Sensor data processing method and device, wearable equipment and storage medium

Abstract

The invention provides a sensor data processing method and device, wearable equipment and a computer readable storage medium. The sensor data processing method comprises the following steps: acquiring a first time interval when a sensor module collects two frames of sensing data; acquiring a marking use timestamp of the previous frame of sensing data of the current frame of sensing data; and based on the annotation use timestamp of the previous frame of sensing data and the first time interval, determining the annotation use timestamp of the current frame of sensing data. According to the method and the device, the error between the use timestamp marked by the sensing data of the wearable equipment and the actual use timestamp of the sensing data of the wearable equipment can be reduced, so that the error accumulation of an operation scene which needs to depend on the marked use timestamp is reduced.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and specifically to a sensor data processing method, apparatus, wearable device and computer-readable storage medium. Background Art

[0002] In recent years, wearable devices have been loved by the general public. In order to better meet people's application needs, the functions of wearable devices have become more and more diversified. Many functions of wearable devices are inseparable from sensor data. Therefore, the accuracy of sensor data and its timestamp information has important guiding significance for wearable devices.

[0003] However, in the related art, there is an error between the usage timestamp annotated with the sensor data of the wearable device and the actual usage timestamp of the sensor data of the wearable device (here, the actual usage timestamp refers to the actual collection time point of the sensor data used, and the annotated usage timestamp refers to the annotated collection time point of the sensor data used), which will cause the errors in some scenarios that rely on the annotated usage timestamps for calculations to continue to accumulate. Summary of the invention

[0004] The present application provides a sensor data processing method, device, smart glasses and computer-readable storage medium, which can reduce the error between the usage timestamp annotated with the sensor data of a wearable device and the actual usage timestamp of the sensor data of the wearable device, thereby reducing the error accumulation in scenarios that rely on the annotated usage timestamp for calculations.

[0005] In a first aspect, the present application provides a sensor data processing method, the method being applied to a wearable device, the wearable device comprising a sensor module, the method comprising:

[0006] Acquire a first time interval for the sensor module to collect two frames of sensing data;

[0007] Obtain the timestamp of the previous frame of sensor data of the current frame;

[0008] The timestamp used for labeling the sensor data of the previous frame is determined based on the timestamp used for labeling the sensor data of the previous frame and the first time interval.

[0009] In a second aspect, the present application provides a sensor data processing device, the sensor data processing device comprising:

[0010] A first acquisition unit, used for acquiring a first time interval for the sensor module to collect two frames of sensing data;

[0011] A second acquisition unit, used for acquiring a timestamp of a previous frame of sensor data of a current frame of sensor data;

[0012] The processing unit is used to determine the timestamp used for labeling the current frame of sensor data based on the timestamp used for labeling the previous frame of sensor data and the first time interval.

[0013] In a third aspect, the present application also provides a wearable device, comprising a main control module and a memory, wherein a computer program is stored in the memory, and when the main control module calls the computer program in the memory, any sensor data processing method provided in the present application is executed.

[0014] In a fourth aspect, the present application also provides a computer-readable storage medium on which a computer program is stored, and the computer program is loaded by a processor to execute the sensor data processing method.

[0015] The present application determines the annotated usage timestamp of the current frame sensor data according to the first time interval between two frames of sensor data collected by the sensor module and the annotated usage timestamp of the previous frame of sensor data of the current frame sensor data. On the one hand, by obtaining the first time interval between two frames of sensor data collected by the sensor module, the actual collection interval can be used as the first time interval. Thus, the annotated usage timestamp of the current frame sensor data is determined in combination with the first time interval and the annotated usage timestamp of the previous frame of sensor data. When the interval between two frames of sensor data collected by the sensor module and the interval between the two frames of sensor data read by the main control module are inconsistent, the annotated usage timestamp of the current frame sensor data can be made closer to the true value of the collection time point, thereby reducing the error between the annotated usage timestamp of the sensor data of the wearable device and the actual usage timestamp of the sensor data of the wearable device, thereby reducing the error accumulation of the scenario that needs to rely on the annotated usage timestamp for calculation, and improving the accuracy of the sensor data timestamp. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 is a schematic block diagram of the structure of a wearable device provided in an embodiment of the present application;

[0018] Figure 2 It is a flow chart of a sensor data processing method provided in an embodiment of the present application;

[0019] Figure 3 This is a schematic diagram of an embodiment of step 202 provided in an embodiment of the present application;

[0020] Figure 4 is another flow chart of the sensor data processing method provided in the embodiment of the present application;

[0021] Figure 5 It is a schematic diagram of an embodiment of the sensor data processing flow provided in an embodiment of the present application;

[0022] Figure 6 It is a schematic diagram of the structure of an embodiment of a sensor data processing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0024] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0025] In the description of the embodiments of the present application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present application, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0026] In order to enable any person skilled in the art to implement and use the present application, the following description is provided. In the following description, details are listed for the purpose of explanation. It should be understood that those of ordinary skill in the art can recognize that the present application can also be implemented without using these specific details. In other examples, the known process will not be elaborated in detail to avoid unnecessary details that make the description of the present application embodiment obscure. Therefore, the present application is not intended to be limited to the embodiments shown, but is consistent with the widest range of principles and features disclosed in accordance with the embodiments of the present application.

[0027] In the related technology, there is an error between the usage timestamp marked on the sensor data of the wearable device and the actual usage timestamp of the sensor data of the wearable device (here, the actual usage timestamp refers to the actual collection time point of the sensor data used, and the marked usage timestamp refers to the marked collection time point of the sensor data used), which will lead to an increasing error in some scenarios that rely on the marked usage timestamps for calculations, and ultimately lead to the inability to be used normally in business scenarios. After research, the inventor of the embodiment of the present application found that one of the reasons for "the error between the usage timestamp marked with the sensor data of the wearable device and the actual usage timestamp of the sensor data of the wearable device" is that: the interval between the sensor module collecting two frames of sensor data (such as the interval between the sensor module collecting the first frame of sensor data and the second frame of sensor data is 0.01 seconds) and the interval between the main control module reading the two frames of sensor data (such as the interval between the main control module reading the first frame of sensor data and the second frame of sensor data is 0.02 seconds) may be inconsistent. If the reading time of the sensor data is marked as the usage time of the sensor data, there will be an error between the usage timestamp marked with the sensor data of the wearable device and the actual usage timestamp of the sensor data of the wearable device.

