Space-time gait feature extraction method based on plantar pressure center trajectory, electronic equipment and readable storage medium

Through the spatial and temporal gait feature extraction method based on the center trajectory of the sole foot pressure, the problem of individual differences in the diagnosis of early Parkinson's disease is solved, and the accurate identification of early lesions is achieved, and the detection accuracy is improved.

CN118873129APending Publication Date: 2024-11-01CHONGQING JIAOTONG UNIV +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410953783.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-16
Publication Date
2024-11-01

AI Technical Summary

Technical Problem

In the diagnosis of early Parkinson's disease, the gait feature extraction method is greatly affected by individual differences, and the calculation is complex and it is difficult to accurately identify early lesions. The traditional linear method is affected by factors such as height and age, and the nonlinear method is complex and the early steps are The difference in state data is not obvious.

Method used

The space-time gait feature extraction method based on the center trajectory of the sole foot pressure, and the Y-axis coordinates of the center trajectory of the sole foot are calculated to determine the significance difference. Spatial-temporal gait features a and d are calculated to reduce individual differences and improve diagnostic accuracy.

Benefits of technology

Through the analysis of the center trajectory of the sole foot pressure center, the impact of individual differences was reduced, and the spatiotemporal gait characteristics a and d with stronger correlation with Parkinson's patients were calculated, which improved the detection accuracy of early Parkinson's disease and provided more accurate data support.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118873129A_ABST
    Figure CN118873129A_ABST
Patent Text Reader

Abstract

The invention provides a space-time gait feature extraction method based on a plantar pressure center track. The space-time gait feature extraction method comprises the following steps: S1, collecting opposite acting forces of a left foot and a right foot of a subject vertical to the ground; s2, respectively calculating Y-axis coordinates of plantar pressure center tracks of the left foot and the right foot according to the opposite acting force vertical to the ground; s3, judging whether the significant difference of the Y-axis coordinates of the plantar pressure center track of the left foot is greater than the significant difference of the Y-axis coordinates of the plantar pressure center track of the right foot, and if so, calculating space-time gait characteristics a and d by adopting the Y-axis coordinates of the plantar pressure center track of the left foot; if not, the Y-axis coordinate of the plantar pressure center track of the right foot is adopted to calculate the space-time gait characteristics a and d. Through the method, the change of the foot center can be reflected, the influence of individual differences is avoided, and the spatio-temporal gait feature a and the spatio-temporal gait feature d which have larger significant differences between early PD patients and healthy people can be obtained.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of feature extraction, and in particular to a spatiotemporal gait feature extraction method based on a plantar pressure center trajectory, an electronic device and a readable storage medium. Background Art

[0002] Time domain gait feature analysis is an effective method to identify the presence of Parkinson's Disease (PD). Time domain gait feature analysis includes the use of linear and nonlinear methods to extract gait features; the research method of using linear methods to extract traditional gait features such as swing time, standing time, and step length for PD diagnosis is subject to the influence of group characteristics such as individual height, age, and weight. The process of extracting multi-level gait features using nonlinear methods is computationally complex. In addition, since the symptoms of early PD are relatively mild, the gait data of early PD patients are not significantly different from those of healthy people. In addition, the interference of individual difference factors poses a huge challenge to the diagnosis decision of early PD.

[0003] Therefore, in order to solve the above technical problems, it is urgent to propose a new technical means. Summary of the invention

[0004] In view of this, in order to more accurately detect early PD patients, the present invention proposes a spatiotemporal gait feature extraction method based on the plantar pressure center trajectory, an electronic device and a readable storage medium.

