Data processing method and system, storage medium and product

Abstract

The invention discloses a data processing method and system, a storage medium and a product, and the method comprises the steps: a data perception module analyzes foresight perception data, and obtains a structured analysis result; based on a structured analysis result, packaging according to a structured data format corresponding to the sensing theme to obtain initial message data of the sensing theme; the data processing module obtains intermediate message data of all perception themes through the data receiving module, and receives function requirements from the vehicle end function module; determining a target sensing theme corresponding to the function demand, and processing the intermediate message data of the target sensing theme according to the function demand to obtain the processed intermediate message data of the target sensing theme; filling data fields in the processed intermediate message data of the target sensing theme into target structured message data corresponding to the vehicle end function module; and the data sending module sends the target structured message data of all the target sensing themes to the corresponding vehicle end function modules.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular to a data processing method and system, a storage medium and a product. BACKGROUND

[0002] With the continuous development of intelligent driving technology, the perception and processing capabilities of vehicles on the surrounding environment are increasingly improved. As the core support of intelligent driving, the Advanced Driver Assistance Systems (ADAS) relies on multiple types of sensors such as cameras and radars to synchronously collect environmental data of roads, vehicles, pedestrians, etc., and through accurate analysis and operation, provides decision basis for key functions such as lane keeping assistance, automatic emergency braking, adaptive cruise, etc. In this process, efficient and flexible data processing and communication mechanisms have become an important foundation for the reliable operation of intelligent driving systems.

[0003] However, in the current application of ADAS technology, there are still obvious bottlenecks in the processing flow of ADAS forward-looking data. Such data is usually transmitted in the form of sensor raw data stream, and subsequent multi-link processing such as data analysis, format conversion, protocol encapsulation is required before it can be called by the function control module at the vehicle end. It should be noted that these processing links often exhibit strong coupling characteristics, and lack clear functional boundaries between steps, making it difficult to perform targeted functional expansion according to the actual needs of different scenarios, such as the differentiated data processing needs of high-speed driving and urban roads, resulting in low overall data flow efficiency and difficulty in matching the high real-time requirements of intelligent driving. SUMMARY

[0004] The present application provides a data processing method, system, storage medium and product.

[0005] The technical solution of the present application is implemented as follows:

[0006] In a first aspect, the present application provides a data processing method applied to a data processing system, the data processing system comprising a data perception module, a data receiving module, a data processing module and a data sending module, the method comprising:

[0007] The data perception module analyzes the forward-looking perception data in the collected vehicle end forward-looking data to obtain structured analysis results of different object types; based on the structured analysis results of each object type, the initial message data of the perception theme is obtained by encapsulating according to the structured data format corresponding to the perception theme associated with the object type;

[0008] The data processing module obtains intermediate message data of all sensing topics through the data receiving module, and receives functional requirements from the vehicle-side functional modules; wherein, the intermediate message data is at least a portion of the message data in all the initial message data;

[0009] The data processing module determines the target perception topic corresponding to the functional requirements, processes the intermediate message data of the target perception topic according to the functional requirements, and obtains the processed intermediate message data of the target perception topic; the data fields in the processed intermediate message data of the target perception topic are filled into the target structured message data corresponding to the vehicle-side functional module.

[0010] The data sending module sends the target structured message data of all target perception topics to the corresponding vehicle-side functional modules.

[0011] Based on the aforementioned technical means, firstly, the data perception module performs structured analysis on the forward-looking data and encapsulates it into initial message data for multiple perception topics, improving the modularity and scalability of data processing; secondly, the data processing module can flexibly filter, complete, or transform the required data fields according to functional requirements, enhancing the system's adaptability and flexibility; finally, the data sending module optimizes the data transmission method and message structure according to the transmission strategy, improving data transmission efficiency and real-time performance, thereby effectively solving the problems of high data coupling, inconsistent communication protocols, and poor scalability in existing systems.

[0012] In some embodiments, the data processing system further includes a message middleware. The vehicle-side forward-looking data includes perception frame metadata. Based on the structured analysis results of each object type, the data is encapsulated according to the structured data format corresponding to the perception topic associated with the object type to obtain the initial message data of the perception topic. This includes: the data perception module encapsulating the data according to the structured analysis results of each object type and the corresponding perception frame metadata according to the structured data format corresponding to the perception topic associated with the object type to obtain the initial message data of the perception topic, and calling the message middleware to publish all perception topics containing the initial message data; correspondingly, the data processing module obtains the intermediate message data of all perception topics through the data receiving module, including: the data receiving module calling the message middleware to subscribe to at least one perception topic and obtain the initial message data of at least one perception topic; the data receiving module verifying the structured analysis results in the initial message data based on the perception frame metadata in the initial message data of each perception topic to obtain the intermediate message data of the perception topic, and sending the intermediate message data corresponding to all perception topics to the data processing module.

[0013] Based on the above technical means, the message publishing and subscription mechanism of the sensing topic is realized by introducing message middleware, which makes data transmission more decoupled and efficient; at the same time, the sensing frame metadata is used to verify the structured analysis results, ensuring the consistency and validity of the data, and further improving the stability and reliability of the system operation.

[0014] In some embodiments, the data receiving module verifies the structured analysis results in the initial message data based on the sensing frame metadata in the initial message data of each sensing topic to obtain intermediate message data of the sensing topic. This includes: the data receiving module verifies the initial message data of the sensing topic based on the timestamps and / or frame numbers corresponding to multiple structured analysis results in the initial message data of each sensing topic; wherein, the sensing metadata includes timestamps and frame numbers; if there are multiple structured analysis results with different timestamp intervals and / or discontinuous frame numbers, the data receiving module removes all structured analysis results with different timestamp intervals and / or discontinuous frame numbers from the initial message data of the sensing topic to obtain intermediate message data of the sensing topic.

[0015] Based on the above technical means, by strictly verifying the timestamp and frame sequence number, abnormal or incomplete data is filtered out, ensuring the quality and consistency of subsequent data processing, avoiding misjudgment and processing failure caused by data abnormalities, thereby improving the overall robustness and stability of the system.

[0016] In some embodiments, the data processing module processes the intermediate message data of the target sensing topic according to functional requirements to obtain the processed intermediate message data of the target sensing topic, including: parsing the intermediate message data of the sensing topic in the order of sensing topics to obtain the topic name and multiple data fields of the target sensing topic; filtering the first part of the data fields among the multiple data fields according to functional requirements, and / or performing semantic completion on the second part of the data fields to obtain the partial data fields that match the functional requirements, wherein the processed intermediate message data of the target sensing topic includes the topic name and the partial data fields that match the functional requirements.

[0017] Based on the above technical means, by parsing the message data of the target-aware topic, dynamic filtering and semantic completion of data fields are supported, which can meet the specific needs of different functional modules, significantly improve the system's flexibility and adaptability, reduce redundant data transmission, and reduce resource consumption.

[0018] In some embodiments, the data sending module sends all target structured message data of the target perception topic to the corresponding vehicle-side functional module, including: the data sending module determines the transmission strategy of the target structured message data according to the data volume of the target structured message data, the corresponding perception topic type and the data transmission requirements of the vehicle-side functional module; wherein, the transmission strategy includes the transmission method and message structure; the data sending module sends the target structured message data to the corresponding vehicle-side functional module according to the transmission strategy.

[0019] Based on the above technical means, by dynamically formulating transmission strategies and combining data volume, perceived topic type and functional requirements, appropriate transmission methods and message structures are selected, which improves the efficiency and stability of data transmission. It is especially suitable for efficient transmission in large data volume scenarios. It can also reduce system response latency and enhance the real-time performance of the system by combining asynchronous transmission mechanisms.

