Sitting posture classification method based on cushion

Abstract

The invention discloses a sitting posture classification method based on a cushion, and belongs to the technical field of sitting posture classification. According to the invention, the problems of low classification accuracy, high cost and poor adaptability of a traditional sitting posture classification method are solved. The force sensitive resistors and the accelerometers corresponding to the ischial tuberosity and the thigh area of the human body are fixed on the cushion main body, so that the sitting posture signal data of the human body can be comprehensively and effectively collected, and the sitting posture classification cost is effectively reduced. By solving a pressure-acceleration sitting posture signal data correlation index matrix, screening out data strongly related to sitting posture classification; the method comprises the following steps: screening signal data to be classified, fusing the screened signal data strongly related to sitting posture classification, effectively establishing a relation between the data, and finally processing the signal data to be classified and a fusion result together by constructing a correlation degree characteristic distance and a correlation degree characteristic distance classification algorithm to obtain a human body sitting posture classification result. The method can be applied to the field of human body sitting posture classification.

Description

Technical Field

[0001] The invention belongs to the technical field of sitting posture classification, and in particular relates to a sitting posture classification method based on a seat cushion. Background Art

[0002] In modern society, people sit for a long time at work, study and in life. However, incorrect sitting posture may lead to a variety of health problems, such as scoliosis, cervical and lumbar spine diseases, and also affect work efficiency and quality of life. Therefore, accurate detection and classification of sitting posture is of great significance for timely reminding people to maintain correct sitting posture and prevent health problems.

[0003] There are many sitting posture detection methods and technologies at present. For example, document 1 (Roland Z, Matteo T, Plüss Stefan, et al. Application of Machine Learning Approaches for Classifying Sitting Posture Based on Force and Acceleration Sensors. BioMed Research International, 2016: 1-9. DOI: 10.1155 / 2016 / 5978489.) uses 16 pressure sensors and 1 inertial sensor. The pressure sensors are arranged in the seat pan, on the backrest and each armrest, and the inertial sensor is arranged behind the backrest. However, these sensors are simply listed, and there is basically no connection between the sensors. Moreover, for the sensors arranged on the backrest and armrest, different people may not cause the data of these sensors to change. Therefore, the error of sensor data measurement is large. At the same time, due to the lack of connection between the data of each sensor, the accuracy of sitting posture classification based on the data collected by the sensor is low, and the use of a large number of pressure sensors will lead to high overall costs. Reference 2 (Kerr J, Carlson J, Godbole S, et al. Improving Hip-Worn Accelerometer Estimates of Sitting Using Machine Learning Methods. Medicine & Science in Sports & Exercise, 2018: 1. DOI: 10.1249 / MSS.0000000000001578.) only involves accelerometers. Because the collected data is often not comprehensive and accurate, it is difficult to accurately distinguish different types of sitting postures. In addition, the accelerometer is worn on the front of the subject's thigh, which is an invasive measurement of the human body.Reference 3 (Ran X, Wang C, Xiao Y, et al. A portable sitting posture monitoring system based on a pressure sensor array and machine learning [J]. Sensors and Actuators A: Physical, 2021, 331: 112900-. DOI: 10.1016 / j.sna.2021.112900.) and Reference 4 (Jawad A, Johan, Sidén, Henrik AA Proposal of Implementation of Sitting Posture Monitoring System for Wheelchair Utilizing Machine Learning Methods. Sensors. 2021 21 (19): 6349. DOI: 10.3390 / s21196349.) both use force sensor arrays, which greatly increases the system cost and the accuracy of sitting posture classification is still relatively limited. In addition, some existing technologies perform poorly in terms of adaptability, cannot effectively reflect changes in sitting posture, and are difficult to adapt to the needs of different groups of people.

[0004] In summary, the traditional sitting posture classification method still has the problems of low classification accuracy, high cost and poor adaptability. Therefore, in order to overcome the shortcomings of the existing technology, providing a more advanced and effective solution for sitting posture detection and classification is an urgent problem to be solved. Summary of the invention

[0005] The purpose of the present invention is to solve the problems of low classification accuracy, high cost and poor adaptability of traditional sitting posture classification methods, and propose a sitting posture classification method based on a cushion.

