A multiple collision protection method and system, medium and electronic device

Abstract

The application relates to a multiple anti-collision protection method and system, a medium and an electronic device, which comprises the following steps: obtaining real-time monitoring data from a collected video stream; the monitoring data comprises the motion state of an obstacle recorded in each frame of the video stream; defining a vehicle safety interval corresponding to the relative distance between the obstacle and the vehicle body, formulating multiple anti-collision measures in different intervals; finally, matching corresponding measures according to the current vehicle safety interval, that is, reducing the reaction interval of the driver by broadcasting the motion details of the obstacle through voice, so that the driver makes a decision in the shortest time; meanwhile, the brake, the throttle and the airbag can be automatically controlled according to the misoperation of the driver's side, the vehicle is assisted to slow down and the collision force is buffered, and the vehicle-mounted equipment controls the anti-collision protection of the personnel in the vehicle. In the above scheme, the safety interval is divided and multiple anti-collision measures are formulated, so that the occurrence of a collision accident in the vehicle driving process is prevented, and the safety in the vehicle driving process is improved.

Description

Technical Field

[0001] This invention relates to the field of automotive collision avoidance technology, specifically to a multi-layered collision avoidance protection method and system, medium, and electronic equipment. Background Technology

[0002] With societal development, automobile production has increased significantly, leading to a surge in the number of vehicles on the road and making road safety a major concern. The rising rate of car accidents in recent years is primarily attributed to drivers failing to notice potential collisions and operational errors, resulting in accidents.

[0003] Currently, there are various patents for automotive collision avoidance both domestically and internationally. Application number "CN2892545Y," titled "Automotive Intelligent Collision Avoidance System," uses laser and microwave ranging, computer calculations, and electrical signals to magnetize magnets, preventing collisions through magnetic repulsion. Application number "CN102642510B," titled "An Image-Based Vehicle Collision Warning Method," uses sensors to collect images of vehicles approaching them, analyzes image changes to calculate distances, and then issues warnings. However, these patents primarily rely on warnings to notify drivers to take action to avoid collisions. While this method can remind drivers to avoid collisions to some extent, it provides very little information, requiring drivers to locate the collision object and observe its current state, which can distract them from focusing on driving. Furthermore, existing solutions offer only a single safety measure and do not provide further protection against unexpected collisions. Summary of the Invention

[0004] Therefore, it is necessary to provide a multi-layered anti-collision protection method and system, medium and electronic equipment to address the above-mentioned technical problems, setting up multiple safety defenses and protection measures to prevent automobile collision accidents and provide safety guarantees for accidental collision events.

[0005] To achieve the above-mentioned objectives, the present invention adopts the following technical solution:

[0006] In a first aspect, the present invention provides a multi-layered anti-collision protection method, the method comprising:

[0007] Real-time monitoring data is obtained from the acquired video stream; wherein, the monitoring data includes: the obstacle motion state contained in each frame of the video stream;

[0008] Define the vehicle safety zone corresponding to the motion state of the obstacle, and formulate multiple collision avoidance measures for different zones;

[0009] Appropriate collision avoidance measures are matched to the current vehicle safety range, and the on-board equipment is controlled to protect the occupants from collisions.

[0010] Optionally, obtaining real-time monitoring data from the acquired video stream includes:

[0011] During vehicle startup, video stream data of the road ahead is collected in real time using video acquisition equipment; wherein, the video acquisition equipment includes: cameras installed at the front and rear of the vehicle respectively; the cameras include one or more of wide-angle cameras, high-definition zoom cameras and infrared cameras;

[0012] The video acquisition device is connected to the detection network and the YOLOx-S processor. The acquired video stream data is sent to the YOLOx-S processor to identify the obstacle information contained in each frame of the video stream. The motion state of the obstacle is obtained by calling the monocular ranging API interface multiple times.

[0013] The obstacle information includes the type of obstacle;

[0014] The motion state of the obstacle includes: the obstacle's speed, direction, and relative distance to the current vehicle body.