[0005] The present invention provides a spatiotemporal gait feature extraction method based on plantar pressure center trajectory, comprising the following steps:

[0006] S1. Collect the vertical ground reaction force of the subject's left and right feet;

[0007] S2. Calculate the Y-axis coordinates of the plantar pressure center trajectories of the left and right feet, respectively, according to the vertical ground reaction force;

[0008] S3. Determine whether the significant difference of the Y-axis coordinate of the plantar pressure center trajectory of the left foot is greater than the significant difference of the Y-axis coordinate of the plantar pressure center trajectory of the right foot. If so, use the Y-axis coordinate of the plantar pressure center trajectory of the left foot to calculate the spatiotemporal gait features a and d; if not, use the Y-axis coordinate of the plantar pressure center trajectory of the right foot to calculate the spatiotemporal gait features a and d;

[0009] Wherein, a represents the ratio coefficient of the value of the Y-axis coordinate of 0 in a plantar pressure center trajectory to the total number of samples, and d represents the difference between the maximum value and the minimum value of the Y-axis coordinate in a plantar pressure center trajectory.

[0010] Furthermore, the spatiotemporal gait feature a is calculated by the following formula:

[0011]

[0012] Where B = {b 1 ,b 2 ,b 3 ,…,b q ,0,0,0,…,0} represents the trajectory of the plantar pressure center in a gait cycle, b q Represents the Y-axis coordinate of the qth plantar pressure center trajectory of the left or right foot, and num(0) represents the number of 0 values.

[0013] Furthermore, the spatiotemporal gait feature d is calculated by the following formula:

[0014] d=|Max(B)-Min(B)|

[0015] Where B = {b 1 ,b 2 ,b 3 ,…,b q ,0,0,0,…,0} represents the trajectory of the plantar pressure center in a gait cycle, b q Represents the Y-axis coordinate of the qth plantar pressure center trajectory of the left or right foot.

[0016] Further, the Y-axis coordinate of the plantar pressure center trajectory of the left foot is determined according to the following formula:

[0017]

[0018] The symmetry center of the left foot is taken as the origin, the symmetry axis of the left foot is taken as the Y axis, and the straight line passing through the origin and perpendicular to the Y axis is taken as the X axis; b q-Left is the Y-axis coordinate of the qth plantar pressure center trajectory of the left foot, i is the number of the left foot pressure sensor, m is the number of left foot pressure sensors, and y i is the ordinate value of the i-th sensor position; F i Indicates the vertical ground reaction force value corresponding to the i-th sensor of the left foot.

[0019] Further, the Y-axis coordinate of the plantar pressure center trajectory of the right foot is determined according to the following formula:

[0020]

[0021] The symmetry center of the right foot is taken as the origin, the symmetry axis of the right foot is taken as the Y axis, and the straight line passing through the origin and perpendicular to the Y axis is taken as the X axis; b q-Rightis the Y-axis coordinate of the qth plantar pressure center trajectory of the right foot, j is the number of the right foot pressure sensor, n is the number of the right foot pressure sensors, and y j is the ordinate value of the jth sensor position, F j Indicates the vertical ground reaction force value corresponding to the jth sensor of the right foot.

[0022] Accordingly, the present invention further provides an electronic device, comprising:

[0023] A memory and a processor, wherein the memory is used to store a computer program, and when the computer program is executed by the processor, the spatiotemporal gait feature extraction method based on the plantar pressure center trajectory is implemented.

[0024] Correspondingly, the present invention also provides a readable storage medium, in which computer instructions are stored. When the computer instructions are executed by a processor, the above-mentioned spatiotemporal gait feature extraction method based on the plantar pressure center trajectory is implemented.

[0025] The beneficial effects of the present invention are as follows: the present invention performs gait analysis through the trajectory of the plantar pressure center, which can reflect the changes in the center of the foot and is not affected by individual differences; and the present invention provides data support for more accurate detection of early PD patients by calculating the spatiotemporal gait features a and d that have more significant differences between early PD patients and healthy people. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0027] Figure 1 It is a flow chart of the present invention;

[0028] Figure 2 Schematic diagram showing COP trajectories of healthy subjects and early PD patients. DETAILED DESCRIPTION

[0029] The present invention is further described below in conjunction with the accompanying drawings:

[0030] The present invention provides a spatiotemporal gait feature extraction method based on plantar pressure center trajectory, comprising the following steps:

[0031] S1. Collect the vertical ground reaction force of the subject's left and right feet;

[0032] S2. Calculate the Y-axis coordinates of the plantar pressure center trajectories of the left and right feet, respectively, according to the vertical ground reaction force;

[0033] S3. Determine whether the significant difference of the Y-axis coordinate of the plantar pressure center trajectory of the left foot is greater than the significant difference of the Y-axis coordinate of the plantar pressure center trajectory of the right foot. If so, use the Y-axis coordinate of the plantar pressure center trajectory of the left foot to calculate the spatiotemporal gait features a and d; if not, use the Y-axis coordinate of the plantar pressure center trajectory of the right foot to calculate the spatiotemporal gait features a and d;

[0034] Among them, a represents the ratio coefficient of the value of the Y-axis coordinate of 0 in a plantar pressure center trajectory to the total number of samples, and d represents the difference between the maximum value and the minimum value of the Y-axis coordinate in a plantar pressure center trajectory. Through the above method, the influence of individual differences on the spatiotemporal gait characteristics can be reduced, and the spatiotemporal gait characteristics a and d that are more closely related to Parkinson's patients can be calculated, thereby providing data support for more accurate detection of early PD patients.

[0035] In this embodiment, in step S1, the vertical ground reverse force (Vertical Ground Reverse Force, VGRF) of the left and right feet of the subject is collected. Preferably, 8 pressure sensors are installed on the left and right feet of each subject to collect VGRF data. FSR (Force Sensing Resistor) sensors are preferably used. The collected VGRF data are preprocessed to remove abnormal values ​​in the VGRF data to reduce the impact on subsequent calculations, and then the Y-axis coordinates of the center of pressure (Center of Pressure, COP) of the left and right feet are calculated using the filtered VGRF data.

[0036] In this embodiment, in step S2, the Y-axis coordinates of the plantar pressure center trajectories of the left foot and the right foot are calculated respectively according to the vertical ground reaction force;

[0037] Determine the Y-axis coordinate of the plantar pressure center trajectory of the left foot according to the following formula:

[0038]

[0039] The symmetry center of the left foot is taken as the origin, the symmetry axis of the left foot is taken as the Y axis, and the straight line passing through the origin and perpendicular to the Y axis is taken as the X axis; b q-Left is the Y-axis coordinate of the qth plantar pressure center trajectory of the left foot, i is the number of the left foot pressure sensor, m is the number of the left foot pressure sensors, m is 8, y i is the ordinate value of the i-th sensor position; F i Indicates the vertical ground reaction force value corresponding to the i-th sensor of the left foot.

[0040] Determine the Y-axis coordinate of the plantar pressure center trajectory of the right foot according to the following formula:

[0041]

[0042] The symmetry center of the right foot is taken as the origin, the symmetry axis of the right foot is taken as the Y axis, and the straight line passing through the origin and perpendicular to the Y axis is taken as the X axis; b q-Right is the Y-axis coordinate of the qth plantar pressure center trajectory of the right foot, j is the number of the right foot pressure sensor, n is the number of the right foot pressure sensors, n is 8, y j is the ordinate value of the jth sensor position, F j Indicates the vertical ground reaction force value corresponding to the jth sensor of the right foot.

[0043] In this embodiment, in step S3, it is determined whether the significant difference of the Y-axis coordinates of the plantar pressure center trajectory of the left foot is greater than the significant difference of the Y-axis coordinates of the plantar pressure center trajectory of the right foot. If so, the Y-axis coordinates of the plantar pressure center trajectory of the left foot are used to calculate the spatiotemporal gait features a and d; if not, the Y-axis coordinates of the plantar pressure center trajectory of the right foot are used to calculate the spatiotemporal gait features a and d; in the case where the significant difference of the Y-axis plantar pressure center trajectory information of the left foot and the right foot cannot be determined, it is preferred to use the Y-axis coordinates of the plantar pressure center trajectory of the left foot to calculate the spatiotemporal gait features a and d;

