In-vehicle living body detection method based on bluetooth phase timing disturbance and related device

Abstract

The application provides an in-vehicle living body detection method based on Bluetooth phase timing disturbance and related devices, relates to the technical field of vehicle safety detection, and is applied to a vehicle equipped with a Bluetooth module supporting Bluetooth channel detection. The method comprises the following steps: in response to configuring the Bluetooth module into an active radar mode, emitting a detection signal through the Bluetooth module, receiving a reflected echo signal after the detection signal is reflected by an in-vehicle environment, and extracting a phase timing sequence between the signals; based on the phase timing sequence, extracting a characteristic parameter representing vital signs; and determining whether there is a living body in the vehicle based on the extracted characteristic parameter. The application reuses an existing Bluetooth module supporting Bluetooth channel detection of a vehicle, realizes double-band joint detection by utilizing weak phase timing disturbance caused by vital signs, and thus reliably detects the living body in the vehicle without increasing the cost and invading privacy.

Description

Technical Field

[0001] This application relates to the field of vehicle safety detection technology, and in particular to an in-vehicle liveness detection method and related device based on Bluetooth phase timing perturbation. Background Technology

[0002] Cases of children dying from heatstroke after being left in vehicles are being reported globally. Statistics show that over 1,000 children have died from this over the past two decades, with approximately 52% of these deaths due to unintentional forgetfulness by caregivers. Therefore, reliably detecting living persons inside vehicles shortly after they are turned off and locked remains a pressing technical challenge.

[0003] Among the related technologies, proposed in-vehicle liveness detection solutions include: embedding pressure sensor arrays in seat cushions to determine the presence of occupants or objects by monitoring pressure distribution; installing ultrasonic or millimeter-wave radar modules on the roof or dashboard to detect micro-moving targets by emitting sound waves or electromagnetic waves and analyzing the Doppler frequency shift of the echoes; installing wide-angle cameras in the vehicle and using deep learning models to identify humans or pets from the captured images; using the vehicle's built-in Wireless Fidelity (Wi-Fi) module to collect channel state information and infer the presence of liveness by analyzing changes in signal amplitude and phase; emitting inaudible linear frequency modulated sound waves through the vehicle's audio system and collecting echoes using microphones to analyze the sound wave's flight time and Doppler frequency shift; emitting nanosecond-level narrow pulse signals based on an ultra-wideband (UWB) chip and achieving liveness detection through pulse flight time and micro-movement feature extraction; and using the vehicle's Bluetooth module to scan surrounding Bluetooth broadcast signals and estimate device distance or detect signal strength changes based on received signal strength indicators. However, the above detection solutions all require the addition of dedicated hardware or rely on non-standard modules, making it difficult to reliably detect living beings inside the vehicle using existing vehicle resources without increasing costs or infringing on privacy. Summary of the Invention

[0004] This application provides a method and related device for in-vehicle liveness detection based on Bluetooth phase timing perturbation, which improves the problem that detection schemes in related technologies all require the addition of dedicated hardware or rely on non-standard modules, making it difficult to reliably detect liveness in the vehicle using existing vehicle resources without increasing costs or infringing on privacy.

[0005] Firstly, this application provides an in-vehicle liveness detection method based on Bluetooth phase timing perturbation, applicable to vehicles equipped with Bluetooth modules supporting Bluetooth channel detection functionality. The in-vehicle liveness detection method includes:

[0006] In response to configuring the Bluetooth module to active radar mode;

[0007] The system transmits detection signals via Bluetooth and receives echo signals reflected from the vehicle's interior environment, extracting the phase timing sequence between the transmitted and echo signals.

[0008] Based on phase time series, feature parameters characterizing vital signs are extracted;

[0009] The presence of a living being inside the vehicle is determined based on the extracted feature parameters.

[0010] In one possible implementation, based on the phase time series, feature parameters characterizing vital signs are extracted, including: performing phase unwinding processing on the phase time series to restore continuous phase; performing high-pass filtering on the unwound phase series to filter out static bias and temperature drift to obtain a dynamic phase perturbation signal; and extracting feature parameters characterizing vital signs from the dynamic phase perturbation signal.

[0011] In one possible implementation, the feature parameters include parameters of the respiratory frequency band and the heart rate frequency band. Extracting feature parameters characterizing vital signs from the dynamic phase perturbation signal includes: performing time-frequency transformation on the dynamic phase perturbation signal to obtain a time-frequency distribution; extracting the respiratory dominant frequency, respiratory energy, and respiratory signal-to-noise ratio from the time-frequency distribution within a preset respiratory frequency band; performing bandpass filtering and envelope extraction on the dynamic phase perturbation signal within a preset heart rate frequency band, and performing frequency domain transformation on the extracted envelope to obtain the heart rate dominant frequency and heart rate energy; and outputting a five-dimensional feature vector containing the respiratory dominant frequency, respiratory energy, respiratory signal-to-noise ratio, heart rate dominant frequency, and heart rate energy.

[0012] In one possible implementation, determining whether a living person exists inside the vehicle based on the extracted feature parameters includes: inputting the extracted feature parameters into a pre-trained classification model, the classification model outputting the probability of the presence of a living person; and determining whether a living person exists inside the vehicle based on the probability of the presence of a living person.

[0013] In one possible implementation, the in-vehicle liveness detection method starts execution in response to a vehicle locking signal, and the execution duration is limited to a preset total detection duration, which includes at least two detection windows; based on the probability of liveness presence, it determines whether there is a liveness presence in the vehicle, including: when the probability of liveness presence is greater than or equal to a preset threshold, it determines that there is a liveness presence in the current detection window; when at least two consecutive detection windows determine that there is a liveness presence, it determines that there is a liveness presence in the vehicle.

[0014] In one possible implementation, the in-vehicle liveness detection method based on Bluetooth phase timing perturbation further includes: in response to determining that a live person is present in the vehicle, performing at least one of the following warning operations: controlling the in-vehicle display device to display warning information and controlling the in-vehicle audio system to play voice prompts; pushing alarm notifications to a preset mobile application and controlling the vehicle to issue an audible and visual alarm; sending help requests to preset emergency contacts or rescue platforms and controlling the vehicle's air conditioning system to turn on external circulation, and controlling the windows to open a preset gap when it is detected that there is no rain outside the vehicle.

[0015] In one possible implementation, the in-vehicle liveness detection method based on Bluetooth phase timing perturbation further includes: in response to a door unlocking signal or a release command received through a mobile application, stopping the current warning operation and controlling the Bluetooth module to enter a standby state; if no liveness is detected in at least two consecutive detection windows within a preset total detection time, controlling the Bluetooth module to enter a sleep state.

[0016] Secondly, this application provides an in-vehicle liveness detection device based on Bluetooth phase timing perturbation, applicable to vehicles equipped with Bluetooth modules supporting Bluetooth channel detection functionality. The in-vehicle liveness detection device includes:

[0017] A response module for responding to configuring the Bluetooth module into active radar mode;

[0018] The phase timing sequence extraction module is used to transmit detection signals via Bluetooth and receive echo signals reflected from the in-vehicle environment, and extract the phase timing sequence between the transmitted signal and the echo signal.

[0019] The feature parameter extraction module is used to extract feature parameters that characterize vital signs based on phase time series sequences;

[0020] The processing module is used to determine whether there is a living person inside the vehicle based on the extracted feature parameters.

[0021] Thirdly, this application provides an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0022] Memory is used to store instructions executed by the computer;

[0023] A processor for executing computer-executable instructions stored in memory to implement any of the methods of the first aspect.

[0024] Fourthly, this application provides a vehicle, including a vehicle body and electronic equipment as described in the third aspect.

[0025] Fifthly, this application provides a computer-readable storage medium storing computer-executable instructions, which, when executed, are used to implement the method of any one of the first aspects.

[0026] Sixthly, this application provides a computer program product, including a computer program that, when executed, implements the method of any one of the first aspects.

[0027] The in-vehicle liveness detection method and related apparatus based on Bluetooth phase timing perturbation provided in this application are applied to vehicles equipped with Bluetooth modules that support Bluetooth channel detection function. The in-vehicle liveness detection method includes: in response to configuring the Bluetooth module to active radar mode, transmitting a detection signal through the Bluetooth module and receiving the echo signal reflected by the in-vehicle environment, and extracting the phase timing sequence between the transmitted signal and the echo signal; extracting feature parameters characterizing vital signs based on the phase timing sequence; and determining whether a live body exists in the vehicle based on the extracted feature parameters.

[0028] In this process, by reusing the vehicle's existing Bluetooth module that supports Bluetooth channel detection, it is automatically woken up and configured into active radar mode. It can actively transmit detection signals and analyze the echo phase timing sequence without adding any new hardware. Based on this, by extracting characteristic parameters representing vital signs, it can utilize the weak phase timing disturbances caused by vital signs to achieve dual-band joint detection, effectively distinguishing between living beings and static objects, mechanical vibrations, and environmental noise, protecting in-vehicle privacy without the need for image acquisition. Simultaneously, it does not rely on the person being left behind carrying any external Bluetooth device, effectively improving the detection blind spot when passive targets such as children are forgotten. Therefore, this application significantly improves the accuracy, anti-interference capability, and scene adaptability of in-vehicle static liveness detection without increasing hardware costs, infringing on privacy, or relying on external devices. It achieves low-cost, highly reliable in-vehicle liveness detection and promptly triggers an early warning when a live being is detected, significantly reducing the risk of safety accidents such as heatstroke caused by forgetfulness. Attached Figure Description

[0029] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0030] Figure 1 A flowchart illustrating an in-vehicle liveness detection method based on Bluetooth phase timing perturbation, provided as an exemplary embodiment of this application;

[0031] Figure 2 Another schematic flowchart of an in-vehicle liveness detection method based on Bluetooth phase timing perturbation provided for an exemplary embodiment of this application;

[0032] Figure 3 A schematic diagram of an in-vehicle liveness detection device based on Bluetooth phase timing perturbation, provided as an exemplary embodiment of this application;

[0033] Figure 4 A schematic diagram of the structure of an electronic device provided as an exemplary embodiment of this application.

