Walking state data collection method, system, device and medium

Abstract

The present invention provides a walking state data collection method, system, device and medium. When the method is executed in a single collection phase, the method includes: determining the corresponding current standard pace template based on the initial movement information of the current collection phase; obtaining each collected walking state data to form a current walking state data stream, and detecting whether the current walking state data stream includes switching identification information. If so, executing the next collection phase; if not, continuing to execute the current collection phase; and simultaneously, based on the current walking state data stream and the current standard pace template, using a sliding window method to perform a pace anomaly detection process to obtain valid walking data matching the standard pace template, thereby quickly and efficiently obtaining valid walking data segments in each walking state data stream.

Description

Technical Field

[0001] The present invention belongs to the field of indoor positioning technology, and more specifically relates to a walking state data collection method, system, device and medium. Background Art

[0002] At present, the collection of pedestrian gait status information is an important part of indoor positioning technology. Before executing data collection, it is usually necessary to pre-set corresponding walking status collection strategies and plans based on the task requirements of data collection, targeting different motion scenarios (such as flat ground, stairs, slopes, etc.), user characteristics (such as gender, height, weight, etc.), and motion postures (such as walking or running). The preset collection strategies are executed by users carrying wearable collection devices (such as inertial sensors) to realize the collection and acquisition of walking status data.

[0003] However, in the actual data collection process, it is inevitable that abnormal walking data will be present in the collected walking status data due to factors such as user-related reasons, external environmental interference, and unstable collection equipment. For example, when collecting the user's running status, the collected motion status data may be abnormal due to interference from external personnel. These abnormal walking data are often difficult to detect and discover directly during the data collection process, resulting in testers being unable to adjust data collection strategies and calibrate equipment in a timely manner based on the abnormal data collection situation. As a result, testers directly use the collected data and need to perform anomaly detection to extract and eliminate abnormal walking data. This leads to problems such as low real-time performance and low data collection efficiency of the collected data, which cannot meet the requirements of real-time data utilization. Some important data also needs to be re-collected, resulting in increased data collection workload and equipment loss.

[0004] Therefore, how to extract effective walking data during the walking state collection process has become a technical problem that needs to be solved urgently in this field. Summary of the Invention

[0005] In view of the above shortcomings in the prior art, the purpose of the present invention is to provide a walking state data collection method, system, device and medium, which is used to solve the problems such as the inability to directly obtain valid walking data matching the standard pace template in the existing walking state collection process, resulting in the inability to timely obtain the efficiency or abnormality rate of data collection.

[0006] To achieve the above-mentioned purpose and other related purposes, the present invention provides a walking state data collection method in a first aspect. When executed in a single collection stage, the method includes: obtaining the initial movement information of the current collection stage, and determining the corresponding standard pace template based on the initial movement information as the current standard pace template; obtaining the walking state data collected at each collection moment to form a current walking state data stream, and detecting whether the current walking state data stream includes switching identification information. If so, the next collection stage is executed; if not, the current collection stage is continued; at the same time, based on the current walking state data stream and the current standard pace template, a sliding window method is used to perform a pace anomaly detection process to obtain valid walking data that matches the standard pace template in the current walking state data stream.

[0007] In one embodiment of the present invention, the implementation method of detecting whether the current walking state data stream includes switching identification information includes: using a peak detection method to detect whether each walking state data contains stomping information; wherein the stomping information is a step state feature corresponding to the stomping action.

[0008] In one embodiment of the present invention, the walking state data includes walking acceleration, and the walking acceleration includes x-axis acceleration and z-axis acceleration; the peak detection method is used to detect whether each walking state data contains stomping information, including: for each walking acceleration data, fusing the x-axis acceleration and the z-axis acceleration to obtain a fused acceleration corresponding to each walking acceleration data; performing peak detection on each fused acceleration, including: obtaining a first peak difference between the current fused acceleration and its previous fused acceleration, and a second peak difference between the current fused acceleration and its next fused acceleration, and detecting whether the first peak difference and the second peak difference are both greater than a preset peak difference threshold; if so, determining that the walking state data corresponding to the current fused acceleration contains the stomping information.

[0009] In one embodiment of the present invention, the gait anomaly detection, when executed once, includes: obtaining a current sliding window, and based on the current sliding window, obtaining a current data segment to be detected in the current walking state data stream; performing normalization processing on each walking state data in the current data segment to be detected to obtain the normalized current data segment to be detected; based on the current standard gait template, using a dynamic time warping algorithm to match the normalized current data segment to be detected with each standard gait template to obtain a current minimum matching difference and a current minimum matching timestamp corresponding to the current minimum matching difference; detecting whether the current minimum matching difference is less than a matching difference threshold, and if not, executing subsequent steps; if so, obtaining a time interval between the current minimum matching timestamp and a previous minimum matching timestamp, and detecting whether the time interval is less than a time interval threshold; if so, marking each walking state data between the current minimum matching timestamp and the previous minimum matching timestamp as valid walking data; updating the sliding window, and returning to execute the above steps.

