Vehicle control method, electronic device, storage medium and program product

Abstract

According to the vehicle control method, the electronic equipment, the storage medium and the program product provided by the embodiment of the invention, the signal acquisition network covering the in-vehicle space is formed through the information acquisition equipment deployed in the vehicle, and the target CSI signal of the target object in the target preset area is acquired in real time; and sending the target CSI signal as input data into a pre-trained preset model, and outputting the action type, so that the vehicle is controlled through the function control instruction corresponding to the action type. According to the invention, traditional optical equipment can be replaced by the signal propagation state information of the CSI signal, and the problem of dependence on a light environment in related technologies is solved. Moreover, the deployment of the information collection equipment covers the whole space in the vehicle, and through the cooperation of signal propagation characteristic analysis and a deep learning model, the high-precision recognition of the motion in the vehicle is realized, a shielding blind area in optical recognition is eliminated, and the recognition reliability of passengers in the back row or a complex motion scene is improved.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and in particular to a vehicle control method, electronic device, storage medium, and program product. Background Technology

[0002] With the development of intelligent driving and in-vehicle electronic systems, the need for in-vehicle perception of occupant behavior is increasing. For example, while the vehicle is in motion, the driver needs to control the air conditioning, entertainment system, and communication functions through actions (such as gestures), while simultaneously monitoring the driver's attention status in real time (such as looking down, distraction, and dangerous behaviors like taking both hands off the steering wheel). Furthermore, the recognition of passenger entertainment needs (such as the front passenger adjusting the seat or audio system) and vehicle safety scenarios (such as airbag deployment during a collision and dangerous action recognition) also places higher demands on the accuracy and real-time performance of in-vehicle action perception.

[0003] In related technologies, vehicle functions can be controlled through motion recognition. Motion recognition mainly relies on optical technologies (such as structured light, binocular stereo imaging, and time-of-flight sensors (ToF)). However, motion recognition is sensitive to the lighting environment. The lighting inside the vehicle changes frequently during the vehicle's operation, resulting in poor recognition stability and blind spots, making it difficult to cover rear passengers or complex motion scenarios. Summary of the Invention

[0004] This application provides vehicle control methods, electronic devices, storage media, and program products for realizing motion recognition throughout the vehicle interior to achieve vehicle control.

[0005] In a first aspect, embodiments of this application provide a vehicle control method, applied to a vehicle, wherein multiple locations within the vehicle are respectively equipped with information acquisition devices for acquiring wireless channel status (CSI) signals in multiple preset areas within the vehicle; the method includes:

[0006] The target CSI signal is acquired by the information acquisition device. The target CSI signal is the CSI signal corresponding to the action of the target object in the target preset area. The target preset area belongs to the plurality of preset areas.

[0007] The target CSI signal is input into the action recognition model for action recognition to obtain the action type of the target object's action within the target preset area. The action recognition model is pre-trained by the preset model based on CSI signals of various user actions.

[0008] Execute the function control command corresponding to the action type, the function control command being used to control the vehicle.

[0009] In one possible implementation, the number of the target preset regions is multiple;

[0010] Execute the function control instructions corresponding to the action type, including:

[0011] If the action types of actions within multiple target preset areas are different, determine the priority of each target preset area;

[0012] Based on the priority of each target preset region, the target action type is determined from the action types of actions within the multiple target preset regions;

[0013] Execute the function control command corresponding to the target action type.

[0014] In one possible implementation, the target action type is determined from the action types of actions within multiple target preset regions based on the priority of each target preset region, including:

[0015] Determine the functional system to which each action type belongs, wherein the functional system is an entertainment system, a driving control system, or an alarm system;

[0016] If the functional systems to which each action type belongs are the same, the action type of the action in the target preset area with the highest priority in each target preset area shall be determined as the target action type.

[0017] If the functional systems to which the various action types belong are different, the functional system with the highest priority is determined as the target functional system based on the priority of the functional systems to which the various action types belong, and it is determined whether the preset area where the action type belonging to the target functional system is located is the priority area of ​​the target functional system; if so, the action type of the target functional system is determined as the target action type.

[0018] In one possible implementation, executing the function control instruction corresponding to the target action type includes:

[0019] If the target action type belongs to the driving control system, acquire the physical state data of the sensors built into the vehicle;

[0020] Determine whether the action type is a false trigger based on the physical state data;

[0021] If not, then execute the function control instruction corresponding to the target action type.

[0022] In one possible implementation, the method further includes:

[0023] When a vehicle collision warning is received, a first CSI signal and a second CSI signal are acquired. The first CSI signal includes the CSI signals of the information collection devices deployed at various locations before the collision, and the second CSI signal includes the CSI signals of the information collection devices deployed at various locations after the collision.

