A sports data detection method with real-time identity authentication and a sports bracelet

Abstract

The application provides a sports data detection method with real-time identity authentication and a sports bracelet. The method comprises the following steps: obtaining a first fingerprint feature of a target user before starting sports, and generating a reference fingerprint template; collecting and analyzing sports data within a preset time length calibration window, and constructing a sports feature template; continuously collecting a second fingerprint feature and physiological data of a wearer during sports, and matching and verifying the second fingerprint with the reference template; if the verification is inconsistent, comparing the physiological data with the sports feature template, and generating an identity consistency deviation; comparing the deviation with a preset fault tolerance threshold, determining whether the associated real-time sports data is valid or not, and executing a safety response if it is invalid. The technical solution provided by the application realizes continuous and highly reliable binding of user identity and sports data, effectively prevents cheating by replacing a runner, and ensures data authenticity.

Description

Technical Field

[0001] This invention relates to the field of smart wearable device technology, and in particular to a method for detecting motion data with real-time identity authentication and a sports bracelet. Background Technology

[0002] With the deep integration of the Internet of Things, biometrics, and health monitoring technologies, wearable electronic devices (such as smart fitness trackers) have gradually become key tools in fields such as sports and health management, medical monitoring, and sports event certification. These devices, by integrating biometric recognition, motion sensors, and wireless communication technologies, enable real-time monitoring and data collection of users' exercise status and physiological indicators, providing crucial support for personalized health management, telemedicine, and competitive sports certification.

[0003] In application scenarios where data authenticity is critical, a reliable fitness tracker's core functions extend beyond basic metrics like step tracking and heart rate monitoring. Its deeper technological mission lies in ensuring that collected exercise data (including activity intensity, duration, and energy consumption) is authentically, uniquely, and indisputably attributed to the wearer. To achieve this, current methods embed fingerprint verification points in the smart tracker's display or the phone's screen. The smart device randomly sends a verification command before the run; verification must be completed within a specified time (usually 30 seconds). Failure to complete the verification within the time limit or an error triggers a "suspected runner" alarm, which is then sent to the device.

[0004] However, current technology only verifies fingerprints, which may lead to situations where the substitute runner wears the device while the person being replaced is only responsible for completing the fingerprint verification. Summary of the Invention

[0005] In view of this, this application provides a method for detecting sports data with real-time identity authentication and a sports bracelet to solve the above problems.

[0006] Firstly, a method for detecting motion data with real-time identity authentication is provided for use in fitness trackers. This method includes: Obtain the first fingerprint feature of the target user and generate a baseline fingerprint template. The first fingerprint feature is the fingerprint feature obtained before the start of the movement. Within the first preset calibration window, the wearer's motion data is continuously collected and analyzed, and a motion feature template is constructed based on the motion data; The wearer's second fingerprint features and physiological data are continuously collected, and the second fingerprint features are matched and verified with the reference fingerprint template. The second fingerprint features are fingerprint features acquired during movement. When the fingerprint matching verification result is inconsistent, the physiological data and the motion feature template are compared and analyzed to generate the identity consistency deviation. Based on the comparison between the identity consistency deviation and the preset fault tolerance threshold, the real-time motion data associated with physiological data is determined to be valid or invalid data, and a preset safety response is executed when it is determined to be invalid data.

[0007] The above technical solution significantly enhances the authenticity and security of exercise data by constructing a dual, continuous identity verification system that integrates initial static fingerprint authentication with dynamic physiological feature comparison during exercise. First, an authoritative user identity benchmark is established using initial fingerprint registration, and the system learns from the user's personalized exercise data to form a dynamic feature template. During exercise, when fingerprint verification fails, the fitness tracker compares real-time physiological data with the pre-stored dynamic template to calculate a quantified "identity consistency deviation." The fitness tracker uses this deviation to determine the validity of the data and triggers corresponding security responses, thus forming a complete closed loop from identity establishment and dynamic modeling to anomaly detection and handling. This series of interconnected steps achieves a continuous and highly reliable binding between user identity and exercise data, effectively preventing cheating and ensuring data authenticity.

[0008] Optionally, motion feature templates can be constructed based on motion data, specifically including: Extract multiple consecutive first data components representing the user's dynamic posture characteristics from motion data; Based on each first data component, the motion data is divided into multiple different motion intensity level intervals; Separate a second data component representing the user's cardiovascular activity characteristics from the exercise data, corresponding to each exercise intensity level range; The average step frequency and average stride length are calculated based on the first data component, and a gait feature parameter set is established based on the average step frequency and average stride length. Based on the second data component, a set of heart rate characteristic parameters associated with each exercise intensity level interval is established. The set of heart rate characteristic parameters includes the heart rate stable interval, heart rate rise rate, and heart rate recovery rate. The gait feature parameter set and the heart rate feature parameter set are combined to generate a motion feature template.

[0009] The above technical solution creates a multi-dimensional, contextualized personal biometric model by linking a user's dynamic posture characteristics (gait) with cardiovascular activity characteristics (heart rate) and further subdividing them according to exercise intensity levels. This model is no longer a single static data set but reflects the user's unique physiological response patterns under different exercise loads. This makes the basis for identity verification richer and more unique, thereby improving the accuracy and robustness of identity recognition.

[0010] Optionally, based on each first data component, the motion data can be divided into multiple different motion intensity level intervals, specifically including: Perform a fast Fourier transform on each first data component to obtain frequency spectrum data corresponding to each first data component. Each frequency spectrum data contains multiple frequency amplitudes. In each frequency spectrum data, the maximum frequency amplitude is determined, and the frequency value corresponding to the maximum frequency amplitude is determined as the main motion frequency; Based on a preset frequency intensity mapping relationship, each first data component is divided into one of multiple motion intensity level intervals according to each main motion frequency.

[0011] The above technical solutions enable precise segmentation of the motion process, providing a foundation for constructing accurate feature templates. By using Fast Fourier Transform (FFT) to analyze the main motion frequency, this method can accurately quantify and identify the current motion intensity, avoiding errors caused by subjective judgment. This ensures that the physiological and gait data extracted in each intensity range have high consistency and comparability, thereby enhancing the accuracy of the entire certified fitness tracker.

[0012] Optionally, the average cadence and average stride length are calculated based on the first data component, and a gait feature parameter set is established based on the average cadence and average stride length, specifically including: A peak detection algorithm is performed on the first data component to obtain multiple consecutive signal peaks in the first data component. The average step frequency is calculated based on the time interval between the multiple consecutive signal peaks. The multiple consecutive signal peaks represent a single gait cycle. The original vertical acceleration signal is separated from the first data component; The original vertical acceleration signal is processed through a high-pass filter to generate a corrected vertical acceleration signal; Perform the first numerical integration operation on the corrected vertical acceleration signal to generate a vertical velocity signal; Perform a second numerical integration operation on the vertical velocity signal to generate a vertical displacement signal; In the vertical displacement signal, the maximum and minimum displacement values ​​of the target are determined, and the difference between the maximum and minimum displacement values ​​of the target is determined as the vertical amplitude. The average stride length is calculated from the average stride frequency and vertical amplitude using a preset biomechanical function. The calculated values ​​of average cadence and average stride are combined to form a gait feature parameter set.

[0013] The above technical solution establishes a reliable and difficult-to-imitate personal gait biometric identifier. This method, through signal processing techniques such as peak detection and quadratic integration of vertical acceleration signals, can accurately extract key parameters reflecting an individual's walking or running habits from raw sensor data. These parameters collectively constitute a unique dynamic biometric feature, further enhancing the uniqueness of the user identity template and the reliability of verification.

