Sitting posture detection method, device, equipment, system and storage medium

Abstract

The embodiment of the invention relates to the technical field of sitting posture detection, in particular to a sitting posture detection method, device, equipment and system and a storage medium. The sitting posture detection method is applied to sitting posture detection equipment, the sitting posture detection equipment is installed on a seat, a preset number of pressure sensors are arranged at preset positions of the seat respectively, and the sitting posture detection method comprises the steps that analog signals collected by the pressure sensors are obtained, the analog signals pass through an amplification circuit, a filtering circuit and other analog circuits, and the analog signals are converted into the sitting posture detection equipment; digital signals used for sitting posture detection and identification are obtained after digital-to-analog conversion processing; determining a current sitting posture feature of the to-be-detected object according to the digital signal; and detecting the current sitting posture characteristics based on a sitting posture detection algorithm, and determining abnormal sitting posture information of the to-be-detected object according to a detection result. According to the method, the accuracy and the cost of sitting posture detection and the use convenience of the user can be considered.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of sitting posture detection, and in particular to a sitting posture detection method, apparatus, device, system and storage medium. Background Art

[0002] Currently, there are three main technical solutions for sitting posture detection: the first is to use computer vision technology based on optical sensors to analyze whether the human sitting posture is abnormal; the second is to place sensors such as accelerometers, multi-axis force sensors or gyroscopes on wearable equipment to achieve sitting posture detection by having the user wear the wearable device; the third is to place a distance sensor on the back of the chair to test the sensor response under different sitting postures.

[0003] However, the first solution is easily affected by ambient light, and in many scenarios (such as being blocked by tables or clothing), it is difficult to monitor the movements of the lower limbs (such as crossing legs or shaking legs). In addition, this solution requires turning on the camera, which raises concerns about the protection of user privacy. The second solution is relatively cumbersome to operate, resulting in a poor user experience and a relatively high technical cost. The distance sensor of the third solution is also easily affected by factors such as user clothing, affecting the accuracy of sitting posture detection.

[0004] It can be seen that how to ensure accuracy in sitting posture detection while improving user convenience has become an urgent problem to be solved. Summary of the Invention

[0005] One purpose of the embodiments of the present application is to provide a sitting posture detection method, device, equipment, system and storage medium to solve the technical problem that the existing technology cannot simultaneously take into account the accuracy of sitting posture detection and user convenience.

[0006] In a first aspect, a sitting posture detection method is provided, which is applied to a sitting posture detection device, wherein the sitting posture detection device is installed on a seat, and a preset number of pressure sensors are respectively configured at preset positions of the seat, and the method comprises: obtaining analog signals collected by each pressure sensor, the analog signals being amplified, filtered, and processed by analog circuits, and subjected to digital-to-analog conversion to obtain digital signals for sitting posture detection and identification; determining the current sitting posture characteristics of the object to be detected based on the digital signals; detecting the current sitting posture characteristics based on a sitting posture detection algorithm, and determining abnormal sitting posture information of the object to be detected based on the detection results.

[0007] In combination with the first aspect, in a possible embodiment, the sitting posture detection device is connected to a host computer; the current sitting posture characteristics of the object to be detected based on the digital signal are determined, including: sending the digital signal to the host computer, so that the host computer determines the current sitting posture characteristics of the object to be detected based on the digital signal; receiving the current sitting posture characteristics; and / or, detecting the current sitting posture characteristics based on a sitting posture detection algorithm, and determining the abnormal sitting posture information of the object to be detected based on the detection result, including: sending the current sitting posture characteristics to the host computer, so that the host computer detects the current sitting posture characteristics based on a sitting posture detection algorithm, and determines the abnormal sitting posture information of the object to be detected based on the detection result; receiving the abnormal sitting posture information.

[0008] In combination with the first aspect, in a possible implementation, before determining the current sitting posture characteristics of the object to be detected based on the digital signal, it also includes: judging whether the sitting posture detection device has completed the baseline calibration processing and / or personal calibration processing for the current time period; if both are yes, executing the determination of the current sitting posture characteristics of the object to be detected based on the digital signal.

[0009] In combination with the first aspect, in a possible embodiment, the method further includes: after the sitting posture detection device is in the turned-on state, performing the baseline calibration process every preset time range: when the standard deviation and the range of the digital signals of each pressure sensor within the preset time range are both within the preset threshold, and there is no quasi-periodic change in the digital signals of each pressure sensor within the preset time range, judging based on the digital signal whether the difference between the average value of the digital signal of each pressure sensor within the preset time range and the current baseline of the sitting posture detection device is less than the preset difference threshold; if the difference between the average value of the digital signal of each pressure sensor within the preset time range and the current baseline of the sitting posture detection device is less than the preset difference threshold, then updating the current baseline with the average value of the digital signal of each pressure sensor within the preset time range.

[0010] In combination with the first aspect, in a possible implementation, the method further includes: if the sitting posture detection device has not completed the personal calibration process, obtaining the biometric information of the object to be detected; calibrating the table according to the biometric information so that the distance between the table and the sitting posture detection device is within a preset distance range; and calibrating the height of the sitting posture detection device according to the biometric information so that the height of the sitting posture detection device meets a preset height condition.

[0011] In combination with the first aspect, in a possible implementation, the biometric information of the object to be detected includes weight information and / or height information; obtaining the biometric information of the object to be detected includes: obtaining the biometric information of the object to be detected based on the digital signal of the pressure sensor; or receiving the biometric information input by the object to be detected.

[0012] In combination with the first aspect, in a possible implementation, the preset height condition is determined based on the height of the object to be detected and the digital signal value of the pressure sensor; the height calibration of the sitting posture detection device based on the biometric information so that the height of the sitting posture detection device meets the preset height condition, including: looking up the proportional relationship between the sitting height and the height of the object to be detected based on the height information in the biometric information; determining the height range of the sitting posture detection device based on the proportional relationship; debugging the sitting posture detection device from small to large according to the height range of the sitting posture detection device, and when the ratio of the digital signal value of the target pressure sensor and the gravity of the object to be detected is within the preset range, determining that the height of the sitting posture detection device at the current moment is the calibration height; wherein, the target pressure sensor includes a pressure sensor at the foot pad and / or seat cushion position.

[0013] In combination with the first aspect, in a possible implementation, the biometric information includes weight information; the method further includes: calculating the time domain mean of the digital signals of each pressure sensor of the object to be detected in a normal sitting position; calculating the total pressure value of the object to be detected perpendicular to the ground in a normal sitting position based on the time domain mean; calculating the gravity exerted on the object to be detected based on the weight information; if the difference between the gravity and the total pressure value is greater than a preset gravity difference threshold, a prompt message is issued, and the prompt message is used to prompt that the pressure sensor reading is incorrect or the input weight information is incorrect.

[0014] In combination with the first aspect, in a possible implementation, before determining the current sitting posture characteristics of each pressure sensor based on the digital signal, it also includes: preprocessing the digital signal according to the time domain mean; and determining the current sitting posture characteristics of the sitting posture detection device based on the preprocessed digital signal.

[0015] In combination with the first aspect, in a possible implementation, the sitting posture detection algorithm includes preset sitting posture features and the correspondence between the preset sitting posture features and abnormal sitting postures; the current sitting posture features are detected based on the sitting posture detection algorithm, and the abnormal sitting posture information of the object to be detected is determined according to the detection results, including: comparing the similarity between the preset sitting posture features and the current sitting posture features; and determining the abnormal sitting posture information of the object to be detected based on the preset sitting posture features with the highest similarity and the correspondence.

[0016] In combination with the first aspect, in a possible implementation, the current sitting posture feature is detected based on the sitting posture detection algorithm, and the abnormal sitting posture information of the object to be detected is determined according to the detection result. Before that, it also includes: collecting the pressure distribution information set of the pressure sensor under each abnormal sitting posture; extracting the abnormal sitting posture feature according to the pressure distribution information set, and marking the abnormal sitting posture feature with label data, and the label data is used to identify the correspondence between the abnormal sitting posture feature and the abnormal sitting posture; completing the training process of the sitting posture detection algorithm according to the label data; the current sitting posture feature is detected based on the sitting posture detection algorithm, and the abnormal sitting posture information of the object to be detected is determined according to the detection result, including: feature classification of the current sitting posture feature based on the sitting posture detection algorithm to obtain the abnormal sitting posture information of the object to be detected.

[0017] In combination with the first aspect, in a possible embodiment, the extracting of abnormal sitting posture features based on the pressure distribution information set includes: performing feature extraction processing on the pressure distribution information set based on physical principles and experimental data to obtain abnormal sitting posture features; and / or, performing feature extraction processing on the pressure distribution information set based on a deep learning model to obtain abnormal sitting posture features.

[0018] In combination with the first aspect, in a possible implementation manner, the preset positions include seat cushion, foot pad and backrest positions.