[0024] Based on the first CSI signal and the second CSI signal, determine whether to trigger an alarm;

[0025] If so, the vehicle's alarm function will be triggered.

[0026] In one possible implementation, the method further includes:

[0027] If the action type is a dangerous action type, acquire the third CSI signal corresponding to the action within the target preset area at multiple consecutive time points;

[0028] The multiple third CSI signals are respectively input into the action recognition model for action recognition to obtain multiple action types;

[0029] If the multiple action types are the same, an alarm message is output, which is used to indicate that the action type of the current target object is unsafe.

[0030] Secondly, embodiments of this application provide a vehicle control device applied to a vehicle, wherein multiple locations within the vehicle are respectively equipped with information acquisition devices for acquiring wireless channel status (CSI) signals in multiple preset areas within the vehicle. The device includes:

[0031] The acquisition module is used to acquire a target CSI signal through the information acquisition device. The target CSI signal is the CSI signal corresponding to the action of the target object in a target preset area. The target preset area belongs to the plurality of preset areas.

[0032] The recognition module is used to input the target CSI signal into the action recognition model for action recognition, and obtain the action type of the target object's action in the target preset area. The action recognition model is pre-trained by the preset model based on the CSI signals of various user actions.

[0033] The control module is used to execute the function control instructions corresponding to the action type, and the function control instructions are used to control the vehicle.

[0034] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0035] The memory stores computer-executed instructions;

[0036] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0037] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0038] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0039] The vehicle control method, electronic device, storage medium, and program product provided in this application form a signal acquisition network covering the vehicle interior space through information acquisition devices deployed within the vehicle. This network acquires the target CSI signal of a target object (such as a driver or passenger) within a preset target area in real time. The target CSI signal is then used as input data and fed into a pre-trained preset model, outputting an action type. This allows for vehicle control via function control commands corresponding to that action type. This application replaces traditional optical devices with CSI signal propagation state information, solving the dependence of related technologies on ambient light. Furthermore, the deployment of the information acquisition device covers the entire vehicle interior space. Through the synergy of signal propagation characteristic analysis and deep learning models, it achieves high-precision recognition of in-vehicle actions, eliminates blind spots in optical recognition, and improves the reliability of recognizing rear passengers or complex action scenarios. Attached Figure Description

[0040] 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.

[0041] Figure 1 This application provides a schematic diagram of the scenario.

[0042] Figure 2 A schematic flowchart illustrating a vehicle control method provided in an embodiment of this application;

[0043] Figure 3 A flowchart for CSI action recognition of the OK gesture as an example;

[0044] Figure 4 A schematic flowchart illustrating another vehicle control method provided in an embodiment of this application;

[0045] Figure 5This is a diagram illustrating the determination of the target action type when the action type within multiple target preset areas belongs to the entertainment system;

[0046] Figure 6 A schematic diagram illustrating the processing flow when the target action type belongs to the alarm system, as provided in the embodiments of this application;

[0047] Figure 7 A flowchart illustrating yet another vehicle control method provided in this application;

[0048] Figure 8 A schematic diagram of the structure of a vehicle control device provided in this application;

[0049] Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0050] 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

[0051] 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.

[0052] First, let me explain the terms used in this application:

[0053] (1) Channel State Information (CSI): Wireless Fidelity (WiFi) CSI technology refers to obtaining information about the wireless channel state by measuring the propagation characteristics experienced by the wireless signal during transmission. This information includes the channel's frequency response, attenuation, multipath effects, etc., and can be used to realize various wireless communication applications, such as wireless positioning, gesture recognition, wireless sensor networks, etc.

[0054] (2) Action recognition: Action recognition refers to the process of analyzing a user’s actions or behaviors to achieve specific functions, such as gesture control and health monitoring.

[0055] (3) Gesture recognition: Gesture recognition is a technology that uses computer vision, sensors, and other technologies to recognize human gestures. It can be used to realize various applications, such as human-computer interaction, virtual reality, and healthcare. Gesture recognition usually requires the use of devices such as cameras and depth sensors to capture human movements and gestures, and then analyzes and recognizes them through algorithms.

[0056] (4) Access Point (AP): An AP is a device that connects to a wired network and provides access services for wireless devices. APs are typically used to connect wireless devices to wired networks for data transmission and communication.

[0057] (5) Triangulation: Triangulation is a technique based on multi-point measurement used to determine the position of an object.

[0058] (6) Deep Convolutional Generative Adversarial Networks (DGGANs): A variant of Generative Adversarial Networks (GANs), DCGANs introduce a Convolutional Neural Network (CNN) structure compared to traditional GANs, making them more suitable for image generation tasks. DCGANs are designed to address the problems of unstable training and low-quality generated results in native GANs. GAN is a deep learning model, and its innovation lies in training the model through competition between two neural networks: a generator and a discriminator. CNN is a deep learning model specifically designed to process data with a grid structure, such as images and sound signals. CNNs have achieved significant success in computer vision, especially in tasks such as image recognition, classification, and object detection.