[0014] Optionally, based on the second data component, a set of heart rate characteristic parameters associated with each exercise intensity level interval is established, specifically including: Within each exercise intensity level range, the second data component is segmented through a second preset time window; Calculate the standard deviation of the second data component within each time window; The standard deviation value is compared with a preset first stable threshold to determine the first continuous time period in the second data component. During the first continuous time period, the standard deviation value is continuously lower than the first stable threshold. The first stable heart rate interval is determined based on the maximum and minimum values ​​of the second data component within the first continuous time period. Identify the moment when the intensity level increases and changes. At the moment when the intensity level increases and changes, the exercise intensity level interval changes from the first previous exercise intensity level interval to the first current exercise intensity level interval. The intensity level of the first current exercise intensity level interval is higher than the intensity level of the first previous exercise intensity level interval. The second data component within the time period between the point when the intensity level changes and the start point of the first continuous time period is identified as the heart rate climb data. The slope of the regression line was calculated using a linear regression algorithm on the heart rate rise data, and the slope of the regression line was determined as the heart rate rise rate. Identify the moment when the intensity level decreases and changes. At the moment when the intensity level decreases and changes, the exercise intensity level interval changes from the second previous exercise intensity level interval to the second current exercise intensity level interval. The intensity level of the second current exercise intensity level interval is lower than the intensity level of the second previous exercise intensity level interval. The time period from the point when the intensity level changes to the point when the standard deviation value is consistently below the first stable threshold is defined as the heart rate recovery data. The heart rate recovery data were analyzed using an exponential decay function to calculate the decay coefficient, which was then used as the heart rate recovery rate. The continuous period of time in which the standard deviation value is consistently lower than the first stable threshold is defined as the second continuous period. The second heart rate stability interval is determined based on the maximum and minimum values ​​of the second data component within the second continuous time period. The first stable heart rate interval, the second stable heart rate interval, the heart rate rise rate, and the heart rate recovery rate are collectively established as a set of heart rate characteristic parameters.

[0015] The above technical solution details how to quantify and establish a dynamic heart rate feature parameter set correlated with changes in exercise intensity. It captures the personalized response patterns of a user's cardiovascular fitness tracker to changes in exercise load, greatly enriching the dimensions of physiological characteristics. By precisely defining the stable heart rate range, rise rate, and heart rate recovery rate, this method focuses not only on the static heart rate value but also on the dynamic change process, thus constructing a more refined and comprehensive heart rate model. This model can more sensitively reflect an individual's physiological characteristics, making identity verification analysis more in-depth and accurate.

[0016] Optionally, the physiological data and motion feature templates are compared and analyzed to generate an identity consistency deviation score, specifically including: Determine the target exercise intensity level range to which the physiological data belong; Based on motion feature templates, extract target heart rate feature parameter sets and target gait feature parameter sets related to the target motion intensity level range; The physiological data is compared with the target heart rate feature parameter set to obtain the heart rate deviation. Physiological data are compared with the target gait feature parameter set to obtain gait deviation. The identity consistency deviation is calculated by using a preset weighted algorithm to calculate the heart rate deviation and gait deviation.

[0017] The above technical solution realizes a comprehensive and dynamic identity consistency assessment mechanism. By comparing real-time physiological data with template data (heart rate and gait) corresponding to the current exercise intensity and fusing them using a weighted algorithm, this method can comprehensively assess the matching degree between the current wearer and the original user, generating an intuitive deviation value, providing a clear and reliable basis for subsequent decision-making.

[0018] Optionally, physiological data can be compared with a target heart rate feature parameter set to obtain heart rate deviation, specifically including: The heart rate value in the physiological data is compared with the target heart rate stable interval in the target heart rate feature parameter set to obtain the interval deviation. Calculate the rate of change of heart rate in physiological data within a preset time window before the current moment, and compare the rate of change of heart rate with the target heart rate rise rate or target heart rate recovery rate in the target heart rate feature parameter set to obtain the deviation of the rate of change. Based on the interval deviation and the rate of change deviation, the heart rate deviation is calculated using a preset weighting function.

[0019] The above technical solution further refines the calculation method for heart rate deviation, taking into account both static values ​​and dynamic trends. This significantly improves the sensitivity and accuracy of heart rate feature comparison. By simultaneously comparing whether the absolute value of the heart rate is within a stable range and whether the rate of change of the heart rate conforms to the expected rise or recovery pattern, this method can detect more subtle anomalies, such as cases where the heart rate value is within the range but the trend of change is abnormal. This effectively reduces the false positive rate and makes identity verification more rigorous.

[0020] Optionally, a preset security response can be executed when the data is determined to be invalid, specifically including: If the identity consistency deviation is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, and the second preset threshold is greater than the first preset threshold, then it is an invalid data association abnormal status flag. If the deviation of identity consistency is greater than the second preset threshold, it is an invalid data association abnormal state flag, and an alarm data packet is generated and sent to a preset external terminal.

[0021] The above technical solution defines a tiered, automated security response strategy. Based on the level of risk, it achieves differentiated and intelligent security management. By setting two different deviation thresholds, the fitness tracker can flag medium-risk cases only at the data level, avoiding excessive user interference; for high-risk cases, it immediately triggers an alarm and notifies external devices. This tiered response mechanism ensures data security while also improving the user experience, making security measures both effective and non-overly sensitive.

[0022] Secondly, a fitness tracker is provided that performs any of the methods described above, the fitness tracker comprising: The main body of the wristband includes a motion sensor, a physiological data sensor, and a processor. The fingerprint recognition finger sleeve is connected to the main body of the wristband via a flexible connecting strip. It is configured to collect a first fingerprint feature and a second fingerprint feature, and send the first fingerprint feature and the second fingerprint feature to the processor. The first fingerprint feature is the fingerprint feature of the target user acquired before the start of the movement, and the second fingerprint feature is the fingerprint feature of the wearer acquired during the movement. The sensor area on the fingerprint recognition finger sleeve used to collect the first fingerprint feature and the second fingerprint feature is covered with a hydrophobic coating. Motion sensor, configured to collect motion data from the wearer; Physiological data sensor, configured to collect physiological data from the wearer; The processor is configured to generate a reference fingerprint template based on a first fingerprint feature; construct a motion feature template based on motion data within a calibration window of a first preset duration; perform fingerprint matching verification between a second fingerprint feature and the reference fingerprint template; when the fingerprint matching verification result is inconsistent, perform a conformity comparison analysis between physiological data and the motion feature template to generate an identity consistency deviation; and, based on the comparison result between the identity consistency deviation and a preset fault tolerance threshold, determine whether the real-time motion data associated with the physiological data is valid or invalid data, and execute a preset security response when it is determined to be invalid data.

[0023] Optionally, the main body of the bracelet may also include: An embedded user identification module is configured to send an alarm data packet generated when real-time motion data is determined to be invalid to a preset external terminal.

[0024] In summary, one or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: By establishing a dual identity benchmark of physical fingerprint and personal motion feature template, the technical challenge of easily spoofed single-point fingerprint verification is solved. During exercise, it first performs periodic fingerprint matching as the first layer of monitoring. If fingerprint verification fails, it seamlessly switches to the second layer of monitoring, instantly comparing the current wearer's motion features with pre-stored motion feature templates. Finally, based on the result of this dual verification, the motion data is authenticated in real time, determining its validity or invalidity. This series of interconnected steps achieves a continuous and highly reliable binding between user identity and motion data, effectively preventing cheating by proxy runners and ensuring data authenticity. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating a motion data detection method with real-time identity authentication in an embodiment of this application; Figure 2 This is a schematic diagram of the structure of a fitness tracker according to an embodiment of this application.

[0026] Figure 3 This is a schematic diagram of the appearance of a fitness tracker according to an embodiment of this application.