[0034] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0035] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0036] The terms “first,” “second,” etc., used in the specification and claims of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, products, or apparatus.

[0037] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with relevant laws, regulations and standards, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0038] In related technologies, millimeter-wave radar, UWB chips, pressure sensor arrays, and vehicle cameras all require additional dedicated hardware, increasing vehicle manufacturing costs. Furthermore, Wi-Fi modules are typically in a low-power or completely off state after the vehicle is turned off, making continuous detection impossible. Secondly, regarding detection reliability, pressure sensors struggle to distinguish between living individuals and static objects like backpacks or shopping bags, leading to false alarms. Cameras show significantly reduced recognition rates in low light, under blanket obstructions, or at different shooting angles. Bluetooth signal strength-based solutions are heavily affected by multipath effects and human occlusion, with ranging errors reaching several meters and unreliable detection of stationary living individuals. Acoustic echo detection is susceptible to interference from wind noise, roadside construction, and other environmental noise, resulting in poor stability. Thirdly, camera solutions require continuous image acquisition from inside the vehicle, raising concerns about occupant privacy and presenting obstacles in terms of user acceptance and regulatory compliance. Finally, Bluetooth signal strength-based solutions rely on left-behind individuals carrying Bluetooth broadcast source devices such as mobile phones or wristbands, which children typically do not carry, rendering the solution completely ineffective in these core application scenarios.

[0039] Therefore, existing solutions struggle to reliably detect stationary living individuals inside a vehicle using existing vehicle resources without increasing hardware costs, infringing on privacy, or relying on external devices. The reasons for this are the following common problems: first, the need for additional dedicated sensors increases costs; second, reliance on external devices or environmental interference leads to insufficient detection reliability; and third, image acquisition raises privacy concerns. The fundamental reason is that existing technologies fail to fully utilize existing vehicle hardware resources to detect subtle signal changes caused by vital signs, particularly the signal characteristics corresponding to minute chest displacements (approximately 0.1mm~1mm) caused by breathing and heartbeat in a stationary state. Furthermore, they cannot simultaneously achieve the comprehensive requirements of detection accuracy, low false alarm rate, and low power consumption within the short window after the vehicle is locked.

[0040] To address the aforementioned technical challenges, this application proposes a novel technical concept that eliminates reliance on dedicated hardware or external devices. Instead, it utilizes Bluetooth modules already widely standard in vehicles that support Bluetooth channel detection. By analyzing the phase and timing perturbations between the transmitted and echo signals, it detects faint signs of life. To solve the aforementioned problems, this application provides an in-vehicle liveness detection scheme based on Bluetooth phase and timing perturbations. By utilizing a Bluetooth module already equipped with Bluetooth channel detection capabilities, the module is activated and configured in active radar mode after the vehicle is locked. The Bluetooth module then transmits a detection signal and receives the echo signal reflected from the vehicle's interior environment. The phase and timing sequence is extracted from the transmitted and echo signals, and characteristic parameters representing vital signs are obtained based on this sequence to determine the presence of a live person inside the vehicle. This technical approach does not rely on cameras, millimeter-wave radar, or other new sensing hardware. Instead, it utilizes the vehicle's existing Bluetooth module to perceive and analyze reflected signals inside the vehicle, enabling the system to identify a live person within a safe window after the vehicle is locked. Since this scheme focuses on the perturbation characteristics at the phase and timing level rather than simple signal intensity changes, it is more suitable for identifying weak periodic changes caused by vital signs in a static state. It can provide a more stable technical foundation for subsequent liveness detection and support timely warnings and safety linkages for vehicles. Thus, it can reliably detect liveness inside vehicles by utilizing existing vehicle resources without increasing costs or infringing on privacy.

[0041] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0042] Figure 1 This is a flowchart illustrating an in-vehicle liveness detection method based on Bluetooth phase timing perturbation, provided as an exemplary embodiment of this application. The in-vehicle liveness detection method based on Bluetooth phase timing perturbation provided in this embodiment is applied to vehicles equipped with Bluetooth modules that support Bluetooth channel detection functionality. Figure 1 As shown, the in-vehicle liveness detection method based on Bluetooth phase timing perturbation includes the following steps:

[0043] S101, In response to configuring the Bluetooth module to active radar mode.

[0044] The Bluetooth module can be understood as a wireless module in the vehicle's original electronic and electrical architecture that has Bluetooth communication capabilities and supports Bluetooth Channel Sounding (CS) function. Its hardware foundation can be an on-board Bluetooth chip with radio frequency transceiver, baseband processing unit and phase measurement capabilities. The active radar mode can be understood as the Bluetooth module no longer performing conventional pairing, connection or data transmission tasks, but periodically transmitting wireless detection signals and receiving environmental reflected echoes according to preset detection parameters to form a working state similar to short-range wireless sensing.

[0045] In practice, after the vehicle control module detects that the vehicle is powered down, all doors are closed, and the locking action is performed, it can send a wake-up command and a mode switching command to the Bluetooth module via the in-vehicle communication bus. The in-vehicle communication bus can be a controller area network bus, a local interconnection network bus, or an Ethernet communication link. After receiving the locking-related message, the Bluetooth module recovers from a low-power standby state to an operating state and loads a detection configuration file for liveness detection. This configuration file can include parameters such as detection duration, transmission power level, sampling period, channel switching strategy, echo buffer depth, and detection window length to ensure the repeatability and comparability of subsequent sampling. In one possible embodiment, the Bluetooth module immediately enters active radar mode within a preset time after locking, for example, starting the detection process within 1 to 5 seconds after the vehicle is locked, so as to complete the detection as early as possible within the high-risk time window after the owner leaves the vehicle.

[0046] To enable the Bluetooth module to perform liveness detection, this embodiment can suspend the Bluetooth module's regular external pairing broadcasts and non-essential data services after mode switching, retaining only the transmission and echo acquisition functions of the probe frames, thereby reducing interference from other wireless communication behaviors on the phase measurement results. Specifically, the Bluetooth module can lock a preset channel group or perform detection according to a predetermined frequency hopping sequence. The transmitter clock and the receiver sampling clock are synchronized based on the same reference time base to reduce phase deviation caused by local oscillator drift. The Bluetooth module can also perform a static environment initialization process before starting detection, that is, first acquire a segment of in-vehicle reflection baseline data under conditions of no active human motion input, and use this baseline data as a reference for subsequent disturbance judgment. Based on the above processing, the system can directly utilize the vehicle's original Bluetooth capabilities to enter a dedicated sensing state without adding additional hardware such as cameras, millimeter-wave radar, or pressure sensors, establishing a reliable triggering relationship between vehicle locking events and in-vehicle liveness detection processes at the system level. Based on the above analysis, this step enables the liveness detection and the vehicle locking process to be automatically linked, reducing the timeliness issues caused by relying on manual detection. Furthermore, by configuring the Bluetooth module to active radar mode, the Bluetooth hardware, originally used for communication, acquires environmental awareness capabilities, laying the sampling foundation for subsequent identification of weak vital signs of stationary live bodies.

[0047] S102. Transmit a detection signal and receive the echo signal reflected by the in-vehicle environment via the Bluetooth module, and extract the phase timing sequence between the transmitted signal and the echo signal.

[0048] The detection signal can be understood as a wireless radio frequency signal transmitted by the Bluetooth module in a predetermined frequency band and timing sequence. It can take the form of detection frames, pilot sequences, reference symbols, or adapted short burst signals supported by the Bluetooth protocol. The echo signal can be understood as the signal that returns to the Bluetooth receiver after the detection signal propagates inside the vehicle and is reflected, scattered, and superimposed by the surfaces of seats, dashboards, ceilings, windows, carpets, child seats, and other objects such as people, children, or pets that may be present in the vehicle. The phase timing sequence can be understood as the time arrangement data of the phase difference between the transmitted signal and the echo signal at multiple consecutive sampling times. It can reflect the small perturbation patterns of the in-vehicle reflection environment over time.

