An all-terrain vehicle and its multi-age mode intelligent safety control method, device, medium and product

Abstract

This application discloses an all-terrain vehicle (ATV) and its multi-age mode intelligent safety control method, equipment, medium, and product, relating to the field of vehicle control. The method includes: constructing a headcount age recognition model based on a YOLOv5s network model; inputting a driver area image into the headcount age recognition model to obtain a headcount age recognition result; determining a primary driver's face detection frame based on the headcount age recognition result, and using the age corresponding to the primary driver's face detection frame as the driver's age; determining a driving mode based on the driver's age; generating ATV control instructions based on the driving mode, and using the control instructions to complete intelligent safety control of the ATV. This application can automatically adjust the vehicle's driving mode based on the recognition results, thereby improving safety and driving experience during driving, and realizing safe and intelligent driving management.

Description

Technical Field

[0001] The present application relates to the field of vehicle control, and in particular to an all-terrain vehicle and its multi-age mode intelligent safety control method, equipment, medium and product. Background Art

[0002] Most ATV safety systems currently only support the safety needs of adult drivers and lack differentiated support for drivers of different age groups. This situation leads to limitations when dealing with children, adolescents, and elderly drivers, potentially failing to meet their specific safety and driving experience needs.

[0003] In particular, insufficient consideration of the safety and driving experience needs of children, teenagers and elderly drivers may lead to driving discomfort and increased safety risks, thereby affecting the overall user experience of all-terrain vehicles and causing insufficient balance between driving experience and safety.

[0004] Traditional all-terrain vehicle systems have limited real-time driver status monitoring and feedback capabilities, making them unable to dynamically adjust vehicle parameters to address the driver's varying physiological and psychological characteristics. This static human-machine interface limitation can lead to insufficient response to the driver's status, thus affecting driving safety and comfort.

[0005] In summary, in the safety control of all-terrain vehicles, in order to meet the safety needs of drivers and passengers, traditional systems often lack refined control for drivers of different age groups, and there are challenges in age recognition and driving mode adaptability. Summary of the Invention

[0006] The purpose of this application is to provide an all-terrain vehicle and its multi-age mode intelligent safety control method, equipment, medium and product, which can automatically adjust the vehicle's driving mode according to the recognition results, thereby improving the safety and driving experience during driving and realizing safe and intelligent driving management.

[0007] To achieve the above objectives, this application provides the following solutions:

[0008] In a first aspect, the present application provides a multi-age mode intelligent safety control method for an all-terrain vehicle, comprising:

[0009] Acquire driver area image;

[0010] Build a population age recognition model based on the YOLOv5s network model;

[0011] Inputting the driver area image into the headcount and age recognition model to obtain a headcount and age recognition result; the headcount and age recognition result includes the coordinates of the face detection frame and the age corresponding to the face detection frame;

[0012] Determine a primary driver's face detection frame based on the age recognition result of the number of people, and use the age corresponding to the primary driver's face detection frame as the driver's age;

[0013] Determine a driving mode based on the driver's age; the driving modes include: child mode, teenager mode, youth mode, adult mode and elderly mode;

[0014] A control instruction for the all-terrain vehicle is generated based on the driving mode, and the control instruction is used to complete intelligent safety control of the all-terrain vehicle.

[0015] Optionally, a person age recognition model is constructed based on the YOLOv5s network model, specifically including:

[0016] Obtain a sample data set; the sample data set includes a training set and a test set; the training set and the test set each include a plurality of training sample pairs; each training sample pair includes a historical image of the driver area and annotated data corresponding to the driver area image; the annotated data includes: coordinates of a face detection frame and an age corresponding to the face detection frame;

[0017] Based on the YOLOv5s network model, use the Model class to create an initial model;

[0018] The initial model is iteratively trained using a sample data set until the set conditions are met, and the trained initial model is used as a model for identifying the age of a person.

[0019] Optionally, determining a primary driver's face detection frame based on the age recognition result of the number of people specifically includes:

[0020] Determining the number of face detection frames in the age recognition result of the person;

[0021] When the number of face detection frames is 1, this face detection frame is used as the main driver's face detection frame;

[0022] When the number of face detection frames is greater than 1, the binocular point cloud distance collection of all pixels in each face detection frame is obtained based on the face detection frame coordinates;

[0023] Determine the average distance of the binocular point cloud distance collection, and determine the face detection frame with the smallest average distance as the main driver's face detection frame.