[0027] Explanation of reference numerals in the attached figures: 201, fingerprint recognition finger sleeve; 202, wristband body; 203, processor; 204, motion sensor; 205, physiological data sensor; 206, embedded user identification module. Detailed Implementation

[0028] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0029] In the description of the embodiments of this application, the words "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design that is described as "for example" or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design options. Rather, the use of the words "for example" or "for instance" is intended to present the relevant concepts in a specific manner.

[0030] In the description of the embodiments of this application, the term "multiple" means two or more. For example, "multiple fitness trackers" refers to two or more fitness trackers, and "multiple screen terminals" refers to two or more screen terminals. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0031] Figure 1 This is a flowchart illustrating a motion data detection method with real-time authentication in an embodiment of this application. This method can be implemented using a computer program or a microcontroller, and can run on a fitness tracker. The computer program can be integrated into an application or run as a standalone utility application. The specific steps of a motion data detection method with real-time authentication are described in detail below.

[0032] S101: Obtain the first fingerprint feature of the target user and generate a baseline fingerprint template. The first fingerprint feature is the fingerprint feature obtained before the start of the movement.

[0033] In this embodiment, the reference fingerprint template refers to a processed and stored original reference standard used for all subsequent fingerprint matching verifications. It uniquely represents the user's fingerprint information in digital form. For example, the reference fingerprint template is like a signature sample reserved in a bank's fitness tracker; all subsequent signatures will be compared with this sample to confirm identity.

[0034] Specifically, before a user uses the device for the first time or before starting a new workout session, the fitness tracker prompts the user to perform an initial identity binding. During this process, the user provides their initial fingerprint feature by placing their finger on the sensor area of ​​the fingerprint recognition sleeve. This initial fingerprint feature refers to the raw, unprocessed fingerprint image or signal acquired before the start of the workout. After acquiring the initial fingerprint feature, the fitness tracker performs a series of image processing and feature extraction operations, such as identifying endpoints, bifurcation points, and other detailed features in the fingerprint image. These unique feature points and their relative positions are then converted into an encrypted, compact digital data format. This digital data, securely stored in the fitness tracker, is established as the baseline fingerprint template, serving as the sole basis for comparison in all subsequent real-time authentication.

[0035] S102: Within the first preset calibration window, continuously collect and analyze the wearer's motion data, and construct a motion feature template based on the motion data.

[0036] For example, to construct a unique motion feature template for each user, the fitness tracker first analyzes their motion data: On one hand, it analyzes the user's posture signals using Fast Fourier Transform to determine the main motion frequency, thereby classifying the motion into different intensity levels such as walking and jogging; simultaneously, it uses algorithms such as peak detection, high-pass filtering, and double integration to accurately calculate the user's average stride frequency, stride length, and vertical amplitude at that intensity, collectively forming their "gait signature." On the other hand, for each intensity level, it uses methods such as standard deviation analysis, linear regression, and exponential decay function fitting to accurately calculate the user's stable heart rate range, heart rate rise rate, and heart rate recovery rate, forming their "cardiac signature." Combining the user's unique "gait signature" and "cardiac signature" generates a highly personalized, multi-dimensional motion feature template bound to their identity.

[0037] In one possible implementation, constructing a motion feature template based on motion data specifically includes: extracting multiple consecutive first data components representing the user's dynamic posture characteristics from the motion data; dividing the motion data into multiple different exercise intensity level intervals based on each first data component; separating a second data component representing the user's cardiovascular activity characteristics corresponding to each exercise intensity level interval from the motion data; calculating the average cadence and average stride length based on the first data components, and establishing a gait feature parameter set based on the average cadence and average stride length; establishing a heart rate feature parameter set associated with each exercise intensity level interval based on the second data components, the heart rate feature parameter set including the heart rate stability interval, heart rate rise rate, and heart rate recovery rate; and combining the gait feature parameter set and the heart rate feature parameter set to generate a motion feature template.

[0038] In this embodiment, the motion feature template refers to a structured personal data profile indexed by different exercise intensity levels, which records in detail the user's unique gait and cardiovascular activity parameters at the corresponding intensity. It is a multi-dimensional biometric benchmark strongly correlated with exercise status, used for subsequent identity consistency comparison analysis. For example, a user's motion feature template might record that at the jogging intensity level, their average cadence is 150 steps / minute, average stride length is 0.9 meters, heart rate is stable in the range of 135-145 beats / minute, and their heart rate rises with a specific slope when switching from walking to jogging.

[0039] Specifically, the process first extracts two core data categories from the raw motion data collected by the fitness tracker's sensors (such as accelerometers, gyroscopes, and PPG (Photoplethysmogram) sensors). The first data component, primarily derived from the accelerometer, refers to physical signals reflecting the user's limb movements, rhythm, and posture, forming the basis for gait analysis. Simultaneously, a second data component is extracted, reflecting the user's physiological state, mainly from the PPG sensor, used to acquire cardiovascular activity characteristics such as heart rate. Based on the analysis of the first data component (e.g., through step frequency or acceleration amplitude), the user's motion state is divided into several different exercise intensity levels, such as "stationary," "walking," "jogging," and "running." For the first data component, the number of steps per unit time is calculated using algorithms such as peak detection to obtain the average step frequency, and the average stride is estimated using a biomechanical model in conjunction with information such as vertical amplitude. These two core indicators together constitute the user's gait characteristic parameter set. For each defined exercise intensity level range, a refined analysis of the corresponding second data component is performed to establish a set of associated heart rate characteristic parameters. These include: the range within which the user's heart rate remains relatively stable after sustained exercise at that intensity (i.e., the heart rate stability range); the rate at which the heart rate rises from low to high as exercise intensity increases (i.e., the heart rate climb rate); and the rate at which the heart rate falls rapidly after exercise intensity decreases (i.e., the heart rate recovery rate). The calculated gait characteristic parameter set, representing the user's general exercise habits, is then structurally combined with the heart rate characteristic parameter set, which reflects the cardiopulmonary function response characteristics at different intensity levels, ultimately generating a comprehensive, multi-layered exercise characteristic template.

[0040] In one possible implementation, based on each first data component, the motion data is divided into multiple different motion intensity level intervals, specifically including: performing a fast Fourier transform on each first data component to obtain frequency spectrum data corresponding to each first data component, each frequency spectrum data containing multiple frequency amplitudes; determining the maximum frequency amplitude in each frequency spectrum data, and determining the frequency value corresponding to the maximum frequency amplitude as the main motion frequency; and dividing each first data component into one of the multiple motion intensity level intervals according to each main motion frequency through a preset frequency intensity mapping relationship.

[0041] In this embodiment, the dominant motion frequency refers to the periodic frequency with the highest energy content extracted from the user's dynamic posture signal during movement. It objectively quantifies the user's current primary motion rhythm. For example, when a user runs at a frequency of 180 steps per minute, the frequency with the highest energy concentration in the dynamic posture signal such as arm swing is 3 Hz (180 times / 60 seconds). At this time, 3 Hz is determined as the dominant motion frequency, serving as a core motion parameter to characterize that the user is in a high-intensity exercise state.

[0042] Specifically, multiple consecutive first data components representing the user's dynamic posture characteristics are acquired. This typically refers to a segment of accelerometer time-series data collected within a specific time window (e.g., 2-3 seconds). To objectively assess the motion intensity corresponding to this data, an FFT is first performed. FFT is an efficient signal analysis algorithm that decomposes the original, time-dimension-based first data component into a superposition of multiple sine waves of different frequencies, generating frequency spectrum data. This frequency spectrum data clearly shows all the frequency components contained in the original motion signal and the intensity of each frequency component, i.e., the frequency amplitude. A search is performed within the obtained frequency spectrum data to determine the largest frequency amplitude. The frequency corresponding to this largest frequency amplitude is determined as the dominant motion frequency in the current motion state, as it represents the most prevalent and energy-concentrated periodic motion pattern within that time window. This calculated dominant motion frequency value is then compared with a preset frequency intensity mapping relationship. This mapping relationship can be a pre-calibrated reference table or function that defines the correspondence between different frequency ranges and different exercise intensity levels (for example, 1.5-2.5Hz corresponds to "jogging", and greater than 2.5Hz corresponds to "sprinting"). Through this mapping process, the current first data component is accurately classified into a specific exercise intensity level range among multiple exercise intensity level ranges, thus completing the automatic classification of exercise intensity.