[0006] The technical solution adopted by the present invention to solve the above technical problems is: a sitting posture classification method based on a seat cushion, the method specifically comprising the following steps:

[0007] Step 1: Obtain the reference data matrix for sitting posture classification from the server;

[0008] Step 2: When the user sits on the cushion, the force-sensitive resistors and accelerometers on the cushion are used to collect the user's signal data, and then the data associated with the sitting posture classification is selected from the collected user signal data;

[0009] Step 3: Obtain the user's sitting posture classification result based on the data selected in step 2 and the reference data matrix.

[0010] The beneficial effects of the present invention are:

[0011] The present invention fixes 4 force-sensitive resistors and 2 accelerometers on the seat cushion body, and these sensors correspond to the ischial tuberosity and thigh area of ​​the human body respectively, so as to comprehensively and effectively collect the sitting posture signal data of the human body, and provide a reliable data source for accurately classifying the sitting posture. By solving the correlation index matrix of the pressure-acceleration sitting posture signal data, the strong correlation index data of the pressure-acceleration sitting posture signal with the absolute value of the correlation index above 0.8 is screened out; by fusing the screened signal data with the strong correlation index of the sitting posture classification, the connection between the data is effectively established, and finally, by constructing the correlation feature distance and the correlation feature distance classification algorithm, the signal data to be classified is processed together with the fusion result, thereby effectively improving the accuracy and efficiency of classifying different human sitting postures, and realizing the effective distinction between a variety of common sitting postures, while improving the adaptability of the classification method.

[0012] At the same time, since the method of the present invention does not need to use a large number of force sensors, the cost of sitting posture classification is effectively reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 is a flow chart of a method for classifying sitting postures based on a cushion of the present invention;

[0014] Figure 2 It is a structural schematic diagram of a seat cushion used for sitting posture classification in the present invention;

[0015] Among them, the cushion body 1, the first force-sensitive resistor 2, the second force-sensitive resistor 3, the third force-sensitive resistor 4, the fourth force-sensitive resistor 5, the first accelerometer 6, the second accelerometer 7, the first fixing belt 8, the second fixing belt 9, the third fixing belt 10, the fourth fixing belt 11, the first central axis 12, and the second central axis 13;

[0016] Figure 3 It is a normalized curve diagram of the numerical values ​​of the signal data collected by the force-sensitive resistor;

[0017] Figure 4 It is a numerical normalized curve diagram of the signal data collected by the accelerometer;

[0018] Figure 5 is the maximum correlation index under various sitting postures of the subjects;

[0019] Figure 6 It is a schematic diagram of the confusion matrix for classifying and predicting the subject data using the association feature distance classification algorithm. DETAILED DESCRIPTION

[0020] Specific implementation method 1: Combination Figure 2 The present embodiment is described as a seat cushion for classifying sitting postures, the seat cushion comprising a seat cushion body 1, a first force-sensitive resistor 2, a second force-sensitive resistor 3, a third force-sensitive resistor 4, a fourth force-sensitive resistor 5, a first accelerometer 6, a second accelerometer 7, a first central axis 12 and a second central axis 13; wherein:

[0021] The first force-sensitive resistor 2, the second force-sensitive resistor 3, the third force-sensitive resistor 4, the fourth force-sensitive resistor 5, the first accelerometer 6 and the second accelerometer 7 are all placed on the surface of the cushion body 1;

[0022] The first force-sensitive resistor 2 and the second force-sensitive resistor 3 are symmetrically arranged on both sides of the first central axis 12 of the seat cushion body 1;

[0023] The third force-sensitive resistor 4 and the fourth force-sensitive resistor 5 are symmetrically arranged on both sides of the first central axis 12 of the seat cushion body 1;

[0024] The first accelerometer 6 and the second accelerometer 7 are symmetrically arranged on both sides of the first central axis 12 of the seat cushion body 1, the first accelerometer 6 is placed between the first force-sensitive resistor 2 and the third force-sensitive resistor 4, the second accelerometer 7 is placed between the second force-sensitive resistor 3 and the fourth force-sensitive resistor 5, and the first accelerometer 6 and the second accelerometer 7 are placed on the second central axis 13 of the seat cushion body 1.

[0025] Specific implementation method 2: This implementation method is different from specific implementation method 1 in that the seat cushion body 1 is in a rectangular shape.

[0026] The other steps and parameters are the same as those in the first embodiment.

