Passenger detection method and device, medium and vehicle

Abstract

The invention provides a passenger detection method and device, a medium and a vehicle. The method comprises the steps of obtaining output values of multiple types of sensors for a target seat on the vehicle; the multiple types of sensors comprise at least two types of pressure sensors, temperature sensors, sound sensors and visual sensors; according to the output value, the seat state of the target seat is determined; under the condition that the seat state indicates that the passenger exists in the target seat, identity verification operation is executed aiming at the output values of the various sensors, and identity information of the passenger is determined from a preset user library; the identity verification operation comprises the step of comparing the output value with a corresponding value of each user in a preset user library. According to the embodiment of the invention, the passenger detection accuracy can be improved.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to an occupant detection method, device, medium and vehicle. Background Art

[0002] As the automotive industry evolves toward higher-level intelligence, occupant detection technology plays a crucial role in improving vehicle safety and providing a personalized experience. As vehicles become increasingly intelligent, the requirements for occupant detection technology are also becoming increasingly stringent. However, most existing occupant detection methods can only detect whether a seat is occupied, resulting in poor accuracy. Summary of the Invention

[0003] The present application provides an occupant detection method, device, medium, and vehicle, which can improve the accuracy of occupant detection.

[0004] In a first aspect, an embodiment of the present application provides an occupant detection method, applied to a vehicle, comprising:

[0005] Obtaining output values ​​of multiple types of sensors on the vehicle for the target seat; the multiple types of sensors include at least two of the following: a pressure sensor, a temperature sensor, a sound sensor, and a visual sensor;

[0006] Determine the seat state of the target seat according to the output value;

[0007] When the seat status indicates that there is an occupant in the target seat, identity authentication operations are performed on the output values ​​of each type of sensor to determine the occupant's identity information from a preset user library; the identity authentication operation includes: comparing the output value with the corresponding value of each user in the preset user library.

[0008] In a second aspect, the present application provides an occupant detection device for use in a vehicle, the device comprising:

[0009] an acquisition module, configured to acquire output values ​​of multiple types of sensors on the vehicle for the target seat; the multiple types of sensors include at least two of the following: a pressure sensor, a temperature sensor, a sound sensor, and a visual sensor;

[0010] a determination module, configured to determine a seat state of the target seat according to the output value;

[0011] A verification module is used to perform identity authentication operations on the output values ​​of each type of sensor when the seat status indicates that there is an occupant in the target seat, and determine the identity information of the occupant from a preset user library; the identity authentication operation includes: comparing the output value with the corresponding value of each user in the preset user library.

[0012] In a third aspect, an embodiment of the present application provides an electronic device, the electronic device comprising: a processor and a memory storing computer program instructions;

[0013] When the processor executes the computer program instructions, it implements the occupant detection method in any one of the embodiments of the first aspect.

[0014] In a fourth aspect, an embodiment of the present application provides a computer storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, an occupant detection method as in any one of the embodiments in the first aspect is implemented.

[0015] In a fifth aspect, an embodiment of the present application provides a computer program product. When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the occupant detection method in any one of the embodiments in the first aspect above.

[0016] In the occupant detection method, device, medium, and vehicle provided in embodiments of the present application, by acquiring the output values ​​of multiple types of sensors on the vehicle for the target seat, including at least two of the following: pressure sensors, temperature sensors, sound sensors, and visual sensors, the seat and occupant status can be comprehensively perceived from multiple dimensions. After determining the seat status, identity verification operations are further performed on the output values ​​of each type of sensor. By comparing these values ​​with the corresponding values ​​for each user in a preset user library, the occupant's identity can be accurately identified. This method not only detects whether a seat is occupied but also further identifies the specific occupant, greatly improving the accuracy of occupant detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0018] Figure 1 This is one of the flow charts of the occupant detection method provided in the embodiment of the present application;

[0019] Figure 2 is a schematic diagram of pressure zones of a pressure sensor provided in an embodiment of the present application;

[0020] Figure 3 is a schematic diagram of a pressure grid of a pressure sensor provided in an embodiment of the present application;

[0021] Figure 4 is a schematic diagram of a seat coordinate system provided in an embodiment of the present application;

[0022] Figure 5This is the second flow chart of the occupant detection method provided in the embodiment of the present application;

[0023] Figure 6 This is the third flow chart of the occupant detection method provided in the embodiment of the present application;

[0024] Figure 7 1 is a schematic structural diagram of an occupant detection device provided in an embodiment of the present application;

[0025] Figure 8 It is a structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0026] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features therein can be combined with each other in the absence of conflict.

[0027] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0028] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprises" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article or device. In the absence of further limitations, an element defined by the sentence "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device that includes the element.

[0029] In order to solve the problems existing in the related art, the embodiments of the present application provide an occupant detection method, device, medium and vehicle.

[0030] The following first introduces the occupant detection method provided by the embodiment of the present application. Figure 1 As shown, the method is applied to a vehicle, and the method specifically comprises the following steps:

[0031] S100, obtaining output values ​​of multiple types of sensors on the vehicle for a target seat; the multiple types of sensors include at least two of the following types: a pressure sensor, a temperature sensor, a sound sensor, and a visual sensor.

[0032] Optionally, in this embodiment of the present application, multiple sensor types refer to a combination of sensors that work together through multiple physical quantity detection methods, covering dimensions such as pressure, temperature, vision, and sound, to achieve comprehensive perception of the occupant's status. The target seat refers to the specific seat for which the occupant's status is to be detected, and can be any seat in the vehicle cabin.

[0033] Optionally, in an embodiment of the present application, the pressure sensor mainly detects the pressure distribution on the seat surface, and its output value reflects whether there is pressure applied by an object or occupant on the seat, as well as the magnitude and distribution range of the pressure, which can be used to preliminarily determine the occupancy status of the seat. The temperature sensor is used to monitor temperature changes in the seat area, and its output value represents the temperature of the seat and the surrounding environment, which helps to sense whether there is a living occupant on the seat, because human body temperature is usually different from ambient temperature. The output value of the sound sensor captures sound signals around the seat, such as the occupant's speech, the sound generated by the movement, etc., which can be used to identify the occupant's voice characteristics and provide voiceprint information for subsequent identity verification. The visual sensor obtains image or video data of the seat and the occupant, and its output value contains visual information such as the appearance and posture of the occupant on the seat, which can be used for facial recognition or other vision-related identity verification. The output values ​​of these multiple types of sensors together constitute a comprehensive perception of the target seat and its occupant status, providing a rich data foundation for the subsequent determination of the seat status and occupant identity verification.

[0034] For example, the sound sensor may be a voiceprint recognition microphone array, and the visual sensor may be a 3D visual acquisition module (structured light camera or Time of Flight (TOF) camera) and a TOF depth sensor.

