Alarm method for vehicle, controller, alarm system and vehicle

Abstract

The invention relates to the technical field of vehicle-mounted intelligent safety, in particular to an alarm method for a vehicle, and the alarm method comprises the steps: S110, receiving detection data related to a passenger of the vehicle; s120, activating a judgment mode corresponding to the detection data based on the detection data; s130, analyzing the detection data by means of the activated judgment mode so as to evaluate the dangerous state of the passenger; and a step S140 of generating and outputting an alarm signal based on the evaluated dangerous state. In addition, the invention also comprises a controller, an alarm system, a vehicle and a computer program product. According to the invention, the analysis and evaluation of the safety of the vehicle occupant are realized by means of different judgment modes, so that the alarm is triggered automatically, flexibly and effectively. Therefore, the method can be suitable for multi-scene and multi-mode alarm, the protection on vehicle passengers is improved, and a timely, hidden and efficient alarm mode is realized.

Description

Technical Field

[0001] This invention relates to the field of vehicle-mounted intelligent safety technology, and more specifically to an alarm method for vehicles. Furthermore, this invention also relates to a controller, an alarm system, a vehicle, and a computer program product. Background Technology

[0002] With the development of in-vehicle intelligent safety technology, vehicle safety is receiving increasing attention.

[0003] Currently, an increasing number of vehicles are equipped with anti-hijacking systems or alarm systems. Existing vehicle alarm systems typically include an alarm button for vehicle occupants, such as the driver, to manually trigger the alarm if a passenger is threatened or hijacked. The alarm system then responds to the passenger's call to contact an emergency call center. However, in a hijacked scenario, occupants are usually unable to actively trigger the button or answer phone calls or voice messages. In particular, occupants may be unable to make a sound due to their mouths being gagged. If occupants were to make a sound or attempt to trigger the alarm, it could provoke the intruder and lead to further threats to their personal safety.

[0004] Furthermore, existing voice alarm systems can only recognize and execute preset alarm commands, resulting in limited voice prompts and an inability to automatically identify potential hazards in dangerous contexts. This also leads to a lack of data integrity and continuity in recording the threat process during periods when occupants' personal safety is threatened.

[0005] In particular, existing alarm operations cannot automatically adopt different identification modes based on different threat situations, which can easily arouse the vigilance of intruders and cause secondary harm.

[0006] Therefore, there is still a need to improve existing technologies to address at least some of the aforementioned problems. Summary of the Invention

[0007] Based on this, the present invention proposes an efficient solution that not only overcomes the shortcomings of existing technical solutions, but also reliably and accurately identifies personal threats to occupants and quickly and discreetly triggers alarms while effectively protecting occupant safety, thereby protecting occupant safety and improving user experience.

[0008] According to a first aspect of the present invention, an alarm method for a vehicle is provided, wherein the alarm method includes: Step S110: Receive detection data related to the occupants of the vehicle; Step S120: Activate the judgment mode corresponding to the detection data based on the detection data; Step S130: Analyze the detection data using the activated judgment mode to assess the occupant's dangerous status; and Step S140: Generate and output an alarm signal based on the assessed hazardous condition.

[0009] The basic concept of this invention lies in using different judgment modes to achieve highly flexible, accurate, and timely automatic alarm triggering for the analysis and assessment of the safety status of vehicle occupants. This not only enables the applicability of this solution in multiple scenarios and provides multi-mode alarms, but also improves the protection of vehicle occupants and achieves a concealed and efficient alarm method.

[0010] Advantageous configurations of the technical solution of the present invention can be obtained from the following optional embodiments.

[0011] In an optional embodiment of the alarm method according to the present invention, the detection data includes first data and second data, wherein the first data represents non-interactive data relating to the occupant and not interacting with the vehicle, and the second data represents interactive data relating to the occupant and interacting with the vehicle, wherein a first judgment mode is activated based on the first data, and a second judgment mode is activated based on the second data.

[0012] In an optional embodiment of the alarm method according to the present invention, in the first judgment mode, the first data is analyzed by means of an artificial intelligence model to generate a first danger probability of the occupant, and an alarm signal is generated and output based on the first danger probability.

