Image recognition and analysis method and system for CAN bus data in automobile field

Abstract

The invention discloses an image recognition and analysis method and system for CAN bus data in the automobile field, and relates to the technical field of image recognition, and the method comprises the steps: synchronously collecting a CAN message data flow of a vehicle and a video data flow in the vehicle operation process in a collection period; according to the signal parameters, the CAN message data flow in the acquisition period is split into a plurality of signal data, a comparison signal set is determined based on all the signal data in the acquisition period, and the comparison signal set is stored in a comparison signal data source memory; according to the video parameters, dynamically identifying a video data stream in the acquisition period, determining a reference signal based on the dynamically identified data in the acquisition period, and storing the reference signal in an operation database; and analyzing the comparison signal set and the reference signal, and storing the comparison signal in the comparison signal set meeting a set condition into a signal database. According to the invention, automatic acquisition and analysis of CAN bus data in the automobile field are realized.

Description

Technical Field

[0001] The present application relates to the field of image recognition technology, and in particular to a method and system for image recognition and analysis of CAN bus data in the automotive field. Background Art

[0002] In the process of analyzing Controller Area Network (CAN) data, technicians face great challenges due to the complexity of data collection and the existence of a large amount of data. It becomes difficult to locate a specific signal because the required information needs to be filtered out from the huge amount of data, and even after finding the target signal, repeated calibration is required to ensure the accuracy of the data.

[0003] In this complex data collection process, technicians need to deal with the following problems: (1) Large amount of data: The amount of data transmitted by the CAN bus is huge and frequent. Technicians need to process a large number of data packets and filter out the required key information. (2) Difficulty in signal positioning: Due to the complexity and diversity of the data, it is difficult for technicians to quickly locate specific signals directly from massive data. It takes a lot of time and effort to screen and analyze. (3) Cumbersome calibration process: Even if a specific signal is found, since the data may be affected by noise, interference and other factors, technicians still need to perform repeated calibration and verification to ensure that the acquired signal is accurate and reliable.

[0004] During the CAN bus data analysis process, the data collection process is relatively complicated and the amount of data collected is complex. It is difficult for technicians to quickly locate the required specific signals from a large amount of data. After finding the specific signal in time, technicians still need to calibrate repeatedly. Summary of the invention

[0005] The purpose of this application is to provide an image recognition and analysis method and system for CAN bus data in the automotive field, so as to solve the problem that CAN bus data in the automotive field cannot be automatically collected and analyzed.

[0006] To achieve the above objectives, this application provides the following solutions.

[0007] In a first aspect, the present application provides an image recognition and analysis method for CAN bus data in the automotive field, comprising: During the collection cycle, the vehicle's CAN message data stream and the video data stream during vehicle operation are synchronously collected; the CAN message data stream includes: diagnostic data, broadcast data and diagnostic reply data; According to the signal parameters, the CAN message data stream within the acquisition period is split into multiple signal data, and a comparison signal set is determined based on all the signal data within the acquisition period, and the comparison signal set is stored in a comparison signal data source memory; Dynamically identifying a video data stream within a collection period according to video parameters, determining a reference signal based on the dynamically identified data within the collection period, and storing the reference signal in a computational database; The comparison signal set and the reference signal are analyzed, and the comparison signals in the comparison signal set that meet the set conditions are stored in a signal database.

[0008] Optionally, the CAN message data stream of the vehicle is collected, including: Collecting diagnostic data and broadcast data through a CAN analyzer; the CAN analyzer is connected to the OBD-II port of the vehicle; Or use a CAN analyzer to capture the diagnostic request message sent by the diagnostic device to the OBD simulator, generate a diagnostic reply based on the captured diagnostic request message, simulate sending diagnostic reply data with different numbers of bytes, and collect the diagnostic reply data.

[0009] Optionally, the video data stream during vehicle operation is collected, including: The video data stream during the vehicle operation is collected through the video acquisition device.

[0010] Optionally, before collecting the video data stream during the operation of the vehicle by using a video collection device, the method further includes: Calibrate the video acquisition device; Set video acquisition parameters; the video acquisition parameters include: acquisition quantity, acquisition interval, recognition area and recognition algorithm.

