Rehabilitation Exercise Information Monitoring Method for Cerebral Palsy Patients Based on Wearable Devices

Abstract

The present invention relates to the technical field of data compression, and particularly to a method for monitoring rehabilitation exercise information of cerebral palsy patients based on wearable devices; obtaining a motion state according to the distribution characteristics of heart rate data; obtaining an acceleration similarity according to the types of acceleration data; obtaining heart rate outliers and normal heart rate points according to the acceleration similarity; obtaining a first acceleration discrete point and an acceleration continuous point according to the heart rate outliers. Obtaining an acceleration cross-region point and a second acceleration discrete point according to the differential characteristics of the acceleration similarity; determining a corresponding motion state according to the acceleration cross-region point; obtaining an acceleration anomaly point according to the first acceleration discrete point and the second acceleration discrete point. The present invention adaptively compresses acceleration data according to the distribution characteristics of acceleration data corresponding to acceleration anomaly points and normal heart rate points, improving the compression efficiency and transmission real-time performance.

Description

Technical Field

[0001] The present invention relates to the technical field of data compression, and particularly to a method for monitoring rehabilitation exercise information of cerebral palsy patients based on wearable devices. Background Art

[0002] With the development of technology, wearable devices for monitoring user status are gradually becoming more portable. For example, smart watches can identify various characteristics of users, such as their exercise behavior, exercise intensity, and heart rate, and then analyze the users' exercise level and health level.

[0003] During the movement of cerebral palsy patients, the watch will obtain acceleration data through a three-axis acceleration sensor. The watch transmits the obtained acceleration data to the terminal. Since the amount of acceleration data is huge and is generally transmitted via Bluetooth, the transmission rate is slow when the data volume is large. Therefore, the data needs to be compressed during transmission to improve the transmission rate. The commonly used run-length encoding method can compress data, but due to numerous changes in the acceleration data, it may lead to an increase in the amount of compressed characters and poor compression effect, resulting in the inability to improve the transmission speed and affecting the real-time monitoring of the user's exercise situation. Summary of the Invention

[0004] In order to solve the technical problem of poor transmission speed when compressing acceleration data by the run-length encoding method, the purpose of the present invention is to provide a method for monitoring rehabilitation exercise information of cerebral palsy patients based on wearable devices. The specific technical solution adopted is as follows:

[0005] Obtain the heart rate data and acceleration data of the watch; obtain different exercise states according to the distribution characteristics of the heart rate data;

[0006] Obtain the acceleration similarity according to the type characteristics of the acceleration data corresponding to the heart rate data; obtain the heart rate outliers and normal heart rate points according to the difference characteristics of the acceleration similarity in the exercise state; obtain the first acceleration discrete points and acceleration continuous points of the corresponding acceleration data according to the timing characteristics of the heart rate outliers;

[0007] Obtain the acceleration cross-region points and the second acceleration discrete points according to the difference characteristics of the acceleration similarity corresponding to the acceleration continuous points; determine the corresponding exercise state according to the acceleration cross-region points; obtain the acceleration abnormal points according to the first acceleration discrete points and the second acceleration discrete points;

[0008] Compress the acceleration data according to the distribution characteristics of the acceleration data corresponding to the acceleration abnormal points and normal heart rate points in the exercise state.

[0009] Further, the step of obtaining different exercise states according to the distribution characteristics of the heart rate data includes:

[0010] Calculate the product of the user's maximum heart rate and a preset first value to obtain the minimum exercise heart rate; calculate the product of the user's maximum heart rate and a preset second value to obtain the maximum exercise heart rate; take zero to the minimum exercise heart rate as the normal exercise state; take the minimum exercise heart rate to the maximum exercise heart rate as the moderate exercise state; take the maximum exercise heart rate to the preset maximum heart rate as the strenuous exercise state.

[0011] Further, the step of obtaining the user's maximum heart rate includes:

[0012] Calculate the difference between a preset heart rate value and the user's age to obtain the user's maximum heart rate.

[0013] Further, the step of obtaining the acceleration similarity according to the type characteristics of the acceleration data corresponding to the heart rate data includes:

[0014] Calculate the information entropy according to the probabilities of different types in the acceleration data corresponding to any heart rate data point and perform a negative correlation mapping to obtain the acceleration similarity of the any heart rate data point.

