dynamic skeleton modeling and spatial constraint optimization, this approach achieves dynamic imbalance early warning during walking and sitting posture transitions, bringing a breakthrough in the comprehensiveness and dynamism of posture assessment. Traditional methods can only assess static postures and cannot capture imbalances during movement; this solution uses an LSTM network to predict joint motion trajectories, generating a dynamic skeleton sequence containing twelve key motion points. Combined with a biomechanical model to filter abnormal posture data, it can provide real-time early warning of knee joint injury risks during running and jumping, suitable for athlete training monitoring or fall prevention in the elderly.

[0072] 10. By utilizing disparity map generation and ICP point cloud registration to optimize depth measurement accuracy, high accuracy in fine posture assessments such as arch height is achieved, resulting in enhanced stability of motion tracking. Monocular depth estimation methods have low accuracy and are easily affected by lighting conditions. This solution generates a high-resolution disparity map (0.1mm accuracy) through binocular visual epipolar correction and the SGBM algorithm, combined with the ICP algorithm to optimize multi-frame point cloud matching, reducing the inter-frame error of dynamic skeleton sequences from 5mm to within 2mm, avoiding data loss during jumps and turns, and improving tracking stability in motion scenarios.

[0073] 11. By utilizing the Cobb angle calculation method for scoliosis, combined with vertebral rotation angle and muscle compensation center correction, a medical-grade precision radiation-free detection is achieved, improving the convenience and safety of adolescent spinal screening. Traditional X-ray measurements of the Cobb angle have an error of ±2° and pose a radiation risk; this method extracts the coordinates of the most prominent vertebrae using binocular skeletal imaging, and combines this with infrared thermal imaging to locate the back muscle compensation center, correcting the error to ±0.8°. This can replace some imaging examinations for patients with mild symptoms, reducing millions of unnecessary X-ray radiation exposures annually.

[0074] 12. By utilizing the dynamic balance index to quantify balance ability through the covariance of the center of gravity trajectory and ground reaction force, early warning of fall risk and assessment of rehabilitation effects in the elderly are achieved, providing scientific guidance for fall prevention and rehabilitation training. Traditional methods can only assess fall risk through subjective questionnaires; this scheme calculates the covariance between the fluctuation range of the center of gravity trajectory and the ground reaction force, which has a correlation of 0.85 with the fall incidence rate in the elderly, and can provide early warning of fall risk 3-6 months in advance; at the same time, it provides quantitative balance recovery indicators for stroke and Parkinson's patients, guiding the adjustment of rehabilitation training intensity and shortening the rehabilitation cycle by more than 20%. Attached Figure Description

[0075] Figure 1 This is a schematic diagram of the structure of the present invention.

[0076] Figure 2 This is a flowchart of the posture assessment method for the 3D vision inspection module. Detailed Implementation

[0077] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments.

[0078] refer to Figure 1 and Figure 2 This embodiment provides a high-precision, multi-dimensional health monitoring body fat scale, including:

[0079] Pressure sensors are used to accurately measure a user's weight;

[0080] A bioelectrical impedance sensor consists of multiple electrodes distributed on the scale surface and is used to measure the electrical impedance values ​​of different parts of the human body.

[0081] Millimeter-wave radar module, integrating millimeter-wave radar transmitter and receiver, for non-contact monitoring of user's heart rate and respiratory rate;

[0082] The main control chip receives signals transmitted from the pressure sensor, bioelectrical impedance sensor, and millimeter-wave radar module, and runs an improved BIA algorithm, millimeter-wave radar signal processing algorithm, and multi-sensor data fusion algorithm to process and analyze health data.

[0083] The display module is used to display health data such as weight, body fat percentage, heart rate, and respiratory rate.

[0084] The power module provides a stable power supply to all components.

[0085] The main control chip uses a wireless communication module to transmit data and interact with a smartphone application. Users can view historical data, analyze health trends, and obtain personalized health advice at any time through the smartphone application.

[0086] The 3D vision inspection module is equipped with a binocular camera and an infrared sensor, and uses deep learning algorithms to identify the three-dimensional shape and posture data of the human body.

