Electronic device and method of processing motion signal

Abstract

The invention discloses an electronic device and a method for processing a motion signal. The method comprises the following steps: detecting by a radar to obtain a dynamic signal; performing continuous wavelet transform on the dynamic signal to obtain a wavelet magnitude map; segmenting the wavelet magnitude map to generate a plurality of samples; grouping the plurality of samples into a first cluster and a second cluster; sampling a motion signal from the dynamic signal according to the first cluster; and outputting the motion signal.

Description

TECHNICAL FIELD

[0001] The present application relates to a signal processing technique, and particularly relates to an electronic device and method for processing motion signals. BACKGROUND

[0002] Currently, medical personnel can use sensors to detect the motion behavior of a subject in a time period, and determine the health condition of the subject according to the detection result. For example, the medical personnel can determine whether the motion ability of the subject is poor according to the detection result, and then determine whether the subject suffers from sarcopenia. However, the conventional method can only evaluate the overall motion behavior of the subject in a time period, and cannot evaluate individual actions of the subject. Therefore, the detection result of the motion behavior generated by the conventional method is not very accurate and reliable. SUMMARY

[0003] The present application provides an electronic device and method for processing motion signals, which can extract motion signals related to the actions of a subject from radar signals.

[0004] One embodiment of a method for processing motion signals of the present application includes: detecting by a radar to obtain a dynamic signal; performing a continuous wavelet transform on the dynamic signal to obtain a wavelet magnitude graph; segmenting the wavelet magnitude graph to generate a plurality of samples; grouping the plurality of samples into a first cluster and a second cluster; sampling a motion signal from the dynamic signal according to the first cluster; and outputting the motion signal.

[0005] One embodiment of an electronic device for processing motion signals of the present application includes a transceiver and a processor. The transceiver is communicatively connected to a radar. The processor is coupled to the transceiver and configured to perform: detecting by the radar to obtain a dynamic signal; performing a continuous wavelet transform on the dynamic signal to obtain a wavelet magnitude graph; segmenting the wavelet magnitude graph to generate a plurality of samples; grouping the plurality of samples into a first cluster and a second cluster; sampling a motion signal from the dynamic signal according to the first cluster; and outputting the motion signal by the transceiver.

[0006] Based on the above, the electronic device of the present application can perform abnormal detection and noise filtering on the radar signal, and can generate an analysis result according to the processed radar signal. The analysis result generated by the method of the present application can accurately indicate the samples of the radar signal corresponding to each action of the subject. Users can evaluate individual actions of the subject based on the analysis result. BRIEF DESCRIPTION OF DRAWINGS

[0007] Figure 1 A schematic diagram of an electronic device for processing motion signals is shown for one embodiment of the present application;

[0008] Figure 2A schematic diagram illustrating the use of radar to detect the motion behavior of a subject according to an embodiment of the present invention is shown;

[0009] Figure 3 This is a schematic diagram illustrating the mapping relationship between the subject's actions and radar signals, as shown in one embodiment of the present invention;

[0010] Figure 4 A flowchart illustrating the sampling motion signal is shown for one embodiment of the present invention;

[0011] Figure 5 A schematic diagram illustrating the reconstructed dynamic signal is shown for one embodiment of the present invention;

[0012] Figure 6 A schematic diagram illustrating the conversion of a dynamic signal into a wavelet quantity graph is shown for one embodiment of the present invention;

[0013] Figure 7 A flowchart illustrating the generation of analysis results is shown for one embodiment of the present invention;

[0014] Figure 8 A schematic diagram illustrating motion signals and slope signals is shown for one embodiment of the present invention;

[0015] Figure 9 A flowchart illustrating the update analysis results is shown for one embodiment of the present invention;

[0016] Figure 10 A schematic diagram illustrating motion signal, slope signal, distance signal, and reference signal is shown for an embodiment of the present invention;

[0017] Figure 11 A schematic diagram illustrating the result of dynamic time warping according to an embodiment of the present invention;

[0018] Figure 12 A flowchart illustrating a method for processing motion signals is shown for one embodiment of the present invention.