[0035] Optionally, in a feasible implementation of the present application, first, the pressure sensor serves as the core detection unit, and the real-time acquisition of the pressure distribution on the target seat surface can be achieved through a 16-point pressure sensor array (4×4 grid layout). These sensors are evenly distributed on the seat cushion and backrest according to a preset grid spacing (such as L=120±5mm), and each sensor node continuously outputs an analog voltage signal, corresponding to the pressure value per unit area (such as kPa). Secondly, the temperature sensor monitors environmental parameters in real time and can accurately detect changes in ambient temperature within the range of -40℃ to 85℃. The visual sensor is responsible for collecting spatial information and biometric features. The visual sensor can scan the target seat with a period of 500ms, capture the facial depth information and texture features of the occupant in real time, and generate a 3D model containing facial key points.

[0036] The sound sensor uses a voiceprint recognition microphone array deployed in a multi-channel layout around the target seat to collect ambient sound signals in real time. When the pressure sensor detects a pressure change rate ΔP / Δt greater than 2kPa / s (identified as an occupant change event), the microphone array initiates voice command acquisition to capture the occupant's voice information.

[0037] At the data synchronization level, the vehicle's main controller uses a hardware timer to synchronize multiple sensor clocks, ensuring a timestamp error of ≤10ms for pressure, temperature, visual, and audio data. All raw data undergoes analog-to-digital conversion and is transmitted via a high-speed bus, providing standardized input for subsequent seat status determination (S200) and identity verification (S300).

[0038] S200: Determine the seat state of the target seat according to the output value.

[0039] Optionally, in the embodiment of the present application, the seat state refers to the current usage of the target seat, specifically including whether the seat is in a load-bearing state and the nature of the load-bearing object. The seat state can reflect whether there is an occupant on the seat or other objects are placed on it.

[0040] Optionally, in one implementation of the present application, the state of the seat can be judged based on the output value of a single sensor. For example, a pressure threshold value, such as Ph, is pre-set. When the pressure value P detected by the pressure sensor is greater than Ph, the seat is judged to be in a load-bearing state. Furthermore, it is also possible to preliminarily judge whether it is a passenger or other object based on the shape of the pressure distribution. For example, when a passenger sits on the seat, the pressure distribution has certain shape characteristics, such as a distribution shape that matches the contours of the human buttocks and legs; while the pressure distribution of non-living objects may be relatively uniform or present other irregular shapes. Through a simple analysis of the pressure distribution shape, it can be preliminarily determined whether the seat is carrying a passenger or other object.

[0041] As another example, image processing and computer vision technologies can be used to analyze data acquired by visual sensors. For example, an object detection algorithm can be used to identify whether an object is on a seat. Furthermore, a facial recognition algorithm can be used to determine whether the object is an occupant. If a face or other human features (such as limb outlines) are detected in the seat area, it is determined that an occupant is on the seat. However, the visual sensor's judgment may be affected by factors such as lighting conditions and obstructions. Therefore, it is necessary to consider the impact of these factors on the judgment results and take appropriate measures (such as fill lighting, multi-angle shooting, etc.) to improve the reliability of the judgment.

[0042] Alternatively, in other embodiments, the outputs of multiple sensors can be combined to provide a more comprehensive understanding of the seat and its surroundings, thereby improving the accuracy and reliability of seat status judgment. Different sensors have their own strengths and limitations, and data fusion can complement and validate each other, reducing the possibility of misjudgment by a single sensor.

[0043] Specifically, data collected by multiple types of sensors, such as pressure sensors, temperature sensors, sound sensors, and visual sensors, can be fused and processed. For example, a weighted fusion algorithm can be used to assign different weights to each sensor based on its reliability and relevance. The sensor output values ​​are then weighted summed or subjected to other fusion operations to generate a comprehensive judgment index. The seat status is determined based on the comparison of this comprehensive index with a preset threshold. For example, the presence of an occupant can be determined when the pressure value detected by the pressure sensor is greater than Ph, the temperature value detected by the temperature sensor exceeds Th, and the visual sensor detects an object in the seat area. The rules can be further refined, such as when the pressure sensor and temperature sensor meet these conditions, and the sound sensor detects sound characteristics associated with an occupant, the presence of an occupant can be further confirmed. In this way, the output values ​​of multiple sensors corroborate each other, improving the accuracy of the judgment and the ability to resist interference.

[0044] S300, when the seat status indicates that there is an occupant in the target seat, identity authentication operations are performed on the output values ​​of each type of sensor respectively, and the identity information of the occupant is determined from the preset user library; the identity authentication operation includes: comparing the output value with the corresponding value of each user in the preset user library.

[0045] Optionally, in an embodiment of the present application, the preset user library is a database stored in the vehicle system, which contains the pre-entered personal information of authorized vehicle users and the multimodal biometric templates corresponding to each user, such as the pressure template, which contains a sitting pressure distribution thermal map and a pressure change dynamic curve; the temperature template, which contains a human body thermal radiation feature library under different ambient temperatures; the visual template, which contains facial feature vectors, iris texture features, and human body contour geometric parameters; and the sound template, which contains a specific voiceprint.

[0046] Identity information refers to the passenger's identity determined after verification. It can include basic information such as user ID and driving permission level (e.g., administrator / normal user). Biometrics, such as successfully matched sensor feature vectors (e.g., pressure fingerprint, voiceprint), and behavioral characteristics, such as sitting habits, voice intonation, and temperature adaptation patterns.

[0047] Corresponding values ​​refer to user-specific feature values ​​corresponding to the output values ​​of various sensors. For example, the output value of the pressure sensor corresponds to the user's pressure waveform fingerprint, the output value of the temperature sensor corresponds to the user's body temperature feature value, the output value of the sound sensor corresponds to the user's voiceprint feature value, and the output value of the visual sensor corresponds to the user's facial feature value. These feature values ​​are collected and stored when the user is added to the preset user library.

[0048] Optionally, in an embodiment of the present application, the specific implementation method of the identity authentication operation is flexible and diverse, and can be adjusted according to actual needs and system design. Specifically, when performing identity authentication, a parallel verification method can be adopted, that is, the output values ​​of multiple sensors are used simultaneously for independent comparison and verification. For example, the output value of the pressure sensor can be compared with the pressure distribution characteristics in the user library, and the output value of the sound sensor can be matched with the voiceprint information. If the results of each independent verification meet a certain similarity threshold, the identity of the occupant can be determined. The advantage of this method is that the verification speed is fast, and it can use data from multiple modalities to make a comprehensive judgment in a short time, thereby improving the accuracy of verification.

[0049] Alternatively, a serial verification approach can be employed, performing verification sequentially in a predetermined order. For example, the pressure sensor output can be used for preliminary screening to determine the range of possible matching users; within this range, the sound sensor output can then be used for further, more precise verification. This verification approach can gradually narrow the matching range and reduce computational effort, making it particularly suitable for applications with large user databases.