[0020] In some embodiments, the data sending module determines the transmission strategy of the target structured message data based on the data volume of the target structured message data, the corresponding perception topic type, and the functional requirements of the vehicle-side functional module. This includes: if the data volume is less than the data volume threshold and the perception topic type is a first perception topic type, the transmission strategy of the target structured message data is determined to be a single transmission method and matched with a lightweight message structure; if the data volume is greater than or equal to the data volume threshold and the perception topic type is a second perception topic type, the transmission strategy of the target structured message data is determined to be a split transmission method and matched with a big data message structure.

[0021] Based on the above technical means, by setting data volume thresholds and classifying perceived topic types, automatic switching of transmission strategies is achieved, which not only ensures efficient transmission in the case of small data volume, but also copes with reliable transmission in the case of large data volume, thus taking into account both transmission efficiency and system stability.

[0022] In some embodiments, the data processing system further includes a cache management module, which includes a first cache area and a second cache area. The method further includes: the first cache area stores intermediate message data of all sensing topics currently being processed by the data processing module, and the second cache area receives and stores new intermediate message data of all sensing topics sent by the data sensing module to the data processing module, so that after the data processing module finishes processing the current intermediate message data, it obtains the next intermediate message data to be processed from the second cache area.

[0023] Based on the above technical means, the dual-buffer design achieves seamless connection and efficient flow in the data processing process, avoids the risk of data processing interruption and loss, ensures the continuity and stability of the data processing flow, and also improves the system's throughput and response speed.

[0024] Secondly, this application provides a data processing system, comprising: a data sensing module, a data receiving module, a data processing module, and a data sending module, wherein...

[0025] The data perception module is used to analyze the forward perception data in the collected vehicle-side forward-looking data to obtain structured analysis results for different object types; based on the structured analysis results of each object type, it is encapsulated according to the structured data format corresponding to the perception topic associated with the object type to obtain the initial message data of the perception topic.

[0026] The data processing module is used to obtain intermediate message data of all sensing topics through the data receiving module, and to receive functional requirements from the vehicle-side functional modules; wherein, the intermediate message data is at least a portion of the message data in all initial message data; determine the target sensing topic corresponding to the functional requirements, process the intermediate message data of the target sensing topic according to the functional requirements, and obtain the processed intermediate message data of the target sensing topic; fill the data fields in the processed intermediate message data of the target sensing topic into the target structured message data corresponding to the vehicle-side functional modules;

[0027] The data sending module is used to send the target structured message data of all target perception topics to the corresponding vehicle-side functional modules.

[0028] Thirdly, this application provides a data processing device, which includes: a transceiver, a memory, and a processor.

[0029] The memory stores computer programs that can run on a processor;

[0030] The transceiver is used to perform the operation of receiving or sending data under the control of the processor;

[0031] When the processor executes the computer program, it implements some or all of the steps in the data processing method as described in the first aspect.

[0032] Fourthly, this application provides a computer-readable storage medium storing one or more computer programs, which can be executed by one or more processors to implement some or all of the steps in the data processing method as described in the first aspect.

[0033] Fifthly, this application provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement some or all of the steps in the data processing method as described in the first aspect.

[0034] The beneficial effects of this application are:

[0035] 1) Modular data processing architecture: The modular design decomposes the ADAS data processing flow into independent functional modules, realizing the independence and flexibility of functions, and facilitating system maintenance, upgrades and expansion;

[0036] 2) Efficient data conversion and sending mechanism: Message encapsulation and conversion, by parsing ADAS data and encapsulating it into a message format conforming to the ROS protocol, supports the transmission of data packets of different sizes, optimizing data transmission efficiency; Dynamic message processing: Based on the data content and functional module requirements, it dynamically selects the appropriate message structure and transmission method, improving the system's flexibility and adaptability; Efficient sending mechanism: Through the ROS protocol, it achieves fast data transmission, supports multi-threaded processing and asynchronous transmission, significantly improving the system's real-time performance and communication efficiency.

[0037] 3) High real-time performance and reliability: The caching mechanism reduces data processing latency and ensures real-time data processing; the middleware is designed with a robust error handling mechanism that can detect and handle abnormal situations in data transmission, thereby improving the reliability of the system. Attached Figure Description

[0038] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with this application and, together with the specification, serve to explain the technical solutions of this application.

[0039] Figure 1 This application provides a schematic diagram of the structure of an optional data processing system according to an embodiment of the present application.

[0040] Figure 2 A schematic diagram of the implementation flow of an optional data processing method provided in this application embodiment. Figure 1 ;

[0041] Figure 3 A schematic diagram of the implementation flow of an optional data processing method provided in this application embodiment. Figure 2 ;

[0042] Figure 4 A schematic diagram of the implementation flow of an optional data processing method provided in this application embodiment. Figure 3 ;

[0043] Figure 4 A schematic diagram of the implementation flow of an optional data processing method provided in this application embodiment. Figure 6 ;

[0044] Figure 5 A schematic diagram of the implementation flow of an optional data processing method provided in this application embodiment. Figure 7 ;

[0045] Figure 1This is a schematic diagram of the hardware structure of an optional data processing device provided in an embodiment of this application. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application are further described in detail below with reference to the accompanying drawings and embodiments. The described embodiments should not be regarded as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0047] In the following description, references to "some embodiments" refer to a subset of all possible embodiments. It is understood that "some embodiments" may be the same or different subsets of all possible embodiments and may be combined with each other without conflict. The terms "first / second / third" are used merely to distinguish similar objects and do not represent a specific ordering of objects. It is understood that "first / second / third" may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0048] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terminology used herein is for descriptive purposes only and is not intended to limit the scope of this application.

[0049] Reference Figure 2 As shown in the figure, this application embodiment provides a data processing system. The data processing system 100 includes: a data sensing module 101, a data receiving module 102, a data processing module 103, and a data sending module 104. Of course, the data processing system may also include a message middleware, a cache management module 105, and a system management module 106.

[0050] The data perception module is used to collect real-time forward-looking data from the vehicle and analyze the forward-looking perception data to obtain structured analysis results for different object types. Based on the structured analysis results, it publishes perception topics containing the structured analysis results by calling a message middleware such as the Robot Operating System (ROS). It should be noted that the data to be collected is the same for different functional or data processing needs of different vehicle-side functional modules. Therefore, for different vehicle-side functional modules, all forward-looking data can be collected through the same data perception module. This forward-looking data includes, but is not limited to, video data and radar data. After the data perception module collects the forward-looking data, it can analyze the data to obtain structured analysis results for different object types and publish perception topics containing initial message data with the structured analysis results in the ROS operating system. This ensures that the initial message data of the published perception topic has information such as timestamps and frame numbers. Perception topics include, but are not limited to, lane line information, obstacle information, and signage information, so that other modules can acquire or subscribe to them.

[0051] The data receiving module subscribes to multiple sensing topics through the ROS system. It employs a data pre-validation mechanism to verify the initial message data from these topics, discarding data with discontinuous timestamp intervals or frame numbers to obtain intermediate message data for the sensing topics. This ensures the validity and continuity of the input. Furthermore, the intermediate message data from the sensing topics is passed to downstream modules, such as the data processing module, for further processing.

[0052] The data processing module primarily receives functional requirements from vehicle-side functional modules, determines the target perception topic corresponding to each requirement, processes the intermediate message data of the target perception topic according to the functional requirements, and obtains processed intermediate message data for the target perception topic. It then fills the data fields from the processed intermediate message data into the target structured message data corresponding to the vehicle-side functional module. Different vehicle-side functional modules have different functional requirements, and the perception topics corresponding to the data to be processed differ. For example, the driving requirement for intelligent assisted driving might be sign information and lane line information; the target perception topic corresponding to this driving requirement could be sign information and lane line information, meaning the required data consists of the intermediate message data corresponding to sign information and lane line information, respectively. Furthermore, even if the required target perception topic is the same for different functional requirements, the data fields of the intermediate message data corresponding to the required target perception topic might differ, resulting in different content from the same data being sent to different vehicle-side functional modules.

