Vehicle behavior warning method and system, electronic device and storage medium

Abstract

The application provides a vehicle behavior early warning method and system, electronic equipment and storage medium, and relates to the field of vehicle behavior monitoring. The method uses a roadside device to collect road data in real time, wherein the road data at least includes vehicle data, pedestrian data and road condition data. Then, driving behavior intention data sent by a vehicle is received, and the driving behavior intention data is used to determine the serialized data of the road data. The behavior of the vehicle is predicted using the serialized data, and the risk evaluation data of the vehicle is determined according to the prediction result. Finally, the risk evaluation data is sent to the vehicle to control the vehicle to perform behavior early warning according to the risk evaluation data. The method converts image data into serialized data through mapping transformation and gridding, reduces the data transmission bandwidth, improves the transmission efficiency, reduces the data transmission cost, improves the overall operation efficiency, and solves the problems of low precision and poor effect in the vehicle behavior early warning process.

Description

Technical Field

[0001] This invention relates to the field of vehicle behavior monitoring, and in particular to a vehicle behavior early warning method, system, electronic device, and storage medium. Background Technology

[0002] Autonomous driving is currently a hot research area, and with the significant advancements in deep learning and vehicle-to-everything (V2X) technologies, it has made rapid progress. However, achieving fully autonomous driving and mature V2X-based autonomous driving still faces many practical challenges. Due to limitations in autonomous driving algorithms, perception, and detection capabilities, the complexity and diversity of road conditions can easily lead to traffic accidents during autonomous driving, resulting in safety hazards.

[0003] Therefore, vehicle behavior warning during vehicle cooperative driving is an effective method to reduce safety hazards. Vehicle behaviors such as reversing lane changes, overtaking, and cooperative lane changes can be simulated and reasoned through algorithms to obtain predicted values ​​of vehicle behavior risk, and timely warnings can be issued for dangerous behaviors.

[0004] However, existing technologies for vehicle behavior warning still suffer from low accuracy and poor effectiveness. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a vehicle behavior warning method, system, electronic device, and storage medium. This method converts image data into standardized non-image data, and transforms image data into serialized data through mapping transformation and meshing, thereby reducing data transmission bandwidth, improving transmission efficiency, reducing data transmission costs, and improving overall operating efficiency. At the same time, this method performs risk prediction analysis of vehicle behavior for vehicle-road cooperation and sends the prediction results to the vehicle end for warning, which can enhance the automation and intelligence of cooperative driving, reduce the occurrence of traffic accidents, and thus alleviate the problems of low accuracy and poor effect in the vehicle behavior warning process.

[0006] In a first aspect, embodiments of the present invention provide a vehicle behavior warning method, the method comprising:

[0007] Real-time acquisition of road data collected by roadside equipment; the road data includes at least: vehicle data, pedestrian data, and road condition data;

[0008] Receive driving behavior intention data from vehicles and use the driving behavior intention data to determine the serialized data of road data;

[0009] Serialized data is used to predict vehicle behavior, and the risk assessment data of the vehicle is determined based on the prediction results.

[0010] The risk assessment data is sent to the vehicle, and the vehicle is controlled to issue behavioral warnings based on the risk assessment data.

[0011] In one embodiment, the step of receiving driving behavior intention data from a vehicle and using the driving behavior intention data to determine serialized data of road data includes:

[0012] The roadside control equipment uses the homography matrix to transform the target category and location information contained in the road data to obtain the mapping result from the pixel coordinate system to the standard area two-dimensional target;

[0013] The driving behavior intent data contained in the road data is obtained, and the mapping results are used to encode the targets in the road area of ​​the road data to obtain the encoding results;

[0014] The serialized data of the road data is determined by overlaying the encoding results with the intent data.

[0015] In one implementation, the step of controlling the roadside equipment to perform position transformation on the target category and location information contained in the road data through a homography matrix to obtain a mapping result from the pixel coordinate system to a standard area two-dimensional target includes:

[0016] Segment the road data into regions to obtain the road region data contained within the road data;

[0017] Determine the road dimensions in the road area data, and use the preset road template dimensions to cut the road area data to obtain the cutting result of the road area data;

[0018] The location of the road area data segmentation results is transformed using the homography matrix, converting the location information into the center point coordinates in the segmentation results;

[0019] The mapping result from the pixel coordinate system to the standard area two-dimensional target is determined based on the center point coordinates.