[0011] Optionally, the video data stream includes: vehicle status data, driving behavior data, instrument panel data and screen data of diagnostic equipment.

[0012] Optionally, the signal parameters include: signal data type, signal length, signal position, coefficient and offset.

[0013] Optionally, determining a comparison signal set based on all signal data in an acquisition period includes: Convert each signal data in the acquisition period into decimal, multiply by the coefficient and add the offset to obtain the corresponding converted signal data; The converted signal data within the acquisition period are formed into an array to obtain the comparison signal set.

[0014] Optionally, the video parameters include: target recognition parameters and the position, size and shape of the target in the video.

[0015] Optionally, analyzing the comparison signal set and the reference signal, and storing the comparison signals in the comparison signal set that meet the set conditions in a signal database, comprises: Determine any comparison signal in the comparison signal set as a current comparison signal; Draw a linear contrast image based on the current contrast signal, and draw a linear reference image based on the reference signal; Comparing and analyzing the linear contrast image and the linear reference image using a data normalization algorithm, a Pearson correlation coefficient algorithm, and a linear regression coefficient to obtain a Pearson correlation coefficient, a linear regression coefficient, and a linear regression offset corresponding to a current contrast signal; Sorting the comparison signals in ascending order of the Pearson correlation coefficient; Update the current contrast signal to any contrast signal in the top 20% of the contrast signals after sorting, and return "Use the data normalization algorithm, the Pearson correlation coefficient algorithm, and the linear regression coefficient to compare and analyze the linear contrast image and the linear reference image to obtain the Pearson correlation coefficient, the linear regression coefficient, and the linear regression offset corresponding to the current contrast signal"; The comparison signals, for which the difference between the linear regression coefficients calculated twice and the difference between the linear regression offsets are both within ±0.02, are stored in the signal database.

[0016] In a second aspect, the present application provides an image recognition and analysis system for CAN bus data in the automotive field, comprising: A data acquisition unit, used to synchronously acquire the vehicle's CAN message data stream and the video data stream during vehicle operation during an acquisition cycle; the CAN message data stream includes: diagnostic data, broadcast data, and diagnostic reply data; A comparison signal set determination unit, used for splitting the CAN message data stream within the acquisition period into a plurality of signal data according to the signal parameters, and determining a comparison signal set based on all the signal data within the acquisition period; A reference signal determination unit, configured to dynamically identify a video data stream within a collection period according to video parameters, and determine a reference signal based on the dynamically identified data within the collection period; The storage unit is used to analyze the comparison signal set and the reference signal, and store the comparison signals in the comparison signal set that meet the set conditions into a signal database.

[0017] According to the specific embodiments provided by the present application, the present application discloses the following technical effects: the present application discloses an image recognition and analysis method and system for CAN bus data in the automotive field. First, within an acquisition period, the CAN message data stream of the vehicle and the video data stream during the operation of the vehicle are synchronously acquired; secondly, according to signal parameters, the CAN message data stream within the acquisition period is split into multiple signal data, and a comparison signal set is determined based on all signal data within the acquisition period, and the comparison signal set is stored in a comparison signal data source memory; then, according to video parameters, the video data stream within the acquisition period is dynamically identified, and a reference signal is determined based on the dynamically identified data within the acquisition period, and the reference signal is stored in a calculation database; finally, the comparison signal set and the reference signal are analyzed, and the comparison signal set that meets the set conditions is stored in the signal database, thereby realizing the automatic acquisition and analysis of CAN bus data in the automotive field. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 A flowchart of an image recognition and analysis method for CAN bus data in the automotive field provided in one embodiment of the present application.

[0020] Figure 2 This is an architecture diagram of the image recognition and analysis method for CAN bus data in the automotive field.

[0021] Figure 3 This is a schematic diagram of the CAN message data stream acquisition process.

[0022] Figure 4 Schematic diagram of the video data stream acquisition process.

[0023] Figure 5 Schematic diagram of signal data calculation and analysis process. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0025] The purpose of this application is to provide an image recognition and analysis method and system for CAN bus data in the automotive field, aiming to realize automatic collection and analysis of CAN bus data in the automotive field.