[0050] Furthermore, to further enhance verification accuracy and reliability, a hybrid verification model can be formed by combining parallel and serial verification. For example, serial verification can be performed on preliminarily screened users in addition to parallel verification, leveraging the advantages of multiple sensors to ensure accurate identification.

[0051] Optionally, in other implementations of the present application, when the seat status indicates that there is no occupant in the target seat, a low-power monitoring mode can be entered to balance the real-time detection and energy consumption. Specifically, the acquisition frequency of multiple types of sensors will be dynamically adjusted according to the preset strategy: for example, the pressure sensor switches from real-time acquisition to periodic polling, the visual sensor and the sound sensor enter a dormant state (only the wake-up trigger mechanism is retained), and the temperature sensor maintains basic ambient temperature monitoring. When a potential wake-up event is detected (such as the instantaneous pressure change of the pressure sensor exceeds the preset change threshold, the door opening signal is triggered, the seat vibration sensor captures obvious displacement, etc.), it will immediately wake up from the low-power mode, restore the full-frequency acquisition state of multiple types of sensors, and re-execute the complete process from S100 to S300 to ensure that it can respond quickly and complete the status judgment and identity authentication when the occupant takes the seat again.

[0052] In an occupant detection method provided in an embodiment of the present application, by acquiring the output values ​​of multiple types of sensors on the vehicle for a target seat, including at least two of the following: pressure sensors, temperature sensors, sound sensors, and visual sensors, the seat and occupant status can be comprehensively perceived from multiple dimensions. After determining the seat status, identity verification operations are further performed on the output values ​​of each type of sensor. By comparing these values ​​with the corresponding values ​​for each user in a preset user library, the occupant's identity can be accurately identified. This method not only detects whether a seat is occupied but also further identifies the specific occupant, greatly improving the accuracy of occupant detection.

[0053] In one embodiment, the plurality of types of sensors include a plurality of pressure sensors arranged in an array on the target seat;

[0054] Before obtaining output values ​​of multiple types of sensors on the vehicle for the target seat, the method further includes:

[0055] Acquiring feature point coordinates of a calibration plate on the target seat and array coordinates of the plurality of pressure sensors; the calibration plate is used to represent a coordinate system reference of the target seat in physical space;

[0056] constructing a mapping relationship between the plurality of pressure sensors and the target seat in physical space based on the feature point coordinates and the array coordinates; the mapping relationship is used to represent the relationship between the pressure data collected by the pressure sensors and the position of the target seat;

[0057] determining, based on the mapping relationship and the detection ranges of the multiple pressure sensors, detection blind areas of the multiple pressure sensors relative to the target seat;

[0058] Filling the detection blind area with data to obtain a pressure sensor that has undergone blind filling operation;

[0059] The obtaining output values ​​of multiple types of sensors on the vehicle for the target seat includes:

[0060] The output value of the pressure sensor on the target seat after the blind spot filling operation is obtained.

[0061] Optionally, in embodiments of the present application, the calibration plate can be a tool used for calibrating a physical space coordinate system. For example, the calibration plate can be a flat device with a specific geometric pattern (e.g., a black and white checkerboard pattern, a circular array, etc.), mounted on the surface of the target seat or at a fixed location around it. By using feature points at known physical locations on the calibration plate, a unified physical space coordinate system (e.g., a 3D Cartesian coordinate system) can be established for the target seat, enabling the data collected by the pressure sensor to be mapped to the actual seat position.

[0062] Feature point coordinates refer to the position information of each feature point on the calibration plate extracted after the calibration plate image is captured by the visual sensor. These coordinate points have corresponding coordinate values ​​in the image space and physical space.

[0063] Array coordinates refer to the two-dimensional index coordinates (expressed as row and column indices) of the multiple pressure sensors arranged on the target seat within the grid array. For example, for a 16-point pressure sensor array with a 4×4 grid layout, each sensor's position is labeled with a (row number, column number) (e.g., the array coordinates for the sensor in row 3, column 2 are (3, 2)). These coordinates are used to identify the relative position of the sensor in the array.

[0064] Optionally, in embodiments of the present application, a mapping relationship describes the correspondence between pressure data collected by the pressure sensor array and the actual physical position of the seat. Specifically, the mapping relationship maps the pressure information detected by the pressure sensor to the actual position coordinates on the seat to accurately determine where the pressure is occurring on the seat. This allows the pressure sensor output to be correlated with the actual physical position on the seat, enabling more accurate detection and analysis.

[0065] The detection blind spot refers to the seat area that the pressure sensor cannot directly detect due to physical layout or technical limitations.

[0066] Optionally, in the embodiment of the present application, Figure 2-4 As shown in Figure 1, the pressure sensors can be arranged in a 16-point 4×4 grid. On the target seat, the sensors are neatly arranged in rows and columns, with a grid spacing of 120±5mm in the X / Y axis direction.

[0067] like Figure 2 As shown, pressure functional zoning divides the seat into three zones: front, middle, and rear. The front zone is associated with the backrest and is marked with "X>0.6." X represents the normalized coordinate in the seat coordinate system, with the total seat length as 1. Values ​​greater than 0.6 correspond to the characteristics of the front zone. The middle zone corresponds to the seat center and is marked with a value of 0.3. The rear zone is associated with the seat edge and is marked with "X<0.3." This zoning method helps analyze pressure conditions at different locations and understand the pressure distribution of occupants on the seat. This zoning method links the pressure data collected by the pressure sensor with the actual functional zones of the seat. Subsequently, combined with the precise coordinates of the 4×4 pressure grid, it allows for more intuitive analysis of pressure distribution in different zones (such as whether the backrest pressure is uniform and whether the center pressure of the seat cushion is ergonomic). This provides a functional reference for blind spot detection (such as insufficient sensor coverage at the edge) and identity verification feature extraction (such as pressure zone differences caused by different occupant sitting postures).

[0068] like Figure 3 As shown in Figure 1, the pressure grid represents a 4×4 grid of sensor nodes with a total of 16 nodes numbered 1 to 16. These nodes can be used to accurately collect pressure data at different locations on the seat surface.

[0069] like Figure 4 As shown in the figure, the seat coordinate system defines three coordinate axes: X, Y, and Z. This provides a spatial positioning reference for seat pressure detection and analysis. Combined with data collected from pressure zones and pressure grids, this coordinate system accurately determines the spatial location and distribution of pressure, facilitating more precise pressure analysis and related calculations.

[0070] Optionally, in one specific implementation of the present application, the coordinates of feature points on a calibration plate on the target seat are first obtained. A visual sensor is used to capture an image of the calibration plate, and a corner detection algorithm is used to identify the feature points on the calibration plate. Based on the intrinsic and extrinsic parameters of the visual sensor, the precise coordinates of these feature points in the target seat's physical space coordinate system are calculated, thereby establishing a coordinate system reference for the seat.