[0043] In one possible implementation, the average step frequency and average stride length are calculated based on the first data component, and a gait feature parameter set is established based on the average step frequency and average stride length. Specifically, this includes: performing a peak detection algorithm on the first data component to obtain multiple consecutive signal peaks in the first data component, and calculating the average step frequency based on the time interval between the multiple consecutive signal peaks, where the multiple consecutive signal peaks represent a single gait cycle; separating the original vertical acceleration signal from the first data component; processing the original vertical acceleration signal through a high-pass filter to generate a corrected vertical acceleration signal; performing a first numerical integration operation on the corrected vertical acceleration signal to generate a vertical velocity signal; performing a second numerical integration operation on the vertical velocity signal to generate a vertical displacement signal; determining the target maximum displacement value and the target minimum displacement value in the vertical displacement signal, and determining the difference between the target maximum displacement value and the target minimum displacement value as the vertical amplitude; calculating the average stride length using the average step frequency and vertical amplitude through a preset biomechanical function; and establishing the calculated values ​​of the average step frequency and average stride length together as a gait feature parameter set.

[0044] In this embodiment, the gait feature parameter set refers to a set of core data indicators used to quantify and uniquely identify an individual's walking or running pattern. This parameter set mainly consists of two key values: average cadence and average stride length. Together, they depict the user's dynamic biomechanical characteristics during movement, much like a person's movement signature. For example, a user's gait feature parameter set might be recorded as {average cadence: 160 steps / minute, average stride length: 1.1 meters}, and this data can serve as one of the unique bases for their personal identity verification.

[0045] Specifically, a gait feature parameter set is established through two parallel computational paths. The first path calculates the average cadence: a peak detection algorithm is performed on the first data component from the accelerometer. Since each step the user takes generates a significant shock wave peak due to the foot striking the ground, multiple consecutive signal peaks identified by the algorithm can be considered as consecutive step nodes, and the time interval between two adjacent peaks represents the time required to complete a single gait cycle. By calculating the average of these time intervals and converting it to the number of steps per minute, the average cadence can be accurately obtained.

[0046] Furthermore, the second path is used to calculate the average stride: first, the raw vertical acceleration signal is precisely separated from the first data component, focusing only on the user's body acceleration in the vertical direction. As mentioned earlier, the first data component mainly comes from an accelerometer. Specifically, each first data component is a data packet containing a time-varying acceleration sequence along three mutually orthogonal coordinate axes (usually denoted as the X-axis, Y-axis, and Z-axis). To calculate the vertical amplitude reflecting the user's body's up-and-down movement, the acceleration signal representing vertical motion, i.e., the "raw vertical acceleration signal," needs to be precisely separated from this three-axis data packet. In a typical way of wearing a wristband, this usually corresponds to the Z-axis acceleration data. However, this raw vertical acceleration signal (i.e., the Z-axis signal) contains a constant gravitational acceleration (approximately 9.8 m / s²), which interferes with subsequent integration calculations. Therefore, it needs to be processed by a high-pass filter to filter out this low-frequency gravitational component, generating a clean, corrected vertical acceleration signal generated solely by the user's movement. Next, a first numerical integration operation is performed on the correction signal. Based on the principle of physics (velocity is the integral of acceleration over time), a vertical velocity signal is obtained. Then, a second numerical integration operation is performed on the vertical velocity signal (displacement is the integral of velocity over time), thereby generating a vertical displacement signal. This signal accurately depicts the up-and-down trajectory of the user's center of mass during walking or running. Within one or more gait cycles of this displacement signal, the highest point (target maximum displacement value) and the lowest point (target minimum displacement value) of its trajectory are determined. The difference between the two is the user's vertical amplitude, which quantifies the bounce height of each step. The average step frequency calculated from the first path and the vertical amplitude obtained from the current path are used as inputs and substituted into a preset biomechanical function. This function is a mathematical model based on a large amount of experimental data, which reveals the intrinsic relationship between step frequency, vertical amplitude, and stride length. This function can be a multiple linear regression function calibrated through experimental data, and its specific form can be expressed as: SL=c1·f+c2·A v +c0 Where: SL represents the average stride length to be calculated (unit: meters), f represents the average step frequency calculated from the first path (unit: Hertz Hz, i.e., steps / second), A v This represents the vertical amplitude calculated for the current path (in meters). c1, c2, and c0 represent the regression coefficients preset in the wristband. These coefficients are empirical constants obtained by collecting stride data from a large number of test subjects of different heights and weights at different speeds (corresponding to different stride frequencies and vertical amplitudes), and fitting the data using a multiple linear regression analysis method. The user's average stride can be estimated using this function.

[0047] In one possible implementation, based on the second data component, a set of heart rate characteristic parameters associated with each exercise intensity level interval is established. Specifically, this includes: segmenting the second data component within each exercise intensity level interval using a second preset time window; calculating the standard deviation of the second data component within each time window; comparing the standard deviation with a preset first stability threshold to determine a first continuous time period within the second data component, where the standard deviation remains consistently below the first stability threshold during the first continuous time period; determining a first stable heart rate interval based on the maximum and minimum values ​​of the second data component within the first continuous time period; identifying the intensity level transition point, where the exercise intensity level interval changes from a first prior exercise intensity level interval to a first current exercise intensity level interval, and the intensity level of the first current exercise intensity level interval is higher than that of the first prior exercise intensity level interval; identifying the second data component within the time period from the intensity level transition point to the start of the first continuous time period as heart rate climb data; and processing the heart rate climb data using linear regression... The algorithm calculates the slope of the regression line and determines it as the heart rate rise rate. It identifies the moment when the intensity level decreases, at which point the exercise intensity level interval changes from the second preceding exercise intensity level interval to the second current exercise intensity level interval, where the intensity level of the second current exercise intensity level interval is lower than that of the second preceding exercise intensity level interval. The time period from the moment when the intensity level decreases until the starting point when the standard deviation value is consistently lower than the first stable threshold is defined as heart rate recovery data. The heart rate recovery data is processed using an exponential decay function to calculate the decay coefficient, which is then determined as the heart rate recovery rate. The continuous time period where the standard deviation value is consistently lower than the first stable threshold is defined as the second continuous time period. Based on the maximum and minimum values ​​of the second data component within the second continuous time period, a second heart rate stable interval is determined. The first heart rate stable interval, the second heart rate stable interval, the heart rate rise rate, and the heart rate recovery rate are collectively established as the heart rate feature parameter set.

[0048] In this embodiment, the heart rate characteristic parameter set refers to a set of key indicators that comprehensively and quantitatively describe the response pattern of a user's cardiovascular exercise bracelet at a specific exercise intensity level. It not only includes the heart rate range when the user reaches physiological steady state at that intensity, i.e., the heart rate stability interval, but more importantly, it also captures the dynamic changes in the user's heart rate from an unstable state to a steady state, such as the rate at which the heart rate increases with increasing intensity (heart rate rise rate) and the rate at which the heart rate decreases with decreasing intensity (heart rate recovery rate). For example, a user's heart rate characteristic parameter set at a fast running intensity might be: {First heart rate stability interval: [150, 160] beats / minute, Second heart rate stability interval: [152, 161] beats / minute, Heart rate rise rate: 3.5, Heart rate recovery rate: 0.9}. This set of parameters collectively constitutes their unique physiological "fingerprint".