[0026] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0027] In an exemplary embodiment, Figure 1 and Figure 2 As shown, the image recognition and analysis method of CAN bus data in the automotive field in this embodiment includes the following steps.

[0028] Step 1: During the collection period, the vehicle's CAN message data stream and the video data stream during vehicle operation are synchronously collected; the CAN message data stream includes: diagnostic data, broadcast data, and diagnostic response data.

[0029] Specifically, Figure 2 As shown in the figure, by using the OBD simulator and CAN analyzer to collect the vehicle's CAN message data stream, and combining it with the visual data (video data stream) collected by the external video acquisition device, the visual data is converted into digital signals and dynamically analyzed with the CAN message data, thereby realizing a new method to accurately find the target signal from the CAN message. This method not only improves the accuracy and efficiency of data analysis, but also opens up a new direction in the field of CAN data analysis.

[0030] During actual collection, the specific operation process includes the following steps.

[0031] Step 1: Message data collection.

[0032] In the process of message data collection, there are two methods: real vehicle data collection and data collection based on diagnostic equipment.

[0033] Real vehicle data collection: In real vehicle testing, by using a CAN analyzer device connected to the vehicle's OBD-II port, message data including diagnostic data and broadcast data can be collected in real time.

[0034] Data collection based on diagnostic equipment: In the process of collecting data based on diagnostic equipment, an OBD simulator device is required. First, the diagnostic device sends a diagnostic request to the OBD simulator, and uses the CAN analyzer to capture the diagnostic request message. Subsequently, the CAN analyzer generates a diagnostic response according to the ISO 14229 standard protocol based on the received diagnostic request message. During the diagnostic response process, the CAN analyzer simulates sending values ​​of different byte numbers (diagnostic response data), and the data generated by this series of operations is the required message data.

[0035] Step 2: Video data collection.

[0036] Video data acquisition is the use of high-definition cameras or other image acquisition devices to capture real-time visual data during vehicle operation, such as vehicle status, driving behavior, dashboards, diagnostic equipment screens, etc.).

[0037] Step 3: Parameter setting.

[0038] The parameter setting process is divided into two parts: signal parameters and video parameters. Signal parameters are used to parse the collected message data, including data signal length, signal position, coefficient, offset and other indicators. Video parameters are used to set the subsequent video signal extraction process, covering the identification of signal targets (such as vehicles, human behavior or screen devices) and parameters such as the position, size and shape of the signal in the video.

[0039] Step 4: Start collecting (start state).

[0040] Start acquisition means that the synchronous acquisition of CAN message data stream and video data stream is started at the same time. This step ensures the time alignment of the data. Click the Start acquisition button to start the synchronous acquisition of CAN message data stream and video data stream.

[0041] Step 5: Parsing, extraction, and calculation of CAN bus data.

[0042] According to the length of the data signal, the CAN message data stream is split into multiple signals, and these signal data are converted to decimal, multiplied by the coefficient and added with the offset. Subsequently, a new array is formed based on all the data values ​​in the acquisition cycle, and the values ​​of the array will be used for the continued operation in step 7.

[0043] Step 6: Identification and extraction of video data.

[0044] The data in the video data stream is dynamically identified according to the video parameters, and a new array is formed according to all the data values ​​in the acquisition cycle. The values ​​of the array will be used for the continued calculation in the seventh step.

[0045] Step 7: Data analysis, calculation and storage.

[0046] In this step, the array output in the sixth step is used as the reference signal, and the array output in the fifth step is used as the comparison signal to compare the similarity of the two arrays. If the similarity meets the similarity parameter setting value, the comparison signal information is stored in the signal database.

[0047] As an optional implementation, collecting the CAN message data stream of the vehicle includes: The diagnostic data and broadcast data are collected through a CAN analyzer; the CAN analyzer is connected to the OBD-II port of the vehicle.

[0048] Or use a CAN analyzer to capture the diagnostic request message sent by the diagnostic device to the OBD simulator, and generate a diagnostic reply based on the captured diagnostic request message, simulate sending diagnostic reply data with different numbers of bytes, and collect the diagnostic reply data.