[0019] Symbol Explanation

[0020] 100: Electronic devices

[0021] 110: Processor

[0022] 120: Storage Media

[0023] 130: Transceiver

[0024] 1040: Reference signal

[0025] 200: Radar

[0026] 300, 631, 632: Time windows

[0027] 310, 1030: Distance signal

[0028] 320: Speed ​​signal

[0029] 330:Acceleration signal

[0030] 41, 42, 43, 44, 45, 46, 47: Sampling signals

[0031] 510, 520: Simulation diagram

[0032] 610: Dynamic signal

[0033] 620: Wavelet Quantity Plot

[0034] 621, 622: Samples

[0035] 630, 810, 1010: Motion signals

[0036] 820, 1020: Slope signal

[0037] S121, S122, S123, S124, S125, S126, S401, S402, S403, S404, S405, S406, S701, S702, S703, S704, S901, S902, S903, S904: Steps Detailed Implementation

[0038] Figure 1 A schematic diagram of an electronic device 100 for processing motion signals is illustrated according to an embodiment of the present invention. The electronic device 100 may include a processor 110, a storage medium 120, and a transceiver 130.

[0039] Processor 110 may be, for example, a central processing unit (CPU), or other programmable general-purpose or special-purpose microcontroller (MCU), microprocessor, digital signal processor (DSP), programmable controller, application-specific integrated circuit (ASIC), graphics processing unit (GPU), image signal processor (ISP), image processing unit (IPU), arithmetic logic unit (ALU), complex programmable logic device (CPLD), field-programmable gate array (FPGA), or other similar elements or combinations thereof. Processor 110 may be coupled to storage medium 120 and transceiver 130, and access and execute multiple modules and various applications stored in storage medium 120.

[0040] Storage medium 120 may be any type of fixed or removable random access memory (RAM), read-only memory (ROM), flash memory, hard disk drive (HDD), solid state drive (SSD), or similar components or combinations thereof, for storing multiple modules or various applications that can be executed by processor 110.

[0041] Transceiver 130 transmits or receives signals wirelessly or via a wired connection. Transceiver 130 can also perform operations such as low-noise amplification, impedance matching, mixing, up- or down-frequency conversion, filtering, amplification, and similar functions. Processor 110 can communicate with a radar (not shown) via transceiver 130, such as a millimeter-wave (mmWave) radar or a frequency-modulated continuous wave (FMCW) radar. Processor 110 can receive radar detection results from the radar.

[0042] The electronic device 100 can detect specific actions performed by a subject using radar and generate analysis results. These results may include a mapping relationship between the subject's actions and samples of radar signals. For example, to determine if a subject suffers from sarcopenia, the user of the electronic device 100 can instruct the subject to repeatedly stand and sit, and use radar to detect the subject's movement behavior to generate analysis results. The analysis results may include a mapping relationship between the subject's standing actions and samples of radar signals, and may also include a mapping relationship between the subject's sitting actions and samples of radar signals.

[0043] Figure 2 A schematic diagram illustrating the detection of a subject's motion behavior using radar 200 is shown according to an embodiment of the present invention. Radar 200 can be configured to detect the subject's motion behavior, thereby generating dynamic signals or distance signals, such as velocity signals, acceleration signals, or Doppler signals. The definition of a Doppler signal is shown in equation (1), where x represents the distance in the X direction. y represents the velocity in the X direction, and y represents the distance in the Y direction. represents the velocity in the Y direction, z represents the distance in the Z direction, and This represents the velocity in the Z direction.

[0044]

[0045] In one embodiment, radar 200 is located at the origin of the coordinate system, and its boresight is the Y-axis. The subject sequentially performs five actions: sitting, standing up, standing, sitting back down, and sitting again. When the subject is sitting, the distance signal detected by radar 200 indicates that the distance between the subject and radar 200 (i.e., the distance on the Y-axis) is at its maximum. When the subject stands up, the velocity signal detected by radar 200 indicates that the subject's velocity is negative (i.e., the subject is moving towards the origin along the Y-axis). When the subject is standing, the distance signal detected by radar 200 indicates that the distance between the subject and radar 200 is at its minimum. When the subject sits back down, the velocity signal detected by radar 200 indicates that the subject's velocity is positive. When the subject is sitting, the distance signal detected by radar 200 indicates that the distance between the subject and radar 200 is at its maximum. The electronic device 100 can generate analysis results based on the detection results of radar signals, and the analysis results can indicate the radar signal sample corresponding to the subject's standing up or sitting back actions.