[0071] Next, a mapping relationship is constructed based on the acquired feature point coordinates. Using a spatial transformation algorithm (such as perspective transformation), the array coordinates of the pressure sensor on the two-dimensional plane are linked to the three-dimensional coordinates of the target seat's physical space, and then the three-dimensional coordinates of each pressure sensor in the seat's physical space coordinate system are obtained, realizing the correspondence between the array position of the pressure sensor → the physical space coordinates of the seat. On this basis, a mapping relationship between pressure data and seat position is further established: Since the raw data collected by the pressure sensor is essentially bound to the sensor's own physical position, combined with the aforementioned obtained sensor physical space coordinates, the pressure value output by each sensor can be mapped to the specific position of the seat's physical space, and then through mathematical modeling (such as constructing a lookup table of pressure values ​​and seat coordinates, fitting spatial functions, etc.), a mapping relationship between pressure data → seat physical position is formed.

[0072] Then, the detection blind area is determined based on the mapping relationship. Specifically, first, based on the array arrangement parameters of the pressure sensor, the theoretical coverage range of a single sensor and the entire array is calculated (defining the spatial area in which the sensor can collect pressure data under ideal conditions); then, with the help of the mapping relationship, these theoretical coverage ranges are accurately projected from the sensor array coordinate system to the seat's physical coordinate system, so that the coverage area of ​​each sensor corresponds to the seat's actual three-dimensional spatial position; finally, all areas in the seat's physical coordinate system are traversed, and the physical space not covered by any sensor is the pressure sensor's detection blind area relative to the target seat. This can be used to fill in the blind area and improve the integrity of pressure data collection.

[0073] Finally, fill in the data of the detection blind area. Interpolation algorithms (such as bilinear interpolation and spline interpolation) can be used to estimate the pressure value in the blind area based on the pressure data collected by the pressure sensors around the detection blind area. Machine learning models (such as convolutional neural networks) can also be used to train the model with the surrounding sensor data as input to predict the pressure distribution in the detection blind area, thereby obtaining a pressure sensor that has undergone blind filling operation. After the above steps are processed, when obtaining the output values ​​of multiple types of sensors for the target seat on the vehicle, the output values ​​of the pressure sensor that has undergone blind filling operation are obtained, providing a more complete and accurate data foundation for subsequent analysis and applications based on pressure data (such as occupant sitting posture analysis, identity verification, etc.).

[0074] In these optional embodiments, obtaining the coordinates of the characteristic points on the calibration plate establishes a precise coordinate system for the target seat, providing accurate spatial reference for the pressure sensor data. Constructing a mapping relationship clearly correlates pressure data with seat position, improving data interpretation accuracy. Identifying and filling in blind spots avoids information gaps caused by sensor layout limitations, providing more complete pressure data. These steps yield the filled-in pressure sensor output value, providing more reliable and comprehensive data support for subsequent identity verification applications, enhancing the accuracy and reliability of occupant status monitoring.

[0075] In one embodiment, the multiple types of sensors include temperature-calibrated pressure sensors and temperature sensors;

[0076] The method further comprises:

[0077] When a temperature calibration operation is triggered, historical pressure information collected by the pressure sensor and real-time temperature information collected by the temperature sensor are obtained;

[0078] Determining a drift compensation coefficient corresponding to the real-time temperature information based on the historical pressure information and the real-time temperature information based on a temperature-pressure coupling model; the temperature-pressure coupling model is used to quantify the degree of influence of temperature changes on the output value of the pressure sensor;

[0079] The sensitivity compensation factor of the pressure sensor is updated according to the drift compensation coefficient to obtain a temperature-calibrated pressure sensor.

[0080] Optionally, in embodiments of the present application, historical pressure information refers to pressure data collected and stored by the pressure sensor over a period of time. This data covers pressure values ​​in different scenarios and time periods, reflecting the output of the pressure sensor under various operating conditions. By analyzing this historical pressure information, the operating characteristics and variation patterns of the pressure sensor can be understood, providing a data foundation for subsequent temperature calibration.

[0081] Optionally, the real-time temperature information is the temperature value of the current environment or the surrounding area of ​​the seat collected instantly by the temperature sensor, which reflects the real-time temperature state of the pressure sensor.

[0082] Alternatively, a temperature-pressure coupling model is a mathematical model used to quantitatively describe the impact of temperature changes on the pressure sensor's output. This model can be established by analyzing the relationship between a large amount of historical pressure data and the corresponding temperature information. It accurately quantifies how temperature changes cause pressure sensor output offsets and is a key tool for determining drift compensation coefficients.

[0083] Optionally, the drift compensation coefficient is calculated using a temperature-pressure coupling model based on historical pressure information and real-time temperature information. The drift compensation coefficient represents the degree to which the pressure sensor output drifts due to temperature changes at the current real-time temperature. It measures the impact of temperature on the pressure sensor output and facilitates compensation.

[0084] Optionally, the sensitivity compensation factor is a parameter used to adjust the sensitivity of the pressure sensor. Updating it based on the drift compensation factor can correct the effects of temperature changes on the pressure sensor's sensitivity, allowing the pressure sensor to more accurately reflect actual pressure conditions in different temperature environments, resulting in a temperature-calibrated pressure sensor.

[0085] Optionally, in a specific implementation of the present application, when the temperature calibration operation is triggered, the temperature calibration process can be started through a dual mechanism of timing cycle and dynamic threshold. First, when the internal clock module counts for 15 minutes (timing calibration cycle), or the temperature sensor detects that the temperature change rate exceeds 2°C / min (dynamic trigger condition), the vehicle's main controller immediately synchronizes the acquisition of the historical pressure information of the pressure sensor and the real-time temperature information of the temperature sensor. Subsequently, the temperature-pressure coupling compensation model is called (which can be a polynomial model based on least squares fitting, with an order of n=3), and the historical pressure data and real-time temperature values ​​are input into the model. The model calculates the drift compensation coefficient at the current temperature by analyzing the temperature-pressure scatter plot.

[0086] After obtaining the drift compensation coefficient, the main controller uses the sliding window parameter update mechanism (the window size can be configured to 15 minutes or 30 minutes) to substitute the coefficient into the sensitivity compensation formula to obtain the updated sensitivity compensation factor. The updated sensitivity compensation factor can be written to the EEPROM register of the pressure sensor via the bus. At the same time, a compensation log is generated in the dynamic calibration engine, and the entire calibration process does not require interruption of real-time detection. The sensitivity compensation formula is as follows:

[0087] S compensated =S nominal ×(1+k×α×ΔT)

[0088] Among them, S compensated is the sensitivity after compensation; S nominal is the current sensitivity; k is the temperature-sensitivity correlation coefficient, which is obtained through experimental calibration (such as fitting based on a large amount of temperature-sensitivity test data); α is the drift compensation coefficient; ΔT is the temperature difference between the current temperature and the calibration temperature.

