Automobile collision test report generation method based on artificial intelligence

Abstract

The invention discloses an artificial intelligence-based automobile collision test report generation method, which comprises the following steps of: firstly, obtaining collision test files in different formats, and adopting an existing chart recognition method, such as a chart recognition method based on chart characteristics or a chart recognition method based on machine learning, to generate a collision test report; and automatically identifying the chart area in the obtained collision test file, and determining the chart area in each collision test file. Therefore, the operation of manually extracting the chart from the test file is reduced, misoperation or omission caused by manual operation is avoided, and the report generation efficiency and accuracy are improved. Then, the independent curve graph file is intercepted from the graph area, the curve graph file is stored according to the number and the category, and when the test report is generated, the curve graph file can be automatically inserted into the corresponding position of the test report. The steps of manually analyzing data are reduced, misoperation or omission caused by manual operation is avoided, and the report generation efficiency and accuracy are further improved.

Description

Technical Field

[0001] The present invention relates to the technical field of information retrieval, and in particular to a method for generating an automobile collision test report based on artificial intelligence. Background Art

[0002] Automobile collision test is an important means to evaluate the safety performance of automobiles. By simulating different collision scenarios, the structure and performance of the automobile are tested, so as to evaluate its ability to protect the safety of passengers. At present, the generation process of automobile collision test reports mostly relies on manual operations, including data extraction, analysis, report generation and other steps. Traditional methods usually require manual extraction of test data, such as dummy data, collision parameters, etc. Due to the large amount of data in the collision test files, manual extraction and analysis may result in omissions or errors, affecting the accuracy of the report. Moreover, report preparation involves a large number of data matching and image insertion operations, and manual operations take a long time. Therefore, there is an urgent need for an efficient and intelligent analysis method to support data processing and result analysis of automobile collision tests. Summary of the invention

[0003] In view of the shortcomings of the existing technology, the present invention proposes a method for generating a car collision test report based on artificial intelligence. The accuracy and generation efficiency of the report can be greatly improved. The specific technical solution is as follows:

[0004] In a first aspect, a method for generating a car collision test report based on artificial intelligence is provided. In a first implementable manner of the first aspect, the method comprises:

[0005] Obtain collision test files in different formats, and detect chart areas in the collision test files;

[0006] A curve graph is cut out from the chart area and inserted into a corresponding position of the test report.

[0007] In combination with the first implementable manner of the first aspect, in a second implementable manner of the first aspect, detecting a chart area in the collision test file includes:

[0008] When the collision test file is an unreadable file, the collision test file is converted into a readable file.

[0009] In combination with the first implementable manner of the first aspect, in a third implementable manner of the first aspect, detecting a chart area in the collision test file includes:

[0010] The page layout and keywords of each page of the collision test file are analyzed respectively, and the page including the curve graph is determined according to the page layout and keywords.

[0011] In combination with the first implementable manner of the first aspect, in a fourth implementable manner of the first aspect, detecting a chart area in the collision test file includes:

[0012] An edge detection algorithm is used to detect the chart area in the page including the curve chart.

[0013] In combination with the first implementable manner of the first aspect, in a fifth implementable manner of the first aspect, extracting a curve graph from the chart area includes:

[0014] A partitioning algorithm is used to extract a curve graph from the graph area.

[0015] In combination with the first implementable manner of the first aspect, in a sixth implementable manner of the first aspect, further comprising:

[0016] The key data points in the curve graph are extracted, converted into numerical data and inserted into corresponding positions of the test report.

[0017] In a second aspect, a method for generating a car collision test report based on artificial intelligence is provided, comprising:

[0018] Obtain collision test files in different formats, and respectively extract key data from the collision test files in different formats using corresponding parsing methods;

[0019] Retrieve the corresponding report template according to the set test type, and retrieve relevant key data from all extracted key data to fill in the corresponding position of the report template to generate a test report;

[0020] Using an artificial intelligence-based automobile collision test report generation method as described in any one of the first to sixth possible implementations of the first aspect, the curve chart in the collision test file is inserted into the corresponding position of the test report.

[0021] In a third aspect, a method for generating a car collision test report based on artificial intelligence is provided, comprising:

[0022] Obtaining a collision test photo of the car, and identifying the vehicle part corresponding to the collision test photo through an image recognition algorithm;

[0023] Determine the reporting step corresponding to the collision test photo according to the corresponding vehicle part, and insert the collision test photo into the corresponding position of the test report according to the reporting step;

[0024] Using an artificial intelligence-based automobile collision test report generation method as described in any one of the first to sixth possible implementations of the first aspect, the curve chart in the collision test file is inserted into the corresponding position of the test report.