[0049] In its implementation, after entering active radar mode, the Bluetooth module continuously transmits detection signals at a preset sampling frequency and collects the corresponding echo signal within a preset receiving window after each transmission. Specifically, the Bluetooth module periodically transmits detection signals through a preset channel group or frequency hopping sequence and collects echo signals within a fixed receiving window. The clocks of the transmitting and receiving ends are synchronized based on the same reference time base to reduce phase deviation caused by local oscillator drift. This synchronization mechanism ensures time alignment between the transmitted signal and the echo signal, thus providing a reliable basis for subsequent phase timing sequence extraction. Since the interior of a vehicle is a relatively enclosed multipath environment, the received signal is typically composed of direct leakage components, static reflection components, and dynamic disturbance components. The dynamic disturbance components mainly originate from the slight displacement changes caused by the rise and fall of the living chest and heartbeat. The baseband processing unit of the Bluetooth module can perform down-conversion, quadrature demodulation, and complex representation on the transmitted reference signal and the received echo signal to obtain in-phase and quadrature components, and calculate the instantaneous phase difference using arctangent operation. For example, the phase difference can be expressed as "the ratio of the arctangent function acting on the received quadrature component to the received in-phase component, minus the transmit reference phase." This processing transforms the original radio frequency waveform into phase information that reflects changes in path length. When a living target is present, breathing and heartbeat cause minute changes in the propagation path. The path change and phase change approximately satisfy a linear mapping relationship; that is, the more stable the path increment, the more pronounced the periodicity in the phase timing. Therefore, phase information is more suitable than simple received intensity information for detecting vital signs in a static state.

[0050] Considering that the original phase measurements may contain double-circular jumps, carrier frequency offsets, and environmental noise, the embodiments of this application can also perform preprocessing on the obtained phase sequence. Specifically, phase unwrapping can be performed first to eliminate discontinuous abrupt changes in phase values ​​at the positive and negative circumferential boundaries; then, detrending processing can be performed to remove slow-changing components caused by temperature drift, device bias, and static structures; subsequently, median filtering, moving average filtering, or adaptive filtering can be used to suppress instantaneous spike noise; in some implementations, principal component separation or static background elimination methods can also be combined to weaken the masking effect of fixed reflectors on dynamic vital signs. The sampling frequency can be set to at least twice the frequency of the highest target vital signs to meet the sampling theorem requirements. For example, when it is necessary to cover the 2.0Hz heartbeat frequency band, the sampling frequency can be set to a value in the range of 5Hz to 50Hz, thereby balancing resolution and power consumption control. The detection window length can be set to 10 seconds to 30 seconds to accumulate a sufficient number of phase samples within a limited time for subsequent frequency band analysis and stability judgment. The result of this step is the generation of a continuous phase-time sequence dataset for subsequent processing, which is then cached in the Bluetooth module's local storage or sent to the vehicle controller for further analysis. Based on the above analysis, it can be seen that by extracting the phase-time sequence between the transmitted signal and the echo signal, the weak periodic displacement changes caused by stationary living objects in the complex reflection environment inside the vehicle can be converted into calculable and separable time-series features, thereby effectively overcoming the problem of insensitivity to stationary targets when relying solely on signal strength.

[0051] S103. Based on the phase time series, extract the characteristic parameters that characterize vital signs.

[0052] Among them, the characteristic parameters that characterize vital signs can be understood as statistical quantities, spectral quantities or time-frequency quantities used to characterize whether there are stable vital sign components in the phase time series within the corresponding frequency band, such as the dominant frequency position, spectral peak energy, in-band energy ratio, periodic stability, peak duration, envelope fluctuation degree and signal-to-noise ratio, etc.

[0053] In specific implementation, the preprocessed phase time series obtained in step S102 can be input into the frequency domain analysis module, and the spectral distribution can be obtained through Fast Fourier Transform or Short-Time Fourier Transform. To accommodate the slow fluctuations of vital signs over time, this embodiment can use a sliding window approach to segment the phase time series, calculate the spectrum of each segment separately, and then fuse the results from multiple windows. For example, the characteristic parameters representing vital signs include, but are not limited to, parameters of the respiratory frequency band and the heart rate frequency band. The respiratory frequency band can be set in the range of approximately 0.1Hz to 0.5Hz, which corresponds to the respiratory rhythm of most adults, children, and some pets in a resting state; the heart rate frequency band can be set in the range of approximately 0.8Hz to 2.0Hz, which covers the frequency range corresponding to common resting heart rates. The purpose of setting the above frequency band range is to focus on the target frequency region where vital signs may appear from complex full-band phase disturbances, reducing the impact of scattered vibrations inside the vehicle, interference from electronic devices, and environmental noise on the judgment results. Based on frequency domain analysis, the dominant frequency, peak amplitude, and peak width corresponding to the largest spectral peak within the respiratory frequency band can be extracted. Similarly, the energy of the main peak, the distribution of secondary peaks, and the energy concentration within the band can be extracted within the heartbeat frequency band. Furthermore, by employing short-time Fourier transform, the drift of the spectral peak over a continuous time window can be obtained, thus characterizing rhythm stability.

[0054] In addition to the spectral method, this embodiment can also perform bandpass filtering and envelope analysis on the phase time series. Specifically, the original phase sequence is first separated using digital bandpass filters designed for the respiratory and cardiac frequency bands, respectively, to obtain the respiratory component sequence and the cardiac component sequence. Then, the instantaneous envelope and instantaneous frequency of each component are obtained through Hilbert transform, thereby further calculating parameters such as time-domain period, fluctuation amplitude, inter-peak interval variance, and continuous period consistency. For the respiratory component, if its envelope changes smoothly and its dominant frequency continuously falls within the target interval, it can be considered that there are valid signs of breathing. For the cardiac component, if its energy is weak but maintains a stable peak value in a continuous window, it can serve as an important basis for further confirming the presence of a living organism. In one possible embodiment, a feature vector can also be constructed, and the respiratory dominant frequency, respiratory energy, cardiac dominant frequency, cardiac energy, signal-to-noise ratio within the frequency band, and multi-window stability index can be jointly input into the subsequent judgment model. For special individuals such as children, pets, or occupants in a curled-up position, their respiratory and heartbeat signals may be weak or partially offset in frequency bands. Therefore, an adaptive frequency band extension mechanism can be introduced, which involves setting an auxiliary search band of a certain width near the base band to capture vital sign components with individual differences. The result of this step is the output of structured feature parameters corresponding to the respiratory and heartbeat frequency bands, forming a feature set for liveness detection. Based on the above analysis, this step, by decomposing the weak periodic perturbations in the phase time sequence into frequency bands with clear physiological significance, achieves the conversion from raw wireless reflection signals to vital sign description parameters. This improves the interpretability of feature expression and enhances the stability and accuracy of static liveness detection.

[0055] S104. Determine whether there is a living person inside the vehicle based on the extracted feature parameters.

[0056] Among them, the liveness detection can be understood as the process of calculating and analyzing whether there is a target with vital signs in the vehicle based on the characteristic parameters corresponding to the respiratory frequency band and the heartbeat frequency band, and outputting a binary result or probability result; the warning can be understood as the prompt, alarm or linkage control action issued by the vehicle's local or remote system after confirming the presence of a live body in the vehicle. Its purpose is to remind the vehicle owner to deal with the risk of leftover live bodies in a timely manner, and to further reduce the probability of safety accidents caused by high temperature and closed environment.

[0057] In its implementation, the controller can receive the feature parameters output in step S103 and complete the liveness determination based on preset rules or a trained classification model. Under rule-based determination, scores can be given for whether the respiratory frequency falls within a set range, whether the respiratory peak energy exceeds a threshold, whether there is a stable peak within the heartbeat frequency band, and whether the consistency of features across multiple consecutive windows meets requirements. When the comprehensive score exceeds the determination threshold, a result indicating the presence of a live person is output. Under model-based determination, the aforementioned feature vectors can be input into a random forest model, support vector machine model, logistic regression model, or lightweight neural network model, and the model outputs the probability of a live person's presence. The probability threshold can be set to a fixed value or dynamically adjusted based on environmental conditions. For example, when the vehicle interior temperature is high, signal stability is good, and obvious respiratory features appear in two consecutive detection windows, the trigger threshold can be appropriately lowered to improve detection timeliness. When there is strong residual mechanical vibration or wireless interference in the detection environment, the threshold can be increased, requiring at least one feature—breathing or heartbeat—to continuously appear in multiple windows to control the false alarm rate. To enhance robustness, a multi-window confirmation mechanism can also be adopted in this embodiment, that is, the judgment results of several consecutive detection windows are voted on or weighted averaged, and the final conclusion that a living organism exists is only output when the cumulative result meets the confirmation condition.

[0058] Upon detecting the presence of a living being, the system can immediately trigger an early warning process. This process can include local vehicle-side audio and visual alerts, horn activation, flashing lights, and infotainment system displays, or sending SMS messages, push notifications, or phone calls to the owner's mobile device via a vehicle-to-everything (V2X) platform. In vehicles equipped with features like air conditioning waste heat management, window control, or thermal management linkage, safety linkages can also be triggered, such as briefly activating fresh air, lowering the interior temperature, or partially opening windows to buy time for rescue. The warning output can also include detection confidence level, detection timestamp, and vehicle location information for quick assessment and response by the owner. In one possible embodiment, detection does not immediately stop after the initial warning trigger but continues for several subsequent monitoring cycles to provide repeated alerts as long as a living being remains. The monitoring process ends when the living being leaves or the door is reopened. If no living being is detected, the active radar mode can be disabled, and the Bluetooth module can be restored to low-power standby mode to reduce overall vehicle static current consumption. Based on the above analysis, it can be seen that this step reduces misjudgments caused by relying solely on a single frequency band or a single sampling result by jointly judging the characteristics of the respiratory frequency band and the heart rate frequency band. At the same time, by combining it with local early warning, remote notification and vehicle linkage control, the detection results can be promptly transformed into safety response actions, thereby forming a complete risk protection closed loop within the limited time window after the vehicle is locked.