[0049] As an optional implementation, collecting the video data stream during the vehicle operation process includes: The video data stream during the vehicle operation is collected through the video acquisition device.

[0050] As an optional implementation manner, before collecting the video data stream during the operation of the vehicle by the video acquisition device, the method further includes: Calibrate the video acquisition equipment.

[0051] Set the video acquisition parameters; video acquisition parameters include: acquisition quantity, acquisition interval, recognition area and recognition algorithm.

[0052] As an optional implementation, the video data stream includes: vehicle status data, driving behavior data, instrument panel data and screen data of the diagnostic device.

[0053] Step 2: According to the signal parameters, the CAN message data stream within the acquisition period is split into multiple signal data, and a comparison signal set is determined based on all the signal data within the acquisition period, and the comparison signal set is stored in the comparison signal data source memory.

[0054] As an optional implementation, the signal parameters include: signal data type, signal length, signal position, coefficient and offset.

[0055] As an optional implementation, determining a comparison signal set based on all signal data within a collection period includes: Each signal data in the acquisition period is converted into decimal, multiplied by a coefficient and added with an offset to obtain the corresponding converted signal data.

[0056] The converted signal data within the acquisition period are formed into an array to obtain a comparison signal set.

[0057] Specifically, in actual operation, Figure 3 As shown, the acquisition process of the CAN message data stream specifically includes the following steps.

[0058] S11: Select a data source.

[0059] There are two ways to select data sources: real vehicle data collection and data collection based on diagnostic equipment.

[0060] Real vehicle data collection: During real vehicle testing, the required message data including diagnostic data and broadcast data are collected in real time by using a CAN analyzer device connected to the vehicle's OBD-II port.

[0061] Data collection based on diagnostic equipment: In the process of collecting data based on diagnostic equipment, an OBD simulator device is required. First, the diagnostic device sends a diagnostic request to the OBD simulator, and uses the CAN analyzer to capture the diagnostic request message. Subsequently, the CAN analyzer generates a diagnostic response according to the ISO 14229 standard protocol based on the received diagnostic request message. During the diagnostic response process, the CAN analyzer simulates sending values ​​of different byte numbers (diagnostic response data), and the data generated by this series of operations is the required message data.

[0062] S12: (Message) data acquisition module.

[0063] The (message) data acquisition module will use the corresponding data processing and acquisition methods according to the selected data source. The (message) data acquisition module dynamically extracts the data into the system memory by reading the CAN message data provided by the CAN analyzer device.

[0064] S13: Set signal parameters.

[0065] Set the signal parameters required for the signal extraction process, including signal data type, signed / unsigned, coefficient, offset, extracted signal binary data bit length value (extraction judgment length value), etc., to facilitate use in subsequent steps.

[0066] S14: Generate a comparison data source.

[0067] In this step, firstly, the original message data collected by the (message) data acquisition module is cut into a series of original signal data of fixed length according to the extraction judgment length value parameter. Then, these original signal data are calculated according to other signal parameter parameters, and then the calculated data are placed in the system memory where the comparison data source is located.

[0068] Step 3: According to the video parameters, dynamically identify the video data stream within the acquisition period, determine the reference signal based on the dynamically identified data within the acquisition period, and store the reference signal in the operation database.

[0069] As an optional implementation, the video parameters include: target recognition parameters and the position, size and shape of the target in the video.

[0070] Specifically, in actual operation, Figure 4 As shown, the video data stream acquisition process specifically includes the following steps.

[0071] S21: Video acquisition equipment calibration.

[0072] Before starting to capture, it is necessary to calibrate the video capture equipment to ensure that the content to be captured is clear and included in the video range.

[0073] S22: Set video acquisition parameters.

[0074] Set the parameters required for video data stream processing. The parameter fields are described as follows.

[0075] Collection quantity: used to identify the number of data collection point information contained in the same data frame.

[0076] Interval: refers to the time interval for processing a data packet during the subsequent dynamic calculation process.