[0024] In a second aspect, the present application provides an all-terrain vehicle, comprising:

[0025] An image acquisition device, used for acquiring an image of the driver area;

[0026] an in-vehicle terminal connected to the image acquisition device and having an embedded headcount age recognition model, configured to use the headcount age recognition model to obtain a headcount age recognition result based on the driver area image; the headcount age recognition result including the coordinates of a face detection frame and the age corresponding to the face detection frame;

[0027] a vehicle controller, connected to the vehicle-mounted terminal, configured to determine a driving mode based on the age recognition result of the person, and to generate control instructions for the all-terrain vehicle based on the driving mode; the driving modes include: child mode, juvenile mode, youth mode, adult mode, and elderly mode;

[0028] A motor controller, connected to the vehicle controller, for generating motor parameter control instructions based on the control instructions;

[0029] The motor is connected to the motor controller and is used to execute the motor parameter control instructions.

[0030] Optionally, the vehicle controller includes: a child lock system;

[0031] The states of the child lock system include: a driving not allowed state, a driving allowed state and a safety protection state.

[0032] Optionally, when the driving mode is the child mode, the child lock system is turned on, and the state of the child lock system is that driving is not allowed;

[0033] The motor parameter control instruction generated by the motor controller is a motor power-off instruction.

[0034] Optionally, the child lock system includes an electronic screen;

[0035] When the driving mode is the youth mode, the child lock system is turned on and the state of the child lock system is a safety protection state;

[0036] In the safety protection state, the electronic screen displays the words "adult confirmation required" and generates a human-computer interaction interface including a confirmation button and a rejection button;

[0037] When the vehicle controller receives feedback information from the confirmation button, the state of the child lock system is changed to a driving permission state;

[0038] When the vehicle controller receives the feedback information of the negative button, the state of the child lock system is changed to the driving permission state.

[0039] In a third aspect, the present application provides a computer device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the multi-age mode intelligent safety control method for an all-terrain vehicle as described above.

[0040] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any one of the above-mentioned methods for intelligent safety control of multi-age modes for all-terrain vehicles.

[0041] In a fifth aspect, the present application provides a computer program product, comprising a computer program, which, when executed by a processor, implements the steps of any one of the above-mentioned methods for intelligent safety control of multi-age modes for all-terrain vehicles.

[0042] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0043] This application provides an all-terrain vehicle and its multi-age mode intelligent safety control method, equipment, medium and product, introduces the YOLOv5 network model for face target detection and age recognition, realizes dynamic recognition and position confirmation of the main driver, and can more efficiently identify the driver's status and perform safety control through real-time monitoring and data processing compared to the static human-computer interaction interface of the traditional system. Determining the driving mode according to the age of the main driver can balance driving responsiveness and safety, and improve safety and driving experience during driving. In addition, this dynamic adjustment can better adapt to the actual needs of the driver and the driving environment. Compared with the traditional fixed parameter setting, it has significant advantages and can achieve safe and intelligent driving management. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 A flowchart of a multi-age mode intelligent safety control method for an all-terrain vehicle provided in one embodiment of the present application;

[0046] Figure 2 A flowchart of an implementation method for multi-age mode intelligent safety control of an all-terrain vehicle provided in another embodiment of the present application;

[0047] Figure 3A diagram of a multi-age mode intelligent safety control architecture for an all-terrain vehicle provided in another embodiment of the present application;

[0048] Figure 4 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0050] In order to make the purpose, features and advantages of this application more obvious and easy to understand, this application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0051] The embodiment of the present application provides a multi-age mode intelligent safety control method for an all-terrain vehicle, such as Figure 1 As shown, a multi-age mode intelligent safety control method for an all-terrain vehicle is provided. The method is executed by a computer device, specifically a computer device such as a terminal or a server, or a terminal and a server. In the embodiment of the present application, the method is described by taking the application of the method to a vehicle terminal as an example, and includes the following steps 101 to 106. Among them:

[0052] Step 101: Acquire a driver area image.

