Subway vehicle sealing joint detection method and system and storage medium
By constructing and training a deep convolutional neural network detection model, the sealed joints of subway vehicles are automatically detected, which solves the problems of low detection efficiency, accuracy and reliability in the existing technology, and achieves efficient and accurate sealed joint detection.
Patent Information
- Application Number
- CN202411299188.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-18
- Publication Date
- 2025-02-07
AI Technical Summary
In the prior art, the detection efficiency, detection accuracy and fault detection rate of subway vehicle sealed joint detection methods are relatively low, the yield detection rate is troublesome, the post-tracement and verification are inconvenient, and the detection reliability and safety are relatively low.
By collecting standard image data of sealed joints and image data to be detected, pre-processing, constructing and training a deep convolutional neural network detection model for detection, and automatic detection is achieved.
It improves detection efficiency, detection accuracy and fault detection rate, and does not require secondary testing and troubleshooting, which facilitates post-tracement and verification, and improves the reliability and safety of detection.
Smart Images
Figure CN119399513A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of nondestructive testing of rail transit, and in particular to a method, system and storage medium for testing sealed joints of subway vehicles. Background Art
[0002] Driven by national policies, the rail transit industry has entered a period of rapid development. Subway vehicles in some cities are undergoing concentrated overhaul and maintenance, and maintenance efficiency and quality have become key factors in ensuring efficient operations and safe services.
[0003] During the inspection and maintenance process, subway vehicle sealing joints are difficult to inspect due to the variety of joint forms and complex operating environment. Existing rail transit sealing joint testing mainly relies on traditional methods such as manual inspection to perform millimeter-level defect detection, which has low inspection efficiency, detection accuracy and fault detection rate, and requires subsequent secondary inspection methods such as rain shower to check the yield rate of sealing joints. During the inspection process, it is difficult for manual inspection to record, save and share inspection data in a timely manner, which is not conducive to subsequent backtracking and verification.
[0004] In addition, the existing detection process is prone to cause certain damage to the detection object, reducing the reliability and safety of the detection, which is particularly evident in long-term and large-scale maintenance. Summary of the invention
[0005] To this end, the technical problem to be solved by the present invention is to overcome the problems of low detection efficiency, detection accuracy and fault detection rate of subway vehicle sealed joint detection methods in the prior art, troublesome inspection of yield rate, inconvenient subsequent backtracking and verification, and low detection reliability and safety. A subway vehicle sealed joint detection method, system and storage medium are provided, which have high detection efficiency, detection accuracy and fault detection rate, do not need to conduct secondary inspection to check yield rate, are convenient for subsequent backtracking and verification, and have high detection reliability and safety.
[0006] In a first aspect, in order to solve the above technical problems, the present invention provides a subway vehicle sealing joint detection method, comprising:
[0007] Step S1: collecting standard image data of different forms of sealed joints to obtain standard boundary data and standard surface data; collecting image data of the sealed joint to be detected to obtain boundary data to be matched and surface data to be matched;
[0008] Step S2: preprocessing the standard boundary data, the standard surface data, the boundary data to be matched and the surface data to be matched;
[0009] Step S3: constructing and training a detection model based on the preprocessed standard boundary data and the preprocessed standard surface data;
[0010] Step S4: input the preprocessed boundary data to be matched and the preprocessed surface data to be matched into the detection model to obtain a first detection result of the boundary data to be matched and a second detection result of the surface data to be matched; when the first detection result and the second detection result are both qualified, the sealing joint to be detected is qualified.
[0011] In one embodiment of the present invention, the method for preprocessing the standard boundary data, the standard surface data, the boundary data to be matched and the surface data to be matched includes:
[0012] In MATLAB, first, the median grayscale value of the neighborhood pixels of the data to be processed is calculated; then, the difference between the median grayscale value and the pixel grayscale value of the data to be processed is calculated; then, the filter window is determined based on the difference; finally, the pixels of the data to be processed are subjected to median filtering based on the filter window to remove random noise in the data to be processed.
[0013] In one embodiment of the present invention, the method for preprocessing the standard boundary data, the standard surface data, the boundary data to be matched and the surface data to be matched includes:
[0014] The first step is to divide the data to be processed into multiple overlapping or non-overlapping local areas, and the shape of the local areas is a regular pattern; the second step is to extract pixels in each of the local areas, calculate the grayscale image histogram based on the extracted pixels, normalize the grayscale image histogram to obtain a cumulative distribution function, equalize the extracted pixels based on the cumulative distribution function, and update the local area based on the equalized pixels; the third step is to merge all the updated local areas into complete image data to achieve a balance on the lighting impact.
[0015] In one embodiment of the present invention, the shape of the local area is a square.