[0027] The first force-sensitive resistor 2 and the second force-sensitive resistor 3 correspond to the left and right ischial tuberosity areas of the user, respectively, and the third force-sensitive resistor 4 and the fourth force-sensitive resistor 5 correspond to the left and right thigh areas of the user, respectively. The length of the cushion body 1 is set to 55 cm and the width is set to 45 cm, wherein the first force-sensitive resistor 2 and the second force-sensitive resistor 3 are 11 cm apart, the third force-sensitive resistor 4 and the fourth force-sensitive resistor 5 are 47.5 cm apart, the first accelerometer 6 and the second accelerometer 7 are symmetrically arranged on both sides of the first central axis 12, and are also located on the second central axis 13, and the first accelerometer 6 and the second accelerometer 7 are 24 cm apart. The parameter values ​​in the present invention can be taken as the above values, but are not limited to the above values. For example, different parameter designs can be made for children and adults, respectively.

[0028] Specific embodiment three: This embodiment is different from specific embodiment one or two in that the seat cushion also includes a first fixing strap 8, a second fixing strap 9, a third fixing strap 10 and a fourth fixing strap 11, and the first fixing strap 8, the second fixing strap 9, the third fixing strap 10 and the fourth fixing strap 11 are used to fix the seat cushion body 1 to the seat surface.

[0029] The other steps and parameters are the same as those in the first or second embodiment.

[0030] The third fixing belt 10 and the fourth fixing belt 11 are respectively arranged on the two sides of the cushion body 1 parallel to the first central axis (corresponding to the wide sides of the rectangular cushion in the present invention). Of the two sides of the cushion body 1 parallel to the second central axis, the first fixing belt 8 and the second fixing belt 9 are arranged on the side close to the back, and the first fixing belt 8, the second fixing belt 9, the third fixing belt 10 and the fourth fixing belt 11 are used to cooperate with each other to fix the cushion body 1 on the seat surface.

[0031] Specific implementation method four: Combination Figure 1 The present embodiment describes a method for classifying sitting postures based on a seat cushion, and the method specifically comprises the following steps:

[0032] Step 1: Obtain the reference data matrix for sitting posture classification from the server;

[0033] Step 2: When the user sits on the cushion, the force-sensitive resistors and accelerometers on the cushion are used to collect the user's signal data, and then the data associated with the sitting posture classification is selected from the collected user signal data (the data associated with the coordinate classification here is the data associated with the sitting posture classification determined when calculating the reference data matrix);

[0034] Step 3: Obtain the user's sitting posture classification result based on the data selected in step 2 and the reference data matrix.

[0035] Specific implementation mode 5: This implementation mode is different from specific implementation mode 4 in that, in step 1, the method for generating the reference data matrix stored on the server side is:

[0036] Step 1: R subjects are selected, and each subject sits on the cushion in various sitting postures, that is, for the selected r-th subject, P groups of signal data are collected in each sitting posture of the r-th subject, and each group of signal data includes force-sensitive resistor signal data and accelerometer signal data, r = 1, 2, ..., R;

[0037] The signal data of the R×P group when there is no subject on the cushion is collected as the baseline signal;

[0038] Step 1 and 2: perform numerical standardization on the signal data of each subject, taking the rth subject as an example:

[0039] Performing numerical standardization processing on the data collected by the first force-sensitive resistor in each group of signal data corresponding to the rth subject;

[0040] Performing numerical standardization processing on the data collected by the second force-sensitive resistor in each group of signal data corresponding to the rth subject;

[0041] Performing numerical standardization processing on the data collected by the third force-sensitive resistor in each group of signal data corresponding to the rth subject;

[0042] Performing numerical standardization processing on the data collected by the fourth force-sensitive resistor in each group of signal data corresponding to the rth subject;

[0043] Performing numerical standardization processing on the acceleration data in the x-axis direction collected by the first accelerometer in each group of signal data corresponding to the rth subject;

[0044] Performing numerical standardization processing on the acceleration data in the y-axis direction collected by the first accelerometer in each group of signal data corresponding to the r-th subject;

[0045] Performing numerical standardization processing on the acceleration data in the z-axis direction collected by the first accelerometer in each group of signal data corresponding to the rth subject;

[0046] Performing numerical standardization processing on the acceleration data in the x-axis direction collected by the second accelerometer in each group of signal data corresponding to the rth subject;

[0047] Performing numerical standardization processing on the acceleration data in the y-axis direction collected by the second accelerometer in each group of signal data corresponding to the r-th subject;