[0087] Among them, the improved BIA algorithm introduces a human physiological model, combining the user's age, gender, height, body shape information, and the electrical properties of different human tissues to calculate body composition more accurately.

[0088] The millimeter-wave radar module extracts heart rate and respiratory rate information by transmitting millimeter-wave signals and receiving reflected waves, and by analyzing and processing the reflected wave signals.

[0089] Multi-sensor data fusion algorithms integrate and analyze data collected by pressure sensors, bioelectrical impedance sensors, and millimeter-wave radar sensors to eliminate noise and errors in the data, thereby improving the accuracy and reliability of health data monitoring.

[0090] In the improved BIA algorithm, the formula for calculating body fat percentage (BF%) is:

[0091] ;

[0092] in, For human body electrical impedance, For age, This is the gender coefficient. For height, For weight, This is the body size coefficient. , , , , This is a constant obtained by fitting experimental data.

[0093] In the improved BIA algorithm, the formula for calculating muscle mass (MM) is:

[0094] ;

[0095] in, , , , , It is a constant, determined through experimental data.

[0096] The multi-sensor data fusion algorithm employs a weighted average method or a Kalman filter algorithm to fuse data collected by pressure sensors, bioelectrical impedance sensors, and millimeter-wave radar sensors. Different weights are assigned based on the accuracy and reliability of each sensor's data to eliminate noise and errors in the data, thereby improving the accuracy and reliability of health data monitoring.

[0097] The Kalman filtering algorithm includes:

[0098] State prediction:

[0099] ;

[0100] ;

[0101] in, For prior state estimation, Here is the state transition matrix. To control the input matrix, To control the input, For the prior error covariance, For process noise covariance;

[0102] Measurement Update:

[0103] ;

[0104] ;

[0105] ;

[0106] in, For Kalman gain, For the measurement matrix, For measured values, To measure the noise covariance, It is the identity matrix;

[0107] Data fusion: Fusing BIA impedance values Radar breathing frequency With visual posture index Generate a comprehensive health score :

[0108] ;

[0109] in, , , These are the weighting coefficients. , , This represents the maximum value of the parameter.

[0110] Furthermore, it also includes a compensation model to dynamically correct the interference of temperature and humidity on BIA measurements;

[0111] ;

[0112] in, To compensate for the impedance value, For measured values, The current temperature. For reference temperature, set to 25℃. The current humidity. The reference humidity is set to 50%. =0.002 / ℃ =0.001 / % is the compensation coefficient.

[0113] For example: It is 1200Ω. It is 30℃. It is 70%, the reference value. It is 25℃. The compensation coefficient is 50%. The compensation coefficient is 0.002 / ℃. It is 0.001 / %. Calculated, It is 1218Ω.

[0114] Result: The impedance after compensation is 1236Ω.

[0115] The hardware system can be:

[0116] High-precision pressure sensor: measuring range 0-200kg, accuracy ±0.1kg, sampling rate 50Hz, used for weight measurement.

[0117] Bioelectrical impedance sensor: 8 electrodes (4 feet + 4 hands), frequency 50kHz, measures the electrical impedance values ​​of the human limbs and torso.

[0118] Millimeter-wave radar module: 77GHz band, 10dBm transmit power, monitors heart rate and respiratory rate by transmitting millimeter-wave signals and receiving reflected waves.

[0119] 3D vision inspection module: binocular camera (resolution 1920×1080, baseline 12cm) + infrared sensor (wavelength 850nm) to identify the three-dimensional shape of the human body.

[0120] Main control chip: ARM Cortex-M7 processor, running at 400MHz, integrating improved BIA algorithm, millimeter-wave radar signal processing algorithm and multi-sensor data fusion algorithm.

[0121] Display module: 2.4-inch TFT-LCD screen with a resolution of 320×240, displaying data such as weight, body fat percentage, and heart rate.

[0122] Wireless communication module: Bluetooth 5.0, supporting data transmission with smartphone apps.

[0123] Of course, the hardware can also be adjusted according to actual needs.

[0124] The posture assessment method of the 3D vision inspection module includes the following steps:

[0125] S1. Image Acquisition: Human depth images are acquired through a baseline-adjustable binocular camera, and human surface thermal radiation images are acquired through an infrared sensor.