[0013] In an optional embodiment of the alarm method according to the present invention, in the second judgment mode, the second data is analyzed by means of preset data and / or artificial intelligence model to generate a second danger probability of the occupant, and an alarm signal is generated and output based on the second danger probability.

[0014] In an optional embodiment of the alarm method according to the present invention, the first data includes at least one of the following: image data, voice data, and vital sign data.

[0015] In an optional embodiment of the alarm method according to the present invention, the second data includes at least one of the following: voice interaction data, image interaction data, tactile interaction data, and driving status data.

[0016] In an optional embodiment of the alarm method according to the present invention, the preset data includes at least one of the following: preset voice data, preset image data, and preset tactile data.

[0017] In an optional embodiment of the alarm method according to the present invention, the artificial intelligence model includes at least one of the following models: a semantic context recognition model, an image recognition model, and a personnel status analysis model.

[0018] In an optional embodiment of the alarm method according to the present invention, the first data and / or the second data are preprocessed via data fusion.

[0019] In an optional embodiment of the alarm method according to the present invention, in step S140, an instruction for sending and / or storing real-time in-vehicle audio data and / or real-time video data and / or the real-time location of the vehicle is additionally output.

[0020] In an optional embodiment of the alarm method according to the present invention, the alarm signal and / or the real-time in-vehicle audio data and / or the real-time video data and / or the real-time location of the vehicle are sent to a remote emergency call center and / or emergency contacts and / or surrounding traffic participants.

[0021] In an optional embodiment of the alarm method according to the present invention, the alarm signal and / or the real-time in-vehicle audio data and / or the real-time video data and / or the real-time location of the vehicle are transmitted wirelessly.

[0022] In an optional embodiment of the alarm method according to the present invention, the alarm signal includes the current location of the vehicle and / or vehicle information and / or occupant information and / or dangerous person information.

[0023] According to a second aspect of the present invention, a controller is provided, wherein the controller includes a processor and a memory, the memory storing computer instructions, which, when executed by the processor, are used to assist in implementing the alarm method according to one of the above embodiments.

[0024] According to a third aspect of the present invention, an alarm system is provided, wherein the alarm system comprises: The vision unit includes an image detection device for detecting in-vehicle images and a vision output device for outputting visual information. The voice unit includes a voice detection device for detecting in-vehicle voice and a voice output device for outputting voice. A vital signs detection unit, configured to detect the vital signs data of the occupant; A tactile unit, the tactile unit including a tactile input device for inputting tactile signals and a tactile output device for outputting tactile signals; A communication unit configured to connect to and output data to an external device; and According to the controller of the second aspect of the present invention, the controller is data-connected to the vision unit, the voice unit, the vital sign detection unit, the tactile unit and the communication unit.

[0025] According to a fourth aspect of the present invention, a vehicle is provided, wherein the vehicle includes an alarm system according to a third aspect of the present invention.

[0026] According to a fourth aspect of the present invention, a computer program product is provided, wherein the computer program product includes computer instructions, which, when executed by a processor, are used to assist in implementing the alarm method according to one of the above embodiments.

[0027] Further features of the invention will become apparent from the claims, drawings, and description of the figures. Features and combinations of features mentioned in the foregoing description, as well as features and combinations of features mentioned in the following description of the figures and / or shown only in the figures, can be used not only in the corresponding specified combinations, but also in other combinations without departing from the scope of the invention. Therefore, the following are also considered to be covered and disclosed by the invention: those not explicitly shown in the figures and not explicitly interpreted, but rather derived from and produced by combinations of separate features derived from the interpreted content. The following combinations of features are also considered to be disclosed: those that do not possess all the features of the originally drafted independent claims. Furthermore, the following combinations of features are considered to be disclosed, especially those exceeding or deviating from the feature combinations defined in the reference relationships of the claims. Attached Figure Description

[0028] The principles, features, and advantages of the invention will be better understood below by referring to the accompanying drawings. In the drawings: Figure 1 A schematic diagram of an embodiment of a vehicle according to the present invention is shown; Figure 2 A schematic diagram of an embodiment of an alarm system according to the present invention is shown; Figure 3 Show Figure 2 A schematic diagram of an alarm system; Figure 4 A schematic diagram illustrating an application scenario of the alarm method for a vehicle according to the present invention; and Figure 5 A schematic flowchart illustrating an embodiment of an alarm method for a vehicle according to the present invention is shown.