[0019] In combination with the first aspect, in a possible implementation manner, the abnormal sitting posture information includes any one or more of the following: left-right tilted sitting posture, crossed-legged sitting posture, cross-legged sitting posture, lying on the table sitting posture, hunched-over sitting posture, and shaking leg sitting posture; when performing the left-right tilted sitting posture detection, the current sitting posture characteristics at least include: the left-right center of gravity of the seat cushion, the left-right relative center of gravity change of the seat cushion, the normalized difference of the left-right pressure of the seat cushion, and the left-right relative pressure change difference of the seat cushion; when performing the crossed-legged sitting posture or cross-legged sitting posture detection, the current sitting posture characteristics at least include: the difference between the left-right pressure ratio of the front side of the seat cushion and the left-right pressure ratio of the rear side of the seat cushion, the relative pressure change of the left front side of the seat cushion and the relative pressure change of the right front side of the seat cushion, the difference between the left-right relative pressure change difference of the front side of the seat cushion and the left-right relative pressure change difference of the rear side of the seat cushion; when performing the leg-extended sitting posture detection, the current sitting posture characteristics at least include: the left-right center of gravity of the seat cushion, the left-right relative center of gravity change of the seat cushion, the normalized difference of the left-right pressure of the seat cushion and the left-right relative pressure change difference of the rear side of the seat cushion The posture characteristics include at least: the pressure ratio of the front edge of the seat cushion, the difference ratio of the relative pressure changes of the front edge of the seat cushion, and the difference ratio of the relative pressure changes of the front side of the seat cushion and the relative pressure changes of the back side of the seat cushion; when performing the lying on the table sitting posture detection, the current sitting posture characteristics include at least: the ratio of the total pressure of the seat cushion and the total pressure of the template perpendicular to the ground, the ratio of the total pressure of the seat cushion and foot pad and the total pressure of the template perpendicular to the ground, the total pressure change of the seat cushion, and the total pressure change of the seat cushion and foot pad; when performing the hunched sitting posture detection, the current sitting posture characteristics include at least: the change of the relative center of gravity of the front and rear of the seat cushion, and the torque characteristics of the human spine; when performing the leg shaking sitting posture detection, the current sitting posture characteristics include at least: the maximum value and maximum frequency of the frequency spectrum of the digital signal of the pressure sensor of the foot pad and / or foot pad position within the preset acquisition time, and the number of leg shaking times per unit time.

[0020] In combination with the first aspect, in a possible implementation, the current sitting posture feature includes a current sitting posture time domain feature and / or a current sitting posture frequency domain feature.

[0021] In combination with the first aspect, in a possible implementation, the pressure sensor is a strain gauge pressure sensor.

[0022] In combination with the first aspect, in a possible implementation, the digital signal includes a digital signal of at least one pressure sensor and / or a combination obtained based on the digital signal of the at least one pressure sensor.

[0023] In a second aspect, a sitting posture detection device is provided, which is applied to a sitting posture detection device, wherein the sitting posture detection device is installed on a seat, and a preset number of pressure sensors are respectively arranged at preset positions of the seat, and the sitting posture detection device includes: an acquisition module, which is used to acquire analog signals collected by each pressure sensor, and the analog signals are subjected to analog circuits such as amplification and filtering, and digital-to-analog conversion processing to obtain digital signals for sitting posture detection and identification; a determination module, which is used to determine the current sitting posture characteristics of the object to be detected based on the digital signal; and a detection module, which is used to detect the current sitting posture characteristics based on a sitting posture detection algorithm, and determine the abnormal sitting posture information of the object to be detected based on the detection results.

[0024] In a third aspect, a sitting posture detection device is provided, which includes a memory and a processor, and a preset number of pressure sensors respectively arranged at preset positions of a seat, each pressure sensor and the memory being connected to the processor, and the processor being used to execute one or more computer programs stored in the memory. When the processor executes the one or more computer programs, the sitting posture detection device implements the method described in the first aspect.

[0025] In a fourth aspect, a sitting posture detection system is provided, which includes a seat and the sitting posture detection device as described in the third aspect.

[0026] In a fifth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor executes the method as described in the first aspect.

[0027] The embodiments of the present application can achieve the following technical effects: the sitting posture detection device is configured with a preset number of pressure sensors at preset positions, obtains analog signals collected by each pressure sensor, and after digital-to-analog conversion, obtains a digital signal for sitting posture detection and identification, determines the current sitting posture characteristics of the object to be detected based on the digital signal, and detects the current sitting posture characteristics based on the sitting posture detection algorithm to determine the abnormal sitting posture information of the object to be detected. In the present application, by setting pressure sensors at preset positions on the seat, extracting sitting posture characteristics and performing sitting posture detection based on the sitting posture detection algorithm, the accuracy of sitting posture detection is effectively guaranteed, and users can perform sitting posture detection without wearing the device. In this way, the accuracy of sitting posture detection and user convenience are taken into account. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0029] Figure 1 A schematic structural diagram of a sitting posture detection device provided in an embodiment of the present application;

[0030] Figure 2 A flowchart of a sitting posture detection method provided in an embodiment of the present application;

[0031] Figure 3 A flowchart of another sitting posture detection method provided in an embodiment of the present application;

[0032] Figure 4 A schematic diagram of a normal sitting posture provided in an embodiment of the present application;

[0033] Figure 5 A schematic diagram of a left-right tilted sitting posture provided in an embodiment of the present application;

[0034] Figure 6 A schematic diagram of a sitting posture with legs crossed or legs crossed provided in an embodiment of the present application;

[0035] Figure 7 A schematic diagram of a sitting posture lying on a table provided in an embodiment of the present application;

[0036] Figure 8 A schematic diagram of a hunched sitting posture provided in an embodiment of the present application;

[0037] Figure 9 A schematic structural diagram of a sitting posture detection device provided in an embodiment of the present application;

[0038] Figure 10 A schematic structural diagram of a sitting posture detection device provided in an embodiment of the present application;

[0039] Figure 11 A schematic diagram of the architecture of a sitting posture detection system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0040] In order to make the purpose, technical solutions and advantages of this application more clearly understood, the present application is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0041] It should be noted that, if there is no conflict, the various features in the embodiments of the present application can be combined with each other and are all within the scope of protection of the present application. In addition, although the functional modules are divided in the device schematic and the logical order is shown in the flow chart, in some cases, the steps shown or described can be performed in a different order than the module division in the device or the order in the flow chart. Furthermore, the words "first", "second", "third", etc. used in this application do not limit the data and execution order, but only distinguish between the same items or similar items with basically the same functions and effects.

[0042] With the progress and development of society, modern people basically sit for a long time when working and studying, but many people have health problems due to poor sitting posture. Abnormal sitting posture can cause excessive load on the spine (including the cervical, thoracic, lumbar, sacral and coccyx), which in turn leads to muscle soreness, poor blood flow, nerve numbness, and even pathological bone structure changes. In the existing technology, there are three main technical solutions for sitting posture detection: the first is to detect and analyze human sitting posture with computer vision technology based on optical sensors; the second is a sitting posture monitoring solution that arranges sensors such as acceleration, multi-axis force sensors or gyroscopes on wearable equipment on the human body; the third is to arrange distance sensors on the back of the chair to test the sensor response under different sitting postures.

[0043] However, in the first approach, the accuracy of the computer vision-based technical solution is easily affected by ambient light and cannot be used in many scenarios (such as being blocked by tables or clothing). In addition, this solution also has difficulties in monitoring the movements of the lower limbs of the human body (such as crossing legs, shaking legs, etc.). The accuracy of sitting posture detection cannot be well guaranteed. In the second approach, the operation of the wearable equipment solution is relatively cumbersome and the user experience is poor. In the third approach, the distance sensor is easily interfered with by factors such as the user's clothing, which can easily lead to inaccurate readings.

[0044] In view of this, the present application proposes a sitting posture detection method, device, equipment, system and storage medium. By installing a sitting posture detection device on a seat, the sitting posture detection device is configured with a pressure sensor at a preset position of the seat, and the user's sitting posture characteristics are determined by the signal collected by the pressure sensor, and sitting posture detection is performed based on the sitting posture detection algorithm. On the one hand, installing a pressure sensor on the seat to collect signals and performing sitting posture detection based on the sitting posture detection algorithm can ensure the accuracy of sitting posture detection. On the other hand, the user does not need to wear the device to achieve sitting posture detection. In this way, the accuracy of sitting posture detection and the convenience of user use are taken into account.

[0045] To make this application easier to understand, first of all, we will introduce a sitting posture detection device involved in this application. Figure 1 , is a structural diagram of a sitting posture detection device provided in an embodiment of the present application. The sitting posture detection device is installed on a seat, and a preset number of pressure sensors are respectively configured at preset positions of the seat. As a feasible implementation method, the preset positions include Figure 1 The seat cushion, foot pad and backrest positions shown in the figure are divided into the following types according to the setting position: Figure 1 The seat cushion sensor, foot mat sensor and backrest sensor are shown.