[0015] Optionally, the real-time acquisition of video stream data of the road ahead via the video acquisition device includes: after receiving the vehicle start-up sensing signal, turning on the cameras installed at the front and rear of the vehicle to enter the monitoring mode, performing video stream acquisition relative to the road ahead of the camera, and obtaining video stream data.

[0016] Optionally, the process of identifying obstacle information contained in each frame of the video stream includes: disassembling the frame images in the video stream data frame by frame according to a preset frequency, converting each frame image from color to grayscale, and then performing adaptive threshold segmentation to obtain a binary image; removing noise from the binary image, and using an improved YOLOx-S algorithm to detect obstacles in the image;

[0017] The image containing obstacles is matched with historical image data in a custom OpenCV image database. Based on the image matching results, it is determined whether the obstacle in the image belongs to a known obstacle type in the database.

[0018] Optionally, obtaining the motion state of the obstacle by repeatedly calling the monocular ranging API interface includes:

[0019] The relative distance, speed, and direction of the obstacle to the current vehicle are obtained from each call to the monocular ranging API interface and are passed to the equalization function interface to output the distance and speed values ​​that are closest to the actual driving speed.

[0020] Optionally, the vehicle safety zone corresponding to the defined motion state of the obstacle includes:

[0021] The interval where the distance to an obstacle exceeds a preset threshold is defined as the safe zone; the interval where the distance to an obstacle is less than the maximum braking distance is defined as the sensitive zone; and the interval where the distance obtained by multiplying the vehicle's airbag deployment time by the obstacle's speed is defined as the danger zone.

[0022] Optionally, matching appropriate anti-collision measures to the current vehicle safety zone and controlling the on-board equipment to provide anti-collision protection for the driver includes:

[0023] When the vehicle is in a safe zone, a command is sent to the in-vehicle voice device to trigger the in-vehicle voice device to provide obstacle information to the driver's side.

[0024] When the car is driving in a sensitive area, commands are sent to the in-vehicle voice device and the ECU module respectively, triggering the in-vehicle voice device to provide voice prompts to the driver's side about the obstacle;

[0025] The ECU module magnetizes the electromagnets inside the vehicle, controlling the gravity control pedal and accelerator pedal to adjust to a preset height to reduce vehicle speed.

[0026] When the car is in a danger zone, commands are sent to the ECU module of the in-vehicle voice device and the airbag, triggering the in-vehicle voice device to provide voice prompts to the driver's side about the obstacle and controlling the deployment of the airbags inside the vehicle.

[0027] Secondly, the present invention provides a multi-layer anti-collision protection system, the system comprising:

[0028] The monitoring module is used to obtain real-time monitoring data from the acquired video stream; wherein, the monitoring data includes: the obstacle motion status contained in each frame of the video stream;

[0029] The definition module is used to define the vehicle safety zone corresponding to the motion state of the obstacle and to formulate multiple collision avoidance measures for different zones;

[0030] The control module is used to match appropriate anti-collision measures to the current vehicle safety zone and control the on-board equipment to protect the occupants from collisions.

[0031] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method as described in any one of the first aspects.

[0032] Fourthly, the present invention provides an electronic device including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the method as described in any one of the first aspects.

[0033] Compared with the closest existing technology, the present invention has the following advantages:

[0034] This invention proposes a multi-layered collision avoidance protection method, system, medium, and electronic device to address the problems of drivers being unable to make decisions based on the alarm content after hearing it and having no further measures to take when an accident is about to occur.

[0035] Compared to current alarm prompts, this invention uses voice broadcasting of detailed obstacle movement information, reducing driver reaction time and enabling drivers to make decisions in the shortest possible time. Furthermore, it intelligently controls the brakes and accelerator to assist in vehicle deceleration in case of driver inattention or misoperation. Compared to adding a bumper, controlling the deployment of airbags can buffer the impact during a collision, reducing acceleration and impact force, thus providing greater safety. Attached Figure Description

[0036] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0037] Figure 1 This is a flowchart of a multi-collision protection method provided in a specific embodiment of the present invention;

[0038] Figure 2 This is a flowchart of a method for controlling an in-vehicle device to perform collision avoidance protection for the driver, provided in an embodiment of the present invention.