[0049] Specifically, this process meticulously constructs a set of heart rate feature parameters for each exercise intensity level interval. First, the second data component (i.e., continuous heart rate readings) within the intensity interval is sliced ​​into segments according to a second preset time window (e.g., every 5 seconds), and the standard deviation of the data within each time window is calculated. This sequence of standard deviation values ​​reflects the real-time fluctuations of heart rate and forms the basis for all subsequent feature extractions. The standard deviation values ​​are compared with a preset, relatively low first stability threshold. When the standard deviation remains below this threshold for a continuous period (the first continuous time period), it indicates that the heart rate has entered a stable state. At this point, the first stable heart rate interval is determined based on the maximum and minimum heart rate values ​​during this period. Then, the transition point of the exercise intensity level is identified. The data from this point to the start of the aforementioned stable period is the heart rate rise data, which is analyzed using a linear regression algorithm. The slope of the calculated regression line is the heart rate rise rate.

[0050] Furthermore, the point in time where the exercise intensity level decreases is identified. From this point onward, the heart rate drops rapidly, and the standard deviation temporarily increases. The fitness tracker continues to monitor until the standard deviation is consistently below the first stable threshold again. All heart rate data from the point of intensity decrease to this new stable point is precisely defined as heart rate recovery data. This definition is logical because it fully captures the entire process from dynamic change to re-stabilization. Subsequently, an exponential decay function is applied to this recovery data for fitting, and the calculated decay coefficient is the heart rate recovery rate.

[0051] Furthermore, after the recovery process described above, the heart rate enters a new stable period. The fitness tracker defines the continuous time during which the standard deviation remains below the first stable threshold as the second continuous time period. This step is entirely consistent with the logic of finding the stable interval in the first step, ensuring methodological consistency. Then, based on the maximum and minimum heart rate values ​​within this new stable time period, the second heart rate stable interval is determined. This second heart rate stable interval represents the user's new steady state after experiencing an intensity change and recovering, and together with the first heart rate stable interval, it constitutes the user's steady-state heart rate characteristics at that intensity level. The key parameters calculated through the above steps—the first heart rate stable interval, the second heart rate stable interval, the heart rate rise rate, and the heart rate recovery rate—are integrated to establish a complete set of heart rate characteristic parameters associated with the current exercise intensity level.

[0052] S103: Continuously collect the wearer's second fingerprint features and physiological data, and perform fingerprint matching verification between the second fingerprint features and the reference fingerprint template. The second fingerprint features are fingerprint features acquired during movement.

[0053] For example, after the first preset calibration window ends, the fitness tracker enters a continuous real-time monitoring phase. During this phase, the fitness tracker continuously collects the wearer's fingerprint information via the fingerprint recognition finger sleeve. When the wearer's finger touches the sensor, the fitness tracker extracts a set of feature points from this real-time, dynamic fingerprint image. This set of features acquired during movement constitutes the second fingerprint feature, a sample used for immediate comparison. Furthermore, other physiological sensors on the fitness tracker (such as the PPG sensor and accelerometer) continuously collect the wearer's physiological data, such as real-time heart rate and posture data. This data is temporarily cached after collection to prepare for potential fingerprint verification failures. After collecting the second fingerprint feature, a fingerprint matching verification operation is immediately performed. Using a built-in matching algorithm, this real-time sample (the second fingerprint feature) is compared at high speed with a pre-stored baseline fingerprint template, calculating the similarity between the two. The result of this comparison directly determines whether the current wearer is the legitimate user.

[0054] S104: When the fingerprint matching verification result is inconsistent, the physiological data and the motion feature template are compared and analyzed to generate the identity consistency deviation.

[0055] For example, when fingerprint verification fails, the fitness tracker initiates a multi-dimensional conformity comparison analysis process to calculate the final identity consistency deviation. Based on the current exercise intensity, the fitness tracker extracts the corresponding personal heart rate and gait parameter sets from the exercise feature template as a reference. The fitness tracker compares the current physiological data with these two parameter sets in parallel to obtain gait deviation and heart rate deviation, respectively. The heart rate deviation calculation employs a composite evaluation standard combining static and dynamic factors. It not only compares whether the current heart rate value is within the user's stable range (static value) but also analyzes whether its recent trend matches the user's personal heart rate rise or recovery pattern (dynamic trend). Through a preset weighted algorithm, these two deviations are fused to obtain an identity consistency deviation that comprehensively and accurately reflects the degree of consistency between the current wearer and the legitimate user's identity.

[0056] In one possible implementation, physiological data is compared with a motion feature template to generate an identity consistency deviation score. Specifically, this includes: determining the target exercise intensity level range to which the physiological data belongs; extracting a target heart rate feature parameter set and a target gait feature parameter set related to the target exercise intensity level range based on the motion feature template; comparing the physiological data with the target heart rate feature parameter set to obtain a heart rate deviation score; comparing the physiological data with the target gait feature parameter set to obtain a gait deviation score; and calculating the identity consistency deviation score by using a preset weighted algorithm to calculate the heart rate deviation score and the gait deviation score.

[0057] In this embodiment, the identity consistency deviation refers to a calculated comprehensive value used to quantify the degree of mismatch between the current wearer's real-time motion physiological data and their pre-stored motion characteristic template. The lower the value, the higher the consistency between the current data and the user's own characteristics, and the stronger the identity consistency; conversely, the higher the value, the greater the discrepancy, and the higher the risk that the wearer may not be the actual person. For example, in a scoring system from 0 to 100, a deviation below 10 may indicate identity matching, while a deviation above 60 may trigger a security alarm.

[0058] Specifically, to ensure fairness and accuracy in the comparison, the fitness tracker needs to perform contextual localization. It analyzes the currently collected physiological data (especially the motion signal portion) in real time to determine the target exercise intensity level range to which the physiological data belongs, for example, determining that the wearer is currently jogging. After determining the current exercise intensity level (e.g., jogging), the fitness tracker retrieves the corresponding benchmark data. It accesses a pre-built motion feature template and precisely extracts a subset of features related to the target exercise intensity level range, specifically the target heart rate feature parameter set for jogging (including the normal heart rate stability range, ascent rate, etc.) and the target gait feature parameter set (including the average cadence and stride length during jogging). Next, the fitness tracker performs independent comparisons. On one hand, it compares real-time heart rate data (such as the current heart rate value and heart rate trend) with the target heart rate feature parameter set, calculates the difference between the two, and quantifies it into a numerical value to obtain the heart rate deviation. On the other hand, it simultaneously compares the current cadence and stride length calculated from real-time motion signals with the target gait feature parameter set, quantifying the differences to obtain the gait deviation. Finally, the fitness tracker performs a comprehensive weighted evaluation. It doesn't simply average the two deviations; instead, it uses a preset weighted algorithm to fuse the heart rate deviation and gait deviation. This algorithm may assign different weights based on the stability and uniqueness of different features. For example, gait features may be more difficult to imitate than heart rate features, and therefore may be given a higher weight (e.g., 60%). Through this weighted calculation, it ultimately arrives at an identity consistency deviation that comprehensively reflects the similarity between the current wearer and the user.

[0059] In one possible implementation, comparing physiological data with a target heart rate feature parameter set to obtain heart rate deviation specifically includes: comparing the heart rate value in the physiological data with the target heart rate stable interval in the target heart rate feature parameter set to obtain interval deviation; calculating the heart rate change rate of the physiological data within a preset time window before the current moment, and comparing the heart rate change rate with the target heart rate rise rate or target heart rate recovery rate in the target heart rate feature parameter set to obtain change rate deviation; and calculating the heart rate deviation based on the interval deviation and the change rate deviation using a preset weighting function.