[0046] Figure 3A schematic diagram illustrating the mapping relationship between a subject's actions and radar signals is provided according to an embodiment of the present invention. The radar signals may include a distance signal 310, a velocity signal 320, or an acceleration signal 330 generated by the radar 200. Based on the analysis results generated by the radar signals, the electronic device 200 can indicate samples of the distance signal 310 or the velocity signal 320 corresponding to the subject's standing-up or sitting-back actions, such as... Figure 3 As shown. After five cycles of standing up and sitting back, the subjects remained seated during time window 300. Therefore, the analysis results indicate that multiple samples within time window 300 did not correspond to any action.

[0047] Figure 4 A flowchart illustrating the sampling of motion signals is shown according to an embodiment of the present invention, the flowchart being as follows: Figure 1 The illustrated electronic device 100 is implemented. In step S401, the processor 110 can detect the subject via the radar 200 to obtain dynamic signals, which may include velocity signals, acceleration signals, or Doppler signals. In one embodiment, the signals detected by the radar 200 may further include distance signals.

[0048] In one embodiment, radar 200 may be configured to increase the value of a dynamic signal (or distance signal) during a specific action performed by the subject, and decrease the value of the dynamic signal (or distance signal) during another action performed by the subject. For example, radar 200 may be configured to decrease the value of a speed signal (or distance signal) when the subject stands up, and increase the value of a speed signal (or distance signal) when the subject sits back down, such as... Figure 3 As shown.

[0049] In step S402, the processor 110 may perform preprocessing on the dynamic signal to filter out noise in the dynamic signal or extract important features of the dynamic signal.

[0050] In one embodiment, processor 110 can perform singular spectrum analysis (SSA) on a dynamic signal to reconstruct the dynamic signal. First, processor 110 can convert the time series of the dynamic signal into a matrix (e.g., a Hankel matrix), and then decompose this matrix using singular value decomposition (SVD) to obtain the singular values ​​and corresponding singular vectors representing the principal components of the sequence. The principal components may contain characteristics such as trends, periods, or random fluctuations of the dynamic signal. Processor 110 can select one or more singular vectors to decompose the time series into the sum of its components to obtain the reconstructed dynamic signal.

[0051] When reconstructing a dynamic signal, the more singular vectors the processor 110 selects, the closer the reconstructed dynamic signal is to the original data. Figure 5 A schematic diagram of a reconstructed dynamic signal is illustrated according to an embodiment of the present invention. Simulation 510 shows a dynamic signal reconstructed using 3 singular vectors, and simulation 520 shows a dynamic signal reconstructed using 100 singular vectors. If too few singular vectors are used, the reconstructed dynamic signal may only contain information about the more important features, lacking the details of the original data. When reconstructing dynamic signals, the user needs to weigh the importance of data processing accuracy against speed to determine the number of singular vectors to use.

[0052] Back Figure 4 In step S403, processor 110 may perform a continuous wavelet transform (CWT) on the dynamic signal to obtain a scalogram or wavelet coefficient power graph. In one embodiment, processor 110 may use a wavelet function including the cgau6 function to perform the continuous wavelet transform on the dynamic signal.

[0053] Figure 6A schematic diagram illustrating the conversion of a dynamic signal into a wavelet magnitude map is shown according to an embodiment of the present invention. Processor 110 performs continuous wavelet transformation on the dynamic signal 610 to generate a wavelet magnitude map 620. The horizontal axis of the wavelet magnitude map 620 represents time (or samples), and the vertical axis can represent frequency. The wavelet magnitude map 620 can use color depth to represent the intensity of samples at specific frequencies. Taking the wavelet magnitude map 620 as an example, the regions corresponding to samples 400 to 600 and frequencies 10 to 15 in the wavelet magnitude map 620 have darker colors. Therefore, processor 110 can determine that the dynamic signal has a higher intensity in these regions.

[0054] Back Figure 4 In step S404, the processor 110 can segment the wavelet value map in the time domain to generate multiple samples. Taking the wavelet value map 620 as an example, the processor 110 can segment the wavelet value map 620 in the time domain to generate multiple samples including sample 621 and sample 622.