[0016] In one embodiment of the present invention, the method for preprocessing the standard boundary data, the standard surface data, the boundary data to be matched and the surface data to be matched includes:
[0017] An opening operation is performed on the data to be processed to smooth the data at the edge of the image, and then a closing operation is performed to fill the data in the holes inside the image.
[0018] In one embodiment of the present invention, the detection model is a deep convolutional neural network model, which includes an input layer, a convolution layer, a pooling layer, a classification decision layer, a fully connected layer and an output layer.
[0019] In one embodiment of the present invention, the standard image data of the sealed joints of different forms and the image data of the sealed joints to be inspected are collected by a CCD camera.
[0020] In a second aspect, in order to solve the above technical problems, the present invention also provides a subway vehicle sealing joint detection system, comprising:
[0021] The acquisition module acquires standard image data of sealing joints of different forms to obtain standard boundary data and standard surface data; acquires image data of sealing joints to be detected to obtain boundary data to be matched and surface data to be matched;
[0022] A preprocessing module, preprocessing the standard boundary data, the standard surface data, the boundary data to be matched and the surface data to be matched;
[0023] A model building and training module, which builds and trains a detection model based on the preprocessed standard boundary data and the preprocessed standard surface data;
[0024] The detection module inputs the preprocessed boundary data to be matched and the preprocessed surface data to be matched into the detection model to obtain a first detection result of the boundary data to be matched and a second detection result of the surface data to be matched; when the first detection result and the second detection result are both qualified, the sealing joint to be detected is qualified.
[0025] In one embodiment of the present invention, an alarm and feedback module is further included. When the detection result of the sealed joint to be detected is unqualified, the alarm and feedback module sends out an acoustic signal or a light signal and generates an unqualified report.
[0026] In a third aspect, in order to solve the above-mentioned technical problem, the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the subway vehicle sealing joint detection method are implemented.
[0027] The above technical solution of the present invention has the following beneficial effects compared with the prior art:
[0028] The present invention provides a subway vehicle sealing joint detection method, system and storage medium, which have high detection efficiency, detection accuracy and fault detection rate, and do not require secondary detection to check the yield rate, which is convenient for subsequent backtracking and verification, and has high detection reliability and safety. BRIEF DESCRIPTION OF THE DRAWINGS
[0029] In order to make the contents of the present invention more clearly understood, the present invention is further described in detail below based on specific embodiments of the present invention in conjunction with the accompanying drawings.
[0030] Figure 1 This is a flow chart of the steps of a subway vehicle sealing joint detection method in a preferred embodiment of the present invention;
[0031] Figure 2 The module flow chart of a subway vehicle sealed joint detection system in an embodiment of the present invention. DETAILED DESCRIPTION
[0032] It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0033] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments.
[0034] In addition, in the embodiments of the present application, descriptions such as "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by this application.
[0035] During the inspection and maintenance process, subway vehicles have many types of sealing joints and a complex operating environment, which makes inspection and maintenance difficult. Existing rail transit sealing joint testing mainly relies on traditional methods such as manual inspection to perform millimeter-level defect detection, which has low inspection efficiency, detection accuracy and fault detection rate, and requires subsequent secondary inspection methods such as rain showers to check the yield rate of sealing joints. During the inspection process, it is difficult for manual inspection to record, save and share inspection data in a timely manner, which is not conducive to subsequent backtracking and verification. In addition, the existing inspection process is prone to cause certain damage to the inspection object, reducing the reliability and safety of the inspection, which is particularly evident in long-term and large-scale inspections.
[0036] To this end, embodiments of the present application provide a subway vehicle sealed joint detection method, system and storage medium.
[0037] Embodiment 1
[0038] This embodiment provides a subway vehicle sealing joint detection method, please refer to Figure 1 As shown, including:
[0039] Step S1: collecting standard image data of different forms of sealed joints to obtain standard boundary data and standard surface data; collecting image data of the sealed joint to be detected to obtain boundary data to be matched and surface data to be matched;
[0040] Step S2: preprocessing the standard boundary data, the standard surface data, the boundary data to be matched and the surface data to be matched;
[0041] Step S3: constructing and training a detection model based on the preprocessed standard boundary data and the preprocessed standard surface data;
[0042] Step S4: input the preprocessed boundary data to be matched and the preprocessed surface data to be matched into the detection model to obtain a first detection result of the boundary data to be matched and a second detection result of the surface data to be matched; when the first detection result and the second detection result are both qualified, the sealing joint to be detected is qualified.
[0043] The present embodiment provides a subway vehicle sealing joint detection method, which (1) detects the to-be-matched boundary data and to-be-matched surface data of the to-be-detected sealing joint through a detection model, and has high detection efficiency, detection accuracy and fault detection rate; (2) does not require secondary detection to check the yield rate, which facilitates subsequent backtracking and verification; (3) does not cause damage to the detection object, and has high detection reliability and safety.