[0059] The in-vehicle liveness detection method based on Bluetooth phase timing perturbation provided in this application reuses the vehicle's existing Bluetooth module that supports Bluetooth channel detection, automatically wakes it up, and configures it in active radar mode. It can actively transmit detection signals and analyze the echo phase timing sequence without adding any new hardware. Based on this, by extracting characteristic parameters representing vital signs, it can utilize the weak phase timing perturbations caused by vital signs to achieve dual-band joint detection, effectively distinguishing between live bodies and static objects, mechanical vibrations, and environmental noise, protecting in-vehicle privacy without the need for image acquisition. Simultaneously, it does not rely on the person being left behind carrying any external Bluetooth device, effectively improving the detection blind spot when passive targets such as children are forgotten. Therefore, this application significantly improves the accuracy, anti-interference capability, and scene adaptability of in-vehicle static liveness detection without increasing hardware costs, infringing on privacy, or relying on external devices. It achieves low-cost, highly reliable in-vehicle liveness detection and promptly triggers an early warning when a live body is detected, significantly reducing the risk of safety accidents such as heatstroke caused by forgetfulness.

[0060] In some embodiments, based on the phase time series, feature parameters characterizing vital signs are extracted, including: performing phase unwinding processing on the phase time series to restore continuous phase; performing high-pass filtering on the unwound phase series to filter out static bias and temperature drift to obtain dynamic phase perturbation signals; and extracting feature parameters from the dynamic phase perturbation signals.

[0061] Among them, the phase timing sequence refers to the phase sampling sequence that changes over time after the Bluetooth module transmits a detection signal and receives the reflected echo from inside the vehicle, and then performs phase calculation. The continuous phase is used to characterize the weak periodic disturbances on the reflection path inside the vehicle. The dynamic phase disturbance signal refers to the low-amplitude phase change component retained after high-pass filtering, which mainly corresponds to the periodic displacement information caused by human vital signs such as breathing and heartbeat.

[0062] In practical processing, phase wrap detection is first performed on the phase time sequence. When the phase difference between adjacent sampling points exceeds a preset threshold, such as π, the phase shift is compensated in multiples of 2π to restore the phase abrupt change that originally spanned the main value interval to a continuous curve. The high-pass filter can be implemented using a digital filter, with the filter cutoff frequency set below the lower limit of the breathing band (0.1Hz), for example, 0.05Hz, to suppress low-frequency drift caused by static reflection, device bias, and gradual temperature changes. The filter can be constructed from first-order or multi-order infinite impulse response (IIR) filters, or it can be implemented from finite impulse response (FIR) filters. In practical applications, other models of this component can also be selected, and this application embodiment does not limit this.

[0063] After obtaining the dynamic phase perturbation signal, characteristic parameters of the respiratory frequency band and the heartbeat frequency band can be extracted based on a preset frequency band range. The respiratory frequency band is used to characterize low-frequency periodic perturbations in the range of approximately 0.1Hz to 0.5Hz, and the heartbeat frequency band is used to characterize high-frequency weak perturbations in the range of approximately 0.8Hz to 2.0Hz. The characteristic parameters may include the dominant frequency position, spectral peak amplitude, in-band energy, spectral peak stability, or time-frequency envelope statistics. They can be extracted by combining short-time Fourier transform, wavelet analysis, or frequency domain transform after bandpass filtering to enhance the ability to separate the vital signs of a static living organism.

[0064] This processing method, by first restoring continuous phase and then suppressing slow-varying bias, can separate the static reflection components in the in-vehicle environment from vital sign disturbances, allowing the frequency band characteristics corresponding to breathing and heartbeat to be stably displayed. Since the extracted objects come from the same phase time sequence, the system can complete the identification of weak vital signs without adding additional sensing hardware, and provide more reliable input for subsequent liveness detection.

[0065] By adopting the above method, encapsulation jumps and low-frequency drifts in the phase sequence are effectively eliminated, improving the separability of vital signs and thus enhancing the accuracy and stability of in-vehicle static liveness detection. Simultaneously, this scheme is highly adaptable to changes in the in-vehicle reflective environment, reducing the probability of false alarms and false negatives, and facilitating rapid determination within the limited time after vehicle locking.

[0066] In some embodiments, the feature parameters include parameters of the respiratory frequency band and the heart rate frequency band. Extracting feature parameters from the dynamic phase perturbation signal includes: performing time-frequency transformation on the dynamic phase perturbation signal to obtain a time-frequency distribution; extracting the respiratory dominant frequency, respiratory energy, and respiratory signal-to-noise ratio from the time-frequency distribution within a preset respiratory frequency band; performing bandpass filtering and envelope extraction on the dynamic phase perturbation signal within a preset heart rate frequency band, and performing frequency domain transformation on the extracted envelope to obtain the heart rate dominant frequency and heart rate energy; and outputting a five-dimensional feature vector containing the respiratory dominant frequency, respiratory energy, respiratory signal-to-noise ratio, heart rate dominant frequency, and heart rate energy.

[0067] The respiratory frequency band can be understood as the range of low-frequency phase disturbances corresponding to the slow, periodic movements of the chest and abdomen caused by breathing in humans or animals; the heartbeat frequency band can be understood as the range of higher-frequency phase disturbances caused by heartbeats and related micro-vibrations. This application utilizes, for example, Bluetooth 6.0 channel detection functionality to continuously acquire the phase timing of a single channel; the dynamic phase disturbance signal refers to the baseband signal reflecting human micro-movements obtained from the phase timing sequence after continuous phase recovery and static bias suppression. It can characterize the weak periodic changes caused by a stationary living body inside the vehicle. Considering that when a living body is present in the vehicle, the micro-displacement of the chest cavity caused by breathing and heartbeat (e.g., 0.1mm-1mm) will cause periodic micro-jitter (e.g., ±1°-±10°) in the phase of the reflection path, the presence of a living body can be determined by analyzing the periodic disturbances in the phase timing. Unlike Bluetooth positioning schemes in related technologies, this application embodiment does not focus on phase difference, but only on the timing jitter of a single phase, thereby achieving a technological leap from a "positioning tool" to a "sensing sensor". Compared to Bluetooth signal strength indicators, phase signals have a higher signal-to-noise ratio and better stability; compared to radar, UWB or Wi-Fi solutions, this solution reuses the vehicle's existing digital key Bluetooth hardware, achieving zero-cost incremental improvements.

[0068] This application extracts periodic features from two frequency bands simultaneously from a phase perturbation sequence to achieve joint detection of respiration and heartbeat across two frequency bands. The time-frequency transformation can be implemented using a short-time Fourier transform, and its output time-frequency distribution characterizes the spectral energy distribution at different times. A preset respiration frequency band covers the low-frequency range (e.g., 0.1Hz~0.5Hz) corresponding to resting respiration, while a preset heartbeat frequency band covers the higher-frequency range (e.g., 0.8Hz~2.0Hz) caused by heartbeats. A five-dimensional feature vector carries the joint characterization results of respiration and heartbeat, facilitating subsequent liveness detection.

[0069] The advantages of using the above-mentioned dual-band joint detection in this application are as follows: First, children's respiratory rate is relatively high (e.g., 0.5Hz~0.8Hz), which partially overlaps with the adult heart rate frequency band. The dual-band setting can cover a wider range of living organisms, including adults, children, and infants. Second, the energy of mechanical vibrations inside the vehicle (e.g., residual fan rotation, slight vehicle shaking) is usually concentrated in the ultra-low frequency band below 0.1Hz. By setting the lower limit of the respiratory frequency band to 0.1Hz, such mechanical interference can be effectively distinguished. Third, the heart rate of dogs is about 1.5Hz~2.5Hz, and the heart rate of cats is about 2Hz~3Hz. The above-mentioned heart rate frequency band range can also cover the detection of vital signs of common pets, further expanding the scope of application of this application.

[0070] Based on the above principles, in specific implementation, after obtaining the dynamic phase perturbation signal, the system first segments it according to a preset window length and performs time-frequency transformation. The time-frequency transformation can be implemented using short-time Fourier transform, and its output time-frequency distribution is used to characterize the spectral energy distribution at different times. The preset breathing frequency band is used to cover the low-frequency range corresponding to human resting respiration (e.g., 0.1Hz~0.5Hz). The preset heartbeat frequency band is used to cover the higher-frequency range caused by heartbeat (e.g., 0.8Hz~2.0Hz). The five-dimensional feature vector is used to carry the joint characterization results of respiration and heartbeat, which is convenient for subsequent liveness determination.

[0071] When extracting respiratory frequency band features, the system searches for continuous peak regions within the respiratory frequency band, determines the frequency corresponding to the peak as the dominant respiratory frequency, and accumulates the spectral energy of the neighborhood of this frequency to obtain the respiratory energy. Simultaneously, the respiratory signal-to-noise ratio (SNR) is calculated as the ratio of the target peak energy to the nearby background energy. When extracting heartbeat frequency band features, the system first performs bandpass filtering on the dynamic phase disturbance signal to suppress low-frequency interference from respiration and the environment. Then, it extracts the envelope from the filtered result, which can be obtained through Hilbert transform or rectification followed by smoothing. Subsequently, it performs frequency domain transformation on the envelope and determines the dominant peak frequency and corresponding spectral energy within the heartbeat frequency band. After normalizing the extracted features, they are arranged into a five-dimensional feature vector in a fixed order and output to the subsequent decision module.