[0029] List of reference numerals 1 vehicle 2 crew members 3. Dangerous personnel 10 Alarm System 20 visual units 21. In-vehicle cameras 22 Vehicle-mounted display devices 30 speech units 31 microphones 32 speakers 40 vital signs detection units 41 Heart rate sensor 50 tactile units 51 Pressure Sensor 52 Vibration Generator 60 communication units 70 Controller 71 processor 72 Memory 80 External devices 100 alarm methods Method steps S110 to S140. Detailed Implementation

[0030] To make the technical problems to be solved, the technical solutions, and the beneficial technical effects of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and several exemplary embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of protection of this invention.

[0031] In this specification, for convenience, terms such as "middle," "upper," "lower," "front," "rear," "vertical," "horizontal," "top," "bottom," "inner," and "outer" are used to indicate orientation or positional relationships in conjunction with the accompanying drawings. This is solely for the purpose of facilitating the description and simplification, and does not imply that the device or element referred to has a specific orientation, or is constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this disclosure. The positional relationships of the constituent elements may be appropriately varied depending on the direction in which each constituent element is described. Therefore, the use of terms not limited to those described in the specification may be appropriately replaced as needed.

[0032] The present invention will now be described in detail with reference to the accompanying drawings.

[0033] Figure 1 A schematic diagram of one embodiment of the vehicle 1 according to the present invention is shown.

[0034] like Figure 1 As shown, vehicle 1 is configured as a passenger car, and is exemplary an electric vehicle with autonomous driving capabilities. In other embodiments, vehicle 1 may also be configured as other types of vehicles, such as gasoline-powered vehicles.

[0035] Vehicle 1 includes an alarm system 10. The alarm system 10 is capable of establishing a data connection with an external device 80 and is used to send alarm signals to the external device 80 in dangerous situations.

[0036] Figure 2 A schematic diagram of an embodiment of the alarm system 10 according to the present invention is shown. Figure 3 Show Figure 2 A schematic diagram of the alarm system 10.

[0037] According to this embodiment, the alarm system 10 includes a visual unit 20, a voice unit 30, a vital sign detection unit 40, and a tactile unit 50.

[0038] The vision unit 20 includes an image detection device for detecting images inside the vehicle, such as an in-vehicle camera 21. Figure 3 The illustration shows an occupant 2. Occupant 2 is the driver and is located in the driver's seat. The in-vehicle camera 21 can capture images and videos of the occupant 2. Furthermore, the vision unit 20 also includes a vision output device for outputting visual information, such as an in-vehicle display device 22. The in-vehicle display device 22 may, exemplarily, be a central control screen and / or a rear-seat display and / or a driver's instrument panel and / or a head-up display. Figure 3 As shown, exemplarily, one in-vehicle camera 21 is positioned at the front of the vehicle facing the interior space of vehicle 1, and another in-vehicle camera 21 is positioned inside the A-pillar at the passenger side and laterally facing the driver's seat. Alternatively, more in-vehicle cameras 21 can be installed to achieve 180° panoramic in-vehicle detection. Exemplarily, the in-vehicle camera 21 can be configured as a high-definition fisheye camera capable of capturing 1080P images and 30fps video. The in-vehicle display device 22 is exemplarily positioned centrally at the front of the vehicle on the dashboard.

[0039] The voice unit 30 includes a voice detection device, such as a microphone 31, for detecting in-vehicle voice, and a voice output device, such as a speaker 32, for outputting voice. Figure 3 As shown, exemplarily, one microphone 31 is arranged at the front of the vehicle interior, another microphone 31 is arranged on the inside of the door, and a speaker 32 is arranged at the in-vehicle display device 22. According to another embodiment, the speaker 32 and microphone 31 can also be integrated into the in-vehicle display device 22. Furthermore, a microphone array, such as an 8-channel microphone array, can be provided to achieve 360° sound field localization within the vehicle.

