Image processing method and device based on automobile gateway, vehicle, medium and product

Abstract

The invention relates to the technical field of vehicles, in particular to an image processing method and device based on an automobile gateway, a vehicle, a storage medium and a program product, and the method comprises the steps: dividing to-be-processed image data into at least one piece of sub-data; determining a target sub-node of each piece of sub-data according to the gateway node and the vehicle-mounted network state; and respectively performing convolution operation and pooling operation on the sub-data in the target sub-nodes to obtain pooling results, and generating a processing result of the image data according to the pooling results of all the target sub-nodes. Therefore, the problems that in a vehicle-mounted environment, computing resources are limited, the requirement for real-time performance is high, and a traditional centralized convolutional neural network computing method is difficult to meet the requirement are solved.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to an image processing method, device, vehicle, storage medium and program product based on an automobile gateway. Background Art

[0002] In the existing technology, CNNs (Convolutional Neural Networks) have demonstrated excellent performance in image processing and classification tasks. In the vehicle environment, CNNs are mainly used in autonomous driving technology and vehicle monitoring systems. Using CNNs to process the collected images can achieve, for example, identifying road markings to help vehicles stay in lanes; identifying obstacles such as pedestrians, bicycles, and other vehicles in front to avoid collisions; monitoring the driver's status, including whether he is tired and whether he is focused, etc., which brings great driving convenience to people and improves vehicle safety.

[0003] Although CNNs have performed well in image processing, they still face challenges in the in-vehicle environment. In-vehicle systems need to quickly process large amounts of real-time data with limited computing resources, which are usually limited in in-vehicle systems. Traditional centralized CNNs computing methods require a lot of computing resources for real-time analysis and are difficult to meet these requirements. Summary of the invention

[0004] The present application provides an image processing method, device, vehicle, storage medium and program product based on an automobile gateway to solve the problem that the traditional centralized CNNs calculation method is difficult to meet the limited computing resources and high real-time requirements in the vehicle environment.

[0005] The first aspect of the present application provides an image processing method based on an automobile gateway, comprising the following steps: dividing the image data to be processed into at least one sub-data; determining the target sub-node of each sub-data according to the gateway node and the vehicle network status; performing convolution operations and pooling operations on the sub-data in the target sub-node to obtain pooling results, and generating a processing result of the image data according to the pooling results of all target sub-nodes.

[0006] Optionally, the target sub-node of each sub-data is determined according to the gateway node and the vehicle network status, including: extracting the data characteristic value of each sub-data; identifying the load status of the gateway node and the sub-node in the vehicle network status; selecting the target data flow path according to the data characteristic value load status, and determining the target sub-node of each sub-data according to the target data flow path.

[0007] Optionally, the convolution operation is calculated as:

[0008] Zij =W ij *X i +b ij ,

[0009] Where i represents the index of the gateway node, j represents the index of the child node, and Z ij represents the result of the convolution operation, X i Indicates the multiple sub-data blocks into which the input image data is divided, X i The i in the block represents the sub-data block index after segmentation, W ij represents the convolution kernel, b ij represents the bias term and * represents the convolution operation.

[0010] Optionally, the calculation formula for the pooling operation is:

[0011] P ij =Pooling(Z ij ),

[0012] Among them, P ij Represents the result of the pooling operation. Pooling(﹒) represents the pooling operation, i represents the index of the gateway node, and j represents the index of the child node.

[0013] Optionally, generating a processing result of the image data according to the pooling results of all target sub-nodes includes: inputting the pooling results of all target sub-nodes into a fully connected layer, and the fully connected layer outputs the processing result of the image data.

[0014] Optionally, the calculation formula of the fully connected layer is:

[0015]

[0016] Where O represents the result of the fully connected layer operation; FC(﹒) represents the fully connected layer operation; i represents the index of the gateway node; m represents the total number of gateway nodes participating in the fully connected calculation; P i is the pooling result of the i-th gateway node.