[0055] In step S405, the processor 110 may cluster multiple samples to obtain two clusters, namely a first cluster and a second cluster. In one embodiment, the processor 110 may cluster multiple samples according to the k-means clustering algorithm.

[0056] The first cluster may contain one or more samples representing the subject's more active movements (e.g., samples when the subject repeatedly performs actions such as standing up or sitting back down), and the second cluster may contain one or more samples representing the subject's less active movements (e.g., samples when the subject is resting in a chair). Taking wavelet value graph 620 as an example, the processor 110 can assign the samples 621 representing the subject's more active movements to the first cluster and the samples 622 representing the subject's less active movements to the second cluster. The number of samples in the first cluster may be less than the number of samples in the second cluster. Therefore, after performing the clustering, the processor 110 can determine that the cluster containing fewer samples is the first cluster and the cluster containing more samples is the second cluster.

[0057] In step S406, the processor 110 samples motion signals from the dynamic signals based on the first cluster. These motion signals represent signals during periods when the subject's movement is relatively active. The processor 110 can output the motion signals via the transceiver 130 for user reference.

[0058] by Figure 6For example, processor 110 can obtain one or more consecutive time periods corresponding to one or more samples in the first cluster, and set the one or more consecutive time periods as time window 631. Processor 110 can sample motion signal 630 from dynamic signal 610 according to time window 631. On the other hand, processor 110 can obtain one or more time periods corresponding to one or more samples in the second cluster, and set the one or more time periods as time window 632, where time window 632 represents the period when the subject's motion behavior is less active.

[0059] After obtaining the motion signal representing the overall motion behavior of the subject during a specific time period, the processor 110 can further match the subject's various actions with the various sampled signals in the motion signal to generate analysis results. The analysis results can indicate the subject's actions corresponding to the various sampled signals in the motion signal. Figure 7 A flowchart illustrating the generation of analysis results is shown in one embodiment of the present invention, the flowchart being as follows: Figure 1 The electronic device 100 shown is implemented.

[0060] In step S701, the processor 110 can calculate the slope of the motion signal to generate a slope signal. For example... Figure 8 As shown, processor 110 can calculate the slope of motion signal 810 to generate slope signal 820. In one embodiment, processor 110 can normalize slope signal 820 so that the value of slope signal 820 is between 0 and 1.

[0061] In step S702, the processor 110 can determine that multiple sampled signals in the slope signal correspond to multiple actions of the subject, and generate analysis results. The processor 110 can output the analysis results through the transceiver 130 for user reference.

[0062] In one embodiment, the processor 110 can categorize multiple sampled signals into a first category corresponding to a first action (e.g., sitting back) and a second category corresponding to a second action (e.g., standing up). If each sampled point of the sampled signal is greater than a threshold, the processor 110 can determine that the sampled signal corresponds to the first category. If each sampled point of the sampled signal is less than or equal to the sampling threshold, the processor 110 can determine that the sampled signal corresponds to the second category.

[0063] Taking slope signal 820 as an example, assuming the motion signal is a Doppler signal, based on the rule that "distance and speed decrease when the subject stands up, and distance and speed increase when the subject sits back down," processor 110 can determine that the four sampled signals corresponding to time periods 81, 83, 84, and 86 are greater than the sampling threshold, respectively, and thus the four sampled signals correspond to the sitting-back action. On the other hand, processor 110 can determine that the two sampled signals corresponding to time periods 82 and 85 are less than or equal to the sampling threshold, respectively, and thus the two signals correspond to the standing-up action. Processor 110 can determine the action corresponding to each sampled signal in the slope signal and generate analysis results.

[0064] In one embodiment, the processor 110 can determine whether a sampled signal is valid based on the time period of the sampled signal. If the time period of the sampled signal is greater than a time threshold T1, the processor 110 can determine that the sampled signal is valid and generate an analysis result. If the time period of the sampled signal is less than or equal to the time threshold T1, the processor 110 can determine that the sampled signal is invalid. Generally, the time for a subject to perform one standing up and sitting back action is about 2 seconds. That is, the standing up action and the sitting back action each take about 1 second. Accordingly, the processor 110 can, for example, set the time threshold T1 to 0.066 seconds. If the time period of a sampled signal corresponding to the sitting back action is greater than 0.066 seconds, the processor 110 can determine that the sampled signal corresponding to the sitting back action is valid. Taking the slope signal 820 as an example, the processor 110 can determine that the sampled signal representing the sitting back action corresponding to the time period 81 is valid based on the time period 81 being greater than the time threshold T1.