[0040] The vital signs detection unit 40 is used to detect the vital signs data of the occupant 2. Exemplarily, the vital signs detection unit 40 may include a heart rate sensor 41 and / or an infrared temperature sensor (not shown). The heart rate sensor 41 is capable of detecting changes in the heart rate of the occupant 2 and... Figure 3The heart rate sensor 41 is exemplarily positioned on the steering wheel, allowing the occupant 2, or driver, to monitor their heart rate in real time while holding the steering wheel. Alternatively or additionally, the heart rate sensor 41 may also be positioned on the seatbelt and / or in the back of the vehicle seat. An infrared temperature sensor can detect changes in the occupant 2's body temperature.

[0041] The haptic unit 50 includes a haptic input device, such as a pressure sensor 51, for inputting haptic signals, and a haptic output device, such as a vibration generator 52, for outputting haptic signals. Furthermore, the haptic unit 50 may also include a touchscreen, such as a central control screen. The touchscreen can receive haptic signals input by the occupant 2 (e.g., clicking an icon on the screen interface) and output haptic feedback signals to the occupant 2. Figure 3 As shown, the pressure sensor 51 and the vibration generator 52 are exemplaryly both arranged in the driver's seat. In other embodiments, the pressure sensor 51 and the vibration generator 52 are arranged in each vehicle seat.

[0042] like Figure 2 As also shown, the alarm system 10 further includes a communication unit 60 and a controller 70. The communication unit 60 is used to connect to and output data, such as alarm signals, to an external device 80. Exemplarily, the external device 80 may include a server for a remote emergency call center, a terminal device for an emergency contact (e.g., a smartphone), or surrounding traffic participants (e.g., other vehicles). Furthermore, the external device 80 may also include a server for a judicial system that receives alarm signals and additional evidence data. Exemplarily, the communication unit 60 can transmit data in a 5G / V2X dual-mode manner with a data transmission latency of less than 20 milliseconds.

[0043] The controller 70 is capable of data connection with the vision unit 20, voice unit 30, vital sign detection unit 40, tactile unit 50, and communication unit 60, and exemplaryly achieves real-time data reception and transmission via vehicle Ethernet or CAN bus. Exemplarily, the controller 70 includes a processor 71 and a memory 72. The memory 72 is capable of storing data transmitted by the vision unit 20, voice unit 30, vital sign detection unit 40, and tactile unit 50, and also stores computer instructions. The processor 71 is capable of analyzing and processing the provided data, generating alarm signals, and outputting the alarm signals to the communication unit 60. Here, the processor 71 can also invoke, for example, an artificial intelligence model stored in the memory 72 or in the cloud, to analyze and process the data.

[0044] Figure 4 A schematic diagram illustrating an application scenario of the alarm method 100 for vehicle 1 according to the present invention is shown.

[0045] like Figure 4In the application scenario shown, occupant 2 is located according to Figure 1 In vehicle 1, and exemplarily, the driver. The driver is seated in the driver's seat. Figure 4 It is also shown that there is a dangerous person 3 in the back seat of vehicle 1, who poses a threat to the driver. Dangerous person 3 has hijacked the driver and vehicle 1. Since the driver and dangerous person 3 are in the enclosed space of vehicle 1 and the doors are locked in this scenario, if the driver tries to break free, shout, or manually trigger the alarm device as in the prior art, it may anger dangerous person 3 and thus potentially cause further harm to the driver.

[0046] Unlike existing technologies, the alarm method 100 of the present invention can not only reliably and quickly identify dangers, but also effectively and covertly output alarm signals to ensure the safety of occupants 2.

[0047] Below, refer to Figure 5 And combined Figures 1 to 4 The alarm method 100 of the present invention is described in detail.

[0048] Figure 5 A schematic flowchart illustrating an embodiment of an alarm method 100 for a vehicle 1 according to the present invention is shown.

[0049] like Figure 5 As shown, according to this embodiment, the alarm method 100 schematically includes steps S110 to S140.

[0050] In step S110, the controller 70 receives detection data related to the occupant 2 of the vehicle 1. This detection data may come from the in-vehicle camera 21 of the vision unit 20, the microphone 31 of the voice unit 30, the heart rate sensor 41 of the vital signs detection unit 40, the pressure sensor 51 of the tactile unit 50, or the touch screen.

[0051] These detection data include first data and second data. The first data represents non-interactive data involving occupant 2 that is not interacting with vehicle 1, and the second data represents interactive data involving occupant 2 that is interacting with vehicle 1.