[0020] In one implementation, the step of segmenting road data into regions to obtain road region data contained in the road data includes:

[0021] Determine whether the road data collection scenario is a fixed scenario;

[0022] If so, then determine the corner points of the standard area based on the preset corner points in the road data, and determine the road area data contained in the road data based on the corner points;

[0023] If not, the road area data contained in the road data is determined based on the pre-installed fixed markers in the road data.

[0024] In one implementation, the steps of predicting vehicle behavior using serialized data and determining vehicle hazard assessment data based on the behavior prediction results include:

[0025] Upload the serialized data to the preset cloud server;

[0026] After receiving the serialized data, the cloud server performs intent reasoning calculations on the serialized data to output the vehicle's driving behavior, and uses the vehicle's driving behavior to determine the behavior prediction result.

[0027] The risk level assessment data of the vehicle is determined based on the behavioral prediction results.

[0028] In one implementation, after receiving serialized data, the cloud server performs intent reasoning calculations on the serialized data to output the vehicle's driving behavior, and uses the vehicle's driving behavior to determine the behavior prediction result, including the following steps:

[0029] The cloud server will receive serialized data and cut it into M*N regions according to preset rules, and then use preset encoding parameters to concatenate the regions to obtain an input matrix with M*N+1 sub-elements.

[0030] The input matrix is ​​fed into the trained Transformer model, which then controls the Transformer model to obtain the vehicle's driving behavior based on the input matrix; the driving behavior includes at least: straight driving, lane changing, red light warning, and collision warning.

[0031] The Transformer model is controlled to determine the degree of danger corresponding to driving behavior, and the behavior prediction result is determined based on the degree of danger.

[0032] In one implementation, hazard assessment data is sent to the vehicle, and the vehicle is controlled to perform behavioral warning steps based on the hazard assessment data, including:

[0033] Control roadside equipment to receive hazard assessment data from a cloud server;

[0034] The risk level assessment data is sent to the vehicle, and the vehicle is controlled to use the risk level assessment data to complete the behavior warning process.

[0035] Secondly, embodiments of the present invention provide a vehicle behavior warning system, the system comprising:

[0036] The roadside data acquisition module is used to acquire road data collected by roadside equipment in real time; the road data includes at least: vehicle data, pedestrian data and road condition data.

[0037] The vehicle location coding module is used to receive driving behavior intention data sent by the vehicle and use the driving behavior intention data to determine the serialized data of the road data.

[0038] The cloud-based algorithm processing module is used to predict vehicle behavior using serialized data and determine vehicle risk assessment data based on the prediction results.

[0039] The vehicle-mounted data transceiver module is used to send hazard assessment data to the vehicle and control the vehicle to issue behavioral warnings based on the hazard assessment data.

[0040] Thirdly, embodiments of the present invention also provide an electronic device, including a processor and a memory, wherein the memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the steps of the vehicle behavior warning method provided in the first aspect.

[0041] Fourthly, embodiments of the present invention also provide a storage medium storing computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement the steps of the vehicle behavior warning method provided in the first aspect.

[0042] This invention provides a vehicle behavior warning method, system, electronic device, and storage medium. The method utilizes roadside equipment to collect road data in real time, wherein the road data includes at least vehicle data, pedestrian data, and road condition data. It then receives driving intention data from vehicles and uses this data to determine serialized road data. Next, it uses the serialized data to predict vehicle behavior and determines vehicle hazard assessment data based on the prediction results. Finally, it sends the hazard assessment data to the vehicle, controlling the vehicle to provide behavioral warnings according to the hazard assessment data. This method converts image data into standardized non-image data through mapping transformation and meshing.

[0043] Converting image data into serialized data reduces data transmission bandwidth, improves transmission efficiency, lowers data transmission costs, and enhances overall operational efficiency. Simultaneously, this method is geared towards vehicle-road cooperation, enabling vehicle movement...

[0044] To conduct hazard prediction analysis and send the prediction results to the vehicle for early warning, the automation and intelligence of collaborative driving can be improved, traffic accidents can be reduced, and the problems of low accuracy and poor effect in the early warning process of vehicle behavior can be alleviated.

[0045] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. Objects and other advantages of the invention

[0046] The points are implemented and obtained by means of the structures specifically pointed out in the specification, claims and drawings.