[0017] The second aspect of the present application provides an image processing device based on an automobile gateway, including: a division module, used to divide the image data to be processed into at least one sub-data; a determination module, used to determine the target sub-node of each sub-data according to the gateway node and the vehicle network status; a processing module, used to perform convolution operations and pooling operations on the sub-data in the target sub-node to obtain pooling results, and generate a processing result of the image data according to the pooling results of all target sub-nodes.

[0018] The third aspect of the present application provides a vehicle, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the image processing method based on the automobile gateway of the first aspect.

[0019] The fourth aspect of the present application provides a computer-readable storage medium on which a computer program or instruction is stored. When the computer program or instruction is executed, the image processing method based on the automobile gateway of the first aspect is implemented.

[0020] The fifth aspect of the present application provides a computer program product, including a computer program or instructions, which, when executed, implements the image processing method based on the automobile gateway of the first aspect.

[0021] Therefore, this application includes the following beneficial effects:

[0022] The embodiment of the present application can divide the image data to be processed into at least one sub-data, and determine the target sub-nodes of these sub-data according to the gateway node and the vehicle network status, perform convolution and pooling operations on the sub-data in the target sub-node respectively, and generate the processing results of the image data according to the pooling results of all target sub-nodes, thereby improving the computing efficiency and adaptive ability of the vehicle-mounted convolutional neural network by introducing the vehicle gateway node and dynamic routing mechanism, and providing an efficient and real-time image processing and classification solution. Thus, the problem that the traditional centralized CNNs computing method is difficult to meet the needs of the vehicle environment due to limited computing resources and high real-time requirements is solved.

[0023] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0025] Figure 1 A flowchart of an image processing method based on an automobile gateway provided according to an embodiment of the present application;

[0026] Figure 2 A flowchart of an image processing method based on an automobile gateway provided according to an embodiment of the present application;

[0027] Figure 3 This is an example diagram of an image processing device based on a car gateway according to an embodiment of the present application;

[0028] Figure 4It is a schematic diagram of the structure of a vehicle according to an embodiment of the present application. DETAILED DESCRIPTION

[0029] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0030] The following describes the image processing method, device, vehicle, storage medium and program product based on the automobile gateway of the embodiment of the present application with reference to the accompanying drawings. In view of the problem that the traditional centralized CNNs calculation method is difficult to meet the needs of the vehicle-mounted environment mentioned in the above background technology, the image data to be processed is divided into at least one sub-data, and the target sub-nodes of these sub-data are determined according to the gateway node and the vehicle-mounted network status, and the convolution and pooling operations are performed on the sub-data in the target sub-node respectively, and the processing results of the image data are generated according to the pooling results of all target sub-nodes. The automobile gateway node and dynamic routing mechanism are introduced, the computing efficiency and adaptability of the vehicle-mounted convolutional neural network are improved, and an efficient and real-time image processing and classification solution is provided. As a result, the problem that the traditional centralized CNNs calculation method is difficult to meet the needs of the vehicle-mounted environment with limited computing resources and high real-time requirements is solved.

[0031] Specifically, Figure 1 A flowchart of an image processing method based on a car gateway provided in an embodiment of the present application.

[0032] like Figure 1 As shown, the image processing method based on the automobile gateway includes the following steps:

[0033] In step S101 , the image data to be processed is divided into at least one sub-data.

[0034] Wherein, a sub-data is a part of data divided from the image data to be processed.

[0035] It is understandable that the embodiment of the present application can process the image data to be processed in blocks, dividing the image data to be processed into at least one sub-data.

[0036] For example, using the formula X i =Split(X), where X represents the image data to be processed; irepresents the sub-data after block division, i represents the index of the sub-data after block division, i is a positive integer from 1 to N, where N represents a positive integer and N is greater than or equal to 1; Split(﹒) represents the block processing of data. This formula can be used to divide the image data to be processed into N sub-data.

[0037] In step S102, the target sub-node of each sub-data is determined according to the gateway node and the vehicle network state.

[0038] Among them, the gateway node is responsible for the routing and protocol conversion of data packets. The gateway node can identify the ECU (Electronic Control Unit) that needs to communicate and determine which network or subnetwork the data should be sent to; the vehicle network status includes but is not limited to network load, error rate and other status.