[0010] In one embodiment of the present invention, the walking state data includes acceleration data and gyroscope data, and the dynamic time warping algorithm is used to match the standardized current data segment to be detected and each standard step template to obtain the current minimum matching difference, including: based on each acceleration data in the current data segment to be detected and each sampling data in the standard step template, using the Euclidean distance method to construct an acceleration distance matrix, and based on each gyroscope data in the current data segment to be detected and each sampling data in the standard step template, using the Euclidean distance method to construct a gyroscope distance matrix; based on the acceleration distance matrix and the gyroscope distance matrix, obtaining a total matching distance matrix; and, based on the total matching distance matrix, using a dynamic programming method to obtain a distance accumulation matrix; determining the minimum matching distance value in the distance accumulation matrix as the minimum matching difference in the current data segment to be detected.

[0011] In one embodiment of the present invention, when executing the detection of whether the current minimum matching difference is less than the matching difference threshold, it also includes: when detecting that the current minimum matching difference is greater than or equal to the matching difference threshold, determining that the current data segment to be detected includes walking abnormality data, and performing zero-crossing rate abnormality detection on the current data segment to be detected.

[0012] In one embodiment of the present invention, before the normalization processing is performed on each walking state data in the current data segment to be detected, the method further includes: performing filtering and noise reduction processing on each walking state data in the current data segment to be detected.

[0013] Secondly, the present invention provides a walking state data collection method in a second aspect, including: a standard pace template acquisition module, which is used to obtain the initial movement information of the current collection stage when executing the current collection stage, and determine the corresponding standard pace template based on the initial movement information to serve as the current standard pace template; a walking state data collection module, which is used to collect walking state data at each moment of the current collection stage to form a current walking state data stream; a pace anomaly detection module, which is used to perform a pace anomaly detection process based on the current walking state data stream and the current standard pace template using a sliding window method to obtain valid walking data matching the standard pace template in the current walking state data stream; a switching identifier detection module, which is used to detect whether the current walking state data stream includes switching identifier information; if so, enabling the standard pace template acquisition module, the walking state data collection module and the pace anomaly detection module to execute the next collection stage; if not, continuing to execute the current collection stage.

[0014] In a third aspect, the present invention further provides an electronic device comprising: a memory, a processor, and a communicator; wherein the memory is used to store computer instructions; the processor executes the computer instructions to implement any of the above-described walking status data collection methods; and the communicator is used to connect to one or more walking status data collection devices.

[0015] In a fourth aspect, the present invention further provides a computer storage medium, wherein the computer storage medium stores a computer program, and when the computer program is executed by a processor, any of the above-mentioned walking state data collection methods is implemented.

[0016] As described above, the walking state data collection method, system, device and medium provided by the present invention determine the current standard pace template of the current collection cycle by obtaining initial movement information; and while collecting the walking state data, detect whether there is switching identification information in each walking state data, and determine whether to switch the collection stage according to the corresponding detection result, thereby realizing an uninterrupted walking state data collection process; and, while collecting data, perform walking anomaly detection on the collected walking state data stream, so that the walking valid data segments in each walking state data stream can be quickly and efficiently identified, and whether there is walking abnormality data in the collected walking state data can be quickly, conveniently and in real time known, and then the data collection strategy and equipment parameters can be timely adjusted according to the walking abnormality data, thereby improving the collection efficiency of the walking state data. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 Schematic diagram of the distribution of the x-axis, y-axis, and z-axis of the walking motion space coordinate system in one embodiment of the present invention;

[0018] Figure 2 A schematic flow chart of the walking state data collection method provided by the present invention in one embodiment;

[0019] Figure 3 A schematic diagram of the distribution of walking acceleration templates in a standard pace template when the initial motion information is walking on flat ground in one embodiment of the present invention;

[0020] Figure 4 A schematic diagram of the distribution of walking gyroscope templates in a standard step template when the initial movement information is walking on flat ground in one embodiment of the present invention;

[0021] Figure 5 A schematic diagram of the distribution of walking acceleration templates in a standard pace template when the initial motion information is flat ground running in one embodiment of the present invention;

[0022] Figure 6 A schematic diagram of the distribution of walking gyroscope templates in a standard pace template when the initial motion information is flat ground running in one embodiment of the present invention;

[0023] Figure 7 A schematic diagram of the process of performing a single execution of the abnormal step detection according to one embodiment of the present invention;

[0024] Figure 8 A schematic structural diagram of the walking status data acquisition system provided by the present invention in one embodiment. DETAILED DESCRIPTION

[0025] The following describes the embodiments of the present invention through specific examples. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments. The details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other unless they conflict.

[0026] It should be noted that the illustrations provided in the following embodiments are merely schematic illustrations of the basic concept of the present invention. Therefore, the illustrations only show components related to the present invention and are not drawn according to the number, shape, and size of components in actual implementation. In actual implementation, the type, quantity, and proportion of each component may be changed arbitrarily, and the component layout may also be more complex.