[0046] In one embodiment, the preset number can vary depending on the sensor installation location. For example, the seat cushion sensor can be divided into six symmetrical blocks, each used to measure pressure perpendicular to the seat cushion. These blocks are designated as sensors 1 through 6, from the front left to the back right. The foot pad sensor can be placed on the foot pad to measure pressure perpendicular to the foot pad, designated as sensor 7. The backrest sensor can be placed on the seat back to measure pressure perpendicular to the seat back, designated as sensor 8. With the center of the seat cushion as the origin, the left-right direction as the x-axis, and the front-to-back direction as the y-axis, the position coordinates of sensors 1 through 6 can be designated as (x1, y1) through (x6, y6).

[0047] Through Figure 1 The sitting posture detection device shown can implement the sitting posture detection method shown in this application. Figure 2 , is a flow chart of a sitting posture detection method provided in an embodiment of the present application, Figure 2 The sitting posture detection method shown may include:

[0048] S201. Acquire analog signals collected by various pressure sensors. The analog signals are processed by analog circuits such as amplification and filtering, and digital-to-analog conversion to obtain digital signals for sitting posture detection and recognition.

[0049] As a feasible implementation method, the pressure sensor involved in the present application can be a strain type pressure sensor, which is a pressure sensor that uses elastic sensitive elements and strain gauges to convert the measured pressure into corresponding resistance value changes. It has the advantages of low cost, high repeatability, and easy promotion.

[0050] As another feasible implementation, the pressure sensor may also be a piezoelectric, piezoresistive, capacitive, magnetostrictive or other type of pressure sensor capable of measuring pressure, and this application does not impose any restrictions on this.

[0051] In this embodiment, the sitting posture detection device can obtain the analog signal in real time, or obtain the analog signal when the user turns on the sitting posture recognition function, or obtain the analog signal when a large change is detected in the analog signal of the pressure sensor. This application does not impose any restrictions on this.

[0052] It should be noted that the sitting posture detection device may take the value before being pressed as a baseline, and the value of the digital signal may be the change in the readings of each pressure sensor after being pressed.

[0053] For example, the analog signals collected by each pressure sensor are processed by analog circuits such as amplification and filtering, and then by digital-to-analog conversion to become digital signals used for sitting posture recognition. The digital signal can be recorded as f (i,t) , where i=1,2,3,4,5,6,7,8 represents the sensor number and t>0 represents the acquisition time.

[0054] As a feasible implementation, the digital signal can be processed for feature extraction and / or posture detection at the embedded end (i.e., the posture detection device itself), or it can be transmitted to a host computer (such as a computer, smartphone, etc.) for feature extraction and / or posture detection. Transmission to the host computer can be wired (such as serial port, USB), wireless (such as Bluetooth, WiFi), or a combination of these.

[0055] As a feasible implementation, to protect privacy, data transmission can be encrypted. Encryption methods such as DES (Data Encryption Standard) and RSA (an encryption algorithm named after its inventors, Rivest, Shamir, and Adleman) can be used. To ensure data integrity, packet verification can be performed during transmission. Verification methods such as CRC (Cyclic Redundancy Check) can be used. When the number of sensors and / or sampling rate are high, the amount of data increases, and data compression may be required during transmission. Compression methods such as DEFLATE (a lossless data compression algorithm that uses both the LZ77 algorithm and Huffman coding) and LZO (Lempel-Ziv-Oberhumer) can be used.

[0056] In one embodiment, the digital signal includes a digital signal of at least one pressure sensor and / or a combination obtained based on the digital signal of the at least one pressure sensor.

[0057] For example, different digital signals may be used to detect different abnormal sitting posture characteristics. For example, detecting an abnormal sitting posture that leans left or right primarily uses the digital signal from the seat cushion sensor, while detecting a sitting posture that lies prone on a desk primarily uses the digital signals from the seat cushion sensor and the footrest sensor. Alternatively, the digital signal may be a combination of the digital signals from the at least one pressure sensor, such as a linear combination.

[0058] S202: Determine the current sitting posture characteristics of the object to be detected according to the digital signal.

[0059] In this embodiment, the object to be detected may be a user who is using the sitting posture detection device.

[0060] As a feasible implementation manner, the sitting posture detection device can determine the pressure distribution information on the sitting posture detection device according to the digital signal, and the current sitting posture characteristics of the object to be detected can be determined according to the pressure distribution information.

[0061] In one embodiment, the current sitting posture feature includes a current sitting posture time domain feature and / or a current sitting posture frequency domain feature.

[0062] It should be noted that the current sitting posture feature can be determined on the sitting posture detection device, or the digital signal can be transmitted to a host computer, which determines the current sitting posture feature of the object to be detected based on the digital signal.

[0063] In one embodiment, the sitting posture detection device can be connected to a host computer, and the step S202 includes: sending the digital signal to the host computer so that the host computer determines the current sitting posture characteristics of the object to be detected based on the digital signal; and receiving the current sitting posture characteristics.

[0064] For example, after obtaining the digital signal, the sitting posture detection device sends the digital signal to the host computer through a communication connection pre-established with the host computer. After receiving the digital signal, the host computer determines the current sitting posture characteristics of the object to be detected based on the digital signal, and sends the current sitting posture characteristics to the sitting posture detection device. The sitting posture detection device can then obtain the current sitting posture characteristics.

[0065] S203: Detect the current sitting posture feature based on a sitting posture detection algorithm, and determine abnormal sitting posture information of the object to be detected according to the detection result.

[0066] It should be noted that the abnormal sitting posture involved in this application is a concept relative to the normal sitting posture. For example, the normal sitting posture may refer to the subject to be detected sitting upright on the seat, with the calf maintaining a first preset angle (such as 90°) with the ground, the knee maintaining a second preset angle (such as 90°) with the thigh, the thigh being horizontal, and the thigh maintaining a third preset angle (such as 90°) with the waist. For example, the normal sitting posture may be as follows: Figure 4 shown.

[0067] In one embodiment, the abnormal sitting posture information includes, but is not limited to, information on leaning left or right sitting posture, sitting with legs crossed, sitting with legs crossed, sitting with the back of the table bent, sitting with a hunched back, and sitting with legs shaking.

[0068] It should be noted that the current sitting posture feature can be detected on the sitting posture detection device based on the sitting posture detection algorithm, or the current sitting posture feature can be transmitted to the host computer, which then detects the current sitting posture feature based on the sitting posture detection algorithm.

[0069] In one embodiment, the sitting posture detection device can be connected to a host computer, and the S203 step includes: sending the current sitting posture characteristics to the host computer so that the host computer detects the current sitting posture characteristics based on the sitting posture detection algorithm, and determines the abnormal sitting posture information of the object to be detected according to the detection results; receiving the abnormal sitting posture information.

[0070] For example, after obtaining the current sitting posture feature, the sitting posture detection device sends the current sitting posture feature to the host computer through a communication connection pre-established with the host computer. After receiving the current sitting posture feature, the host computer determines the abnormal sitting posture information of the object to be detected based on the current sitting posture feature, and sends the abnormal sitting posture information of the object to be detected to the sitting posture detection device. The sitting posture detection device can then obtain the abnormal sitting posture information.

[0071] In one embodiment, the sitting posture detection algorithm includes preset sitting posture features and a correspondence between the preset sitting posture features and abnormal sitting postures. Detecting the current sitting posture features based on the sitting posture detection algorithm and determining abnormal sitting posture information of the subject to be detected based on the detection results includes: comparing the preset sitting posture features with the current sitting posture features for similarity; and determining abnormal sitting posture information of the subject to be detected based on the preset sitting posture feature with the highest similarity and the correspondence.

[0072] It should be noted that the preset sitting posture feature can be obtained based on prior knowledge or experimental data, and the sitting posture detection can be performed based on the preset sitting posture feature and judgment rules for different abnormal sitting posture features.

[0073] For example, if the current sitting posture feature has the highest similarity to the preset sitting posture feature of the abnormal sitting posture with legs extended forward, it can be considered that the current abnormal sitting posture of the subject to be detected is the sitting posture with legs extended forward.

[0074] In one embodiment, the method of detecting the current sitting posture characteristics based on the sitting posture detection algorithm and determining the abnormal sitting posture information of the object to be detected based on the detection results further includes: collecting a set of pressure distribution information of pressure sensors under various abnormal sitting postures; extracting abnormal sitting posture characteristics based on the pressure distribution information set, and marking the abnormal sitting posture characteristics with label data, wherein the label data is used to identify the correspondence between abnormal sitting posture characteristics and abnormal sitting postures; and completing the training process of the sitting posture detection algorithm based on the label data. The method of detecting the current sitting posture characteristics based on the sitting posture detection algorithm and determining the abnormal sitting posture information of the object to be detected based on the detection results includes: performing feature classification on the current sitting posture characteristics based on the sitting posture detection algorithm to obtain the abnormal sitting posture information of the object to be detected.

[0075] In one embodiment, the extracting of abnormal sitting posture features based on the pressure distribution information set includes: performing feature extraction processing on the pressure distribution information set based on physical principles and experimental data to obtain abnormal sitting posture features; and / or, performing feature extraction processing on the pressure distribution information set based on a deep learning model to obtain abnormal sitting posture features.