[0044] Where a represents the ratio coefficient of the value of the Y-axis coordinate of 0 in a plantar pressure center trajectory to the total number of samples, and d represents the difference between the maximum value and the minimum value of the Y-axis coordinate in a plantar pressure center trajectory;

[0045] Methods for determining the size of significant differences include calculating P values, T tests, chi-square tests, F tests, etc. Preferably, the Y-axis plantar pressure center trajectory information of the foot and the mean and standard deviation of the Y-axis coordinates of the plantar pressure center trajectory of the right foot are calculated by the T test method, and then the P value of the mean and standard deviation is further calculated. When the P value of the mean and standard deviation of the Y-axis coordinates of the plantar pressure center trajectory of the left foot is less than the P value of the mean and standard deviation of the Y-axis coordinates of the plantar pressure center trajectory of the right foot, it indicates whether the significant difference of the Y-axis coordinates of the plantar pressure center trajectory of the left foot is greater than the significant difference of the Y-axis coordinates of the plantar pressure center trajectory of the right foot. The specific processes of calculating the P value, T test, chi-square test, and F test are existing technologies and will not be repeated here.

[0046] The spatiotemporal gait feature a is calculated by the following formula:

[0047]

[0048] Where B = {b 1 ,b 2 ,b 3 ,…,b q,0,0,0,…,0} represents the trajectory of the plantar pressure center in a gait cycle, b q It represents the Y-axis coordinate of the qth plantar pressure center trajectory of the left or right foot, and num(0) represents the number of 0 values ​​in the plantar pressure center trajectory of a gait cycle.

[0049] The spatiotemporal gait feature d is calculated by the following formula:

[0050] d=|Max(B)-Min(B)|

[0051] Where B = {b 1 ,b 2 ,b 3 ,…,b q ,0,0,0,…,0} represents the trajectory of the plantar pressure center in a gait cycle, b q It represents the Y-axis coordinate of the qth plantar pressure center trajectory of the left or right foot, where q represents the number of swings of the left or right foot in one gait cycle.

[0052] By acquiring the above data, it is possible to detect subtle differences in the posture of individual limbs, thereby more accurately detecting early-stage Parkinson's patients.

[0053] Accordingly, the present invention further provides an electronic device, comprising:

[0054] A memory and a processor, wherein the memory is used to store a computer program, and when the computer program is executed by the processor, the spatiotemporal gait feature extraction method based on the plantar pressure center trajectory is implemented.

[0055] Correspondingly, the present invention also provides a readable storage medium, in which computer instructions are stored. When the computer instructions are executed by a processor, the above-mentioned spatiotemporal gait feature extraction method based on the plantar pressure center trajectory is implemented.

[0056] Data description of the present invention being more accurate than the existing COP trajectory: The PD (Parkinson's Disease) gait data set in the Physionet database is used for testing. The data set contains 55 early PD gait data and 73 healthy gait data. In the data set, 8 FSR sensors are respectively set on the left foot and the right foot to collect VGRF data. The position coordinates of the 8 FSR sensors on the left foot and the right foot are shown in Table 1 and Table 2 respectively:

[0057] Table 1

[0058]

[0059] Table 2

[0060]

[0061] Based on the 8 FSR sensors on the left and right feet and the collected VGRF data, the COP trajectory (Center of Pressure) of the subject can be calculated;

[0062] Determine the COP trajectory information of the left foot according to the following formula:

[0063]

[0064] Among them, the COP trajectory information of the left foot includes COP X-Left and COP Y-Left , COP X-Left is the X-axis coordinate of the left foot COP trajectory, COP Y-Left is the Y-axis coordinate of the left foot COP trajectory, i is the number of the left foot pressure sensor: 1-8, m is the number of left foot pressure sensors, m is 8, x i Indicates the horizontal coordinate value of the i-th sensor position, y i is the ordinate value of the i-th sensor position; F i represents the VGRF signal value corresponding to the i-th sensor of the left foot;

