Sleeping posture recognition system

Abstract

The invention provides a sleeping posture recognition system, which comprises a pressure pad, a signal acquisition system and an upper computer, a flexible piezoresistive sensing unit is arranged on the pressure pad, the flexible piezoresistive sensing unit is connected with the signal acquisition system, the signal acquisition system is connected with the upper computer through a plurality of serial ports, acquired pressure signals are sent to the upper computer, and the upper computer is connected with the signal acquisition system. A sleeping posture recognition model is integrated in the upper computer, and the current sleeping posture is presented in real time through the sleeping posture recognition model. The pressure pad adopts a three-layer structural design, and has the characteristics of wide sensing range, high sensitivity, high response speed, good durability and the like. The invention provides a voltage source cross exchange strategy, so that all parasitic charges are released in time, the cross coupling phenomenon caused by a high-density large-scale sensor can be well solved, high measurement precision among rows and columns and good stability under high-density arrangement are ensured, and the precision of analog quantity measured by the sensor is greatly improved.

Description

Technical Field

[0001] The invention belongs to the technical field of human health detection, and in particular relates to a sleeping posture recognition system. Background Art

[0002] Bedsores refer to skin and subcutaneous tissue damage caused by long-term pressure on the skin. If not treated in time or if underlying health conditions hinder healing, they can be life-threatening, especially for disabled elderly people who have been bedridden for a long time. The most common sleep monitoring system in the commercial field is smart wearable devices, which are easy to wear and can monitor at any time and anywhere. However, such wearable devices cannot fully and accurately reflect the sleeping posture of the whole body, especially movements such as turning over. In addition, as a typical restrained monitoring device, patients may be affected by the foreign body sensation caused by wearing it and affect their normal sleep. Vision-based sleeping posture recognition methods can achieve a high sleeping posture recognition accuracy and provide data support for the prevention of bedsores. However, there are some limitations, such as being easily blocked and sensitive to light. In addition, due to privacy issues, vision-based solutions are often difficult to accept in the field of patient behavior monitoring. The accuracy of sleeping posture recognition cannot be guaranteed due to limitations in sensor density. For example, only approximate supine and side-lying postures can be detected, and movements such as turning over cannot be accurately recognized. The current pressure-based sleeping posture recognition devices or systems in the commercial and academic fields at home and abroad have the following shortcomings: (1) The limitation of sensor density or the number of sensor points leads to poor sleeping posture recognition performance. The sensing density of current state-of-the-art systems is generally less than 1 point / cm 2 . The low sensing density will hinder the large-scale integration of sensing points and further lead to the lack of precise collection of body pressure data; (2) The severe cross-coupling crosstalk phenomenon caused by high-density and large-scale sensor arrays. The complex cross-coupling crosstalk problem between rows and columns caused by high-density sensor arrays needs to be taken into account. This phenomenon will cause mutual interference between signals between different axes, resulting in the inability to accurately obtain the measurement value of a single axis and distortion of the sensor measurement results; when the sensor is subject to external interference, cross-coupling will aggravate the fluctuation of the system, making it difficult for the system to maintain a stable state. As a result, there is no longer a mapping relationship between the pressure at the sensor point and the analog value obtained. Summary of the invention

[0003] In view of this, the present invention aims to overcome the deficiencies of the above-mentioned problems in the prior art and proposes a sleeping posture recognition system.

[0004] To achieve the above object, the technical solution of the present invention is achieved as follows:

[0005] A sleeping posture recognition system comprises a pressure pad, a signal acquisition system, and a host computer. The pressure pad is provided with a flexible piezoresistive sensing unit, and the flexible piezoresistive sensing unit is connected to the signal acquisition system. The signal acquisition system is connected to the host computer via multiple serial ports and sends the collected pressure signals to the host computer. A sleeping posture recognition model is integrated in the host computer, and the sleeping posture recognition model is used to present the current sleeping posture in real time.

[0006] Furthermore, the pressure pad includes 6 small pressure pads of 200 cm², each of which includes a top electrode layer, a bottom electrode layer and an intermediate fiber matrix dielectric layer, the top electrode layer includes a top polyimide film, copper electrodes are arranged laterally below the top polyimide film, the bottom electrode layer includes a bottom polyimide film, copper electrodes are arranged vertically above the bottom polyimide film, the intermediate fiber matrix dielectric layer includes conductive fabric, the top electrode layer and the bottom electrode layer are cross-assembled to clamp the conductive fabric to form a small pressure pad, and the intersection of the top electrode layer and the bottom electrode layer constitutes a flexible piezoresistive sensing unit.