[0028] To this end, the embodiments of the present application provide a sensor data processing method, apparatus, wearable device, and computer-readable storage medium, which can reduce the error between the usage timestamp annotated by the sensor data of the wearable device and the actual usage timestamp of the sensor data of the wearable device, thereby reducing the error accumulation in the calculation scenario that relies on the annotated usage timestamp and improving the accuracy of the sensor data timestamp. The sensor data processing apparatus can be integrated in the wearable device.

[0029] The executor of the sensor data processing method of the embodiment of the present application may be the sensor data processing device provided in the embodiment of the present application, or a wearable device integrating the sensor data processing device, wherein the sensor data processing device may be implemented in hardware or software.

[0030] Some embodiments of the present application are described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0031] Figure 1 It is a schematic block diagram of the structure of a wearable device provided in an embodiment of the present application.

[0032] like Figure 1As shown, the wearable device 100 includes a main control module 101 and a sensor module 102. The main control module 101 and the sensor module 102 can communicate through communication protocols such as I2C protocol, SPI protocol, UART protocol, etc. The main control module 101 uses this to write registers to the sensor module 102 or read register data from the sensor module 102. When reading data, the main control module 101 can also read out the timestamp inside the sensor module 102. The wearable device 100 can be smart glasses, smart watches, smart rings, etc. Smart glasses can include but are not limited to: AR (augmented reality) glasses, VR (Virtual Reality) glasses, etc.

[0033] Specifically, the main control module 101 is used to provide computing and control capabilities to support the operation of the entire wearable device 100. The main control module 101 can be a central processing unit (CPU), and the main control module 101 can also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0034] Furthermore, the wearable device may further include a memory 103, and the main control module 101 and the memory 103 are connected via a bus 104, such as an I2C (Inter-integrated Circuit) bus. Specifically, the memory 103 may be a Flash chip, a read-only memory (ROM) disk, an optical disk, a USB flash drive, or a mobile hard disk.

[0035] Those skilled in the art will understand that Figure 1 The structure shown in the figure is merely a block diagram of a partial structure related to the embodiment scheme of the present application, and does not constitute a limitation on the wearable device to which the embodiment scheme of the present application is applied. The specific wearable device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0036] The main control module 101 is used to run the computer program stored in the memory 103, and implement any one of the sensor data processing methods provided in the embodiments of the present application when executing the computer program. For example, the main control module 101 is used to run the computer program stored in the memory 103, and can implement the following steps when executing the computer program:

[0037] Obtain a first time interval for a sensor module to collect two frames of sensor data; obtain a timestamp for annotating and using the previous frame of sensor data of a current frame of sensor data; and determine a timestamp for annotating and using the current frame of sensor data based on the timestamp for annotating and using the previous frame of sensor data and the first time interval.

[0038] It should be noted that technicians in the relevant field can clearly understand that for the convenience and conciseness of description, the specific working process of the wearable device described above can refer to the corresponding process in the following sensor data processing method embodiment, and will not be repeated here.

[0039] In the following, the sensor data processing method is applied to Figure 1 Taking the wearable device shown in as an example, the sensor data processing method provided by the embodiment of the present application is described in detail. It should be noted that, Figure 1 The scenarios are only used to explain the sensor data processing method provided in the embodiments of the present application, but do not constitute a limitation on the application scenarios of the sensor data processing method provided in the embodiments of the present application.

[0040] See also Figure 2 , Figure 2 1 is a flow chart of a sensor data processing method provided in an embodiment of the present application. The sensor data processing method is applied to a wearable device, the wearable device includes a sensor module and a main control module, the main control module and the sensor module can communicate through communication protocols such as I2C protocol, SPI protocol, UART protocol, etc., the wearable device can be smart glasses (such as AR glasses, VR glasses, etc.), smart watches, smart rings, etc.; the sensor data processing method includes steps 201 to 203, wherein:

[0041] 201. Obtain a first time interval for a sensor module to collect two frames of sensing data.

[0042] The first time interval is a collection interval between two frames of sensing data of the sensor module.

[0043] Among them, the sensor module is a sensor on the wearable device, and the sensor module may include but is not limited to: an inertial sensor (Inertial Measurement Unit, IMU sensor, a sensor mainly used to detect and measure acceleration and rotational motion), a magnetometer (used to test the strength and direction of the magnetic field, locate the orientation of the device, and can measure the angle between the current device and the four directions of east, south, west and north), a position sensor (used to measure or detect the position of an object), etc.

[0044] The sensor data is data collected by the sensor module. For example, if the sensor module is an IMU sensor, the sensor data may be acceleration collected by the IMU sensor.

[0045] In this embodiment, the sensor data collected at each time point is recorded as a frame of sensor data. Since the sensor data is sampled according to a certain sampling rate, there will be a certain interval between two adjacent frames of sensor data.

[0046] There are many ways to obtain the first time interval for the sensor module to collect two frames of sensing data in step 201, illustratively including:

[0047] (1) Before acquiring the current frame of sensor data each time, the main control module first reads the current sampling rate of the sensor module, and calculates the first time interval for the sensor module to collect two frames of sensor data based on the current sampling rate of the sensor module. At this time, step 201 may specifically include the following steps 2011A to 2012A:

[0048] 2011A. Get the current sampling rate of the sensor module.

[0049] 2012A. Based on the current sampling rate of the sensor module, determine a first time interval for the sensor module to collect two frames of sensing data.

[0050] For example, the reciprocal of the current sampling rate of the sensor module can be used as the first time interval for the sensor module to collect two frames of sensor data. For example, if the current sampling rate of the sensor module is 50 Hz (i.e., 50 frames of sensor data / second), the reciprocal of the current sampling rate of the sensor module (i.e., 1 / 50 = 0.02 seconds) can be used as the first time interval (i.e., 0.02 seconds) for the sensor module to collect two frames of sensor data.