[0065] Determine the COP trajectory information of the right foot according to the following formula:

[0066]

[0067] Among them, the COP trajectory information of the right foot includes COP X-Right and COP Y-Right , COP X-Right is the X-axis coordinate of the COP trajectory of the right foot, COP Y-Right is the Y-axis coordinate of the COP trajectory of the right foot; j is the number of the right foot pressure sensor: 1-8, n is the number of the right foot pressure sensors, n is 8, x j Indicates the horizontal coordinate value of the jth sensor position, y j is the ordinate value of the jth sensor position, F j represents the VGRF signal value corresponding to the jth sensor of the right foot;

[0068] Through the above formula, the COP trajectory of the left foot and the COP trajectory of the right foot of all subjects in the data set are calculated, as shown in Figure 2 As shown in the figure, (a) represents the COP trajectory of 73 healthy subjects, (b) represents the COP trajectory of 55 early PD patients, Leftfoot represents the left foot, Right foot represents the right foot; X-axis coordinate value represents the X-axis coordinate value, i.e., COPX-Left or COP X-Right ; Y-axis coordinate value represents the Y-axis coordinate value, i.e. COP Y-Left or COP Y-Right ;from Figure 2 It can be concluded that the COP trajectories of healthy subjects are highly overlapped, while the COP trajectories of early PD patients are divergent; and the front-to-back offset of the Y-axis coordinate signal of the left foot COP trajectory of healthy subjects is evenly distributed, while the front-to-back offset of the Y-axis coordinate signal of the left foot COP trajectory of early PD patients is uneven.

[0069] Furthermore, a T-test was performed on the X-axis and Y-axis coordinate values ​​of the COP of the left and right feet of 55 early PD patients and 73 healthy subjects. The results are shown in Table 3:

[0070] Table 3

[0071] COP trajectory CO(73) Early PD (55) P-value <![CDATA[COP X-Left ]]> -497.16±123.45 -477.72±152.96 0.4297 <![CDATA[COP X-Right ]]> 498.01±121.46 484.55±147.24 0.5737 <![CDATA[COP Y-Left ]]> -25.49±59.81 18.99±69.98 <0.001 <![CDATA[COP Y-Right ]]> -19.88±56.73 17.40±88.47 0.0078

[0072] From Table 3, we can see that COP Y-Left The results were extremely significant (P<0.001); and Figure 2 (C) It can be seen that the COP of early PD patients Y-Left The trajectory is shorter than that of healthy people; because the Physionet database used does not state in advance whether the subject is left-side weight bearing or has left-side disease, that is to say, the COP Y-Left It is general and adopts COP Y-Left Further judgment of early PD is more accurate.

[0073] Subsequently, the COP Y-Left Perform spatiotemporal gait feature extraction. In a gait cycle, the unilateral COP trajectory contains non-zero values ​​and zero values, which respectively reflect the COP trajectory during the current single-leg support period and the unstressed condition during the single-leg swing period; therefore, the spatiotemporal gait features a and d are obtained in the following way:

[0074]

[0075] d=|Max(B)-Min(B)|

[0076] B = {b 1 ,b 2 ,b 3 ,…,bq,0,0,0,…,0}, where B represents the Y-axis plantar pressure center trajectory coordinates of the left foot in a gait cycle;

[0077] The spatiotemporal gait feature a was calculated to verify whether the subject reached a swing phase of 40% of the gait cycle. The spatiotemporal gait feature d was calculated to verify whether there was a difference in trajectory length between healthy people and early PD. The spatiotemporal gait features a and d were further subjected to a T test, as shown in Table 4:

[0078] Table 4

[0079] Spatiotemporal gait characteristics CO(73) Early PD (55) P-value d 1485.628±44.458 1434.99±116.449 P<0.001 a 0.374±0.026 0.351±0.028 P<0.001

[0080] It can be concluded from Table 4 that the P values ​​of the spatiotemporal gait features d and a are both less than 0.001, which is extremely significant, indicating that these two features can well distinguish early PD patients from healthy subjects, confirming that the spatiotemporal gait features d and a can detect subtle single-limb running / posture differences, which is beneficial for the detection of early PD patients.