[0065] In one embodiment, if there is a short interval between two sampled signals corresponding to the same action and a sampled signal corresponding to another action, the sampled signal may be due to noise. The processor 110 may combine this sampled signal with the two sampled signals to form a single sampled signal. Specifically, assume the slope signal includes a first sampled signal, a second sampled signal, and a third sampled signal, where the first and third sampled signals correspond to a classification of a first action, and the second sampled signal between the first and second sampled signals corresponds to a classification of a second action. The processor 110 may update the slope signal in response to the second sampled signal's duration being less than or equal to a time threshold T2 to combine the first, second, and third sampled signals into a single sampled signal to match the first action; the time threshold T2 may be 0.5 times the time threshold T1.

[0066] Taking slope signal 820 as an example, processor 110 can determine that the sampling signals corresponding to time periods 84 and 86 represent a sitting-back action, and the sampling signal corresponding to time period 85 represents a standing-up action. If time period 85 is less than or equal to the time threshold T2, processor 110 can determine that the sampling signal of time period 85 is affected by noise. Processor 110 can determine that time periods 84, 85, and 86 correspond to a single sitting-back action, and can update the analysis results based on the determination results.

[0067] In one embodiment, if the slope signal is a distance signal, a velocity signal, or a Doppler signal, the processor 110 can determine whether the orientation signal in the slope signal is valid based on the acceleration signal. Taking the sampled signal corresponding to time period 81 in the slope signal 820 as an example, the processor 110 can determine that the sampled signal corresponds to a sitting-back action based on the sampled signal being greater than a sampling threshold. Since the sitting-back action causes the subject to move away from the radar 200, the acceleration signal detected by the radar 200 should be positive. Accordingly, if the sampled signal of the acceleration signal in time period 81 is positive, it means that the sampled signal of the acceleration signal in time period 81 matches the sampled signal of the slope signal 820 in time period 81. The processor 110 can thus determine that the sampled signal of the slope signal 820 in time period 81 is valid. On the other hand, if the sampled signal of the acceleration signal in time period 81 is negative, it means that the sampled signal of the acceleration signal in time period 81 does not match the sampled signal of the slope signal 820 in time period 81. The processor 110 can thus determine that the slope signal 820 is invalid during time period 81.

[0068] Back Figure 7 To assess the health status of test subjects, they are typically asked to perform a fixed number of specific actions. If the number of sampled signals corresponding to the actions identified by the processor 110 based on the slope signal is greater than the aforementioned fixed number, it indicates that the slope signal may contain noise. To improve the accuracy of the analysis results, the processor 110 may execute steps S703 and S704 to update the analysis results.

[0069] Specifically, in step S703, the processor 110 can count the number of multiple sampled signals (or multiple actions) identified from the slope signal and determine whether the number is greater than a preset value. If the number is greater than the preset value, it means that some sampled signals are noise. Accordingly, the processor 110 can determine in step S704 that one or more sampled signals in the slope signal are invalid, and update the analysis results. If the number is less than or equal to the preset value, it means that these sampled signals should not be noise. Accordingly, the processor 110 may not update the analysis results.

[0070] Figure 9 A flowchart illustrating the updated analysis results is shown according to an embodiment of the present invention, the flowchart being as follows: Figure 1The electronic device 100 shown is implemented. In step S901, the processor 110 can detect the subject via radar to obtain a distance signal. In step S902, the processor 110 can normalize the slope signal and multiply the normalized slope signal by the distance signal to generate a reference signal.

[0071] Figure 10 A schematic diagram illustrating a motion signal 1010, a slope signal 1020, a distance signal 1030, and a reference signal 1040 is shown according to an embodiment of the present invention. After the processor 110 calculates the slope of the motion signal 1010 to obtain the slope signal 1020 and normalizes the slope signal 1020, the processor 110 can multiply the normalized slope signal 1020 by the distance signal 1030 to generate the reference signal 1040.