[0047] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0048] 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. Obviously, the accompanying drawings described below are some embodiments of the present invention, and are for those with ordinary knowledge in the art.

[0049] For technicians, without expending any creative effort, they can obtain other drawings based on these drawings.

[0050] Figure 1 A flowchart of a vehicle behavior warning method provided in an embodiment of the present invention;

[0051] Figure 2 A flowchart of step S102 of a vehicle behavior warning method provided in an embodiment of the present invention;

[0052] Figure 3 A flowchart of step S201 of a vehicle behavior warning method provided in an embodiment of the present invention is shown in Figure 5.

[0053] Figure 4 A flowchart of step S301 of a vehicle behavior warning method provided in an embodiment of the present invention;

[0054] Figure 5 A flowchart of step S103 of a vehicle behavior warning method provided in an embodiment of the present invention;

[0055] Figure 6 A flowchart of step S502 of a vehicle behavior warning method provided in an embodiment of the present invention;

[0056] Figure 7 A flowchart of step S104 of a vehicle behavior warning method provided in an embodiment of the present invention;

[0057] Figure 8 This is a schematic diagram of the data flow in a vehicle behavior warning method provided in an embodiment of the present invention;

[0058] Figure 9 This is a schematic diagram of the coding process in a vehicle behavior warning method provided in an embodiment of the present invention;

[0059] Figure 10This is a flowchart of a vehicle behavior warning method according to an embodiment of the present invention, showing the predicted value of behavioral hazards.

[0060] Figure 11 This is a schematic diagram of the structure of a vehicle behavior warning system provided in an embodiment of the present invention;

[0061] Figure 12 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0062] icon:

[0063] 1110 - Roadside data acquisition module; 1120 - Vehicle location coding module; 1130 - Cloud algorithm processing module; 1140 - Vehicle data transceiver module;

[0064] 101 - Processor; 102 - Memory; 103 - Bus; 104 - Communication interface. Detailed Implementation

[0065] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0066] Autonomous driving is currently a hot research area, and with the significant advancements in deep learning and vehicle-to-everything (V2X) technologies, it has made rapid progress. However, achieving fully autonomous driving and mature V2X-based autonomous driving still faces many practical challenges. Due to limitations in autonomous driving algorithms, perception, and detection capabilities, the complexity and diversity of road conditions can easily lead to traffic accidents during autonomous driving, resulting in safety hazards.

[0067] Therefore, vehicle behavior warning during vehicle cooperative driving is an effective method to reduce safety hazards in vehicle-road cooperative driving. Vehicle behaviors such as reverse lane changing, overtaking, and cooperative lane changing can be simulated and reasoned through algorithms to obtain predicted values ​​of vehicle behavior risk, and timely warnings can be issued for dangerous behaviors. However, existing technologies for vehicle behavior warning still suffer from low accuracy and poor effectiveness. Based on this, this invention provides a vehicle behavior warning method, system, electronic device, and storage medium. This method converts image data into standardized non-image data, and transforms image data into serialized data through mapping transformation and gridding, reducing data transmission bandwidth, improving transmission efficiency, reducing data transmission costs, and improving overall operational efficiency. At the same time, this method performs risk prediction analysis of vehicle behavior for vehicle-road cooperation and sends the prediction results to the vehicle end for warning, which can enhance the automation and intelligence of cooperative driving, reduce the occurrence of traffic accidents, and thus alleviate the problems of low accuracy and poor effectiveness in vehicle behavior warning.

[0068] To facilitate understanding of this embodiment, a vehicle behavior warning method disclosed in this invention will first be described in detail, such as... Figure 1 As shown, the method includes:

[0069] Step S101: Acquire road data collected by roadside equipment in real time; wherein, the road data includes at least: vehicle data, pedestrian data and road condition data.

[0070] Roadside units (RSUs) mainly consist of equipment such as cameras or lidar deployed on roads. Roadside units can acquire road data in real time and combine them with image recognition algorithms to acquire vehicle data, pedestrian data, and road condition data, thereby obtaining target category and location data of road data in the road area.

[0071] Step S102: Receive driving behavior intention data sent by the vehicle, and use the driving behavior intention data to determine the serialized data of the road data.