[0081] According to the above data, COP is generally preferred. Y-Left Calculate the spatiotemporal gait features a and d; when COP Y-Right COP Y-Left When the significant difference is more obvious, COP is used Y-Right Calculate spatiotemporal gait features a and d.

[0082] To further illustrate that the spatiotemporal gait features a and d are more accurate, the Gaussian kernel support vector machine, K nearest neighbor and linear discriminant models were built using Python language to classify the spatiotemporal gait features a and d proposed in this paper into early PD and health. In addition, for comparison, the four traditional gait features (right foot standing time V1, left foot standing time V2, right foot standing time coefficient of variation V3 and left foot standing time coefficient of variation V4) characterized by significant differences (P<0.05) and their fusion features [V1, V2, V3, V4] were used as controls. Acc (accuracy), Recall (recall rate) and Spe (Specificity) were used as evaluation indicators;

[0083] The coefficient of variation of stance time was calculated using the following formula:

[0084]

[0085] Among them, M mean (A) represents the average standing time, S std (A) represents the standard deviation of standing time; when calculating the coefficient of variation of left foot standing time, M mean (A) and S std (A) Substitute the mean and standard deviation of the left foot stance time; do the same for the right foot.

[0086] Gaussian kernel support vector machine, K nearest neighbor and linear discriminant model are all existing technologies and will not be described in detail here; the process of calculating Acc, Recall and Spe indicators is also existing technology and will not be described in detail here.

[0087] According to the Physionet dataset, four traditional gait features, fusion features, spatiotemporal gait features a, spatiotemporal gait features d and spatiotemporal gait fusion features [a, d] are obtained (the spatiotemporal gait features are based on COP Y-Left get);

[0088] The above data were input into Gaussian kernel support vector machine, K nearest neighbor and linear discriminant model for classification, and the Acc, Recall and Spe indicators were calculated according to the classification results; the indicators in Gaussian kernel support vector machine, K nearest neighbor and linear discriminant model are shown in Table 5, Table 6 and Table 7 respectively:

[0089] Table 5

[0090] Characteristic parameters Acc(%) Recall(%) Spe(%) V1 52.46 55.33 64.44 V2 56.31 54.22 57.47 V3 55.32 55.41 58.74 V4 71.59 69.55 72.23 [V1,V2,V3,V4] 84.67 86.65 88.34 a 77.03 75.99 79.32 d 97.93 98.45 97.45 [a,d] 99.41 99.22 97.62

[0091] Table 6

[0092] Characteristic parameters Acc(%) Recall(%) Spe(%) V1 52.31 54.14 56.54 V2 56.23 52.82 56.71 V3 54.43 56.62 57.94 V4 69.52 70.91 73.55 [V1,V2,V3,V4] 83.35 86.61 87.76 a 75.11 73.89 78.93 d 96.53 96.36 96.52 [a,d] 98.21 98.76 97.14

[0093] Table 7

[0094]

[0095]

[0096] It can be concluded from Table 5 that the performance indicators of features a and d based on plantar pressure trajectory and their fusion feature [a, d] are higher than those of traditional gait feature parameters. It is worth mentioning that the accuracy of the [a, d] fusion feature reaches 99.41%, which is 1.48% higher than that of the single d feature input.

[0097] It can be concluded from Table 6 that although the classification effect of each feature parameter in the K-nearest neighbor algorithm is not as good as that of the Gaussian kernel support vector machine algorithm, the features d and a extracted in this paper are significantly better than the traditional gait feature parameters in terms of performance indicators. This result emphasizes that even in the relatively simple K-nearest neighbor algorithm framework, features a and d can effectively highlight their classification potential.