[0048] Performing numerical standardization processing on the acceleration data in the z-axis direction collected by the second accelerometer in each group of signal data corresponding to the rth subject;

[0049] Then the numerical normalization result of the i-th group of signal data corresponding to the r-th subject is recorded as in, is the numerical normalized value of the data collected by the first force-sensitive resistor in the i-th group of signal data, is the numerical normalized value of the data collected by the second force-sensitive resistor in the i-th group of signal data, is the numerical normalized value of the data collected by the third force-sensitive resistor in the i-th group of signal data, is the numerical normalized value of the data collected by the fourth force-sensitive resistor in the i-th group of signal data, is the numerical normalized value of the acceleration data in the x-axis direction collected by the first accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the y-axis direction collected by the first accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the z-axis direction collected by the first accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the x-axis direction collected by the second accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the y-axis direction collected by the second accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the z-axis direction collected by the second accelerometer in the i-th group of signal data;

[0050] The numerical standardization results of the signal data collected by the force sensitive resistor are as follows: Figure 3 As shown in the figure, the numerical normalization result of the accelerometer signal data is as follows Figure 4 As shown; Similarly, the data of each subject were numerically standardized;

[0051] Step 13: using the numerical standardized data corresponding to any subject, determine the data whose correlation index with the sitting posture classification is greater than the threshold, and use the determined data as the data associated with the sitting posture classification in each group;

[0052] Step 14: retain the data associated with the sitting posture classification in all groups, and then fuse the force-sensitive resistor data and accelerometer data retained in each group (here, a dynamic filtering method is used for data fusion) to obtain the fused data of each group;

[0053] Each set of fused data is used as a row of the reference data matrix, and the reference data matrix is ​​composed of the fused sets of data.

[0054] The other steps and parameters are the same as those in the fourth embodiment.

[0055] Specific implementation method six: This implementation method is different from specific implementation methods four or five in that the sitting posture of the subject specifically includes sitting upright, leaning left, leaning right, leaning forward, leaning backward, shaking the left leg, shaking the right leg, crossing the left legs and crossing the right legs.

[0056] The other steps and parameters are the same as those in the fourth or fifth embodiment.

[0057] The characteristics of sitting upright are: the spine is in a neutral position, the head maintains a natural balance, and there is no obvious tilt or rotation.

[0058] The characteristics of left-leaning are: the body's center of gravity shifts to the left, the left waist muscles contract, and the right muscles stretch.

[0059] The characteristics of right leaning are: the body's center of gravity shifts to the right, the right waist muscles contract, and the left muscles stretch.

[0060] The characteristics of forward leaning are: the body's center of gravity moves forward, the back muscles stretch, and the abdominal muscles contract.

[0061] The characteristics of backward tilt are: the body's center of gravity moves backward, the abdominal muscles stretch, and the back muscles contract.

[0062] The characteristics of left leg shaking are: the left leg muscles contract and relax rapidly, producing a shaking effect, while the right leg and upper body remain stable.

[0063] The characteristics of right leg shaking are: the right leg muscles contract and relax rapidly, producing a shaking effect, while the left leg and upper body remain stable.

[0064] The characteristics of left crossed legs are: the left leg is abducted, the right leg is adducted, and the body's center of gravity is slightly shifted to the right.

[0065] The characteristics of right crossed legs are: the right leg is abducted, the left leg is adducted, and the body's center of gravity is slightly shifted to the left.

[0066] Specific implementation method 7: This implementation method is different from any one of specific implementation methods 4 to 6 in that the specific process of step 13 is as follows:

[0067] For any sitting posture, the correlation index between the data collected by the first force-sensitive resistor and the acceleration data in the x-axis direction collected by the first accelerometer for the rth subject in this sitting posture is for:

[0068]

[0069] in,

[0070] Then the correlation index matrix Q corresponding to the rth subject in this sitting posture is:

[0071]

[0072] Calculate the absolute value of each element in the matrix Q respectively, obtain all elements in the matrix Q with an absolute value greater than 0.8, and then determine the data associated with this sitting posture based on all the obtained elements. a,b If the absolute value of is greater than 0.8, then a and b are data associated with this sitting posture.

[0073] Similarly, the data associated with each sitting posture of the r-th subject are determined respectively, and all the determined data are used as the data associated with the sitting posture classification.

