Sitting posture detection system and method

Abstract

The invention discloses a sitting posture detection system and method, and relates to the technical field of sitting posture detection, and the system comprises a pressure sensor which is used for measuring the sitting posture pressure of a human body according to the deformation degree of a seat; the microcontroller is used for collecting pressure distribution information on the surface of the seat and removing noise in the information through a Kalman filtering algorithm; the cloud server is used for carrying out real-time online analysis on the sitting posture condition and providing a data storage function; the terminal equipment and the application program are used for presenting the human body sitting posture analysis result. According to the technical scheme of the invention, real-time detection and analysis of the human body sitting posture condition can be realized, real-time auxiliary reminding can be carried out for unhealthy sitting postures, the recognition precision of the human body sitting posture type is improved, and the method has a relatively high application value.

Description

Technical Field

[0001] The present invention relates to the technical field of sitting posture detection, and in particular to a sitting posture detection system and method. Background Art

[0002] Nowadays, due to the needs of work and study, more and more people need to sit at a desk for a long time. As the working hours increase, the fatigue of the human body gradually increases, and the sitting posture is more likely to be deformed, so the human body will inadvertently maintain a bad sitting posture for a long time. Long-term bad sitting posture will cause compression of the human nerves, leading to deformation of the soft tissues of the waist and neck, thereby increasing the incidence of scoliosis, cervical spondylosis, lumbar and back muscle strain and other musculoskeletal system diseases.

[0003] In daily life, maintaining a healthy sitting posture and shortening the sitting time are of great significance to people's physical and mental health and disease prevention. Most of the existing sitting posture correction technologies are carried out in formal places such as hospitals. These treatment methods often have short treatment time and a high possibility of rebound. Therefore, it is very important to be able to quickly detect and remind the sitting posture of the human body in daily life. Let people understand their long-term sitting posture, which helps people change bad sitting posture from a subjective consciousness and prevent related health problems. At present, most of the existing sitting posture detection methods have defects in some aspects. For example, the sitting posture detection method based on piezoelectric information and image information can quickly provide the sitting posture of the human body, but the detection method of this method is too single, and the detection effect will be affected by factors such as personal weight and sitting habits, and the detection result is likely to have errors. Therefore, on this basis, the present application proposes a sitting posture detection system and method that ensures real-time detection while maintaining high precision. Summary of the invention

[0004] Technical Purpose In order to solve the above problems, the purpose of the present invention is to provide a sitting posture detection system and method, which can not only provide accurate human sitting posture detection, but also have low cost and low energy consumption, and can be easily deployed on various forms of seats and similar scenes. The system and method have a wide range of application scenarios, making it possible for the sitting posture detection method with pressure as the detection target to enter the consumer product market.

[0005] Technical Solution In order to achieve the above-mentioned purpose, the present invention provides a sitting posture detection system and method, which adopts polyvinylidene fluoride material combined with conductive silver paste with anti-stretching properties to design a pressure sensor, uses a Kalman filter algorithm to process the noise in the data, and sets a voltage divider to convert the resistivity change caused by the pressure change into a voltage signal. Finally, the sitting posture is detected by measuring the pressure distribution information on the seat surface, thereby realizing the recognition of the human body's sitting posture.

[0006] In a first aspect, the present invention provides a sitting posture detection system, comprising: A pressure sensor is used to measure the sitting pressure of a human body according to the degree of deformation of the seat; A microcontroller, which is used to collect pressure distribution information on the seat surface and remove noise from the information through a Kalman filter algorithm; Frequency domain analysis module, used to extract key features related to human sitting posture in pressure information; A sitting posture recognition module is used to identify the sitting posture type based on the extracted key features; Cloud server, used for real-time online analysis of sitting posture and providing data storage function; The terminal device and application are used to present the human sitting posture analysis results.

[0007] Further, the pressure sensor includes polyvinylidene fluoride, conductive silver paste, insulating coating material, PVC coated fabric and sponge cushion; Among them, polyvinylidene fluoride has high conductivity and stability, and can maintain its performance as the base fabric bends and stretches, enabling the sensor to accurately capture pressure changes and adapt to different usage conditions.

[0008] The conductive silver paste is a conductive wire, which is printed on the insulating oil coating of the base fabric and is used to connect the pressure sensor and the microcontroller; The conductive silver paste has a resistance of less than 10 ohms per square foot; The insulating coating material is a substrate; The tensile strength of the conductive silver paste and the insulating coating material is not less than 50%. The tensile strength of the conductive silver paste and the insulating coating material helps to improve the durability of the sensor and reduce problems such as wear and tear and short circuit caused by use; The PVC coated fabric is the base fabric, and the insulating coating material is coated on the PVC coated fabric to prevent the conductive silver paste from being lost during use.