[0072] Driving intent data is collected by the On-Board Unit (OBU), including intentions such as lane changing and overtaking in the wrong direction. This data is uploaded through data exchange with the Roadside Unit (RSU). The RSU receives broadcast data from the OBU, such as vehicle ID, vehicle location, vehicle speed, and driving intent. The RSU then overlays the driving intent data uploaded by the OBU with the OBU's data to establish a correlation between driving intent and changes in vehicle location.

[0073] Step S103: Predict the behavior of the vehicle using serialized data, and determine the vehicle's risk assessment data based on the prediction results.

[0074] This process utilizes standardized sequence data to perform relevant algorithmic inference, outputting a prediction of the vehicle's intended hazard level, and then obtaining hazard assessment data based on the prediction results. This can be achieved using a relevant artificial intelligence model, which is pre-trained and directly used to obtain prediction results by inputting sequenced data, and then using these prediction results to determine the hazard assessment data.

[0075] Step S104: Send the hazard assessment data to the vehicle and control the vehicle to issue behavioral warnings based on the hazard assessment data.

[0076] After the hazard assessment data is acquired, it needs to be sent to the vehicle. The hazard assessment data is sent to the vehicle through the roadside unit (RSU), thereby controlling the vehicle to implement a behavior warning process based on the hazard assessment data.

[0077] In one implementation, step S102 involves receiving driving intention data from a vehicle and using that driving intention data to determine the serialized data of road data, as follows: Figure 2 As shown, it includes:

[0078] Step S201: Control the roadside equipment to perform position transformation on the target category and location information contained in the road data through homography matrix to obtain the mapping result from pixel coordinate system to standard area two-dimensional target.

[0079] Step S202: Obtain driving behavior intention data contained in the road data, and use the mapping results to encode the targets in the road area of ​​the road data to obtain the encoding results.

[0080] Step S203: Determine the serialized data of the road data based on the superposition result of the encoding result and the intent data.

[0081] The Roadside Unit (RSU) transforms the acquired target category and location information using a homography matrix, mapping it from a pixel coordinate system to a standard two-dimensional target area. This mapping enables standardized encoding of targets within the road area, which in turn overlays driving intent information, establishing a link between driving intent and vehicle position changes.

[0082] In one implementation, the roadside equipment controls the target category and location information contained in the road data to perform position transformation using a homography matrix, obtaining a mapping result from the pixel coordinate system to a standard area two-dimensional target. (Step S201 is as follows...) Figure 3 As shown, it includes:

[0083] Step S301: Perform region segmentation on the road data to obtain the road region data contained in the road data;

[0084] Step S302: Determine the road dimensions in the road area data, and cut the road area data using the preset road template dimensions to obtain the cutting result of the road area data;

[0085] Step S303: Use the homography matrix to perform position transformation on the road area data cutting results, converting the position information into the center point coordinates in the cutting results;

[0086] Step S304: Determine the mapping result from the pixel coordinate system to the standard area two-dimensional target based on the center point coordinates.

[0087] In the process of determining the serialized road data, the road region is first segmented to separate the road from the surrounding area. The road region has a length of h and a width of w. A standardized transformation template is selected based on the length-width ratio, and the standard region is segmented. The preset road template has a length of h1 and a width of w1, with a long side cutting ratio of h1 / M and a wide side cutting ratio of w1 / N, thus forming an M*N grid. Homography matrix transformation is performed in conjunction with camera intrinsic parameters to map the image region to the standard template image. The vehicle position information is replaced by the center point (x,y), and coded coordinates are formed based on the transformed vehicle position coordinates. In the M*N grid, each small square has a fixed code x11~xnn, responsible for encoding the vehicle points falling into this region.

[0088] Standardized data mapping primarily targets straight roads and can be categorized into two scenarios: fixed cameras and non-fixed cameras. In one implementation, step S301 involves segmenting the road data into regions to obtain road region data contained within the road data, as follows: Figure 4 As shown, it includes:

[0089] Step S401: Determine whether the road data collection scenario is a fixed scenario.

[0090] Step S402: If yes, then determine the corner points of the standard area based on the preset corner points in the road data, and determine the road area data contained in the road data based on the corner points.

[0091] The camera is positioned in a fixed scene, and the image area captured by the fixed camera remains constant. Four points representing the road location in the fixed image are mapped to the four corner points of the standard region. A homography matrix mapping is then performed from the image region to the standard region using an intrinsic parameter matrix.

[0092] Step S403: If not, determine the road area data contained in the road data based on the pre-installed fixed markers in the road data.