[0025] In a fourth aspect, a method for generating a car collision test report based on artificial intelligence is provided, comprising:

[0026] Generate a test report using the method for generating a car collision test report based on artificial intelligence as described in the third aspect;

[0027] Obtaining a collision test photo of the car, and identifying the vehicle part corresponding to the collision test photo through an image recognition algorithm;

[0028] The reporting step corresponding to the collision test photo is determined according to the corresponding vehicle part, and the collision test photo is inserted into the corresponding position of the test report according to the reporting step.

[0029] In a fifth aspect, a method for generating a car collision test report based on artificial intelligence is provided, comprising:

[0030] Obtain collision test files in different formats, and respectively extract key data from the collision test files in different formats using corresponding parsing methods;

[0031] Retrieve the corresponding report template according to the set test type, and retrieve relevant key data from all extracted key data to fill in the corresponding position of the report template to generate a test report;

[0032] Obtaining a collision test photo of the car, and identifying the vehicle part corresponding to the collision test photo through an image recognition algorithm;

[0033] The reporting step corresponding to the collision test photo is determined according to the corresponding vehicle part, and the collision test photo is inserted into the corresponding position of the test report according to the reporting step.

[0034] Beneficial effects: The method for generating a car collision test report based on artificial intelligence of the present invention automatically obtains collision test files and automatically detects the chart area in the collision test files in batches, thereby reducing the manual extraction of charts from the test files and avoiding errors or omissions caused by manual operations. This improves the efficiency and accuracy of report generation. Moreover, by automatically cutting out a curve chart from the chart area and inserting it into the corresponding position of the test report, the steps of manual data analysis are reduced and errors or omissions caused by manual operations are avoided, further improving the efficiency and accuracy of report generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the specific implementation of the present invention, the following will briefly introduce the drawings required for use in the specific implementation. In all the drawings, each element or part is not necessarily drawn according to the actual scale.

[0036] Figure 1A flowchart of a method for generating a car collision test report based on artificial intelligence provided by an embodiment of the present invention;

[0037] Figure 2 A flowchart of a method for generating a car collision test report based on artificial intelligence provided by another embodiment of the present invention;

[0038] Figure 3 A flowchart of a method for generating a car collision test report based on artificial intelligence provided by another embodiment of the present invention;

[0039] Figure 4 A flowchart of a method for generating a car collision test report based on artificial intelligence provided by another embodiment of the present invention;

[0040] Figure 5 A flowchart of a method for generating an automobile collision test report based on artificial intelligence is provided as another embodiment of the present invention. DETAILED DESCRIPTION

[0041] The following embodiments of the technical solution of the present invention are described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and are therefore only used as examples, and cannot be used to limit the protection scope of the present invention.

[0042] Embodiment 1

[0043] like Figure 1 A flow chart of a method for generating a car collision test report based on artificial intelligence is shown, and the generating method comprises:

[0044] Step 1: Obtain collision test files in different formats, and detect the chart area in the collision test files;

[0045] Step 2: extract a curve chart from the chart area and insert the curve chart into a corresponding position of the test report.

[0046] Specifically, firstly, collision test files of different formats can be obtained, and existing chart recognition methods, such as chart recognition methods based on chart features, can be used to automatically identify the chart area in the obtained collision test files and determine the chart area in each collision test file. This reduces the manual operation of extracting charts from test files and avoids manual misoperation or omissions. This improves the efficiency and accuracy of report generation.

[0047] With the rapid development of artificial intelligence technology, especially the breakthrough of deep learning in the field of image processing and data analysis, new opportunities have been provided for the field of automotive safety. Deep learning can not only process large-scale data efficiently, but also shows great potential in aspects such as image classification, opening up a new direction for the intelligent processing of collision test data. Therefore, a chart recognition method based on machine learning can also be used to automatically identify the chart area in the acquired collision test file and determine the chart area in each collision test file.

[0048] Then, you can extract independent curve chart files from the chart area and save them according to numbers and categories. When generating a test report, you can automatically insert the curve chart files into the corresponding positions of the test report. This reduces the steps of manual data analysis and avoids manual errors or omissions, further improving the efficiency and accuracy of report generation.

[0049] In this embodiment, optionally, in step 1, detecting the chart area in the collision test file includes:

[0050] When the collision test file is an unreadable file, the collision test file is converted into a readable file.