[0009] Furthermore, the resistivity of the pressure sensor changes with the change of the piezoresistive material.

[0010] Furthermore, the polyvinylidene fluoride is a piezoelectric material having a bending modulus of 40,000-50,000 pounds per square inch, a thickness of 104 microns, a volume resistivity of less than 500 ohm·meter, and a tensile strength of 1,700-2,000 pounds per square inch. Furthermore, the non-connected part circuits of the pressure sensor and the microcontroller are coated with an insulating coating material to avoid wear, short circuit, etc. during use.

[0011] Furthermore, the PVC coated fabric is bonded onto the sponge cushion to prevent sliding during sitting posture detection.

[0012] Furthermore, when the pressure sensor is subjected to pressure, it becomes thinner as a whole, its shape bends, and current flows through the pressure sensor sheet to the microcontroller.

[0013] The sensor is low-cost and suitable for mass production; the design of the sensor takes into account deployment and maintenance issues in practical applications and is suitable for use in various forms of seats and similar scenarios.

[0014] Further, the system also includes a voltage divider for measuring and configuring the resistance change of the pressure sensor and matching the impedance; The output voltage of the pressure sensor is calculated as follows:

[0015] In the formula, is the output voltage of the pressure sensor; is the resistance of the pressure sensor; is the power supply voltage; is the resistance of the voltage divider.

[0016] Furthermore, Preferably +3.7 volts, Preferably 1000 ohms.

[0017] Furthermore, the total area of ​​the pressure sensor is 450*450 mm, and the area of ​​a single pressure sensor piece is 30*30 mm.

[0018] Furthermore, the pressure range of each pressure sensor sheet is 4.86 kg–9.72 kg.

[0019] Furthermore, the microcontroller removes noise sources including flicker, mechanical and stretching stimulation, temperature change and calibration error-induced noise through the Kalman filter algorithm.

[0020] Furthermore, the step of removing noise by the Kalman filter algorithm includes: estimating the current state variable and updating the estimated value using a weighted average value, giving more weight to the estimated value with higher certainty.

[0021] Furthermore, the Kalman filter algorithm takes the current input measurement value and the previous calculation state as input.

[0022] Furthermore, the system establishes communication with the IoT platform via WiFi and communicates with the application via Bluetooth.

[0023] Furthermore, the core processor of the microcontroller is ESP32; The microcontroller is preset with a time threshold. When the system determines that the duration of a human body being in an unhealthy sitting posture exceeds the time threshold, the microcontroller issues a vibration reminder.

[0024] Furthermore, the system includes 16 pressure collection points, which are connected to the mainboard via an analog multiplexer. The multiplexer allows the data of the pressure sensor to be transmitted to the ESP32 processor via five analog input pins in sequence.

[0025] The pressure sensor can accurately detect the pressure distribution on the seat surface, which improves the accuracy of human sitting posture detection, realizes real-time conversion and reading of pressure data, effectively reduces the noise received by the sensor when working, and improves the stability and reliability of the signal.

[0026] Furthermore, the data acquisition frequency of the system is 5 Hz.

[0027] Furthermore, the power source used by the system is a DC 3.7V 3000mAh lithium battery. The duty cycle between WiFi and Bluetooth is reduced to reduce system energy consumption, and data collection is performed in a sleep and work cycle.

[0028] Furthermore, the system supports reporting data in real-time or scheduled form.

[0029] Furthermore, the system activates a sleep mode to reduce energy consumption if no change in pressure acting on the sensor is detected within a threshold time except for the time when the system reports data.

[0030] The system can provide users with real-time visual information, historical information and feedback reports through terminal devices and applications, allowing users to intuitively understand their sitting posture and make adjustments accordingly.

[0031] Furthermore, the frequency domain analysis module performs frequency domain characteristic analysis on the pressure information processed by the Kalman filter algorithm to extract key features that have an impact on human sitting posture recognition; By collecting the pressure distribution information of the pressure sensor and recording the continuous pressure degrees to form time series data; The collected pressure information is preliminarily processed through the Kalman filter algorithm, and the data is standardized to eliminate the influence of different dimensions; The preprocessed pressure time series data are converted from the time domain to the frequency domain by fast Fourier transform, and the frequency characteristics of the sitting posture are represented by the frequency components and the corresponding amplitudes; The frequency components representing the sitting posture characteristics are determined based on the calculation results of the fast Fourier transform, and the frequency components are matched with the specific sitting posture.