[0051] Thus, the first time interval for the sensor module to collect two frames of sensor data is determined based on the current sampling rate of the sensor module; on the one hand, when the sampling rate of the sensor module fluctuates, the first time interval can be matched with the actual sampling rate of the sensor module, and finally the timestamp used for the annotation of the current frame of sensor data is closer to the real value at the time of its collection. On the other hand, when the interval for the sensor module to collect two frames of sensor data is inconsistent with the interval for the main control module to read the two frames of sensor data, the actual collection interval can be used as the first time interval, so that the timestamp used for the annotation of the current frame of sensor data is closer to the real value at the time of its collection.

[0052] (2) The main control module reads the number of sensor data frames within a period of time (e.g., 100 seconds), and calculates the average interval between every two frames of sensor data of the sensor module based on the number of sensor data frames within a period of time, as the first time interval for the sensor module to collect two frames of sensor data. At this time, step 201 may specifically include the following steps 2011B to 2012B:

[0053] 2011B. Obtain the number of sensor data frames collected by the sensor module within the statistical period.

[0054] 2012B. Based on the number of sensing data frames, determine an average interval between every two frames of sensing data of the sensor module as the first time interval.

[0055] The sensor data frame number refers to the number of frames of sensor data collected by the sensor module during the statistical period.

[0056] For example, if the number of sensor data frames collected by the sensor module in the statistical period (such as 10:00:00 to 10:01:40) is 1000 frames, the number of sensor data frames (i.e., 1000 frames) and the duration corresponding to the statistical period (i.e., 100 seconds) can be used for calculation to obtain the average interval between every two frames of sensor data of the sensor module (i.e., 100 / 1000=0.1 seconds) as the first time interval. The statistical period here is only an example, and the statistical period can be set according to the actual business scenario requirements and is not limited here.

[0057] Thus, by obtaining the number of sensor data frames collected by the sensor module within the statistical period, the average interval between every two frames of sensor data of the sensor module is determined as the first time interval; on the one hand, when the sampling rate of the sensor module fluctuates slightly, the first time interval can be made to match the actual sampling rate of the sensor module more closely, and finally the timestamp used for the annotation of the current frame of sensor data is closer to the real value at the time of its collection. On the other hand, when the interval between the two frames of sensor data collected by the sensor module is inconsistent with the interval between the two frames of sensor data read by the main control module, the actual collection interval can be used as the first time interval, so that the timestamp used for the annotation of the current frame of sensor data is closer to the real value at the time of its collection.

[0058] 202. Obtain a timestamp for labeling a previous frame of sensor data of a current frame of sensor data.

[0059] The current frame sensor data is the sensor data with a timestamp to be determined and marked.

[0060] The timestamp used for labeling the sensor data is used to indicate the time point when the transmitter module collects the sensor data used.

[0061] The annotation of the previous frame of sensor data of the current frame of sensor data uses a timestamp to indicate the time point when the previous frame of sensor data of the current frame of sensor data was collected.

[0062] There are many ways to obtain the annotation of the previous frame of sensor data of the current frame of sensor data in step 202 using the timestamp, which include, for example:

[0063] (1) Referring to steps 201 to 203, the timestamp used for labeling each frame of sensor data is determined by using the timestamp used for labeling the previous frame of sensor data (recorded as the t-1 frame of sensor data) and the first time interval (recorded as Δt1) between the sensor module collecting two frames of sensor data, that is, the timestamp used for labeling the t frame of sensor data = the timestamp used for labeling the t-1 frame of sensor data + the first time interval Δt1. Exemplarily, assuming that the current frame sensor data is the t-th frame sensor data, the previous frame sensor data of the current frame sensor data is the t-1th frame sensor data, the previous frame sensor data of the t-1th frame sensor data is the t-2th frame sensor data, and so on, then "the annotation usage timestamp of the previous frame sensor data of the current frame sensor data" can be specifically determined with reference to the following steps A1 to A3. In step 202, the annotation usage timestamp of the t-1th frame sensor data determined by referring to the following steps A1 to A3 can be read as the annotation usage timestamp of the previous frame sensor data of the current frame sensor data.

[0064] A1. Acquire the first time interval Δt1 between two frames of sensing data collected by the sensor module.

[0065] A2. Obtain the timestamp of the sensor data of frame t-2.

[0066] A3. Obtain the sum of the timestamp for labeling the sensing data of the t-2th frame and the first time interval Δt1 as the timestamp for labeling the sensing data of the t-1th frame.

[0067] For example, the timestamp used for labeling the t-2 frame of sensor data is 15 hours 40 minutes 36 seconds, and the first time interval Δt1 is 1 second, then the timestamp used for labeling the t-1 frame of sensor data is 15 hours 40 minutes 37 seconds.

[0068] (2) After research, it was found that another reason for "the error between the usage timestamp of the sensor data of the wearable device and the actual usage timestamp of the sensor data of the wearable device" is that the clock of the sensor module is not necessarily accurate. If the sensor data collection time point recorded by the clock of the sensor module is marked as the usage time of the sensor data, there will still be an error between the usage timestamp of the sensor data of the wearable device and the actual usage timestamp of the sensor data of the wearable device. For this reason, in the embodiments of the present disclosure, in some cases (for example, when the last frame of sensor data read is the first frame of sensor data collected by the sensor module), the time difference between the sensor module collecting the sample sensor data and the main control module reading the sample sensor data can be used to determine the annotated usage timestamp of the last frame of sensor data, thereby improving the accuracy of the annotated usage timestamp of the last frame of sensor data, and then improving the accuracy of the annotated usage timestamp of the current frame of sensor data. Please refer to Figure 3 At this time, step 202 may specifically include the following steps 2021 to 2023:

[0069] 2021. Acquire a second time interval between when the sensor module collects sample sensor data and when the main control module reads the sample sensor data.

[0070] 2022. Obtain a first reading timestamp when the main control module reads a previous frame of sensing data from the sensor module.

[0071] 2023. Determine a timestamp for marking the previous frame of sensing data based on the first reading timestamp and the second time interval.

[0072] The second time interval is the time difference between when the sensor module collects the sample sensing data and when the main control module reads the sample sensing data.