[0051] Specifically, the file formats corresponding to different automobile collision test data may be different. For example, detailed collision test data, such as dummy damage data, collision parameters, etc., are often recorded in DataTable files in TXT format. Collision curve graphs, reports and other image data are often recorded in PDF files. The basic information of the vehicle, test number, test parameters and other information are often recorded in the test inspection information table in Excel format.

[0052] Some of the test files are unreadable, so before detecting the chart area, it is necessary to use appropriate methods to convert the unreadable collision test files into readable files for subsequent chart area recognition. For example, the PyMuPDF library can be used to scan PDF files page by page and convert the content of each page of the test file into a readable object.

[0053] In this embodiment, optionally, detecting the chart area in the collision test file includes:

[0054] The page layout and keywords of each page of the collision test file are analyzed respectively, and the page including the curve graph is determined according to the page layout and keywords.

[0055] Specifically, in order to quickly identify all the curves in the collision test file and further improve the efficiency of test report generation, before identifying the chart area, the page layout and keywords of each page of the test file can be analyzed to filter out the pages containing curves. Then, by identifying the chart area of ​​the page containing the curve, the amount of data for chart area identification can be reduced, further improving the efficiency of report generation.

[0056] Specifically, the page can be divided into multiple areas, such as header, body, footer, etc. Then the content type of each area can be determined by analyzing the physical layout of the page, such as area coordinates, size, etc. In addition, a trained convolutional neural network model can be used for analysis to automatically distinguish between text, pictures, tables and other areas.

[0057] In this embodiment, optionally, detecting the chart area in the collision test file includes:

[0058] An edge detection algorithm is used to detect the chart area in the page including the curve chart.

[0059] Specifically, when identifying the chart area in the page, an edge detection algorithm can be used to identify the chart area in the page. Specifically, the page is first grayed and denoised. Then, the Canny algorithm is used to detect the edge of the image. Finally, the potential chart area is found through the existing contour extraction algorithm, and the chart area is screened by geometric features such as area and shape.

[0060] In this embodiment, optionally, extracting a curve chart from the chart area includes:

[0061] A partitioning algorithm is used to extract a curve graph from the graph area.

[0062] Specifically, a partitioning algorithm, such as the Louvain algorithm, the Girvan-Newman algorithm, etc., may be used to partition the chart area and separate the curve chart from other page elements to accurately extract the curve chart.

[0063] For example, the Louvain algorithm divides the regions in the image by maximizing the modularity, which can effectively separate the chart region in the image, such as the curve graph, from other page elements. The specific steps include: first, constructing the similarity relationship between the image regions into a graph. Then, applying the Louvain algorithm to divide the graph into communities. Then, the curve graph in the divided region is identified by the image features. Finally, the chart region is accurately extracted based on the identified curve graph.

[0064] In this embodiment, optionally, it further includes:

[0065] The key data points in the curve graph are extracted, converted into numerical data and inserted into corresponding positions of the test report.

[0066] Specifically, after intercepting the curve graph, optical character recognition technology can be used to extract key data points in the curve graph, and the key data points can be converted into numerical data and inserted into the corresponding positions of the test report to ensure the data integrity of the test report and improve the accuracy of the report.

[0067] Embodiment 2

[0068] like Figure 2 A flow chart of a method for generating a car collision test report based on artificial intelligence is shown, and the generating method comprises:

[0069] Step S1, obtaining collision test files in different formats, and respectively extracting key data from the collision test files in different formats using corresponding parsing methods;

[0070] Step S2: retrieve the corresponding report template according to the set test type, retrieve relevant key data from all the extracted key data and fill them into the corresponding position of the report template to generate a test report;

[0071] Step S3: using the above-mentioned method for generating a car collision test report based on artificial intelligence, insert the curve graph in the collision test file into the corresponding position of the test report.

[0072] Specifically, first, collision test files in different formats can be obtained from various data sources, such as obtaining DataTable files containing detailed collision test data, obtaining PDF files containing collision data graphs, reports and other image data, and obtaining test inspection information tables in Excel format that record basic vehicle information, test numbers, test parameters and other information.

[0073] After obtaining collision test files in various formats, key data can be extracted from them using parsing methods that match the various collision test files.

[0074] For example, use the Python regular expression module to extract key data fields from the DataTable file, such as the dummy's head acceleration, chest compression displacement, etc.

[0075] The PDF file is parsed through the PyMuPDF library, the file is scanned page by page to extract key pages, and optical character recognition technology is used to extract text data in the chart to ensure data integrity.