[0072] For example, the preprocessed phase signal Perform a short-time Fourier transform with a window length of, for example, 10 seconds and an overlap rate of 50%; extract the dominant respiratory frequency within the respiratory frequency band (0.1Hz~0.5Hz). Breathing energy and signal-to-noise ratio Simultaneously, the signal is bandpass filtered from 0.8 to 2.0 Hz, the envelope is extracted using Hilbert transform, and the heart rate is extracted using fast Fourier transform. and heartbeat energy Output a five-dimensional feature vector for each 10-second time window. .

[0073] It should be noted that the window length of 10 seconds, the overlap rate of 50%, and the output of the five-dimensional feature vector described in the above embodiments are merely specific examples provided to facilitate understanding of the technical solution of this application, and are not necessary conditions for implementing this application. In actual implementation, the above parameters can be reasonably selected and adjusted according to factors such as the characteristics of the in-vehicle environment, the sampling frequency of the Bluetooth module, the target detection accuracy, and the system power consumption budget, and are not limited here.

[0074] The above processing method enhances the distinguishability of respiratory features by utilizing time-frequency distribution, while improving the detectability of heartbeat features through bandpass filtering and envelope analysis. This allows the dominant respiratory frequency, respiratory energy, respiratory signal-to-noise ratio, dominant heartbeat frequency, and heartbeat energy to be stably expressed in a unified feature space, thereby improving the accuracy, anti-interference ability, and scene adaptability of in-vehicle static liveness detection. Since feature extraction is completed using Bluetooth phase perturbation signals, there is no need to deploy additional cameras, radar, or pressure sensors, which helps reduce overall vehicle costs and maintain in-vehicle privacy.

[0075] Based on the above embodiments, in some embodiments, determining whether there is a living person inside the vehicle based on the extracted feature parameters includes: inputting the extracted feature parameters into a pre-trained classification model, the classification model outputting the probability of the presence of a living person; and determining whether there is a living person inside the vehicle based on the probability of the presence of a living person.

[0076] Among them, the feature parameters are the liveness discrimination parameters extracted based on the in-vehicle phase temporal disturbance signals, which are used to characterize the periodic micro-motion features in the breathing frequency band and the heartbeat frequency band; the pre-trained classification model is a supervised model trained in an offline environment using labeled in-vehicle live and non-live samples. The model input is a feature vector composed of feature parameters, and the output is the probability value of the presence of a live body in the current detection window.

[0077] During the model training phase, the system pre-collects a large number of positive and negative samples for model training. Positive samples are collected from scenarios with people inside the vehicle, covering different seat positions, occupant postures, ages, and body types. Negative samples are collected from scenarios with empty vehicles, covering different ambient temperatures, parking lot types, and vehicle models. Each sample corresponds to a five-dimensional feature vector extracted within a detection window (e.g., a 10-second time window) and its corresponding label. There are no fewer than 5000 positive and negative samples. Based on the collected training dataset, a random forest algorithm is used for model training to learn the mapping relationship between feature parameters and the probability of a live object. In practical applications, the classification model can also use a support vector machine, lightweight neural network, or gradient boosting tree model to perform probability inference at the vehicle end with lower computational overhead. This embodiment does not limit the specific implementation.

[0078] Accordingly, after inputting the feature parameters into the classification model, the model performs feature mapping and discrimination calculations on the input vector and outputs a probability value between 0 and 1. The higher the probability value, the higher the likelihood of a live object being present in the current detection window. The system pre-sets a threshold, such as 0.7. When the probability of a live object being present output by the model reaches or exceeds this threshold, the current detection window is marked as containing a live object, thus converting the continuous probability result into a discrete window judgment result. The threshold can be configured in conjunction with the false alarm rate and the false negative rate, and can be calibrated according to different vehicle models, different in-vehicle temperature conditions, or different signal quality. In practical applications, the threshold can also be set in other ways according to calibration requirements; this embodiment does not limit this.

[0079] With this specific implementation method, the system can perform probabilistic identification of live objects inside the vehicle using a pre-trained classification model without relying on additional cameras or millimeter-wave radar. It can also improve detection reliability through threshold determination, thereby outputting more stable detection results within a limited time after the car door is locked, further enhancing the practicality and safety of detecting live objects left inside the vehicle.

[0080] In some embodiments, the in-vehicle liveness detection method begins execution in response to a vehicle locking signal, and the execution duration is limited to a preset total detection duration, which includes at least two detection windows. If no liveness is detected in the vehicle within the preset total detection duration, the Bluetooth module is controlled to enter a sleep state. Determining whether a liveness exists in the vehicle based on the probability of liveness includes: determining that a liveness exists in the current detection window when the probability of liveness is greater than or equal to a preset threshold; and determining that a liveness exists in the vehicle when at least two consecutive detection windows detect the presence of a liveness.

[0081] The vehicle locking signal is used to characterize the security detection scenario after the vehicle has been locked and exited. This signal can be generated by the body control module based on remote locking commands, automatic locking results from the keyless entry system, or locking commands from the central control unit. After the signal is valid, the Bluetooth module is activated and enters the detection working state. The Bluetooth module can use an in-vehicle communication unit that supports Bluetooth channel detection. In the working state, it transmits detection signals and receives echo signals reflected from the cabin interior, windows, and the surface of the person's body. Then, it samples and analyzes the phase timing disturbance between the transmission and the echo. The preset total detection time consists of multiple continuous detection windows. Each detection window corresponds to a fixed length or an acquisition cycle that is adaptively set according to the vehicle status. Multiple detection windows are used to cover the weak phase changes caused by breathing and heartbeat, so as to improve the stability of identifying stationary living bodies.

[0082] This application features a targeted design for the detection time window. Specifically, the preset total detection duration is matched to the rate at which the thermal risk inside the vehicle increases after the engine is turned off, ensuring that the Bluetooth module maintains detection only within the necessary timeframe, thus balancing timely response with overall vehicle power consumption control. The selection of this preset total detection duration (e.g., 3 minutes) is based on the following considerations: this duration is sufficient to complete 10 to 20 full breathing cycles of sampling, meeting the frequency resolution requirements of frequency domain analysis (e.g., Fast Fourier Transform); within this duration, even in high-temperature environments, the interior temperature has not significantly risen to a dangerous level; and the owner is typically not far from the vehicle after locking it, enabling timely response to warnings. Unlike existing solutions that continuously monitor for several hours, this application emphasizes "instant detection and instant response," with the system automatically going into sleep mode after the detection window ends, resulting in negligible power consumption.

[0083] Based on the aforementioned time window design, when the feature parameters obtained within any detection window meet the liveness detection criteria, the system maintains detection or outputs an early warning. For example, after completing the judgment of the current detection window, the system continuously confirms the judgment results of adjacent detection windows. To reduce false triggering due to instantaneous noise, the final conclusion that a liveness exists in the vehicle is only output when at least two consecutive detection windows are determined to have a liveness. For instance, the presence of a liveness is only confirmed when the judgment results of three consecutive detection windows are all ≥0.7. This continuous confirmation mechanism can suppress single-window misjudgments caused by vehicle vibration, short-term wireless interference, or occasional noise, resulting in higher stability of the final judgment result. The detection window can be divided into several time slices of fixed duration. Windows can be continuously spliced ​​or generated using an overlapping sliding method. In practical applications, other forms of window organization can also be selected, and this embodiment does not limit this.

[0084] The working principle of the above method is that the classification model first gives the probability of a liveness detection within a single window based on feature parameters, then converts the probability result into a window-level judgment through a threshold, and finally completes the vehicle-level liveness detection by confirming the consistency of at least two consecutive windows. This processing method allows the system to maintain a fast response speed while helping to reduce the false alarm probability caused by single noise fluctuations and enhancing the stability of identifying stationary liveness. Through the above time window design, the proposed solution achieves low-power operation while ensuring detection reliability. Compared with the "continuous monitoring for several hours" method in related technologies, it is more suitable for short-term, real-time detection scenarios after the vehicle is turned off and locked.

[0085] In this embodiment, detection can be initiated immediately after the vehicle is locked, and by limiting the total duration and the number of detection windows, the liveness detection can be completed within a controllable time period. In addition, since the detection process relies on existing Bluetooth communication hardware, no additional sensors are required, which helps to reduce the overall vehicle cost and improve the convenience of vehicle system integration.

[0086] In some embodiments, the in-vehicle liveness detection method based on Bluetooth phase timing perturbation further includes: in response to determining that a live person exists in the vehicle, performing at least one of the following warning operations: controlling the in-vehicle display device to display warning information and controlling the in-vehicle audio system to play voice prompts; pushing alarm notifications to a preset mobile application and controlling the vehicle to issue an audible and visual alarm; sending help requests to preset emergency contacts or rescue platforms and controlling the vehicle's air conditioning system to turn on external circulation, and controlling the windows to open a preset gap when it is detected that there is no rain outside the vehicle.