[0027] To solve the technical problems existing in the prior art, the present invention provides a walking state data collection method in one embodiment, which is suitable for performing an uninterrupted collection process of walking state data during walking movement, and is used to obtain valid walking data that matches a standard step template in each walking state data stream while collecting and obtaining each walking state data stream.

[0028] In this embodiment, the walking state data is motion state information collected by a collection device (such as an inertial sensor) during walking motion, including walking acceleration data and walking gyroscope data; the walking acceleration data includes x-axis acceleration, y-axis acceleration, and z-axis acceleration; the walking gyroscope data includes x-axis gyroscope data, y-axis gyroscope data, and z-axis gyroscope data;

[0029] See also Figure 1 , which is a schematic diagram showing the distribution of the x-axis, y-axis and z-axis in the walking motion space coordinate system; Figure 1 As shown, the x-axis is the same as the forward and backward direction of walking and parallel to the ground; the y-axis is perpendicular to the forward and backward direction of walking and parallel to the ground; and the z-axis is perpendicular to the ground.

[0030] The uninterrupted collection process is to execute the walking state data collection process at different stages without re-adjusting the motion state collection device. The collection stage can be switched, and a continuous walking state data stream can be obtained.

[0031] Exemplarily, the uninterrupted collection process is that the test person wears an acceleration sensor, performs walking movement in a walking posture in the first collection stage, and performs walking movement in a running posture in the second collection stage; during the test person's movement, the acceleration sensor continuously and uninterruptedly performs the collection of acceleration data.

[0032] See also Figure 2 , which is a flow chart of the walking state data collection method in this embodiment.

[0033] like Figure 2 As shown, the method, when executed in a single acquisition phase, comprises the following steps:

[0034] S100, obtaining initial movement information of the current acquisition phase, and determining a standard pace template corresponding to the current acquisition phase based on the initial movement information as the current standard pace template;

[0035] The initial movement information includes at least movement posture information and movement environment information of walking movement; the movement posture information includes walking, running, jumping or other walking movement postures; the movement environment information includes flat land, sloped land, etc.

[0036] Optionally, the initial exercise information also includes physiological characteristic information of the test person, such as height, weight, etc.

[0037] The standard gait template includes distribution characteristics of walking state data within a single gait phase, and is used for changes in walking state characteristics within a single gait phase;

[0038] Specifically, the standard step template includes the distribution characteristics of walking acceleration data and the distribution characteristics of walking gyroscope data in a single step phase; that is, the standard step template includes the walking acceleration template and the walking gyroscope template.

[0039] For example, when the initial movement information is walking on flat ground, the walking acceleration template in the standard pace template is distributed as follows: Figure 3 As shown, the walking gyroscope template distribution in its standard step template is as follows Figure 4 As shown; when the initial movement information is flat ground running, the walking acceleration template distribution in its standard pace template is as follows Figure 5 As shown, the walking gyroscope template distribution in its standard step template is as follows Figure 6 shown.

[0040] S200, at each collection moment in the current collection phase, collecting walking state data to form a current walking state data stream, and detecting whether the collected current walking state data stream includes switching identification information. If so, executing the next collection phase; if not, continuing to execute this step; at the same time, based on the current walking state data stream and the current standard pace template, using a sliding window method to perform a pace anomaly detection process to obtain valid walking data in the current walking state data stream that matches the current standard pace template.

[0041] Specifically, walking state data is collected at each collection moment to form a current walking state data stream, which is:

[0042] D={acc xi ,acc yi ,acc zi ,gyro xi ,gyro yi ,gyro zi , t i}

[0043] Among them, D is the current walking state data stream, acc xi ,acc yi ,acc zi , are the x-axis acceleration, y-axis acceleration, and z-axis acceleration in the walking state data obtained at each acquisition moment; gyro xi ,gyro yi,gyro zi are the x-axis gyroscope data, y-axis gyroscope data, and z-axis gyroscope data in the walking state data acquired at each collection moment; i Indicates the collection time of each motion state data.

[0044] It is detected whether the current walking state data stream includes switching identification information; the switching identification information is a walking state feature used to identify a change in the movement initial information.

[0045] In this embodiment, stomping information is used as the switching identification information to indicate that the initial motion information has changed, such as the motion posture switches from walking to running; that is, when the test person changes from walking to running, he or she first performs a stomping action; the stomping information is the pace state feature corresponding to the stomping action, that is, when the test person performs the stomping action, the pace state feature collected by the collection device includes acceleration features.

[0046] The implementation method of detecting whether the collected motion state data includes the switching identification information includes: using a peak detection method to detect whether the collected walking state data includes stomping information.