[0089] After calibration, the pressure sensor output value can be compensated to eliminate the ±15% temperature drift error and maintain a detection accuracy of ±1% within a wide temperature range of -40°C to 85°C. For example, when the ambient temperature rises sharply from -20°C to 60°C, the pressure sensor output error is corrected from +12% to -0.8% through dynamic calibration, ensuring that the error in occupant weight calculation is ≤1.5kg. This mechanism, combined with the coordinate system mapping of the checkerboard calibration (error ≤1.5mm), forms a complete link of "spatial positioning-temperature compensation-dynamic calibration", effectively improving environmental adaptability and providing a reliable data basis for functions such as occupant posture recognition and identity authentication.

[0090] In these optional embodiments, by triggering temperature calibration, combining historical pressure and real-time temperature data, and using the temperature-pressure coupling model to calculate the drift compensation coefficient and update the sensitivity compensation factor, the pressure data accuracy is improved, providing a reliable basis for occupant status judgment, identity authentication, etc., and significantly enhancing the stability and detection reliability in extreme temperature environments.

[0091] In one embodiment, the output values ​​of the multiple types of sensors include pressure information, sound information, and visual information;

[0092] The performing of identity authentication operations on the output values ​​of the various sensors to determine the identity information of the occupant from a preset user database includes:

[0093] Performing feature extraction on the pressure information, sound information, and visual information to obtain pressure waveforms, voiceprint information, and facial biometric information;

[0094] The identity verification operation is performed on the facial biometric information, the voiceprint information, and the pressure waveform, and the identity information of the occupant is determined from the preset user library.

[0095] Optionally, in an embodiment of the present application, the pressure information may be seat surface pressure distribution data collected by a pressure sensor array, reflecting physical characteristics such as the occupant's weight distribution and sitting posture changes, such as the pressure value of each grid point (unit: kPa) and the coordinates of the pressure center.

[0096] The sound information may be an audio signal collected by a voiceprint recognition microphone array, which includes acoustic features such as voice commands and speaking intonation, and is used to extract a voiceprint feature vector.

[0097] Visual information can be images and depth data obtained by 3D vision acquisition modules and TOF depth sensors, including facial images, three-dimensional point clouds of human body contours, etc., which are used for biometric recognition.

[0098] The pressure waveform is waveform data obtained after time-domain analysis of pressure information. Characteristics such as the pressure change rate (ΔP / Δt) and pressure distribution entropy are extracted to reflect the occupant's seat entry / exit dynamics or sitting posture pattern.

[0099] Voiceprint information is the Mel Frequency Cepstral Coefficients (MFCC), log-likelihood ratio and other feature vectors extracted after preprocessing the sound information, which is used for voiceprint model matching.

[0100] Facial biometric information is the 68-point facial feature coordinates, iris texture coding and other biometric features extracted from visual information, and a standardized feature vector is generated through 3D structured light technology.

[0101] In these optional embodiments, multimodal data cross-verification can be achieved by extracting features from pressure, sound, and visual information separately and performing multi-level verification (first facial biometric information, then voiceprint information, and finally pressure waveform). This forms a progressive verification chain from "biometric characteristics to behavioral characteristics." This hierarchical mechanism can significantly enhance reliability under extreme working conditions, thereby improving the accuracy of identity recognition and effectively preventing misidentification.

[0102] In one embodiment, the performing of feature extraction operations on the pressure information, sound information, and visual information to obtain pressure waveforms, voiceprint information, and facial biometric information includes:

[0103] The pressure information is subjected to noise reduction processing by a wavelet transform algorithm to obtain noise-reduced pressure. The noise-reduced pressure is subjected to spatiotemporal analysis to extract the ratio of the short-time average energy to the long-time average energy of the noise-reduced pressure to obtain the pressure waveform.

[0104] For the sound information, extracting Mel-frequency cepstral coefficients of the sound information to obtain original voiceprint feature coefficients, mining the spectrum-time series joint features of the original voiceprint feature coefficients to obtain the voiceprint information;

[0105] For the visual information, 3D facial point cloud processing is performed on the visual information to obtain the facial biological information.

[0106] Optionally, in a specific implementation of the present application, when extracting features from pressure information, a wavelet transform algorithm is first used to perform noise reduction processing on it. Wavelet transform can effectively identify and remove noise components in the pressure signal, thereby obtaining a purer noise-reduced pressure signal. Subsequently, a spatiotemporal analysis is performed on the noise-reduced pressure signal to calculate the ratio of its short-time average energy to its long-time average energy. This ratio can reflect the energy distribution characteristics of the pressure signal at different time scales, and then construct the corresponding pressure waveform, providing key feature information for subsequent identity verification operations.

[0107] Mel-frequency cepstral coefficients (MFCCs) are first extracted from the voice information to obtain the original voiceprint feature coefficients. MFCC extraction converts the voice signal into a set of coefficients that characterize the speaker. Based on this, the combined spectral and temporal features of the original voiceprint feature coefficients are further mined to obtain richer and more discriminative voiceprint information. This voiceprint information reflects the uniqueness of an individual's voice and can be used for identity recognition.

[0108] For visual information, 3D facial point cloud processing technology is used to extract facial biometric information. 3D facial point cloud processing reconstructs the three-dimensional structure of the face from images captured by the visual sensor, accurately extracting key facial features and contour information to form facial biometric information. 3D facial point cloud processing not only provides facial geometric features but also reflects facial depth information, helping to improve the accuracy and robustness of facial recognition.

[0109] In these optional embodiments, wavelet transform noise reduction and timely spatial domain analysis are performed on the pressure information to extract the pressure waveform, which can effectively filter out noise interference, accurately capture the pressure change characteristics, provide stable pressure feature input for identity recognition, and significantly improve the accuracy and security of identity recognition.

[0110] In one embodiment, performing the identity verification operation on the facial biometric information, the voiceprint information, and the pressure waveform to determine the occupant's identity information from the preset user library includes:

[0111] Comparing the facial biometric information with facial information in a preset user database for structural similarity;

[0112] When the similarity between the facial biometric information and the facial information of any occupant in the preset user database is greater than or equal to a first threshold, performing a cosine similarity comparison between the voiceprint information and the voiceprint information in the preset user database;

[0113] When the cosine similarity between the voiceprint information and the voiceprint information of any occupant in the preset user library is greater than or equal to a second threshold, performing dynamic time warping matching on the pressure waveform and the pressure data in the preset user library;

[0114] When the degree of match between the pressure waveform and the pressure data of any occupant in the preset user library is greater than or equal to a third threshold, the identity information of the occupant is determined from the preset user library based on the facial biometric information, voiceprint information, and pressure waveform.

[0115] Optionally, in a specific implementation of the present application, identity authentication may be implemented through a three-level progressive decision tree architecture, with the following specific steps:

[0116] 1. First level verification: facial biometric information comparison

[0117] First, the real-time extracted facial biometric feature vector (including geometric and texture features) is compared for structural similarity with the facial templates stored in the preset user library. A coherent point drift algorithm can be used during this comparison to calculate the Hamming distance between the feature point position deviation and the texture descriptor. When the Hamming distance exceeds a first threshold, the voiceprint verification phase begins. If it fails, a visual recheck is triggered (such as adjusting the camera angle) or the system skips to pressure waveform-assisted verification (applicable when the occupant is not facing forward).