[0059] Among related technologies, motion recognition (such as gesture recognition) primarily relies on optical techniques, such as structured light, binocular stereo imaging, or Time-of-Flight (ToF). Structured light uses infrared lasers to project structured light patterns, combined with a camera capturing reflected light to calculate depth information, thus enabling gesture or motion recognition. This technology is sensitive to light and environmental noise and requires operation within a specific distance, making it difficult to adapt to dynamic scenarios involving moving vehicles. Binocular stereo imaging uses two cameras to calculate object positions using triangulation principles, but requires high-precision cameras and complex algorithms, resulting in high costs and susceptibility to occlusion. ToF emits modulated light pulses and calculates round-trip time, but requires high-power infrared emitters, leading to high energy consumption and the inability to penetrate non-transparent objects, limiting its application in complex vehicle structures.

[0060] Furthermore, optical equipment is susceptible to obstruction from vehicle interior structures (such as A-pillars and B-pillars) or human bodies, resulting in blind spots for recognition of rear passengers or complex action scenarios, making it difficult to achieve full coverage of the entire vehicle space. In other words, the limitations of optical recognition technology make it dependent on specific lighting conditions and subject to blind spots. During vehicle movement, the lighting inside the vehicle is constantly changing, and optical-based motion recognition technology needs to increase lens costs or recognition time to improve the success rate.

[0061] Alternatively, current vehicle gesture recognition can also use visual recognition, which has a limited recognition area. It requires gestures to be performed at a specific distance from the in-vehicle camera and takes a certain amount of time to be recognized. During vehicle operation, this means that the driver needs to take one hand off the steering wheel for a period of time, which poses a certain risk at high speeds.

[0062] WIFI CSI technology requires no optical equipment and can penetrate non-metallic obstacles, making it suitable for enclosed spaces inside vehicles. Building upon this, it can be further combined with deep learning models (such as DCGANs) to achieve action recognition and improve the success rate.

[0063] Therefore, this application provides a vehicle control method. By deploying information acquisition devices inside the vehicle, the signal collection area can cover the entire vehicle space, enabling the acquisition of CSI signals from occupants in multiple preset areas within the vehicle. These CSI signals are then processed using a pre-trained action recognition model to achieve action recognition, thereby enabling vehicle control via function control commands corresponding to the action type. This application replaces traditional optical devices with CSI signal propagation state information, solving the dependence on ambient light in existing technologies. Furthermore, the deployment of the information acquisition devices covers the entire vehicle space, eliminating blind spots in optical recognition and ensuring reliable recognition of rear-seat passengers or complex action scenarios. This method achieves high-precision, non-contact recognition of in-vehicle actions through the synergy of signal propagation characteristic analysis and deep learning models, adapting to the complex needs of the enclosed space of a vehicle.

[0064] Figure 1 The scenario diagram provided for this application is as follows: Figure 1 As shown, at least three information collection devices can be deployed inside vehicle 10.

[0065] In one possible implementation, the information acquisition device can be a WiFi antenna array, or it can be an access point (AP), or other devices that can be used to acquire CSI signals. This application does not limit the scope of the application.

[0066] For example, with Figure 1Taking a small passenger vehicle as an example, an information collection device can be deployed on the center console, left B-pillar and right B-pillar of the vehicle 10 to form a signal collection network covering the interior space. Each information collection device can transmit and receive CSI signals to different areas inside the vehicle, and the actions of all occupants in the vehicle can be identified by the changes in the CSI signals.

[0067] In one possible implementation, the target object (i.e., the occupants inside the vehicle) can be located using three information acquisition devices. A three-dimensional reference coordinate system based on the vehicle interior space can be established, and multiple preset regions can be divided, for example, such as... Figure 1 As shown, multiple preset areas can include the driver's area, the front passenger area, and the rear passenger area.

[0068] Taking a WiFi antenna array as an example, based on triangulation, a reference coordinate system is established, and the three-dimensional coordinates of each WiFi antenna array in this reference coordinate system are determined. The vehicle interior is then divided into zones within the reference coordinate system, such as the driver's area, the front passenger area, and the rear passenger area. After the zones are defined, once the three WiFi antenna arrays acquire the CSI signal corresponding to the target object's action (e.g., a gesture), the action can be located based on this CSI signal, determining the area where the action occurred. This allows for precise vehicle control based on the priority of each preset zone.

[0069] It should be noted that the vehicle in this application can also be a medium or large passenger vehicle or a freight vehicle.

[0070] 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.