[0015] Further, the step of obtaining the heart rate outlier points and heart rate normal points according to the difference characteristics of the acceleration similarity in the exercise state includes:

[0016] Calculate the average value of all acceleration similarities in the exercise state to obtain the similarity mean value; calculate the product of the standard deviation of the acceleration similarity in the exercise state and a preset constant to obtain the similarity standard deviation; calculate the absolute value of the difference between the acceleration similarity of any heart rate data point and the similarity mean value to obtain the similarity difference; calculate the difference between the similarity difference and the similarity standard deviation to obtain the discrete judgment coefficient;

[0017] When the discrete judgment coefficient exceeds a preset discrete threshold, the any heart rate data point is a heart rate outlier point; otherwise, it is a heart rate normal point.

[0018] Further, the preset discrete threshold is 0.

[0019] Further, the step of obtaining the first acceleration discrete points and acceleration continuous points of the corresponding acceleration data according to the timing characteristics of the heart rate outlier points includes:

[0020] Sort the heart rate outlier points from small to large according to the time characteristics, take the acceleration data corresponding to the heart rate outlier points with discontinuous time characteristics in the sequence as the first acceleration discrete points, and take the acceleration data corresponding to the heart rate outlier points with continuous time characteristics in the sequence as the acceleration continuous points.

[0021] Further, the step of obtaining the acceleration cross-region point and the second acceleration discrete point according to the difference feature of the acceleration similarity corresponding to the acceleration continuous points includes:

[0022] Calculate the average value of the acceleration similarity corresponding to the acceleration continuous points in any time period and the acceleration similarity of the heart rate normal points in any other motion state to obtain the cross-region similarity mean value. Calculate the standard deviation of the acceleration similarity corresponding to the acceleration continuous points in any time period and the acceleration similarity of the heart rate normal points in any other motion state to obtain the cross-region standard deviation. Calculate the product of the cross-region standard deviation and a preset constant to obtain the cross-region similarity standard deviation. Calculate the absolute value of the difference between the acceleration similarity corresponding to any acceleration continuous point in the acceleration continuous points in any time period and the cross-region similarity mean value to obtain the cross-region similarity difference. Calculate the difference between the cross-region similarity difference and the cross-region similarity standard deviation to obtain the cross-region discrete judgment coefficient.

[0023] When the cross-region discrete judgment coefficient exceeds the preset discrete threshold, the arbitrary acceleration continuous point is the second acceleration discrete point; otherwise, it is the acceleration cross-region point.

[0024] Further, the step of determining the corresponding motion state according to the acceleration cross-region point includes:

[0025] Judge the minimum value of the cross-region discrete judgment coefficient obtained from the acceleration cross-region point and any other motion state. The acceleration cross-region point belongs to any other motion state corresponding to the minimum value of the cross-region discrete judgment coefficient, and move the acceleration cross-region point to any other motion state.

[0026] Further, the step of compressing the acceleration data according to the distribution characteristics of the acceleration data corresponding to the acceleration abnormal points and the heart rate normal points in the motion state includes:

[0027] In the motion state, compress the acceleration data corresponding to each segment of heart rate normal points with the acceleration mean value as the repeated data of run-length encoding, and compress the acceleration data corresponding to the acceleration abnormal points through the ASCII encoding algorithm.

[0028] The present invention has the following beneficial effects:

[0029] In the embodiments of the present invention, obtaining different motion states can classify different acceleration characteristics based on heart rate characteristics, which is convenient for determining the motion characteristics corresponding to the acceleration and provides a basis for improving the compression efficiency; calculating the acceleration similarity can characterize the degree of chaos of the acceleration data corresponding to the heart rate data, and then judge the encoding method of the acceleration data corresponding to that moment. Obtaining heart rate outliers and normal heart rate points initially determines the range of acceleration data that can be encoded; obtaining acceleration continuous points can further analyze the range of acceleration data that can be encoded based on the characteristic of the lag between heart rate data and acceleration data, further improving the compression efficiency. Obtaining acceleration cross-region points and the corresponding motion states can accurately determine the encoding method for this type of acceleration data and avoid reducing the compression efficiency; obtaining acceleration anomalies can determine the range of acceleration data for selecting the encoding method. Finally, the acceleration data is compressed according to the distribution characteristics of the acceleration data corresponding to the acceleration anomaly points and normal heart rate points in the motion state, improving the compression efficiency and transmission real-time performance while ensuring that the change characteristic trend of the overall acceleration data remains unchanged. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0031] Figure 1 It is a flowchart of a method for monitoring the rehabilitation movement information of cerebral palsy patients based on a wearable device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0032] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following will, in conjunction with the drawings and preferred embodiments, describe in detail the specific implementation manner, structure, characteristics, and effects of a method for monitoring the rehabilitation movement information of cerebral palsy patients based on a wearable device proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0033] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0034] The following will specifically describe the specific solution of a method for monitoring the rehabilitation movement information of cerebral palsy patients based on a wearable device provided by the present invention in conjunction with the drawings.