[0053] Step 102: Build a person age recognition model based on the YOLOv5s network model.

[0054] Step 103: Input the driver area image into the headcount and age recognition model to obtain the headcount and age recognition result. The headcount and age recognition result includes the coordinates of the face detection frame and the age corresponding to the face detection frame.

[0055] Step 104: Determine the main driver's face detection frame based on the age recognition results, and use the age corresponding to the main driver's face detection frame as the driver's age.

[0056] Step 105: Determine a driving mode based on the driver's age. The driving modes include: child mode, teenager mode, youth mode, adult mode, and elderly mode.

[0057] Step 106: Generate control instructions for the all-terrain vehicle based on the driving mode, and use the control instructions to complete intelligent safety control of the all-terrain vehicle.

[0058] In another exemplary embodiment of the present application, in order to further improve the accuracy of number and age detection, the implementation process of the above step 102 may include:

[0059] Step 1021: Obtain a sample dataset. The sample dataset includes a training set and a test set. Each training set and test set includes multiple training sample pairs. Each training sample pair includes a historical image of the driver's area and annotated data corresponding to the driver's area image. The annotated data includes the coordinates of the face detection frame and the age corresponding to the face detection frame.

[0060] Step 1022: Based on the YOLOv5s network model, use the Model class to create an initial model.

[0061] Step 1023: The initial model is iteratively trained using the sample data set until the set conditions are met, and the trained initial model is used as the age recognition model for people.

[0062] In another exemplary embodiment of the present application, in order to further improve the accuracy of number and age detection, the implementation process of the above step 104 includes:

[0063] Step 1041: Determine the number of face detection frames in the age recognition result. For example, the number of face detection frames is n.

[0064] Step 1042: When the number of face detection frames is 1, this face detection frame is used as the main driver's face detection frame.

[0065] Step 1043: When the number of face detection frames is greater than one, obtain the binocular point cloud distances for all pixels within each face detection frame based on the face detection frame coordinates. For example, each face detection frame is numbered {id_0, id_1...id_n}, and each face detection frame has corresponding lower left coordinates (xleft_n, ydown_n) and upper right coordinates (xright_n, yup_n). The binocular point cloud distances for all pixels within each face detection frame are expressed as: d_n = {d(xleft_n, ydown_n)...d(xright_n, yup_n)}. The average distance of the binocular point cloud distances is calculated as dmean_n.

[0066] Step 1044: Determine the average distance of the binocular point cloud distance collection, and determine the face detection frame with the smallest average distance as the main driver's face detection frame.

[0067] Based on the same inventive concept, the present application also provides an all-terrain vehicle for implementing the aforementioned multi-age intelligent safety control method for an all-terrain vehicle. The solution provided by this all-terrain vehicle is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more of the following all-terrain vehicle embodiments can be found in the aforementioned limitations of the multi-age intelligent safety control method for an all-terrain vehicle, and will not be further elaborated here.

[0068] In an exemplary embodiment, an all-terrain vehicle is provided, comprising: an image acquisition device, a vehicle-mounted terminal, a vehicle controller, a motor controller, and a motor.

[0069] The image acquisition device is used to acquire images of the driver area. For example, the image acquisition device used is a binocular camera.

[0070] The vehicle-mounted electronic control unit (ECU) is connected to the image acquisition device and is embedded with a headcount and age recognition model. It uses this model to generate headcount and age recognition results based on the driver area image. The headcount and age recognition results include the coordinates of the face detection frame and the age corresponding to the face detection frame.

[0071] The vehicle control unit (VCU) is connected to the vehicle terminal and is primarily responsible for determining the driving mode based on age recognition and generating control commands for the ATV based on the driving mode. Driving modes include: Child Mode, Teenager Mode, Youth Mode, Adult Mode, and Senior Mode.

[0072] The motor controller is connected to the vehicle controller and is mainly used to generate motor parameter control instructions based on control instructions.

[0073] The motor is connected to the motor controller, which is mainly used to execute motor parameter control instructions.

[0074] In another exemplary embodiment of the present application, the vehicle controller includes a child lock system. The child lock system has states including a driving-disallowed state, a driving-allowed state, and a safety protection state.