[0007] Furthermore, there are 3,600 flexible piezoresistive sensing unit points on each small pressure pad, totaling 21,600 points, forming a high-density large-scale sensing unit arrangement pressure pad.

[0008] Furthermore, the signal acquisition system includes 6 circuit boards to realize the pressure signal acquisition of different pressure pads. Each circuit board is provided with a signal measurement circuit, and the signal measurement circuit includes an MCU processor, a multi-decoder, an analog switch, a voltage follower circuit, and a median filter. The MCU processor controls the analog switch through the multi-decoder to connect different rows and columns of the small pressure pads to determine the position of the sensor unit point that needs to be collected. The generated signal passes through a voltage follower circuit, and the analog signal is converted into a digital signal through an ADC. The acquired original sensor unit data is uploaded to the host computer through USB after median filtering.

[0009] Furthermore, the sleeping posture recognition model includes: an input module, the input module is connected to convolution kernels of sizes 1×1, 3×3 and 5×5 and a maximum pooling layer; the outputs of the three convolution kernels are connected to the fusion module, the output of the maximum pooling layer is connected to the input of the fusion module through a 3×3 convolution kernel, the output of the fusion module is connected to a 5×5 convolution kernel, the output of the convolution kernel is connected to a maximum pooling layer, the output of the maximum pooling layer is connected to a 3×3 convolution kernel, the output of the convolution kernel is connected to a maximum pooling layer, the output of the maximum pooling layer is connected to a global average pooling layer, the global average pooling layer is connected to a batch normalization layer; the batch normalization layer is connected to a global average pooling layer, and finally connected to the output layer.

[0010] Furthermore, the implementation process of the sleeping posture recognition model is as follows:

[0011] The image is inputted through the input module, and the image data is inputted into three convolution kernels of different sizes to extract sleeping posture information of different scales;

[0012] The image data is input into the maximum pooling layer, and the maximum pooling layer is used to extract features with spatial invariance at the same time;

[0013] The fusion module is used to calculate the output of three convolution kernels of different sizes and the maximum pooling layer in the depth dimension to extract temporal features;

[0014] Use a global average pooling to map multi-dimensional features to a one-dimensional vector;

[0015] The output layer is used to obtain the probability distribution of different sleeping positions.

[0016] Furthermore, a voltage source cross-interchange circuit is provided between each small pressure pad and the circuit board, and the voltage source cross-interchange circuit includes a flexible flat cable, and the copper electrodes of the top electrode layer and the copper electrodes of the bottom electrode layer of the small pressure pad are respectively connected to the circuit board through the flexible flat cable and form an electrical connection with the multi-way analog switch.

[0017] Compared with the prior art, the sleeping posture recognition system described in the present invention has the following advantages:

[0018] (1) The present invention realizes the development of high spatial density flexible piezoresistive sensor units and the layout strategy of large-scale sensor units, conducts sensitivity analysis between the overall sleeping posture recognition accuracy, and determines the layout density of the sensor units on the basis of ensuring an acceptable sleeping posture recognition accuracy.

[0019] (2) The present invention proposes a solution strategy for the cross-coupling crosstalk caused by high-density and large-scale sensor unit arrays. The complex cross-coupling crosstalk between rows and columns caused by high-density sensor unit arrays cannot be solved from the sensor body structure. This phenomenon causes the mapping relationship between the sensor unit point pressure and the acquired analog quantity to no longer exist. The present invention ensures that the signals of different sensor units are acquired in time by designing a voltage source cross-interchange circuit.

[0020] (3) Realize the construction of a high-precision unrestrained sleeping posture recognition system based on body pressure analysis data and deep learning. The acquired body pressure information is uploaded to the host system in real time based on multi-port and real-time communication. The host system developed in Python includes a real-time display platform for sleeping posture pressure images, a deep learning engine, and a telemedicine system. The pattern recognition algorithm based on deep learning technology will perform high-precision recognition of the patient's current sleeping posture, and the rehabilitation physician can provide timely guidance and adjustments to the patient's bed status through the telemedicine system, creating conditions for home rehabilitation. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the accompanying drawings:

[0022] Figure 1 A schematic diagram of the materials used for the small pressure pad of the present invention;

[0023] Figure 2 This is a schematic diagram of the flexible piezoresistive sensing unit of the present invention;