[0077] Identification area: represents the area where a collection point is located, drawn as a square. The database stores the coordinate pixel values ​​of the upper left corner and lower right corner of the video.

[0078] Recognition algorithm: It is used to process the recognition algorithm within the recognition area in the video data stream, including: digital recognition algorithm and state recognition algorithm.

[0079] S23: Determine whether it is a dynamic parameter.

[0080] Determine whether the data type field in the video acquisition parameters is dynamic data. If it is a fixed value, the video will process the first frame of video data and pass the processed data as a fixed value to the subsequent steps; if it is dynamic data, all data in the video data stream will be passed to the subsequent steps.

[0081] S24: Dynamic flow calculation.

[0082] This step is only executed when the data is dynamic. In this step, according to the video acquisition parameters, the continuously acquired video data stream is parsed frame by frame, and a value is obtained by calling the corresponding recognition algorithm from each frame of the video data stream. Then, the data is integrated into a data packet according to the interval parameters, and the data packet is stored in the operation database. Then, the data of the next data packet is collected.

[0083] S25: Reference signal data source.

[0084] The data source information after video signal processing is stored so that it can be used as a reference signal data source for subsequent operations.

[0085] Step 4: Analyze the comparison signal set and the reference signal, and store the comparison signals in the comparison signal set that meet the set conditions into the signal database.

[0086] As an optional implementation, step 4 includes: Step 41: Determine any comparison signal in the comparison signal set as the current comparison signal.

[0087] Step 42: Draw a linear contrast image based on the current contrast signal, and draw a linear reference image based on the reference signal.

[0088] Step 43: Use the data normalization algorithm, the Pearson correlation coefficient algorithm, and the linear regression coefficient to compare and analyze the linear contrast image and the linear reference image to obtain the Pearson correlation coefficient, the linear regression coefficient, and the linear regression offset corresponding to the current contrast signal.

[0089] Step 44: Sort the comparison signals in order of the Pearson correlation coefficient from small to large.

[0090] Step 45: Update the current comparison signal to any comparison signal in the top 20% of the comparison signals after sorting, and return to step 43.

[0091] Step 44: storing the comparison signal in which the difference between the linear regression coefficients calculated twice and the difference between the linear regression offsets are both within ±0.02 in the signal database.

[0092] Specifically, in actual operation, Figure 5 As shown, the process of signal data calculation and analysis specifically includes the following steps.

[0093] S31: Data collection.

[0094] During the data analysis process, the data acquisition function dynamically extracts data from the comparison signal data source and the reference signal source. Since both data are dynamic data, the amount of data will continue to increase over time.

[0095] When data is collected for the first time, all data from the start time to the end time are extracted based on the end time of the reference signal data, and these data are output to the interface to form two comparison curves, so as to facilitate better analysis and comparison. After outputting the dynamic curve, the extracted comparison signal data and reference signal data are output together to the subsequent calculation steps.

[0096] When collecting subsequent data, the end time of the last collection is used as the start time of this collection, and all data from the start time to the maximum end time of the reference signal source are collected, and these data are output to the interface and added to the interface ratio curve to facilitate better analysis and comparison. After outputting the dynamic curve, the extracted comparison signal data and reference signal data are output again to the subsequent calculation steps, forming a continuous collection process.

[0097] S32: Data comparison.

[0098] After obtaining the reference signal and the comparison signal set, a comparison analysis between the reference signal and the comparison signal is performed.

[0099] First, take a comparison signal from the comparison signal set and determine whether its first value is consistent with the first value of the reference signal. If not, the start time is the time of whichever occurs last; then determine whether the end time is consistent. If consistent, the end time is the time of whichever ends first. Next, plot the data between the start time and the end time into linear images. Finally, the two image data are compared and analyzed using the data standardization algorithm, the Pearson correlation coefficient algorithm, and the linear regression coefficient, respectively, to obtain the values ​​of the similarity, coefficient, and offset parameters. The general description of the above analysis method is as follows.

[0100] Data normalization algorithm: for data sets , the standardized dataset It can be calculated by the following formula: .

[0101] in, Dataset Dataset The mean of For the dataset The standard deviation of .