[0032] The formula for the fast Fourier transform is as follows:

[0033] In the formula, For sequence No. The complex representation of the frequency components; is the first samples, indicating that the pressure sensor is The number of degrees at a time point; For the The weight factor of each sample is used to adjust the importance of pressure readings at different time points; is the total number of samples, indicating the number of data points collected; is an imaginary unit; is a complex exponential function, which represents the rotation factor, where is the frequency index.

[0034] By analyzing the periodicity and frequency characteristics of sitting posture, the recognition rate of sitting posture detection is improved; the fast Fourier algorithm is efficient and can quickly process large amounts of data, which helps to improve the real-time performance of the system; by identifying important frequency components in the signal, the amount of data stored and transmitted is reduced.

[0035] Furthermore, the sitting posture recognition module defines an optimal hyperplane to separate different types of sitting posture data in the sitting posture feature space, thereby identifying different sitting posture types. The application of this module enables the system to more accurately identify and classify different sitting postures, thereby improving the accuracy of detection; the support vector machine has a certain robustness to outliers in the data, and can provide relatively stable classification results even in the case of poor data quality; the implementation and deployment costs are relatively low, which helps to reduce the cost of the entire system.

[0036] In a second aspect, the present invention further provides a sitting posture detection method, which is based on the system described in the first aspect, and comprises: The method comprises: Measure the sitting pressure of the human body according to the deformation degree of the seat; Collect the pressure distribution information of the seat surface and remove the noise in the information through Kalman filter algorithm; Extract key features related to human sitting posture in pressure information; Identify the sitting posture type based on the extracted key features; Analyze the sitting posture in real time and present the analysis results of human sitting posture.

[0037] In a third aspect, the present invention also provides a computer device, comprising a processor and a memory, wherein the processor is connected to the memory, the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory, so that the computer device performs at least one step of the aforementioned sitting posture detection method.

[0038] In a fourth aspect, the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, at least one step of the aforementioned sitting posture detection method is implemented.

[0039] The present invention uses polyvinylidene fluoride material combined with conductive silver paste with anti-stretching properties to design a pressure sensor; uses the Kalman filter algorithm to process the noise in the data, sets a voltage divider to convert the resistivity change caused by the pressure change into a voltage signal, and detects the sitting posture by measuring the pressure distribution information on the seat surface; performs frequency domain characteristic analysis on the pressure information processed by the Kalman filter algorithm to extract key features that affect human sitting posture recognition; defines the optimal hyperplane to separate different types of sitting posture data in the sitting posture feature space, and then distinguishes different types of sitting postures. The system and method have a wide range of application scenarios, making it possible for the sitting posture detection method with pressure as the detection target to enter the consumer product market.

[0040] Beneficial Effects By implementing the sitting posture detection system and method provided by the present invention, the following technical effects are achieved: (1) The present invention uses polyvinylidene fluoride material combined with conductive silver paste with stretch resistance to design a pressure sensor. Polyvinylidene fluoride has high conductivity and stability, and can maintain its performance as the base fabric bends and stretches, allowing the sensor to accurately capture pressure changes and adapt to different usage conditions; the stretch resistance of the conductive silver paste and the insulating coating material helps to improve the durability of the sensor and reduce problems such as wear and tear and short circuits caused by use; the cost of the sensor is low and meets the conditions for large-scale production; the design of the sensor takes into account deployment and maintenance issues in actual applications, and is suitable for use in various forms of seats and similar scenarios.

[0041] (2) Use the Kalman filter algorithm to process the noise in the data, set a voltage divider to convert the resistivity change caused by pressure change into a voltage signal, and detect the sitting posture by measuring the pressure distribution information on the seat surface. The pressure sensor accurately detects the pressure distribution on the seat surface, improves the accuracy of human sitting posture detection, realizes real-time conversion and reading of pressure data, effectively reduces the noise received by the sensor when working, and improves the stability and reliability of the signal; the system can provide users with real-time visual information, historical information and feedback reports, so that users can intuitively understand their sitting posture and make adjustments accordingly.