[0052] For example, the first data may include at least one of the following: image data, voice data, and vital sign data. The second data may include at least one of the following: voice interaction data, image interaction data, and tactile interaction data. That is, for example, if the microphone 31 detects threatening or fearful words between the dangerous person 3 and the occupant 2, or the in-vehicle camera 21 detects a dangerous action made by the dangerous person 3 towards the occupant 2, or the heart rate sensor 41 detects that the occupant 2's heart rate is too fast, or the pressure sensor 51 detects a sudden change in the pressure of the occupant 2, then the first data is detected. If the microphone 31 detects that the occupant 2 interacts with the vehicle 1 via voice, such as the occupant 2 outputting a voice command to the vehicle system, then voice interaction data is detected; or if the in-vehicle camera 21 detects that the occupant 2 outputs a specific action or expression towards the in-vehicle camera 21, then image interaction data is detected.

[0053] In addition, tactile interaction data can be detected via the touchscreen. For example, if occupant 2 touches the central control screen, such as clicking an icon on the screen, tactile interaction data is detected.

[0054] In another embodiment, the second data may further include driving status data. Here, occupant 2 is the driver. The second data represents data relating to the driver's operational interaction with vehicle 1 and reflects the driver's driving status. This driving status data can be detected by sensors of vehicle 1, such as speed sensors, acceleration sensors, wheel angle sensors, accelerator pedal sensors, brake pedal sensors, etc.

[0055] According to one embodiment, the first data and / or the second data may be preprocessed via data fusion, such as data denoising, normalization, etc.

[0056] In step S120, a judgment mode corresponding to the detection data is activated based on the detection data. According to this embodiment, the judgment mode includes a first judgment mode and a second judgment mode. If the first data is detected, the first judgment mode is activated; if the second data is detected, the second judgment mode is activated.

[0057] exist Figure 4In this scenario, microphone 31 detects in-vehicle audio in real time, for example, at 44.1kHz and 128kbps; in-vehicle camera 21 detects in-vehicle images at 30 frames per second; and heart rate sensor 41 detects the heart rate changes of occupant 2 in real time. If, for example, a dangerous person 3 is detected saying to occupant 2 or the driver, "Don't shout, hand over the money," and / or in-vehicle camera 21 detects a dangerous person 3 holding a knife and holding occupant 2 hostage, and / or heart rate sensor 41 detects that occupant 2's heart rate exceeds 120 beats per minute, then first data is detected, and controller 70 activates a first judgment mode based on this first data. For example, if microphone 31 detects occupant 2, for example, conversing with the vehicle's infotainment system and / or in-vehicle camera 21 detects occupant 2 continuously blinking at the in-vehicle camera 21, or if occupant 2's clicks on screen icons are received via touchscreen, then second data is detected, and controller 70 activates a second judgment mode based on this second data.

[0058] Then, in step S130, the detection data is analyzed using the activated judgment mode to assess the dangerous state of occupant 2.

[0059] According to one embodiment, in a first judgment mode, the first data is analyzed using an artificial intelligence model to generate a first probability of danger for occupant 2. Here, the artificial intelligence model is exemplarily stored in the memory 72 of the controller 70. This artificial intelligence model may include a semantic context recognition model, such as a BERT+BiLSTM hybrid neural network model (BERT: Bidirectional Encoder Representations from Transformers, BiLSTM: Bidirectional Long Short-Term Memory), and / or an image recognition model, such as a convolutional neural network model (CNN), and / or a person state analysis model. Other models suitable for semantic context recognition and image recognition are also considered. Using these artificial intelligence models, data such as speech and images in the first data can be analyzed and processed. Alternatively, the artificial intelligence model may also be stored in a cloud server and can be invoked by the controller 70.

[0060] For example, the AI ​​model can match and analyze keywords in speech and assign corresponding weights to these keywords. For instance, the keyword "don't move" is assigned a weight of 0.8, while the keyword "be careful" is assigned a weight of 0.5. For example, the weights of these keywords can be summed and compared to a corresponding danger level threshold. Here, the danger level threshold can correspond to the probability of danger. If the sum of weights is greater than 0.5 and less than 1, it represents a danger level of 1, i.e., a danger probability of 30% to 50%; if the sum of weights is greater than or equal to 1 and less than 1.5, it represents a danger level of 2, i.e., a danger probability of 50% to 70%; and if the sum of weights is greater than or equal to 1.5, it represents a danger level of 3, i.e., a danger probability of 70% to 95%.