[0071] Figure 2 This is a flowchart illustrating a vehicle control method provided in an embodiment of this application, applied to a vehicle. For example, the vehicle may be as follows: Figure 1 The vehicle 10 shown is, for example Figure 2 As shown, the method includes:

[0072] S201. Obtain the target CSI signal through the information acquisition device. The target CSI signal is the CSI signal corresponding to the action of the target object within the target preset area.

[0073] The information acquisition device can be a signal acquisition device deployed on the vehicle's center console, left B-pillar, and right B-pillar. This information acquisition device, for example, is a WiFi antenna array used to collect in-vehicle CSI signals.

[0074] The target audience is users who operate inside the vehicle (such as drivers and / or passengers).

[0075] The target preset area is a pre-defined area inside the vehicle that can be used for motion recognition (such as the driver's area and / or the passenger's area).

[0076] Specifically, when a target object performs an action within a pre-defined target area while the vehicle is in motion, the information acquisition device can collect CSI signals from inside the vehicle. Then, it filters these CSI signals to identify the CSI signal corresponding to the action performed by the target object within the pre-defined target area, using this as the target CSI signal, and excluding invalid CSI signals from areas without action or outside the target area. For example, the vehicle can compare the CSI signals collected by each information acquisition device with the CSI signals emitted by that device, and use the CSI signal showing a change as the target CSI signal.

[0077] S202. Input the target CSI signal into the motion recognition model to perform motion recognition and obtain the motion type of the target object's actions within the target preset area.

[0078] The action recognition model is an algorithmic model used to identify the action type corresponding to the CSI signal. It is pre-trained from a preset model based on CSI signals of various user actions. The preset model is a basic training model (such as DCGAN, CNN, etc.). Action type refers to preset user actions that can trigger vehicle control, such as hand gestures, fist clenching, swiping, or actions such as prolonged head-down posture or hands off the steering wheel.

[0079] In one possible implementation, the vehicle can input the acquired target CSI signal into a pre-trained motion recognition model. The motion recognition model analyzes the feature values ​​of the CSI signal (such as environmental attenuation value, path loss attenuation value, amplitude attenuation value, and power attenuation value), matches them with a preset motion feature library, and outputs the corresponding motion type.

[0080] In one possible implementation, the vehicle can preprocess the acquired target CSI signal (e.g., normalize it) and then input the preprocessed target CSI signal into a pre-trained action recognition model for action recognition, which can improve the recognition accuracy of the action recognition model.

[0081] Specifically, taking DCGANs as a preset model, the target CSI signal can be input into the discriminator of the action recognition model. The discriminator compares the discriminator with the feature values ​​recorded in the feature template library, thereby realizing the recognition of the action type corresponding to the target CSI signal.

[0082] The feature template library is generated during the training of the preset model, and it stores feature values ​​of CSI signals corresponding to various actions. Specifically, the training of the preset model is as follows:

[0083] When the motion recognition function is first activated in a vehicle, a preset model can be trained to obtain a motion recognition model. Alternatively, when a new motion is added, the currently used motion recognition model can be retrained to optimize the motion recognition model.

[0084] Specifically, taking DCGANs as a preset model, the CSI signals corresponding to the actions of occupants inside the vehicle are used as conditions and input into the preset model for conditional training. Further, the features of the occupants' actions are extracted and recognized. The recognized feature values ​​are recorded in the discriminator of the preset model, thus obtaining the action recognition model. In subsequent use, the discriminator of the action recognition model compares the features of the collected CSI signals with those of the recorded CSI signals to identify the type of action of the occupants inside the vehicle. Taking the collection of the OK gesture as an example, the process of CSI action recognition for the OK gesture and the training process of the preset model are as follows: Figure 3 As shown.

[0085] S203. Execute the function control command corresponding to the action type. The function control command is used to control the vehicle.

[0086] Among them, function control commands are control commands that are bound to action types and can realize specific vehicle functions (such as adjusting the air conditioner, changing music, opening windows, etc.).

[0087] The vehicle can obtain the corresponding function control command based on the action type output by the action recognition model by calling the preset action-command mapping relationship. The vehicle control system then executes the command to complete the corresponding vehicle function control.

[0088] In one possible implementation, if the action type is classified as a dangerous action, such as "hands off the steering wheel," CSI signals can be continuously collected for verification to avoid false alarms, promptly alert drivers to unsafe actions, and ensure driving safety. Specifically, the vehicle can acquire third-party CSI signals corresponding to actions within a preset target area at multiple consecutive time points. These multiple third-party CSI signals are then input into an action recognition model for action recognition, resulting in multiple action types. If multiple action types are identical, an alarm message is output to indicate that the current target object's action type is unsafe.

[0089] After initially identifying the type of dangerous action, CSI signals are continuously collected and multiple action recognitions are performed. Alarm information is only output when the continuous recognition results are consistent. This effectively avoids invalid alarms caused by single signal interference and misidentification, and can promptly identify continuous dangerous actions and issue reminders to urge the target to stop unsafe operations. This significantly improves the safety of vehicles during operation and fits the actual application scenario of in-vehicle action recognition.