[0024] Figure 3 It is an overall schematic diagram of a pressure pad used for collecting sleeping posture information of the present invention;

[0025] Figure 4 It is a schematic diagram of the structure of the signal measurement circuit of the present invention;

[0026] Figure 5 This is a schematic diagram of the pressure pad signal acquisition hardware system of the present invention;

[0027] Figure 6 This is a schematic diagram of the real-time display system of pressure information of the present invention;

[0028] Figure 7 is a schematic diagram of a sleeping posture recognition model used in the present invention;

[0029] Figure 8 It is an overall schematic diagram of the voltage source cross-interchange strategy of the present invention;

[0030] Fig. 9 This is a schematic diagram of supine posture pressure recognition according to the present invention;

[0031] Fig.10 This is a schematic diagram of prone posture pressure recognition according to the present invention;

[0032] Fig.11 This is a schematic diagram of pressure recognition in the left side lying posture of the present invention;

[0033] Fig.12 It is a schematic diagram of pressure recognition in the right side lying posture according to the present invention. DETAILED DESCRIPTION

[0034] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0035] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and the like are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, features defined as "first", "second", and the like may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "multiple" means two or more.

[0036] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood by specific circumstances.

[0037] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0038] The present invention provides a sleeping posture recognition system, including a pressure pad, a signal acquisition system, and a host computer. The pressure pad is provided with a flexible piezoresistive sensing unit, and the flexible piezoresistive sensing unit is connected to the signal acquisition system. The signal acquisition system is connected to the host computer through multiple serial ports, and the collected pressure signals are sent to the host computer. A sleeping posture recognition model is integrated in the host computer, and the sleeping posture recognition model is used to present the current sleeping posture in real time.

[0039] In the present invention, a complete pressure pad is composed of 6 small pressure pads of 400 cm², such as Figure 1 As shown, each small pressure pad includes a top electrode layer, a bottom electrode layer and an intermediate fiber matrix dielectric layer, the top electrode layer includes a top polyimide film, copper electrodes are arranged horizontally below the top polyimide film, the bottom electrode layer includes a bottom polyimide film, copper electrodes are arranged vertically above the bottom polyimide film, the intermediate fiber matrix dielectric layer includes a conductive fabric, the top electrode layer and the bottom electrode layer are cross-assembled to clamp the conductive fabric to form a small pressure pad, and the intersection of the top electrode layer and the bottom electrode layer constitutes a flexible piezoresistive sensing unit. Among them, Figure 1(a) shows the structure of each small pressure pad. Figure 1 (b) describes the shape of the sensor unit in the non-compressed and compressed states. Specifically, the fabric is compressed, resulting in a smaller distance between the upper and lower electrode layers, further reducing the resistance between the upper and lower electrode layers.

[0040] The number of sensing unit points on each small pressure pad of the present invention is as high as 3,600, and the total number of sensing unit points of the pressure pad is as high as 21,600, with a density of up to 9 points within 1 cm², achieving a sleeping posture recognition accuracy of 97.29%. Figure 2 A schematic diagram of the sensing unit is shown, where Figure 2 (a) SEM images of nonwoven fabric impregnated with MXene dispersion at low and high magnifications. Figure 2 (b) is the sensitivity curve of the sensor unit to external mechanical pressure. The red line represents the sensitivity curve of the sensor unit to external mechanical pressure, and the blue line represents the sensitivity of the sensor at different stages. S1, S2, and S3 represent the sensitivity of the sensor at different stages, respectively. Figure 2 (c) is the response and recovery time of the sensor unit tested by 28kPa loading / unloading. The blue line is the response curve of the sensor unit tested by 28kPa loading / unloading. The red line represents the response and recovery time of the sensor unit. Figure 2 (d) is the corresponding behavior of the sensing unit under different compression rates. Figure 2 (e) The cyclic behavior of the sensor unit was tested by applying more than 10,000 continuous dynamic loading / unloading cycles at 28 kPa. The left red line represents the local enlarged image of the sensor unit cyclic test 1700-1708 cycles, and the right red line represents the local enlarged image of the sensor unit cyclic test 8300-8308 cycles. The pressure pad for sleeping posture recognition is prepared as shown in FIG. Figure 3 shown. Figure 3 (a) is a schematic diagram of the entire pressure pad and a partial schematic diagram; Figure 3 (b) is the signal response image of the pressure pad actually obtained when the hand actually presses on the pressure pad.