[0061] Furthermore, keywords such as "help" or "alarm" are directly and explicitly associated with dangerous situations, therefore these keywords are assigned a weight of 1. In other words, a single keyword "help" or "alarm" represents a 50% probability of danger.

[0062] In this invention, these keywords are used only to illustrate the solution of the invention and may also include other keywords not described herein.

[0063] For example, an artificial intelligence model may also include a keyword lexicon and each keyword may be pre-assigned a corresponding weight.

[0064] Furthermore, the AI ​​model can also additionally analyze the tone of voice and assign additional weights to keywords in the voice. For example, the tone of "shouting" and the tone of "calmly stating" are assigned different weights for keywords. For instance, if "don't move" is detected and the tone of voice is analyzed as "shouting," the weight of the keyword "don't move" can be adjusted from 0.8 to 1.

[0065] In addition, the AI ​​model can also identify stress characteristics such as voice tremor and abnormal speech rate, analyze these stress characteristics, and similarly assign corresponding weights to them.

[0066] Alternatively, AI models can analyze speech based on these stress features and adjust the weights of keywords in the speech accordingly.

[0067] Similarly, the AI ​​model can also match and analyze detected images and assign corresponding weights to them. For example, an image of a stranger (dangerous person 3) inside the car is assigned a weight of 0.2, an image of the terrified expression on occupant 2's face is assigned a weight of 0.5, and an image of the weapon in dangerous person 3 is assigned a weight of 0.8. These image weights can be summed and compared to corresponding danger level thresholds. If the sum of weights is greater than 0.2 and less than 0.5, it represents a danger level of 1, with a probability of danger of 30% to 50%; if the sum of weights is greater than or equal to 0.5 and less than 1.5, it represents a danger level of 2, with a probability of danger of 50% to 70%; and if the sum of weights is greater than or equal to 1.5, it represents a danger level of 3, with a probability of danger of 70% to 95%.

[0068] Furthermore, the artificial intelligence model can analyze the context and generate a danger probability when relevant keywords appear multiple times within a predetermined time. For example, if the keywords "money" and "not allowed" appear twice consecutively within 10 seconds, the danger probability is 60%. If they appear three times, the danger probability is 80%. According to one embodiment, in the second judgment mode, second data is analyzed using preset data and / or the artificial intelligence model to generate a second danger probability for occupant 2. The preset data exemplarily includes at least one of the following: preset voice data, preset image data, and preset tactile data. The artificial intelligence model, as described above, can include at least one of the following models: a semantic context recognition model, an image recognition model, and a personnel state analysis model.

[0069] For example, occupant 2 interacts with the vehicle's infotainment system and says, "Please order me a Hawaiian pizza." The second judgment mode is based on the activation of this interactive voice and compares the voice interaction data with preset voice data. If the preset voice data matches the voice interaction data, the generated second hazard probability is 95%. If the voice interaction data only partially matches the preset voice data (e.g., "Please order me takeout"), the generated second hazard probability is 50%.

[0070] For example, occupant 2 interacts with the vehicle's infotainment system and says, "It's too hot, please turn on the air conditioning." Although matching using preset data fails, the voice data can be further analyzed using an artificial intelligence model, and logical judgments can be made based on temperature data from the vehicle's interior / exterior temperature sensors. If there is a clear contradiction, such as the exterior temperature being 0 degrees Celsius and the interior temperature being below 20 degrees Celsius, there is a logical problem, and the generated second danger probability is 80%. If there is no clear logical problem, the generated second danger probability is 20%.

[0071] For example, occupant 2 can also output specific actions or expressions to the in-vehicle camera 21, such as frowning, sticking out the tongue, or blinking three times in succession. The processor 71 of the controller 70 compares the specific action or expression data, or image interaction data, with preset image data. If the image interaction data matches the preset image data, the generated second hazard probability is 90%; if they do not match, the generated second hazard probability is 20%.