[0076] It should be noted that this feature extraction processing method can be obtained based on physical principles and experimental data, or it can be extracted using a deep learning model.

[0077] For example, the sitting posture detection algorithm can be a traditional machine learning method (such as support vector machines, random forests, naive Bayes, etc.). The sitting posture detection algorithm can be first trained based on training data. The sitting posture detection algorithm automatically learns the discrimination rules and classifies the sitting posture features, thereby achieving sitting posture detection. For example, for the multi-classification problem of normal sitting posture, hunched over, and lying on the desk, labeled data can be collected first, sitting posture features can be extracted, and then a multi-class random forest classifier can be trained to perform sitting posture detection.

[0078] As another feasible implementation, the sitting posture detection device can also utilize deep learning methods to further mine features based on the collected sensor signals and / or extracted sitting posture features, thereby achieving sitting posture detection. For example, to detect crossed legs, data can be collected first for both crossed and uncrossed legs. Then, the manually extracted features and the original pressure signal can be input into a trained neural network. The neural network (e.g., a fully connected network) can be used to further mine features and achieve sitting posture detection.

[0079] In one embodiment, the sitting posture detection algorithm can combine the above schemes. For example, some postures can be detected using a threshold comparison method, while others can be detected using a deep learning algorithm. Alternatively, a threshold algorithm can be used for coarse classification, followed by a deep learning model for fine classification. This application does not impose any restrictions on this.

[0080] It can be seen that through the method shown in the embodiment of the present application, the sitting posture detection device is respectively configured with a preset number of pressure sensors at preset positions, and after obtaining the analog signals collected by each pressure sensor and processing them through analog circuits and digital-to-analog conversion to obtain digital signals for sitting posture detection and identification, the current sitting posture characteristics of the object to be detected are determined based on the digital signals, and the current sitting posture characteristics are detected based on the sitting posture detection algorithm to determine the abnormal sitting posture information of the object to be detected. The present application sets pressure sensors at preset positions on the seat, extracts sitting posture characteristics, and performs sitting posture detection based on the sitting posture detection algorithm, so that the detection accuracy of the sitting posture detection algorithm is effectively guaranteed, and the user can achieve sitting posture detection without wearing the device. In this way, the accuracy of sitting posture detection and the convenience of user use are taken into account.

[0081] See below Figure 3 , which is a flow chart of another sitting posture detection method provided in an embodiment of the present application. Figure 3The sitting posture detection method shown is applied to a sitting posture detection device installed on a seat, and a preset number of pressure sensors are respectively configured at the seat cushion, foot pad, and backrest of the seat. The method may include:

[0082] S301. Acquire analog signals collected by various pressure sensors. The analog signals are processed by analog circuits such as amplification and filtering, and digital-to-analog conversion to obtain digital signals for sitting posture detection and recognition.

[0083] It should be noted that the implementation process of step S301 can refer to the corresponding description of the aforementioned method embodiment, and will not be repeated here.

[0084] S302: Determine whether the sitting posture detection device has completed the baseline calibration process and / or the personal calibration process for the current time period.

[0085] As a feasible implementation, after the sitting posture detection device is turned on, a personal calibration process may be performed, and / or the baseline calibration process may be performed every preset time range.

[0086] In this embodiment, if the sitting posture detection device has completed the baseline calibration process and / or the personal calibration process for the current time period, step S305 and step S306 may be directly executed.

[0087] S303: If the sitting posture detection device has not completed the baseline calibration process for the current time period, perform the baseline calibration process.

[0088] It can be understood that the value of the digital signal of each pressure sensor is based on the change in the baseline after being pressurized. Calibration of the baseline can ensure the accuracy of the pressure sensor data.

[0089] Among them, after the sitting posture detection device is in the turned-on state, the step of performing the baseline calibration process every preset time range may include: when the standard deviation and the range of the digital signals of each pressure sensor within the preset time range are both within the preset threshold value, and there is no quasi-periodic change in the digital signals of each pressure sensor within the preset time range, judging according to the digital signal whether the difference between the average value of the digital signals of each pressure sensor within the preset time range and the current baseline of the sitting posture detection device is less than the preset difference threshold value; if the difference between the average value of the digital signals of each pressure sensor within the preset time range and the current baseline of the sitting posture detection device is less than the preset difference threshold value, then updating the current baseline with the average value of the digital signals of each pressure sensor within the preset time range.

[0090] In one embodiment, the method further includes: if the difference between the average value of the digital signals of each pressure sensor within the preset time range and the current baseline of the sitting posture detection device is greater than or equal to the preset difference threshold, an interference prompt message is displayed, and the interference prompt message is used to prompt that there is foreign body interference information on the pressure sensor.

[0091] For example, the preset time range can be 5 minutes. The sitting posture detection device can first perform activity detection: detect whether the standard deviation and range of the digital signals of each pressure sensor within 5 minutes are within the preset threshold. If not, no calibration will be performed and the current baseline will still be used. If it is within the preset threshold, then non-living body detection can be further performed: detect whether the digital signals of each pressure sensor within 5 minutes have quasi-periodic changes within a specific period range (in one embodiment, the specific period range can be the human heart rate period of 0.2 to 2 seconds, and the quasi-periodic changes can be measured by calculating its frequency domain distribution). If it is a quasi-periodic change, no calibration will be performed and the current baseline will still be used. If it is not a quasi-periodic change, then consistency detection can be further performed: detect whether the difference between the average value of the digital signals of each pressure sensor within 5 minutes and the current baseline is too large. If the difference is less than the preset difference threshold, the average value is used as the new baseline. Otherwise, the calibration process is terminated, and at the same time, an interference prompt message is issued to prompt that the sensor is interfered with by foreign objects.

[0092] S304: If the sitting posture detection device has not completed the personal calibration process, perform the personal calibration process.

[0093] It should be noted that the personal calibration process can be a personalized setting for each subject to be detected. It is understandable that, taking the subject to be detected as a human as an example, since each subject to be detected has different heights, weights, body shapes, etc., performing personal calibration can better improve the accuracy of sitting posture detection.

[0094] Among them, the personal calibration processing step includes: obtaining the biometric information of the object to be detected; calibrating the table according to the biometric information so that the distance between the table and the sitting posture detection device is within a preset distance range; and calibrating the height of the sitting posture detection device according to the biometric information so that the height of the sitting posture detection device meets the preset height condition.

[0095] For example, the personal calibration process can be mainly divided into two parts: table calibration and sitting posture detection device height calibration. In the table calibration part, the sitting posture detection device can obtain the biometric information of the subject to be detected, such as height and weight, manually input by the user. Then, based on the height, it can find the optimal ratio between the table height and the height, and obtain the appropriate table height. The distance between the table and the sitting posture detection device can be adjusted to within a preset distance range (for example, the distance between the human body and the edge of the table when sitting upright is 10-15 cm).

[0096] The sitting posture detection device may acquire biometric information by: obtaining the biometric information of the subject to be detected based on the digital signal of the pressure sensor; or receiving biometric information input by the subject to be detected. That is, the sitting posture detection device may automatically determine the biometric information based on the digital signal of the pressure sensor, or the biometric information may be manually input by the user.

[0097] As a feasible embodiment, the height of the tabletop and the distance between the table and the sitting posture detection device can be adjusted automatically. For example, the sitting posture detection device, the chair, and the table are interconnected, and when the sitting posture detection device detects the presence of an object to be detected, the device can automatically adjust the height of the tabletop and the distance between the table and the sitting posture detection device. Alternatively, the height of the tabletop and the distance between the table and the sitting posture detection device can be adjusted manually by the user after the sitting posture detection device prompts the user with an appropriate distance range.

[0098] The height calibration part of the sitting posture detection device (in this embodiment, it can be equivalent to the height calibration of the seat) can be completed based on the body height and / or the pressure values of each pressure sensor.

[0099] In one embodiment, height calibration can be performed directly based on the height information. Specifically, the optimal ratio between the sitting height and the height of the person is found based on the height information, and then the height of the sitting posture detection device can be calculated, and the height calibration is performed using the calculated height.

[0100] In one embodiment, the calibration can also be completed directly based on the pressure value (i.e., the value of the digital signal) of the target pressure sensor, which includes the pressure sensor at the foot pad and / or seat cushion position. For example, the following formula is used to calculate the pressure ratio of the front side of the seat cushion: and / or foot pad pressure ratio

[0101]

[0102]

[0103] The sitting posture detection device can determine the pressure ratio of the front side of the seat cushion Is it within the pre-set range? If it is higher than the range, the seat height can be lowered. Lower to the preset range. If it is lower than the range, the seat height can be raised so that When the When it is within the pre-set range, the seat height calibration is considered complete. Or, determine the proportion of foot pad pressure Is it within the pre-set range? If it is higher than the range, the seat height can be raised so that Lower to the preset range. If it is lower than the range, the seat height can be lowered. When the When the seat height is within the pre-set range, the seat height calibration is considered complete.