[0093] The camera is not in a fixed scene. Four fixed markers are installed beside the road. These four fixed markers are acquired in real time through an image detection algorithm. Then, the homography matrix mapping from the image region to the standard region is completed by combining the intrinsic parameter matrix.

[0094] In the process of predicting vehicle behavior using serialized data, a cloud server can be used. Specifically, in one implementation, step S103 involves predicting vehicle behavior using serialized data and determining the vehicle's risk assessment data based on the behavior prediction results, as follows: Figure 5 As shown, it includes:

[0095] Step S501: Upload the serialized data to the preset cloud server.

[0096] The cloud server contains a relevant cloud algorithm library and is connected to the roadside equipment.

[0097] In step S502, after receiving the serialized data, the cloud server performs intent reasoning calculation on the serialized data to output the vehicle's driving behavior, and uses the vehicle's driving behavior to determine the behavior prediction result.

[0098] The cloud-based algorithm library in the cloud server receives location-encoded serialized data, performs vehicle behavior prediction through relevant behavior prediction algorithms, and calculates and outputs the vehicle's driving behavior through intent reasoning, thereby determining the behavior prediction result.

[0099] Step S503: Determine the vehicle's risk level assessment data based on the behavior prediction results.

[0100] Vehicle behavior prediction can be performed using a Transformer-based NLP algorithm. In one implementation, after receiving serialized data, the cloud server performs intent reasoning calculations on the serialized data to output the vehicle's driving behavior, and uses the vehicle's driving behavior to determine the behavior prediction result in step S502. Figure 6 As shown, it includes:

[0101] In step S601, the cloud server will cut the received serialized data into M*N regions according to preset rules, and use preset encoding parameters to concatenate the regions to obtain an input matrix with M*N+1 sub-elements.

[0102] Step S602: Input the input matrix into the trained Transformer model and control the Transformer model to obtain the vehicle's driving behavior based on the input matrix.

[0103] Step S603: Control the Transformer model to determine the degree of danger corresponding to the driving behavior, and determine the behavior prediction result based on the degree of danger.

[0104] The standard region is divided into fixed-size M*N squares after graph region segmentation. Each square has a fixed encoding parameter. The M*N squares are concatenated to form an M*N+1 input matrix, which, together with the one-dimensional vehicle intent encoding, serves as the input to the Transformer model. The output is the predicted hazard value for each vehicle behavior, such as lane change, lane switching, red light warning, and collision warning. The Transformer model outputs the hazard level for each behavior; if a hazard is detected, it returns a warning to the vehicle.

[0105] In one implementation, hazard assessment data is sent to the vehicle, and the vehicle is controlled to perform a behavioral warning step S104 based on the hazard assessment data, such as... Figure 7 As shown, it includes:

[0106] Step S701: Control the roadside equipment to receive hazard assessment data from the cloud server.

[0107] Step S702: The hazard assessment data is sent to the vehicle, and the vehicle is controlled to use the hazard assessment data to complete the behavior warning process.

[0108] The cloud server receives location-encoded serialized data and uses a behavior prediction algorithm to predict vehicle behavior. The vehicle behavior prediction uses a Transformer-based NLP algorithm to predict the vehicle's intended hazard level, which is then sent back to the RSU and OBU devices to complete the vehicle behavior warning process.

[0109] like Figure 8 As shown in the data flow diagram of the vehicle behavior warning method, the roadside unit (RSU) acts as the central device, receiving vehicle-coded data from cameras, radar, and integrated radar-visual systems. The on-board unit (OBU) acquires the vehicle's behavioral intent and uploads it to the roadside RSU. The roadside RSU integrates the vehicle intent data with standardized sequence data to form new serialized data. This integrated serialized data is then uploaded to a cloud-based algorithm library for algorithm inference. The algorithm inference results are downloaded to the roadside RSU and then sent to the on-board OBU, completing the feedback of the vehicle behavior hazard prediction results and thus enabling warnings for unsafe driving behaviors.