[0039] It can be understood that the embodiments of the present application can detect the status of each gateway node and the status of the vehicle network, and find nodes that meet the sending conditions by analyzing the status of the two, such as nodes with lower loads, and determine the target sub-node of each sub-data by the status of the node.

[0040] In an embodiment of the present application, the target sub-node of each sub-data is determined according to the gateway node and the vehicle network status, including: extracting the data characteristic value of each sub-data; identifying the load status of the gateway node and the sub-node in the vehicle network status; selecting the target data flow path according to the data characteristic value load status, and determining the target sub-node of each sub-data according to the target data flow path.

[0041] The data characteristic value of the sub-data may represent the characteristics of the sub-data, which may include data type, priority, size, timeliness, etc.

[0042] It can be understood that the embodiment of the present application obtains the characteristics of the sub-data by extracting the data feature values ​​of each sub-data, and identifies the load status of the gateway node and the sub-node in the vehicle network status, selects the target data flow path according to the characteristics of each sub-data and the real-time status of the vehicle network, and determines the target sub-node of each sub-data through the target data flow path, ensuring that the computing task of each sub-node matches its current load and resource status, thereby realizing dynamic adjustment of the data flow path and the allocation of computing resources, and ensuring computing efficiency and performance optimization.

[0043] In step S103, convolution operations and pooling operations are respectively performed on the sub-data in the target sub-node to obtain pooling results, and a processing result of the image data is generated according to the pooling results of all the target sub-nodes.

[0044] Among them, convolution operation and pooling operation are two data processing operations. The convolution operation is mainly used to extract local features in the input image. The pooling operation is usually performed after the convolution operation, which is mainly used to reduce the spatial dimension of the feature map and reduce the amount of calculation.

[0045] It can be understood that the embodiments of the present application can perform convolution operations and pooling operations on each sub-data within the target sub-node, and generate processing results of the image data based on the pooling results of all target sub-nodes. By performing convolution calculations and pooling operations on different sub-nodes, parallel processing and resource optimization of computing tasks are achieved.

[0046] In the embodiment of the present application, the calculation formula of the convolution operation is:

[0047] Z ij =W ij *X i +b ij ,

[0048] Where i represents the index of the gateway node; j represents the index of the child node; Z ij Represents the result of the convolution operation; X i Indicates the multiple sub-data blocks into which the input image data is divided, X i The i in the block represents the sub-data block index after segmentation; W ij represents the convolution kernel; b ij represents the bias term; * represents the convolution operation.

[0049] In the embodiment of the present application, the calculation formula of the pooling operation is:

[0050] P ij =Pooling(Z ij ),

[0051] Among them, P ij Indicates the result of the pooling operation; Pooling(﹒) indicates the pooling operation; i indicates the index of the gateway node; j indicates the index of the child node.

[0052] It can be understood that the embodiment of the present application uses the above-mentioned convolution operation formula and pooling operation formula to perform convolution operation and pooling operation on the sub-data in the target sub-node in sequence, extract local features in the input image, reduce the spatial dimension of the feature map, and achieve optimized data processing.

[0053] In an embodiment of the present application, a processing result of image data is generated according to the pooling results of all target sub-nodes, including: inputting the pooling results of all target sub-nodes into a fully connected layer, and the fully connected layer outputs the processing result of the image data.

[0054] Among them, the nodes in the fully connected layer are connected to all the child nodes of the previous layer to form a fully connected network structure, which can integrate the features extracted by the previous layer. The fully connected layer can be implemented through the automotive gateway node, which uses the automotive gateway node to summarize the pooling results of each child node and perform the calculation of the fully connected layer in the gateway node to improve the efficiency and accuracy of the fully connected calculation.

[0055] It can be understood that the embodiment of the present application can integrate the pooling results of all target sub-nodes through the fully connected layer and output the processing results of the image data.