[0053] The data sending module is used to send the target structured message data of all target perception topics to the corresponding vehicle-side functional modules.

[0054] The system management module connects to the data sensing module, data receiving module, data processing module, and message middleware. It can also connect to other modules, such as the communication management module, and is responsible for the lifecycle management of the entire process. During system startup, the system management module initializes all necessary modules, including the data processing and communication management modules, ensuring collaborative operation. The initialization process is as follows: load configuration → start ROS nodes → register topics → pre-allocate cache → start daemon threads. During system operation, the system management module continuously monitors the system's running status to ensure the stability of data processing and transmission. During runtime, it monitors the Central Processing Unit (CPU) utilization, message latency, and packet loss rate in real time, triggering dynamic frequency reduction or log alarms when limits are exceeded. When the system stops, the system management module cleans up all resources to ensure a clean system exit. Thus, through the collaborative work of these modules, the data processing system provided in this embodiment can achieve efficient parsing, stable transmission, and flexible expansion of multi-source data, making it suitable for real-time sensing and control needs in complex driving environments.

[0055] Reference Figure 2 , Figure 1 This is a schematic diagram illustrating the implementation flow of a data processing method provided in an embodiment of this application. Applied to a data processing system, the data processing system can... Figure 2 As shown, the data processing system includes: a data sensing module, a data receiving module, a data processing module, and a data sending module. Here, we will combine... Figure 3 The steps shown are explained.

[0056] Step 201: The data perception module analyzes the forward perception data in the collected vehicle-side forward-looking data to obtain structured analysis results for different object types; based on the structured analysis results for each object type, it encapsulates the data according to the structured data format corresponding to the perception topic associated with the object type to obtain the initial message data of the perception topic.

[0057] In this embodiment, the data perception module is typically a camera or other sensor such as radar installed at the front of the vehicle. It is responsible for collecting real-time information about the environment in front of the vehicle, i.e., the vehicle-side forward-looking data. The vehicle-side forward-looking data includes forward-looking perception data and perception frame metadata. The forward-looking perception data may include video image data and other radar data collected by the radar sensor. The perception frame metadata may include the timestamp and frame number of the forward-looking perception data collection.

[0058] In this embodiment, the data perception module can have a built-in processing chip, enabling it to directly perform preliminary detection and identification of objects in the forward-looking perception data at the hardware level, obtaining structured analysis results for different object types. It should be noted that the structured analysis results can be data content with clearly defined fields and semantics, formed after preliminary processing. Examples include obstacle bounding boxes, confidence levels, and type labels extracted from images. This data is encapsulated in a structured manner within messages for easy subsequent processing and use.

[0059] In this embodiment, the object type can be different categories of entities or targets distinguished in the pre-sensing data, used for classifying and processing different sensing content. Object types include, but are not limited to, obstacles, lane lines, signs, traffic lights, and ground markings. Each object type has a corresponding specific structured analysis result, which can be used for subsequent processing and distribution.

[0060] Here, obstacles include, but are not limited to, steps, bollards, pedestrians, other vehicles, or animals. The structured analysis results of obstacle object types include, but are not limited to, the obstacle's category, its bounding box, its detection confidence level, and its motion information, such as speed and direction. The structured analysis results of obstacle object types are also known as obstacle information.

[0061] Here, lane lines can be lines on a road used to delineate lane boundaries. The structured analysis results of lane line object types include, but are not limited to, lane line categories (such as solid lines, dashed lines, double yellow lines) and lane line coordinates, used to determine whether the current vehicle is deviating from the lane center, thus assisting the autonomous driving system in making corresponding adjustments. The structured analysis results of lane line object types are also known as lane line information.

[0062] Here, signs represent traffic signs on roads, such as speed limit signs and no-stopping signs. The structured analysis results of sign object types include, but are not limited to, traffic sign category, location coordinates, and detection confidence levels. These structured analysis results can be used to assist drivers or autonomous driving systems in making correct decisions. The structured analysis results of sign object types are also known as sign information.

[0063] In this embodiment, a perception topic can be a data set or information category defined for a certain type of object, used for unified management and distribution of perception data. For example, obstacle information, lane line information, and sign information correspond to different perception topics. Each perception topic has a unified structured data format and can be obtained by subscribers on demand.

[0064] In this embodiment, the data perception module can analyze the forward-looking perception data collected from the vehicle's forward-looking data to obtain structured analysis results for different object types, thereby transforming the raw perception data into structured data. Furthermore, the structured data transformed by the data perception module is encapsulated into corresponding perception themes according to different object types. For example, obstacle information, lane line information, and sign information correspond to different perception themes, each with a unified data format definition. By encapsulating structured data into different perception themes, the subsequent data receiving module can subscribe to multiple perception themes, allowing the data processing module to process the data corresponding to multiple perception themes according to different functional requirements, achieving flexible data processing and accurate distribution.

[0065] Step 202: The data processing module obtains intermediate message data of all sensing topics through the data receiving module, and receives functional requirements from the vehicle-side functional modules.

[0066] The intermediate message data consists of at least a portion of all the initial message data.

[0067] In this embodiment, intermediate message data refers to temporary data generated from initial message data after verification or preprocessing during the data processing flow, which can be used for further processing. Intermediate message data typically has invalid or abnormal content removed from the initial message data, ensuring data quality and consistency.

[0068] In this embodiment, the data receiving module is used to obtain initial message data for each sensing topic from the data sensing module. Since not all initial message data will be used, the data receiving module only receives the data that requires further processing, i.e., the intermediate message data for the sensing topic. The intermediate message data for the sensing topic is typically pre-verified data, i.e., invalid or abnormal content has been removed to ensure data consistency and reliability. For example, during data reception, if the timestamp interval of a frame of initial message data is too large or the frame sequence number is discontinuous, the data receiving module will discard that frame of initial message data to avoid affecting subsequent processing efficiency.

[0069] In this embodiment, the vehicle-side functional module can be a software or hardware component installed on an intelligent driving vehicle to perform specific functions, such as a microcontroller unit (MCU) control module, an ADAS decision module, an instrument panel display module, etc. The vehicle-side functional module relies on structured perception data to complete operations such as decision-making, control, or feedback.

[0070] In this embodiment of the application, the functional requirements of the vehicle-side functional module include driving requirements, safety protection requirements, and instrument display requirements.

[0071] In this embodiment, the data processing module can also receive functional requirements from vehicle-side functional modules. Vehicle-side functional modules refer to software or hardware components within the vehicle that perform specific control tasks, such as MCUs and ADAS control modules. These modules may propose different data processing requirements based on the current driving scenario; for example, they may need to obtain information about the current lane lines for lane keeping control, or they may need to obtain information about obstacles ahead for automatic emergency braking. Therefore, the data processing module needs to be able to dynamically respond to the functional requirements of the vehicle-side functional modules and, based on the functional requirements from different modules, process the intermediate message data of the perception subject accordingly to obtain data corresponding to different functional requirements.

[0072] In actual implementation, the data processing module first obtains intermediate message data for all sensing subjects through the data receiving module, and then receives specific functional requirements from the vehicle-side functional modules. Because the data processing module first obtains intermediate message data for all sensing subjects through the data receiving module, and then receives specific functional requirements from the vehicle-side functional modules, the system can quickly respond to the input requirements of downstream modules while ensuring data quality, thus improving the overall system's flexibility and adaptability.

[0073] Step 203: The data processing module determines the target perception topic corresponding to the functional requirements, processes the intermediate message data of the target perception topic according to the functional requirements, and obtains the processed intermediate message data of the target perception topic; the data fields in the processed intermediate message data of the target perception topic are filled into the target structured message data corresponding to the vehicle-side functional module.