[0090] It is understandable that the aforementioned multiple consecutive moments can be multiple moments after the current moment.

[0091] In this embodiment, by acquiring CSI signals corresponding to the actions of a target object inside the vehicle within a preset target area using information acquisition equipment, identifying the action type using a pre-trained action recognition model, and then executing the corresponding vehicle control command, high-precision, low-cost, and non-contact recognition of in-vehicle actions is achieved through the collaboration of signal propagation characteristic analysis and deep learning models, adapting to the complex scenario requirements of the vehicle's enclosed space.

[0092] In one possible implementation, there are multiple target objects, and correspondingly, there are also multiple target preset areas, which are the same as the number of target objects. That is to say, at the same time, multiple occupants in the vehicle make different action instructions. For the driving safety of the vehicle, the vehicle can be controlled in the following way.

[0093] Specifically, Figure 4 This is a flowchart illustrating another vehicle control method provided in an embodiment of this application, applied to a vehicle. For example, the vehicle may be as follows: Figure 1 The vehicle 10 shown is, for example Figure 4 As shown, the method includes:

[0094] S401. If the action types of actions within multiple target preset areas are different, determine the priority of each target preset area.

[0095] At the same time, or within a preset duration (e.g., 3 seconds), when multiple target preset areas have different action types, the priority of each target preset area can be determined.

[0096] by Figure 1 Taking a scenario as an example, if the target preset areas include the driver's seat area, the front passenger seat area, and the rear passenger seat area, the priority of each target preset area from high to low is as follows: driver's seat area, front passenger seat area, and rear passenger seat area. That is to say, the driver's seat area has the highest priority, and the rear passenger seat area has the lowest priority.

[0097] For example, if the target preset area includes the front passenger area and the rear passenger area, then the front passenger area has the highest priority.

[0098] It can be understood that this priority refers to a pre-set level used to distinguish the execution order of actions in different target preset areas (actions in preset areas with higher priority are responded to first).

[0099] The vehicle detects the action types in each target preset area. If user actions are detected in multiple target preset areas and the action types corresponding to each action are different, the priority of each target preset area can be determined first.

[0100] S402. Based on the priority of each target preset area, determine the target action type from the action types of actions within multiple target preset areas.

[0101] After determining the priority of each target preset area, the action type of the action in the target preset area with the highest priority can be determined as the target action type. In other words, the action type of the target preset area with lower priority can be ignored according to the priority, so as to avoid conflicts between multiple action commands and improve the driving safety of the vehicle.

[0102] In one possible implementation, when determining the target action type based on the priority of each target preset area, the functional system to which each action type belongs can be further considered. The target action type is then determined based on the functional system, further improving vehicle safety. Specifically, this can be achieved in the following ways:

[0103] The vehicle can determine the functional system to which each action type belongs. This functional system could be, for example, an entertainment system, a driving control system, or a warning system. If the functional systems to which the action types belong are the same, the action type within the highest priority target preset area is determined as the target action type. If the functional systems to which the action types belong are different, the target functional system is determined based on the priority of each action type's functional system. The vehicle then checks whether the preset area containing the action type corresponding to the highest priority functional system is a priority area of ​​the target functional system; if so, the action type of the target functional system is determined as the target action type.

[0104] It is understandable that a vehicle's driving control system is a system used for controlling functions such as navigation, driving assistance, and vehicle settings; a vehicle's entertainment system is a system used for controlling functions such as multimedia playback, air conditioning adjustment, and ambient lighting control; and a vehicle's warning system is a system used for safety reminders such as seat belt reminders, tire pressure warnings, and collision warnings.

[0105] The aforementioned priority preset areas are preset areas with the highest response authority under a certain functional system. For example, the priority preset area for an entertainment system is the passenger side area, the priority preset area for a driving control system is the driver side area, and the alarm system may not have area priority restrictions.

[0106] Specifically, the vehicle can determine the functional system to which each action type belongs. For example, "adjusting the volume" belongs to the entertainment system, "turning on navigation" belongs to the driving control system, and "taking both hands off the steering wheel" belongs to the warning system. Among them, adjusting the volume can be further divided into mute, increase volume, and decrease volume. When the action type in multiple target preset areas is "mute gesture," "increase volume gesture," and "decrease volume gesture," then these action types all belong to the entertainment system.

[0107] If the action types within multiple target preset areas all belong to the functional system, for example, such as Figure 5 As shown, when the driver makes a "mute gesture" in the driver's area, the passenger in the passenger's area also makes a "volume up gesture," both of which belong to the entertainment system. However, the driver's area has a higher priority than the passenger's area, so the vehicle can respond to the driver's area's action first, thus determining the driver's area's action type as the target action type.