[0098] It can be concluded from Table 7 that the features d and a extracted in this paper show their excellent performance, which once again verifies the effectiveness of the proposed method and proves the important value and application potential of features d and a in the early diagnosis of Parkinson's disease.

[0099] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

Claims

1. A spatiotemporal gait feature extraction method based on plantar pressure center trajectory, characterized in that: The following steps are involved: S1. Collect the vertical ground reaction force of the subject's left and right feet; S2. Calculate the Y-axis coordinates of the plantar pressure center trajectories of the left and right feet, respectively, according to the vertical ground reaction force; S3. Determine whether the significant difference of the Y-axis coordinate of the plantar pressure center trajectory of the left foot is greater than the significant difference of the Y-axis coordinate of the plantar pressure center trajectory of the right foot. If so, use the Y-axis coordinate of the plantar pressure center trajectory of the left foot to calculate the spatiotemporal gait features a and d; if not, use the Y-axis coordinate of the plantar pressure center trajectory of the right foot to calculate the spatiotemporal gait features a and d; Wherein, a represents the ratio coefficient of the value of the Y-axis coordinate of 0 in a plantar pressure center trajectory to the total number of samples, and d represents the difference between the maximum value and the minimum value of the Y-axis coordinate in a plantar pressure center trajectory.

2. The spatiotemporal gait feature extraction method based on plantar pressure center trajectory according to claim 1, characterized in that: The spatiotemporal gait feature a is calculated by the following formula: Where B = {b1, b2, b3, …, b q ,0,0,0,…,0} represents the trajectory of the plantar pressure center in a gait cycle, b q Represents the Y-axis coordinate of the qth plantar pressure center trajectory of the left or right foot, and num(0) represents the number of 0 values.

3. The spatiotemporal gait feature extraction method based on plantar pressure center trajectory according to claim 1, characterized in that: The spatiotemporal gait feature d is calculated by the following formula: d=|Max(B)-Min(B)| Where B = {b1, b2, b3, …, b q ,0,0,0,…,0} represents the trajectory of the plantar pressure center in a gait cycle, b q Represents the Y-axis coordinate of the qth plantar pressure center trajectory of the left or right foot.

4. The spatiotemporal gait feature extraction method based on plantar pressure center trajectory according to claim 2 or claim 3, characterized in that: Determine the Y-axis coordinate of the plantar pressure center trajectory of the left foot according to the following formula: The symmetry center of the left foot is taken as the origin, the symmetry axis of the left foot is taken as the Y axis, and the straight line passing through the origin and perpendicular to the Y axis is taken as the X axis; b q-Left is the Y-axis coordinate of the qth plantar pressure center trajectory of the left foot, i is the number of the left foot pressure sensor, m is the number of left foot pressure sensors, and y i is the ordinate value of the i-th sensor position; F i Indicates the vertical ground reaction force value corresponding to the i-th sensor of the left foot.

5. The method for extracting spatiotemporal gait features based on the plantar pressure center trajectory according to claim 2 or claim 3, characterized in that: Determine the Y-axis coordinate of the plantar pressure center trajectory of the right foot according to the following formula: The symmetry center of the right foot is taken as the origin, the symmetry axis of the right foot is taken as the Y axis, and the straight line passing through the origin and perpendicular to the Y axis is taken as the X axis; b q-Right is the Y-axis coordinate of the qth plantar pressure center trajectory of the right foot, j is the number of the right foot pressure sensor, n is the number of the right foot pressure sensors, and y j is the ordinate value of the position of the th sensor, F j Indicates the vertical ground reaction force value corresponding to the jth sensor of the right foot.

6. An electronic device, characterized in that: include: A memory and a processor, wherein the memory is used to store a computer program, and when the computer program is executed by the processor, the spatiotemporal gait feature extraction method based on the plantar pressure center trajectory described in any one of claims 1 to 5 is implemented.

7. A readable storage medium, characterized in that: The readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the spatiotemporal gait feature extraction method based on the plantar pressure center trajectory according to any one of claims 1 to 5 is implemented.