[0072] Processor 110 may extract one or more sampled signals from reference signal 1040. For example, processor 110 may extract the portion of reference signal 1040 that is greater than a sampling threshold as sampled signals, such as sampled signals 41, 42, 43, 44, 45, 46, or 47. The sampled signals extracted by processor 110 correspond to the subject's actions or noise.

[0073] Back Figure 9 In step S903, multiple similarities are calculated between multiple sampled signals of the reference signal, and the multiple sampled signals are divided into a first category and a second category according to the multiple similarities. Sampled signals in the first category are valid, and sampled signals in the second category are invalid. Valid sampled signals can be retained in the analysis results, while invalid sampled signals can be deleted from the analysis results.

[0074] Specifically, the processor 110 can perform dynamic time warping (DTW) on multiple sampled signals to obtain multiple similarities. Figure 11 According to an embodiment of the present invention, a schematic diagram is drawn showing the result of performing dynamic time warping on sampled signals 41, 42, 43, 44, 45, 46 and 47. Darker colors represent higher similarity, and lighter colors represent lower similarity.

[0075] After obtaining multiple similarity scores, the processor 110 can divide the multiple sampled signals into a first category and a second category based on the multiple similarity scores. The processor 110 can perform agglomerative hierarchical clustering on the multiple sampled signals based on the multiple similarity scores to ensure that the number of samples assigned to the first category reaches a preset value (e.g., five). Once the number of samples in the first category reaches the preset value, the processor 110 can stop executing the agglomerative hierarchical clustering. One or more sampled signals that have not yet been assigned to the first category can be assigned to the second category.

[0076] In step S904, the processor 110 can determine that one or more sampling signals in the first category are valid, and update the analysis results accordingly. Taking sampling signals 41, 42, 43, 44, 45, 46, and 47 as an example, after the aggregation-based hierarchical clustering method is completed, the number of sampling signals 42, 43, 44, 45, and 46 included in the first category has reached a preset value. Therefore, the processor 110 can assign sampling signals 41 and 47, which have not yet been assigned to the first category, to the second category. The processor 110 can determine that sampling signals 41 and 47 are invalid and delete them from the analysis results. The analysis results can retain only the relevant information of sampling signals 42, 43, 44, 45, and 46.

[0077] Figure 12 A flowchart illustrating a method for processing motion signals according to an embodiment of the present invention is shown. The method may be performed by, for example... Figure 1 The electronic device 100 shown is implemented. In step S121, a dynamic signal is acquired by radar detection. In step S122, a continuous wavelet transform is performed on the dynamic signal to obtain a wavelet magnitude map. In step S123, the wavelet magnitude map is segmented to generate multiple samples. In step S124, the multiple samples are grouped into a first cluster and a second cluster. In step S125, a motion signal is sampled from the dynamic signal according to the first cluster. In step S126, the motion signal is output through a transceiver.

[0078] In summary, the electronic device of this invention can acquire dynamic signals by detecting a moving subject using radar. The electronic device can then group multiple samples generated from the dynamic signals using methods such as singular spectrum analysis or continuous wavelet transform to determine the subject's actions corresponding to each sample. In other words, the electronic device can acquire motion signals corresponding to each action performed by the subject, thereby generating analysis results that record the mapping relationship between the motion signals and the subject's actions. To assess the subject's health, the subject is typically asked to perform a fixed number of specific actions. Accordingly, the electronic device can determine whether the signal corresponding to the specific action in the analysis results is valid based on the aforementioned fixed number, and can update the analysis results if the signal is determined to be invalid (e.g., the signal is caused by noise). Furthermore, the electronic device can also improve the analysis results based on distance and acceleration signals detected by radar, making the mapping relationship between the signals and actions in the analysis results more accurate.

Claims

1. A method for processing motion signals, comprising: Detection using radar to obtain dynamic signals; Perform continuous wavelet transform on the dynamic signal to obtain a wavelet magnitude map; The wavelet magnitude map is segmented to generate multiple samples; The multiple samples are grouped into a first cluster and a second cluster; Motion signals are sampled from the dynamic signals based on the first cluster; as well as The motion signal is output.

2. The method of claim 1, wherein sampling the motion signal from the dynamic signal according to the first cluster comprises: The time window is determined based on the first cluster; as well as The motion signal is sampled from the dynamic signal according to the time window.