[0056] In the embodiment of the present application, the calculation formula of the fully connected layer is:

[0057]

[0058] Where O represents the result of the fully connected layer operation; FC(﹒) represents the fully connected layer operation; i represents the index of the gateway node; m represents the total number of gateway nodes participating in the fully connected calculation; P i is the pooling result of the i-th gateway node.

[0059] It can be understood that the embodiment of the present application performs a fully connected layer operation through the calculation formula of the above-mentioned fully connected layer to summarize the pooling results of each sub-node.

[0060] According to the image processing method based on automobile gateway proposed in the embodiment of the present application, the image data to be processed is divided into at least one sub-data, and the target sub-nodes of these sub-data are determined according to the gateway node and the vehicle network status, and convolution and pooling operations are performed on the sub-data in the target sub-node respectively. The processing result of the image data is generated according to the pooling results of all target sub-nodes. The automobile gateway node and dynamic routing mechanism are introduced, the computing efficiency and adaptability of the vehicle-mounted convolutional neural network are improved, and an efficient and real-time image processing and classification solution is provided.

[0061] The following describes the image processing method based on the automobile gateway proposed in the embodiment of the present application through a specific embodiment. Figure 2 As shown, the convolutional neural network calculation method based on the automobile gateway includes the following steps:

[0062] Step S201: data preprocessing.

[0063] The input image data (X) is processed in blocks:

[0064] X i =Split(X)

[0065] The image data to be processed is divided into blocks, and the image data to be processed is divided into at least one sub-data X i .

[0066] Step S202: Gateway node calculation.

[0067] Each automotive gateway node selects child nodes for convolution calculation and pooling operations based on a dynamic routing mechanism:

[0068]

[0069] Among them, Y i Indicates that each car gateway node selects a subnode for convolution calculation and pooling operation according to the dynamic routing mechanism, i represents the index of the gateway node; G ij represents the dynamic routing matrix; X i Indicates the multiple sub-data blocks into which the input image data is divided, X i The i in the block represents the sub-data block index after segmentation; W ij represents the convolution kernel; b ij represents the bias term; * represents the convolution operation.

[0070] Step S203: Dynamic routing mechanism.

[0071] The dynamic routing mechanism adjusts the data flow path according to the characteristics of the input data and the status of the vehicle network:

[0072] G ij =DynamicRouting(X i ,State)

[0073] Among them, G ij Represents the dynamic routing selection matrix; State represents the state of the vehicle network; DynamicRouting(﹒) represents the dynamic routing mechanism.

[0074] Step S204: convolution calculation.

[0075] The formula for convolution calculation is:

[0076] Z ij =W ij *X i +b ij

[0077] Through the above convolution formula, the sub-data is convolved in the target sub-node to extract the local features of the sub-data.

[0078] Step S205: Pooling operation.

[0079] The formula for the pooling operation is:

[0080] P ij =Pooling9Z ij )

[0081] Through the above pooling formula, the sub-data is pooled in the target sub-node to reduce the spatial dimension of the feature map and achieve optimized data processing.

[0082] Step S206: Full connection calculation.

[0083] The formula for full connection calculation is:

[0084]

[0085] Where O represents the result of the fully connected layer operation; FC(﹒) represents the fully connected layer operation; i represents the index of the gateway node; m represents the total number of gateway nodes participating in the fully connected calculation; P i is the pooling result of the i-th gateway node.

[0086] The fully connected operation is performed through the above-mentioned fully connected calculation formula, the pooling results of all target sub-nodes are input into the fully connected layer, and the fully connected layer outputs the processing results of the image data.

[0087] Through the above steps, the convolutional neural network calculation method based on the automobile gateway proposed in the embodiment of the present application can effectively improve the computing efficiency and real-time performance in the vehicle environment, and is suitable for application scenarios such as autonomous driving and vehicle monitoring.

[0088] For example, the embodiments of the present application are applied to the scenario of processing image data collected by a vehicle-mounted camera. Compared with the traditional centralized convolutional neural network, the method of the embodiments of the present application has obvious advantages in terms of fast calculation time and high resource utilization.