[0073] The sample sensing data is a sensing data test sample, which is used to test the time difference between when the sensor module collects the sensing data and when the main control module reads the sensing data.

[0074] There are multiple implementation methods for steps 2021 to 2023, illustratively including the following method 1 and method 2:

[0075] Mode 1: In some embodiments (for example, when the clock of the sensor module is more accurate than the clock of the main control module), in step 2021, the time point when the sensor module collects sample sensor data and the time point when the main control module reads the sample sensor data can be recorded respectively by the clock of the sensor module; then, the difference between the time point when the sensor module collects sample sensor data recorded by the clock of the sensor module and the time point when the main control module reads the sample sensor data recorded by the clock of the sensor module is used as the second time interval; for example, the clock of the sensor module records that the sensor module collects sample sensor data at 20:30:30, and the clock of the sensor module records that the main control module reads the sample sensor data at 20:30:31, then the second time interval is 1 second. In step 2022, the first reading timestamp refers to the reading time recorded by the clock of the sensor module when the sensor module reads the previous frame of sensor data of the current frame of sensor data; at this time, the time point when the main control module reads the previous frame of sensor data of the current frame of sensor data can be recorded by the clock of the sensor module as the first reading timestamp. In step 2023, the difference between the first read timestamp and the second time interval can be directly used as the timestamp for marking the previous frame of sensor data of the current frame. Similarly, the timestamp for marking the sensor data of the t-1th frame = the first read timestamp T of the sensor data of the t-1th frame read by the main control module from the sensor module. t-1 - A second time interval Δt2.

[0076] Therefore, on the one hand, since the timestamp used to mark the previous frame of sensor data is recorded using the clock of the sensor module, and the clock of the sensor module is more accurate than the clock of the main control module, the accuracy of the timestamp used to mark the previous frame of sensor data can be improved; on the other hand, since the second time interval is the difference between the time point when the sensor module collects sample sensor data recorded by the clock of the sensor module and the time point when the main control module reads the sample sensor data recorded by the clock of the sensor module, the second time interval can accurately reflect the time difference between the time point when the sensor module collects the sample sensor data and the time point when the main control module reads the sample sensor data. By using the difference between the first reading timestamp and the second time interval as the timestamp used to mark the previous frame of sensor data of the current frame of sensor data, the problem that the timestamp used to mark the previous frame of sensor data is not the time point when the sensor module collects the sample sensor data can be avoided, thereby improving the accuracy of the timestamp used to mark the previous frame of sensor data.

[0077] Mode 2: In some other embodiments (for example, when the clock of the main control module is more accurate than the clock of the sensor module), in step 2021, the time point when the sensor module collects sample sensor data and the time point when the main control module reads the sample sensor data can be recorded respectively by the clock of the main control module; then, the difference between the time point when the sensor module collects sample sensor data recorded by the clock of the main control module and the time point when the main control module reads the sample sensor data recorded by the clock of the main control module is used as the second time interval; for example, the clock of the main control module records that the sensor module collects sample sensor data at 10:00:00, and the clock of the main control module records that the main control module reads the sample sensor data at 10:00:01, then the second time interval is 1 second. In step 2022, the first reading timestamp refers to the reading time recorded by the clock of the main control module when the main control module reads the previous frame of sensor data of the current frame of sensor data; at this time, the time point when the main control module reads the previous frame of sensor data of the current frame of sensor data can be recorded by the clock of the sensor module as the first reading timestamp. In step 2023, the difference between the first read timestamp and the second time interval may be directly used as the timestamp for marking the previous frame of sensing data of the current frame of sensing data.

[0078] Therefore, on the one hand, since the timestamp used for the annotation of the previous frame of sensor data is recorded by the clock of the main control module, and the clock of the main control module is more accurate than the clock of the sensor module, the accuracy of the timestamp used for the annotation of the previous frame of sensor data can be improved; on the other hand, since the second time interval is the difference between the time point when the sensor module collects sample sensor data recorded by the clock of the main control module and the time point when the main control module reads the sample sensor data recorded by the clock of the main control module, the second time interval can accurately reflect the time difference between the time point when the sensor module collects the sample sensor data and the time point when the main control module reads the sample sensor data. By using the difference between the first reading timestamp and the second time interval as the timestamp used for the annotation of the previous frame of sensor data of the current frame of sensor data, the problem that the timestamp used for the annotation of the previous frame of sensor data is not the time point when the sensor module collects the sample sensor data can be avoided, thereby improving the accuracy of the timestamp used for the annotation of the previous frame of sensor data.

[0079] 203. Determine a timestamp for labeling the current frame of sensor data based on the timestamp for labeling the previous frame of sensor data and the first time interval.

[0080] The timestamp used for labeling the current frame sensor data is used to indicate the time point when the current frame sensor data is collected.

[0081] Exemplarily, the sum of the annotation usage timestamp of the previous frame of sensor data of the current frame of sensor data and the first time interval can be used as the annotation usage timestamp of the current frame of sensor data. For example, assuming that the annotation usage timestamp of the previous frame of sensor data of the current frame of sensor data is 17:00:30, and the first time interval Δt1 is 1 second, then the annotation usage timestamp of the current frame of sensor data = the annotation usage timestamp of the previous frame of sensor data of the current frame of sensor data + the first time interval = 17:00:30 + 1 second = 17:00:31. Similarly, refer to steps 201 to 203 to determine the annotation usage timestamp of the t-th frame of sensor data, that is, the annotation usage timestamp of the t-th frame of sensor data = the annotation usage timestamp of the t-1th frame of sensor data + the first time interval Δt1.

[0082] Therefore, the annotated usage timestamp of the current frame sensor data is determined according to the first time interval between the two frames of sensor data collected by the sensor module and the annotated usage timestamp of the previous frame sensor data of the current frame sensor data; on the one hand, by obtaining the first time interval between the two frames of sensor data collected by the sensor module, the actual collection interval can be used as the first time interval; thereby combining the first time interval and the annotated usage timestamp of the previous frame sensor data to determine the annotated usage timestamp of the current frame sensor data, when the interval between the two frames of sensor data collected by the sensor module and the interval between the two frames of sensor data read by the main control module are inconsistent, the annotated usage timestamp of the current frame sensor data can be made closer to the true value of its collection time point, thereby reducing the error between the annotated usage timestamp of the sensor data of the wearable device and the actual usage timestamp of the sensor data of the wearable device, thereby reducing the error accumulation of the calculation scenario that needs to rely on the annotated usage timestamp, and improving the accuracy of the sensor data timestamp.