[0108] If the action types within multiple target preset areas belong to different functional systems, the vehicle can identify the highest priority driving control system as the target functional system, and then determine whether the action corresponding to the target functional system occurs within its priority area. If the condition is met, the action is identified as the final target action type to ensure the safety of the control logic.

[0109] For example, the vehicle identifies actions such as "turning on windshield wipers" (belonging to the driving control system), "raising volume" (belonging to the entertainment system), and "turning on music" (belonging to the entertainment system) within the vehicle. At this point, the target function system can be determined as the highest priority driving control system (relative to the entertainment system). The priority area of ​​the driving control system is the driver's area. The vehicle can determine whether the preset area containing the action type belonging to the driving control system is the driver's area, that is, whether the "turning on windshield wipers" action occurs in the driver's area. If so, then the "turning on windshield wipers" action can be determined as the target action type.

[0110] It is understandable that if the driving control system's action, such as "turning on the windshield wipers," occurs in the rear passenger area or the front passenger area, the vehicle may not respond to this type of action in order to ensure the safety of driving.

[0111] By classifying action types according to functional systems, and then selecting target action types based on the highest priority target preset area for each action type belonging to the same functional system; and selecting target action types based on the priority of functional systems for each action type belonging to different functional systems, safety control under multiple preset areas and multiple action conflicts is achieved, effectively avoiding command confusion, prioritizing driving safety-related operations, and significantly improving the reliability of vehicle control and user experience.

[0112] S403, Execute the function control command corresponding to the target action type.

[0113] The vehicle can invoke the corresponding function control command based on the final determined target action type, and execute the function control command through the corresponding function system to ensure the orderliness and accuracy of vehicle control.

[0114] In one possible implementation, such as Figure 6 As shown, during vehicle operation, if the target action type belongs to the alarm system, such as "both hands off the steering wheel", the target action type can be determined to be an unsafe action and an alarm prompt can be output. For example, the alarm prompt can be output by voice. This application does not limit the alarm prompt method.

[0115] In this embodiment, when multiple target preset areas exhibit different action types, the priority of each target preset area is first determined, and then the target action type is selected and the corresponding control command is executed based on the priority. This effectively solves the problem of command conflict when multiple preset areas exhibit actions simultaneously, ensuring the orderly control of the vehicle. At the same time, it prioritizes the response to actions of high-priority preset areas (such as the driver's area), improving driving safety and user experience.

[0116] In one possible implementation, if the target action type belongs to the driving control system, the vehicle can acquire physical state data from its built-in sensors (such as vehicle speed, steering wheel angle, and acceleration), and determine whether the action type is a false trigger based on the physical state data. If not, the function control command corresponding to the target action type is executed. For example, if the target action type is "turn on navigation," and the physical state data indicates that the vehicle is in a stable driving state and the driver has the conditions to perform safe operation, then it can be determined that it is not a false trigger. As another example, if the target action type is "switch driving mode," and the physical state data indicates that the driver is urgently maneuvering the vehicle, the gesture is highly likely to be a false action, then the function control command corresponding to this action type can be avoided to ensure driving safety.

[0117] By adding sensor physical state verification before executing driving control actions, it can determine whether the trigger is false based on the actual operating state of the vehicle. The corresponding command is only executed in safe and stable scenarios, which effectively avoids misoperation caused by bumps, interference, and unconscious actions, and greatly improves the safety, reliability and practicality of vehicle gesture control.

[0118] When a vehicle is involved in an accident, damage to the vehicle structure or airbag deployment may occur, causing changes in the CSI signal in a preset area. During actual vehicle operation, if a collision or other accident occurs, resulting in airbag deployment or structural deformation of critical safety structures such as the A-pillar, the system can match the characteristics of the CSI signals collected before the accident to identify changes in critical safety structures and / or airbags, thus triggering the vehicle's warning system.

[0119] Specifically, Figure 7 This application provides a flowchart illustrating another vehicle control method, applicable to a vehicle, specifically, applicable to the vehicle's control components. For example, the vehicle may be as follows: Figure 1 The vehicle 10 shown is, for example Figure 7 As shown, the method includes:

[0120] S701. When receiving vehicle collision warning information, acquire the first CSI signal and the second CSI signal.

[0121] The first CSI signal includes the CSI signals of information collection devices deployed at various locations before the collision occurs. In other words, when the vehicle is in a normal state without a collision, the CSI signals collected by information collection devices at various locations inside the vehicle (such as the center console, left B-pillar, and right B-pillar) are collected.

[0122] The second CSI signal includes the CSI signals from information collection devices deployed at various locations after a collision.