[0039] Figure 3 This is a schematic diagram of a multi-collision protection system structure provided in a specific embodiment of the present invention;

[0040] Figure 4 This is an internal structural diagram of the electronic device provided by the present invention. Detailed Implementation

[0041] The embodiments of the technical solution of the present invention will now be described in detail with reference to the accompanying drawings. These embodiments are only used to more clearly illustrate the technical solution of the present invention and are therefore merely examples, and should not be construed as limiting the scope of protection of the present invention.

[0042] It should be noted that, unless otherwise stated, the technical or scientific terms used in this application should have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0043] To address the aforementioned issues, this invention provides a multi-layered anti-collision protection method and system, medium, and electronic device suitable for automotive anti-collision applications.

[0044] The embodiments of the present invention will now be described with reference to the accompanying drawings.

[0045] In one embodiment, such as Figure 1 As shown, a multi-layered anti-collision protection method is provided, the specific steps of which include:

[0046] S101 obtains real-time monitoring data from the acquired video stream; wherein, the monitoring data includes: the obstacle motion state contained in each frame of the video stream;

[0047] S102 defines the vehicle safety zone corresponding to the motion state of the obstacle and formulates multiple collision avoidance measures for different zones;

[0048] S103 matches corresponding collision avoidance measures to the current vehicle safety zone and controls the on-board equipment to protect the occupants from collisions.

[0049] In the above embodiment, step S101, obtaining real-time monitoring data from the acquired video stream, includes:

[0050] During vehicle startup, video stream data of the road ahead is collected in real time using video acquisition equipment; wherein, the video acquisition equipment includes: cameras installed at the front and rear of the vehicle respectively; the cameras include one or more of wide-angle cameras, high-definition zoom cameras and infrared cameras;

[0051] The video acquisition device is connected to the detection network and the YOLOx-S processor. The acquired video stream data is sent to the YOLOx-S processor to identify the obstacle information contained in each frame of the video stream. The motion state of the obstacle is obtained by calling the monocular ranging API interface multiple times.

[0052] Among them, the YOLOX-S processor is a system developed primarily based on the YOLOX-S software library;

[0053] The obstacle information includes the type of obstacle;

[0054] The motion state of the obstacle includes: the obstacle's speed, direction, and relative distance to the current vehicle body.

[0055] In the above embodiments, the real-time acquisition of video stream data of the road ahead by the video acquisition device includes: after receiving the vehicle start-up perception signal, turning on the cameras installed at the front and rear of the vehicle to enter the monitoring mode, performing video stream acquisition relative to the road ahead of the camera, and obtaining video stream data.

[0056] In the above embodiments, identifying obstacle information contained in each frame of the video stream includes: disassembling the frame images in the video stream data frame by frame according to a preset frequency, converting each frame image from color to grayscale image and then performing adaptive threshold segmentation to obtain a binary image; removing noise from the binary image, and using an improved YOLOx-S algorithm to detect obstacles in the image;

[0057] The image containing obstacles is matched with historical image data in a custom OpenCV image database. Based on the image matching results, it is determined whether the obstacle in the image belongs to a known obstacle type in the database.

[0058] In this embodiment, noise interference in the image can be removed using algorithms such as Candy or Sobel, so that objects in the image frame can be displayed.

[0059] In the above embodiments, obtaining the motion state of the obstacle by repeatedly calling the monocular ranging API interface includes:

[0060] The relative distance, speed, and direction of the obstacle to the current vehicle are obtained from each call to the monocular ranging API interface and are passed to the equalization function interface to output the distance and speed values ​​that are closest to the actual driving speed.

[0061] like Figure 2 As shown, based on the above embodiments, the following specific embodiments are provided:

[0062] High-definition sensors are installed at the front and rear of the vehicle, electromagnets and springs are installed on the brake and accelerator pedals, and airbags are embedded around the vehicle.