[0110] like Figure 9As shown in the flowchart of the vehicle behavior warning method, the vehicle location encoding process includes two parts: first, the driving intent information reported by the vehicle; and second, the mapping, transformation, and serialization of data from the roadside unit (RSU). The RSU transforms the acquired target category and location information using a homography matrix, realizing the mapping from the pixel coordinate system to a standard two-dimensional target area. This mapping enables standardized encoding of targets within the road area, which is then superimposed with driving intent information to establish a link between driving intent and vehicle location changes. Therefore, this method, based on the vehicle-road cooperative system, divides the road area into a spatial grid, converting road image coordinate information into standard serialized grid encoding, avoiding image data transmission and significantly improving system operating efficiency.

[0111] like Figure 10 As shown in the flowchart of the behavior hazard prediction value in the vehicle behavior warning method, image data is coded into a grid to form standardized serialized information with a fixed size, reducing the alignment process of the Transformer model input. The encoded data can be used for behavior prediction using NLP algorithms, avoiding direct image behavior prediction and reducing algorithm complexity and resource consumption.

[0112] As can be seen from the vehicle behavior warning method mentioned in the above embodiments, this method transforms image data into standardized non-image data, and transforms image data into serialized data through mapping transformation and meshing, thereby reducing data transmission bandwidth, improving transmission efficiency, reducing data transmission costs, and improving overall operating efficiency. At the same time, this method performs risk prediction analysis on vehicle behavior for vehicle-road cooperation and sends the prediction results to the vehicle end for warning, which can enhance the automation and intelligence of cooperative driving, reduce the occurrence of traffic accidents, and thus alleviate the problems of low accuracy and poor effect in the vehicle behavior warning process.

[0113] Regarding the vehicle behavior warning method provided in the foregoing embodiments, this invention provides a vehicle behavior warning system, such as... Figure 11 As shown, the system includes:

[0114] The roadside data acquisition module 1110 is used to acquire road data collected by roadside equipment in real time; the road data includes at least: vehicle data, pedestrian data and road condition data.

[0115] The vehicle location coding module 1120 is used to receive driving behavior intention data sent by the vehicle and use the driving behavior intention data to determine the serialized data of the road data.

[0116] The cloud-based algorithm processing module 1130 is used to predict the behavior of vehicles using serialized data and determine the risk assessment data of vehicles based on the prediction results.

[0117] The vehicle data transceiver module 1140 is used to send hazard assessment data to the vehicle and control the vehicle to issue behavioral warnings based on the hazard assessment data.

[0118] In one embodiment, the vehicle location encoding module 1120 is further configured to: control the roadside equipment to perform location transformation on the target category and location information contained in the road data through a homography matrix to obtain a mapping result from the pixel coordinate system to the two-dimensional target in the standard area; acquire driving behavior intention data contained in the road data, and use the mapping result to encode the target in the road area in the road data to obtain an encoding result; and determine the serialized data of the road data based on the superposition result of the encoding result and the intention data.

[0119] In one embodiment, the vehicle location encoding module 1120, while controlling the roadside equipment to perform position transformation on the target category and location information contained in the road data through a homography matrix to obtain the mapping result from the pixel coordinate system to the standard area two-dimensional target, is also used for: segmenting the road data to obtain road area data contained in the road data; determining the road size in the road area data and cutting the road area data using a preset road template size to obtain the cutting result of the road area data; performing position transformation on the cutting result of the road area data using a homography matrix to convert the location information into the center point coordinates in the cutting result; and determining the mapping result from the pixel coordinate system to the standard area two-dimensional target based on the center point coordinates.

[0120] In one embodiment, the vehicle location coding module 1120, during the process of segmenting road data into regions and obtaining road area data contained in the road data, is further configured to: determine whether the road data collection scenario is a fixed scenario; if so, determine the corner points of the standard area based on the preset corner points in the road data, and determine the road area data contained in the road data based on the corner points; if not, determine the road area data contained in the road data based on the pre-installed fixed marker points in the road data.

[0121] In one embodiment, the cloud algorithm processing module 1130 is further configured to: upload serialized data to a preset cloud server; after receiving the serialized data, the cloud server performs intent reasoning calculation on the serialized data to output the driving behavior of the vehicle, and uses the driving behavior of the vehicle to determine the behavior prediction result; and determine the vehicle's danger level evaluation data based on the behavior prediction result.