[0087] For example, in response to the determination of a living being inside the vehicle, a Level 1 warning is executed: the in-vehicle display device is controlled to display a warning message, and the in-vehicle audio system is controlled to play a voice prompt. The in-vehicle display device can be a central control screen, instrument panel display, or head-up display terminal, used to display warning icons and text (e.g., "Children in the vehicle"); the in-vehicle audio system can include central control speakers, door panel speakers, or independent buzzer units, used to play the voice prompt; the cancellation command can be issued by the vehicle owner through a digital key application, vehicle-mounted interface, or remote control terminal, used to terminate subsequent warnings after confirming the risk has been eliminated; the preset mobile application can be a vehicle-linked security management application, and the preset emergency contacts or rescue platform can include the vehicle owner's preset contacts, a roadside assistance platform, or the vehicle brand's remote service center.

[0088] In this solution, once the liveness detection module outputs a result indicating the presence of a live subject, the warning control module immediately invokes the body controller and infotainment system to execute a Level 1 warning, without waiting for the detection window to close. Specifically, the central control screen displays a warning icon and text (such as "Child in vehicle"), and the speakers play a voice prompt, providing the driver with the most intuitive way to take timely action.

[0089] Accordingly, if no cancellation command is received within the first preset time after the Level 1 warning is issued, a Level 2 warning is issued: an alarm notification is pushed to a preset mobile application, and the vehicle is controlled to emit an audible and visual alarm. For example, if no cancellation command is received through a mobile application or door unlocking within the first preset time (e.g., 30 seconds), the warning control module issues a Level 2 warning: an alarm notification is pushed to a preset mobile application, containing information such as the vehicle's location and the type of any living person left behind; simultaneously, the vehicle horn is controlled to sound intermittently, and the hazard lights are controlled to flash to attract the attention of the vehicle owner and surrounding personnel.

[0090] Furthermore, if no release command is received within the second preset time after the Level 1 warning is executed, or if the total duration after the vehicle is locked exceeds the preset time and a living person still exists, a Level 3 warning is executed: the vehicle's air conditioning system is controlled to turn on external circulation; if no rain is detected outside the vehicle, the windows are controlled to open to a preset gap; and a request for help is sent to a preset emergency contact or rescue platform. For example, if no release command is received within the second preset time (e.g., 3 minutes), or if the total duration after the vehicle is locked exceeds the preset time (e.g., 3 minutes) and a living person is still detected, the warning control module executes a Level 3 warning: the vehicle's air conditioning system is controlled to turn on external circulation and set a target temperature (e.g., 22°C); if no rain is detected outside the vehicle, the windows are controlled to lower to a predetermined height (which can be set to 2cm or a limited opening); simultaneously, a request for help is sent to a preset emergency contact or third-party rescue platform. The request for help includes the vehicle's Global Positioning System (GPS) coordinates, vehicle identification number, and alarm time, so that rescue personnel can quickly locate the vehicle.

[0091] Correspondingly, if no living entity is confirmed within the preset total detection time (e.g., 3 minutes after the vehicle is locked), the Bluetooth module automatically enters sleep mode, reducing power consumption to the microwatt level. In the above-mentioned tiered warning mechanism, after a Level 1 warning is triggered, if the vehicle owner returns promptly and unlocks the door or confirms the alarm cancellation via a mobile application, the warning stops and the vehicle returns to standby mode; if the vehicle owner does not respond within the specified time, the warning intensity gradually escalates.

[0092] It should be noted that there is no strict order or hierarchy among the various warning operations mentioned above. In practical applications, one or more of the warning operations can be selected and combined based on vehicle configuration, environmental conditions, and user-preset strategies. For example, when a child is detected lingering and the outside temperature is high, the system can simultaneously push an alarm notification to the mobile application, activate the vehicle's audible and visual alarms, and turn on the air conditioning's external circulation and window ventilation. When a pet is detected lingering and the ambient temperature is suitable, only the alarm notification to the mobile application needs to be pushed. These warning operations can be executed in parallel or sequentially at preset time intervals. The specific execution method can be adaptively determined by the system based on the actual scenario or pre-configured by the user; this application does not impose any limitations on this.

[0093] This embodiment of the application utilizes the vehicle's existing human-machine interaction, body control, and remote communication capabilities to transform the results of liveness detection into a coordinated control system that combines local alerts, remote notifications, and emergency response. This allows vehicle owners to quickly become aware of risks and continuously enhances warning intensity and buys time for rescue when no one responds. By adopting this specific implementation method, risk escalation processing from minor to severe can be achieved without relying on additional sensing hardware. This helps reduce the probability of safety accidents caused by live objects left inside the vehicle due to factors such as high temperatures and lack of oxygen, and improves the timeliness, accessibility, and completeness of vehicle-wide warnings and responses.

[0094] In some embodiments, the in-vehicle liveness detection method based on Bluetooth phase timing perturbation further includes: in response to a door unlocking signal or a release command received through a mobile application, stopping the current warning operation and controlling the Bluetooth module to enter a standby state; or, if no situation occurs within a preset total detection time where at least two consecutive detection windows determine the presence of a liveness, controlling the Bluetooth module to enter a sleep state.

[0095] The door unlock signal can be an unlock status signal output by the vehicle door lock controller, or an unlock confirmation signal issued by the digital key, mobile phone Bluetooth authorization, or in-vehicle central control system; the release command received through the mobile application is a warning release command sent to the vehicle controller via the vehicle network communication link after the owner completes identity verification on the mobile terminal.

[0096] For example, when the vehicle detects that a door has been legally unlocked, or receives a valid unlock command, the warning control unit immediately terminates any currently running audible and visual alarms, voice prompts, remote push notifications, or vehicle linkage controls, and cancels the timers and recurring trigger conditions associated with that warning level. Subsequently, a control command is sent to the Bluetooth module, causing it to disable continuous detection and echo acquisition, retaining only minimal wake-up listening capabilities, thus putting the Bluetooth module into a low-power standby mode. In standby mode, the Bluetooth module retains its communication address, detection configuration parameters, and the most recent detection result, allowing for quick resumption of operation the next time the vehicle is locked.

[0097] In practical applications, the warning cancellation control can be executed by the body controller, communication controller, or independent liveness detection control unit. The controller clears the warning level through a software state machine and sends a mode switching command to the Bluetooth module. To avoid misunderstandings, the cancellation command sent by the mobile application can undergo signature verification or token verification before entering the vehicle control system. Only after successful verification is the current warning level allowed to stop. After the Bluetooth module enters standby mode, it can still respond to new locking signals and switch back to active radar mode. This application embodiment does not limit this.

[0098] Through this mechanism, when the vehicle owner has confirmed that the vehicle is safe or has opened the door to handle the situation, the continuous alarm can be stopped in a timely manner, and the Bluetooth module can be switched to a low-power state. This reduces unnecessary power consumption and communication overhead, while also minimizing the interference caused by continued alarm output. This solution also allows the vehicle to quickly return to a standby state ready for re-detection after the alarm is deactivated, contributing to improved response efficiency and ease of use of the overall vehicle safety protection system.

[0099] Furthermore, if no liveness detection result is obtained within the preset total detection time (e.g., 3 minutes), the control module sends a sleep command to the Bluetooth module, causing it to exit the active sensing state and enter a low-power sleep mode. Conversely, if no liveness is confirmed within the preset total detection time, the Bluetooth module automatically enters sleep mode, reducing power consumption to the microwatt level. This achieves low-power operation while ensuring detection reliability, making it more suitable for short-term, real-time detection scenarios after the vehicle is turned off and locked, compared to the "continuous monitoring for several hours" method in related technologies.

[0100] Figure 2 Another flowchart illustrating an in-vehicle liveness detection method based on Bluetooth phase timing perturbation, provided as an exemplary embodiment of this application. Figure 2 As shown, the in-vehicle liveness detection method based on Bluetooth phase timing perturbation includes the following steps:

[0101] S201. In response to the vehicle locking signal, the Bluetooth module is woken up and configured to active radar mode.

[0102] For example, the process begins in response to a vehicle locking signal, and the execution duration is limited to a preset total detection time, which includes at least two detection windows. The preset total detection time is, for example, 3 minutes.

[0103] S202. Transmit a detection signal and receive the echo signal reflected by the in-vehicle environment via the Bluetooth module, and extract the phase timing sequence between the transmitted signal and the echo signal.

[0104] S203. Perform preprocessing operations on the phase time sequence, including phase unwinding and high-pass filtering, to obtain the dynamic phase perturbation signal.

[0105] S204. Extract the characteristic parameters of the respiratory frequency band and the heartbeat frequency band from the dynamic phase perturbation signal, respectively.

[0106] In some embodiments, feature parameters of the respiratory frequency band and the heartbeat frequency band are extracted from the dynamic phase perturbation signal, including: performing time-frequency transformation on the dynamic phase perturbation signal to obtain a time-frequency distribution; extracting the respiratory dominant frequency, respiratory energy, and respiratory signal-to-noise ratio from the time-frequency distribution within a preset respiratory frequency band; performing bandpass filtering and envelope extraction on the dynamic phase perturbation signal within a preset heartbeat frequency band, and performing frequency domain transformation on the extracted envelope to obtain the heartbeat dominant frequency and heartbeat energy; and outputting a five-dimensional feature vector containing the respiratory dominant frequency, respiratory energy, respiratory signal-to-noise ratio, heartbeat dominant frequency, and heartbeat energy.

[0107] S205. Determine whether there is a living person inside the vehicle based on the extracted feature parameters.

[0108] Specifically, the extracted feature parameters are input into a pre-trained classification model, and the classification model outputs the probability of the presence of a live object. When the probability of the presence of a live object is greater than or equal to a preset threshold, it is determined that there is a live object in the current detection window. When at least two consecutive detection windows are determined to have a live object, it is determined that there is a live object in the vehicle.