[0104] In one embodiment, the height calibration may also incorporate calibration methods based on height and pressure distribution.

[0105] Specifically, the preset height condition can be determined based on the height of the object to be detected and the pressure value of the pressure sensor. The height calibration of the sitting posture detection device based on the biometric information so that the height of the sitting posture detection device meets the preset height condition includes: finding the proportional relationship between the sitting height and the height of the object to be detected based on the height information in the biometric information; determining the height range of the sitting posture detection device based on the proportional relationship; debugging the sitting posture detection device from small to large height range, and when the ratio of the digital signal value of the target pressure sensor and the gravity of the object to be detected is within the preset range, determining the height of the sitting posture detection device at the current moment as the calibration height; wherein the target pressure sensor includes a pressure sensor at the foot pad and / or seat cushion position.

[0106] For example, the sitting posture detection device can first determine an approximate height range based on height information, and then adjust the device based on the target pressure sensor's pressure range to achieve the optimal seat height. The determination of the approximate height range based on height information and the adjustment based on the target pressure sensor's pressure range can be found in the previous examples and will not be further elaborated here.

[0107] In one embodiment, after completing the height calibration, it also includes: calculating the time domain mean of the digital signals of each pressure sensor of the object to be detected in a normal sitting position; calculating the total pressure value of the object to be detected perpendicular to the ground in a normal sitting position based on the time domain mean; calculating the gravity exerted on the object to be detected based on the weight information; if the difference between the gravity and the total pressure value is greater than a preset gravity difference threshold, a prompt message is issued, and the prompt message is used to prompt that the pressure sensor reading is incorrect or the input weight information is incorrect.

[0108] For example, the sitting posture detection device can collect digital signals of various pressure sensors in a normal sitting posture for a period of time (such as 10 seconds), and calculate the time domain mean of these signals as template parameters, such as reference Figure 1 The distribution of pressure sensors, denoted as f (i,0) , where i = 1, 2, 3, 4, 5, 6, 7, 8. Based on the template parameters, the total pressure value of the object to be detected perpendicular to the ground in a normal sitting posture can be calculated: Furthermore, the sitting posture detection device can also calculate the gravity of the object to be detected based on the weight information in the biometric information, and compare it with the f (g,0) When the difference between the two is greater than a preset gravity difference threshold, a prompt message may be issued to indicate that the sensor reading is incorrect or the input weight is inaccurate.

[0109] In one embodiment, before executing step S304, the method may further include: preprocessing the digital signal according to the time domain mean; and determining the current sitting posture feature of the sitting posture detection device according to the preprocessed digital signal.

[0110] It should be noted that preprocessing the digital signal according to the time domain mean can improve the generalization of the sitting posture detection algorithm.

[0111] For example, suppose the preprocessed digital signal is denoted as s (i,t) , the preprocessing process can satisfy the following formula:

[0112]

[0113] S305: Determine the current sitting posture characteristics of the object to be detected according to the digital signal.

[0114] In a feasible embodiment, the sitting posture characteristics can be divided according to the seat cushion, foot pad and backrest positions. For example, the sitting posture characteristics of the seat cushion position can include: the sum of the normalized absolute values of the difference between the seat cushion pressure distribution and the template pressure distribution The sum of the absolute values of the differences between the normalized distribution of cushion pressure and the normalized distribution of template pressure The sum of the absolute values of the relative pressure changes of the seat cushion The above values can be expressed by the following formula:

[0115]

[0116]

[0117]

[0118] In one embodiment, when the left and right tilted sitting posture detection is performed, the current sitting posture characteristics include at least: the left and right center of gravity of the seat cushion, the change in the left and right relative center of gravity of the seat cushion, the normalized difference in the left and right pressures of the seat cushion, and the difference in the left and right relative pressure changes of the seat cushion; when the crossed-legged sitting posture or the cross-legged sitting posture detection is performed, the current sitting posture characteristics include at least: the difference between the left and right pressure ratios of the front side of the seat cushion and the left and right pressure ratios of the rear side of the seat cushion, the relative pressure change of the left front side of the seat cushion and the relative pressure change of the right front side of the seat cushion, and the difference between the left and right relative pressure changes of the front side of the seat cushion and the left and right relative pressure changes of the rear side of the seat cushion; when the legs-stretched sitting posture detection is performed, the current sitting posture characteristics include at least: the pressure ratio of the front edge of the seat cushion, the difference in the relative pressure changes of the front edge of the seat cushion Ratio, the difference ratio of the relative pressure change on the front side of the seat cushion and the relative pressure change on the back side of the seat cushion; when performing the lying-over-the-table sitting posture detection, the current sitting posture characteristics at least include: the ratio of the total pressure of the seat cushion and the total pressure of the template vertical to the ground, the ratio of the total pressure of the seat cushion and foot pad and the total pressure of the template vertical to the ground, the total pressure change of the seat cushion, the total pressure change of the seat cushion and foot pad; when performing the hunched-over sitting posture detection, the current sitting posture characteristics at least include: the change of the front and rear relative center of gravity of the seat cushion, and the torque characteristics of the human spine; when performing the leg-shaking sitting posture detection, the current sitting posture characteristics at least include: the maximum value and maximum frequency of the frequency spectrum of the digital signal of the pressure sensor at the foot pad and / or foot pad position within the preset acquisition time, and the number of leg shaking times per unit time.

[0119] To describe this application in more detail, the following examples illustrate the current sitting posture characteristics under different abnormal sitting postures. Figure 5 In the abnormal sitting posture of left and right tilt, the center of gravity of the seat cushion will deviate from the midline. At this time, the current sitting posture features that can be used to detect the left and right tilt sitting posture include: the left and right center of gravity of the seat cushion Changes in the relative center of gravity of the seat cushion Normalized difference in pressure between the left and right sides of the seat cushion Relative pressure change difference between left and right sides of seat cushion Any one or more of . The above values satisfy the following formulas:

[0120]

[0121]

[0122]

[0123]

[0124] See also Figure 6In the abnormal sitting posture of crossing legs or sitting cross-legged, one leg is lifted and placed on the other leg, causing the pressure on the front side of the seat cushion on that side to be significantly smaller than the pressure on the front side of the other side, and the pressure on the front side of the seat cushion on that side to be significantly smaller than the pressure on the back side of the seat cushion on that side. Therefore, the current sitting posture features used to detect the sitting posture of crossing legs or sitting cross-legged include: the difference between the left-right pressure ratio of the front side of the seat cushion and the left-right pressure ratio of the back side of the seat cushion Relative pressure change on the left front side of the seat cushion Relative pressure change on the right front side of the seat cushion The difference between the left and right relative pressure change difference on the front side of the seat cushion and the left and right relative pressure change difference on the back side of the seat cushion Any one or more of . The above values satisfy the following formulas:

[0125]

[0126]

[0127]

[0128]

[0129] In the extended leg sitting posture, the current sitting posture characteristics that can be used to detect the extended leg sitting posture include: the pressure ratio of the front edge of the seat cushion The difference ratio of the relative pressure change at the front edge of the seat cushion The difference ratio between the relative pressure change on the front side of the seat cushion and the relative pressure change on the back side of the seat cushion Any one or more of them satisfy the following formulas:

[0130]

[0131]

[0132]

[0133] See also Figure 7 When sitting on the table, part of the vertical pressure will be transferred from the seat cushion to the foot pad and / or the table. The current sitting posture characteristics that can be used to detect the sitting posture on the table include: the ratio of the total pressure of the seat cushion to the total pressure of the template vertical to the ground The ratio of the total pressure of the seat and foot pads to the total pressure of the template vertical to the ground Total cushion pressure change Total pressure change of seat cushion and foot pad Any one or more of the above values satisfy the following formula:

[0134]

[0135]

[0136]

[0137]

[0138] See also Figure 8 , in a hunched sitting posture, the pressure distribution and total pressure of the seat cushion or foot pad will change. The current sitting posture features that can be used to detect hunched sitting posture include: the relative center of gravity change of the seat cushion Human spine torque characteristics Any one or more of the following. Among them, the relative center of gravity of the seat cushion changes Satisfies the following formula:

[0139]

[0140] In one embodiment, to solve the torque of the human spine, the following parameters need to be obtained first: the distance l_s between the center of the seat cushion and the backrest, which can be determined according to the actual seat size; the total pressure f_s of the seat cushion t and the total foot pad pressure f_f t It can be calculated based on the sensor signal; the gravity on the thigh f_t t and the gravity on the calf f_c t According to the total pressure f of the template vertical to the ground (g,0) The average weight proportions of the human thigh and calf r_t and r_c are calculated.