[0089] The embodiment of the present application can also be applied to the scenario of real-time processing of in-vehicle camera data to detect the driver's fatigue state. Similarly, compared with the traditional centralized convolutional neural network, the method of the embodiment of the present application has obvious advantages in fast calculation time and high resource utilization.

[0090] Next, the image processing device based on the automobile gateway proposed in accordance with the embodiment of the present application is described with reference to the accompanying drawings.

[0091] Figure 3 It is a block diagram of an image processing device based on a car gateway according to an embodiment of the present application.

[0092] like Figure 3 As shown, the image processing device 10 based on the automobile gateway includes: a division module 301 , a determination module 302 and a processing module 303 .

[0093] Among them, the division module 301 is used to divide the image data to be processed into at least one sub-data; the determination module 302 is used to determine the target sub-node of each sub-data according to the gateway node and the vehicle network status; the processing module 303 is used to perform convolution operations and pooling operations on the sub-data in the target sub-node to obtain pooling results, and generate the processing results of the image data according to the pooling results of all target sub-nodes.

[0094] In an embodiment of the present application, the determination module 302 is further used to: extract the data characteristic value of each sub-data; identify the load status of the gateway node and the sub-node in the vehicle network status; select the target data flow path according to the data characteristic value load status, and determine the target sub-node of each sub-data according to the target data flow path.

[0095] In the embodiment of the present application, the calculation formula of the convolution operation is:

[0096] Z ij =W ij *X i +b ij ,

[0097] Where i represents the index of the gateway node, j represents the index of the child node, and Z ij represents the result of the convolution operation, X i Indicates the multiple sub-data blocks into which the input image data is divided, X i The i in the block represents the sub-data block index after segmentation, W ij represents the convolution kernel, b ij represents the bias term and * represents the convolution operation.

[0098] In the embodiment of the present application, the calculation formula of the pooling operation is:

[0099] P ij =Pooling(Z ij ),

[0100] Among them, P ij Represents the result of the pooling operation. Pooling(﹒) represents the pooling operation, i represents the index of the gateway node, and j represents the index of the child node.

[0101] In the embodiment of the present application, the processing module 303 is further used to: input the pooling results of all target sub-nodes into the fully connected layer, and the fully connected layer outputs the processing results of the image data.

[0102] In the embodiment of the present application, the calculation formula of the fully connected layer is:

[0103]

[0104] Where O represents the result of the fully connected layer operation; FC(﹒) represents the fully connected layer operation; i represents the index of the gateway node; P i is the pooling result of the i-th gateway node; m represents the total number of gateway nodes participating in the full connection calculation.

[0105] It should be noted that the aforementioned explanation of the embodiment of the image processing method based on the automobile gateway is also applicable to the image processing device based on the automobile gateway of this embodiment, and will not be repeated here.

[0106] According to the image processing device based on the automobile gateway proposed in the embodiment of the present application, the division module, the determination module and the processing module collaborate to complete the division of the image data to be processed into at least one sub-data, determine the target sub-nodes of these sub-data, perform convolution and pooling operations on the sub-data in the target sub-nodes, and generate the processing results of the image data according to the pooling results of all target sub-nodes. The automobile gateway node and the dynamic routing mechanism are introduced, the computing efficiency and adaptability of the vehicle-mounted convolutional neural network are improved, and an efficient and real-time image processing and classification solution is provided.

[0107] Figure 4 A schematic diagram of the structure of a vehicle provided in an embodiment of the present application. The vehicle may include:

[0108] Memory 401 , processor 402 , and a computer program stored in the memory 401 and executable on the processor 402 .

[0109] When the processor 402 executes the program, the image processing method based on the automobile gateway provided in the above embodiment is implemented.

[0110] Furthermore, the vehicle also includes:

[0111] The communication interface 403 is used for communication between the memory 401 and the processor 402 .

[0112] The memory 401 is used to store computer programs that can be executed on the processor 402 .

[0113] The memory 401 may include a high-speed RAM (Random Access Memory) memory, and may also include a non-volatile memory, such as at least one disk memory.