3. The method of claim 1, wherein performing the continuous wavelet transform on the dynamic signal to obtain the wavelet magnitude map comprises: Perform singular spectrum analysis on the dynamic signal to reconstruct the dynamic signal; as well as The continuous wavelet transform is performed on the reconstructed dynamic signal to obtain the wavelet magnitude map.

4. The method of claim 1, further comprising: Calculate the slope of the motion signal to generate a slope signal; Determine that the first sampled signal in the slope signal corresponds to the first action to generate an analysis result; as well as Output the analysis results.

5. The method of claim 4, wherein determining that the first sampling signal in the slope signal corresponds to the first action to generate the analysis result comprises: In response to the first sampling signal having a time period greater than a time threshold, the first sampling signal is determined to be valid in order to generate the analysis result.

6. The method of claim 4, wherein the slope signal further comprises a second sampling signal and a third sampling signal, wherein the first sampling signal and the third sampling signal correspond to a first category, wherein the second sampling signal between the first sampling signal and the third sampling signal corresponds to a second category, wherein the method further comprises: In response to the second sampling signal having a time period less than or equal to a time threshold, the slope signal is updated to match the first sampling signal, the second sampling signal, and the third sampling signal with the first action.

7. The method of claim 6, further comprising: In response to each sampling point of the first sampling signal being greater than a sampling threshold, it is determined that the first sampling signal corresponds to the first category.

8. The method of claim 6, further comprising: In response to each sampling point of the second sampling signal being less than or equal to the sampling threshold, it is determined that the second sampling signal corresponds to the second category.

9. The method of claim 4, further comprising: The radar is used to detect and obtain distance signals. Normalize the slope signal; The normalized slope signal is multiplied by the distance signal to generate a reference signal; as well as The analysis results are updated based on the reference signal.

10. The method of claim 9, wherein updating the analysis result based on the reference signal comprises: Multiple sampled signals are extracted from the reference signal, and multiple similarities between the multiple sampled signals are calculated; The multiple sampled signals are divided into a first category and a second category based on the multiple similarities; and Determine that at least one sampled signal in the first category is valid to update the analysis results.

11. The method of claim 10, wherein calculating the plurality of similarities among the plurality of sampled signals comprises: Dynamic time warping is performed on the multiple sampled signals to obtain the multiple similarities.

12. The method of claim 10, wherein dividing the plurality of sampled signals into the first category and the second category based on the plurality of similarities comprises: Based on the multiple similarities, an aggregated hierarchical clustering method is performed on the multiple sampling signals to ensure that the number of samples assigned to the first category reaches a preset value.

13. The method of claim 9, further comprising: It is determined that the multiple sampled signals in the slope signal correspond to multiple actions respectively; as well as In response to the number of the plurality of actions exceeding a preset value, the analysis result is updated based on the reference signal.

14. The method of claim 4, further comprising: The radar is used to detect and obtain a distance signal, wherein the distance signal includes a second sample signal corresponding to the first sample signal; as well as In response to the second sampling signal being less than or equal to the sampling threshold, the first sampling signal is determined to be valid in order to generate the analysis result.

15. The method of claim 4, wherein the method further comprises: The radar is used to detect and obtain an acceleration signal, wherein the acceleration signal includes a second sample signal corresponding to the first sample signal; as well as In response to the second sampling signal being greater than the sampling threshold, the first sampling signal is determined to be valid in order to generate the analysis result.

16. The method of claim 1, wherein the wavelet function of the continuous wavelet transform includes the cgau6 function.

17. The method of claim 1, further comprising: The multiple samples are grouped according to the k-means algorithm.

18. The method of claim 1, wherein the dynamic signal includes one of the following: a velocity signal, an acceleration signal, and a Doppler signal.

19. The method of claim 1, further comprising: The radar is configured such that the value of the dynamic signal increases during the execution of the first action.

20. An electronic device for processing motion signals, comprising: Transceiver, communication connection to radar; as well as A processor, coupled to the transceiver, and configured to perform: Detection is performed using the radar to obtain dynamic signals; Perform continuous wavelet transform on the dynamic signal to obtain a wavelet magnitude map; The wavelet magnitude map is segmented to generate multiple samples; The multiple samples are grouped into a first cluster and a second cluster; Motion signals are sampled from the dynamic signals based on the first cluster; as well as The motion signal is output through the transceiver.

Images