[0141]

[0142]

[0143] f_f t =f (7,t)

[0144] f_t t =r_t×f (g,0)

[0145] f_c t =r_c×f (g,0)

[0146]

[0147] When shaking your legs, the pressure sensor will collect quasi-periodic pressure fluctuations (optionally, the pressure sensor on the foot pad or the front of the seat cushion will have more obvious fluctuations). Considering that the upper limit of the frequency of human leg shaking is about 7Hz, the sensor sampling rate needs to be greater than 14Hz. The pressure signal collected from t0 to t1 is f i ={f (i,t) │t∈(t0,t1)}. The current sitting posture features that can be used to detect leg shaking include: fi The spectrum of fre i ={fre (i,w) │fre_max, the maximum value of w∈(w0,w1)} i and maximum frequency w_fre_max i , the number of leg shaking times per unit time Any one or more of . The above values satisfy the following formula:

[0148] fre i =FFT(f i )

[0149]

[0150] Calculate the number of leg shakes per unit time During the process, the preset leg shaking waveform template can be recorded as k = {k τ │τ∈(τ0,τ1)}, use k in f i Perform sliding convolution on the result, extract the maximum point of the result, and then perform non-maximum suppression with a specific window size (such as 0.2 seconds). Then count the number of extreme points exceeding the threshold, and finally divide by f i The length of time (t1-t0) can be used to get the number of leg shaking times per unit time.

[0151] In one embodiment, the backrest is not an abnormal sitting posture, but can help detect abnormal sitting postures. When sitting with the backrest, a pressure perpendicular to the backrest is generated, so a sensor needs to be arranged on the backrest. The current sitting posture characteristics that can be used to detect the backrest sitting posture are: backrest pressure f (8,t) and / or the ratio of the backrest pressure to the total pressure of the formwork perpendicular to the ground Among them, the ratio of total pressure Satisfies the following formula:

[0152]

[0153] For example, when the sitting posture detection device detects that the object to be detected is using the sitting posture detection device, it can calculate the characteristic values corresponding to each abnormal sitting posture based on the digital signals of each pressure sensor, and use the characteristic values as the current sitting posture features to perform further abnormal sitting posture detection and judgment.

[0154] S306: Detect the current sitting posture feature based on a sitting posture detection algorithm, and determine abnormal sitting posture information of the object to be detected according to the detection result.

[0155] It should be noted that the implementation process of step S306 can refer to the corresponding description of the aforementioned method embodiment and will not be repeated here.

[0156] It can be seen that in the sitting posture detection method shown in the embodiment of the present application, the sitting posture detection device is respectively configured with a preset number of pressure sensors at preset positions of the seat cushion, foot pad and backrest. By acquiring the analog signal collected by each pressure sensor, a digital signal for sitting posture detection and identification is obtained based on the analog signal. After completing the baseline calibration and personal calibration, the current sitting posture characteristics of the object to be detected are determined based on the digital signal, and the current sitting posture characteristics are detected based on the sitting posture detection algorithm to determine the abnormal sitting posture information of the object to be detected. The baseline calibration and personal standard make the sitting posture detection results more consistent with the object to be detected, and different sitting posture characteristics are compared according to different abnormal sitting postures. While taking into account the accuracy of sitting posture detection and user convenience, the reliability of the algorithm results is further improved.

[0157] It should be noted that, in each of the above-mentioned embodiments, there is not necessarily a certain order between the above-mentioned steps. A person skilled in the art can understand, based on the description of the embodiments of this application, that in different embodiments, the above-mentioned steps may have different execution orders, that is, they may be executed in parallel, may be executed interchangeably, and so on.

[0158] As another aspect of the present invention, an embodiment of the present invention provides a sitting posture detection device. The sitting posture detection device may be a software module comprising a plurality of instructions stored in a memory. A processor may access the memory and execute the instructions to implement the sitting posture detection method described in each of the above embodiments.

[0159] In some embodiments, the sitting posture detection device can also be constructed by hardware devices. For example, the sitting posture detection device can be constructed by one or more chips, and the chips can work in coordination with each other to complete the sitting posture detection methods described in the above embodiments. For another example, the sitting posture detection device can also be constructed by various logic devices, such as a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), a single-chip microcomputer, an ARM (Acorn RISC Machine) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination of these components.

[0160] See below Figure 9 , is a structural diagram of a sitting posture detection device provided in an embodiment of the present application. The sitting posture detection device is applied to a sitting posture detection device, which is installed on a seat and has a preset number of pressure sensors respectively configured at preset positions of the seat. Figure 9 The sitting posture detection device 90 shown includes:

[0161] The acquisition module 901 is used to acquire the analog signals collected by each pressure sensor. The analog signals are processed by analog circuits such as amplification and filtering, and digital-to-analog conversion to obtain digital signals for sitting posture detection and recognition.

[0162] The determination module 902 is configured to determine the current sitting posture characteristics of the subject to be detected based on the digital signal.

[0163] The detection module 903 is used to detect the current sitting posture characteristics based on a sitting posture detection algorithm, and determine the abnormal sitting posture information of the object to be detected according to the detection result.

[0164] As a possible design, the sitting posture detection device 90 may be connected to a host computer;

[0165] The determination module 902 is used to determine the current sitting posture characteristics of the object to be detected based on the digital signal, and is specifically used to: send the digital signal to the host computer so that the host computer determines the current sitting posture characteristics of the object to be detected based on the digital signal; and receive the current sitting posture characteristics.

[0166] And / or, the detection module 903 is used to detect the current sitting posture characteristics based on the sitting posture detection algorithm, and determine the abnormal sitting posture information of the object to be detected according to the detection results, and is specifically used to: send the current sitting posture characteristics to the host computer, so that the host computer detects the current sitting posture characteristics based on the sitting posture detection algorithm, and determines the abnormal sitting posture information of the object to be detected according to the detection results; and receive the abnormal sitting posture information.

[0167] As a possible design, the determination module 902 is used to determine the current sitting posture characteristics of the object to be detected based on the digital signal, and is also used to: determine whether the sitting posture detection device has completed baseline calibration processing and / or personal calibration processing; if both are yes, execute the determination of the current sitting posture characteristics of the object to be detected based on the digital signal.

[0168] As a possible design, the sitting posture detection device 90 further includes: a calibration module (not shown in the figure), which is used to perform the baseline calibration process every preset time range after the sitting posture detection device is in the turned-on state: when the standard deviation and the range of the digital signals of each pressure sensor in the preset time range are both within the preset threshold value, and there is no quasi-periodic change in the digital signals of each pressure sensor in the preset time range, it is determined based on the digital signal whether the difference between the average value of the digital signals of each pressure sensor in the preset time range and the current baseline of the sitting posture detection device is less than the preset difference threshold; if the difference between the average value of the digital signals of each pressure sensor in the preset time range and the current baseline of the sitting posture detection device is less than the preset difference threshold, the current baseline is updated with the average value of the digital signals of each pressure sensor in the preset time range.

[0169] As a possible design, the calibration module is further used to: obtain the biometric information of the object to be detected if the sitting posture detection device has not completed the personal calibration process; calibrate the table according to the biometric information so that the distance between the table and the sitting posture detection device is within a preset distance range; and calibrate the height of the sitting posture detection device according to the biometric information so that the height of the sitting posture detection device meets a preset height condition.

[0170] As a possible design, the biometric information of the object to be detected includes weight information and / or height information; when the calibration module is used to obtain the biometric information of the object to be detected, it is specifically used to: obtain the biometric information of the object to be detected based on the digital signal of the pressure sensor; or, receive the biometric information input by the object to be detected.

[0171] As a possible design, the preset height condition is determined based on the height information of the object to be detected and the digital signal value of the pressure sensor. The calibration module is used to calibrate the height of the sitting posture detection device based on the biometric information, so that when the height of the sitting posture detection device meets the preset height condition, it is specifically used to: find the proportional relationship between the sitting height and the height of the object to be detected based on the height information in the biometric information; determine the height range of the sitting posture detection device based on the proportional relationship; debug the sitting posture detection device from small to large according to the height range of the sitting posture detection device, and when the ratio of the digital signal value of the target pressure sensor and the gravity of the object to be detected is within the preset range, determine the height of the sitting posture detection device at the current moment as the calibration height; wherein, the target pressure sensor includes a pressure sensor at the foot pad and / or seat cushion position.

[0172] As a possible design, the method also includes: calculating the time domain mean of the digital signals of each pressure sensor of the object to be detected in a normal sitting position; calculating the total pressure value of the object to be detected perpendicular to the ground in a normal sitting position based on the time domain mean; calculating the gravity exerted on the object to be detected based on the weight information; if the difference between the gravity and the total pressure value is greater than a preset gravity difference threshold, issuing a prompt message, wherein the prompt message is used to prompt that the pressure sensor reading is incorrect or the input weight information is incorrect.

[0173] As a possible design, the determination module 902 is used to: preprocess the digital signal according to the time domain mean before determining the current sitting posture characteristics of each pressure sensor based on the digital signal; and determine the current sitting posture characteristics of the sitting posture detection device based on the preprocessed digital signal.