[0063] High-definition cameras at the front and rear of the vehicle capture image data in real time while the vehicle's infotainment system is running. The data is then transmitted to the vehicle's intelligent system, where the video stream is split into frames at 120 FPS (one frame is one image). These frames are then sent to the OpenCV preprocessing module, where the images are converted to grayscale. The CLAHE (Contrast Limiting Adaptive Histogram Equalization) algorithm and binarization are used to sharpen the images. Gaussian filtering and salt-and-pepper filtering algorithms are employed to remove noise interference. Candy and Sobel algorithms are used to expose objects in each frame, identifying obstacles. The data is then transmitted to a YOLOx-S processor, which repeatedly calls the monocular ranging and monocular velocity APIs. The APIs retrieve the speed of objects in the video, their relative distance to the vehicle, their direction, and their type. Each distance and speed is then fed into the equalization function, ultimately outputting a distance and speed that most closely approximates reality.

[0064] In one embodiment, detecting obstacles in an image using the improved YOLOv-S algorithm can be performed through the following steps, specifically:

[0065] Let F be the input image. Perform max pooling and average pooling operations on F, and feed the pooled feature information into a Multi-Layer Perceptron (MLP). After stacking the two output features, apply the sigmoid activation function and calculate the channel attention weight coefficient MC. The output feature is generated by multiplying MS with the feature map F' using the following formula:

[0066] M S (F′)=σ(f 7×7 (Concat[AvgPool(F′),MaxPool(F′)]))

[0067] In the formula, Concat represents the concatenation operation; σ represents the Sigmoid non-linear activation function; MaxPool and AvgPool represent max pooling and average pooling, respectively; F' represents the feature map generated by multiplying MC with the input image F; and f represents the convolution operation using a 7×7 kernel.

[0068] In one embodiment, the relative distance of the current vehicle body is obtained using the following vehicle ranging method:

[0069] The YOLOX-S model can output the confidence score, class, and vertices v of the detected object in each orientation of the bounding box. max u min u max The information is obtained by selecting the midpoint p(u,v) at the bottom of the vehicle detection frame as the feature point for distance measurement, and the calculation formula is as follows:

[0070]

[0071] In vehicle ranging, O is the center point of the camera target surface, OC is the optical center of the camera, the camera installation height is H, the focal length is f, the pitch angle is β, ∠aOCO is γ, the angle between point P and the projection of the optical axis is ε; w is the horizontal distance from point a to point p, w=(u-u0)dx; h is the vertical distance from point a to the image center point O, h=(v-v0)dy.

[0072] Let P be the distance measurement feature point of the vehicle being detected on the road surface, D be the longitudinal distance between the vehicle and the camera, K be the lateral distance AP, and L be the actual distance.

[0073]

[0074] Then the longitudinal distance D:

[0075]

[0076] Lateral distance K:

[0077]

[0078] In summary, this invention presents an improved target detection method and monocular ranging method based on the YOLOx-S principle. Compared to detection algorithms based on other principles, the improved YOLOx-S method demonstrates superior accuracy, recall, and mean precision in distance measurement. Furthermore, the monocular velocity measurement method can measure the speed of an object while the high-definition probe is moving, saving on the cost of previous radar and sensor hardware.

[0079] This invention prevents operation in cases where the driver does not operate or operates the system incorrectly, and provides protective measures against dangerous accidents, thus greatly improving safety.

[0080] In the above embodiment, step S102, which defines the vehicle safety zone corresponding to the motion state of the obstacle, includes:

[0081] The interval where the distance to an obstacle exceeds a preset threshold is defined as the safe zone; the interval where the distance to an obstacle is less than the maximum braking distance is defined as the sensitive zone; and the interval where the distance obtained by multiplying the vehicle's airbag deployment time by the obstacle's speed is defined as the danger zone.

[0082] As can be seen, the present invention uses a segmented approach to implement multiple measures to prevent collisions from occurring.

[0083] In the above embodiment, step S103 matches corresponding anti-collision measures for the current vehicle safety zone and controls the vehicle-mounted equipment to protect the driver from collision. The vehicle-mounted equipment includes voice equipment, ECU module, electromagnet, gravity control pedal, accelerator pedal, airbag, etc.