[0076] The vehicle information and test parameters in the Excel file are read through the OpenPyXL library. The parsed data can be automatically stored in the system database for subsequent test report generation. This can reduce the manual extraction of key data from the test file and avoid manual errors or omissions. This improves the efficiency and accuracy of report generation.

[0077] Then, according to the set test type, such as frontal collision or side collision, the corresponding report template can be selected. The report template predefines the positions of various types of data and fields and the order of image insertion. The system will automatically retrieve the corresponding key data from the system database and fill it into the corresponding position of the report template based on the extracted data information. The report generated in this way has a unified format and structure, meeting the requirements of industry standards. Finally, the method described in Example 1 can be used to insert the curve chart involved in the collision test into the test report, reducing the manual extraction and parsing of charts from the test file, and further improving the efficiency and accuracy of report generation.

[0078] Embodiment 3

[0079] like Figure 3 A flow chart of a method for generating a car collision test report based on artificial intelligence is shown, and the generating method comprises:

[0080] Step D1, obtaining a collision test photo of the car, and identifying the vehicle part corresponding to the collision test photo through an image recognition algorithm;

[0081] Step D2, determining the reporting step corresponding to the collision test photo according to the corresponding vehicle part, and inserting the collision test photo into the corresponding position of the test report according to the reporting step;

[0082] Step D3: using the artificial intelligence-based automobile collision test report generation method described in the first embodiment, insert the curve graph in the collision test file into the corresponding position of the test report.

[0083] Specifically, in a vehicle collision test, photos of the vehicle are usually taken from various angles before and after the collision. In order to insert these photos into the test report. First, existing image recognition algorithms, such as convolutional neural networks, can be used to detect the vehicle landmark features corresponding to the acquired test photos, and then determine the vehicle part corresponding to the photo. For example, by detecting the vehicle headlights and rearview mirrors in the test photos, it can be determined whether the test photo belongs to the front or side of the vehicle.

[0084] Then, based on the report steps set in the report template, the report steps corresponding to each test photo can be determined according to the vehicle part corresponding to the test photo, and each test photo can be inserted into the corresponding position of the test report according to the report steps. In this way, the operation of manually analyzing the test photos can be reduced, and the misoperation or omission caused by manual operation can be avoided. Thus, the efficiency and accuracy of report generation can be improved.

[0085] Finally, the method described in the first embodiment can be used to insert the curve graph involved in the collision test into the test report, thereby reducing the manual extraction and parsing of graphs from the test file and further improving the efficiency and accuracy of report generation.

[0086] Embodiment 4

[0087] like Figure 4 A flow chart of a method for generating a car collision test report based on artificial intelligence is shown, and the generating method comprises:

[0088] Step A1: Generate a test report using the method for generating a car collision test report based on artificial intelligence described in the second embodiment;

[0089] Step A2: obtaining a collision test photo of the car, and identifying the vehicle part corresponding to the collision test photo through an image recognition algorithm;

[0090] Step A3: determine the reporting step corresponding to the collision test photo according to the corresponding vehicle part, and insert the collision test photo into the corresponding position of the test report according to the reporting step.

[0091] Specifically, first, the generation method described in the second embodiment can be used to insert key data and curve graphs extracted from collision test files of different formats into the test report, so as to reduce the manual operation of extracting key data and curve graphs from the test file, thereby avoiding manual operation errors or omissions, and improving the efficiency and accuracy of report generation.

[0092] Then, the existing image recognition algorithm can be used to detect the vehicle sign features corresponding to the acquired test photos, and then determine the vehicle part corresponding to the photo. Finally, based on the order of image insertion of different vehicle parts set in the report template, the report step corresponding to each test photo is determined according to the vehicle part corresponding to the test photo, and each test photo is inserted into the corresponding position of the test report according to the report step. In this way, the operation of manually analyzing the test photos can be reduced, and the misoperation or omission caused by manual operation can be avoided, further improving the efficiency and accuracy of report generation.

[0093] Embodiment 5

[0094] like Figure 5A flow chart of a method for generating a car collision test report based on artificial intelligence is shown, and the generating method comprises:

[0095] Step C1, obtaining collision test files in different formats, and respectively extracting key data from the collision test files in different formats using corresponding parsing methods;

[0096] Step C2: retrieve the corresponding report template according to the set test type, retrieve relevant key data from all the extracted key data and fill them into the corresponding position of the report template to generate a test report;

[0097] Step C3, obtaining a collision test photo of the car, and identifying the vehicle part corresponding to the collision test photo through an image recognition algorithm;

[0098] Step C4: determine the reporting step corresponding to the collision test photo according to the corresponding vehicle part, and insert the collision test photo into the corresponding position of the test report according to the reporting step.