[0083] In order to avoid the problem that the sensor module clock may not be accurate, if the sensor data collection time recorded by the sensor module clock is marked as the sensor data usage time, there will still be a discrepancy between the usage timestamp marked by the wearable device's sensor data and the actual usage timestamp of the wearable device's sensor data. Figure 4 , the usage time of the current frame sensor data can also be determined by referring to steps 401 to 403, wherein:

[0084] 401. Acquire a second time interval between when a sensor module collects sample sensor data and when a main control module reads the sample sensor data.

[0085] The second time interval is the time difference between when the sensor module collects the sample sensing data and when the main control module reads the sample sensing data.

[0086] There are many ways to "obtain the second time interval between the sensor module collecting the sample sensor data and the main control module reading the sample sensor data" in step 401, illustratively including:

[0087] (1) In some embodiments (for example, when the clock of the sensor module is more accurate than the clock of the main control module), the clock of the sensor module can be used to record the time point when the sensor module collects sample sensor data and the time point when the main control module reads the sample sensor data. Then, the difference between the time point when the sensor module collects sample sensor data recorded by the clock of the sensor module and the time point when the main control module reads the sample sensor data recorded by the clock of the sensor module is used as the second time interval. For example, if the clock of the sensor module records that the sensor module collects sample sensor data at 20:30:30, and the clock of the sensor module records that the main control module reads sample sensor data at 20:30:31, then the second time interval is 1 second.

[0088] (2) In some other embodiments (for example, when the clock of the main control module is more accurate than the clock of the sensor module), the clock of the main control module can be used to record the time point when the sensor module collects the sample sensor data and the time point when the main control module reads the sample sensor data. Then, the difference between the time point when the sensor module collects the sample sensor data recorded by the clock of the main control module and the time point when the main control module reads the sample sensor data recorded by the clock of the main control module is used as the second time interval. For example, if the clock of the main control module records that the sensor module collects the sample sensor data at 10:00:00, and the clock of the main control module records that the main control module reads the sample sensor data at 10:00:01, then the second time interval is 1 second.

[0089] 402. Obtain a second reading timestamp when the main control module reads the current frame of sensing data from the sensor module.

[0090] The second read timestamp is the read time recorded by the clock of the main control module when the main control module reads the current frame sensor data. In step 402, when the main control module reads the current frame sensor data from the sensor module, the time point of the clock of the main control module at this moment is used as the second read timestamp.

[0091] 403. Determine a timestamp for labeling the current frame of sensing data based on the second reading timestamp and the second time interval.

[0092] Specifically, the difference between the second read timestamp and the second time interval (denoted as Δt2) can be directly used as the timestamp for marking the current frame sensor data. For example, the main control module reads the current frame sensor data from the sensor module at 10:00:01, and the second time interval Δt2 is 1 second, then the timestamp for marking the current frame sensor data is 10:00:02. Similarly, the timestamp for marking the t-th frame sensor data can be determined with reference to steps 401 to 403, that is: the timestamp for marking the t-th frame sensor data = the second read timestamp T when the main control module reads the t-th frame sensor data from the sensor module t - A second time interval Δt2.

[0093] Therefore, on the one hand, since the clock of the main control module is used to record the timestamp of the annotation of the current frame sensor data, it is avoided to directly use the acquisition time recorded by the clock of the sensor as the timestamp of the annotation of the current frame sensor data, thereby avoiding the problem of inaccurate timestamp of the annotation of the current frame sensor data due to the inaccuracy of the clock of the sensor module, so that the accuracy of the timestamp of the annotation of the previous frame sensor data can be improved. On the other hand, by using the difference between the second read timestamp and the second time interval as the timestamp of the annotation of the current frame sensor data, it is avoided to directly use the second read timestamp of the current frame sensor data read from the sensor module by the main control module as the timestamp of the annotation of the current frame sensor data, thereby avoiding the problem that the timestamp of the annotation of the current frame sensor data is not the time point when the sensor module collects the sample sensor data, thereby improving the accuracy of the timestamp of the annotation of the current frame sensor data; reducing the error between the timestamp of the sensor data of the wearable device and the actual timestamp of the sensor data of the wearable device, thereby reducing the error accumulation of the scene that needs to rely on the timestamp of the annotation.

[0094] Further, after determining the timestamp for labeling the current frame of sensor data by using steps 401 to 403, the timestamp for labeling the next frame of sensor data (recorded as the t+1th frame of sensor data) of the current frame of sensor data can also be determined by referring to the following steps B1 to B2:

[0095] B1. Obtaining the first time interval for the sensor module to collect two frames of sensing data.

[0096] In some embodiments, in step B1, the current sampling rate of the sensor module can be obtained; based on the current sampling rate of the sensor module, the first time interval for the sensor module to collect two frames of sensing data is determined. The detailed implementation process is similar to steps 2011A to 2012A above, and the details can be referred to the relevant description above, which will not be repeated here.

[0097] In some other embodiments, in step B1, the number of sensor data frames collected by the sensor module within the statistical period can be obtained; based on the number of sensor data frames, the average interval between every two frames of sensor data of the sensor module is determined as the first time interval. The detailed implementation process is similar to steps 2011B to 2012B above, and the details can be referred to the relevant description above, which will not be repeated here.

[0098] B2. Determine a timestamp for labeling a next frame of sensor data of the current frame of sensor data based on the timestamp for labeling the current frame of sensor data and the first time interval.

[0099] The annotation of the next frame of sensor data of the current frame of sensor data uses a timestamp to indicate the time point of collecting the next frame of sensor data of the current frame of sensor data.