[0102] Pearson Correlation Coefficient Algorithm: Dataset and dataset Pearson correlation coefficient between It can be calculated by the following formula: .

[0103] in, For the dataset or dataset The amount of data in For the dataset The i-th data in; For the dataset The mean of For the dataset The i-th data in; For the dataset The mean of .

[0104] Linear regression coefficients: The general form of a linear regression model is , in this algorithm, the linear regression coefficient and the linear regression offset The calculation is based on the standardized data, and the specific formula is as follows: .

[0105] .

[0106] in, is the dependent variable; is the independent variable; For the dataset The standard deviation of For the dataset The standard deviation of For the dataset The mean of For the dataset The mean of .

[0107] S33: Dynamic calculation.

[0108] According to the comparison signal The value is calculated and the reverse sorting operation is performed on all the comparison signals in the comparison signal set. Then, the 20% comparison signals at the front end after sorting are extracted and compared with the reference signals extracted from the video data stream for a second data comparison calculation. If the data obtained after the second calculation is consistent with the data obtained in the first calculation, and If the difference is kept within ±0.02, then these data are the final signal data required and are stored in the signal database.

[0109] The method of the present application ensures the integrity and stability of the CAN signal data source and the video acquisition data source by storing them in different data source libraries, which helps to maintain and manage different types of data; at the same time, the OBD simulator and the CAN analyzer are connected to realize the collection of multi-source vehicle data, which can provide a richer data source; the video acquisition module does not limit the collection object, and can obtain vehicle screen data and instrument panel data, etc. This flexible design can meet different data requirements; data signal search and comparison are performed based on the data template extracted from the video data stream. Once the required signal comparison is completed, the system will start the continuous optimization and automatic verification function, and perform continuous verification according to the preset verification parameters, which increases system performance and improves the accuracy and efficiency of data analysis; once the verification is met, the verified data will be directly stored in the signal library, which improves the efficiency and reliability of data management and facilitates subsequent data processing, decision-making and system development; before signal comparison, the real-time data signal source and the real-time video signal source are specially collected, and the comparison signal source and the reference signal source are generated through processing. This meticulous processing improves the accuracy and reliability of the data; in the signal comparison process, the reference signal is used as a benchmark to find a change curve data set similar to the reference signal from the comparison signal source and bring it out, effectively screening out relevant data and improving the efficiency and accuracy of signal matching.

[0110] In an exemplary embodiment, a system for image recognition and analysis of CAN bus data in the automotive field is provided, comprising: The data acquisition unit is used to synchronously acquire the vehicle's CAN message data stream and the video data stream during vehicle operation during the acquisition cycle; the CAN message data stream includes: diagnostic data, broadcast data and diagnostic reply data.

[0111] The comparison signal set determination unit is used to split the CAN message data stream within the acquisition period into multiple signal data according to the signal parameters, and determine the comparison signal set based on all the signal data within the acquisition period.

[0112] The reference signal determination unit is used to dynamically identify the video data stream within the acquisition period according to the video parameters, and determine the reference signal based on the dynamically identified data within the acquisition period.

[0113] The storage unit is used to analyze the comparison signal set and the reference signal, and store the comparison signals in the comparison signal set that meet the set conditions into the signal database.

[0114] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0115] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for image recognition and analysis of CAN bus data in the automotive field, characterized in that: The image recognition and analysis method of CAN bus data in the automotive field includes: During the collection cycle, the vehicle's CAN message data stream and the video data stream during vehicle operation are synchronously collected; the CAN message data stream includes: diagnostic data, broadcast data and diagnostic reply data; According to the signal parameters, the CAN message data stream within the acquisition period is split into multiple signal data, and a comparison signal set is determined based on all the signal data within the acquisition period, and the comparison signal set is stored in a comparison signal data source memory; Dynamically identifying a video data stream within a collection period according to video parameters, determining a reference signal based on the dynamically identified data within the collection period, and storing the reference signal in a computational database; The comparison signal set and the reference signal are analyzed, and the comparison signals in the comparison signal set that meet the set conditions are stored in a signal database.