[0060] In this embodiment, heart rate deviation refers to a comprehensive evaluation score that not only measures whether the user's instantaneous heart rate is within the normal range, but more importantly, it assesses whether the dynamic trend of heart rate changes conforms to its inherent physiological response pattern. This is a composite index generated by combining static value comparison and dynamic trend comparison, aiming to more comprehensively and accurately determine whether there are abnormalities in the current heart rate data. For example, a wearer's heart rate value may happen to be within their normal range, but if this heart rate value is caused by a sudden and drastic fluctuation rather than a smooth rise or fall, then their heart rate deviation will still be high, thus identifying this potential abnormality.

[0061] Specifically, after obtaining the target heart rate feature parameter set, the fitness tracker executes two independent conformity analysis processes in parallel: First, it extracts the current heart rate value from real-time physiological data. Then, it compares this single value with the target heart rate stability interval (a pre-defined normal range, such as 140-150 beats / minute) retrieved from the target heart rate feature parameter set. The fitness tracker determines whether the current heart rate value falls within this interval and calculates a quantified difference value based on the distance or degree to which it deviates from the interval; this value is the interval deviation. If the heart rate value is within the interval, the deviation is 0 or a very small value. Second, the fitness tracker examines a series of heart rate readings within a preset time window prior to the current moment (e.g., the past 10 seconds) and calculates the recent heart rate change rate, i.e., whether the heart rate is rising or falling, and the speed of the change. Next, the fitness tracker compares the real-time calculated heart rate variability with the target heart rate rise rate or target heart rate recovery rate corresponding to the target heart rate feature parameter set, based on the current exercise context (whether it's in an increasing or decreasing intensity phase). For example, if the user is increasing exercise intensity, the fitness tracker compares their real-time heart rate rise rate with the rise rate in the template. The difference between the two is quantified, resulting in the variability deviation. To obtain a comprehensive indicator that fully reflects heart rate consistency, the fitness tracker calculates the final heart rate deviation based on the interval deviation and the variability deviation using a preset weighted function. This function may assign different weights to the interval deviation and the variability deviation depending on the exercise state. For example, the specific formula for the preset weighted function is: Heart Rate Deviation = (Interval Deviation × W1) + (Variation Deviation × W2), where W1 and W2 are preset weight coefficients, and W1 + W2 = 1. For example, a fitness tracker can assign a higher weight (e.g., W1=0.7) to the deviation from a static value range and a lower weight (e.g., W2=0.3) to the deviation from a rate of change representing a dynamic trend, thus achieving differentiated consideration of deviations in different dimensions. In a more preferred implementation, the weighting coefficients W1 and W2 can also be dynamically adjusted according to the context of the current exercise. For example, when the user is in the heart rate ramp-up or recovery phase, the weight W2 of the rate of change deviation can be dynamically increased to more sensitively capture abnormal dynamic trends.

[0062] S105: Based on the comparison result of the identity consistency deviation degree and the preset fault tolerance threshold, determine whether the real-time motion data associated with the physiological data is valid or invalid data, and execute the preset safety response when it is determined to be invalid data.

[0063] For example, after calculating the final identity consistency deviation, it is compared with a preset fault tolerance threshold (e.g., this threshold can be set to 30). If the identity consistency deviation (e.g., the calculated result is 25) is lower than or equal to the preset fault tolerance threshold, the fitness tracker determines that the current wearer's identity matches that of a legitimate user, and therefore determines that the real-time motion data temporally associated with the physiological data used in this verification is valid data. Conversely, if the identity consistency deviation (e.g., the calculated result is 80) is higher than the preset fault tolerance threshold, the fitness tracker determines that the current wearer's identity does not match that of a legitimate user, determines that the corresponding real-time motion data is invalid data, and initiates a preset security response.

[0064] In one possible implementation, when invalid data is determined, a preset security response is executed, specifically including: if the identity consistency deviation is greater than or equal to a first preset threshold and less than or equal to a second preset threshold, and the second preset threshold is greater than the first preset threshold, then an invalid data association anomaly status flag is generated; if the identity consistency deviation is greater than the second preset threshold, then an invalid data association anomaly status flag is generated, an alarm data packet is generated, and the alarm data packet is sent to a preset external terminal.

[0065] In this application embodiment, tiered security response refers to an intelligent security processing mechanism that executes different levels of pre-set response measures based on the quantified severity of risk (identity consistency deviation). For example, tiered security response is like an intelligent weather warning fitness tracker; for a blue alert (lower risk), it may only issue a notification in the background; while for a red alert (extremely high risk), it will trigger an emergency broadcast and contingency plan throughout the city. Similarly, a low identity consistency deviation may only result in a silent data tagging, while an extremely high deviation will trigger an active, real-time user alert.

[0066] Specifically, after the fitness tracker determines that the real-time exercise data is invalid, it initiates a tiered security response process. The fitness tracker first compares the calculated identity consistency deviation with two preset thresholds of progressively increasing intensity: a first preset threshold and a second preset threshold, where the second preset threshold is numerically greater than the first preset threshold. The fitness tracker executes different response actions depending on the range the identity consistency deviation falls into: If the identity consistency deviation is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, the fitness tracker executes a low-level response, simply associating an abnormal status flag with the invalid data in local storage. This operation is silent and does not actively disturb the user. If the identity consistency deviation is greater than the second preset threshold, the fitness tracker executes a high-level response. In addition to associating the invalid data with an abnormal status flag, it also generates an alarm data packet and sends it to a preset external terminal (such as the user's smartphone) to proactively alert the user in real time.

[0067] Figure 2 This is a schematic diagram of the structure of a fitness tracker according to an embodiment of this application.

[0068] Figure 3 This is a schematic diagram of the external shape of a fitness tracker according to an embodiment of this application. Figure 2-3 As shown, the fitness tracker includes: The main body of the wristband 202 includes a motion sensor 204, a physiological data sensor 205, and a processor 203. The fingerprint recognition finger sleeve 201 is connected to the main body 202 of the wristband via a flexible connecting strip. It is configured to collect a first fingerprint feature and a second fingerprint feature, and send the first fingerprint feature and the second fingerprint feature to the processor 203. The first fingerprint feature is the fingerprint feature of the target user acquired before the start of the movement, and the second fingerprint feature is the fingerprint feature of the wearer acquired during the movement. The sensor area on the fingerprint recognition finger sleeve 201 used to collect the first fingerprint feature and the second fingerprint feature is covered with a hydrophobic coating. Motion sensor 204, configured to collect motion data of the wearer; Physiological data sensor 205, configured to collect physiological data from the wearer; The processor 203 is configured to generate a reference fingerprint template based on a first fingerprint feature; construct a motion feature template based on motion data within a preset calibration window; perform fingerprint matching verification between a second fingerprint feature and the reference fingerprint template; when the fingerprint matching verification result is inconsistent, use physiological data as physiological data and perform a conformity comparison analysis between the physiological data and the motion feature template to generate an identity consistency deviation; and determine whether the real-time motion data associated with the physiological data is valid or invalid based on the comparison result of the identity consistency deviation and a preset fault tolerance threshold, and execute a preset security response when it is determined to be invalid data.

[0069] Optionally, the main body 202 of the wristband also includes: The embedded user identification module 206 is configured to send an alarm data packet generated when the real-time motion data is determined to be invalid to a preset external terminal.