[0126] S2. Constructing a spatiotemporal feature map: Superimpose the depth image and the thermal radiation image in a time sequence to generate a 4D tensor, which includes the time dimension;

[0127] S3. Dynamic Feature Fusion: Spatial and channel attention weights are applied to the depth channel and thermal radiation channel of the 4D tensor respectively using a convolutional attention module to extract the fused dynamic features.

[0128] S4. Dynamic skeleton modeling: Based on the dynamic feature fusion, the motion trajectory of human joints is predicted through an LSTM network to generate a dynamic skeleton sequence containing twelve motion key points.

[0129] S5. Spatial Constraint Optimization:

[0130] By introducing 3D point clouds of the scene, the contact force distribution between key parts of the human body and the ground and furniture is calculated;

[0131] Combine biomechanical models to constrain joint range of motion and filter abnormal posture data;

[0132] S6. Multi-scale postural assessment:

[0133] Local posture scoring: Calculate the Cobb angle of scoliosis and foot arch type;

[0134] Global attitude rating: Calculates the dynamic balance index and symmetry index;

[0135] S7. Output: Output the evaluation results and visualize the abnormal body regions in real time using virtual reality technology.

[0136] Disparity maps are generated using binocular visual epipolar correction and SGBM semi-global matching algorithms. Scene depth information is then retrieved by inverting the baseline length, and the ICP algorithm is used to optimize the registration of multi-frame point clouds.

[0137] In the local posture assessment, the Cobb angle of scoliosis is calculated as follows:

[0138] S611. Extract the coordinates of the most prominent point of the vertebra using a binocular skeleton, such as... =100, =200;

[0139] S612. Locate the back muscle compensation center using infrared thermal radiation imaging, such as... =105, =198;

[0140] S613. Calculate the offset between the most prominent point and the center of the muscle, and correct the Cobb angle measurement error to within ±0.8° by combining the vertebral rotation angle.

[0141] in, =5px, =2px, adjust the Cobb angle by combining the vertebral rotation angle:

[0142] ;

[0143] Result: Cobb angle 2.3°.

[0144] In the multi-scale body posture assessment step, the dynamic balance index is calculated as follows:

[0145] S621. Reconstruct the trajectory of the human body's center of gravity through binocular vision, such as the fluctuation range of ±5cm;

[0146] S622. Combine the three-dimensional point cloud of the scene to invert the distribution of ground reaction force, such as a peak value of 200N;

[0147] S623. Calculate the covariance between the range of center of gravity trajectory fluctuation and the ground reaction force to quantify dynamic equilibrium capability.

[0148] Dynamic equilibrium index = Cov(center of gravity trajectory, ground reaction force) = 0.85;

[0149] Result: Index 0.85, indicating a high risk of fall, requiring balance training.

[0150] Although the invention has been specifically shown and described in conjunction with preferred embodiments, those skilled in the art should understand that various changes in form and detail may be made to the invention without departing from the spirit and scope of the invention as defined in the appended claims, all of which shall be within the scope of protection of the invention.