[0075] When the driving mode is child mode, the child lock system is turned on and the status of the child lock system is not allowed to drive.

[0076] The motor parameter control instruction generated by the motor controller is a motor power-off instruction.

[0077] Furthermore, the child lock system may also include an electronic screen.

[0078] When the driving mode is youth mode, the child lock system is turned on and the status of the child lock system is safety protection status.

[0079] In the safety protection state, the electronic screen displays the words "adult confirmation is required" and generates a corresponding human-computer interaction interface including a confirmation button and a rejection button. In addition, the confirmation button and the rejection button can also be a key unit external to the electronic screen.

[0080] When the vehicle controller receives the feedback information of the confirmation button, the status of the child lock system is changed to the driving permission status.

[0081] When the vehicle controller receives the feedback information of the negative button, the state of the child lock system is changed to the driving permission state.

[0082] In another exemplary embodiment of the present application, Figure 2 and Figure 3 Taking the implementation process shown as an example, the multi-age mode intelligent safety control process of the all-terrain vehicle provided by this application is described. Specifically:

[0083] Step 1: Camera installation and image collection.

[0084] Step 1.1: Camera installation and field of view setting:

[0085] Install a binocular camera at the front of the vehicle to ensure a clear image of the driver's area. Adjust the camera's viewing angle to ensure a complete image of the driver and passengers.

[0086] Step 1.2, image collection:

[0087] Get real-time image frames captured by the camera. Transmit and save the images to a cloud server and create a dataset of the collected images.

[0088] Step 2: Train the age recognition model on the cloud, that is, the training process is carried out on the cloud server.

[0089] Step 2.1, Initialization settings:

[0090] Set the total number of training epochs to 300, the batch size to 64, and the initial weights to opt.weights.

[0091] Use a random seed to initialize the random number generator to ensure the repeatability of the training process.

[0092] Determine the number of classes and model parameter settings.

[0093] Step 2.2, create the initial model:

[0094] Create an initial model using the Model class based on the YOLOv5s network model.

[0095] Step 2.3. Verify image size:

[0096] Calculate the maximum step size of the initial model and ensure that the input image size is a multiple of the maximum step size to meet the input requirements of the initial model.

[0097] Step 2.4, select the optimizer:

[0098] Select the adam optimizer according to the value of opt.adam, set parameters such as learning rate (lr) and momentum (momentum), and add the parameter group to the optimizer.

[0099] Step 2.5, load the pre-trained model weights:

[0100] Use torch.load to load the checkpoint of the pre-trained model into ckpt.

[0101] Convert the loaded initial model parameters (i.e., the checkpoint of the pre-trained model) to float type and load it into the current model (i.e., the initial model) to ensure compatibility with the current model.

[0102] If the checkpoint contains the optimizer state, the optimizer state and the best performance indicators are loaded.

[0103] Step 2.6, set the training start:

[0104] Set the starting epoch to epoch+1 recorded in the checkpoint and continue training from where the last training ended.

[0105] Step 2.7, Learning Rate Scheduler:

[0106] Use the cosine annealing scheduler to adjust the learning rate according to the current epoch.

[0107] Step 2.8, create a data loader:

[0108] Create a data loader and batch load the previously collected image dataset for training. Set the number of processes for parallel data loading to 2 and set data shuffling.

[0109] Step 2.9, training loop:

[0110] From the beginning to the end of the epoch, each epoch performs the following operations:

[0111] (1) Set the initial model to training mode and enable dropout and batch normalization.

[0112] (2) Calculate the weight of the category weights as needed and reselect the indexes in the dataset for weighted sampling.

[0113] (3) Initialize a tensor to store the average loss.

[0114] (4) Print key training information such as epoch, GPU memory usage, loss, etc.

[0115] (5) Traverse each batch of data in the training data loader:

[0116] (6) Adjust the learning rate and momentum as needed.

[0117] (7) Preprocess the input image and pass it to the initial model for forward propagation to obtain the predicted value.

[0118] (8) Calculate the loss between the model prediction value and the target label.

[0119] (9) Perform backpropagation and perform gradient updates of the optimizer based on the accumulated gradient steps.

[0120] (10) Update and print the average loss, GPU memory usage, and other training metrics.