[0070] Based on the above embodiments, as an optional embodiment, the processor 203 is further configured to: Extract multiple consecutive first data components representing the user's dynamic posture characteristics from motion data; Based on each first data component, the motion data is divided into multiple different motion intensity level intervals; Separate a second data component representing the user's cardiovascular activity characteristics from the exercise data, corresponding to each exercise intensity level range; The average step frequency and average stride length are calculated based on the first data component, and a gait feature parameter set is established based on the average step frequency and average stride length. Based on the second data component, a set of heart rate characteristic parameters associated with each exercise intensity level interval is established. The set of heart rate characteristic parameters includes the heart rate stable interval, heart rate rise rate, and heart rate recovery rate. The gait feature parameter set and the heart rate feature parameter set are combined to generate a motion feature template.

[0071] Based on the above embodiments, as an optional embodiment, the processor 203 is further configured to: Perform a fast Fourier transform on each first data component to obtain frequency spectrum data corresponding to each first data component. Each frequency spectrum data contains multiple frequency amplitudes. In each frequency spectrum data, the maximum frequency amplitude is determined, and the frequency value corresponding to the maximum frequency amplitude is determined as the main motion frequency; Based on a preset frequency intensity mapping relationship, each first data component is divided into one of multiple motion intensity level intervals according to each main motion frequency.

[0072] Based on the above embodiments, as an optional embodiment, the processor 203 is further configured to: A peak detection algorithm is performed on the first data component to obtain multiple consecutive signal peaks in the first data component. The average step frequency is calculated based on the time interval between the multiple consecutive signal peaks. The multiple consecutive signal peaks represent a single gait cycle. The original vertical acceleration signal is separated from the first data component; The original vertical acceleration signal is processed through a high-pass filter to generate a corrected vertical acceleration signal; Perform the first numerical integration operation on the corrected vertical acceleration signal to generate a vertical velocity signal; Perform a second numerical integration operation on the vertical velocity signal to generate a vertical displacement signal; In the vertical displacement signal, the maximum and minimum displacement values ​​of the target are determined, and the difference between the maximum and minimum displacement values ​​of the target is determined as the vertical amplitude. The average stride length is calculated from the average stride frequency and vertical amplitude using a preset biomechanical function. The calculated values ​​of average cadence and average stride are combined to form a gait feature parameter set.

[0073] Based on the above embodiments, as an optional embodiment, the processor 203 is further configured to: Within each exercise intensity level range, the second data component is segmented through a second preset time window; Calculate the standard deviation of the second data component within each time window; The standard deviation value is compared with a preset first stable threshold to determine the first continuous time period in the second data component. During the first continuous time period, the standard deviation value is continuously lower than the first stable threshold. The first stable heart rate interval is determined based on the maximum and minimum values ​​of the second data component within the first continuous time period. Identify the moment when the intensity level increases and changes. At the moment when the intensity level increases and changes, the exercise intensity level interval changes from the first previous exercise intensity level interval to the first current exercise intensity level interval. The intensity level of the first current exercise intensity level interval is higher than the intensity level of the first previous exercise intensity level interval. The second data component within the time period between the point when the intensity level changes and the start point of the first continuous time period is identified as the heart rate climb data. The slope of the regression line was calculated using a linear regression algorithm on the heart rate rise data, and the slope of the regression line was determined as the heart rate rise rate. Identify the moment when the intensity level decreases and changes. At the moment when the intensity level decreases and changes, the exercise intensity level interval changes from the second previous exercise intensity level interval to the second current exercise intensity level interval. The intensity level of the second current exercise intensity level interval is lower than the intensity level of the second previous exercise intensity level interval. The time period from the point when the intensity level changes to the point when the standard deviation value is consistently below the first stable threshold is defined as the heart rate recovery data. The heart rate recovery data were analyzed using an exponential decay function to calculate the decay coefficient, which was then used as the heart rate recovery rate. The continuous period of time in which the standard deviation value is consistently lower than the first stable threshold is defined as the second continuous period. The second heart rate stability interval is determined based on the maximum and minimum values ​​of the second data component within the second continuous time period. The first stable heart rate interval, the second stable heart rate interval, the heart rate rise rate, and the heart rate recovery rate are collectively established as a set of heart rate characteristic parameters.

[0074] Based on the above embodiments, as an optional embodiment, the processor 203 is further configured to: Determine the target exercise intensity level range to which the physiological data belong; Based on motion feature templates, extract target heart rate feature parameter sets and target gait feature parameter sets related to the target motion intensity level range; The physiological data is compared with the target heart rate feature parameter set to obtain the heart rate deviation. Physiological data are compared with the target gait feature parameter set to obtain gait deviation. The identity consistency deviation is calculated by using a preset weighted algorithm to calculate the heart rate deviation and gait deviation.

[0075] Based on the above embodiments, as an optional embodiment, the processor 203 is further configured to: The heart rate value in the physiological data is compared with the target heart rate stable interval in the target heart rate feature parameter set to obtain the interval deviation. Calculate the rate of change of heart rate in physiological data within a preset time window before the current moment, and compare the rate of change of heart rate with the target heart rate rise rate or target heart rate recovery rate in the target heart rate feature parameter set to obtain the deviation of the rate of change. Based on the interval deviation and the rate of change deviation, the heart rate deviation is calculated using a preset weighting function.

[0076] Based on the above embodiments, as an optional embodiment, the processor 203 is further configured to: If the identity consistency deviation is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, and the second preset threshold is greater than the first preset threshold, then it is an invalid data association abnormal status flag. If the deviation of identity consistency is greater than the second preset threshold, it is an invalid data association abnormal state flag, and an alarm data packet is generated and sent to a preset external terminal.

[0077] It should be noted that the above embodiments of the fitness tracker are only illustrated by the division of the functional modules described above. In actual applications, the functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the fitness tracker and method embodiments provided above belong to the same concept, and their specific implementation process can be found in the method embodiments, which will not be repeated here.

Claims

1. A method for detecting motion data with real-time identity authentication, characterized in that, Applied to fitness trackers, the method includes: Obtain the first fingerprint feature of the target user and generate a baseline fingerprint template. The first fingerprint feature is the fingerprint feature obtained before the start of the movement. Within a first preset calibration window, the wearer's motion data is continuously collected and analyzed, and a motion feature template is constructed based on the motion data; The wearer's second fingerprint features and physiological data are continuously collected, and the second fingerprint features are matched and verified with the reference fingerprint template. The second fingerprint features are fingerprint features acquired during movement. When the fingerprint matching verification result is inconsistent, the physiological data and the motion feature template are compared and analyzed to generate an identity consistency deviation. Based on the comparison result between the identity consistency deviation and the preset fault tolerance threshold, the real-time motion data associated with the physiological data is determined to be valid or invalid data, and a preset security response is executed when it is determined to be invalid data; wherein, The construction of motion feature templates based on the motion data specifically includes: Extract a series of first data components representing the user's dynamic posture characteristics from the motion data; Based on each of the first data components, the motion data is divided into multiple different motion intensity level intervals; Separate a second data component representing the cardiovascular activity characteristics of the user from the exercise data, corresponding to each exercise intensity level interval; The average step frequency and average stride length are calculated based on the first data component, and a gait feature parameter set is established based on the average step frequency and the average stride length. Based on the second data component, a set of heart rate characteristic parameters associated with each exercise intensity level interval is established. The set of heart rate characteristic parameters includes heart rate stability interval, heart rate rise rate, and heart rate recovery rate. The gait feature parameter set and the heart rate feature parameter set are combined to generate the motion feature template; The step of comparing and analyzing the physiological data with the motion feature template to generate an identity consistency deviation score specifically includes: Determine the target exercise intensity level range to which the physiological data belongs; Based on the motion feature template, extract the target heart rate feature parameter set and the target gait feature parameter set related to the target motion intensity level range; The physiological data is compared with the target heart rate feature parameter set to obtain the heart rate deviation. The physiological data is compared with the target gait feature parameter set to obtain the gait deviation. The identity consistency deviation is calculated by using a preset weighted algorithm to combine the heart rate deviation and the gait deviation.