[0174] As a possible design, the sitting posture detection algorithm includes preset pressure distribution information and a correspondence between the preset pressure distribution information and abnormal sitting postures. The detection module 903 is configured to detect the current sitting posture characteristics based on the sitting posture detection algorithm, and when determining the abnormal sitting posture information of the subject to be detected based on the detection results, specifically, to compare the preset sitting posture characteristics with the current sitting posture characteristics for similarity; and determine the abnormal sitting posture information of the subject to be detected based on the preset sitting posture characteristics with the highest similarity and the correspondence.

[0175] As a possible design, when the detection module 903 is used to extract abnormal sitting posture features based on the pressure distribution information set, it is specifically used to: perform feature extraction processing on the pressure distribution information set based on physical principles and experimental data to obtain abnormal sitting posture features; and / or, perform feature extraction processing on the pressure distribution information set based on a deep learning model to obtain abnormal sitting posture features.

[0176] As a possible design, the detection module 903 is used to detect the current sitting posture characteristics based on the sitting posture detection algorithm, and before determining the abnormal sitting posture information of the object to be detected based on the detection results, it is also used to: collect a set of pressure distribution information of pressure sensors under various abnormal sitting postures; extract abnormal sitting posture characteristics based on the pressure distribution information set, and mark the abnormal sitting posture characteristics with label data, the label data is used to identify the correspondence between abnormal sitting posture characteristics and abnormal sitting postures; and complete the training process of the sitting posture detection algorithm based on the label data. The detection module 903 is used to detect the current sitting posture characteristics based on the sitting posture detection algorithm, and when determining the abnormal sitting posture information of the object to be detected based on the detection results, it is specifically used to: feature classify the current sitting posture characteristics based on the sitting posture detection algorithm to obtain the abnormal sitting posture information of the object to be detected.

[0177] As a possible design, the preset positions include seat cushion, foot pad and backrest positions.

[0178] As a possible design, the abnormal sitting posture information includes any one or more of the following: left-right tilted sitting posture, crossed-legged sitting posture, cross-legged sitting posture, lying on the table sitting posture, hunched-over sitting posture, and leg-shaking sitting posture; when performing the left-right tilted sitting posture detection, the current sitting posture characteristics at least include: the left-right center of gravity of the seat cushion, the change in the left-right relative center of gravity of the seat cushion, the normalized difference in the left-right pressure of the seat cushion, and the difference in the left-right relative pressure change of the seat cushion; when performing the crossed-legged sitting posture or cross-legged sitting posture detection, the current sitting posture characteristics at least include: the difference between the left-right pressure ratio of the front side of the seat cushion and the left-right pressure ratio of the rear side of the seat cushion, the relative pressure change of the left front side of the seat cushion and the relative pressure change of the right front side of the seat cushion, the difference between the left-right relative pressure change difference of the front side of the seat cushion and the left-right relative pressure change difference of the rear side of the seat cushion; when performing the legs-extended sitting posture detection, the current sitting posture characteristics to At least include: the pressure ratio of the front edge of the seat cushion, the difference ratio of the relative pressure changes of the front edge of the seat cushion, the difference ratio of the relative pressure changes of the front side of the seat cushion and the relative pressure changes of the back side of the seat cushion; when performing the lying on the table sitting posture detection, the current sitting posture characteristics at least include: the ratio of the total pressure of the seat cushion and the total pressure of the template vertical to the ground, the ratio of the total pressure of the seat cushion and foot pad and the total pressure of the template vertical to the ground, the total pressure change of the seat cushion, the total pressure change of the seat cushion and foot pad; when performing the hunched sitting posture detection, the current sitting posture characteristics at least include: the change of the relative center of gravity of the front and rear of the seat cushion, the human spine torque characteristics; when performing the leg shaking sitting posture detection, the current sitting posture characteristics at least include: the maximum value and maximum frequency of the frequency spectrum of the digital signal of the pressure sensor of the foot pad and / or foot pad position within the preset acquisition time, and the number of leg shaking times per unit time.

[0179] As a possible design, the current sitting posture feature includes a current sitting posture time domain feature and / or a current sitting posture frequency domain feature.

[0180] As a possible design, the pressure sensor is a strain gauge pressure sensor.

[0181] As a possible design, the digital signal includes a digital signal of at least one pressure sensor and / or a combination obtained based on the digital signal of the at least one pressure sensor.

[0182] It can be seen that the above-mentioned device obtains the analog signals collected by each pressure sensor, obtains the digital signal for sitting posture detection and identification based on the analog signal, determines the current sitting posture characteristics of the object to be detected based on the digital signal, and detects the current sitting posture characteristics based on the sitting posture detection algorithm to determine the abnormal sitting posture information of the object to be detected. The present application sets a pressure sensor at a preset position on the seat, extracts the sitting posture characteristics, and performs sitting posture detection based on the sitting posture detection algorithm, so that the detection accuracy of the sitting posture detection algorithm is effectively guaranteed. The user does not need to wear the device to perform sitting posture detection. In this way, the accuracy of sitting posture detection and the convenience of user use are taken into account.

[0183] It should be noted that the above-mentioned sitting posture detection device can execute the sitting posture detection method provided in the embodiments of this application, and has the corresponding functional modules and beneficial effects of executing the method. For technical details not fully described in the embodiments of the sitting posture detection device, please refer to the sitting posture detection method provided in the embodiments of this application.

[0184] See also Figure 10 , Figure 10 1 is a schematic diagram of the structure of a sitting posture detection device provided in an embodiment of the present application. The sitting posture detection device 100 includes a memory 1002 and a processor 1003, and a preset number of pressure sensors 1001 respectively configured on the seat cushion, foot pad, and backrest of the seat. Each pressure sensor 1001 and the memory 1002 are connected to one or more processors 1003, for example, via a bus.

[0185] The pressure sensor 1001 is used to collect analog signals generated by pressure.

[0186] The processor 1002 is configured to support the sitting posture detection device in executing the corresponding functions of the method in the above-mentioned method embodiment. The processor can be a central processing unit (CPU), a network processor (NP), a hardware chip, or any combination thereof. The above-mentioned hardware chip can be an application specific integrated circuit (ASIC), a programmable logic device (PLD), or any combination thereof. The above-mentioned PLD can be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0187] Memory y03 is used to store program code, etc. Memory y03 may include volatile memory (VM), such as random access memory (RAM); non-volatile memory (NVM), such as read-only memory (ROM), flash memory, a hard disk drive (HDD), or a solid-state drive (SSD); or a combination of these types of memory.

[0188] The memory 1003 can be used to store non-volatile software programs, non-volatile computer executable programs, and modules, such as the program instructions / modules corresponding to the sitting posture detection method in the embodiments of the present application. The processor executes the non-volatile software programs, instructions, and modules stored in the memory to execute various functional applications and data processing of the sitting posture detection method and the sitting posture detection device, thereby implementing the functions of the various modules or units of the sitting posture detection method and the sitting posture detection device provided in the above-mentioned method embodiments.

[0189] The memory may include a program storage area and a data storage area. The program storage area may store an operating system and application programs required for at least one function. The data storage area may store data generated based on the use of the sitting posture detection device. In some embodiments, the memory may optionally include a remote memory located relative to the processor. Such remote memory may be connected to the sitting posture detection device via a network. Examples of such networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0190] The one or more modules are stored in the memory, and when executed by the one or more processors, they execute the sitting posture detection method in any of the above method embodiments, for example, execute the method steps described in the above method embodiments, and realize the functions of the modules described in the above device embodiments.

[0191] See also Figure 11 , is a structural diagram of a sitting posture detection system provided in an embodiment of the present application. The sitting posture detection system 110 in the figure includes a seat 1101 and a sitting posture detection device 1102, which can be used to execute the sitting posture detection method described in the present application.

[0192] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a computer, the computer executes the method as described in the above embodiment.

[0193] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing related hardware through a computer program. The program can be stored in a computer-readable storage medium, and when executed, the program can include the processes in the above-described method embodiments. The storage medium can be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0194] The above disclosure is only a preferred embodiment of the present application, and certainly cannot be used to limit the scope of rights of the present application. Therefore, equivalent changes made according to the claims of the present application are still within the scope covered by the present application.

Claims

1. A sitting posture detection method, characterized in that: Applied to a sitting posture detection device, the sitting posture detection device is installed on a seat, and a preset number of pressure sensors are respectively configured at preset positions of the seat. The method includes: The analog signals collected by each pressure sensor are obtained, and the analog signals are amplified, filtered, and processed by analog circuits, and then subjected to digital-to-analog conversion to obtain digital signals for sitting posture detection and recognition; determining the current sitting posture characteristics of the object to be detected according to the digital signal; The current sitting posture feature is detected based on a sitting posture detection algorithm, and abnormal sitting posture information of the object to be detected is determined according to the detection result.

2. The method according to claim 1, wherein The sitting posture detection device is connected to the host computer; The determining of the current sitting posture feature of the object to be detected according to the digital signal includes: Sending the digital signal to the host computer so that the host computer determines the current sitting posture characteristics of the object to be detected according to the digital signal; receiving the current sitting posture feature; and / or, The detecting the current sitting posture feature based on the sitting posture detection algorithm and determining the abnormal sitting posture information of the object to be detected according to the detection result includes: Sending the current sitting posture feature to the host computer, so that the host computer detects the current sitting posture feature based on a sitting posture detection algorithm, and determines abnormal sitting posture information of the subject to be detected according to the detection result; Receive the abnormal sitting posture information.