[0100] Exemplarily, the sum of the timestamp used for the annotation of the current frame sensor data and the first time interval can be used as the timestamp used for the annotation of the next frame sensor data. For example, assuming that the timestamp used for the annotation of the current frame sensor data is 17:00:31, and the first time interval Δt1 is 1 second, then the timestamp used for the annotation of the next frame sensor data = the timestamp used for the annotation of the current frame sensor data + the first time interval = 17:00:31 + 1 second = 17:00:32. Similarly, after determining the timestamp used for the annotation of the t-th frame sensor data with reference to steps 401 to 403, the timestamp used for the annotation of the t+1-th frame sensor data can be determined with reference to steps B1 to B2, that is: the timestamp used for the annotation of the t+1-th frame sensor data = the timestamp used for the annotation of the t-th frame sensor data + the first time interval Δt1.

[0101] Therefore, the annotated usage timestamp of the next frame of sensor data is determined according to the first time interval between the two frames of sensor data collected by the sensor module and the annotated usage timestamp of the current frame of sensor data. On the one hand, by obtaining the first time interval between the two frames of sensor data collected by the sensor module, the actual collection interval can be used as the first time interval. The annotated usage timestamp of the next frame of sensor data is determined in combination with the first time interval and the annotated usage timestamp of the current frame of sensor data. When the interval between the two frames of sensor data collected by the sensor module and the interval between the two frames of sensor data read by the main control module are inconsistent, the annotated usage timestamp of the next frame of sensor data can be closer to the actual value of its collection time point.

[0102] Furthermore, after determining the timestamp used for the annotation of the current frame sensor data with reference to steps 201 to 203 above (or determining the timestamp used for the annotation of the current frame sensor data with reference to steps 401 to 403 above), the timestamp used for the annotation of the current frame sensor data can also be used for further processing, so as to avoid the problem of inaccurate processed data caused by directly using the reading time point of the current frame sensor data for further processing, thereby improving the accuracy of the further processed data. For example, the timestamp used for the annotation of the current frame sensor data can be used for further processing in the following situations (i) to (iii):

[0103] (a) The current frame sensor data and the annotation use timestamp of the current frame sensor data can be provided to the upper rendering module of the wearable device, so that the upper rendering module can display the status of the sensor module based on the current frame sensor data and the annotation use timestamp of the current frame sensor data.

[0104] For example, if the sensor module is an inertial sensor and the current frame sensing data is inertial sensor data (such as acceleration, etc.), the inertial sensor data and the annotation of the inertial sensor data can be provided to the upper rendering module of the wearable device using a timestamp, so as to display the status of the inertial sensor of the wearable device at each moment through the upper rendering module.

[0105] For another example, if the sensor module is a magnetometer and the current frame sensor data is magnetometer data (such as the angles in the four directions of east, south, west and north, etc.), the magnetometer data and the annotation of the magnetometer data can be provided to the upper rendering module of the wearable device using a timestamp, so that the status of the magnetometer of the wearable device at each moment can be displayed through the upper rendering module.

[0106] For another example, if the sensor module is a position sensor and the current frame sensing data is position sensor data, the position sensor data and the annotation of the position sensor can be provided to the upper rendering module of the wearable device using a timestamp, so that the status of the position sensor of the wearable device at each moment can be displayed through the upper rendering module.

[0107] (ii) The current frame sensor data and the timestamp used to annotate the current frame sensor data may also be provided to an application of the wearable device, so that the application may perform an integral operation based on the current frame sensor data and the timestamp used to annotate the current frame sensor data to obtain a state of the wearable device.

[0108] There are multiple ways to “provide the current frame sensor data and the timestamp of the annotation of the current frame sensor data to the application of the wearable device, so that the application performs an integral operation based on the current frame sensor data and the timestamp of the annotation of the current frame sensor data to obtain the state of the wearable device”, exemplarily including the following cases 1 to 3:

[0109] Case 1: The current frame sensing data includes the acceleration of the wearable device, and the state of the wearable device includes the displacement of the wearable device. At this time, the acceleration of the wearable device and the timestamp of the acceleration of the wearable device can be provided to the business application of the wearable device, so that the business application can perform an integral operation based on the acceleration of the wearable device and the timestamp of the acceleration of the wearable device to obtain the displacement of the wearable device. For example, taking the business application as a navigation application, the acceleration within a period of time (such as 1 minute) can be collected by the sensor module. The acceleration of the wearable device is collected 60 times within 1 minute. The acceleration of the wearable device collected each time can be used as a current frame sensing data. Refer to the determination method of "the timestamp of the current frame sensing data" in steps 201 to 203 above or refer to the determination method of "the timestamp of the current frame sensing data" in steps 301 to 303 above to determine the timestamp of the acceleration of the wearable device collected each time; thereby, 60 wearable devices can be obtained. The acceleration of the wearable device and the annotated use timestamp of the acceleration of 60 wearable devices are collected; finally, the acceleration of the 60 wearable devices and the annotated use timestamp of the acceleration of the 60 wearable devices are sent to the navigation application, and the navigation application uses the acceleration of the 60 wearable devices and the annotated use timestamp of the acceleration of the 60 wearable devices to perform integration operations, and the displacement of the wearable device within 1 minute can be obtained; therefore, since the accuracy of the annotated use timestamp of the acceleration of the wearable device is not directly set to the reading time of the acceleration of the wearable device, the recognition accuracy of the displacement of the wearable device can be improved.

[0110] Case 2: The current frame sensor data includes the magnetometer data of the wearable device, and the state of the wearable device includes the rotation angle of the wearable device. At this time, the magnetometer data of the wearable device and the timestamp of the magnetometer data of the wearable device can be provided to the business application of the wearable device, so that the business application can perform an integral operation based on the magnetometer data of the wearable device and the timestamp of the magnetometer data of the wearable device to obtain the rotation angle of the wearable device. For example, taking the business application as a compass application, the magnetometer data within a period of time (such as 1 minute) can be collected through the sensor module. A total of 120 magnetometer data are collected within 1 minute. Each collected magnetometer data can be used as a current frame sensor data. Refer to the determination method of "the timestamp of the current frame sensor data" in steps 201 to 203 above or refer to the determination method of "the timestamp of the current frame sensor data" in steps 301 to 303 above to determine the timestamp of the magnetometer data collected each time; thus, 1 can be obtained. 20 magnetometer data and 120 magnetometer data are annotated with timestamps; finally, the 120 magnetometer data and the annotated with timestamps are sent to the compass application, and the compass application uses the 120 magnetometer data and the annotated with timestamps to perform integration operations, and the rotation angle of the wearable device within 1 minute can be obtained; therefore, since the accuracy of the annotated with timestamps of the magnetometer data is not directly set to the reading time of the magnetometer data, the recognition accuracy of the rotation angle of the wearable device can be improved.