2. The method according to claim 1, characterized in that, The step of dividing the motion data into multiple different motion intensity level intervals based on each of the first data components specifically includes: Perform a fast Fourier transform on each of the first data components to obtain frequency spectrum data corresponding to each of the first data components, and each frequency spectrum data contains multiple frequency amplitudes; In each of the frequency spectrum data, the maximum frequency amplitude is determined, and the frequency value corresponding to the maximum frequency amplitude is determined as the main motion frequency; Based on a preset frequency intensity mapping relationship, each of the first data components is divided into one of the multiple motion intensity level intervals.

3. The method according to claim 1, characterized in that, The step of calculating the average step frequency and average stride length based on the first data component, and establishing the gait feature parameter set based on the average step frequency and average stride length, specifically includes: A peak detection algorithm is performed on the first data component to obtain multiple consecutive signal peaks in the first data component, and the average step frequency is calculated based on the time interval between the multiple consecutive signal peaks, wherein the multiple consecutive signal peaks represent a single gait cycle; The original vertical acceleration signal is separated from the first data component; The original vertical acceleration signal is processed through a high-pass filter to generate a corrected vertical acceleration signal; Perform a first numerical integration operation on the corrected vertical acceleration signal to generate a vertical velocity signal; Perform a second numerical integration operation on the vertical velocity signal to generate a vertical displacement signal; In the vertical displacement signal, the target maximum displacement value and the target minimum displacement value are determined, and the difference between the target maximum displacement value and the target minimum displacement value is determined as the vertical amplitude; The average step length is calculated from the average step frequency and the vertical amplitude using a preset biomechanical function. The calculated values ​​of the average cadence and the average stride are used together to establish the gait feature parameter set.

4. The method according to claim 1, characterized in that, The step of establishing a set of heart rate feature parameters associated with each exercise intensity level interval based on the second data component specifically includes: Within each of the exercise intensity level intervals, the second data component is segmented through a second preset time window; Calculate the standard deviation of the second data component within each time window; The standard deviation value is compared with a preset first stable threshold to determine a first continuous time period in the second data component. During the first continuous time period, the standard deviation value is continuously lower than the first stable threshold. The first heart rate stability interval is determined based on the maximum and minimum values ​​of the second data component within the first continuous time period. Identify the moment when the intensity level increases and changes. At the moment when the intensity level increases and changes, the motion intensity level interval changes from the first previous motion intensity level interval to the first current motion intensity level interval, and the intensity level of the first current motion intensity level interval is higher than the intensity level of the first previous motion intensity level interval. The second data component within the time period between the point when the intensity level changes upward and the starting point of the first continuous time period is determined as heart rate rise data; The slope of the regression line is calculated using a linear regression algorithm on the heart rate rise data, and the slope of the regression line is determined as the heart rate rise rate. Identify the moment when the intensity level decreases and changes. At the moment when the intensity level decreases and changes, the motion intensity level interval changes from the second previous motion intensity level interval to the second current motion intensity level interval, and the intensity level of the second current motion intensity level interval is lower than the intensity level of the second previous motion intensity level interval. The time period from the point when the intensity level changes to the point when the standard deviation value is consistently below the first stable threshold is defined as the heart rate recovery data. The heart rate recovery data is processed using an exponential decay function to calculate the decay coefficient, and the decay coefficient is determined as the heart rate recovery rate. The continuous time period during which the standard deviation value is consistently lower than the first stable threshold is defined as the second continuous time period. The second heart rate stability interval is determined based on the maximum and minimum values ​​of the second data component within the second continuous time period. The first stable heart rate interval, the second stable heart rate interval, the heart rate rise rate, and the heart rate recovery rate are collectively established as the heart rate feature parameter set.

5. The method according to claim 1, characterized in that, The step of comparing the physiological data with the target heart rate feature parameter set to obtain the heart rate deviation specifically includes: The heart rate value in the physiological data is compared with the target heart rate stable interval in the target heart rate feature parameter set to obtain the interval deviation. Calculate the rate of change of the heart rate of the physiological data within a preset time window before the current moment, and compare the rate of change of the heart rate with the target heart rate rise rate or target heart rate recovery rate in the target heart rate feature parameter set to obtain the deviation of the rate of change. Based on the interval deviation and the rate of change deviation, the heart rate deviation is calculated using a preset weighting function.

6. The method according to claim 1, characterized in that, The step of executing a preset security response when data is determined to be invalid specifically includes: If the identity consistency deviation is greater than or equal to the first preset threshold and less than or equal to the second preset threshold, and the second preset threshold is greater than the first preset threshold, then it is the invalid data association abnormal state flag. If the identity consistency deviation is greater than the second preset threshold, the invalid data is associated with the abnormal state flag, an alarm data packet is generated, and the alarm data packet is sent to a preset external terminal.

7. A fitness tracker, characterized in that, The fitness tracker, comprising any one of claims 1-6, includes: The main body of the wristband includes a motion sensor, a physiological data sensor, and a processor. A fingerprint recognition finger sleeve is connected to the main body of the wristband via a flexible connecting strap. It is configured to collect a first fingerprint feature and a second fingerprint feature, and send the first fingerprint feature and the second fingerprint feature to the processor. The first fingerprint feature is the fingerprint feature of the target user acquired before the start of exercise, and the second fingerprint feature is the fingerprint feature of the wearer acquired during exercise. The sensor area on the fingerprint recognition finger sleeve used to collect the first fingerprint feature and the second fingerprint feature is covered with a hydrophobic coating. The motion sensor is configured to collect motion data from the wearer. The physiological data sensor is configured to collect physiological data from the wearer. The processor is configured to: generate a baseline fingerprint template based on the first fingerprint feature; construct a motion feature template based on the motion data within a first preset calibration window; perform fingerprint matching verification between the second fingerprint feature and the baseline fingerprint template; when the fingerprint matching verification result is inconsistent, perform a conformity comparison analysis between the physiological data and the motion feature template to generate an identity consistency deviation; and, based on the comparison result of the identity consistency deviation and a preset fault tolerance threshold, determine whether the real-time motion data associated with the physiological data is valid or invalid data, and execute a preset security response when it is determined to be invalid data. The construction of motion feature templates based on the motion data specifically includes: Extract a series of first data components representing the user's dynamic posture characteristics from the motion data; Based on each of the first data components, the motion data is divided into multiple different motion intensity level intervals; Separate a second data component representing the cardiovascular activity characteristics of the user from the exercise data, corresponding to each exercise intensity level interval; The average step frequency and average stride length are calculated based on the first data component, and the gait feature parameter set is established based on the average step frequency and the average stride length. Based on the second data component, a set of heart rate characteristic parameters associated with each exercise intensity level interval is established. The set of heart rate characteristic parameters includes heart rate stability interval, heart rate rise rate, and heart rate recovery rate. The gait feature parameter set and the heart rate feature parameter set are combined to generate the motion feature template; The step of comparing and analyzing the physiological data with the motion feature template to generate an identity consistency deviation score specifically includes: Determine the target exercise intensity level range to which the physiological data belongs; Based on the motion feature template, extract the target heart rate feature parameter set and the target gait feature parameter set related to the target motion intensity level range; The physiological data is compared with the target heart rate feature parameter set to obtain the heart rate deviation. The physiological data is compared with the target gait feature parameter set to obtain the gait deviation. The identity consistency deviation is calculated by using a preset weighted algorithm to combine the heart rate deviation and the gait deviation.

8. The fitness tracker according to claim 7, characterized in that, The main body of the bracelet also includes: An embedded user identification module is configured to send an alarm data packet generated when the real-time motion data is determined to be invalid to a preset external terminal.

Images