[0118] 2. Second level verification: cosine similarity matching of voiceprint information

[0119] For candidate users who have passed facial verification, the cosine similarity calculation is performed between the real-time voiceprint feature vector and the corresponding voiceprint template in the user library. When the cosine similarity is greater than the second threshold, the voiceprint match is considered successful; if it fails, the voice command re-collection is initiated (such as "Please repeat the password"), and a maximum of 3 retries are allowed.

[0120] 3. Third level verification: Dynamic time warping matching of pressure waveform

[0121] For users who pass voiceprint verification, the real-time pressure waveform characteristics are dynamically time-warped and matched with the pressure distribution patterns pre-stored in the user library. The pressure center trajectory is first time-normalized, then the Euclidean distance matrix is ​​calculated. The optimal alignment path is found through cumulative path search to obtain a matching score. When the matching score exceeds the third threshold, the three-level verification results are combined to confirm the occupant's identity from the preset user library. If the match fails, the user is determined to be an "unknown user" and a temporary authorization process (such as vehicle app authentication) is triggered.

[0122] Optionally, in the embodiment of the present application, the identity authentication order of "facial biometric information → voiceprint information → pressure waveform" is adopted in order to achieve the best balance among recognition accuracy, computational efficiency, and robustness: facial biometric information features high dimensionality and strong uniqueness, which can exclude a large number of non-matching users in the shortest time, greatly reducing the amount of subsequent calculations; voiceprint information relies on the microphone array, and although it also has high discrimination, it is affected by the noise in the car and whether the occupants open their mouths, so it is placed in the second place as a cross-verification to avoid false rejection caused by using the voiceprint first; the pressure waveform has the largest amount of calculation and is easily affected by sitting posture and clothing thickness, so it is placed in the last place, and only the candidate users who have passed the first two levels are finally confirmed, which not only ensures the overall recognition accuracy, but also avoids wasting computing power in obvious mismatch scenarios. Compared with the reverse or parallel order, the average verification time can be significantly shortened and the system power consumption can be reduced without reducing the accuracy.

[0123] In these optional embodiments, by sequentially performing a structural similarity comparison of facial biometric information, a cosine similarity comparison of voiceprint information, and a dynamic time warping matching of pressure waveforms, the range of potential matching users can be gradually screened and narrowed down at each stage. This hierarchical verification mechanism not only improves the accuracy of identity recognition, but also reduces the risk of misjudgment, as multiple strict conditions must be met simultaneously to confirm identity. In addition, setting a threshold ensures that the next step is only taken when there is sufficient similarity or a match, avoiding unnecessary calculations and resource waste. This approach combines the advantages of facial recognition, voiceprint recognition, and pressure waveform analysis, and utilizes multimodal recognition technology to greatly enhance the reliability and security of identity verification, providing a solid foundation for vehicle occupant detection.

[0124] In one embodiment, after performing identity authentication operations on the output values ​​of the various sensors and determining the identity information of the occupant from a preset user database, the method further includes:

[0125] Formulate a personalized strategy based on the occupant information, wherein the personalized strategy includes at least one of the following:

[0126] adjusting the target seat to a preset position corresponding to the occupant information;

[0127] Play multimedia resources corresponding to the occupant information.

[0128] Optionally, in a specific implementation of the present application, after completing the authentication operation for the output values ​​of various sensors and determining the occupant's identity information, a personalized strategy can be formulated based on this identity information to improve the occupant's comfort and experience. Specifically, the personalized strategy may include the following aspects:

[0129] First, the target seat can be adjusted to a preset position based on the occupant's information. This means the vehicle can pre-store seat preferences for different occupants, including parameters such as fore-aft position, height, backrest tilt, and lumbar support. When a specific occupant is identified, the vehicle automatically triggers the seat adjustment mechanism, using the seat's motor and adjustment mechanism to move the seat to the occupant's preset position. This automatic adjustment function not only improves convenience but also ensures that occupants quickly achieve the most comfortable seating position upon entering the vehicle, reducing the need for manual adjustments.

[0130] Secondly, the vehicle can also play multimedia resources tailored to the occupant's information. This involves synchronizing the vehicle app with the individual user account. For example, the vehicle can automatically play a specific music playlist, radio station, podcast, or audiobook based on the occupant's preferences. Furthermore, the vehicle may recommend new multimedia content based on the occupant's usage history and preferences, providing a more personalized and enjoyable in-car entertainment experience.

[0131] In these optional embodiments, personalized strategies based on occupant information can significantly enhance the user experience and vehicle intelligence. Automatically adjusting the target seat to a preset position corresponding to the occupant's information eliminates manual adjustments and improves convenience and comfort. Playing multimedia resources tailored to the occupant's preferences makes in-car entertainment more personalized, enhancing user satisfaction.

[0132] It should be noted that the various optional implementation methods introduced in the embodiments of the present application can be implemented in combination with each other or separately if they do not conflict with each other, and the embodiments of the present application do not limit this.

[0133] To facilitate understanding of the occupant detection method provided by the above embodiment, the above occupant detection method is described below using a specific scenario embodiment.

[0134] Alternatively, in another embodiment, an occupant detection system is provided, comprising:

[0135] The sensor network module includes: a 16-point pressure sensor array (4x4 grid layout); a temperature compensation sensor group; a TOF depth sensor; a 3D vision acquisition module; and a voiceprint recognition microphone array.

[0136] Intelligent processing module, including: main controller; AI accelerator; dynamic calibration engine; multimodal verification decision tree;

[0137] Communication module, including: Controller Area Network (CAN) bus interface (communication with the vehicle's Electronic Control Unit (ECU)); 5G NR module (supporting Over-the-Air (OTA) upgrades).

[0138] like Figure 5 As shown, in one embodiment, the steps are as follows:

[0139] 1. The pressure sensor collects pressure waveform data and transmits it to the main controller. Specifically, the pressure sensor parameters are shown in Table 1 below.

[0140] Table 1: Pressure sensor parameters

[0141]

[0142] The pressure sensor layout parameters are: grid spacing: L = 120 ± 5 mm (X / Y axis); detection blind area compensation: ΔL = 15 mm (compensated by TOF sensor); coordinate mapping error: ≤ 1.5 mm (after chessboard calibration).

[0143] 2. The main controller sends a feature extraction request to the AI ​​processor.

[0144] 3. The AI ​​processor analyzes the pressure waveform data to determine whether a seat change has occurred. If a seat change is confirmed, the process continues. If it is determined to be an interference action (such as an occupant adjusting their sitting position, which is not a seat change), the verification system sends an ignore signal to the main controller, and the process ends.