[0041] In the present invention, the signal acquisition system includes 6 circuit boards to realize the pressure signal acquisition of different pressure pads. Each circuit board is provided with a signal measurement circuit, which includes an MCU processor, a multi-decoder, an analog switch, a voltage follower circuit, and a median filter. The MCU processor controls the analog switch through the multi-decoder to connect different rows and columns of the small pressure pads to determine the position of the sensor unit point to be collected. The generated signal passes through a voltage follower circuit and an ADC to convert the analog signal into a digital signal. The original sensor unit data obtained is uploaded to the host computer through USB after median filtering. The signal measurement circuit structure is as follows Figure 4As shown in the figure, after the system is initialized, the MCU (STM32F103VET6) controls the analog switch through the multi-decoder to connect the copper electrodes of different rows and columns of the top electrode layer and the bottom electrode layer to determine the position of the sensor unit point to be collected. The generated signal passes through a voltage follower circuit and is converted into a digital signal through an ADC. The acquired raw sensor data is median filtered and uploaded to the software system developed by the Pycharm editor in the host computer through USB.

[0042] The pressure pad signal acquisition hardware system of the present invention is as follows Figure 5 The hardware system is modularly designed into 6 independent circuit boards to collect pressure signals from different pressure pads and upload them to the host computer through serial communication. Figure 6 The upper computer software developed in Python language is shown, which presents the current sleeping posture in real time by receiving the pressure pad signal.

[0043] The host computer software of the present invention integrates a sleeping posture recognition model, and the implementation process is as follows: Figure 7 As shown in the figure. The input image data is first fed into convolution kernels of different sizes at the same time. The convolution kernel sizes are set to 1×1, 3×3 and 5×5 respectively to fully extract sleeping posture information of different scales. The 3×3 maximum pooling layer is used to extract features with spatial invariance at the same time. The outputs of these layers are calculated in the depth dimension, which is used to extract temporal features. The next layer is two convolution layers with 64 and 128 filters respectively. And each convolution layer is followed by a maximum pooling layer. Subsequently, a global average pooling (GAP) is used to map the multi-dimensional features to a one-dimensional vector. This step is to replace the fully connected layer, thereby greatly reducing the number of weights. There is a global average pooling (GAP) followed by a batch normalization layer (BN). The last output layer (a dense layer with a Softmax function classifier) ​​obtains the probability distribution of different sleeping postures.

[0044] In high-density and large-scale sensor arrays, due to the small spacing between sensor elements, the electromagnetic fields of each element are prone to overlap with each other, and the electromagnetic fields of each sensitive unit interact with each other, resulting in unnecessary coupling of signals during transmission. To solve this problem, the present invention proposes a voltage source cross-interchange strategy, in which a voltage source cross-interchange circuit is set between each small pressure pad and the circuit board. The voltage source cross-interchange circuit includes a flexible flat cable, and the copper electrodes of the top electrode layer and the bottom electrode layer of the small pressure pad are respectively connected to the circuit board through the flexible flat cable, and are electrically connected to the multi-way analog switch. Figure 8As shown, select any sensing unit point P, and when measuring the resistance at point P, the single-pole double-throw switch S1 is closed. In the first measurement, S0 is connected to GND and S2 is connected to VCC. Take the voltage output Vout=Vout1 at this time, then ;

[0045] Where VCC is the power supply voltage (3.3V is adopted), R ref is the reference resistor, R point is the resistance at point P, R othor is the equivalent resistance caused by cross coupling. In the second measurement, S0 is connected to VCC and S2 is connected to GND. The voltage output V out =V out2 ,but ;

[0046] By interchanging the results of the two measurements before and after the voltage source, an accurate resistance at point P can be obtained, which is expressed as ;

[0047] The resistance R at point P is found by measuring point Only with V out1 and V out2 Related, avoiding the R generated by other points other Therefore, the value of the sensor unit directly reflects the resistance of point P.

[0048] A set of experiments has been conducted to test the sleeping posture recognition performance of the proposed system. A total of 16 subjects (10 males and 6 females aged between 21 and 53 years old) participated in the experiment. Each subject signed an informed consent form before the experiment. Each patient was required to perform four sleeping postures: supine, prone, left side, and right side. Figures 9 to 12 Displays image data collected for four sleeping positions.