[0109] If so, execute S206;

[0110] If not, proceed to S208.

[0111] S206. Implement tiered early warning systems.

[0112] Specifically, in response to the determination of a living being inside the vehicle, a Level 1 warning is executed: the vehicle display device is controlled to show a warning message, and the vehicle audio system is controlled to play a voice prompt; if no release command is received within the first preset time after the Level 1 warning is executed, a Level 2 warning is executed: an alarm notification is pushed to a preset mobile application, and the vehicle is controlled to issue an audible and visual alarm; if no release command is received within the second preset time after the Level 1 warning is executed, or if the total duration after the vehicle is locked exceeds the preset time and a living being still exists, a Level 3 warning is executed: the vehicle's air conditioning system is controlled to turn on external circulation, and if no rain is detected outside the vehicle, the windows are controlled to open to a preset gap, and a request for help is sent to a preset emergency contact or rescue platform.

[0113] S207, in response to a door unlock signal or a release command received via a mobile application, stops executing the current warning level and controls the Bluetooth module to enter standby mode.

[0114] S208. If no living being is detected in the vehicle within the preset total detection time, the Bluetooth module is controlled to enter sleep mode.

[0115] In summary, this application has at least the following advantages:

[0116] I. By utilizing the Bluetooth module that already comes standard with the vehicle and supports Bluetooth channel detection, in-vehicle liveness detection can be achieved without adding any dedicated hardware such as millimeter-wave radar, UWB chip, pressure sensor array, or vehicle camera. In existing solutions, millimeter-wave radar, UWB chip, pressure sensor, and vehicle camera all require additional dedicated hardware. This application, however, achieves liveness detection without adding any hardware by reusing the vehicle's existing digital key Bluetooth hardware.

[0117] Second, this application focuses on the perturbation characteristics at the phase timing level rather than simple signal strength changes. When a living person is present in the vehicle, the micro-displacement of the chest cavity caused by breathing and heartbeat (approximately 0.1 to 1 mm) will cause periodic micro-jitter (approximately ±1° to ±10°) in the phase of the reflection path. By extracting the phase timing sequence between the transmitted signal and the echo signal, these subtle changes can be effectively captured. Compared with Bluetooth signal strength indication schemes, which are affected by multipath effects and human body obstruction, have ranging errors of up to several meters, and cannot reliably detect static living persons, the phase signal used in this application has a higher signal-to-noise ratio and stability, and can reliably detect breathing and heartbeat characteristics in a static state.

[0118] Third, this application simultaneously extracts periodic features of the respiratory frequency band (0.1Hz to 0.5Hz) and the heart rate frequency band (0.8Hz to 2.0Hz) from the phase perturbation sequence, and inputs them into the classification model to determine the presence of a living being. This dual-frequency band design has the following advantages: First, children's respiratory frequency is relatively high (0.5Hz to 0.8Hz), which partially overlaps with the adult heart rate frequency band. The dual-frequency band can cover a wider range of living beings, including adults, children, and infants. Second, the energy of mechanical vibrations inside the vehicle (such as residual fan rotation) is concentrated in the ultra-low frequency band below 0.1Hz, which can be effectively distinguished by high-pass filtering, reducing the false alarm rate. Third, the heart rate of dogs is approximately 1.5Hz to 2.5Hz, and the heart rate of cats is approximately 2Hz to 3Hz. The heart rate frequency band range of this application can cover the detection of vital signs in common pets, expanding the application scenarios.

[0119] Fourth, by introducing an active radar mode based on Bluetooth channel detection, the vehicle's Bluetooth module actively transmits detection signals and receives echoes reflected from the human body, analyzing the phase change between transmission and reception to detect vital signs. This mode does not rely on the abandoned person carrying any external Bluetooth devices (such as mobile phones or wristbands), fundamentally improving the application blind spot of existing Bluetooth signal strength indication schemes that fail due to children not carrying devices, and achieving reliable detection of passive targets. Furthermore, being entirely based on phase perception of Bluetooth wireless signals, it eliminates the need for in-vehicle cameras and does not collect any image information, fundamentally reducing the risk of occupant privacy leaks. Compared to camera-based solutions that require continuous in-vehicle image acquisition and suffer from significantly reduced recognition rates in dim lighting or obstructed conditions, this application is privacy-friendly and unaffected by lighting conditions or obstructions.

[0120] Fifth, by limiting the liveness detection window to 3 minutes after the vehicle is locked, unlike existing solutions that monitor continuously for several hours, this application emphasizes immediate detection and response. The system automatically enters hibernation after the detection window ends, reducing power consumption to the microwatt level, achieving low-power operation while ensuring detection reliability. Furthermore, by adopting a tiered early warning system and utilizing the vehicle's existing human-machine interaction, body control, and remote communication capabilities, a progressive risk handling approach is implemented, from minor to severe, effectively improving the timeliness of early warnings and the completeness of response.

[0121] The following are embodiments of the apparatus described in this application, which can be used to execute the embodiments of the method described in this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the method described in this application.

[0122] Figure 3 This is a schematic diagram of an in-vehicle liveness detection device based on Bluetooth phase timing perturbation, provided as an exemplary embodiment of this application. Figure 3 As shown, the in-vehicle liveness detection device 30 based on Bluetooth phase timing perturbation includes a response module 31, a phase timing sequence extraction module 32, a feature parameter extraction module 33, and a processing module 34, wherein:

[0123] Response module 31 is configured to respond to configuring the Bluetooth module into active radar mode;

[0124] The phase timing sequence extraction module 32 is used to transmit a detection signal and receive an echo signal reflected by the in-vehicle environment via a Bluetooth module, and extract the phase timing sequence between the transmitted signal and the echo signal.

[0125] The feature parameter extraction module 33 is used to extract feature parameters representing vital signs based on the phase time series sequence;

[0126] Processing module 34 is used to determine whether there is a living person inside the vehicle based on the extracted feature parameters.

[0127] In one possible implementation, the phase time series extraction module 32 can be specifically used to: perform phase unwinding processing on the phase time series to restore continuous phase; perform high-pass filtering on the unwound phase series to filter out static bias and temperature drift to obtain dynamic phase perturbation signal; and extract feature parameters characterizing vital signs from the dynamic phase perturbation signal.

[0128] In one possible implementation, the feature parameters include parameters of the respiratory frequency band and the heartbeat frequency band. The phase time sequence extraction module 32 can also be used to: perform time-frequency transformation on the dynamic phase disturbance signal to obtain a time-frequency distribution; extract the respiratory dominant frequency, respiratory energy, and respiratory signal-to-noise ratio from the time-frequency distribution within a preset respiratory frequency band; perform bandpass filtering and envelope extraction on the dynamic phase disturbance signal within a preset heartbeat frequency band, and perform frequency domain transformation on the extracted envelope to obtain the heartbeat dominant frequency and heartbeat energy; and output a five-dimensional feature vector containing the respiratory dominant frequency, respiratory energy, respiratory signal-to-noise ratio, heartbeat dominant frequency, and heartbeat energy.

[0129] In one possible implementation, the processing module 34 can be specifically used to: input the extracted feature parameters into a pre-trained classification model, and the classification model outputs the probability of the presence of a live body; based on the probability of the presence of a live body, determine whether there is a live body inside the vehicle.

[0130] In one possible implementation, the response module 31 can also be used to: start execution in response to a vehicle locking signal, and the execution time is limited to a preset total detection time, which includes at least two detection windows; determine that a live body exists in the current detection window when the probability of a live body is greater than or equal to a preset threshold; and determine that a live body exists in the vehicle when at least two consecutive detection windows are determined to have a live body.

[0131] In one possible implementation, the processing module 34 can also be used to: in response to determining that a living person is in the vehicle, perform at least one of the following warning operations: control the vehicle display device to display warning information and control the vehicle audio system to play voice prompts; push alarm notifications to a preset mobile application and control the vehicle to issue an audible and visual alarm; send help requests to preset emergency contacts or rescue platforms and control the vehicle's air conditioning system to turn on external circulation, and control the windows to open a preset gap when it is detected that there is no rain outside the vehicle.

[0132] In one possible implementation, the processing module 34 can also be used to: stop executing the current warning operation and control the Bluetooth module to enter standby mode in response to the door unlocking signal or the release command received through the mobile application; or, if no situation occurs within the preset total detection time where at least two consecutive detection windows determine the presence of a live body, control the Bluetooth module to enter sleep mode.

[0133] The in-vehicle liveness detection device based on Bluetooth phase timing perturbation provided in this application embodiment can execute the technical solution shown in the above-described in-vehicle liveness detection method embodiment based on Bluetooth phase timing perturbation. Its implementation principle and beneficial effects are similar, and will not be repeated here.

[0134] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0135] It should be further noted that although the steps in the flowchart are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowchart may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0136] It should be noted that the above-described device embodiments are merely illustrative, and the device of this application can also be implemented in other ways; and it should be understood that the division of the various modules of the above device is only a logical functional division, and in actual implementation, they can be fully or partially integrated into a single physical entity, or they can be physically separated. Furthermore, these modules can all be implemented in software through processing element calls; they can all be implemented in hardware; or some modules can be implemented by processing element calls to software, and some modules can be implemented in hardware. For example, a processing module can be a separately established processing element, or it can be integrated into a chip of the above device. Alternatively, it can be stored as program code in the memory of the above device, and its functions can be called and executed by a processing element of the above device. The implementation of other modules is similar. In addition, these modules can be fully or partially integrated together, or they can be implemented independently. The processing element here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed by the integrated logic circuit in the hardware of the processor element or by software instructions.