[0074] In this embodiment, the target perception topic can be a perception topic corresponding to the functional requirements of the vehicle-side functional modules. The target perception topics corresponding to the functional requirements of different vehicle-side functional modules may be the same or different, and there can generally be one or more target perception topics. For example, the vehicle-side functional modules include vehicle-side functional module 1 and vehicle-side functional module 2, and the perception topics include perception topic A, perception topic B, perception topic C, and perception topic D; wherein, the perception topics corresponding to the functional requirements of vehicle-side functional module 1 include perception topic A and perception topic D, and the perception topics corresponding to the functional requirements of vehicle-side functional module 2 include perception topic B, perception topic C, and perception topic D.

[0075] In this embodiment, the data processing module first determines the perception topic corresponding to the received functional requirements. For example, if the functional requirement is lane keeping control, the corresponding perception topic might be lane line information; if the functional requirement is automatic emergency braking, the corresponding perception topic might be obstacle information. Once the target perception topic is determined, the data processing module further processes the intermediate message data under the target perception topic, including but not limited to data filtering, field cleaning, and structure filling, to obtain the processed intermediate message data for each target perception topic.

[0076] It should be noted that the processed intermediate message data will be organized into a structured message data suitable for downstream modules. For example, for an MCU module, which may require a specific CAN format data structure, the data processing module will fill the processed data fields into a message structure conforming to the CAN format data structure. This structured message data not only contains the necessary key information but also excludes irrelevant fields, reducing data packet size and improving transmission efficiency.

[0077] In this embodiment, the target structured message data can be a data structure that is ultimately formed and sent to the corresponding vehicle-side functional module after intermediate message data of the target perception topic has been filtered, cleaned, and supplemented according to functional requirements. This data structure is adapted to the input requirements of downstream modules, ensuring the validity and usability of the data.

[0078] It should be noted that different vehicle-side functional modules support different data formats. The data processing module needs to encapsulate the cleaned and completed data into target structured message data corresponding to the vehicle-side functional module. For example, if the vehicle-side functional module is an MCU control module, the structured data format of the MCU control module can be binary data in CAN bus format, which facilitates real-time data transmission and control; if the vehicle-side functional module is an ADAS decision module, the structured data format of the ADAS decision module can be a lightweight serialization format, suitable for cross-language transmission and storage, such as a protobuf structure; if the vehicle-side functional module is an instrument panel display module, the structured data format of the instrument panel display module can be JSON format. Of course, on-demand expansion is also supported. If a new vehicle-side functional module requires XML format, only the corresponding encapsulation logic needs to be added, without modifying the upstream processing flow.

[0079] As described above, the data processing module can dynamically select target-aware topics based on functional requirements and adapt relevant data to generate structured message data that meets the requirements of downstream modules. This not only improves data processing efficiency but also enhances the system's adaptability to different application scenarios.

[0080] Step 204: The data sending module sends the target structured message data of all target perception topics to the corresponding vehicle-side functional modules.

[0081] In this embodiment, the data sending module is responsible for sending the processed target structured message data to the corresponding vehicle-side functional module. The data sending module supports multiple communication protocols and message structures, and can select the appropriate method based on the data size and transmission scenario. For example, for small data volume scenarios, a compact message structure can be used to reduce resource consumption and speed up transmission; while for large data volume scenarios, a large data message structure can be used, supporting data splitting and compression to improve throughput.

[0082] In practical implementation, the data perception module, data receiving module, data processing module, and data sending module work closely together. The data perception module, as the source, is responsible for collecting and initially analyzing data, generating structured analysis results, and encapsulating them according to the perception theme. The data receiving module is responsible for filtering and receiving intermediate message data from the perception theme, and simultaneously receiving functional requirements from the vehicle-side functional modules. The data processing module processes the data of the target perception theme according to the received functional requirements and generates structured message data that meets the requirements of downstream modules. The data sending module is responsible for efficiently and reliably sending the processed data to the target functional modules. This process, consisting of these four modules, realizes a closed-loop data processing link from front-end perception to back-end control, ensuring the stable operation of the intelligent driving system in various complex scenarios.

[0083] Reference Figure 3 , Figure 2 This is a schematic diagram illustrating the implementation flow of a data processing method provided in an embodiment of this application, applied to a data processing system. The data processing system may further include a message middleware, based on... Figure 2 , Figure 2 Step 201 in the text can be replaced with step 301. Figure 3 Step 202 in the text can be replaced by steps 302 and 304. Here, we will combine... Figure 4 The steps shown are explained.

[0084] Step 301: The data perception module analyzes the forward perception data in the collected vehicle-side forward-looking data to obtain structured analysis results for different object types; based on the structured analysis results of each object type and the corresponding perception frame metadata, it encapsulates the data according to the structured data format corresponding to the perception topic associated with the object type to obtain the initial message data of the perception topic, and calls the message middleware to publish all perception topics containing the initial message data.

[0085] In this embodiment, the data perception module is responsible for processing the structured perception data acquired and initially analyzed by the forward-looking camera. The data perception module encapsulates and converts the detection results of different types of objects (such as vehicles, pedestrians, lane lines, etc.) for use by subsequent modules. The data perception module can ensure the temporal consistency and integrity of the data processed by it by reading metadata such as the timestamp, frame number, image resolution, and sensor status of the acquired forward-looking data. The structured analysis results refer to the data structures generated after algorithmic processing, such as obstacle bounding boxes, confidence scores, and semantic labels. These results are typically stored in structured data formats such as JSON, XML, or Protobuf.

[0086] In this embodiment, a message middleware is a system component used to implement inter-module communication. The message middleware allows different modules to send and receive data asynchronously. The message middleware used in this embodiment supports multiple communication protocols, such as the topic mechanism in the ROS system. The message middleware can efficiently transmit data between multiple modules on the vehicle side. Publishing the encapsulated initial message data to the message middleware ensures that other modules can obtain and process the data in a timely manner.

[0087] In this embodiment, the data perception module analyzes the forward-looking perception data collected from the vehicle's forward-looking data to obtain structured analysis results for different object types. Based on the structured analysis results of each object type and the corresponding perception frame metadata, the data is encapsulated according to the structured data format corresponding to the perception topic associated with the object type to obtain the initial message data of the perception topic. Thus, the data perception module achieves unified encapsulation and standardized processing of the structured perception data collected and initially parsed by the forward-looking camera. Furthermore, the data perception module calls a message middleware such as the ROS system to publish all perception topics containing the initial message data. In this way, cross-module data sharing is achieved through the message middleware, thereby improving the system's real-time performance and flexibility in data processing.

[0088] Step 302: The data receiving module calls the message middleware to subscribe to at least one sensing topic and obtains the initial message data of at least one sensing topic.

[0089] In this embodiment, the data receiving module is a module in the system used to receive sensing topic data from the message middleware. The function of the data receiving module is to listen to a specific sensing topic according to a preset subscription logic and obtain the data corresponding to that sensing topic. A sensing topic is a data channel managed by the message middleware, and each sensing topic corresponds to a type of data stream.

[0090] Step 303: The data receiving module verifies the structured analysis results in the initial message data based on the sensing frame metadata in the initial message data of each sensing topic, obtains the intermediate message data of the sensing topic, and sends the intermediate message data corresponding to all sensing topics to the data processing module.

[0091] In this embodiment of the application, after the data receiving module receives the initial message data, the data receiving module extracts the sensing frame metadata in the initial message data, such as timestamps and frame numbers. Based on the sensing frame metadata, the data receiving module performs verification processing on the structured analysis results in the initial message data to determine the continuity and validity of the data, thereby obtaining the intermediate message data of each sensing topic, and sending the intermediate message data corresponding to all sensing topics to the data processing module.