[0047] Specifically, in order to further enhance the amplitude characteristics of the stomping signal and ensure the accuracy of the results detected by the peak detection method; for each of the collected walking acceleration data, the x-axis acceleration and the z-axis acceleration are fused to obtain each fused acceleration acc mod，l ,Right now:

[0048]

[0049] Among them, acc_mod l is the acceleration data after fusion; acc x，l is the acceleration of each x-axis; acc z，l is the z-axis acceleration; l is the sequence number of each data in the current data segment to be detected; N is the sliding window length.

[0050] Based on the fused acceleration, the fused acceleration corresponding to the walking state data obtained at each acquisition moment is obtained in the current walking state data stream, that is, the fused acceleration corresponding to each acquisition moment is obtained; and peak detection is performed on each fused acceleration, including:

[0051] Obtain a first peak difference between the current fused acceleration and its previous fused acceleration, and a second peak difference between the current fused acceleration and its next fused acceleration, and detect whether both the first peak difference and the second peak difference are greater than a preset peak difference threshold. If so, determine that the current fused acceleration is the stomping information, and thus determine that the walking state data corresponding to the current fused acceleration contains the stomping information; that is:

[0052]

[0053] Among them, acc_mod p acc_mod l The last fused acceleration is the fused acceleration collected at the last acquisition moment relative to the current acquisition moment; acc_mod q acc_mod l The next fused acceleration is the fused acceleration collected at the next acquisition moment relative to the current acquisition moment; peak_thr is the preset acceleration peak difference threshold. l The peak difference between the acceleration after the previous fusion and acc_mod l When the peak difference between the acceleration and the next fused acceleration is greater than the preset peak difference threshold peak_thr, the point is determined to be the peak point generated by the stomping.

[0054] When it is detected that the walking state data contains the stomping information, the process returns to step S100 and executes the next acquisition phase; when it is detected that the motion state data does not contain the stomping information, the process continues to execute this step to obtain the current walking state data stream; at the same time, based on the current walking state data stream and the current standard pace template corresponding to the current acquisition phase, the sliding window method is used to repeat the pace anomaly detection process.

[0055] In this embodiment, if Figure 7 As shown, the abnormal step detection includes the following sub-steps when executed once:

[0056] S301, obtaining a current sliding window, and based on the current sliding window, obtaining an undetected walking state data segment in the current walking state data stream as a current data segment to be detected;

[0057] Specifically, a sliding window with a width of N and a sliding window step size of m is used to obtain a single walking state data segment from the undetected current walking state data stream as the current data segment to be detected; that is:

[0058] data win ={acc xi ,accyi ,acc zi ,gyro xi ,gyro yi ,gyro zi , t i}, i=1,2,...,N (3)

[0059] Among them, data win Represents the current data to be detected, which is a multi-dimensional data vector, including N motion state data; acc xi ,acc yi ,acc zi Respectively represent the x-axis acceleration data, y-axis acceleration data and single z-axis acceleration data in the single walking state data in the current data to be detected; gyro xi ,gyro yi ,gyro zi Respectively represent the x-axis gyroscope data, y-axis gyroscope data, and z-axis gyroscope data in a single walking state data; t i Indicates the collection time of a single walking state data.

[0060] Optionally, N is the data length of 1 to 2 standard steps; m is 1 data length.

[0061] S302, performing normalization processing on each walking state data in the current data segment to be detected to obtain the normalized current data segment to be detected;

[0062] Specifically, each of the fused acceleration data and the y-axis gyroscope data in the current data segment to be detected is subjected to standardization processing, namely:

[0063]

[0064]

[0065]

[0066]

[0067] in, is the sample mean of the fused acceleration data in the sliding window; Sacc_mod is the sample mean of each acc_mod in the current sliding window i Unbiased estimate of the variance of gyro yi is the y-axis gyroscope data in the sliding window; Sgyro y For each gyro in the current sliding window yi An unbiased estimate of the variance of .

[0068] In another specific embodiment, the normalization processing of each walking state data in the current data segment to be detected may also be:

[0069] Each walking acceleration data and each walking gyroscope data in the current data segment to be detected is normalized respectively; the normalization process of the walking acceleration data and the normalization process of the walking gyroscope data are the same as the normalization process of the above-mentioned fused acceleration data, and will not be repeated here.

[0070] S303, based on the current standard pace template, using a dynamic time warping algorithm to match the standardized current data segment to be detected with each of the current standard pace templates to obtain a current minimum matching difference;

[0071] Wherein, a single standard pace template includes sampling data obtained by discretely sampling a plurality of standard pace signals and calculating the average value.

[0072] The minimum matching difference is the degree of similarity of features between the current data segment to be detected and the current standard step template when the features are matched; the greater the minimum matching difference, the greater the feature difference between the two, and vice versa.