[0035] Please refer to Figure 1 , which shows a flowchart of a method for monitoring rehabilitation exercise information of cerebral palsy patients based on a wearable device provided by an embodiment of the present invention. The method includes the following steps:

[0036] Step S1, obtain the heart rate data and acceleration data of the watch; obtain different exercise states according to the distribution characteristics of the heart rate data.

[0037] In the embodiment of the present invention, the implementation scenario is the compression of the acceleration data collected by the smart watch during the user's exercise. Because the acceleration data will change in real time during the user's exercise, even though the acceleration data is related to the exercise intensity, it will not be stable within a small range, which will result in poor compression effect of the acceleration data; while the user's heart rate will maintain at different levels according to different exercise intensities, the higher the exercise intensity, the higher the heart rate, and it will not change easily; therefore, in order to improve the compression efficiency, the acceleration data at the corresponding moment can be quantified according to the heart rate characteristics of the user, and the similar acceleration data can be quantified into the same value, thereby improving the compression efficiency of the run-length encoding method and the data transmission efficiency. It should be noted that the run-length encoding compression algorithm belongs to the prior art, and the specific compression steps will not be described in detail. Among them, the user in the embodiment of the present invention represents a cerebral palsy patient.

[0038] Further, obtain the heart rate data and acceleration data of the watch, and denoise the data through the median filtering algorithm during the obtaining process; the data can be obtained through the built-in sensor of the watch. In the embodiment of the present invention, the heart rate data is collected once per minute, and the acceleration data is collected once per second, that is, one heart rate data corresponds to 60 acceleration data; the collected acceleration data is three-axis data. Since the compression methods of the three-axis acceleration data are the same, only the acceleration data of any one axis is analyzed in the embodiment of the present invention, and the implementer can determine the collection frequency according to the implementation scenario.

[0039] In order to standardize the quantification of the acceleration data of the user under different exercise intensities, it is necessary to analyze the exercise intensity of the user at different times; since the exercise intensity of the user can be characterized by the heart rate, the higher the heart rate during exercise, the greater the exercise intensity, so different exercise states can be obtained according to the distribution characteristics of the heart rate data.

[0040] Preferably, in an embodiment of the present invention, obtaining the motion state includes: calculating the difference between a preset heart rate value and the age of the user to obtain the maximum heart rate of the user. In the embodiment of the present invention, the maximum heart rate of the user can reflect the maximum heart rate value of the user under normal activities. The preset heart rate value is a heart rate index under a preset standard condition. Optionally, the preset heart rate value is 220 beats. That is, when the user is 20 years old, the maximum heart rate value of the user is 220 - 20 = 200 beats. Calculate the product of the maximum heart rate of the user and a preset first value to obtain the minimum exercise heart rate; calculate the product of the maximum heart rate of the user and a preset second value to obtain the maximum exercise heart rate; in the embodiment of the present invention, the preset first value is 0.6 and the preset second value is 0.8. Take zero to the minimum exercise heart rate as the normal exercise state; take the minimum exercise heart rate to the maximum exercise heart rate as the moderate exercise state; take the maximum exercise heart rate to the preset maximum heart rate as the intense exercise state, and the preset maximum heart rate is set to 280. For example, for a certain user who is 20 years old, the heart rate range of the normal exercise state is ; the heart rate range of the moderate exercise state is ; the heart rate range of the intense exercise state is ; the implementer can determine it according to the implementation scenario by himself.