[0121] (11) Update the learning rate scheduler.

[0122] (12) Calculate the average precision (mAP).

[0123] (13) Update the properties of the initial model, the moving average parameter EMA.

[0124] (14) Determine whether it is the last epoch. If it is not the last epoch, perform model testing to evaluate the performance of the initial model after training.

[0125] (15) The optimal performance index is updated according to the fitness of the trained initial model, and the checkpoint of the trained initial model is saved to obtain a headcount age recognition model.

[0126] (16) Convert the saved age recognition model from a .pth model to a .onnx model, and then from the .onnx model to a .rknn model. Deploy the converted model to the lightweight computing board of the ECU.

[0127] Step 3: Driver identification and age inference, i.e. YOLOv5 age recognition.

[0128] Step 3.1, number detection:

[0129] Initialize RKNNAPI in ECU and initialize the operating environment.

[0130] Initialize an AgeEstimator object and pass in the specified .rknn model path.

[0131] Use OpenCV to read image frames from a camera or other source. If the image is successfully read, call the rknn model for inference. The inference output is the coordinates of the face detection frame and the corresponding age.

[0132] Traverse each face detection frame coordinate and extract the coordinates of each detected face detection frame.

[0133] Returns a list Agelist. Each element in the list Agelist is a dictionary containing a 'box' key. The value corresponding to the 'box' key is a tuple containing the coordinates of the face detection box (x1, y1, x2, y2).

[0134] Based on the face detection frames and age predictions output by the age recognition model, the length of Agelist is calculated to obtain the number of face detection frames. The number of face detection frames represents the number of detected passengers.

[0135] Step 3.2: Determine the location and age group of the primary driver:

[0136] If the number of detected face detection frames is 1, it is considered that the person corresponding to the detected face detection frame is the main driver.

[0137] If the detected face detection frame data is greater than 1, obtain the binocular point cloud distance collection d_n = {d(xleft_n, ydown_n)...d(xright_n, yup_n)} of all pixels in each face detection frame, calculate the average distance dmean_n of the collection, and the face detection frame with the smallest average distance dmean_n is determined to be the face detection frame of the main driver, that is, the main driver's face detection frame.

[0138] Step 4: Determine the multi-age system mode and adjust the driving mode and motor parameters.

[0139] Step 4.1: Determine the driving mode based on the age in the detection result.

[0140] Step 4.2: Driving mode and motor parameter adjustment.

[0141] (1) Driving mode classification and settings:

[0142] The driving mode to be adopted is determined according to the age group of the main driver: child mode (0-8 years old), teenager mode (8-12 years old), youth mode (12-18 years old), adult mode (18-60 years old) and elderly mode (60+ years old).

[0143] (2) Motor parameter adjustment.

[0144] 1) Children's Mode (0-8 years old):

[0145] A. Startup control:

[0146] Child lock system on: The VCU's child lock system receives the system on signal, ensuring that children cannot accidentally start the vehicle. When the child lock system is on, the external child lock screen lights up to indicate the child lock system status.

[0147] Children are not allowed to start the vehicle: The child lock system is set to the "Not Allowed" state, and the motor controller cuts off power to the motor, preventing children from driving the vehicle under any circumstances. The external child lock screen displays "Not Allowed" based on the VCU child lock system's instructions.

[0148] B. Motor parameter control:

[0149] Motor controller power off: When the child lock system is set to not allow driving, the motor controller immediately controls the motor power off to ensure that the vehicle stops running. The motor controller controls the motor output voltage to 0V, that is, V_motor = 0, where V_motor is the motor output voltage.

[0150] 2) Teenager Mode (8-12 years old):

[0151] A. Startup control:

[0152] Child lock system is on: The child lock system of the VCU receives the on signal and sets the child lock system to a safety protection state. When the child lock system is on, the external child lock screen (i.e., the electronic screen) lights up to display the status of the child lock system. The external child lock screen displays the words "adult confirmation required" according to the instructions issued by the child lock system. The two buttons have confirmation and denial respectively. If the user presses the confirmation button, the instruction is uploaded to the VCU, the child lock system changes to the driving permission state, and the instruction is sent to display the words "driving permission allowed". If the user presses the denial button, the instruction is uploaded to the VCU, the child lock system changes to the driving not allowed state, and the instruction is sent to display the words "driving not allowed".