[0073] Specifically, based on the normalized acceleration data in the current data segment to be detected and the sampled data in the current standard step template, the acceleration distance matrix d_acc is constructed using the Euclidean distance method. n，m ,for:

[0074]

[0075] Among them, acc_norm n For the nth fused acceleration data, tplt_acc_norm m is the mth fused acceleration data of the current standard step template, and M is the length of the current standard step template;

[0076] And, based on the standardized gyroscope data in the current data segment to be detected and the sampled data in the current standard step template, the gyroscope distance matrix d_gyro is constructed using the Euclidean distance method n，m ,for:

[0077]

[0078] Among them, gyro_norm n For the nth gyroscope data, tplt_gyro_norm m is the mth gyroscope data of the current standard step template, and M is the length of the current standard step template.

[0079] Based on the acceleration distance matrix d_acc n，m and the gyroscope distance matrix d_gyro n，m , get the total matching distance matrix D N，M for:

[0080] D N，M =d_acc N，M +d_gyro N，M (10)

[0081] Based on the total matching distance matrix, the distance accumulation matrix C is obtained according to the dynamic programming method. N，M ,for:

[0082]

[0083] Among them, C i，j is the cumulative distance matrix C N，M Each element in the sliding window, that is, the sequence from the starting point to the i-th data point {acc_norm0, acc_norm1, ..., acc_norm i-1} and the sequence from the starting point to the jth data point in the current standard step template {tplt0, tplt1, ..., tplt j-1 The minimum value of the sum of the distances between the matching points after matching the data points of i-1，j-1 Represents the element in row i-1 and column j-1; C i-1，j represents the element in row i-1 and column j, C i，j-1 represents the element in row i and column j-1; D n，m is the total matching distance matrix D N，M The element in row n and column m; min{C i-1，j-1 , C i-1， j, C i，j-1} is {C i-1，j-1 , C i-1，j , C i，j-1} minimum value in .

[0084] In the distance accumulation matrix C M，N The minimum matching distance value is determined as the minimum matching difference in the current data segment to be detected, and is used as the current minimum matching difference.

[0085] Specifically, in the distance accumulation matrix C M，N In the last row of , the minimum value in the row of values ​​is selected as the minimum matching difference value when the current data segment to be detected is matched with the current standard step template.

[0086] S304: Detect whether the current minimum matching difference is less than a preset matching difference threshold. If not, directly execute the subsequent steps. If so, extract the timestamp corresponding to the current minimum matching difference as the current minimum matching timestamp. Detect whether the time interval between the current minimum matching timestamp and the previous minimum matching timestamp is less than a preset time interval threshold. If so, mark each walking state data between the current minimum matching timestamp and the previous minimum matching timestamp as valid walking data. If not, do not mark it.

[0087] The matching difference threshold is a threshold value determined according to the initial motion information when executing step S100.

[0088] Specifically, detect whether the current minimum matching difference is less than a preset matching difference threshold; if not, skip the current step directly and execute the subsequent step S305; if so, it indicates that the current data segment to be detected includes a matching data segment that matches the current standard pace template, extract the walking state data corresponding to the current minimum matching difference in the data segment to be detected as the minimum matching data in the current data segment to be detected, and extract the acquisition time corresponding to the minimum matching data as the current minimum matching timestamp; obtain the minimum matching timestamp that was extracted before the current pace anomaly detection and the most recent time as the previous minimum matching timestamp; extract the time interval between the current minimum matching timestamp and the previous minimum matching timestamp, and detect whether the time length of the time interval is less than the preset time interval size. If so, the walking state data segment within the time interval is regarded as the data segment that matches the current standard pace template and is marked as valid walking data; if not, do not mark it.

[0089] S305, updating the sliding window based on the step size of the sliding window, and returning to step S301;

[0090] Specifically, according to the step size of the sliding window, the current sliding window is slid to generate a new sliding window, and the process returns to step S301, and steps S301 to S305 are re-executed until all walking state data in the current walking state data stream complete the above-mentioned abnormal pace detection process, thereby obtaining, in the current acquisition phase, identification information of each valid walking data segment that matches the current standard pace template in the collected current walking state data stream; based on the identification information of each valid walking data segment, identification information of each abnormal walking data segment in the current walking state data stream is obtained.

[0091] Furthermore, the walking state data collection method, when executing step S304, further includes:

[0092] When it is detected that the current minimum matching difference is greater than or equal to a preset matching difference threshold, it is determined that the current data segment to be detected includes walking abnormality data, and zero-crossing rate abnormality detection is performed on the current data segment to be detected;

[0093] The zero-crossing rate anomaly detection is a detection of whether there is a zero-crossing rate anomaly in the walking state data segment; the zero-crossing rate anomaly is an abnormal situation in which the walking state data is distorted due to the temperature drift of the gyroscope.

[0094] Specifically, based on the gyroscope data in the current data segment to be detected, the zero crossing rate ZCR of the current gyroscope data is calculated as:

[0095]

[0096] Among them, s t for gyro y The collected signal corresponding to the collection time t, s t-1 for gyro y The collected signal corresponding to the collection time t-1; N is the sliding window length, and the function II{A} is 1 when the parameter A is true, otherwise it is 0.