[0041] Step S2, obtain the acceleration similarity according to the type characteristics of the acceleration data corresponding to the heart rate data; obtain the heart rate outlier points and heart rate normal points according to the difference characteristics of the acceleration similarity in the motion state; obtain the first acceleration discrete points and acceleration continuous points of the corresponding acceleration data according to the time series characteristics of the heart rate outlier points.

[0042] After obtaining the different motion states of the user, the change characteristics of the acceleration data in this motion state can be analyzed, such as the normal characteristics during normal exercise, the moderate characteristics during moderate exercise, and the intense characteristics during intense exercise. During the encoding process, the acceleration data with relatively small change differences can be corresponding with the same value to improve the compression efficiency, and the acceleration data with large change differences can be encoded separately, thereby improving the compression and transmission effects; therefore, the acceleration similarity can be obtained according to the type characteristics of the acceleration data corresponding to the heart rate data.

[0043] Preferably, in an embodiment of the present invention, obtaining the acceleration similarity includes: calculating the information entropy according to the probabilities of different types in the acceleration data corresponding to any heart rate data point and performing a negative correlation mapping. It should be noted that the information entropy belongs to the prior art, and the specific calculation steps will not be elaborated. Obtain the acceleration similarity of this arbitrary heart rate data point, that is, analyze the probability of each type appearing in the 60 acceleration data corresponding to a heart rate data point. The more types there are, it means that the acceleration data in this period is more chaotic, the information entropy is larger, and the acceleration similarity is smaller.

[0044] Because although there are certain differences in the acceleration data in the same motion state, the overall change range and numerical interval are determined. There may be individual time periods with motion changes that cause the acceleration data to deviate from the normal level under this motion intensity. Therefore, the acceleration similarity is the same at most times, while the acceleration similarity at individual times is quite different from the overall situation. The heart rate outliers and normal heart rate points can be obtained based on the difference characteristics of the acceleration similarity in the motion state.

[0045] Preferably, in an embodiment of the present invention, obtaining the heart rate outliers and normal heart rate points includes: calculating the average value of all acceleration similarities in this motion state to obtain the similarity mean value; calculating the product of the standard deviation of the acceleration similarity in this motion state and a preset constant to obtain the similarity standard deviation. The larger the similarity standard deviation, the more discrete the acceleration similarity. In the embodiment of the present invention, the preset constant is 3. Calculate the absolute value of the difference between the acceleration similarity of any heart rate data point and the similarity mean value to obtain the similarity difference; calculate the difference between the similarity difference and the similarity standard deviation to obtain the discrete judgment coefficient. The larger the discrete judgment coefficient, the more chaotic the acceleration data corresponding to the heart rate data point at this moment, and the greater the difference from the change characteristics of the acceleration data at other times in this motion state. When the discrete judgment coefficient exceeds the preset discrete threshold, any heart rate data point is a heart rate outlier, and the heart rate outlier represents that the change of the acceleration data at this moment is relatively chaotic; otherwise, it is a normal heart rate point. In the embodiment of the present invention, the preset discrete threshold is 0, and the implementer can determine it according to the implementation scenario. The formula for obtaining the discrete judgment coefficient includes:

[0046]

[0047] In the formula, represents the discrete judgment coefficient, represents the acceleration similarity, represents the similarity mean value, represents the preset constant, represents the standard deviation of the acceleration similarity in this motion state, represents the similarity difference, represents the similarity standard deviation.

[0048] Furthermore, during the user's exercise, there is a lag between the change in heart rate and the change in acceleration data. For example, when the user starts warming up, the acceleration data will start to change directly. At this time, the heart rate is still in the normal exercise state range. After a while, the heart rate will enter the moderate exercise state. Therefore, when calculating the heart rate outliers, there are two cases for the heart rate outliers. One is that at any moment during the exercise state, there is a certain change in the user's exercise behavior, resulting in an abnormal change in the acceleration data corresponding to that heart rate moment. The other case is that at the end of the exercise state, the heart rate has not changed yet, but the acceleration data has already changed significantly. For this case, it is necessary to classify this type of heart rate outliers into other exercise states to improve the coding accuracy. Therefore, the first acceleration discrete points and acceleration continuous points of the corresponding acceleration data can be obtained according to the timing characteristics of the heart rate outliers.