[0137] For example, these modules can be one or more integrated circuits configured to implement the above methods, such as one or more Application Specific Integrated Circuits (ASICs), one or more Digital Signal Processors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs). As another example, when a module is implemented using processing element scheduler code, the processing element can be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. Furthermore, these modules can be integrated together as a System-On-a-Chip (SOC).

[0138] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., Digital Video Discs, DVDs), or semiconductor media (e.g., solid-state disks (SSDs)).

[0139] Figure 4 A schematic diagram of the structure of an electronic device provided as an exemplary embodiment of this application. For example... Figure 4 As shown, the electronic device 40 in this embodiment includes:

[0140] At least one processor 41; and a memory 42 communicatively connected to the at least one processor;

[0141] The memory 42 stores instructions that can be executed by at least one processor 41 to cause the electronic device to perform the method as described in any of the above embodiments.

[0142] Alternatively, the memory 42 can be either standalone or integrated with the processor 41.

[0143] The memory 42 may include high-speed random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device.

[0144] The processor 41 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application. Specifically, in implementing the in-vehicle liveness detection method based on Bluetooth phase timing perturbation described in the foregoing method embodiments, the electronic device may be, for example, an electronic device with processing capabilities such as a server.

[0145] Optionally, the electronic device may also include a communication interface 43. In specific implementations, if the communication interface 43, memory 42, and processor 41 are implemented independently, they can be interconnected via a bus to complete communication. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc., but this does not imply that there is only one bus or one type of bus.

[0146] Optionally, in a specific implementation, if the communication interface 43, memory 42 and processor 41 are integrated on a single chip, then the communication interface 43, memory 42 and processor 41 can communicate through an internal interface.

[0147] In practical implementation, electronic devices can be integrated into the body control module, cockpit controller, or vehicle communication control unit. The memory stores program instructions corresponding to vehicle locking triggering, Bluetooth module wake-up, detection signal transmission and reception, phase timing extraction, respiratory and heartbeat frequency band analysis, and liveness detection and warning control. The processor calls and executes these instructions, enabling existing vehicle hardware platforms to perform in-vehicle liveness detection. Because the processor and memory work together to implement the above method, it is possible to identify weak vital signs of stationary living beings inside the vehicle using existing Bluetooth modules without adding additional sensors such as cameras or millimeter-wave radar. This makes system integration simpler, reduces cost and power consumption, and improves detection efficiency and judgment stability within a limited time window after locking the vehicle. Therefore, it is more suitable for passenger vehicle in-vehicle security scenarios.

[0148] The implementation principle and technical effects of the electronic device provided in this embodiment can be found in the foregoing embodiments, and will not be repeated here.

[0149] This application also provides a vehicle, including a vehicle body and electronic devices as described above.

[0150] In this embodiment, the vehicle itself serves as the mounting carrier and power supply, communication, and early warning execution platform for the electronic equipment, enabling its deployment within the vehicle's existing electronic and electrical architecture. The electronic equipment can work in conjunction with the vehicle's body control, lock triggering, wireless communication, and warning functions to perform liveness detection within a limited timeframe after the vehicle is locked, turned off, or the passenger leaves the vehicle. Because the electronic equipment can be directly integrated into the vehicle itself, eliminating the need for additional dedicated sensors such as cameras or millimeter-wave radar, it reduces incremental hardware costs and layout complexity, making the system easier to implement in passenger vehicle scenarios. Furthermore, leveraging the vehicle's integrated capabilities, it can promptly output warning information when a live person is detected inside the vehicle, thus improving the safety of people, children, or pets left behind.

[0151] This application also provides a computer program product, including a computer program, which, when executed, implements the method steps as described in the above method embodiments. The specific implementation and technical effects are similar and will not be repeated here.

[0152] This application also provides a computer-readable storage medium storing computer-executable instructions. When the computer-executable instructions are executed, they are used to implement the method steps as described in the above method embodiments. The specific implementation methods and technical effects are similar and will not be repeated here.

[0153] The aforementioned computer-readable storage media can be implemented from any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0154] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an application-specific integrated circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in an in-vehicle liveness detection device based on Bluetooth phase timing perturbations.

[0155] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0156] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0157] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0158] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0159] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0160] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for in-vehicle liveness detection based on Bluetooth phase timing perturbation, characterized in that, The in-vehicle liveness detection method, applicable to vehicles equipped with Bluetooth modules that support Bluetooth channel detection, includes: In response to configuring the Bluetooth module to active radar mode; The Bluetooth module transmits a detection signal and receives an echo signal reflected from the vehicle's interior environment, and extracts the phase timing sequence between the transmitted signal and the echo signal. Based on the phase time sequence, feature parameters characterizing vital signs are extracted; The presence of a living being inside the vehicle is determined based on the extracted feature parameters.

2. The in-vehicle liveness detection method based on Bluetooth phase timing perturbation according to claim 1, characterized in that, The extraction of feature parameters characterizing vital signs based on the phase time sequence includes: The phase timing sequence is subjected to phase unwrapping processing to restore continuous phase; The unwound phase sequence is high-pass filtered to remove static bias and temperature drift, thus obtaining the dynamic phase perturbation signal. The characteristic parameters are extracted from the dynamic phase perturbation signal.

3. The in-vehicle liveness detection method based on Bluetooth phase timing perturbation according to claim 2, characterized in that, The characteristic parameters include parameters of the respiratory frequency band and the heart rate frequency band. Extracting the characteristic parameters from the dynamic phase perturbation signal includes: The dynamic phase disturbance signal is subjected to time-frequency transformation to obtain the time-frequency distribution; Within a preset breathing frequency band, the dominant breathing frequency, breathing energy, and breathing signal-to-noise ratio are extracted from the time-frequency distribution. Within a preset heartbeat frequency band, the dynamic phase disturbance signal is bandpass filtered and envelope extracted, and the extracted envelope is frequency domain transformed to obtain the heartbeat frequency and heartbeat energy. The output is a five-dimensional feature vector containing the respiratory frequency, respiratory energy, respiratory signal-to-noise ratio, heart rate frequency, and heart rate energy.

4. The in-vehicle liveness detection method based on Bluetooth phase timing perturbation according to any one of claims 1 to 3, characterized in that, The step of determining whether a living person exists inside the vehicle based on the extracted feature parameters includes: The extracted feature parameters are input into a pre-trained classification model, and the classification model outputs the probability of the presence of a live organism. Based on the probability of the presence of a living being, it is determined whether there is a living being inside the vehicle.

5. The in-vehicle liveness detection method based on Bluetooth phase timing perturbation according to claim 4, characterized in that, The in-vehicle liveness detection method starts execution in response to the vehicle locking signal, and the execution time is limited to a preset total detection time, which includes at least two detection windows. The determination of whether a living being is present inside the vehicle based on the probability of the presence of a living being includes: When the probability of the presence of a live organism is greater than or equal to a preset threshold, it is determined that a live organism exists in the current detection window; When at least two consecutive detection windows detect the presence of a living being, it is determined that a living being exists inside the vehicle.

6. The in-vehicle liveness detection method based on Bluetooth phase timing perturbation according to claim 5, characterized in that, Also includes: In response to the determination that a living person is present in the vehicle, at least one of the following warning actions shall be executed: Control the in-vehicle display device to show warning information and control the in-vehicle audio system to play voice prompts; Push alarm notifications to a pre-set mobile application to control the vehicle to emit audible and visual alarms; Send a request for help to a pre-set emergency contact or rescue platform, and control the vehicle's air conditioning system to turn on external circulation. If no rain is detected outside the vehicle, control the windows to open to a pre-set gap.

7. The in-vehicle liveness detection method based on Bluetooth phase timing perturbation according to claim 6, characterized in that, Also includes: In response to a door unlock signal or a release command received via a preset mobile application, the current warning operation is stopped, and the Bluetooth module is put into standby mode. Alternatively, if no two consecutive detection windows detect the presence of a live subject within the preset total detection time, the Bluetooth module is controlled to enter a sleep state.

8. An in-vehicle liveness detection device based on Bluetooth phase timing perturbation, characterized in that, The in-vehicle liveness detection device, applicable to vehicles equipped with Bluetooth modules that support Bluetooth channel detection, includes: A response module for responding to configuring the Bluetooth module to active radar mode; The phase timing sequence extraction module is used to transmit a detection signal and receive an echo signal reflected from the in-vehicle environment via the Bluetooth module, and extract the phase timing sequence between the transmitted signal and the echo signal. The feature parameter extraction module is used to extract feature parameters representing vital signs based on the phase time sequence; The processing module is used to determine whether there is a living person inside the vehicle based on the extracted feature parameters.

9. An electronic device, characterized in that, include: A processor, and a memory communicatively connected to the processor; The memory is used to store computer-executed instructions; The processor is configured to execute the computer execution instructions to implement the method as described in any one of claims 1 to 7.

10. A vehicle, characterized in that, It includes the vehicle body and the electronic equipment as described in claim 9.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed, are used to implement the method as described in any one of claims 1 to 7.

Images