[0097] Detect whether the zero-crossing rate of the current gyroscope data is less than the set zero-crossing rate threshold thresh_dr. If so, determine that the current data segment to be detected includes the zero-crossing rate anomaly; if not, determine that the current data segment to be detected includes other walking anomaly types.

[0098] Optionally, to improve the performance of the walking state data collection method and improve the accuracy of detecting and obtaining valid walking data in the walking state data stream, the walking state data collection method, when executing step S302, further includes:

[0099] Performing filtering and noise reduction processing on each walking state data in the current data segment to be detected;

[0100] Specifically, calculating the mean of each acceleration data and the mean of each gyroscope data in the current data segment to be detected to obtain the acceleration mean and the gyroscope data mean respectively;

[0101] Based on the acceleration mean, each acceleration data in the current data segment to be detected is mean filtered, and based on the gyroscope data mean, each gyroscope data in the current data segment to be detected is mean filtered to obtain new acceleration data and new gyroscope data respectively, thereby obtaining a new current data segment to be detected.

[0102] For example, when N is 5, the mean filtering is performed on the gyroscope numbers in the current data segment to be detected, which is:

[0103]

[0104] Among them, gyro_y_lowpass k is the number of gyroscopes after mean filtering;

[0105] Similarly, the x-axis acceleration data and the z-axis acceleration data are mean filtered to obtain the new current detection data after mean filtering, which is:

[0106]

[0107] Among them, gyro_y_lowpass k is the new current data to be detected after mean filtering; acc_x_lowpass j is the x-axis acceleration data after mean filtering; acc_z_lowpass j is the acceleration data of each y-axis after mean filtering.

[0108] It should be noted that, in each subsequent step, the operation of each subsequent step is performed based on the new current data to be detected.

[0109] In order to solve the technical problems existing in the prior art, the present invention also provides a walking state data acquisition system in one embodiment, which is suitable for performing an uninterrupted acquisition process of walking state data during walking movement, and is used to obtain valid walking data that matches the current standard pace template in each walking state data stream while acquiring each walking state data stream.

[0110] See also Figure 8 , which is a schematic structural diagram of the walking state data acquisition system in this embodiment; Figure 8 As shown, the walking state data collection system 600 includes: a standard step template acquisition module 610, a walking state data collection module 620, a step abnormality detection module 630 and a switching mark detection module 640;

[0111] The standard step template acquisition module 610 is used to acquire initial movement information of the current acquisition phase when executing the current acquisition phase, and determine a corresponding standard step template based on the initial movement information to serve as the current standard step template;

[0112] The walking state data collection module 620 is used to collect walking state data at each moment of the current collection phase to form a current walking state data stream;

[0113] The pace anomaly detection module 630 is used to perform a pace anomaly detection process based on the current walking state data stream and the current standard pace template using a sliding window method to obtain valid walking data that matches the current standard pace template in the current walking state data stream.

[0114] The switching identifier detection module 640 is used to detect whether the current walking state data stream includes switching identifier information; if so, the standard step template acquisition module, the walking state data acquisition module and the step abnormality detection module are enabled to execute the next acquisition phase; if not, the current acquisition phase is continued.

[0115] In one embodiment, the abnormal pace detection module 630 includes the following submodules:

[0116] a current data segment to be detected submodule, configured to obtain a current sliding window, and to obtain an undetected walking state data segment in the current walking state data stream based on the current sliding window as the current data segment to be detected;

[0117] A standardization processing submodule is used to perform standardization processing on each walking state data in the current data segment to be detected to obtain the current data segment to be detected after standardization processing;

[0118] a minimum matching difference obtaining submodule, configured to match the standardized current data segment to be detected with each current standard pace template using a dynamic time warping algorithm based on the current standard pace template to obtain a current minimum matching difference;

[0119] Among them, the implementation method of using the dynamic time warping algorithm to match the standardized current data segment to be detected and each current standard pace template based on the current standard pace template to obtain the current minimum matching difference is the same as the implementation method in the above embodiment and will not be repeated here.

[0120] The walking valid data marking submodule is used to detect whether the current minimum matching difference is less than a preset matching difference threshold. If not, the sliding window update submodule (described later) is enabled to perform sliding window update to obtain a new sliding window after update, and other submodules are enabled to perform corresponding operations based on the new sliding window; if so, the minimum matching timestamp corresponding to the current minimum matching difference is obtained and stored as the current minimum matching timestamp according to the current minimum matching difference; the time interval between the current minimum matching timestamp and the previous minimum matching timestamp is extracted, and the time interval is detected to be less than a preset time interval threshold; if so, each walking state data between the current minimum matching timestamp and the previous minimum matching timestamp is marked as walking valid data; if not, no marking is performed.

[0121] The sliding window updating submodule is used to update the sliding window based on the sliding window step size to obtain an updated new sliding window, and after the sliding window is updated, enable other submodules to perform the operation processing corresponding to the submodule.