[0099] Specifically, first, collision test files of different formats can be obtained from various data sources, and key data can be extracted from them using parsing methods that match various collision test files. Then, the corresponding report template can be selected according to the set test type. The report template predefines the positions of various types of data and fields and the order of image insertion. The system will automatically retrieve the corresponding key data from the system database and fill it into the corresponding position of the report template based on the extracted data information. The report generated in this way has a unified format and structure, meeting the requirements of industry standards.

[0100] After that, the existing image recognition algorithm can be used to detect the vehicle sign features corresponding to the acquired test photos, and then determine the vehicle part corresponding to the photo. Finally, based on the order of image insertion of different vehicle parts set in the report template, the report steps corresponding to each test photo are determined according to the vehicle part corresponding to the test photo, and each test photo is inserted into the corresponding position of the test report according to the report steps. In this way, the operation of manually analyzing test photos can be reduced, and errors or omissions caused by manual operation can be avoided, further improving the efficiency and accuracy of report generation.

[0101] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A method for generating a car collision test report based on artificial intelligence, characterized in that: include: Obtain collision test files in different formats, and detect chart areas in the collision test files; A curve graph is cut out from the chart area and inserted into a corresponding position of the test report.

2. The method for generating a car collision test report based on artificial intelligence according to claim 1, characterized in that: Detecting a chart area in the collision test file includes: When the collision test file is an unreadable file, the collision test file is converted into a readable file.

3. The method for generating a car collision test report based on artificial intelligence according to claim 1, characterized in that: Detecting a chart area in the collision test file includes: The page layout and keywords of each page of the collision test file are analyzed respectively, and the page including the curve graph is determined according to the page layout and keywords.

4. The method for generating a car collision test report based on artificial intelligence according to claim 1, characterized in that: Detecting a chart area in the collision test file includes: An edge detection algorithm is used to detect the chart area in the page including the curve chart.

5. The method for generating a car collision test report based on artificial intelligence according to claim 1, characterized in that: Extracting a curve graph from the graph area includes: A partitioning algorithm is used to extract a curve graph from the graph area.

6. The method for generating a car collision test report based on artificial intelligence according to claim 1, characterized in that: Also includes: The key data points in the curve graph are extracted, converted into numerical data and inserted into corresponding positions of the test report.

7. A method for generating a car collision test report based on artificial intelligence, characterized in that: include: Obtain collision test files in different formats, and respectively extract key data from the collision test files in different formats using corresponding parsing methods; Retrieve the corresponding report template according to the set test type, and retrieve relevant key data from all extracted key data to fill in the corresponding position of the report template to generate a test report; By adopting the method for generating a car collision test report based on artificial intelligence as described in any one of claims 1 to 6, the curve chart in the collision test file is inserted into the corresponding position of the test report.

8. A method for generating a car collision test report based on artificial intelligence, characterized in that: include: Obtaining a collision test photo of the car, and identifying the vehicle part corresponding to the collision test photo through an image recognition algorithm; Determine the reporting step corresponding to the collision test photo according to the corresponding vehicle part, and insert the collision test photo into the corresponding position of the test report according to the reporting step; By adopting the method for generating a car collision test report based on artificial intelligence as described in any one of claims 1 to 6, the curve chart in the collision test file is inserted into the corresponding position of the test report.

9. A method for generating a car collision test report based on artificial intelligence, characterized in that: include: Generating a test report using the method for generating a car collision test report based on artificial intelligence as claimed in claim 7; Obtaining a collision test photo of the car, and identifying the vehicle part corresponding to the collision test photo through an image recognition algorithm; The reporting step corresponding to the collision test photo is determined according to the corresponding vehicle part, and the collision test photo is inserted into the corresponding position of the test report according to the reporting step.

10. A method for generating a car collision test report based on artificial intelligence, characterized in that: include: Obtain collision test files in different formats, and respectively extract key data from the collision test files in different formats using corresponding parsing methods; Retrieve the corresponding report template according to the set test type, and retrieve relevant key data from all extracted key data to fill in the corresponding position of the report template to generate a test report; Obtaining a collision test photo of the car, and identifying the vehicle part corresponding to the collision test photo through an image recognition algorithm; The reporting step corresponding to the collision test photo is determined according to the corresponding vehicle part, and the collision test photo is inserted into the corresponding position of the test report according to the reporting step.

Images