[0153] B. Adult confirmation is required: An adult must go through additional safety start controls to start the vehicle, ensuring a safe and controllable driving process.

[0154] C. Motor parameter control:

[0155] Child lock system status confirmation: Determine whether driving is allowed based on the status of the child lock system.

[0156] In the safety protection state, the motor controller powers off the motor to ensure a safe vehicle stop. In the non-driving state, if an adult confirms that driving is not permitted, the child lock system switches to the non-driving state, and the motor controller powers off the motor to ensure a safe vehicle stop. In the driving permission state, if an adult confirms that driving is permitted, the child lock system switches to the driving permission state, and the motor controller powers on the motor, but limits the motor power to no more than 1.5 kW.

[0157] a. Motor power limit: Limit motor power to 1.5 kW to ensure safety and controllability during driving. P_motor <= 1500W, where P_motor is the motor power.

[0158] b. Speed ​​and acceleration settings: Use the controller to adjust the motor speed and set a lower acceleration to ensure that the driving force is suitable for the teenager's driving skills.

[0159] c. Maximum driving speed: Set the maximum driving speed to 15 km / h to stay within a safe range and avoid accidents caused by excessive speed. v_max = 20 km / h, where v_max is the maximum allowed driving speed.

[0160] 3) Youth mode (12-18 years old).

[0161] A. Startup control:

[0162] Normal start conditions: Allows young drivers to start the vehicle independently without special safety start controls.

[0163] B. Motor parameter control:

[0164] a. Motor power setting: Motor power is controlled at 2.5 kW to provide adequate power response while ensuring driving comfort and safety. P_motor <= 2500W.

[0165] b. Acceleration regulation: Allows moderate acceleration to support the comfortable and safe driving needs of young drivers.

[0166] c. Maximum driving speed setting: Set the maximum driving speed for the medium level to 20 km / h, allowing young drivers to move faster while maintaining safety. v_max = 20 km / h.

[0167] 4) Adult mode (18-60 years old).

[0168] A. Startup control:

[0169] Normal start conditions: An adult driver can start the vehicle autonomously without additional confirmation or safety controls.

[0170] B. Motor parameter control:

[0171] a. Motor power adjustment: Motor power is controlled between 3.5-4 kW to provide sufficient power to support higher speeds and dynamic driving requirements. P_motor <= 3500W or P_motor <= 4000W.

[0172] b. Acceleration setting: Allows higher acceleration to ensure the vehicle has the ability to accelerate quickly and drive dynamically.

[0173] c. Maximum driving speed setting: Set the maximum driving speed to 35 km / h to fully utilize the vehicle's speed performance and meet the driving needs of adult drivers. v_max = 35 km / h.

[0174] 5) Elderly Mode (60+ years old)

[0175] A. Startup control:

[0176] Normal starting conditions: Elderly drivers are allowed to start the vehicle independently without any special starting restrictions.

[0177] B. Motor parameter control:

[0178] a. Motor power adjustment: Motor power is controlled at 2 kW, reducing power output to ensure smooth and safe driving. P_motor <= 2000W.

[0179] b. Acceleration limit: Limit the acceleration of the motor to avoid overly sensitive responses and reduce the risk of accidents during driving.

[0180] c. Maximum driving speed reduction: Reduce the maximum driving speed to 15 km / h, keeping it within a safe range and improving control and safety during driving. v_max = 15 km / h.

[0181] Step 5: Control signal transmission and processing.

[0182] Step 5.1, ECU processing of age recognition system:

[0183] The ECU receives the data processed by the camera image, including the recognition result of the main driver's age group.

[0184] Step 5.2: VCU communicates with the motor controller:

[0185] The recognition result of the main driver's age group is transmitted to the VCU.

[0186] The VCU adjusts the vehicle's driving mode and parameter settings based on the data received.

[0187] The VCU's child lock system receives driving mode status information via the CAN bus communication protocol. When the driving mode is updated, the system automatically detects and updates the child lock system status in real time. The VCU is connected to an external electronic screen and two buttons. When the child lock system is engaged, the screen illuminates to display the child lock system status. The VCU's child lock system status is divided into: Not Allowed, Allowed, and Safety Protection.