[0111] Case 3: The current frame sensing data includes the inertial sensor data of the wearable device, and the state of the wearable device includes the posture of the wearer of the wearable device. At this time, the inertial sensor data of the wearable device and the timestamp of the annotation of the inertial sensor data of the wearable device can be provided to the business application of the wearable device, so that the business application can perform an integral operation based on the inertial sensor data of the wearable device and the timestamp of the annotation of the inertial sensor data of the wearable device to obtain the posture of the wearer of the wearable device. For example, taking the business application as a game application, the inertial sensor data within a period of time (such as 30 seconds) can be collected through the sensor module. A total of 100 inertial sensor data are collected within 30 seconds. Each collected inertial sensor data can be used as a current frame sensor data. Refer to the determination method of "labeling the current frame sensor data using the timestamp" in steps 201 to 203 above or refer to the determination method of "labeling the current frame sensor data using the timestamp" in steps 301 to 303 above to determine the labeling use timestamp of each collected inertial sensor data; thereby, 100 inertial sensor data and 100 inertial sensor data labeling use timestamps can be obtained; finally, 1 100 inertial sensor data and 100 inertial sensor data are annotated with a timestamp and sent to a game application. The game application uses the 100 inertial sensor data and the annotated timestamp of the 100 inertial sensor data to perform integration operations, and can obtain the gesture movements of the wearer of the wearable device (such as a smart bracelet) within 30 seconds, so that the game application can further control further game operations according to the gesture movements of the wearer of the wearable device within 30 seconds. Therefore, since the accuracy of the annotated timestamp of the inertial sensor data is not directly set to the reading time of the inertial sensor data, the recognition accuracy of the wearer's posture can be improved, thereby improving the control accuracy of the game operation.

[0112] (iii) running the trained posture recognition model, and performing recognition based on the current frame sensor data and the annotation of the current frame sensor data using a timestamp to obtain the posture of the wearer of the wearable device. For example, the inertial sensor data within a period of time (such as 10 seconds) can be collected through the sensor module. A total of 50 inertial sensor data are collected within 10 seconds. Each collected inertial sensor data can be used as a current frame sensor data. The labeling and use timestamp of the inertial sensor data collected each time is determined by referring to the determination method of "labeling and use timestamp of current frame sensor data" in steps 201 to 203 above or referring to the determination method of "labeling and use timestamp of current frame sensor data" in steps 301 to 303 above; thus, 50 inertial sensor data and 50 labeling and use timestamps of inertial sensor data can be obtained; finally, the trained gesture recognition model is run, and the gesture movements of the wearer of the wearable device (such as a smart bracelet) within 10 seconds are recognized according to the 50 inertial sensor data and the labeling and use timestamps of 50 inertial sensor data; thus, since the accuracy of the labeling and use timestamp of the inertial sensor data is not directly set to the reading time of the inertial sensor data, the recognition accuracy of the wearer's gesture can be improved.

[0113] The trained posture recognition model is learned based on sample sensor data and the annotated posture of the sample sensor data, wherein the sample sensor data includes multiple frames of sensor data and corresponding annotated use timestamps. For example, an open source regression model, such as a recurrent neural network model (RNN), an index-based model, a similarity-based model, etc., can be used to learn the relationship between sensor data and the wearer's posture. For example, taking the training of the RNN model to obtain a trained posture recognition model as an example, the training process of the trained posture recognition model can be as follows: obtain training sample data, a training sample data includes sample sensor data and annotated posture corresponding to the sample sensor data, the sample sensor data includes multiple frames of sensor data, and the annotation of each frame in the multiple frames of sensor data uses a timestamp; use the sample sensor data and the annotated posture corresponding to the sample sensor data to fit the RNN model to learn, and the trained posture recognition model can be obtained.

[0114] In order to better understand the sensor data processing method in the embodiment of the present application, the specific process of sensor data processing is described below by taking "the wearable device is AR glasses, and the head rotation angle of the wearer of the AR glasses is detected to control the game operation" as an example. Figure 5 As shown, the specific process of the sensor data processing includes the following steps 501 to 504:

[0115] 501. Initialize the AR glasses system. Initialization includes: (1) powering on the hardware of the entire AR glasses system as required according to the timing requirements; (2) configuring the registers of the sensor module of the AR glasses, i.e., initializing the sensor module;

[0116] 502. Obtain the current data stream of the sensor module.

[0117] 503. Determine the timestamp used for the annotation of each frame of sensor data in the data stream. Specifically, for each frame of sensor data in the data stream, as current sensor data, determine the timestamp used for the annotation of each frame of sensor data in accordance with steps 201 to 203, or determine the timestamp used for the annotation of each frame of sensor data in accordance with steps 401 to 403.

[0118] 504. The data stream and the annotation of each frame of sensor data in the data stream are packaged with a timestamp and provided to the game application of the AR glasses, so that the game application of the AR glasses can determine the head rotation angle of the wearer of the AR glasses based on the data stream and each frame of sensor data in the data stream, and trigger corresponding game operations based on the head rotation angle of the wearer of the AR glasses.

[0119] In addition, in order to better implement the sensor data processing method in the embodiment of the present application, based on the sensor data processing method, a sensor data processing device is also provided in the embodiment of the present application, such as Figure 6 FIG. 1 is a schematic diagram of a structure of a sensor data processing device according to an embodiment of the present application. The sensor data processing device 600 includes:

[0120] The first acquisition unit 601 is used to acquire the first time interval between two frames of sensor data collected by the sensor module; the second acquisition unit 602 is used to acquire the annotated usage timestamp of the previous frame of sensor data of the current frame of sensor data; the processing unit 603 is used to determine the annotated usage timestamp of the current frame of sensor data based on the annotated usage timestamp of the previous frame of sensor data and the first time interval.