[0092] In some embodiments, step 303, which verifies the structured analysis results in the initial message data based on the sensing frame metadata in the initial message data of each sensing topic to obtain the intermediate message data of the sensing topic, can be implemented through the following process:

[0093] Based on the timestamps and / or frame numbers corresponding to multiple structured analysis results in the initial message data of each sensing topic, the initial message data of the sensing topic is verified. The sensing metadata includes timestamps and frame numbers. If there are multiple structured analysis results with different timestamp intervals and / or discontinuous frame numbers, all structured analysis results with different timestamp intervals and / or discontinuous frame numbers are removed from the initial message data of the sensing topic to obtain the intermediate message data of the sensing topic.

[0094] In this embodiment, the timestamp refers to the acquisition time of the vehicle-side forward-looking data, typically measured in milliseconds or microseconds. The vehicle-side forward-looking data may include video data, and the frame number may be the sequential number of an image frame or data frame within the acquired video data. By combining the timestamp and frame number, the data receiving module can determine whether the perceived data is continuous and whether there are any abnormal jumps or losses. For example, after the forward-looking camera acquires an image frame, the data perception module assigns an incrementing frame number to the image frame corresponding to the structured analysis result in the initial message data of each perceived topic and records the acquisition timestamp of the image frame. The data receiving module can use the timestamp and frame number to verify the integrity and synchronization of the data.

[0095] It should be noted that inconsistent timestamp intervals or discontinuous frame sequence numbers indicate potential communication delays, data loss, or other anomalies. Continuing to use the structured analysis results in subsequent data processing modules may lead to misjudgments or calculation errors. Therefore, when inconsistent timestamp intervals or discontinuous frame sequence numbers are detected, the data receiving module will automatically discard these undesirable structured analysis results, retaining only valid results that meet the time logic and sequence requirements. This discarding operation can be achieved by traversing the dataset and checking each structured analysis result's timestamp and frame sequence number against preset conditions. For example, the maximum allowable time difference between two adjacent frames can be set to 50ms; if this is exceeded, the data receiving module considers the structured analysis result unreliable and should discard it. Similarly, if the frame sequence number is not continuously increasing, it may indicate data loss, and the data receiving module should discard the structured analysis result.

[0096] Furthermore, after removing invalid structured analysis results, the data receiving module generates a new intermediate message data set corresponding to the perceived topic. This new intermediate message data set contains only verified, temporally coherent, and sequentially numbered structured analysis results. This step is crucial for ensuring the accuracy of subsequent data processing (such as target tracking and behavior prediction). Discontinuous or abnormal structured analysis results may distort the algorithm model's input, thus affecting the final decision.

[0097] In this embodiment, by introducing timestamps and frame numbers, structured analysis results with inconsistent timestamp intervals or discontinuous frame numbers are eliminated. This reduces the impact of invalid structured analysis results on system performance, ensures the quality and consistency of intermediate message data, and thus improves the overall reliability and real-time performance of the intelligent driving system.

[0098] Step 304: The data processing module receives functional requirements from the vehicle-side functional modules.

[0099] Step 203: The data processing module determines the target perception topic corresponding to the functional requirements, processes the intermediate message data of the target perception topic according to the functional requirements, and obtains the processed intermediate message data of the target perception topic; the data fields in the processed intermediate message data of the target perception topic are filled into the target structured message data corresponding to the vehicle-side functional module.

[0100] Step 204: The data sending module sends the target structured message data of all target perception topics to the corresponding vehicle-side functional modules.

[0101] In some embodiments, for the data processing module, step 203 involves processing the intermediate message data of the target-aware topic according to functional requirements to obtain the processed intermediate message data of the target-aware topic. Figure 5 Please provide an explanation.

[0102] Step 401: According to the order of the perceived topics, parse the intermediate message data of the target perceived topic to obtain the topic name and multiple data fields of the target perceived topic.

[0103] In this embodiment of the application, the topic name of the target perception topic can be the target perception topic identifier to which the currently processed data belongs. The topic name is used to distinguish the sources of different types of perception data. The data fields are the specific parameters parsed, such as vehicle position, speed, confidence level, frame number, etc.

[0104] In this embodiment, the data processing module parses the intermediate message data of the sensing topics according to the order of sensing topics, thereby decomposing the intermediate message data into more granular information units, such as the data fields of the target sensing topic name and specific parameters, for subsequent processing.

[0105] In some embodiments, the data receiving module can generate a first checksum for the data corresponding to different sensing topics and add it to the message data of the sensing topic. After the data receiving module sends the intermediate message data of all sensing topics to the data processing module, and the data processing module parses and obtains multiple data fields, it can perform a second checksum verification on the data of the multiple data fields. For example, the verification method can be to generate a second checksum based on the data of the multiple data fields, compare the second checksum with the first checksum carried in the intermediate message data, and if they match, it indicates that the data is correct, thus ensuring the integrity of the data transmission process.

[0106] Step 402: Based on the functional requirements, filter the first part of the data fields among multiple data fields, and / or perform semantic completion on the second part of the data fields to obtain the partial data fields that match the functional requirements. Among them, the intermediate message data after processing of the target-aware topic includes the topic name and the partial data fields that match the functional requirements.

[0107] In this embodiment, the first data field can be a data field that matches the functional requirements from among multiple data fields; the second data field can be a data field required by the functional requirements, and the second data field is different from the first data field. It should be noted that the field value corresponding to the second data field can be predicted based on the field values ​​of multiple data fields.

[0108] In this embodiment, functional requirements can be the specific usage scenarios and processing requirements of vehicle-side functional modules for message data of different perception themes. For example, some control modules only need the vehicle's position and direction information, and not complete image data. Filtering refers to removing data fields irrelevant to functional requirements, reducing the transmission and processing overhead of redundant information during the filtering process, and retaining only the fields required by downstream modules, i.e., vehicle-side functional modules, thus reducing data transmission volume. Semantic completion refers to supplementing missing but necessary information based on existing data, such as adding the current time when a timestamp is missing, or predicting lane line information in the vehicle's rear-view data based on lane line information collected from front-end perception data; or, if the upstream perception module outputs a flag ID=5, the data processing module can complete it to flag type=speed limit 60km / h according to the functional requirements of the dashboard display module and a preset mapping table, so that the downstream dashboard display module can directly display text instead of the ID. Thus, by dynamically adapting functional requirements, the intelligent driving system can provide customized data output without changing the original data structure, thereby improving data processing efficiency, reducing communication bandwidth usage, and enhancing the system's flexibility and responsiveness.

[0109] In this embodiment, the data processing module parses the intermediate message data of the target perception topic according to the perception topic order, and filters the first part of the data fields in multiple data fields according to the functional requirements of the downstream vehicle-side functional modules, and / or performs semantic completion on the second part of the data fields. This can effectively separate the data processing logic from the data content, thereby improving the modularity and processing efficiency of the system, and further realizing support for multi-platform compatibility and flexible expansion.

[0110] In some embodiments, step 204, where the data sending module sends the target structured message data of all target perception topics to the corresponding vehicle-side functional modules, is combined with... Figure 1 Please provide an explanation.

[0111] Step 501: The data sending module determines the transmission strategy of the target structured message data based on the data volume of the target structured message data, the corresponding perception topic type, and the data transmission requirements of the vehicle-side functional modules; wherein, the transmission strategy includes the transmission method and message structure;

[0112] Step 502: The data sending module sends the target structured message data to the corresponding vehicle-side functional module according to the transmission strategy.

[0113] In this embodiment, the data volume of the target structured message data is used to determine whether it needs to be split; the perceived topic type is used to affect the complexity and priority of the message data; the data transmission requirements of the vehicle-side functional modules include, but are not limited to, real-time requirements and data integrity requirements; and the data transmission requirements determine the direction of the transmission strategy selection.