[0122] Optionally, the abnormal pace detection module further includes:

[0123] The filtering and denoising submodule is configured to perform filtering and denoising processing on each walking state data after the normalization processing submodule completes the normalization processing on each walking state data.

[0124] The specific implementation of the filtering and noise reduction process is the same as that in the above embodiment, and will not be repeated here.

[0125] In other embodiments, the walking state data collection system further includes:

[0126] The zero-crossing rate anomaly detection submodule is configured to perform corresponding operations in the walking valid data marking submodule, and when it is detected that the current minimum matching difference is greater than or equal to a preset matching difference threshold, determine that the current data segment to be detected includes walking abnormality data, and execute this submodule to perform zero-crossing rate anomaly detection on the current data segment to be detected to determine whether the current data segment to be detected includes a zero-crossing rate anomaly.

[0127] The implementation method of performing zero-crossing rate anomaly detection on the current data segment to be detected is the same as that in the above-mentioned embodiment, and will not be described in detail here.

[0128] To solve the technical problems existing in the prior art, the present invention also provides an electronic device in one embodiment, which is suitable for performing an uninterrupted collection process of walking state data during walking movement, and is used to obtain valid walking data that matches the current standard pace template in each walking state data stream while collecting and obtaining each walking state data stream.

[0129] The electronic device comprises: a processor, a communicator and a memory; wherein the memory is used to store computer instructions; the processor executes the computer instructions to implement the following Figure 1 The various steps of the walking state data collection method; the communicator is used to establish a connection with one or more collection devices to achieve interaction.

[0130] In some embodiments, the number of memories in the data terminal device can be one or more, the number of processors can be one or more, and the number of communicators can be one or more.

[0131] The memory may include random access memory (RAM) or non-volatile memory, such as at least one disk storage device. The memory stores an operating system and operating instructions, executable modules, or data structures, or subsets thereof, or extended sets thereof. The operating instructions may include various operating instructions for implementing various operations. The operating system may include various system programs for implementing various basic services and processing hardware-based tasks.

[0132] The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, and discrete hardware components.

[0133] The communicator is preferably a communication port.

[0134] In some specific applications, the various components of the electronic device are coupled together via a bus system, wherein the bus system includes not only a data bus but also a power bus, a control bus, and a status signal bus.

[0135] In one embodiment of the present invention, the present application further provides a computer-readable storage medium having a computer program stored thereon, which, when called by a processor, implements the following Figure 1 The various steps of the walking state data collection method.

[0136] Among them, a computer-readable storage medium can be a tangible device that can hold and store instructions used by an instruction execution device. The computer-readable storage medium can be, for example, (but not limited to) an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, and a mechanical encoding device.

[0137] The computer-readable program described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing / processing device.

[0138] The computer program instructions for performing the operations of the present application may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, integrated circuit configuration data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and procedural programming languages ​​such as "C" or similar programming languages.

[0139] The computer-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In some embodiments, the state information of the computer-readable program instructions is utilized to personalize an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), which can execute the computer-readable program instructions, thereby implementing various aspects of the present application.

[0140] In summary, the walking state data collection method, system, device, and medium provided by the present invention determine the current standard step template for the current collection cycle by obtaining initial motion information; and while collecting the walking state data, detect whether switching identification information exists in each walking state data, and determine whether to switch the collection phase based on the corresponding detection result, thereby achieving an uninterrupted walking state data collection process; and, while collecting data, perform walking anomaly detection on the collected walking state data stream, thereby quickly and efficiently identifying the valid walking data segments in each walking state data stream, and quickly, conveniently, and in real time determining whether there is walking anomaly data in the collected walking state data, and then timely adjusting the data collection strategy and device parameters based on the walking anomaly data, thereby improving the collection efficiency of the walking state data; in addition, by performing zero-crossing rate anomaly detection on the walking anomaly data, the walking anomaly type can be quickly determined, so that corresponding adjustments can be made based on the walking anomaly type, thereby further improving the collection efficiency of the walking state data.

[0141] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the present invention. Anyone skilled in the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by one of ordinary skill in the art without departing from the spirit and technical principles disclosed herein are intended to be covered by the claims of the present invention.

Claims

1. A walking state data collection method, characterized in that: When executed in a single acquisition phase, it includes: Acquiring initial movement information of the current acquisition phase, and determining a corresponding standard pace template based on the initial movement information to serve as the current standard pace template; Acquire walking state data collected at each collection moment to form a current walking state data stream; the walking state data includes walking acceleration data, and the walking acceleration data includes x-axis acceleration and z-axis acceleration; A peak detection method is used to detect whether each of the walking state data contains stomping information. If so, the next acquisition phase is executed; if not, the current acquisition phase is continued. At the same time, based on the current walking state data stream and the current standard step template, a sliding window method is used to perform a step anomaly detection process to obtain valid walking data that matches the current standard step template in the current walking state data stream; wherein, The peak detection method is used to detect whether each walking state data contains stomping information, including: For each of the walking acceleration data, the x-axis acceleration and the z-axis acceleration are fused to obtain each fused acceleration; a first peak difference between the current fused acceleration and its previous fused acceleration, and a second peak difference between the current fused acceleration and its next fused acceleration are obtained, and it is detected whether the first peak difference and the second peak difference are both greater than a preset peak difference threshold; if so, it is determined that the walking state data corresponding to the current fused acceleration contains the stomping information; the stomping information is a step state feature corresponding to the stomping action.