[0121] In some embodiments, the wearable device further includes a main control module, and the second acquisition unit 602 is specifically used to:

[0122] Acquire a second time interval between when the sensor module collects sample sensor data and when the main control module reads the sample sensor data; acquire a first reading timestamp when the main control module reads a previous frame of sensor data from the sensor module; and determine a timestamp for marking the previous frame of sensor data based on the first reading timestamp and the second time interval.

[0123] In some embodiments, the first acquiring unit 601 is specifically used to:

[0124] Acquire a current sampling rate of the sensor module; and determine a first time interval for the sensor module to collect two frames of sensing data based on the current sampling rate of the sensor module.

[0125] In some embodiments, the first acquiring unit 601 is specifically used to:

[0126] The number of sensor data frames collected by the sensor module within a statistical period is obtained; based on the number of sensor data frames, an average interval between every two frames of sensor data of the sensor module is determined as the first time interval.

[0127] In some embodiments, the processing unit 603 is specifically configured to:

[0128] An integration operation is performed based on the current frame sensing data and the annotation of the current frame sensing data using a timestamp to obtain the state of the wearable device.

[0129] In some embodiments, the processing unit 603 is specifically configured to:

[0130] Based on the current frame sensing data and the current frame sensing data, the state of the sensor module is displayed.

[0131] In some embodiments, the processing unit 603 is specifically configured to:

[0132] Run the trained posture recognition model, and perform recognition based on the current frame sensor data and the annotated usage timestamp of the current frame sensor data to obtain the wearer's posture of the wearable device, wherein the posture recognition model is learned based on sample sensor data and the annotated posture of the sample sensor data, and the sample sensor data includes multiple frames of sensor data and corresponding annotated usage timestamps.

[0133] In specific implementation, the above units can be implemented as independent entities, or can be arbitrarily combined and implemented as the same or several entities. The specific implementation of the above units can refer to the above sensor data processing method embodiment, which will not be repeated here.

[0134] A person skilled in the art will appreciate that all or part of the steps in the above-mentioned sensor data processing method may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0135] To this end, an embodiment of the present application provides a computer-readable storage medium, in which multiple computer programs are stored. The computer programs can be loaded by a processor to execute any sensor data processing method provided in the embodiment of the present application.

[0136] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0137] In the above-mentioned sensor data processing device, computer-readable storage medium, and wearable device embodiments, the description of each embodiment has its own emphasis. For parts not described in detail in a certain embodiment, reference can be made to the relevant description of other embodiments. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process and beneficial effects of the sensor data processing device, computer-readable storage medium, wearable device and its corresponding units described above can refer to the description of the sensor data processing method in the above embodiment, and will not be repeated here.

[0138] The above is a detailed introduction to a sensor data processing method, apparatus, wearable device and computer-readable storage medium provided in the embodiments of the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the ideas of the present application, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A sensor data processing method, characterized in that: The method is applied to a wearable device, the wearable device includes a sensor module, and the method includes: Acquire a first time interval for the sensor module to collect two frames of sensing data; Obtain the timestamp of the previous frame of sensor data of the current frame; The timestamp used for labeling the sensor data of the previous frame is determined based on the timestamp used for labeling the sensor data of the previous frame and the first time interval.

2. The sensor data processing method according to claim 1, characterized in that: The wearable device further includes a main control module, and the tagging of the previous frame of sensor data of the current frame of sensor data using a timestamp includes: Acquiring a second time interval between the sensor module collecting sample sensor data and the main control module reading the sample sensor data; Obtaining a first reading timestamp when the main control module reads the last frame of sensing data from the sensor module; Based on the first reading timestamp and the second time interval, a timestamp for labeling the previous frame of sensing data is determined.

3. The sensor data processing method according to claim 1, characterized in that: The acquiring sensor module collects a first time interval of two frames of sensing data, including: Get the current sampling rate of the sensor module; Based on the current sampling rate of the sensor module, a first time interval for the sensor module to collect two frames of sensing data is determined.

4. The sensor data processing method according to claim 1, characterized in that: The acquiring sensor module collects a first time interval of two frames of sensing data, including: Obtain the number of sensor data frames collected by the sensor module within the statistical period; Based on the number of sensing data frames, an average interval between every two frames of sensing data of the sensor module is determined as the first time interval.

5. The sensor data processing method according to claim 1, characterized in that: The method further comprises: An integration operation is performed based on the current frame sensing data and the annotation of the current frame sensing data using a timestamp to obtain the state of the wearable device.

6. The sensor data processing method according to claim 1, characterized in that: The method further comprises: The state of the sensor module is displayed based on the current frame sensor data and the annotation of the current frame sensor data using a timestamp.

7. The sensor data processing method according to claim 1, characterized in that: The method further comprises: Run the trained posture recognition model, and perform recognition based on the current frame sensor data and the annotated usage timestamp of the current frame sensor data to obtain the wearer's posture of the wearable device, wherein the posture recognition model is learned based on sample sensor data and the annotated posture of the sample sensor data, and the sample sensor data includes multiple frames of sensor data and corresponding annotated usage timestamps.

8. A sensor data processing device, characterized in that: The sensor data processing device comprises: A first acquisition unit, used for acquiring a first time interval for the sensor module to collect two frames of sensing data; A second acquisition unit, used for acquiring a timestamp of the annotation use of a previous frame of sensor data of a current frame of sensor data; The processing unit is used to determine the timestamp used for labeling the current frame of sensor data based on the timestamp used for labeling the previous frame of sensor data and the first time interval.

9. A wearable device, characterized in that: The method comprises a sensor module, a main control module and a memory, wherein a computer program is stored in the memory, and when the main control module calls the computer program in the memory, the sensor data processing method according to any one of claims 1 to 7 is executed.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and the computer program is loaded by a processor to execute the sensor data processing method according to any one of claims 1 to 7.

Images