[0114] In this embodiment, the transmission strategy includes the transmission method and message structure of the target structured message data. The transmission method may involve splitting the target structured message data or not splitting it. Furthermore, when there are multiple vehicle-side functional modules, multiple target perception topics, and multiple sets of target structured message data, the target structured message data can be transmitted asynchronously. The message structure can be in ROS message format or CAN frame format, and can also be a lightweight message structure or a large data message structure; this application does not impose specific limitations on these aspects.

[0115] In this embodiment of the application, before sending the target structured message data of the target perception topic to the vehicle-side functional module, the data sending module first needs to evaluate the target structured message data and formulate an appropriate transmission strategy based on the specific needs of the vehicle-side functional module.

[0116] For example, in an intelligent driving system, for high real-time data such as obstacle detection results, the data sending module may choose to use asynchronous transmission to avoid blocking the main thread, and the data sending module will adopt a lightweight message structure for fast parsing. Conversely, for data requiring high accuracy, such as traffic sign recognition results, the data sending module may choose synchronous transmission to ensure data consistency, and the data sending module will adopt a more complete message format to retain more semantic information. Thus, dynamically adjusting the transmission strategy for the target structured message data can optimize data transmission efficiency, reduce resource consumption, and improve system response speed and stability.

[0117] In some embodiments, the transmission strategy for the target structured message data can be determined through the following process.

[0118] If the data volume is less than the data volume threshold and the perceived topic type is the first perceived topic type, the transmission strategy for the target structured message data is determined to be a single transmission method and matched with a lightweight message structure; if the data volume is greater than or equal to the data volume threshold and the perceived topic type is the second perceived topic type, the transmission strategy for the target structured message data is determined to be a split transmission method and matched with a big data message structure.

[0119] In this embodiment, the data volume threshold can be a preset numerical standard used by the data sending module to determine the data size. When the data volume of the target structured message data is lower than the data volume threshold, it indicates that the target structured message data is small in size and suitable for processing using a lightweight transmission method. The data volume threshold can be configured according to the actual application scenario, for example, set to 1KB, 2KB or higher, to adapt to different communication bandwidths and processing capabilities.

[0120] In this embodiment, the first perception subject type can be a type of perception information that has high real-time requirements but low data complexity, such as lane line recognition results or traffic sign detection results. The perception information type corresponding to the first perception subject type usually has a fixed format and a short lifespan, so it is suitable for processing using an efficient and low-overhead transmission mechanism.

[0121] In this embodiment, the second sensing subject type can be a sensing information type with high data complexity, rich content, or frequent changes, such as obstacle detection results, multi-frame image data, etc. The data of the second sensing subject type often contains a large amount of dynamic information, and the data of the second sensing subject type has high requirements for transmission reliability and integrity.

[0122] In this embodiment, when the target structured message data is small in volume and the perception topic type is the first perception topic type, the data sending module selects a single-transmission method and a lightweight message structure. This significantly reduces communication latency and system load while ensuring real-time performance, thereby improving overall data processing efficiency. When the target structured message data is large in volume and the perception topic type is the second perception topic type, the data sending module selects a split synchronous transmission method and a large data message structure. This transmission method effectively addresses the challenges of transmitting complex data, ensuring data integrity and reliability, and reducing system resource consumption, preventing performance degradation or failure due to high load. Furthermore, the data sending module sends the target structured message data to the corresponding vehicle-side functional module according to the transmission strategy.

[0123] As can be seen from the above, by combining the data volume, the type of perception subject, and the data transmission requirements of the vehicle-side functional modules, the system can dynamically select the optimal transmission strategy and message structure to flexibly adapt to different types of data transmission requirements. This can optimize communication efficiency and system stability, thereby improving the processing performance and overall operating efficiency of forward perception data of intelligent driving vehicles.

[0124] In some embodiments, the data processing system further includes: a cache management module, the cache management module including a first cache and a second cache, and the method further includes:

[0125] The first buffer stores intermediate message data for all sensing topics currently being processed by the data processing module. The second buffer receives and stores new intermediate message data for all sensing topics sent from the data sensing module to the data processing module, so that after the data processing module finishes processing the current intermediate message data, it can obtain the next intermediate message data to be processed from the second buffer.

[0126] In this embodiment, the cache management module is a key component for coordinating the data processing flow and ensuring data stability and consistency. The cache management module achieves efficient data processing through a double-buffering mechanism. It includes a first cache area and a second cache area. The first cache area stores intermediate message data of the perception topic currently being parsed and transformed by the data processing module, while the second cache area receives and caches intermediate message data for the next batch of perception topics. The second cache area achieves a pre-loading effect, meaning that new intermediate message data is ready before the data processing module has finished processing the current data. This reduces the waiting time for the data processing module when acquiring new data and improves the overall system processing efficiency.

[0127] In this embodiment, a close data interaction exists between the cache management module and the data processing module. When the data processing module receives the first batch of intermediate message data for a sensing topic, it temporarily stores the processing results of the intermediate message data for that sensing topic in a first cache. For subsequent batches of intermediate message data for sensing topics, the data is stored in a second cache. After the data processing module completes the processing task for the current batch and sends the processing results to the data sending module, it reads the next set of intermediate message data to be processed from the second cache and continues to perform subsequent parsing and encapsulation operations. This dual-caching mechanism not only improves the system's throughput but also enhances the stability of data processing.

[0128] Below, refer to Figure 6 As shown, the flow of the data processing method is illustrated by way of a specific embodiment.

[0129] With the rapid development of intelligent driving technology, Advanced Driver Assistance Systems (ADAS) are being used more and more widely in vehicles. ADAS uses various sensors such as cameras, radar, and lidar to collect real-time environmental data around the vehicle to achieve functions such as lane keeping, automatic emergency braking, and adaptive cruise control. In existing technologies, the processing and communication flow of ADAS forward-looking data has the following main problems: ADAS data is usually transmitted in the form of raw sensor data, requiring multiple steps such as parsing, conversion, and encapsulation before it can be processed by the target module. In existing systems, these multiple steps are often coupled together, lacking a modular design. This makes it difficult to flexibly expand and optimize the entire system according to specific needs, resulting in a lack of flexibility and maintainability.

[0130] In related technologies, ADAS forward-looking data is typically transmitted using a single serialization protocol, such as CDR format for data encapsulation and FastDDS shared memory for transmission to the data acquisition program. While this solution meets the real-time communication requirements of the vehicle, its lack of modular design results in tight coupling between data parsing, conversion, and encapsulation functions, making it difficult to flexibly adjust or expand according to actual needs. Furthermore, due to its reliance on a specific serialization method, cross-platform systems (such as Java backends) cannot directly parse the data, requiring the development of an additional adaptation layer, which impacts system compatibility and development efficiency.

[0131] Based on one or more of the above issues, the developers have designed a modular, efficient, reliable and easy-to-manage solution for ADAS data processing and communication transmission to meet the requirements of intelligent driving systems for high real-time performance, high reliability and flexibility.

[0132] This application provides a method and system for processing and transmitting forward-looking data of intelligent driving vehicles. Through modular design and an efficient real-time data processing and communication mechanism, the method and system solve the problems of complex data processing, insufficient real-time performance, inconsistent communication protocols, and poor system scalability in the prior art.