[0084] To match appropriate collision avoidance measures to the current vehicle safety range, the specific steps for controlling onboard equipment to provide collision avoidance protection for the driver include:

[0085] When the vehicle is in a safe zone, a command is sent to the in-vehicle voice device to trigger the in-vehicle voice device to provide obstacle information to the driver's side.

[0086] When the car is driving in a sensitive area, commands are sent to the in-vehicle voice device and the ECU module respectively, triggering the in-vehicle voice device to provide voice prompts to the driver's side about the obstacle;

[0087] The ECU module magnetizes the electromagnets inside the vehicle, controlling the gravity control pedal and accelerator pedal to adjust to a preset height to reduce vehicle speed.

[0088] When the car is in a danger zone, commands are sent to the ECU module of the in-vehicle voice device and the airbag, triggering the in-vehicle voice device to provide voice prompts to the driver's side about the obstacle and controlling the deployment of the airbags inside the vehicle.

[0089] Based on the above steps S102 and S103, the following embodiments are provided:

[0090] Based on the distance range, the system is divided into several zones. The area beyond M1 (greater than 100 meters) is called the safe zone, the area from M1 to M2 (maximum braking distance) is called the sensitive zone, and the area from M2 to M3 (distance between the time it takes for the airbag to deploy and the current speed) is called the danger zone.

[0091] When the car is in a safe zone, the system sends commands to the voice system module, informing the driver via voice of the information about the nearest objects to the vehicle. When the car is in a sensitive zone, the system sends commands to the voice system module and the ECU module controlling the electromagnet, instructing the voice system to play information about the nearest objects to the vehicle. The ECU module then magnetizes the electromagnet, using gravity to control the pedals. Based on the system commands, the ECU module pulls the brake pedal down and raises the accelerator pedal up accordingly, controlling the vehicle speed and, to some extent, preventing driver misoperation. When the car is in a danger zone, the system sends commands to the voice module, the ECU module, and the ECU module controlling the airbags. The processes are similar in both cases. Upon receiving the system commands, the airbag module quickly inflates and deploys, protecting the vehicle body, increasing the time before a collision, delaying the collision, and reducing acceleration. According to Newton's second law F=ma, for a given m (mass), the smaller a (acceleration), the smaller the force, and the less damage caused, thus protecting the safety of the car and occupants.

[0092] Based on the data obtained from the above, we analyze it to determine the safety level range in which the current minimum distance acquired by the vehicle falls, and then take different actions depending on whether the vehicle is currently in a safe state.

[0093] Based on the same inventive concept, this application also provides a multi-collision avoidance protection system for implementing the above-described multi-collision avoidance protection method. The solution provided by this system is similar to the solution described in the above-described embodiments. Therefore, the specific limitations of one or more multi-collision avoidance protection embodiments provided below can be found in the limitations of the multi-collision avoidance protection method described above, and will not be repeated here.

[0094] In one embodiment, the present invention also provides a multi-layered anti-collision protection system, such as... Figure 3 As shown, it includes: a monitoring module 110, a definition module 120, and a control module 130; wherein:

[0095] The monitoring module 110 is used to obtain real-time monitoring data from the acquired video stream; wherein, the monitoring data includes: the obstacle motion state contained in each frame of the video stream;

[0096] The definition module 120 is used to define the vehicle safety zone corresponding to the motion state of the obstacle and to formulate multiple anti-collision measures for different zones.

[0097] The control module 130 is used to match corresponding anti-collision measures to the current vehicle safety zone and control the on-board equipment to protect the occupants of the vehicle from collisions.

[0098] Meanwhile, this application also proposes a computer-readable storage medium and an electronic device, wherein the computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of any one of steps S101-S103 of the method.

[0099] In one embodiment, an electronic device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 4 As shown. The electronic device includes a processor, memory, communication interface, display screen, and input device connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements the method described in any one of steps S101 to S103. The display screen can be a liquid crystal display (LCD) or an e-ink display. The input device can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the device's casing, or an external keyboard, touchpad, or mouse.