[0123] When a vehicle exhibits collision characteristics (such as a sudden increase in acceleration or severe vibration of the vehicle body), the control component can receive a collision warning message. The vehicle can acquire the CSI signals stored by each information acquisition device within a preset time period before the collision (e.g., 10 seconds before the collision) as the first CSI signal. At the same time, it can acquire the CSI signals of each information acquisition device within a preset time period after the collision (e.g., 10 seconds after the collision) as the second CSI signal, ensuring that complete channel state data before and after the collision is obtained, providing support for subsequent alarm determination.

[0124] S702. Determine whether to trigger an alarm based on the first CSI signal and the second CSI signal.

[0125] If so, then execute S703.

[0126] After a collision, the vehicle can preprocess the acquired first CSI signal and second CSI signal (noise removal, normalization, etc.), extract the feature values ​​of the two signals (such as environmental attenuation value, path loss attenuation value, amplitude attenuation value, and power attenuation value, etc.), and perform comparative analysis to determine whether to trigger an alarm.

[0127] For example, if the second CSI signal changes significantly compared to the first CSI signal, and the change exceeds a preset threshold, it indicates that a real collision has indeed occurred, and the collision may cause serious changes in the in-vehicle channel environment (such as changes in the position of the occupants or damage to the interior). In this case, it is determined that an alarm needs to be triggered. If the second CSI signal changes slightly compared to the first CSI signal, and the change does not exceed a preset threshold, it indicates that the collision warning information may be falsely triggered (such as severe vehicle shaking or sensor failure). In this case, it is determined that an alarm does not need to be triggered to avoid interference caused by false alarms.

[0128] S703, triggers the vehicle's alarm function.

[0129] When it is determined that an alarm needs to be triggered, the preset alarm function will be activated immediately. Based on the severity of the collision and the actual scenario, the corresponding alarm operation will be executed to ensure that a warning can be issued in a timely manner after the collision, thus providing protection for rescue and personnel safety.

[0130] In this embodiment, after receiving a collision warning message, the system acquires and compares the CSI signals before and after the collision to accurately determine the authenticity of the collision and the necessity of the alarm, and then triggers the corresponding alarm function. This effectively avoids invalid alarms caused by false triggering of collision sensors, while ensuring that timely warnings can be issued in real collision scenarios. This improves the accuracy and reliability of vehicle collision alarms and provides strong support for the safety of occupants and subsequent rescue.

[0131] Figure 8 A schematic diagram of the structure of a vehicle control device provided in this application is shown below. Figure 8 As shown, the vehicle control device 80 includes: an acquisition module 801, an identification module 802, and a control module 803.

[0132] The acquisition module 801 is used to acquire the target CSI signal through the information acquisition device. The target CSI signal is the CSI signal corresponding to the action of the target object in the target preset area. The target preset area belongs to multiple preset areas.

[0133] The recognition module 802 is used to input the target CSI signal into the action recognition model for action recognition, and to obtain the action type of the target object's action within the target preset area. The action recognition model is pre-trained by the preset model based on the CSI signals of various user actions.

[0134] The control module 803 is used to execute the function control instructions corresponding to the action type, which are used to control the vehicle.

[0135] In one possible implementation, the number of target preset areas is multiple, and the control module 803 is specifically used for:

[0136] If the action types of actions within multiple target preset areas are different, the priority of each target preset area is determined.

[0137] Based on the priority of each target preset area, the target action type is determined from the action types of actions within multiple target preset areas.

[0138] Execute the function control instructions corresponding to the target action type.

[0139] In one possible implementation, the control module 803 is specifically used for:

[0140] Identify the functional system to which each action type belongs, such as the entertainment system, driving control system, or alarm system.

[0141] If the functional systems to which each action type belongs are the same, the action type of the action in the target preset area with the highest priority in each target preset area shall be determined as the target action type.

[0142] If the functional systems to which the various action types belong are different, the system with the highest priority is determined as the target functional system based on the priority of the functional systems to which the various action types belong. It is then determined whether the preset area where the action type belonging to the target functional system is located is the priority area of ​​the target functional system. If so, the action type of the target functional system is determined as the target action type.

[0143] In one possible implementation, the control module 803 is specifically used for:

[0144] If the target action type belongs to the driving control system, acquire the physical state data of the vehicle's built-in sensors.

[0145] Determine whether the action type is a false trigger based on the physical state data.

[0146] If not, execute the function control instruction corresponding to the target action type.

[0147] In one possible implementation, device 80 further includes an alarm module 804, specifically used for:

[0148] When a vehicle collision warning is received, a first CSI signal and a second CSI signal are acquired. The first CSI signal includes the CSI signals of the information collection devices deployed at various locations before the collision, and the second CSI signal includes the CSI signals of the information collection devices deployed at various locations after the collision.

[0149] Determine whether to trigger an alarm based on the first CSI signal and the second CSI signal.

[0150] If so, the vehicle's alarm function will be triggered.