[0042] (3) Perform frequency domain characteristic analysis on the pressure information processed by the Kalman filter algorithm to extract key features that affect human sitting posture recognition. By analyzing the periodicity and frequency characteristics of sitting posture, the recognition rate of sitting posture detection is improved; the fast Fourier algorithm is efficient and can quickly process a large amount of data, which helps to improve the real-time performance of the system; by identifying the important frequency components in the signal, the amount of data stored and transmitted is reduced.

[0043] (4) By defining the optimal hyperplane to separate different types of sitting posture data in the sitting posture feature space, different sitting posture types can be distinguished. This enables the system to more accurately identify and classify different sitting postures, thereby improving the accuracy of detection. The support vector machine has a certain robustness to outliers in the data and can provide relatively stable classification results even when the data quality is poor. The implementation and deployment costs are relatively low, which helps to reduce the cost of the entire system. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to make the above-mentioned sitting posture detection system and method of the present invention more obvious and easy to understand, the drawings required for use in the specific implementation of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 A flowchart showing a sitting posture detection method is shown; Figure 2 It represents the schematic diagram of external interface; Figure 3 It shows the schematic diagram of the circuit from Vabt to 3.3V; Figure 4 Represents the schematic diagram of esp32 mode circuit; Figure 5 It shows the schematic diagram of Bluetooth control circuit; Figure 6 A schematic diagram of a charging control circuit is shown; Figure 7Schematic diagram of overcharge protection circuit. DETAILED DESCRIPTION

[0046] Embodiment 1: A sitting posture detection system and method are provided. The flow chart of the sitting posture detection method is as follows: Figure 1 As shown, the sitting posture detection system includes: a pressure sensor, a microcontroller, a frequency domain analysis module, a sitting posture recognition module, a cloud server, a terminal device and an application program, and specifically the sitting posture detection system and method of the steps described below.

[0047] The pressure sensor is used to measure the sitting pressure of the human body according to the deformation degree of the seat.

[0048] The pressure sensor comprises polyvinylidene fluoride, conductive silver paste, insulating coating material, PVC coated fabric and sponge cushion; The conductive silver paste is a conductive wire, which is printed on the insulating oil coating of the base fabric and is used to connect the pressure sensor and the microcontroller; The conductive silver paste has a resistance of less than 10 ohms per square foot; The insulating coating material is a substrate; The tensile strength of the conductive silver paste and the insulating coating material is not less than 50%; The PVC coated fabric is the base fabric, and the insulating coating material is coated on the PVC coated fabric to prevent the conductive silver paste from being lost during use.

[0049] The resistivity of the pressure sensor varies with the piezoresistive material.

[0050] The polyvinylidene fluoride is a piezoelectric material with a bending modulus of 40,000-50,000 pounds per square inch, a thickness of 104 microns, a volume resistivity of less than 500 ohm·meter, and a tensile strength of 1,700-2,000 pounds per square inch. The non-connected part of the circuits of the pressure sensor and the microcontroller are coated with an insulating coating material to avoid wear, short circuit, etc. during use.

[0051] The PVC coated fabric is bonded onto the sponge cushion to prevent sliding during the sitting posture detection.

[0052] When the pressure sensor is subjected to pressure, the entire pressure sensor becomes thinner and its shape bends, and current flows through the pressure sensor sheet to the microcontroller.

[0053] The system also includes a voltage divider for measuring and configuring the resistance change of the pressure sensor and matching the impedance; The output voltage of the pressure sensor is calculated as follows:

[0054] In the formula, is the output voltage of the pressure sensor; is the resistance of the pressure sensor; is the power supply voltage; is the resistance of the voltage divider.

[0055] Preferably +3.7 volts, Preferably 1000 ohms.

[0056] The total area of ​​the pressure sensor is 450*450 mm, and the area of ​​a single pressure sensor piece is 30*30 mm.

[0057] The pressure range of each pressure sensor sheet is 4.86 kg – 9.72 kg.

[0058] The microcontroller is used to collect the pressure distribution information of the seat surface and remove the noise in the information through the Kalman filter algorithm.

[0059] The microcontroller removes noise sources including flicker, mechanical and stretching stimulation, temperature changes and calibration error-induced noise through the Kalman filter algorithm.

[0060] The step of removing noise by the Kalman filter algorithm includes: estimating the current state variable and updating the estimated value using a weighted average value, giving more weight to the estimated value with higher certainty.

[0061] The Kalman filter algorithm takes as input the current input measurements and the previous calculated state.

[0062] The cloud server is used to perform real-time online analysis of sitting posture and provide data storage function.

[0063] The terminal device and the application are used to present the human sitting posture analysis results.