2. The walking state data collection method according to claim 1, characterized in that: The step anomaly detection, when executed once, includes: Obtaining a current sliding window, and obtaining a current data segment to be detected in the current walking state data stream based on the current sliding window; performing normalization processing on each walking state data in the current data segment to be detected to obtain the normalized current data segment to be detected; Based on the current standard pace template, a dynamic time warping algorithm is used to match the standardized current data segment to be detected with each current standard pace template to obtain a current minimum matching difference; Detect whether the current minimum matching difference is less than the matching difference threshold; if not, skip the current step and execute the subsequent step; if so, obtain the timestamp corresponding to the current minimum matching difference as the current minimum matching timestamp; detect whether the time interval between the current minimum matching timestamp and the previous minimum matching timestamp is less than the time interval threshold; if so, mark each walking state data between the current minimum matching timestamp and the previous minimum matching timestamp as valid walking data; Update the sliding window and return to execute the above steps.

3. The walking state data collection method according to claim 2, characterized in that: The walking state data includes walking acceleration data and walking gyroscope data, and the use of the dynamic time warping algorithm to match the standardized current data segment to be detected with each current standard step template to obtain the current minimum matching difference includes: Based on each walking acceleration data in the current data segment to be detected and each sampled data in the current standard step template, an acceleration distance matrix is ​​constructed using a Euclidean distance method, and based on each walking gyroscope data in the current data segment to be detected and each sampled data in the current standard step template, a gyroscope distance matrix is ​​constructed using a Euclidean distance method; Obtaining a total matching distance matrix based on the acceleration distance matrix and the gyroscope distance matrix; and obtaining a distance accumulation matrix based on the total matching distance matrix using a dynamic programming method; A minimum matching distance value is determined in the distance accumulation matrix as a current minimum matching difference in the current data segment to be detected.

4. The walking state data collection method according to claim 2, characterized in that: When executing the detection of whether the current minimum matching difference is less than the matching difference threshold, it also includes: when detecting that the current minimum matching difference is greater than or equal to the matching difference threshold, determining that the current data segment to be detected includes walking abnormality data, and performing zero-crossing rate abnormality detection on the current data segment to be detected.

5. The walking state data collection method according to claim 2, characterized in that: Before the normalization processing is performed on each walking state data in the current data segment to be detected, the method further includes: Perform filtering and noise reduction processing on each walking state data in the current data segment to be detected.

6. A walking state data acquisition system, characterized in that: include: a standard step template acquisition module, configured to acquire initial movement information of the current acquisition phase when executing the current acquisition phase, and determine a corresponding standard step template based on the initial movement information to serve as the current standard step template; a walking state data collection module, configured to collect walking state data at each moment in the current collection phase to form a current walking state data stream; The walking state data includes walking acceleration data, and the walking acceleration data includes x-axis acceleration and z-axis acceleration; a gait anomaly detection module configured to perform a gait anomaly detection process based on the current walking state data stream and the current standard gait template using a sliding window method to obtain valid walking data matching the current standard gait template in the current walking state data stream; a switching identification detection module, configured to detect whether each of the walking state data contains stomping information by using a peak detection method; If yes, the current standard step template acquisition module, the walking state data acquisition module and the step abnormality detection module are enabled to execute the next acquisition phase; if no, the current acquisition phase is continued; wherein, The peak detection method is used to detect whether each walking state data contains stomping information, including: For each of the walking acceleration data, the x-axis acceleration and the z-axis acceleration are fused to obtain each fused acceleration; a first peak difference between the current fused acceleration and its previous fused acceleration, and a second peak difference between the current fused acceleration and its next fused acceleration are obtained, and it is detected whether the first peak difference and the second peak difference are both greater than a preset peak difference threshold; if so, it is determined that the walking state data corresponding to the current fused acceleration contains the stomping information; the stomping information is a step state feature corresponding to the stomping action.

7. An electronic device, characterized in that: include: A memory, a processor and a communicator; wherein the memory is used to store computer instructions; the processor runs the computer instructions to implement the walking state data collection method as described in any one of claims 1 to 5; the communicator is used to connect to one or more walking state data collection devices.

8. A computer storage medium storing a computer program, wherein: When the computer program is executed by a processor, the walking state data collection method according to any one of claims 1 to 5 is implemented.

Images