[0133] Firstly, in terms of data processing architecture, a modular design is adopted, decomposing the ADAS data processing flow into independent functional modules. Each module is responsible for specific processing tasks, including:

[0134] The data perception module includes a forward-facing camera mounted at the front of the vehicle. This camera collects real-time forward-facing perception data, including image streams (corresponding to the video data mentioned above), vehicle detection results, lane line recognition information, and traffic sign recognition information. Vehicle detection results include, but are not limited to, vehicle type, obstacle bounding boxes, detection confidence levels, and speed; lane line recognition results include, but are not limited to, lane line types (e.g., dashed lines, solid lines, double yellow lines) and detection confidence levels; traffic sign recognition information includes, but is not limited to, traffic sign type, detection confidence levels, and the content information contained in the traffic sign. After preliminary analysis of the video data, the data perception module publishes the analysis results in a structured format, such as JSON, through a Topic defined in the ROS system, ensuring that the data includes information such as the timestamp and frame number of the captured image. In this way, by parsing ADAS data and encapsulating it into a message format conforming to the ROS protocol, it supports the transmission of data packets of different sizes, optimizing data transmission efficiency.

[0135] In one possible implementation, the data perception module is typically a camera or other sensor installed at the front of the vehicle, responsible for real-time acquisition of information about the surrounding environment, such as image streams, obstacle detection results, and lane line recognition results. After initial analysis, the system transforms the raw data into structured data with clearly defined fields and semantics. For example, for the obstacle object type, the system's structured analysis results might include obstacle category (e.g., pedestrian, vehicle), bounding box coordinates, detection confidence, and motion state (e.g., speed, direction); for lane lines, the system's structured analysis results might include lane line category (e.g., solid line, dashed line) and lane line coordinates; for signs, the system's structured analysis results might include traffic sign category (e.g., speed limit sign, no-stopping sign), sign location coordinates, and detection confidence. Based on this structured approach, the system can ensure more efficient and accurate subsequent data processing and transmission.

[0136] The data receiving module subscribes to multiple perception topics via ROS, such as lane line information, obstacle information, and sign information. During implementation, a data pre-validation mechanism is employed: data with discontinuous timestamp intervals or frame numbers is directly discarded to ensure input validity and continuity. Once the current visual perception data is published through the ROS system, the data receiving module immediately subscribes to and receives the data, then passes it to the data processing module for further processing.

[0137] The data processing module is responsible for parsing and transforming the received data to adapt to the different needs of various vehicle-side functional modules for perception data. This module performs structured parsing, semantic completion, and target structure encapsulation on the received data. Internally, it sets up multiple processing items (corresponding to the aforementioned perception themes or topics), each dedicated to a specific type of data processing, parsing the received perception data. First, referring to… Figure 7 As shown, the data processing module can receive sensing data 601 corresponding to all sensing topics. Based on the functional requirements received from the vehicle-side functional modules, it filters out the target sensing topic matching the functional requirement from all sensing topics, i.e., filters out the processing item matching the functional requirement 602. It should be noted that if the filtered processing item is empty, meaning there is no processing item matching the functional requirement among all sensing topics, all received data is discarded 603. If the filtered processing item is not empty, meaning there is a processing item matching the functional requirement among all sensing topics, the message data corresponding to the matching processing item is parsed 604 to obtain the Topic name and data fields, and data validation is performed on the content of the message data based on the data fields 605. If the verification fails, the data is considered invalid and discarded (603). If the verification passes, meaning the content of the message data is valid, the data fields related to the functional requirements are extracted, and / or the data fields required for the functional requirements are calculated. Thus, data fields in the original message that are irrelevant to the functional requirements are discarded, thereby obtaining the fields required by the downstream module and reducing the data packet size. Then, the obtained data is filled into a data structure (such as CAN format, protobuf structure, etc.) that the vehicle-side functional module, such as the MCU, can recognize (606), supporting multiple output formats and expandable as needed, completing the data conversion process. Thus, the data processing module converts the parsed data into a structure format that can be processed by the MCU or other target vehicle-side functional modules, and further outputs the message data (607) to the vehicle-side functional module (608).

[0138] In addition, to ensure the stability and reliability of data during the conversion process, the system incorporates a cache management module. A dual-buffering mechanism (A / B areas) is employed: area A processes the current frame data, while area B receives new data and stores intermediate processing results, ensuring data consistency and integrity throughout the process. The cache management module provides temporary storage at various stages of data processing to prevent data loss or inconsistency. In this way, the cache management module optimizes the data processing flow through the caching mechanism, guaranteeing both real-time performance and stability in the data processing process.

[0139] The data transmission module sends the converted data to the target vehicle's functional modules, such as the MCU. The data transmission module selects an appropriate message structure based on the data size to achieve efficient data transmission.

[0140] Furthermore, the system management module is responsible for the lifecycle management of the entire process. During system startup, the system management module initializes all necessary modules, including the data processing module and the communication management module, ensuring that all modules work collaboratively. The initialization process is as follows: load configuration → start ROS nodes → register topics → pre-allocate cache → start daemon threads. During system operation, the system management module continuously monitors the system's running status to ensure the stability of data processing and transmission. During runtime, it monitors CPU utilization, message latency, and packet loss rate in real time, triggering dynamic frequency reduction or log alarms when limits are exceeded. When the system stops, the system management module cleans up all resources to ensure a clean system exit.

[0141] In summary, by adopting a modular design and through the collaborative work of the aforementioned modules, efficient parsing, stable transmission, and flexible expansion of multi-source data are achieved. This makes the system suitable for real-time perception and control requirements in complex driving environments, thereby achieving functional independence and flexibility, and facilitating system maintenance, upgrades, and expansion. Specific effects are as follows:

[0142] 1) Regarding data conversion and transmission mechanisms, the system significantly improves data transmission efficiency and system flexibility through message encapsulation and conversion, dynamic message processing, and efficient transmission mechanisms. Specifically, the system parses ADAS data and encapsulates it into a message format conforming to the ROS protocol, thereby supporting the transmission of data packets of different sizes; the system dynamically selects the appropriate message structure and transmission method based on the data content and the needs of the target node; the system achieves fast data transmission through the ROS protocol and has the ability to perform multi-threaded processing and asynchronous transmission, thereby significantly improving the system's real-time performance and communication efficiency.

[0143] 2) In terms of system performance, the real-time performance and reliability of the system have been further improved through caching optimization and error handling mechanisms. The caching mechanism reduces data processing latency and ensures the real-time performance of data processing; at the same time, the middleware is designed with a robust error handling mechanism that can detect and handle abnormal situations in data transmission, thereby improving the reliability of the system.

[0144] This application provides a hardware entity diagram of a data processing device, such as... ​ As shown, the data processing device can be a vehicle-mounted device. The hardware entity of the data processing device 7 includes a processor 701, a memory 702, and a transceiver 703. The memory 702 stores a computer program that can run on the processor 701. The transceiver 703 is used to perform operations of receiving or sending data under the control of the processor. When the processor 701 executes the computer program, it implements some or all of the steps in the data processing method in the above embodiments.

[0145] The memory 702 stores computer programs that can run on the processor. The memory 702 is configured to store instructions and applications that can be executed by the processor 701. It can also cache data to be processed or already processed by the processor 701 and the various modules in the data processing device 7 (e.g., image data, audio data, voice communication data and video communication data). It can be implemented by flash memory or random access memory (RAM).

[0146] The processor 701 executes the steps of the data processing execution method described above when executing the program. The processor 701 typically controls the overall operation of the data processing device 7.

[0147] The transceiver 703 receives or sends data under the control of the processor.

[0148] This application provides a computer-readable storage medium storing one or more computer programs, which can be executed by one or more processors to implement some or all of the steps in the above-described method. The storage medium can be transient or non-transient.

[0149] This application provides a computer program including computer-readable code, wherein when the computer-readable code is run in a data processing device, a processor in the data processing device performs some or all of the steps for implementing the above-described method.

[0150] This application provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program. When the computer program is read and executed by a computer, it implements some or all of the steps in the above-described method. This computer program product can be implemented specifically through hardware, software, or a combination thereof. In some embodiments, the computer program product is specifically embodied as a computer storage medium; in other embodiments, the computer program product is specifically embodied as a software product, such as a software development kit (SDK), etc.