[0188] In summary, in the all-terrain vehicle provided above in this application, a camera is installed at the front of the vehicle to shoot the driving area. Real-time image frames of the driver and passengers are obtained through the camera and saved to the cloud server to establish a data set. The YOLOv5 model is used for training on the cloud server to achieve face target detection and age estimation. The trained model is deployed to the ECU, and the camera identifies the number of passengers through the headcount and age recognition model, and identifies the main driver in the front position through the binocular algorithm. The age of the main driver is identified by the headcount and age recognition model, and is divided into children (0-8 years old), teenagers (8-12 years old), youth (12-18 years old), adults (18-60 years old) and the elderly (60+ years old). Different driving modes are matched to different age groups, and the child lock system is automatically adjusted to determine whether the vehicle is allowed to start, motor power, motor speed, acceleration, maximum allowable driving speed and other driving data.

[0189] Based on the above description, compared with the prior art, this application has the following advantages:

[0190] 1. Intelligent Safety Control: This application incorporates the YOLOv5 model for facial target detection and age recognition, enabling dynamic identification and location confirmation of the primary driver. Compared to the static human-computer interaction interface of traditional systems, this approach, through real-time monitoring and data processing, can more efficiently identify the driver's status and implement safety control.

[0191] 2. Dynamic Motor Parameter Adjustment: Intelligently adjusts motor power, acceleration, and maximum driving speed based on the driver's age group to balance driving responsiveness and safety. This dynamic adjustment better adapts to the driver's actual needs and driving environment, offering significant advantages over traditional fixed parameter settings.

[0192] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 4As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O) and a communication interface. The processor, memory and input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store multi-age mode intelligent safety control data of an all-terrain vehicle. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a multi-age mode intelligent safety control method for an all-terrain vehicle is implemented.

[0193] Those skilled in the art will understand that Figure 4 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present application and does not constitute a limitation on the computer device to which the solution of the present application is applied. A specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps of the above-mentioned method embodiments when executing the computer program.

[0194] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0195] In an exemplary embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0196] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0197] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0198] The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processors involved in the various embodiments provided herein may include, but are not limited to, general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic units, data processing logic units based on quantum computing, and the like.

[0199] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0200] This document uses specific examples to illustrate the principles and implementation methods of this application. The description of the above examples is only intended to help understand the method and core concept of this application. At the same time, for those skilled in the art, based on the concept of this application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. An intelligent safety control method for all-terrain vehicles in multiple age modes, characterized by: The multi-age mode intelligent safety control method for an all-terrain vehicle includes: Acquire driver area image; Build a crowd age recognition model based on the YOLOv5s network model; convert the crowd age recognition model from a .pth model to a .onnx model, and then from the .onnx model to a .rknn model; Inputting the driver area image into the headcount and age recognition model to obtain a headcount and age recognition result; the headcount and age recognition result includes the coordinates of the face detection frame and the age corresponding to the face detection frame; Determine the main driver's face detection frame based on the age recognition result of the number of people, and use the age corresponding to the main driver's face detection frame as the driver's age; determine the main driver's face detection frame based on the age recognition result of the number of people, including: determining the number of face detection frames in the age recognition result of the number of people; when the number of face detection frames is 1, use this face detection frame as the main driver's face detection frame; when the number of face detection frames is greater than 1, obtain the binocular point cloud distance collection of all pixels in each face detection frame based on the face detection frame coordinates; determine the average distance of the binocular point cloud distance collection, and determine the face detection frame with the smallest average distance as the main driver's face detection frame; wherein, initialize an AgeEstim ator object, passing in the specified .rknn model path; using OpenCV to read the image frame; if the image is successfully read, calling the .rknn model for inference, the inference outputs the face detection frame coordinates and the corresponding age; traversing each face detection frame coordinate, extracting the coordinates of each detected face detection frame; returning a list Agelist; each element in the list Agelist is a dictionary containing a 'box' key; the value corresponding to the 'box' key is a tuple containing the face detection frame coordinates (x1, y1, x2, y2); based on the face detection frame and age prediction output by the headcount age recognition model, calculate the length of Agelist to obtain the number of face detection frames; Determine a driving mode based on the driver's age; the driving modes include: child mode, teenager mode, youth mode, adult mode and elderly mode; A control instruction for the all-terrain vehicle is generated based on the driving mode, and the control instruction is used to complete intelligent safety control of the all-terrain vehicle.