[0122] In one implementation, the cloud algorithm processing module 1130, after receiving serialized data from the cloud server, performs intent reasoning calculation on the serialized data to output the vehicle's driving behavior, and uses the vehicle's driving behavior to determine the behavior prediction result, is further configured to: the cloud server cuts the received serialized data into M*N regions according to preset rules, and concatenates the regions using preset encoding parameters to obtain an input matrix of M*N+1 sub-elements; input the input matrix into a trained Transformer model, and control the Transformer model to obtain the vehicle's driving behavior based on the input matrix; wherein the driving behavior includes at least: straight driving, lane changing, red light warning, and collision warning; control the Transformer model to determine the danger level value corresponding to the driving behavior, and determine the behavior prediction result based on the danger level value.

[0123] In one embodiment, the vehicle-mounted data transceiver module 1140 is further configured to: control the roadside equipment to receive hazard assessment data from a cloud server; send the hazard assessment data to the vehicle; and control the vehicle to use the hazard assessment data to complete a behavior warning process.

[0124] The vehicle behavior warning system provided in this invention converts image data into standardized non-image data, and transforms image data into serialized data through mapping transformation and meshing. This reduces data transmission bandwidth, improves transmission efficiency, reduces data transmission costs, and improves overall operational efficiency. Simultaneously, this method performs hazard prediction analysis of vehicle behavior for vehicle-road cooperation and sends the prediction results to the vehicle for warning. This can enhance the automation and intelligence of cooperative driving, reduce traffic accidents, and thus alleviate the problems of low accuracy and poor effectiveness in vehicle behavior warning processes.

[0125] The vehicle behavior warning system provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned vehicle behavior warning method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0126] This embodiment also provides an electronic device, the structural schematic diagram of which is shown below. Figure 12 As shown, the device includes a processor 101 and a memory 102; wherein, the memory 102 is used to store one or more computer instructions, which are executed by the processor to implement the above-mentioned vehicle behavior warning method.

[0127] Figure 12 The electronic device shown also includes a bus 103 and a communication interface 104, with the processor 101, communication interface 104 and memory 102 connected via the bus 103.

[0128] The memory 102 may include high-speed random access memory (RAM), and may also include non-volatile memory, such as at least one disk storage device. The bus 103 may be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 12 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0129] The communication interface 104 is used to connect to at least one user terminal and other network units through a network interface, and to send encapsulated IPv4 packets or IPv4 packets to the user terminal through the network interface.

[0130] Processor 101 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of processor 101 or by instructions in software form. The processor 101 can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this disclosure. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this disclosure can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory 102. The processor 101 reads the information in memory 102 and, in conjunction with its hardware, completes the steps of the method described in the foregoing embodiments.

[0131] This invention also provides a storage medium storing a computer program, which, when executed by a processor, performs the steps of the methods described in the foregoing embodiments.

[0132] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0133] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0134] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0135] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0136] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A vehicle behavior early warning method, characterized in that, The method includes: Real-time acquisition of road data collected by roadside equipment; wherein the road data includes at least: vehicle data, pedestrian data and road condition data; Receive driving behavior intention data sent by the vehicle, and use the driving behavior intention data to determine the serialized data of the road data; The behavior of the vehicle is predicted using the serialized data, and the risk assessment data of the vehicle is determined based on the prediction results. The hazard assessment data is sent to the vehicle, and the vehicle is controlled to issue behavioral warnings based on the hazard assessment data. The step of receiving driving behavior intention data from a vehicle and using the driving behavior intention data to determine the serialized data of the road data includes: The roadside equipment is controlled to perform position transformation on the target category and location information contained in the road data through a homography matrix to obtain a mapping result from the pixel coordinate system to the standard area two-dimensional target; The driving behavior intent data contained in the road data is obtained, and the target within the road area in the road data is encoded using the mapping result to obtain the encoding result; The serialized data of the road data is determined based on the superposition result of the encoding result and the intent data; The step of controlling the roadside device to perform position transformation on the target category and location information contained in the road data through a homography matrix to obtain a mapping result from the pixel coordinate system to a standard area two-dimensional target includes: The road data is segmented into regions to obtain the road region data contained in the road data; Determine the road dimensions in the road area data, and cut the road area data using a preset road template size to obtain the cutting result of the road area data; The location information of the road area data is converted into the center point coordinates in the cutting result by using the homography matrix. The mapping result from the pixel coordinate system to the standard area two-dimensional target is determined based on the center point coordinates; The step of performing region segmentation on the road data to obtain road region data contained in the road data includes: Determine whether the road data collection scenario is a fixed scenario; If so, then the corner points of the standard area are determined based on the preset corner points in the road data, and the road area data contained in the road data are determined based on the corner points; If not, then the road area data contained in the road data is determined based on the pre-installed fixed markers in the road data.