[0100] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0101] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0102] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0103] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0104] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0105] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of the claims of the present invention pending approval.

Claims

1. A multi-layered anti-collision protection method, characterized in that, The method includes: Real-time monitoring data is obtained from the acquired video stream; wherein, the monitoring data includes: the motion state of the obstacle contained in each frame of the video stream; Define the vehicle safety zone corresponding to the motion state of the obstacle, and formulate multiple collision avoidance measures for different zones; Match appropriate collision avoidance measures to the current vehicle safety zone, and control the on-board equipment to protect the occupants from collisions. The relative distance to the vehicle is obtained using the following vehicle distance measurement method: The YOLOX-S model can output the confidence score, class, and vertices of the bounding box for the detected object. , The information is obtained by selecting the midpoint p(u,v) at the bottom of the vehicle detection frame as the feature point for distance measurement, and the calculation formula is as follows: / 2； In vehicle ranging, O is the center point of the camera target surface, OC is the optical center of the camera, the camera installation height is H, the focal length is f, the pitch angle is β, ∠aOCO is γ, and the angle between point P and the projection of the optical axis is ε; w is the horizontal distance from point a to point p, w=(u-u0)dx; h is the vertical distance from point a to the image center point O, h=(v-v0)dy; Let P be the distance measurement feature point of the vehicle being detected on the road surface, D be the longitudinal distance between the vehicle and the camera, K be the lateral distance AP, and L be the actual distance; ， Then the longitudinal distance D: ， Lateral distance K: ， ； High-definition cameras at the front and rear of the vehicle capture image data in real time while the vehicle's infotainment system is running. The data is then transmitted to the vehicle's intelligent system, where the video stream is split into frames at 120 FPS. The frames are then sent to the OpenCV preprocessing module to convert the images to grayscale and enhance their clarity. Gaussian and salt-and-pepper filters are used to remove noise interference from the images, and Candy and Sobel algorithms are used to expose objects in the frames, identifying obstacles. The data is then transmitted to a YOLOx-S processor, and the monocular ranging and monocular velocity APIs are called multiple times. The speed of objects in the video, their relative distance to the vehicle, their direction, and their type are obtained from the APIs. Each distance and speed is then fed into the equalization function interface, ultimately outputting a distance and speed that most closely approximates reality.

2. The method as described in claim 1, characterized in that, The real-time monitoring data obtained from the acquired video stream includes: During vehicle startup, video stream data of the road ahead is collected in real time using video acquisition equipment; wherein, the video acquisition equipment includes: cameras installed at the front and rear of the vehicle respectively; the cameras include one or more of wide-angle cameras, high-definition zoom cameras and infrared cameras; The video acquisition device is connected to the detection network and the YOLOx-S processor. The acquired video stream data is sent to the YOLOx-S processor to identify the obstacle information contained in each frame of the video stream. The motion state of the obstacle is obtained by calling the monocular ranging API interface multiple times. The obstacle information includes the type of obstacle; The motion state of the obstacle includes: the obstacle's speed, direction, and relative distance to the current vehicle body.

3. The method as described in claim 2, characterized in that, The real-time acquisition of video stream data of the road ahead via video acquisition equipment includes: after receiving the vehicle start-up sensing signal, turning on the cameras installed at the front and rear of the vehicle to enter the monitoring mode, performing video stream acquisition relative to the road ahead of the camera, and obtaining video stream data.

4. The method as described in claim 2, characterized in that, The method for identifying obstacle information contained in each frame of the video stream includes: disassembling the frame images in the video stream data frame by frame according to a preset frequency, converting each frame image from color to grayscale image and then performing adaptive threshold segmentation to obtain a binary image; removing noise from the binary image and using an improved YOLOx-S algorithm to detect obstacles in the image; The image containing obstacles is matched with historical image data in a custom OpenCV image database. Based on the image matching results, it is determined whether the obstacle in the image belongs to a known obstacle type in the database.