[0145] 4. After confirming the seat change, the verification system activates the camera for scanning, prepares for voiceprint collection, and synchronously triggers the seat pressure sensor for high-frequency data collection. Utilizing the pre-configured sensor layout, it accurately obtains the complete pressure waveform data of the current occupant after taking their seat.

[0146] 5. Enter the multimodal verification cycle and proceed in sequence:

[0147] Face recognition (primary verification): As the primary verification method, it recognizes the passenger's facial information.

[0148] Voiceprint matching (for backup verification): As an auxiliary verification method, the collected voiceprint information is matched with the preset voiceprint information.

[0149] Pressure fingerprint comparison: Compare the current pressure waveform data with the preset pressure fingerprint data.

[0150] 6. The verification system feeds back the results of the multimodal verification to the main controller so that the main controller can perform subsequent operations based on the verification results (such as adjusting the seat position according to the occupant's identity, playing personalized multimedia resources, etc.).

[0151] This process uses pressure sensors to preliminarily determine the seat change situation, and then uses multimodal means such as cameras and voiceprint collection to authenticate the identity, which improves the accuracy and reliability of identity authentication, and can effectively deal with interfering actions that may occur in actual scenarios and avoid misjudgment.

[0152] Alternatively, in one embodiment, a comparative experiment between the present application and a traditional solution demonstrated significant advantages. The test conditions included a vibration environment in accordance with the GB / T 28046.3-2011 standard, as well as a temperature cycling test (-40°C to 85°C for 2 hours each, 5 cycles), using 200 sets of occupant data covering children, adults, and pets as samples. The following experimental comparison data, Table 2, was obtained. The traditional solution is based on single-modality (and pressure data) identity verification.

[0153] Table 2: Experimental comparison data table

[0154]

[0155] In terms of misjudgment rate, the traditional solution is 1.7% (according to GB / T 38186), while this application is reduced to 0.25%, an improvement of 85.3%. The calibration time is sharply reduced from the traditional 3.2s to 280±20ms, a decrease of 91.3%. The temperature range is expanded from the traditional -20℃~70℃ to -40℃~85℃, a widening of 35℃. The multimodal accuracy rate is increased from 92.4% to 99.7%, an increase of 7.3%. The parameter drift rate is greatly reduced from 1.8% / h to 0.15% / h, a decrease of 91.7%. The stability under extreme working conditions jumps from 73.5% to 98.2%, an increase of 33.6%. These data fully demonstrate that this application has made a qualitative leap in accuracy, efficiency, environmental adaptability, stability and other aspects compared with traditional solutions, and has higher application value.

[0156] In one embodiment, if Figure 6 As shown, the calibration and dynamic calibration process of the pressure sensor is as follows:

[0157] 1. Use the calibration tool to install the calibration plate to provide a reference for subsequent coordinate calculations.

[0158] 2. The calibration camera sends a scanning command to the pressure sensor. After receiving the command, the pressure sensor returns the original coordinate information to the calibration camera.

[0159] 3. The calibration camera calculates the transformation matrix based on the received raw coordinate information. This step is to establish a mapping relationship between the pressure sensor coordinates and the actual physical space coordinates.

[0160] 4. Start the dynamic calibration process and enter the loop step.

[0161] 5. The calibration camera requests real-time data from the pressure sensor, and the pressure sensor sends the current pressure distribution data to the calibration camera.

[0162] 6. The calibration camera performs coordinate transformation calculations based on the current pressure distribution data obtained to further optimize the coordinate mapping relationship.

[0163] 7. The calibration camera sends the updated parameters to the memory, which writes the new coordinate mapping information, completing a dynamic calibration. The next round of dynamic calibration continues according to the settings to ensure the accuracy and real-time performance of the pressure sensor coordinate mapping.

[0164] The entire process establishes a foundation by installing a calibration plate. Through steps such as calculating the transformation matrix and dynamic calibration, the coordinate mapping relationship of the pressure sensor is continuously optimized to ensure the accuracy of its measurement data.

[0165] Figure 7 A structural schematic diagram of an occupant detection device provided in another embodiment of the present application is shown. For ease of explanation, only the parts related to the embodiment of the present application are shown.

[0166] Reference Figure 7 , the occupant detection device may include:

[0167] An acquisition module 701 is configured to acquire output values ​​of multiple types of sensors on the vehicle for a target seat; the multiple types of sensors include at least two of the following: a pressure sensor, a temperature sensor, a sound sensor, and a visual sensor;

[0168] A determination module 702 is configured to determine a seat state of the target seat according to the output value;

[0169] The verification module 703 is used to perform identity authentication operations on the output values ​​of each type of sensor when the seat status indicates that there is an occupant in the target seat, and determine the identity information of the occupant from a preset user library; the identity authentication operation includes: comparing the output value with the corresponding value of each user in the preset user library.

[0170] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application, and are devices corresponding to the above-mentioned method. All implementation methods in the above-mentioned method embodiment are applicable to the embodiment of the device. Its specific functions and the technical effects brought about can be found in the method embodiment part, which will not be repeated here.

[0171] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0172] Figure 8 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application is shown.

[0173] The device may include a processor 801 and a memory 802 storing program instructions.

[0174] When the processor 801 executes the program, the steps in any of the above method embodiments are implemented.

[0175] For example, the program can be divided into one or more modules / units, one or more modules / units are stored in the memory 802 and executed by the processor 801 to complete the present application. One or more modules / units can be a series of program instruction segments that can perform specific functions, and the instruction segments are used to describe the execution process of the program in the device.

[0176] Specifically, the processor 801 may include a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0177] The memory 802 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 802 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory 802 may include removable or non-removable (or fixed) media. Where appropriate, the memory 802 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 802 is a non-volatile solid-state memory.

[0178] The memory may include read-only memory (ROM), random access memory (RAM), magnetic disk storage media devices, optical storage media devices, flash memory devices, electrical, optical or other physical / tangible memory storage devices. Thus, generally, the memory includes one or more tangible (non-transitory) readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to an aspect of the present disclosure.

[0179] The processor 801 implements any one of the methods in the above embodiments by reading and executing program instructions stored in the memory 802 .

[0180] In one example, the electronic device may further include a communication interface 803 and a bus 810. The processor 801, the memory 802, and the communication interface 803 are connected via the bus 810 and communicate with each other.

[0181] The communication interface 803 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0182] Bus 810 includes hardware, software or both, and the components of online data flow metering equipment are coupled to each other. For example, but not limitation, bus can include accelerated graphics port (AGP) or other graphics bus, enhanced industry standard architecture (EISA) bus, front side bus (FSB), hypertransport (HT) interconnection, industry standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 810 can include one or more buses. Although the present application embodiment describes and shows specific bus, the application considers any suitable bus or interconnection.

[0183] In addition, in combination with the methods in the above embodiments, embodiments of the present application may provide a storage medium for implementation. The storage medium stores program instructions; when the program instructions are executed by a processor, any one of the methods in the above embodiments is implemented.