2. The multi-age mode intelligent safety control method for all-terrain vehicles according to claim 1, characterized in that: Build a crowd age recognition model based on the YOLOv5s network model, specifically including: Obtain a sample data set; the sample data set includes a training set and a test set; the training set and the test set each include a plurality of training sample pairs; each training sample pair includes a historical image of the driver area and annotated data corresponding to the driver area image; the annotated data includes: coordinates of a face detection frame and an age corresponding to the face detection frame; Based on the YOLOv5s network model, use the Model class to create an initial model; The initial model is iteratively trained using a sample data set until the set conditions are met, and the trained initial model is used as a model for identifying the age of a person.

3. An all-terrain vehicle, characterized in that: The all-terrain vehicle comprises: An image acquisition device, used for acquiring an image of the driver area; The vehicle terminal is connected to the image acquisition device and is implanted with a headcount and age recognition model, which is used to adopt the headcount and age recognition model to obtain a headcount and age recognition result based on the driver area image; the headcount and age recognition result includes the coordinates of the face detection frame and the age corresponding to the face detection frame; the number of face detection frames in the headcount and age recognition result is determined; when the number of face detection frames is 1, this face detection frame is used as the main driver's face detection frame; when the number of face detection frames is greater than 1, the binocular point cloud distance collection of all pixels in each face detection frame is obtained based on the face detection frame coordinates; the average distance of the binocular point cloud distance collection is determined, and the face detection frame with the smallest average distance is determined as the main driver's face detection frame; wherein, the initial Initialize an AgeEstimator object and pass in the specified .rknn model path; use OpenCV to read the image frame; if the image is successfully read, call the .rknn model for inference, and the inference outputs the face detection frame coordinates and the corresponding age; traverse each face detection frame coordinate and extract the coordinates of each detected face detection frame; return a list Agelist; each element in the list Agelist is a dictionary containing a 'box' key; the value corresponding to the 'box' key is a tuple containing the face detection frame coordinates (x1, y1, x2, y2); based on the face detection frames and age predictions output by the headcount age recognition model, calculate the length of Agelist to obtain the number of face detection frames; a vehicle controller, connected to the vehicle-mounted terminal, configured to determine a driving mode based on the age recognition result of the person, and to generate control instructions for the all-terrain vehicle based on the driving mode; the driving modes include: child mode, juvenile mode, youth mode, adult mode, and elderly mode; A motor controller, connected to the vehicle controller, for generating motor parameter control instructions based on the control instructions; The motor is connected to the motor controller and is used to execute the motor parameter control instructions.

4. The all-terrain vehicle according to claim 3, characterized in that The vehicle controller includes: a child lock system; The states of the child lock system include: a driving not allowed state, a driving allowed state and a safety protection state.

5. The all-terrain vehicle according to claim 4, characterized in that When the driving mode is the child mode, the child lock system is turned on, and the state of the child lock system is that driving is not allowed; The motor parameter control instruction generated by the motor controller is a motor power-off instruction.

6. The all-terrain vehicle according to claim 4, characterized in that The child lock system includes an electronic screen; When the driving mode is the youth mode, the child lock system is turned on and the state of the child lock system is a safety protection state; In the safety protection state, the electronic screen displays the words "adult confirmation required" and generates a human-computer interaction interface including a confirmation button and a rejection button; When the vehicle controller receives feedback information from the confirmation button, the state of the child lock system is changed to a driving permission state; When the vehicle controller receives the feedback information of the negative button, the state of the child lock system is changed to the driving permission state.

7. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the multi-age mode intelligent safety control method for an all-terrain vehicle according to any one of claims 1 to 2.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the multi-age mode intelligent safety control method for an all-terrain vehicle according to any one of claims 1 to 2 is implemented.

9. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the multi-age mode intelligent safety control method for an all-terrain vehicle according to any one of claims 1 to 2 is implemented.

Images