5. The method as described in claim 2, characterized in that, The process of obtaining the motion state of the obstacle by repeatedly calling the monocular ranging API interface includes: The relative distance, speed, and direction of the obstacle to the current vehicle are obtained from each call to the monocular ranging API interface and are passed to the equalization function interface to output the distance and speed values ​​that are closest to the actual driving speed.

6. The method as described in claim 5, characterized in that, The vehicle safety zone corresponding to the defined motion state of the obstacle includes: The interval where the distance to an obstacle exceeds a preset threshold is defined as the safe zone; the interval where the distance to an obstacle is less than the maximum braking distance is defined as the sensitive zone; and the interval where the distance obtained by multiplying the vehicle's airbag deployment time by the obstacle's speed is defined as the danger zone.

7. The method as described in claim 1, characterized in that, The method of matching appropriate anti-collision measures to the current vehicle safety zone and controlling the on-board equipment to provide anti-collision protection for the driver includes: When the vehicle is in a safe zone, a command is sent to the in-vehicle voice device to trigger the in-vehicle voice device to provide obstacle information to the driver's side. When the car is driving in a sensitive area, commands are sent to the in-vehicle voice device and the ECU module respectively, triggering the in-vehicle voice device to provide voice prompts to the driver's side about the obstacle; The ECU module magnetizes the electromagnets inside the vehicle, controlling the gravity control pedal and accelerator pedal to adjust to a preset height to reduce vehicle speed. When the car is in a danger zone, commands are sent to the ECU module of the in-vehicle voice device and the airbag, triggering the in-vehicle voice device to provide voice prompts to the driver's side about the obstacle and controlling the deployment of the airbags inside the vehicle.

8. A multi-layered anti-collision protection system, characterized in that, The system includes: The monitoring module is used to obtain real-time monitoring data from the acquired video stream; wherein, the monitoring data includes: the obstacle motion status contained in each frame of the video stream; The definition module is used to define the vehicle safety zone corresponding to the motion state of the obstacle and to formulate multiple collision avoidance measures for different zones; The control module is used to match appropriate anti-collision measures to the current vehicle safety zone and control the on-board equipment to protect the occupants from collisions. The relative distance to the vehicle is obtained using the following vehicle distance measurement method: The YOLOX-S model can output the confidence score, class, and vertices of the bounding box for the detected object. , The information is obtained by selecting the midpoint p(u,v) at the bottom of the vehicle detection frame as the feature point for distance measurement, and the calculation formula is as follows: / 2； In vehicle ranging, O is the center point of the camera target surface, OC is the optical center of the camera, the camera installation height is H, the focal length is f, the pitch angle is β, ∠aOCO is γ, and the angle between point P and the projection of the optical axis is ε; w is the horizontal distance from point a to point p, w=(u-u0)dx; h is the vertical distance from point a to the image center point O, h=(v-v0)dy; Let P be the distance measurement feature point of the vehicle being detected on the road surface, D be the longitudinal distance between the vehicle and the camera, K be the lateral distance AP, and L be the actual distance; ， Then the longitudinal distance D: ， Lateral distance K: ， ； High-definition cameras at the front and rear of the vehicle capture image data in real time while the vehicle's infotainment system is running. The data is then transmitted to the vehicle's intelligent system, where the video stream is split into frames at 120 FPS. The frames are then sent to the OpenCV preprocessing module to convert the images to grayscale and enhance their clarity. Gaussian and salt-and-pepper filters are used to remove noise interference from the images, and Candy and Sobel algorithms are used to expose objects in the frames, identifying obstacles. The data is then transmitted to a YOLOx-S processor, and the monocular ranging and monocular velocity APIs are called multiple times. The speed of objects in the video, their relative distance to the vehicle, their direction, and their type are obtained from the APIs. Each distance and speed is then fed into the equalization function interface, ultimately outputting a distance and speed that most closely approximates reality.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method described in any one of claims 1-7.

10. An electronic device, comprising a memory and a processor, characterized in that, The memory stores a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 7.

Images