[0072] For example, occupant 2 can also operate the touchscreen, such as the central control screen, in a specific manner. For instance, if occupant 2 taps the same icon on the screen three times consecutively, processor 71 compares this tactile interaction data with preset tactile data. If the tactile interaction data matches the preset tactile data, the generated second hazard probability is 90%; if they do not match, the generated second hazard probability is 20%. Here, the touchscreen can also output three tactile feedback signals to occupant 2 accordingly, to inform occupant 2 that the touchscreen has received the three taps input by occupant 2.

[0073] Furthermore, the AI ​​model can also analyze driving status data and similarly generate corresponding secondary hazard probabilities for abnormal driving conditions, such as sudden braking, rapid acceleration, alternating braking and acceleration, and non-active steering. Alternatively, these abnormal driving conditions can be combined with other secondary data for analysis.

[0074] Next, in step S140: the controller 70 generates an alarm signal based on the assessed dangerous state and outputs the alarm signal to the communication unit 60.

[0075] According to one embodiment, an alarm signal is generated and output based on a first hazard probability or a second hazard probability. This can be done, for example, by comparing the first hazard probability or the second hazard probability with a hazard probability threshold. For instance, if the first hazard probability or the second hazard probability is greater than or equal to 50%, an alarm signal is generated and output.

[0076] like Figure 4As shown, for example, if the microphone 31 only detects the keyword "Don't move," the AI ​​model assigns a weight of 0.8 to this keyword. This weight is between 0.5 and 1, and represents a danger probability of less than 50%. Therefore, the controller 70 does not generate an alarm signal at this time, thus avoiding false alarms in dangerous situations. If the microphone 31 detects the keyword "Help," the AI ​​model assigns a weight of 1 to this keyword. The single keyword "Help" or "Alarm" represents a danger probability of 50%. Therefore, the controller 70 can directly generate an alarm signal based on this single keyword and output the alarm signal to the communication unit, thereby clearly identifying the danger and quickly outputting an alarm signal to ensure the protection of occupant 2. The alarm signal is wirelessly transmitted to the external device 80 via the communication unit 60. This can be achieved through 5G / 4G cellular communication or NB-IoT low-power transmission.

[0077] For example, the alarm signal includes the current location of vehicle 1 and / or vehicle information and / or occupant information and / or information about dangerous persons, and is sent to a remote emergency call center. Therefore, the remote emergency call center can promptly obtain the relevant information from the alarm signal and take appropriate rescue measures.

[0078] Alternatively or additionally, the alarm signal may also be sent to emergency contacts and / or surrounding road users. Emergency contacts and / or surrounding road users will also be able to receive the alarm information and promptly call the emergency number and / or take rescue measures for occupant 2.

[0079] According to another embodiment, in step S140, the controller 70 may additionally output instructions for sending and / or storing real-time in-vehicle audio data and / or real-time video data and / or the real-time location of vehicle 1. Thus, the external device 80 can receive real-time in-vehicle audio data and / or real-time video data and / or the real-time location of vehicle 1 and can understand the in-vehicle situation and vehicle location in real time, thereby providing further support for rescue efforts. Furthermore, it can provide strong data support for subsequent evidence collection. Exemplarily, this data is transmitted in an encrypted manner using the national cryptographic standard SM4 and stored, for example, in a blockchain. This further improves the security of data transmission and the integrity of data storage.

[0080] Furthermore, if the second probability of danger exceeds the danger probability threshold, in addition to generating an alarm signal, the vehicle's infotainment system can generate normal voice interaction data in response to the occupant's (2) voice interaction data, such as outputting "Okay, we have matched you with the nearest pizza delivery shop" via speaker 32, and displaying the corresponding interaction information on the in-vehicle display device 22. This allows for a more concealed alarm and lowers the guard of the potentially dangerous person (3), preventing further harm to the occupant (2).

[0081] According to an embodiment not shown, verification can also be performed using different types of data in the first data, or by using different types of data in the second data.

[0082] Additionally, in the second judgment mode, the second data can be verified using the first data. For example, if the second data representing "Please order me a Hawaiian pizza" is detected, while the first data representing the image data from the in-vehicle camera 21 indicates that there are no strangers in the vehicle and no nervous expression of occupant 2 is detected, then even if the voice interaction data matches the preset voice data, the second danger probability will be corrected to 30%. This further improves the accuracy of occupant danger identification.