[0049] During the experiment, this study provided specific posture instructions to the subjects. To ensure that the system's performance is applicable to different body types and sleeping positions, this study collected a large amount of sleeping position data and did not restrict the subjects' posture differences. For example, the subjects' limb angles were different when performing the supine position, such as Fig. 9 and Fig.10 As shown, the waist and arm angles were different. In addition to different limb angles, variations of the same posture were also tried, such as when performing the prone sleeping position, Fig.11 It shows the symmetrical support of the elbows on both sides. Fig.12The left elbow is shown to be in an upward position, while the right elbow is in a downward position. In this study, each subject collected 30 sets of data for each sleeping position, so the database contains a total of 1920 sets of sleeping position data.

[0050] Any matters not described in the present invention are applicable to the prior art.

[0051] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A sleeping posture recognition system, characterized in that: The invention comprises a pressure pad, a signal acquisition system, and a host computer. The pressure pad is provided with a flexible piezoresistive sensing unit, which is connected to the signal acquisition system. The signal acquisition system is connected to the host computer through multiple serial ports to send the collected pressure signal to the host computer. The host computer is integrated with a sleeping posture recognition model, and the sleeping posture recognition model is used to present the current sleeping posture in real time. The pressure pad comprises 6 small pressure pads of 200 cm², each of which comprises a top electrode layer, a bottom electrode layer and an intermediate fiber matrix medium layer. The top electrode layer comprises a top layer of polyimide. The invention relates to a membrane, wherein copper electrodes are arranged at intervals laterally below the top polyimide membrane, the bottom electrode layer comprises a bottom polyimide membrane, copper electrodes are arranged at intervals vertically above the bottom polyimide membrane, the intermediate fiber matrix medium layer comprises a conductive fabric, the top electrode layer and the bottom electrode layer are cross-assembled to clamp the conductive fabric to form a small pressure pad, and the intersection of the top electrode layer and the bottom electrode layer constitutes a flexible piezoresistive sensing unit; the number of flexible piezoresistive sensing unit points on each small pressure pad is 3,600, and the total number of points is 21,600, so as to construct a high-density large-scale sensing unit arrangement pressure pad.

2. A sleeping posture recognition system according to claim 1, characterized in that: The signal acquisition system includes 6 circuit boards to realize the pressure signal acquisition of different pressure pads. Each circuit board is provided with a signal measurement circuit, which includes an MCU processor, a multi-decoder, an analog switch, a voltage follower circuit, and a low-pass filter. The MCU processor controls the analog switch through the multi-decoder to connect different rows and columns of the small pressure pads to determine the position of the sensor unit point that needs to be collected. The generated signal passes through a voltage follower circuit and is converted into a digital signal through an ADC. The acquired original sensor unit data is uploaded to the host computer through a USB after low-pass filtering.

3. A sleeping posture recognition system according to claim 1, characterized in that: The sleeping posture recognition model includes: an input module, the input module is connected to convolution kernels of sizes 1×1, 3×3 and 5×5 and a maximum pooling layer; the outputs of the three convolution kernels are connected to a fusion module, the output of the maximum pooling layer is connected to the input of the fusion module through a 3×3 convolution kernel, the output of the fusion module is connected to a 5×5 convolution kernel, the output of the convolution kernel is connected to a maximum pooling layer, the output of the maximum pooling layer is connected to a 3×3 convolution kernel, the output of the convolution kernel is connected to a maximum pooling layer, the output of the maximum pooling layer is connected to a global average pooling layer, the global average pooling layer is connected to a batch normalization layer; the batch normalization layer is connected to a global average pooling layer, and finally connected to the output layer.

4. A sleeping posture recognition system according to claim 3, characterized in that: The implementation process of the sleeping posture recognition model is as follows: The image is inputted through the input module, and the image data is inputted into three convolution kernels of different sizes to extract sleeping posture information of different scales; The image data is input into the maximum pooling layer, and the pooling layer is used to extract spatially invariant features at the same time; The fusion module is used to calculate the output of three convolution kernels of different sizes and the maximum pooling layer in the depth dimension to extract temporal features; Use a global average pooling to map multi-dimensional features to a one-dimensional vector; The output layer is used to obtain the probability distribution of different sleeping positions.

5. A sleeping posture recognition system according to claim 2, characterized in that: A voltage source cross-interchange circuit is provided between each small pressure pad and the circuit board. The voltage source cross-interchange circuit includes a flexible flat cable. The copper electrodes of the top electrode layer and the copper electrodes of the bottom electrode layer of the small pressure pad are respectively connected to the circuit board through the flexible flat cable and form an electrical connection with the multi-way analog switch.

Images