[0114] If the memory 401, the processor 402 and the communication interface 403 are implemented independently, the communication interface 403, the memory 401 and the processor 402 can be connected to each other through a bus and communicate with each other. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0115] Optionally, in a specific implementation, if the memory 401, the processor 402 and the communication interface 403 are integrated on a chip, the memory 401, the processor 402 and the communication interface 403 can communicate with each other through an internal interface.

[0116] The processor 402 may be a CPU (Central Processing Unit), or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present application.

[0117] An embodiment of the present application also provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the above-mentioned image processing method based on the automobile gateway is implemented.

[0118] An embodiment of the present application also provides a computer program product, including a computer program or instructions, which, when executed, implements the above-mentioned image processing method based on the automobile gateway.

[0119] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms are not necessarily directed to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0120] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of the features. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0121] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, fragment or portion of code comprising one or N executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0122] It should be understood that the various parts of the present application can be implemented in hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, the steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array, a field programmable gate array, etc.

[0123] A person of ordinary skill in the art may understand that all or part of the steps carried by the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the above-mentioned program may be stored in a computer-readable storage medium, which, when executed, includes one of the steps of the method embodiment or a combination thereof.

[0124] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. An image processing method based on an automobile gateway, characterized in that: The following steps are involved: Dividing the image data to be processed into at least one sub-data; Determine the target sub-node for each sub-data according to the gateway node and the vehicle network status; Convolution operations and pooling operations are respectively performed on the sub-data in the target sub-node to obtain pooling results, and a processing result of the image data is generated according to the pooling results of all target sub-nodes.

2. The image processing method based on automobile gateway according to claim 1 is characterized in that: The step of determining the target sub-node of each sub-data according to the gateway node and the vehicle network state includes: Extracting data feature values ​​of each sub-data; Identifying the load status of the gateway node and the sub-nodes in the vehicle network status; A target data flow path is selected according to the data characteristic value load state, and a target sub-node of each sub-data is determined according to the target data flow path.

3. The image processing method based on automobile gateway according to claim 1 is characterized in that: The calculation formula of the convolution operation is: Z ij =W ij *X i +b ij , Where i represents the index of the gateway node, j represents the index of the child node, and Z ij represents the result of the convolution operation, X i Indicates the multiple sub-data blocks into which the input image data is divided, X i The i in the block represents the sub-data block index after segmentation, W ij represents the convolution kernel, b ij represents the bias term and * represents the convolution operation.

4. The image processing method based on automobile gateway according to claim 3 is characterized in that: The calculation formula of the pooling operation is: P ij =Pooling(Z ij ), Among them, P ij Represents the result of the pooling operation. Pooling(﹒) represents the pooling operation, i represents the index of the gateway node, and j represents the index of the child node.

5. The image processing method based on automobile gateway according to claim 1 is characterized in that: The processing result of generating the image data according to the pooling results of all target sub-nodes includes: The pooling results of all the target sub-nodes are input into a fully connected layer, and the fully connected layer outputs the processing result of the image data.

6. The image processing method based on automobile gateway according to claim 5 is characterized in that: The calculation formula of the fully connected layer is: Where O represents the result of the fully connected layer operation; FC(﹒) represents the fully connected layer operation; i represents the index of the gateway node; P i is the pooling result of the i-th gateway node; m represents the total number of gateway nodes participating in the full connection calculation.

7. An image processing device based on an automobile gateway, characterized in that: include: A division module, used for dividing the image data to be processed into at least one sub-data; A determination module, used to determine the target sub-node of each sub-data according to the gateway node and the vehicle network status; A processing module is used to perform convolution operations and pooling operations on the sub-data in the target sub-node to obtain pooling results, and generate processing results of the image data according to the pooling results of all target sub-nodes.

8. A vehicle, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the image processing method based on the automobile gateway as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program or instruction stored thereon, characterized in that: When the computer program or instruction is executed, the image processing method based on the automobile gateway according to any one of claims 1 to 6 is implemented.

10. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instruction is executed, the image processing method based on the automobile gateway as described in any one of claims 1 to 6 is implemented.

Images