[0083] The present invention also protects a computer program product, wherein the computer program product includes computer instructions, which, when executed by processor 71, are used at least to assist in implementing the alarm method 100 described in one of the above embodiments.

[0084] Other advantages and alternative embodiments of the invention will be apparent to those skilled in the art. Therefore, the invention is not, in its broader sense, limited to the specific details, representative structures, and exemplary embodiments shown and described. Rather, those skilled in the art can make various modifications and substitutions without departing from the basic spirit and scope of the invention.

Claims

1. An alarm method (100) for a vehicle (1), wherein, The alarm method (100) includes: Step S110: Receive detection data related to the occupants (2) of the vehicle (1); Step S120: Activate the judgment mode corresponding to the detection data based on the detection data; Step S130: Analyze the detection data using the activated judgment mode to assess the dangerous state of the occupant (2); and Step S140: Generate and output an alarm signal based on the assessed hazardous condition.

2. The alarm method (100) according to claim 1, wherein, The detection data includes first data and second data. The first data represents non-interactive data involving the occupant (2) that is not interacting with the vehicle (1), and the second data represents interactive data involving the occupant (2) that is interacting with the vehicle (1). A first judgment mode is activated based on the first data, and a second judgment mode is activated based on the second data.

3. The alarm method (100) according to claim 2, wherein, In the first judgment mode, the first data is analyzed using an artificial intelligence model to generate a first danger probability of the occupant (2), and an alarm signal is generated and output based on the first danger probability; and / or In the second judgment mode, the second data is analyzed with the help of preset data and / or artificial intelligence model to generate the second danger probability of the occupant (2), and an alarm signal is generated and output based on the second danger probability.

4. The alarm method (100) according to claim 3, wherein, The first data includes at least one of the following: image data, voice data, vital sign data; and / or The second data includes at least one of the following: voice interaction data, image interaction data, haptic interaction data, driving status data; and / or The preset data includes at least one of the following: preset voice data, preset image data, and preset tactile data; and / or The artificial intelligence model includes at least one of the following models: semantic context recognition model, image recognition model, personnel state analysis model; and / or The first data and / or the second data are preprocessed via data fusion.

5. The alarm method (100) according to any one of claims 1 to 4, wherein, In step S140, additional instructions are output for sending and / or storing real-time in-vehicle audio data and / or real-time video data and / or the real-time location of the vehicle (1).

6. The alarm method (100) according to claim 5, wherein, The alarm signal and / or the real-time in-vehicle audio data and / or the real-time video data and / or the real-time location of the vehicle are sent to the remote emergency call center and / or emergency contacts and / or surrounding traffic participants. and / or The alarm signal and / or the real-time in-vehicle audio data and / or the real-time video data and / or the vehicle's real-time location are transmitted wirelessly; and / or The alarm signal includes the current location of the vehicle (1) and / or vehicle information and / or occupant information and / or dangerous personnel information.

7. A controller (70), wherein, The controller (70) includes a processor (71) and a memory (72) storing computer instructions that, when executed by the processor (71), are used to assist in implementing the alarm method (100) according to any one of claims 1 to 6.

8. An alarm system (10), wherein, The alarm system (10) includes: The vision unit (20) includes an image detection device for detecting in-vehicle images and a vision output device for outputting visual information; The voice unit (30) includes a voice detection device for detecting in-vehicle voice and a voice output device for outputting voice. A vital signs detection unit (40) is configured to detect the vital signs data of the occupant (2); The tactile unit (50) includes a tactile input device for inputting tactile signals and a tactile output device for outputting tactile signals; Communication unit (60), the communication unit (60) being configured to be data connected to an external device (80) and to output data to the external device (80); and According to claim 7, the controller (70) is data connected to the vision unit (20), the voice unit (30), the vital sign detection unit (40), the tactile unit (50) and the communication unit (60).

9. A vehicle (1), wherein, The vehicle (1) includes an alarm system (10) according to claim 8.

10. A computer program product, wherein, The computer program product includes computer instructions that, when executed by a processor (71), are used to assist in implementing the alarm method (100) according to any one of claims 1 to 6.

Images