[0049] Preferably, in an embodiment of the present invention, obtaining the first acceleration discrete points and acceleration continuous points includes: sorting the heart rate outliers in ascending order according to the time characteristics, and taking the acceleration data corresponding to the heart rate outliers with discontinuous time characteristics in the sequence as the first acceleration discrete points. The first acceleration discrete points are that at any moment during the exercise state, there is a certain change in the user's exercise behavior, resulting in an abnormal change in the acceleration data corresponding to that heart rate moment. Taking the acceleration data corresponding to the heart rate outliers with continuous time characteristics in the sequence as the acceleration continuous points. The acceleration continuous points refer to the moments when the corresponding heart rate data is continuous within a period of time. The acceleration continuous points may represent the situation where the user is about to enter other exercise states, but the heart rate has not changed yet. There may be multiple periods of the first acceleration discrete points and multiple acceleration continuous points in any exercise state interval.

[0050] Step S3, obtaining the acceleration cross-region points and the second acceleration discrete points according to the difference characteristics of the acceleration similarity corresponding to the acceleration continuous points; determining the corresponding exercise state according to the acceleration cross-region points; obtaining the acceleration abnormal points according to the first acceleration discrete points and the second acceleration discrete points.

[0051] Since it is analyzed in step S2 that the acceleration characteristics of the possible acceleration continuous points belong to other exercise states, in order to avoid recognizing such data points as abnormal points and increasing the compression time, it is necessary to analyze whether such acceleration continuous points belong to other exercise states, and obtaining the acceleration cross-region points and the second acceleration discrete points according to the difference characteristics of the acceleration similarity corresponding to the acceleration continuous points.

[0052] Preferably, in the embodiment of the present invention, obtaining the acceleration cross-region point and the second acceleration discrete point includes: calculating the average value of the acceleration similarity corresponding to the continuous acceleration points in any time period and the acceleration similarity of the heart rate normal points in any other motion state, obtaining the cross-region similarity mean value; calculating the standard deviation of the acceleration similarity corresponding to the continuous acceleration points in any time period and the acceleration similarity of the heart rate normal points in any other motion state, obtaining the cross-region standard deviation; calculating the product of the cross-region standard deviation and a preset constant, obtaining the cross-region similarity standard deviation; wherein the preset constant is the same as the preset constant in step S2. Calculating the absolute value of the difference between the acceleration similarity corresponding to any continuous acceleration point in any time period of the continuous acceleration points and the cross-region similarity mean value, obtaining the cross-region similarity difference; calculating the difference between the cross-region similarity difference and the cross-region similarity standard deviation, obtaining the cross-region discrete judgment coefficient. When the cross-region discrete judgment coefficient exceeds the preset discrete threshold, the arbitrary continuous acceleration point is the second acceleration discrete point; otherwise, it is the acceleration cross-region point. The analysis method of the acceleration cross-region point and the second acceleration discrete point is the same as the process of analyzing the heart rate outlier in step S2; the second acceleration discrete point means that the acceleration characteristics of the corresponding continuous acceleration point do not belong to any type of motion state; the acceleration cross-region point means that the user's motion behavior belongs to another type of motion state, but the heart rate state has not changed, and it is necessary to continue to analyze the motion state to which the acceleration cross-region point belongs.

[0053] Further, determining the corresponding motion state according to the acceleration cross-region point specifically includes: judging the minimum value of the cross-region discrete judgment coefficient obtained by the acceleration cross-region point and any other motion state, and the acceleration cross-region point belongs to any other motion state corresponding to the minimum value of the cross-region discrete judgment coefficient; and moving the acceleration cross-region point to any other motion state; because this type of acceleration cross-region point is usually close to the beginning or the end of the any other motion state, that is, splicing this type of acceleration cross-region point to the beginning or the end of the any other motion state. The smaller the cross-region discrete judgment coefficient of the acceleration cross-region point, the closer the acceleration characteristics of the acceleration cross-region point are to the acceleration characteristics in the any other motion state; in the subsequent coding and compression process, it is not necessary to perform separate compression, improving the compression efficiency.

[0054] For the first acceleration discrete point and the second acceleration discrete point, the two are regarded as acceleration abnormal points. The acceleration characteristics of this type of acceleration data point are relatively abnormal and are quite different from the common acceleration data in the motion process. Therefore, in order to improve the compression efficiency, it is necessary to extract this type of data for separate compression.