[0151] It should be noted that the descriptions of the various embodiments above tend to emphasize the differences between them, while their similarities or commonalities can be referred to interchangeably. The descriptions of the above embodiments of the device, storage medium, computer program, and computer program product are similar to the descriptions of the above method embodiments and have similar beneficial effects. For technical details not disclosed in the embodiments of the device, storage medium, computer program, and computer program product of this application, please refer to the descriptions of the method embodiments of this application for understanding.

[0152] The aforementioned processor can be at least one of the following: Application Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Digital Signal Processing Device (DSPD), Programmable Logic Device (PLD), Field Programmable Gate Array (FPGA), Central Processing Unit (CPU), Controller, Microcontroller, and Microprocessor. It is understood that other electronic devices can also implement the functions of the aforementioned processor, and this application does not specifically limit the specific implementation.

[0153] The aforementioned computer storage media / memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), magnetic random access memory (FRAM), flash memory, magnetic surface memory, optical disc, or compact disc read-only memory (CD-ROM), etc.; or it can be various terminals that include one or any combination of the above-mentioned memories, such as mobile phones, computers, tablet devices, personal digital assistants, etc.

[0154] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification do not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. It should be understood that in the various embodiments of this application, the sequence numbers of the above steps / processes do not imply a sequential order of execution; the execution order of each step / process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application. The sequence numbers of the above embodiments of this application are merely descriptive and do not represent the superiority or inferiority of the embodiments.

[0155] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0156] In the several embodiments provided in this application, it should be understood that the disclosed 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, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0157] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0158] In addition, each functional unit in the various embodiments of this application can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0159] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as mobile storage devices, read-only memory (ROM), magnetic disks, or optical disks.

[0160] Alternatively, if the integrated units described above are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence or the part that contributes to related technologies, 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 an in-vehicle terminal (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, magnetic disks, or optical disks.

[0161] The above description is merely an embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A data processing method, characterized by, The method is applied to a data processing system, and the data processing system comprises a data sensing module, a data receiving module, a data processing module and a data sending module, and the method comprises the following steps: The data sensing module analyzes the front-view sensing data in the collected vehicle-end front-view data to obtain structured analysis results of different object types; based on the structured analysis results of each object type, the initial message data of the sensing theme is obtained by encapsulating the structured data format corresponding to the sensing theme associated with the object type; The data processing module obtains the intermediate message data of all sensing themes through the data receiving module, and receives the functional requirements from the vehicle-end functional module; wherein the intermediate message data is at least part of the initial message data; The data processing module determines the target sensing theme corresponding to the functional requirement, processes the intermediate message data of the target sensing theme according to the functional requirement, and obtains the processed intermediate message data of the target sensing theme; the data field in the processed intermediate message data of the target sensing theme is filled into the target structured message data corresponding to the vehicle-end functional module; The data sending module sends the target structured message data of all target sensing themes to the corresponding vehicle-end functional module.

2. The method of claim 1, wherein, The data processing system further comprises a message middleware, the vehicle-end front-view data comprises sensing frame metadata, and the initial message data of the sensing theme is obtained by encapsulating the structured data format corresponding to the sensing theme associated with the object type based on the structured analysis results of each object type; The data sensing module encapsulates the structured data format corresponding to the sensing theme associated with the object type based on the structured analysis results of each object type and the corresponding sensing frame metadata, and obtains the initial message data of the sensing theme; and the message middleware is called to publish all sensing themes containing the initial message data; Correspondingly, the data processing module obtains the intermediate message data of all sensing themes through the data receiving module, which comprises the following steps: The data receiving module subscribes to at least one sensing theme by calling the message middleware, and obtains the initial message data of at least one sensing theme; The data receiving module verifies the structured analysis results in the initial message data based on the sensing frame metadata in the initial message data of each sensing theme, obtains the intermediate message data of the sensing theme, and sends the intermediate message data corresponding to all sensing themes to the data processing module.

3. The method of claim 2, wherein, The data receiving module verifies the structured analysis results in the initial message data based on the sensing frame metadata in the initial message data of each sensing theme, obtains the intermediate message data of the sensing theme, which comprises the following steps: The data receiving module verifies the initial message data of each sensing theme based on the timestamps and / or frame sequence numbers corresponding to multiple structured analysis results in the initial message data of the sensing theme; wherein the sensing metadata comprises timestamps and frame sequence numbers; The data receiving module, if there is a time stamp interval between the plurality of structured analysis results is not the same and / or frame number is not continuous, eliminates all time stamp interval is not the same and / or frame number is not continuous structured analysis result from the initial message data of the perception topic, obtains the intermediate message data of the perception topic.

4. The method according to any one of claims 1 to 3, characterized in that, The data processing module processes the intermediate message data of the target perception topic according to the functional requirement, obtains the processed intermediate message data of the target perception topic, comprising: According to the order of the perception topic, the intermediate message data of the target perception topic is parsed to obtain the topic name and a plurality of data fields of the target perception topic; According to the functional requirement, a first part of the data fields is screened, and / or a second part of the data fields is semantically completed, to obtain a part of the data fields matched with the functional requirement, wherein the processed intermediate message data of the target perception topic includes the topic name and the part of the data fields matched with the functional requirement.

5. The method according to any one of claims 1 to 3, characterized in that, The data sending module sends the target structured message data of all target perception topics to the corresponding vehicle-end functional module, comprising: The data sending module determines the transmission strategy of the target structured message data according to the data amount, the corresponding perception topic type and the data transmission requirement of the vehicle-end functional module of the target structured message data; wherein the transmission strategy includes transmission mode and message structure; The data sending module sends the target structured message data to the corresponding vehicle-end functional module according to the transmission strategy.

6. The method of claim 5, wherein, The data sending module determines the transmission strategy of the target structured message data according to the data amount, the corresponding perception topic type and the functional requirement of the vehicle-end functional module, comprising: If the data amount is less than the data amount threshold value, and the perception topic type is the first perception topic type, the transmission strategy of the target structured message data is determined as single transmission mode and matching lightweight message structure; If the data amount is greater than or equal to the data amount threshold value, and the perception topic type is the second perception topic type, the transmission strategy of the target structured message data is determined as split transmission mode and matching large data message structure.

7. The method according to any one of claims 1 to 3, characterized in that, The data processing system further comprises a cache management module, the cache management module comprises a first cache area and a second cache area, and the method further comprises: The first cache area stores the intermediate message data of all perception topics currently processed by the data processing module, and the second cache area receives and stores the new intermediate message data of all perception topics sent by the data sensing module to the data processing module, so that the data processing module obtains the next intermediate message data to be processed from the second cache area after processing the current intermediate message data.

8. A data processing system, characterized by The system comprises a data sensing module, a data receiving module, a data processing module and a data sending module, wherein The data perception module is configured to analyze front-view perception data in the collected vehicle-end front-view data to obtain structured analysis results of different object types, and encapsulate the structured analysis results of each object type according to a structured data format corresponding to a perception theme associated with the object type to obtain initial message data of the perception theme. The data processing module is configured to obtain intermediate message data of all perception themes through the data receiving module, and receive a function requirement from a vehicle-end function module. The intermediate message data is at least part of the initial message data. The data processing module is further configured to determine a target perception theme corresponding to the function requirement, process the intermediate message data of the target perception theme according to the function requirement to obtain processed intermediate message data of the target perception theme, and fill data fields in the processed intermediate message data of the target perception theme into target structured message data corresponding to the vehicle-end function module. The data sending module is configured to send the target structured message data of all target perception themes to the corresponding vehicle-end function module.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores one or more computer programs that can be executed by one or more processors to implement the data processing method of any one of claims 1 to 7.

10. A computer program product comprising computer programs or instructions, characterized in that, The computer program or instruction, when executed by the processor, implements the data processing method of any one of claims 1 to 7.

Images