2. The image recognition and analysis method of CAN bus data in the automotive field according to claim 1 is characterized in that: Collect the vehicle's CAN message data stream, including: Collecting diagnostic data and broadcast data through a CAN analyzer; the CAN analyzer is connected to the OBD-II port of the vehicle; Or use a CAN analyzer to capture the diagnostic request message sent by the diagnostic device to the OBD simulator, generate a diagnostic reply based on the captured diagnostic request message, simulate sending diagnostic reply data with different numbers of bytes, and collect the diagnostic reply data.

3. The image recognition and analysis method of CAN bus data in the automotive field according to claim 1, characterized in that: Collect video data streams during vehicle operation, including: The video data stream during the vehicle operation is collected through the video acquisition device.

4. The image recognition and analysis method of CAN bus data in the automotive field according to claim 1 is characterized in that: Before the video data stream of the vehicle operation process is collected by the video acquisition device, it also includes: Calibrate the video acquisition device; Set video acquisition parameters; the video acquisition parameters include: acquisition quantity, acquisition interval, recognition area and recognition algorithm.

5. The image recognition and analysis method of CAN bus data in the automotive field according to claim 1, characterized in that: The video data stream includes: vehicle status data, driving behavior data, instrument panel data and screen data of diagnostic equipment.

6. The image recognition and analysis method of CAN bus data in the automotive field according to claim 1 is characterized in that: The signal parameters include: signal data type, signal length, signal position, coefficient and offset.

7. The image recognition and analysis method of CAN bus data in the automotive field according to claim 6 is characterized in that: The comparison signal set is determined based on all signal data within the acquisition period, including: Convert each signal data in the acquisition period into decimal, multiply by the coefficient and add the offset to obtain the corresponding converted signal data; The converted signal data within the acquisition period are formed into an array to obtain the comparison signal set.

8. The image recognition and analysis method of CAN bus data in the automotive field according to claim 1, characterized in that: The video parameters include: target recognition parameters and the position, size and shape of the target in the video.

9. The image recognition and analysis method of CAN bus data in the automotive field according to claim 1, characterized in that: Analyzing the comparison signal set and the reference signal, and storing the comparison signals in the comparison signal set that meet the set conditions in a signal database, includes: Determine any comparison signal in the comparison signal set as a current comparison signal; Draw a linear contrast image based on the current contrast signal, and draw a linear reference image based on the reference signal; Comparing and analyzing the linear contrast image and the linear reference image using a data normalization algorithm, a Pearson correlation coefficient algorithm, and a linear regression coefficient to obtain a Pearson correlation coefficient, a linear regression coefficient, and a linear regression offset corresponding to a current contrast signal; Sorting the comparison signals in ascending order of the Pearson correlation coefficient; Update the current contrast signal to any contrast signal in the top 20% of the contrast signals after sorting, and return "Use the data normalization algorithm, the Pearson correlation coefficient algorithm, and the linear regression coefficient to compare and analyze the linear contrast image and the linear reference image to obtain the Pearson correlation coefficient, the linear regression coefficient, and the linear regression offset corresponding to the current contrast signal"; The comparison signals, for which the difference between the linear regression coefficients calculated twice and the difference between the linear regression offsets are both within ±0.02, are stored in the signal database.

10. An image recognition and analysis system for CAN bus data in the automotive field, characterized in that: The image recognition and analysis system for CAN bus data in the automotive field includes: A data acquisition unit, used to synchronously acquire the vehicle's CAN message data stream and the video data stream during vehicle operation during an acquisition cycle; the CAN message data stream includes: diagnostic data, broadcast data, and diagnostic reply data; A comparison signal set determination unit, used for splitting the CAN message data stream within the acquisition period into a plurality of signal data according to the signal parameters, and determining a comparison signal set based on all the signal data within the acquisition period; A reference signal determination unit, configured to dynamically identify a video data stream within a collection period according to video parameters, and determine a reference signal based on the dynamically identified data within the collection period; The storage unit is used to analyze the comparison signal set and the reference signal, and store the comparison signals in the comparison signal set that meet the set conditions into a signal database.

Images