[0184] An embodiment of the present application further provides a chip, which includes a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned method embodiment and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0185] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0186] An embodiment of the present application provides a computer program product, which is stored in a storage medium. The program product is executed by at least one processor to implement the various processes of the above-mentioned method embodiment and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0187] An embodiment of the present application further provides a vehicle, the vehicle comprising at least one of the following:

[0188] As in the occupant detection device of the aforementioned embodiment;

[0189] The electronic device as in the aforementioned embodiment;

[0190] The computer-readable storage medium in the aforementioned embodiment;

[0191] As the computer program product in the aforementioned embodiment.

[0192] It should be understood that the present application is not limited to the specific configurations and processes described above and illustrated in the figures. For the sake of brevity, a detailed description of known methods is omitted here. In the above embodiments, several specific steps are described and illustrated as examples. However, the method process of the present application is not limited to the specific steps described and illustrated. Those skilled in the art can make various changes, modifications, and additions, or change the order of the steps after understanding the spirit of the present application.

[0193] The functional modules shown in the above block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), suitable firmware, a plug-in unit, a function card or the like. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. The program or code segment can be stored in a machine-readable medium, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable medium" can include any medium that can store or transmit information. The example of a machine-readable medium includes an electronic circuit, a semiconductor memory device, a ROM, a flash memory, an erasable ROM (EROM), a floppy disk, a CD-ROM, an optical disk, a hard disk, an optical fiber medium, a radio frequency (RF) link, or the like. The code segment can be downloaded via a computer grid such as the Internet, an intranet, etc.

[0194] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps. In other words, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0195] Aspects of the present disclosure have been described above with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and program products according to the embodiments of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer or other programmable data processing device to produce a machine so that these instructions executed via the processor of the computer or other programmable data processing device enable the implementation of the function / action specified in one or more boxes of the flowchart and / or block diagram. This processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor or a field programmable logic circuit. It is also understood that each box in the block diagram and / or the flowchart and the combination of the boxes in the block diagram and / or the flowchart can also be implemented by the dedicated hardware that performs the specified function or action, or can be implemented by the combination of dedicated hardware and computer instructions.

[0196] The above is only a specific implementation method of the present application. Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the scope of protection of the present application is not limited to this. Any technician familiar with this technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the scope of protection of this application.

Claims

1. A passenger detection method, characterized in that: Applied to a vehicle, the method comprises: Obtaining output values ​​of multiple types of sensors on the vehicle for the target seat; the multiple types of sensors include at least two of the following types: a pressure sensor, a temperature sensor, a sound sensor, and a visual sensor; determining a seat state of the target seat according to the output value; When the seat status indicates that there is an occupant in the target seat, identity authentication operations are performed on the output values ​​of each type of sensor respectively, and the identity information of the occupant is determined from a preset user library; the identity authentication operation includes: comparing the output value with the corresponding value of each user in the preset user library.

2. The method according to claim 1, characterized in that The multiple types of sensors include a plurality of pressure sensors arranged in an array on the target seat; Before obtaining output values ​​of multiple types of sensors on the vehicle for the target seat, the method further includes: Acquiring feature point coordinates of a calibration plate on the target seat and array coordinates of the plurality of pressure sensors; the calibration plate is used to represent a coordinate system reference of the target seat in physical space; constructing a mapping relationship between the plurality of pressure sensors and the target seat in physical space based on the feature point coordinates and the array coordinates; the mapping relationship is used to represent the relationship between the pressure data collected by the pressure sensors and the position of the target seat; determining, based on the mapping relationship and the detection ranges of the multiple pressure sensors, detection blind areas of the multiple pressure sensors relative to the target seat; The detection blind area is filled with data to obtain a pressure sensor that has undergone the blind filling operation.

3. The method according to claim 1, characterized in that The multiple types of sensors include a temperature-calibrated pressure sensor and the temperature sensor; The method further comprises: When a temperature calibration operation is triggered, historical pressure information collected by the pressure sensor and real-time temperature information collected by the temperature sensor are obtained; Determining a drift compensation coefficient corresponding to the real-time temperature information based on the historical pressure information and the real-time temperature information based on a temperature-pressure coupling model; the temperature-pressure coupling model is used to quantify the degree of influence of temperature changes on the output value of the pressure sensor; The sensitivity compensation factor of the pressure sensor is updated according to the drift compensation coefficient to obtain the temperature-calibrated pressure sensor.

4. The method according to claim 1, wherein The output values ​​of the multiple types of sensors include pressure information, sound information, and visual information; The performing of identity authentication operations on the output values ​​of the various sensors to determine the identity information of the occupant from a preset user database includes: Performing feature extraction on the pressure information, sound information, and visual information to obtain pressure waveforms, voiceprint information, and facial biometric information; The identity verification operation is performed on the facial biometric information, the voiceprint information, and the pressure waveform, and the identity information of the occupant is determined from the preset user library.

5. The method according to claim 4, characterized in that The feature extraction operation is performed on the pressure information, sound information, and visual information to obtain pressure waveform, voiceprint information, and facial biometric information, including: The pressure information is subjected to noise reduction processing by a wavelet transform algorithm to obtain noise-reduced pressure. The noise-reduced pressure is subjected to spatiotemporal analysis to extract the ratio of the short-time average energy to the long-time average energy of the noise-reduced pressure to obtain the pressure waveform.

6. The method according to claim 4, characterized in that The performing the identity authentication operation on the facial biometric information, the voiceprint information, and the pressure waveform to determine the occupant's identity information from the preset user library includes: Comparing the facial biometric information with facial information in a preset user database for structural similarity; When the similarity between the facial biometric information and the facial information of any occupant in the preset user database is greater than or equal to a first threshold, performing a cosine similarity comparison between the voiceprint information and the voiceprint information in the preset user database; When the cosine similarity between the voiceprint information and the voiceprint information of any occupant in the preset user library is greater than or equal to a second threshold, performing dynamic time warping matching on the pressure waveform and the pressure data in the preset user library; When the degree of match between the pressure waveform and the pressure data of any occupant in the preset user library is greater than or equal to a third threshold, the identity information of the occupant is determined from the preset user library based on the facial biometric information, voiceprint information, and pressure waveform.

7. The method according to claim 1, characterized in that After performing identity authentication operations on the output values ​​of the various sensors and determining the identity information of the occupant from a preset user database, the method further includes: Formulate a personalized strategy based on the identity information, where the personalized strategy includes at least one of the following: adjusting the target seat to a preset position corresponding to the occupant information; Play multimedia resources corresponding to the occupant information.

8. An electronic device, characterized in that: The device includes: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the occupant detection method according to any one of claims 1 to 7 is implemented.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer program instructions, which, when executed by a processor, implement the occupant detection method according to any one of claims 1 to 7.

10. A vehicle, characterized in that: Include at least one of the following: The electronic device according to claim 8; The computer-readable storage medium of claim 9.

Images