[0064] The system establishes communication with the IoT platform via WiFi and communicates with the application via Bluetooth.

[0065] The core processor of the microcontroller is ESP32; The microcontroller is preset with a time threshold. When the system determines that the duration of a human body being in an unhealthy sitting posture exceeds the time threshold, the microcontroller issues an interactive reminder.

[0066] Sitting posture detection system circuit Figures 2 to 7As shown, the system includes 16 pressure collection points, which are connected to the mainboard through an analog multiplexer. The multiplexer allows the data of the pressure sensor to be transmitted to the ESP32 processor through 5 analog input pins in sequence.

[0067] The data acquisition frequency of the system is 5 Hz.

[0068] The power source used by the system is a DC 3.7V 3000mAh lithium battery. The duty cycle between WiFi and Bluetooth is reduced to reduce system energy consumption, and data collection is performed in a sleep and work cycle.

[0069] The system supports reporting data in real-time or scheduled form.

[0070] During the time when the system reports data, if no change in the pressure acting on the sensor is detected within a threshold time, the sleep mode is enabled to reduce energy consumption.

[0071] Embodiment 2: On the basis of the above-mentioned embodiment, the system adds a frequency domain analysis module for performing frequency domain characteristic analysis on the pressure information processed by the Kalman filter algorithm, and extracting key features that have an impact on human sitting posture recognition.

[0072] The pressure distribution information of the pressure sensor is collected and continuous pressure readings are recorded to form time series data.

[0073] The collected pressure information is preliminarily processed by the Kalman filter algorithm, and the data is standardized to eliminate the influence of different dimensions.

[0074] The preprocessed pressure time series data is converted from the time domain to the frequency domain by fast Fourier transform, and the frequency characteristics of the sitting posture are represented by the frequency components and the corresponding amplitudes. The fast Fourier transform formula is as follows:

[0075] In the formula, For sequence No. The complex representation of the frequency components; is the first samples, indicating that the pressure sensor is The number of degrees at a time point; For the The weight factor of each sample is used to adjust the importance of pressure readings at different time points; is the total number of samples, indicating the number of data points collected; is an imaginary unit; is a complex exponential function, which represents the rotation factor, where is the frequency index.

[0076] The frequency components representing the sitting posture characteristics are determined based on the calculation results of the fast Fourier transform, and the frequency components are matched with the specific sitting posture.

[0077] By extracting key features that have an impact on human sitting posture recognition, the workload of data collection and analysis can be reduced. The algorithm is relatively simple, which reduces the complexity of the algorithm. Verification shows that when obtaining an average error similar to that of the above embodiment, only about 60% of the collected data is required to complete the task. This shows that adding a frequency domain analysis module helps to reduce the workload of pressure information collection while maintaining high accuracy of sitting posture detection, effectively expand the applicable working conditions of the system, reduce the time cost of collection, and improve the real-time performance of the system.

[0078] Embodiment 3: On the basis of the above-mentioned embodiment, the system adds a sitting posture recognition module for defining an optimal hyperplane to separate different types of sitting posture data in the sitting posture feature space, thereby distinguishing different sitting posture types.

[0079] First, the pressure distribution information of the human body in a sitting position is collected through a pressure sensor, and the pressure information includes information when different users are in different sitting positions.

[0080] The Kalman filter algorithm is used to remove noise from the pressure information to obtain clear and stable data.

[0081] The key features are extracted from the pressure information through the frequency domain analysis module, and the key features include the mean, variance, maximum value, minimum value, asymmetry, etc. of the pressure distribution.

[0082] The collected pressure information is divided into a training set and a test set in a ratio of 8:2, and the data in the training set is labeled, that is, each sitting posture is classified and labeled.

[0083] Among them, the kernel function uses a linear kernel.

[0084] Use the training set data to train the support vector machine model, where the decision function formula of the support vector machine is as follows:

[0085] In the formula, is the decision function used to calculate new data points Classification of is the sitting posture weight factor, which is used to adjust the sensitivity of sitting posture detection; is the Lagrange multiplier; For training data points The category labels represent two different categories; is the kernel function used to calculate new data points and training data points The similarity between is a bias term, which is used to represent the distance between the classification hyperplane and the origin.

[0086] The sitting posture types include forward leaning, backward leaning, left leaning, right leaning and normal.

[0087] The mean square error indicator is used to ensure that the prediction error of the model is within an acceptable range.

[0088] For example, suppose there are 3 training samples, and the features and labels are: ; ; .