[0055] Step S4, compressing the acceleration data according to the distribution characteristics of the acceleration data corresponding to the acceleration abnormal points and the heart rate normal points in the motion state.

[0056] During the motion state, the average acceleration of the acceleration data corresponding to each normal heart rate point is used as the repeated data for run-length encoding, where the normal heart rate points include the heart rate data corresponding to the acceleration cross-region points; the acceleration data corresponding to the acceleration abnormal points is compressed through the ASCII encoding algorithm. It should be noted that run-length encoding and the ASCII encoding algorithm belong to the prior art, and the specific compression steps will not be elaborated. For example, in a certain motion state lasting for 5 minutes, an acceleration abnormal point appears at the 3rd minute. Then, the average acceleration from the start of this motion state to the acceleration abnormal point is used as the repeated data for run-length encoding. The types of this section of acceleration data are relatively close and can be replaced by the same data, encoded as , where represents the motion state type and is convenient to be used as the segmentation symbol for different operation types in the compressed data. The acceleration data at the 3rd minute is encoded through ASCII, and the acceleration data after the acceleration abnormal point is then encoded by run-length encoding with the average acceleration of the acceleration data at the normal heart rate points in this section. By encoding with this method, the chaotic acceleration data can be standardized and quantified through the heart rate characteristics, and the compression efficiency and transmission timeliness can be improved without changing the overall change trend of the acceleration data.

[0057] In summary, the embodiments of the present invention provide a method for monitoring the rehabilitation exercise information of cerebral palsy patients based on a wearable device; obtaining the motion state according to the distribution characteristics of the heart rate data; obtaining the acceleration similarity according to the types of the acceleration data; obtaining the heart rate outlier points and the normal heart rate points according to the acceleration similarity; obtaining the first acceleration discrete point and the acceleration continuous point according to the heart rate outlier points. Obtaining the acceleration cross-region point and the second acceleration discrete point according to the difference characteristics of the acceleration similarity; determining the corresponding motion state according to the acceleration cross-region point; obtaining the acceleration abnormal point according to the first acceleration discrete point and the second acceleration discrete point. The present invention adaptively compresses the acceleration data according to the distribution characteristics of the acceleration data corresponding to the acceleration abnormal points and the normal heart rate points, improving the compression efficiency and transmission real-time performance.

[0058] It should be noted that: the above sequence of the embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0059] Each embodiment in this specification is described in a progressive manner. The same or similar parts among the embodiments can be referred to each other, and the key points of each embodiment are to illustrate the differences from other embodiments.

Claims

1. A rehabilitation exercise information monitoring method for cerebral palsy patients based on wearable devices, characterized in that, The method includes the following steps: Obtain the heart rate data and acceleration data of the watch; obtain different motion states according to the distribution characteristics of the heart rate data; Obtain the acceleration similarity according to the type characteristics of the acceleration data corresponding to the heart rate data; obtain the heart rate outlier points and heart rate normal points according to the difference characteristics of the acceleration similarity in the motion states; obtain the first acceleration discrete points and acceleration continuous points of the corresponding acceleration data according to the timing characteristics of the heart rate outlier points; Obtain the acceleration cross-region points and the second acceleration discrete points according to the difference characteristics of the acceleration similarity corresponding to the acceleration continuous points; determine the corresponding motion state according to the acceleration cross-region points; obtain the acceleration abnormal points according to the first acceleration discrete points and the second acceleration discrete points; Compress the acceleration data according to the distribution characteristics of the acceleration abnormal points and the acceleration data corresponding to the heart rate normal points in the motion states; The step of obtaining the heart rate outlier points and heart rate normal points according to the difference characteristics of the acceleration similarity in the motion states includes: Calculate the average value of all acceleration similarities in the motion state to obtain the similarity mean value; calculate the product of the standard deviation of the acceleration similarity in the motion state and a preset constant to obtain the similarity standard deviation; calculate the absolute value of the difference between the acceleration similarity of any heart rate data point and the similarity mean value to obtain the similarity difference; calculate the difference between the similarity difference and the similarity standard deviation to obtain the discrete judgment coefficient; When the discrete judgment coefficient exceeds the preset discrete threshold, the any heart rate data point is a heart rate outlier point; otherwise, it is a heart rate normal point; The step of obtaining the first acceleration discrete points and acceleration continuous points of the corresponding acceleration data according to the timing characteristics of the heart rate outlier points includes: Sort the heart rate outlier points from small to large according to the time characteristics, use the acceleration data corresponding to the heart rate outlier points with discontinuous time characteristics in the sequence as the first acceleration discrete points, and use the acceleration data corresponding to the heart rate outlier points with continuous time characteristics in the sequence as the acceleration continuous points; Obtaining the acceleration cross-region points and the second acceleration discrete points according to the difference characteristics of the acceleration similarity corresponding to the acceleration continuous points includes: Calculate the average value of the acceleration similarity corresponding to the acceleration continuous points in any time period and the acceleration similarity of the heart rate normal points in any other motion state to obtain the cross-region similarity mean value, calculate the standard deviation of the acceleration similarity corresponding to the acceleration continuous points in any time period and the acceleration similarity of the heart rate normal points in any other motion state to obtain the cross-region standard deviation, calculate the product of the cross-region standard deviation and a preset constant to obtain the cross-region similarity standard deviation; calculate the absolute value of the difference between the acceleration similarity corresponding to any acceleration continuous point in any time period of the acceleration continuous points and the cross-region similarity mean value to obtain the cross-region similarity difference, calculate the difference between the cross-region similarity difference and the cross-region similarity standard deviation to obtain the cross-region discrete judgment coefficient; when the cross-region discrete judgment coefficient exceeds the preset discrete threshold, the any acceleration continuous point is the second acceleration discrete point; otherwise, it is the acceleration cross-region point.