[0074] The other steps and parameters are the same as those in Specific Implementation Methods 4 to 6.

[0075] Specific implementation eight: This implementation differs from specific implementations four to seven in that, in step one four, the reference data matrix is ​​formed using the fused groups of data, specifically:

[0076]

[0077] Where n is the total number of data sets collected in step 1, M is the reference data matrix, G is the number of data types associated with sitting posture classification, g = 1, 2, ..., G, (x i1 ,x i2 ,…,x iG ) is the i-th group of data after fusion.

[0078] The other steps and parameters are the same as those in Specific Implementation Methods 4 to 7.

[0079] Since the data associated with the sitting posture classification determined in the embodiment of the present invention is Therefore, the value of G is 8 at this time.

[0080] Specific implementation method 9: This implementation method is different from any one of specific implementation methods 4 to 8 in that the specific process of step 3 is as follows:

[0081] Step 3.1. Data associated with sitting posture classification selected from the collected user signal data is recorded as

[0082]

[0083] Step 3.2: Calculate separately The characteristic distance of association with each row vector in the reference data matrix M;

[0084]

[0085] Among them, x i is the i-th row vector in the reference data matrix M, x i =(x i1 ,x i2 ,…,x iG ), T represents the transpose operation, S is the covariance matrix, S -1 is the inverse of the covariance matrix, yes The characteristic distance of the degree of association with the i-th row vector in the reference data matrix M;

[0086] Step 3: According to Get the sitting posture classification results:

[0087]

[0088] Among them, C is the sitting posture classification set, is the user's sitting posture classification result, I(y i =c) is the indicator function, y i is the true category of the i-th row vector in the reference data matrix M. i = c, I(y i =c) is 1, when y i ≠c, I(y i =c) has a value of 0.

[0089] The other steps and parameters are the same as those in Specific Implementation Methods 4 to 8.

[0090] The following further describes step 33 of this embodiment. Each category c in the sitting posture classification set C is substituted into the formula for calculation. For example, when category c is left-leaning, if the true category of the i-th row vector is left-leaning, then I(y i =c) is 1, otherwise I(y i =c) is 0, and combined with the previously calculated You can get the corresponding value of the current category c After traversing all the categories in set C, by comparing the corresponding value, to select the largest value, the maximum The category corresponding to the value is taken as the classification result of the user's sitting posture.

[0091] Specific implementation method ten: This implementation method is different from any one of specific implementation methods four to nine in that the covariance matrix S is:

[0092]

[0093] Among them, σ jk is an element in the covariance matrix, j = 1, 2, ..., G, k = 1, 2, ..., G;

[0094]

[0095] in,

[0096] The other steps and parameters are the same as those in Specific Implementation Methods 4 to 9.

[0097] Example

[0098] The following is a further detailed description of a sitting posture classification method based on a seat cushion of the present invention in conjunction with the accompanying drawings. The sitting posture classification method specifically comprises the following steps:

[0099] The present invention stores the reference data matrix for sitting posture classification on the server side for subsequent retrieval by different users, and the generation method of the reference data matrix is:

[0100] First, 30 subjects were selected, including 15 boys and 15 girls. Each subject included unattended (baseline data) + 9 sitting posture data. 200 sets of stable output data were selected for each sitting posture of each subject. The data acquisition frequency was 10Hz, and each set included 10 data points (6 columns of data from 2 accelerometers and 4 columns of data from 4 force-sensitive resistors). That is, the present invention collected n=2000×30 sets of data in total. The collected data was processed as follows:

[0101] Step 1: R subjects are selected, and each subject sits on the cushion in various sitting postures, that is, for the selected r-th subject, P groups of signal data are collected in each sitting posture of the r-th subject, and each group of signal data includes force-sensitive resistor signal data and accelerometer signal data, r = 1, 2, ..., R;

[0102] The signal data of the R×P group when there is no subject on the cushion is collected as the baseline signal;

[0103] Step 1 and 2: perform numerical standardization on the signal data of each subject, taking the rth subject as an example:

[0104] Performing numerical standardization processing on the data collected by the first force-sensitive resistor in each group of signal data corresponding to the rth subject;

[0105] Performing numerical standardization processing on the data collected by the second force-sensitive resistor in each group of signal data corresponding to the rth subject;