[0151] In one possible implementation, the alarm module 804 is also used for:

[0152] If the action type is a dangerous action type, acquire the third CSI signal corresponding to the action within the target preset area at multiple consecutive time points.

[0153] Multiple third-party CSI signals are input into the action recognition model to perform action recognition and obtain multiple action types.

[0154] If multiple actions are of the same type, an alarm message will be output, which will indicate that the action type of the current target object is unsafe.

[0155] The vehicle control device provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0156] Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 9 As shown, the electronic device 90 provided in this embodiment includes at least one processor 901 and a memory 902. Optionally, the device 90 further includes a communication component 903. The processor 901, memory 902, and communication component 903 are connected via a bus 904.

[0157] In a specific implementation, at least one processor 901 executes computer execution instructions stored in memory 902, causing at least one processor 901 to perform the above-described method.

[0158] The specific implementation process of processor 901 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0159] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0160] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0161] 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. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0162] In one possible implementation, the electronic device can be a vehicle or a control component within a vehicle, and this application does not limit this.

[0163] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0164] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0165] The aforementioned readable storage medium can be implemented by 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.

[0166] 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 the device.

[0167] 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.

[0168] 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.

[0169] 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.

[0170] 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.

[0171] 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.

[0172] 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 controlling a vehicle, characterized in that, Applied to vehicles, the method comprises: information acquisition devices deployed at multiple locations within the vehicle to acquire wireless channel status (CSI) signals in multiple preset areas within the vehicle; The target CSI signal is acquired by the information acquisition device. The target CSI signal is the CSI signal corresponding to the action of the target object in the target preset area. The target preset area belongs to the plurality of preset areas. The target CSI signal is input into the action recognition model for action recognition to obtain the action type of the target object's action within the target preset area. The action recognition model is pre-trained by the preset model based on CSI signals of various user actions. Execute the function control command corresponding to the action type, the function control command being used to control the vehicle.

2. The method according to claim 1, characterized in that, The number of the target preset areas is multiple; Execute the function control instructions corresponding to the action type, including: If the action types of actions within multiple target preset areas are different, determine the priority of each target preset area; Based on the priority of each target preset region, the target action type is determined from the action types of actions within the multiple target preset regions; Execute the function control command corresponding to the target action type.

3. The method according to claim 2, characterized in that, Based on the priority of each target preset region, the target action type is determined from the action types of actions within the multiple target preset regions, including: Determine the functional system to which each action type belongs, wherein the functional system is an entertainment system, a driving control system, or an alarm system; If the functional systems to which each action type belongs are the same, the action type of the action in the target preset area with the highest priority in each target preset area shall be determined as the target action type. If the functional systems to which the various action types belong are different, the functional system with the highest priority is determined as the target functional system based on the priority of the functional systems to which the various action types belong, and it is determined whether the preset area where the action type belonging to the target functional system is located is the priority area of ​​the target functional system; if so, the action type of the target functional system is determined as the target action type.

4. The method according to claim 3, characterized in that, Execute the function control instruction corresponding to the target action type, including: If the target action type belongs to the driving control system, acquire the physical state data of the sensors built into the vehicle; Determine whether the action type is a false trigger based on the physical state data; If not, then execute the function control instruction corresponding to the target action type.

5. The method according to any one of claims 1-4, characterized in that, The method further includes: When a vehicle collision warning is received, a first CSI signal and a second CSI signal are acquired. The first CSI signal includes the CSI signals of the information collection devices deployed at various locations before the collision, and the second CSI signal includes the CSI signals of the information collection devices deployed at various locations after the collision. Based on the first CSI signal and the second CSI signal, determine whether to trigger an alarm; If so, the vehicle's alarm function will be triggered.

6. The method according to claim 1, characterized in that, The method further includes: If the action type is a dangerous action type, acquire the third CSI signal corresponding to the action within the target preset area at multiple consecutive time points; The multiple third CSI signals are respectively input into the action recognition model for action recognition to obtain multiple action types; If the multiple action types are the same, an alarm message is output, which is used to indicate that the action type of the current target object is unsafe.

7. A vehicle control device, characterized in that, Applied to vehicles, the device comprises information acquisition equipment deployed at multiple locations within the vehicle to acquire wireless channel status (CSI) signals in multiple preset areas within the vehicle. The acquisition module is used to acquire a target CSI signal through the information acquisition device. The target CSI signal is the CSI signal corresponding to the action of the target object in a target preset area, and the target preset area belongs to the plurality of preset areas. The recognition module is used to input the target CSI signal into the action recognition model for action recognition, and obtain the action type of the target object's action in the target preset area. The action recognition model is pre-trained by the preset model based on the CSI signals of various user actions. The control module is used to execute the function control instructions corresponding to the action type, and the function control instructions are used to control the vehicle.

8. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-6.

10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method described in any one of claims 1-6.

Images