2. The vehicle behavior warning method according to claim 1, characterized in that, The step of predicting the vehicle's behavior using the serialized data and determining the vehicle's risk assessment data based on the behavior prediction results includes: The serialized data is uploaded to a preset cloud server; After receiving the serialized data, the cloud server performs intent reasoning calculation on the serialized data to output the driving behavior of the vehicle, and uses the driving behavior of the vehicle to determine the behavior prediction result. The risk level assessment data of the vehicle is determined based on the behavior prediction results.

3. The vehicle behavior warning method according to claim 2, characterized in that, The steps of the cloud server receiving the serialized data, performing intent reasoning calculations on the serialized data to output the driving behavior of the vehicle, and using the driving behavior of the vehicle to determine the behavior prediction result include: The cloud server receives the serialized data and cuts it into M*N regions according to preset rules, and then uses preset encoding parameters to concatenate the regions to obtain an input matrix with M*N+1 sub-elements. The input matrix is ​​input into the trained Transformer model, which is then controlled to obtain the driving behavior of the vehicle based on the input matrix; wherein the driving behavior includes at least: straight driving, lane changing, red light warning, and collision warning; The Transformer model is controlled to determine the degree of danger corresponding to the driving behavior, and the behavior prediction result is determined based on the degree of danger.

4. The vehicle behavior warning method according to claim 2, characterized in that, The step of sending the hazard assessment data to the vehicle and controlling the vehicle to provide behavioral warnings based on the hazard assessment data includes: The roadside equipment is controlled to receive the hazard assessment data from the cloud server. The danger level assessment data is sent to the vehicle, and the vehicle is controlled to use the danger level assessment data to complete the behavior warning process.

5. A vehicle behavior warning system, characterized in that, The system includes: The roadside data acquisition module is used to acquire road data collected by roadside equipment in real time; wherein, the road data includes at least: vehicle data, pedestrian data and road condition data; The vehicle location encoding module is used to receive driving behavior intention data sent by the vehicle and use the driving behavior intention data to determine the serialized data of the road data. The cloud-based algorithm processing module is used to predict the behavior of the vehicle using the serialized data, and to determine the risk assessment data of the vehicle based on the prediction results. The vehicle-mounted data transceiver module is used to send the hazard assessment data to the vehicle and control the vehicle to issue behavioral warnings based on the hazard assessment data. The vehicle location encoding module is further configured to: control the roadside equipment to perform location transformation on the target category and location information contained in the road data through a homography matrix to obtain a mapping result from the pixel coordinate system to a standard area two-dimensional target; acquire driving behavior intention data contained in the road data, and use the mapping result to encode the target within the road area in the road data to obtain an encoding result; and determine the serialized data of the road data based on the superposition result of the encoding result and the intention data. The vehicle location encoding module, in the process of controlling the roadside equipment to perform position transformation on the target category and location information contained in the road data through a homography matrix to obtain a mapping result from the pixel coordinate system to a standard area two-dimensional target, is further configured to: segment the road data to obtain road area data contained in the road data; determine the road size in the road area data, and cut the road area data using a preset road template size to obtain the cutting result of the road area data; perform position transformation on the cutting result of the road area data using the homography matrix to convert the location information into the center point coordinates in the cutting result; and determine the mapping result from the pixel coordinate system to the standard area two-dimensional target based on the center point coordinates. In the process of segmenting the road data and obtaining the road area data contained in the road data, the vehicle location encoding module is also used to: determine whether the road data collection scenario is a fixed scenario; if so, determine the corner point of the standard area based on the preset corner point in the road data, and determine the road area data contained in the road data based on the corner point; if not, determine the road area data contained in the road data based on the pre-installed fixed marker points in the road data.

6. An electronic device, characterized in that, include: Processors and storage devices; The storage device stores computer-executable instructions that can be executed by a processor, and the processor executes the computer-executable instructions to implement the steps of the vehicle behavior warning method according to any one of claims 1 to 4.

7. A storage medium, characterized in that, The storage medium stores computer-executable instructions, which, when invoked and executed by a processor, cause the processor to implement the steps of the vehicle behavior warning method according to any one of claims 1 to 4.

Images