2. The method for monitoring the rehabilitation exercise information of cerebral palsy patients based on a wearable device according to claim 1, wherein, The step of obtaining different exercise states according to the distribution characteristics of the heart rate data includes: Calculating the product of the user's maximum heart rate and a preset first value to obtain the minimum exercise heart rate; calculating the product of the user's maximum heart rate and a preset second value to obtain the maximum exercise heart rate; taking zero to the minimum exercise heart rate as the normal exercise state; taking the minimum exercise heart rate to the maximum exercise heart rate as the moderate exercise state; taking the maximum exercise heart rate to the preset maximum heart rate as the intense exercise state.

3. The monitoring method for the rehabilitation exercise information of cerebral palsy patients based on a wearable device according to claim 2, wherein The step of obtaining the user's maximum heart rate includes: Calculating the difference between a preset heart rate value and the user's age to obtain the user's maximum heart rate.

4. A monitoring method for the rehabilitation exercise information of cerebral palsy patients based on a wearable device according to claim 1, characterized in that, The type characteristics of the acceleration data include normal characteristics during normal exercise, moderate characteristics during moderate exercise, and intense characteristics during intense exercise. The step of obtaining the acceleration similarity according to the type characteristics of the acceleration data corresponding to the heart rate data includes: Calculating the information entropy based on the probabilities of different types in the acceleration data corresponding to any heart rate data point and performing a negatively correlated mapping to obtain the acceleration similarity of the any heart rate data point.

5. A method for monitoring the rehabilitation exercise information of cerebral palsy patients based on a wearable device according to claim 1, characterized in that, The preset discrete threshold is 0.

6. The monitoring method for the rehabilitation exercise information of cerebral palsy patients based on a wearable device according to claim 1, wherein The step of determining the corresponding exercise state according to the acceleration cross-region point includes: Judging the minimum value of the cross-region discrete judgment coefficient obtained by the acceleration cross-region point and any other exercise state. The acceleration cross-region point belongs to any other exercise state corresponding to the minimum value of the cross-region discrete judgment coefficient, and moving the acceleration cross-region point to the any other exercise state.

7. A method for monitoring the rehabilitation exercise information of cerebral palsy patients based on a wearable device according to claim 1, characterized in that, The step of compressing the acceleration data according to the distribution characteristics of the acceleration data corresponding to the acceleration abnormal point and the heart rate normal point in the exercise state includes: In the exercise state, using the acceleration mean value of the acceleration data corresponding to each segment of heart rate normal points as the repeated data for run-length encoding for compression, and compressing the acceleration data corresponding to the acceleration abnormal point through the ASCII encoding algorithm.

Images