[0106] Performing numerical standardization processing on the data collected by the third force-sensitive resistor in each group of signal data corresponding to the rth subject;

[0107] Performing numerical standardization processing on the data collected by the fourth force-sensitive resistor in each group of signal data corresponding to the rth subject;

[0108] Performing numerical standardization processing on the acceleration data in the x-axis direction collected by the first accelerometer in each group of signal data corresponding to the rth subject;

[0109] Performing numerical standardization processing on the acceleration data in the y-axis direction collected by the first accelerometer in each group of signal data corresponding to the r-th subject;

[0110] Performing numerical standardization processing on the acceleration data in the z-axis direction collected by the first accelerometer in each group of signal data corresponding to the rth subject;

[0111] Performing numerical standardization processing on the acceleration data in the x-axis direction collected by the second accelerometer in each group of signal data corresponding to the rth subject;

[0112] Performing numerical standardization processing on the acceleration data in the y-axis direction collected by the second accelerometer in each group of signal data corresponding to the r-th subject;

[0113] Performing numerical standardization processing on the acceleration data in the z-axis direction collected by the second accelerometer in each group of signal data corresponding to the rth subject;

[0114] Then the numerical normalization result of the i-th group of signal data corresponding to the r-th subject is recorded as in, is the numerical normalized value of the data collected by the first force-sensitive resistor in the i-th group of signal data, is the numerical normalized value of the data collected by the second force-sensitive resistor in the i-th group of signal data, is the numerical normalized value of the data collected by the third force-sensitive resistor in the i-th group of signal data, is the numerical normalized value of the data collected by the fourth force-sensitive resistor in the i-th group of signal data, is the numerical normalized value of the acceleration data in the x-axis direction collected by the first accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the y-axis direction collected by the first accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the z-axis direction collected by the first accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the x-axis direction collected by the second accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the y-axis direction collected by the second accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the z-axis direction collected by the second accelerometer in the i-th group of signal data;

[0115] The numerical standardization results of the signal data collected by the force sensitive resistor are as follows: Figure 3 As shown in the figure, the numerical normalization result of the accelerometer signal data is as follows Figure 4 As shown; Similarly, the data of each subject were numerically standardized;

[0116] Step 13: Using the numerical standardized data corresponding to any subject, determine the data whose correlation index with the sitting posture classification is greater than the threshold, and use the determined data as the data associated with the sitting posture classification in each group; specifically:

[0117] For any sitting posture, the correlation index for the sitting posture is calculated using the data collected by the first force-sensitive resistor and the acceleration data in the x-axis direction collected by the first accelerometer for the rth subject in the sitting posture. for:

[0118]

[0119] in,

[0120] Then the correlation index matrix Q corresponding to the rth subject in this sitting posture is:

[0121]

[0122] Calculate the absolute value of each element in the matrix Q respectively, obtain all elements in the matrix Q with an absolute value greater than 0.8, and then determine the data associated with this sitting posture based on all the obtained elements. a,b If the absolute value of is greater than 0.8, then a and b are data associated with this sitting posture.

[0123] Similarly, the data associated with each sitting posture of the r-th subject are determined respectively, and all the determined data are used as the data associated with the sitting posture classification.

[0124] It should be noted that as long as the absolute value of any element corresponding to data a in the correlation index matrix of any sitting posture is greater than 0.8, data a is the data associated with the sitting posture classification. The data associated with the sitting posture classification of the rth subject is taken as the data associated with the sitting posture classification of each subject, that is, for each subject, the data type associated with the sitting posture classification is the same.

[0125] The maximum correlation index of the subjects in various sitting positions is as follows Figure 5 As shown, in the present invention, the data associated with the sitting posture classification is determined by calculation as follows:

[0126] Step 14: retain the data associated with the sitting posture classification in all groups, and then fuse the force-sensitive resistor data and accelerometer data retained in each group (here, a dynamic filtering method is used for data fusion) to obtain the fused data of each group;

[0127] Each group of fused data is used as a row of the reference data matrix, and the reference data matrix is ​​composed of each group of fused data:

[0128]

[0129] Where n is the total number of data sets collected in step 1, M is the reference data matrix, g = 1, 2, ..., 8, (x i1 ,x i2 ,…,x i8 ) is the i-th group of data after fusion.