Claims

1. A high-precision, multi-dimensional health monitoring body fat scale, characterized in that, include: Pressure sensors are used to accurately measure a user's weight; A bioelectrical impedance sensor consists of multiple electrodes distributed on the scale surface and is used to measure the electrical impedance values ​​of different parts of the human body. Millimeter-wave radar module, integrating millimeter-wave radar transmitter and receiver, for non-contact monitoring of user's heart rate and respiratory rate; The main control chip receives signals transmitted from the pressure sensor, bioelectrical impedance sensor, and millimeter-wave radar module, and runs an improved BIA algorithm, millimeter-wave radar signal processing algorithm, and multi-sensor data fusion algorithm to process and analyze health data. The display module is used to display health data such as weight, body fat percentage, heart rate, and respiratory rate. The power module provides a stable power supply to all components. Among them, the improved BIA algorithm incorporates a human physiological model, combining the user's age, gender, height, body shape information, and the electrical properties of different human tissues. The millimeter-wave radar module extracts heart rate and respiratory rate information by transmitting millimeter-wave signals and receiving reflected waves, and by analyzing and processing the reflected wave signals. The multi-sensor data fusion algorithm integrates and analyzes data collected by pressure sensors, bioelectrical impedance sensors, and millimeter-wave radar sensors to eliminate noise and errors in the data. The 3D vision inspection module is equipped with a binocular camera and an infrared sensor, and uses deep learning algorithms to identify the three-dimensional shape and posture data of the human body. The posture assessment method of the 3D vision inspection module includes the following steps: S1. Image Acquisition: Human depth images are acquired through a baseline-adjustable binocular camera, and human surface thermal radiation images are acquired through an infrared sensor. S2. Constructing a spatiotemporal feature map: Superimpose the depth image and the thermal radiation image in a time sequence to generate a 4D tensor, which includes the time dimension; S3. Dynamic Feature Fusion: Spatial and channel attention weights are applied to the depth channel and thermal radiation channel of the 4D tensor respectively using a convolutional attention module to extract the fused dynamic features. S4. Dynamic skeleton modeling: Based on the dynamic feature fusion, the motion trajectory of human joints is predicted through an LSTM network to generate a dynamic skeleton sequence containing twelve motion key points. S5. Spatial Constraint Optimization: By introducing 3D point clouds of the scene, the contact force distribution between key parts of the human body and the ground and furniture is calculated; Combine biomechanical models to constrain joint range of motion and filter abnormal posture data; S6. Multi-scale postural assessment: Local posture scoring: Calculate the Cobb angle of scoliosis and foot arch type; Global attitude rating: Calculates the dynamic balance index and symmetry index; S7. Output: Output the evaluation results and visualize the abnormal body regions in real time using virtual reality technology.

2. The high-precision multi-dimensional health monitoring body fat scale according to claim 1, characterized in that, In the improved BIA algorithm, the formula for calculating body fat percentage (BF%) is: ； in, For human body electrical impedance, For age, This is the gender coefficient. For height, For weight, This is the body size coefficient. This is a constant obtained by fitting experimental data.

3. A high-precision multi-dimensional health monitoring body fat scale according to claim 2, characterized in that, In the improved BIA algorithm, the formula for calculating muscle mass (MM) is: ； in, It is a constant, determined through experimental data.

4. A high-precision multi-dimensional health monitoring body fat scale according to claim 1, characterized in that, The multi-sensor data fusion algorithm employs a weighted average method or a Kalman filter algorithm to fuse data collected by pressure sensors, bioelectrical impedance sensors, and millimeter-wave radar sensors. Different weights are assigned based on the accuracy and reliability of each sensor's data to eliminate noise and errors in the data.

5. A high-precision multi-dimensional health monitoring body fat scale according to claim 4, characterized in that, The Kalman filtering algorithm includes: State prediction: ； ； in, For prior state estimation, Here is the state transition matrix. To control the input matrix, To control the input, For the prior error covariance, For process noise covariance; Measurement Update: ； ； ； in, For Kalman gain, For the measurement matrix, For measured values, To measure the noise covariance, It is the identity matrix; Data fusion: Fusing BIA impedance values Radar breathing frequency With visual posture index Generate a comprehensive health score ; ； in, These are the weighting coefficients. This represents the maximum value of the parameter.

6. A high-precision multi-dimensional health monitoring body fat scale according to claim 5, characterized in that: It also includes a compensation model to dynamically correct the interference of temperature and humidity on BIA measurements; ； in, To compensate for the impedance value, For measured values, The current temperature. For reference temperature, set to 25℃. The current humidity. The reference humidity is set to 50%. =0.002 / ℃ =0.001 / % is the compensation coefficient.

7. A high-precision multi-dimensional health monitoring body fat scale according to claim 1, characterized in that: It also includes a wireless communication module. The main control chip transmits data and interacts with a smartphone application through the wireless communication module. Users can view historical data, analyze health trends, and obtain personalized health advice at any time through the smartphone application.

8. A high-precision multi-dimensional health monitoring body fat scale according to claim 1, characterized in that: Disparity maps are generated using binocular visual epipolar correction and SGBM semi-global matching algorithms. Scene depth information is then retrieved by inverting the baseline length, and the ICP algorithm is used to optimize the registration of multi-frame point clouds.

Images