[0089] Assume that in the model parameters, , , , , kernel function , sitting weight factor .

[0090] Kernel Function The calculation process is:

[0091]

[0092]

[0093] Decision function The calculation process is:

[0094]

[0095] According to the calculation results, is a positive number. According to the classification rules of the support vector machine, the new sitting posture data point Classify as positive class, e.g., normal sitting posture.

[0096] The effect of the sitting posture recognition module is shown in Table 1.

[0097] Table 1. Summary of the effects of the sitting posture recognition module

[0098] According to the experimental table, the support vector machine decision function can be used to distinguish different types of sitting postures, which can improve the accuracy and real-time performance of the system's sitting posture detection, and help users adjust their sitting postures in time to prevent health problems.

[0099] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable non-transient storage media containing computer-usable program code.

[0100] The present invention can provide computer program instructions to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce means for implementing the system.

[0101] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device that implements the functions of the system.

[0102] These computer program instructions may also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide steps for implementing the functions of the system.

Claims

1. A sitting posture detection system, characterized in that: include: A pressure sensor is used to measure the sitting pressure of a human body according to the degree of deformation of the seat; A microcontroller, which is used to collect pressure distribution information on the seat surface and remove noise from the information through a Kalman filter algorithm; Frequency domain analysis module, used to extract key features related to human sitting posture in pressure information; A sitting posture recognition module is used to identify the sitting posture type based on the extracted key features; Cloud server, used for real-time online analysis of sitting posture and providing data storage function; The terminal device and application are used to present the human sitting posture analysis results.

2. The system according to claim 1, characterized in that: The system also includes a voltage divider for measuring and configuring the resistance change of the pressure sensor and matching the impedance; The output voltage of the pressure sensor is calculated as follows: In the formula, is the output voltage of the pressure sensor; is the resistance of the pressure sensor; is the power supply voltage; is the resistance of the voltage divider.

3. The system according to claim 1, characterized in that: The microcontroller removes noise sources including flicker, mechanical and stretching stimulation, temperature changes and calibration error-induced noise through the Kalman filter algorithm.

4. The system according to claim 1, characterized in that: The microcontroller is preset with a time threshold. When the system determines that the duration of a human body being in an unhealthy sitting posture exceeds the time threshold, the microcontroller issues an interactive reminder.

5. The system according to claim 1, characterized in that: The frequency domain analysis module performs frequency domain characteristic analysis on the pressure information processed by the Kalman filter algorithm to extract key features that have an impact on human sitting posture recognition; By collecting the pressure distribution information of the pressure sensor and recording the continuous pressure degrees to form time series data; The collected pressure information is preliminarily processed through the Kalman filter algorithm, and the data is standardized to eliminate the influence of different dimensions; The preprocessed pressure time series data are converted from the time domain to the frequency domain by fast Fourier transform, and the frequency characteristics of the sitting posture are represented by the frequency components and the corresponding amplitudes; The frequency components representing the sitting posture characteristics are determined based on the calculation results of the fast Fourier transform, and the frequency components are matched with the specific sitting posture.

6. The system according to claim 5, characterized in that: The formula for the fast Fourier transform is as follows: In the formula, For sequence No. The complex representation of the frequency components; In the time domain samples, indicating that the pressure sensor is The number of degrees at a time point; For the The weight factor of each sample is used to adjust the importance of pressure readings at different time points; is the total number of samples, indicating the number of data points collected; is an imaginary unit; is a complex exponential function, which represents the rotation factor, where is the frequency index.

7. The system according to claim 1, characterized in that: The sitting posture recognition module defines an optimal hyperplane to separate different types of sitting posture data in the sitting posture feature space, thereby identifying different sitting posture types.

8. A sitting posture detection method, characterized in that: The method is implemented based on the system according to any one of claims 1 to 7: The method comprises: Measure the sitting pressure of the human body according to the deformation degree of the seat; Collect the pressure distribution information of the seat surface and remove the noise in the information through Kalman filter algorithm; Extract key features related to human sitting posture in pressure information; Identify the sitting posture type based on the extracted key features; Analyze the sitting posture in real time and present the analysis results of human sitting posture.

9. A computer device comprising a processor and a memory, wherein the processor is connected to the memory, and the memory is used to store a computer program, wherein: The processor is configured to execute the computer program stored in the memory, so that the computer device performs at least one step of the method according to claim 8.

10. A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, characterized in that: The computer program implements at least one step of the method of claim 8 when executed.

Images