[0130] When the user sits on the cushion, the force-sensitive resistors and accelerometers on the cushion are used to collect the user's signal data, and then the data associated with the sitting posture classification is selected from the collected user signal data (the data associated with the coordinate classification here is the data associated with the sitting posture classification determined when calculating the reference data matrix);

[0131] And according to the data selected in step 2 and the reference data matrix, the user's sitting posture classification result is obtained; the specific process is:

[0132] The data associated with the sitting posture classification selected from the collected user signal data is recorded as

[0133]

[0134] Calculate separately The characteristic distance of association with each row vector in the reference data matrix M;

[0135]

[0136] Among them, x i is the i-th row vector in the reference data matrix M, x i =(x i1 ,x i2 ,…,x i8 ), T represents the transpose operation, S is the covariance matrix, S -1 is the inverse of the covariance matrix, yes The characteristic distance of the degree of association with the i-th row vector in the reference data matrix M;

[0137] according to Get the sitting posture classification results:

[0138]

[0139] Among them, C is the sitting posture classification set, is the user's sitting posture classification result, I(y i =c) is the indicator function, y i is the true category of the i-th row vector in the reference data matrix M. i = c, I(y i =c) is 1, when y i ≠c, I(y i=c) has a value of 0.

[0140] The present invention collects n=60000 sets of data from the subjects. After the data set number n′ is initialized to 1, all the data sets except the n′th set of data are used as reference data. The sitting posture represented by the n′th set of data is predicted based on the reference data. Then the data set number n′=n′+1 is set, and the iteration is stopped when the data set number n′=60000. That is, the sitting posture prediction results for each set of data of the subjects are obtained respectively. The classification results of the subject data are as follows: Figure 6 As shown, labels 0 to 9 represent no one, sitting upright, leaning left, leaning right, leaning forward, leaning backward, shaking left leg, shaking right leg, crossing left legs and crossing right legs respectively. TPR is the true positive rate, which quantifies the model's ability to correctly identify positive cases among all actual positive cases. FNR is the false negative rate, which reflects the proportion of actual positive cases that the model fails to correctly classify as positive.

[0141] The above calculation examples of the present invention are only used to explain the calculation model and calculation process of the present invention in detail, and are not intended to limit the implementation methods of the present invention. For ordinary technicians in the relevant field, other different forms of changes or modifications can be made based on the above description. It is impossible to list all the implementation methods here. All obvious changes or modifications derived from the technical solution of the present invention are still within the scope of protection of the present invention.

Claims

1. A sitting posture classification method based on a cushion, characterized in that: The method specifically comprises the following steps: Step 1: Obtain the reference data matrix for sitting posture classification from the server; Step 2: When the user sits on the cushion, the force-sensitive resistors and accelerometers on the cushion are used to collect the user's signal data, and then the data associated with the sitting posture classification is selected from the collected user signal data; Step 3: Obtain the user's sitting posture classification result based on the data selected in step 2 and the reference data matrix.