3. The method according to claim 1, wherein Before determining the current sitting posture feature of the object to be detected according to the digital signal, the method further includes: Determining whether the sitting posture detection device has completed baseline calibration processing and / or personal calibration processing for the current time period; If both are yes, then the step of determining the current sitting posture feature of the object to be detected based on the digital signal is executed.

4. The method according to claim 3, wherein The method further comprises: After the sitting posture detection device is turned on, the baseline calibration process is performed every preset time range: When the standard deviation and the range of the digital signals of the pressure sensors within the preset time range are both within the preset thresholds, and there is no quasi-periodic change in the digital signals of the pressure sensors within the preset time range, determining, based on the digital signals, whether a difference between an average value of the digital signals of the pressure sensors within the preset time range and a current baseline of the sitting posture detection device is less than a preset difference threshold; If the difference between the average value of the digital signals of each pressure sensor within the preset time range and the current baseline of the sitting posture detection device is less than a preset difference threshold, the current baseline is updated with the average value of the digital signals of each pressure sensor within the preset time range.

5. The method according to claim 3 or 4, wherein: The method further comprises: If the sitting posture detection device has not completed the personal calibration process, obtaining the biometric information of the object to be detected; Calibrate the table according to the biometric information so that the distance between the table and the sitting posture detection device is within a preset distance range; The height of the sitting posture detection device is calibrated according to the biometric information so that the height of the sitting posture detection device meets a preset height condition.

6. The method according to claim 5, wherein The biometric information of the subject to be detected includes weight information and / or height information; The acquiring of the biometric information of the object to be detected includes: Obtaining biometric information of the object to be detected according to the digital signal of the pressure sensor; or receiving biometric information input by the object to be detected.

7. The method according to claim 6, wherein The preset height condition is determined according to the height information of the object to be detected and the digital signal value of the pressure sensor; The step of calibrating the height of the sitting posture detection device according to the biometric information so that the height of the sitting posture detection device meets a preset height condition includes: Finding the ratio between sitting height and height based on the height information in the biometric information; determining a height range of the sitting posture detection device according to the proportional relationship; Debugging the sitting posture detection device according to a height range from small to large, and when the ratio of the digital signal value of the target pressure sensor and the gravity of the object to be detected is within a preset range, determining the current height of the sitting posture detection device as the calibration height; Wherein, the target pressure sensor includes a pressure sensor at the foot pad and / or seat cushion position.

8. The method according to claim 6 or 7, wherein: The method further comprises: Calculating the time domain mean of the digital signals of each pressure sensor of the subject to be detected in a normal sitting posture; Calculating the total pressure value of the object to be detected perpendicular to the ground in a normal sitting posture according to the time domain mean; Calculating the gravity exerted on the object to be detected according to the weight information; If the difference between the gravity and the total pressure value is greater than a preset gravity difference threshold, a prompt message is issued, where the prompt message is used to prompt that the pressure sensor reading is incorrect or the input weight information is incorrect.

9. The method according to claim 8, wherein Before determining the current sitting posture characteristics of each pressure sensor according to the digital signal, the method further includes: Preprocessing the digital signal according to the time domain mean; The current sitting posture feature of the sitting posture detection device is determined according to the preprocessed digital signal.

10. The method according to claim 1, wherein The sitting posture detection algorithm includes preset sitting posture features and the corresponding relationship between the preset sitting posture features and abnormal sitting postures; The detecting the current sitting posture feature based on the sitting posture detection algorithm and determining the abnormal sitting posture information of the object to be detected according to the detection result includes: Comparing the preset sitting posture feature with the current sitting posture feature for similarity; The abnormal sitting posture information of the object to be detected is determined according to the preset sitting posture feature with the highest similarity and the corresponding relationship.

11. The method according to claim 1, wherein Before detecting the current sitting posture feature based on the sitting posture detection algorithm and determining the abnormal sitting posture information of the object to be detected according to the detection result, the method further includes: Collect pressure distribution information of pressure sensors under various abnormal sitting postures; Extracting abnormal sitting posture features according to the pressure distribution information set, and marking the abnormal sitting posture features with label data, wherein the label data is used to identify the corresponding relationship between the abnormal sitting posture features and the abnormal sitting posture; completing a training process of a sitting posture detection algorithm according to the label data; The detecting the current sitting posture feature based on the sitting posture detection algorithm and determining the abnormal sitting posture information of the object to be detected according to the detection result includes: The current sitting posture features are classified based on the sitting posture detection algorithm to obtain abnormal sitting posture information of the object to be detected.

12. The method according to claim 11, wherein The extracting abnormal sitting posture features according to the pressure distribution information set includes: Perform feature extraction processing on the pressure distribution information set based on physical principles and experimental data to obtain abnormal sitting posture features; and / or, The pressure distribution information set is subjected to feature extraction processing based on a deep learning model to obtain abnormal sitting posture features.

13. The method according to claim 1, wherein The preset positions include the seat cushion, foot pad and backrest positions.

14. The method according to claim 13, wherein The abnormal sitting posture information includes any one or more of the following: leaning left or right sitting posture, sitting with legs crossed, sitting with legs crossed, sitting with the back of the table bent, sitting with the back hunched, and sitting with legs shaking; When performing the left-right tilt sitting posture detection, the current sitting posture features include: any one or more of the left-right center of gravity of the seat cushion, the left-right relative center of gravity change of the seat cushion, the normalized difference of the left-right pressure of the seat cushion, and the left-right relative pressure change difference of the seat cushion; When performing the cross-legged sitting posture detection, the current sitting posture feature includes: any one or more of the difference between the left and right pressure ratios of the front side of the seat cushion and the left and right pressure ratios of the rear side of the seat cushion, the relative pressure change of the left front side of the seat cushion and the relative pressure change of the right front side of the seat cushion, and the difference between the left and right relative pressure change differences of the front side of the seat cushion and the left and right relative pressure change differences of the rear side of the seat cushion; When performing the legs-extended sitting posture detection, the current sitting posture characteristics include: any one or more of: a pressure ratio of a front edge of a seat cushion, a difference ratio of relative pressure changes of the front edge of the seat cushion, and a difference ratio of relative pressure changes of the front side of the seat cushion and relative pressure changes of the back side of the seat cushion; When performing the lying-on-the-desk sitting posture detection, the current sitting posture characteristics include: any one or more of the ratio of the total pressure of the seat cushion to the total pressure of the template perpendicular to the ground, the ratio of the total pressure of the seat cushion and foot pad to the total pressure of the template perpendicular to the ground, the change in the total pressure of the seat cushion, and the change in the total pressure of the seat cushion and foot pad; When performing the hunched sitting posture detection, the current sitting posture characteristics include: any one or more of the front-to-back relative center of gravity change of the seat cushion and the human spine torque characteristics; When performing the leg-shaking sitting posture detection, the current sitting posture characteristics include: any one or more of the maximum value and maximum frequency of the spectrum of the digital signal of the foot pad and / or the pressure sensor at the foot pad position within the preset acquisition time, and the number of leg shaking times per unit time.

15. The method according to claim 1, wherein The current sitting posture feature includes a current sitting posture time domain feature and / or a current sitting posture frequency domain feature.

16. The method according to claim 1, wherein The pressure sensor is a strain gauge pressure sensor.

17. The method according to claim 1, wherein The digital signal includes a digital signal of at least one pressure sensor and / or a combination obtained based on the digital signal of the at least one pressure sensor.

18. A sitting posture detection device, characterized in that: Applied to a sitting posture detection device, the sitting posture detection device is installed on a seat, and a preset number of pressure sensors are respectively configured at preset positions of the seat. The sitting posture detection device includes: An acquisition module is used to acquire analog signals collected by each pressure sensor. The analog signals are processed by analog circuits such as amplification and filtering, and then digital-to-analog conversion to obtain digital signals for sitting posture detection and recognition; a determination module, configured to determine a current sitting posture feature of the object to be detected based on the digital signal; The detection module is used to detect the current sitting posture characteristics based on the sitting posture detection algorithm, and determine the abnormal sitting posture information of the object to be detected according to the detection result.

19. A sitting posture detection device, characterized in that: The sitting posture detection device includes a memory and a processor, and a preset number of pressure sensors respectively arranged at preset positions of the seat, each pressure sensor and the memory are connected to the processor, and the processor is used to execute one or more computer programs stored in the memory. When the processor executes the one or more computer programs, the sitting posture detection device implements the method described in any one of claims 1-17.

20. A sitting posture detection system, characterized in that: The sitting posture detection system includes a seat and the sitting posture detection device as claimed in claim 19.

21. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor is caused to perform the method according to any one of claims 1 to 17.

Images