2. The method for classifying sitting postures based on a cushion according to claim 1, characterized in that: In step 1, the method for generating the reference data matrix stored on the server is: Step 1: R subjects are selected, and each subject sits on the cushion in various sitting postures, that is, for the selected r-th subject, P groups of signal data are collected in each sitting posture of the r-th subject, and each group of signal data includes force-sensitive resistor signal data and accelerometer signal data, r = 1, 2, ..., R; The signal data of the R×P group when there is no subject on the cushion is collected as the baseline signal; Step 1 and 2: perform numerical standardization on the signal data of each subject, taking the rth subject as an example: Performing numerical standardization processing on the data collected by the first force-sensitive resistor in each group of signal data corresponding to the rth subject; Performing numerical standardization processing on the data collected by the second force-sensitive resistor in each group of signal data corresponding to the rth subject; Performing numerical standardization processing on the data collected by the third force-sensitive resistor in each group of signal data corresponding to the rth subject; Performing numerical standardization processing on the data collected by the fourth force-sensitive resistor in each group of signal data corresponding to the rth subject; Performing numerical standardization processing on the acceleration data in the x-axis direction collected by the first accelerometer in each group of signal data corresponding to the rth subject; Performing numerical standardization processing on the acceleration data in the y-axis direction collected by the first accelerometer in each group of signal data corresponding to the r-th subject; Performing numerical standardization processing on the acceleration data in the z-axis direction collected by the first accelerometer in each group of signal data corresponding to the rth subject; Performing numerical standardization processing on the acceleration data in the x-axis direction collected by the second accelerometer in each group of signal data corresponding to the rth subject; Performing numerical standardization processing on the acceleration data in the y-axis direction collected by the second accelerometer in each group of signal data corresponding to the r-th subject; Performing numerical standardization processing on the acceleration data in the z-axis direction collected by the second accelerometer in each group of signal data corresponding to the rth subject; Then the numerical normalization result of the i-th group of signal data corresponding to the r-th subject is recorded as in, is the numerical normalized value of the data collected by the first force-sensitive resistor in the i-th group of signal data, is the numerical normalized value of the data collected by the second force-sensitive resistor in the i-th group of signal data, is the numerical normalized value of the data collected by the third force-sensitive resistor in the i-th group of signal data, is the numerical normalized value of the data collected by the fourth force-sensitive resistor in the i-th group of signal data, is the numerical normalized value of the acceleration data in the x-axis direction collected by the first accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the y-axis direction collected by the first accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the z-axis direction collected by the first accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the x-axis direction collected by the second accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the y-axis direction collected by the second accelerometer in the i-th group of signal data, is the numerical normalized value of the acceleration data in the z-axis direction collected by the second accelerometer in the i-th group of signal data; Similarly, the data of each subject were numerically normalized; Step 13: using the numerical standardized data corresponding to any subject, determine the data whose correlation index with the sitting posture classification is greater than the threshold, and use the determined data as the data associated with the sitting posture classification in each group; Step 14: retain the data associated with the sitting posture classification in all groups, and then fuse the force-sensitive resistor data and accelerometer data retained in each group to obtain each group of fused data; Each set of fused data is used as a row of the reference data matrix, and the reference data matrix is ​​composed of the fused sets of data.

3. The method for classifying sitting postures based on a cushion according to claim 2, characterized in that: The sitting postures of the subject specifically include sitting upright, leaning left, leaning right, leaning forward, leaning backward, shaking left legs, shaking right legs, crossing left legs and crossing right legs.

4. The method for classifying sitting postures based on a cushion according to claim 3, characterized in that: The specific process of step 1-3 is as follows: For any sitting posture, the correlation index between the data collected by the first force-sensitive resistor and the acceleration data in the x-axis direction collected by the first accelerometer for the rth subject in this sitting posture is for: in, Then the correlation index matrix Q corresponding to the rth subject in this sitting posture is: Calculate the absolute value of each element in the matrix Q respectively, obtain all elements in the matrix Q with an absolute value greater than 0.8, and then determine the data associated with this sitting posture based on all the obtained elements. a,b If the absolute value of is greater than 0.8, then a and b are data associated with this sitting posture. Similarly, the data associated with each sitting posture of the r-th subject are determined respectively, and all the determined data are used as the data associated with the sitting posture classification.

5. The method for classifying sitting postures based on a cushion according to claim 4, characterized in that: In the step 1-4, the reference data matrix is ​​formed by using the fused data sets, specifically: Where n is the total number of data sets collected in step 1, M is the reference data matrix, G is the number of data types associated with sitting posture classification, g = 1, 2, ..., G, (x i1 ,x i2 ,…,x iG ) is the i-th group of data after fusion.

6. The method for classifying sitting postures based on a seat cushion according to claim 5, characterized in that: The specific process of step three is: Step 3.

1. Data associated with sitting posture classification selected from the collected user signal data is recorded as Step 3.2: Calculate separately The characteristic distance of association with each row vector in the reference data matrix M; Among them, x i is the i-th row vector in the reference data matrix M, x i =(x i1 ,x i2 ,…,x iG ), T represents the transpose operation, S is the covariance matrix, S -1 is the inverse of the covariance matrix, yes The characteristic distance of the degree of association with the i-th row vector in the reference data matrix M; Step 3: According to Get the sitting posture classification results: Among them, C is the sitting posture classification set, is the user's sitting posture classification result, I(y i =c) is the indicator function, y i is the true category of the i-th row vector in the reference data matrix M. i = c, I(y i =c) is 1, when y i ≠c, I(y i =c) has a value of 0.

7. The method for classifying sitting postures based on a cushion according to claim 6, characterized in that: The covariance matrix S is: Among them, σ jk is an element in the covariance matrix, j = 1, 2, ..., G, k = 1, 2, ..., G; in,

Images