[0044] Next, a subway vehicle sealing joint detection method provided in this embodiment is described in detail:
[0045] Step S1
[0046] Specifically, standard image data of sealed joints of different forms are collected to obtain standard boundary data and standard surface data; image data of sealed joints to be inspected are collected to obtain boundary data to be matched and surface data to be matched.
[0047] Optionally, the standard image data of the sealed joints of different forms and the image data of the sealed joints to be inspected are collected by a CCD camera.
[0048] The CCD camera is a camera that uses a charge coupled device (CCD for short) as an image sensor.
[0049] Optionally, an LED light source is used to provide a uniform and stable lighting environment to ensure that the surface details of the sealed joint to be inspected are clearly visible, thereby reducing the impact of shadows and reflections on image quality.
[0050] Step S2
[0051] Specifically, the standard boundary data, the standard surface data, the boundary data to be matched, and the surface data to be matched are preprocessed.
[0052] Preprocessing 1: In order to eliminate random noise in the data to be processed, in MATLAB, first, the median grayscale value of the neighborhood pixels of the data to be processed is calculated; then, the difference between the median grayscale value and the pixel grayscale value of the data to be processed is calculated; then, the filtering window is determined based on the difference; finally, the pixels of the data to be processed are subjected to median filtering based on the filtering window.
[0053] Performing random noise removal on the processed data can preserve more edge details of the image.
[0054] Preprocessing 2: In order to eliminate the influence of uneven illumination, the first step is to divide the data to be processed into multiple overlapping or non-overlapping local areas, and the shape of the local areas is a regular pattern; the second step is to extract pixels in each of the local areas, calculate the grayscale image histogram based on the extracted pixels, normalize the grayscale image histogram to obtain the cumulative distribution function, equalize the extracted pixels based on the cumulative distribution function, and update the local area based on the equalized pixels; the third step is to merge all the updated local areas into complete image data.
[0055] Preferably, the shape of the local area is a square, which can reduce obvious boundaries between different local areas.
[0056] Eliminating the uneven illumination of the processed data can balance the overall brightness of the image.
[0057] Preprocessing 3: Perform an opening operation on the data to be processed to smooth the data at the edge of the image, and then perform a closing operation to fill the data in the holes inside the image; this processing can remove irrelevant background information and improve image clarity and contrast.
[0058] Optionally, the above three preprocessing operations may be performed only one or in combination.
[0059] Step S3
[0060] Specifically, a detection model is constructed and trained based on the preprocessed standard boundary data and the preprocessed standard surface data.
[0061] Preferably, the detection model is a deep convolutional neural network (CNN) model.
[0062] Specifically, the deep convolutional neural network model includes an input layer, a convolution layer, a pooling layer, a classification decision layer, a fully connected layer and an output layer.
[0063] The classification decision layer identifies and classifies the types of the sealing joints to be detected.
[0064] Optionally, the nonlinear activation function of the CNN model is a ReLU activation function.
[0065] Optionally, the convolution layer uses a convolution kernel of size 3*3.
[0066] Optionally, the convolutional layer is configured as three layers.
[0067] Optionally, the pooling layer is a maximum pooling layer, and the maximum pooling layer uses a pooling kernel of a size of 2*2.
[0068] Specifically, sample data is obtained based on the preprocessed standard boundary data and the preprocessed standard surface data, and the sample data is labeled; features are extracted based on the labeled sample data; and the CNN model is trained and learned based on the extracted sample data features.
[0069] Step S4
[0070] Specifically, the preprocessed boundary data to be matched and the preprocessed surface data to be matched are input into the detection model to obtain a first detection result of the boundary data to be matched and a second detection result of the surface data to be matched; when the first detection result and the second detection result are both qualified, the sealing joint to be detected is qualified.
[0071] For example, taking 400 sealed joints to be inspected as inspection objects, the accuracy rate is lower than 70% when manual inspection is adopted, while the accuracy rate can be higher than 90% when the subway vehicle sealed joint inspection method provided in this embodiment is adopted for inspection.
[0072] Embodiment 2
[0073] This embodiment provides a subway vehicle sealing joint detection system, please refer to Figure 2 As shown, including:
[0074] The acquisition module acquires standard image data of sealing joints of different forms to obtain standard boundary data and standard surface data; acquires image data of sealing joints to be detected to obtain boundary data to be matched and surface data to be matched;
[0075] A preprocessing module, preprocessing the standard boundary data, the standard surface data, the boundary data to be matched and the surface data to be matched;
[0076] A model building and training module, which builds and trains a detection model based on the preprocessed standard boundary data and the preprocessed standard surface data;
[0077] The detection module inputs the preprocessed boundary data to be matched and the preprocessed surface data to be matched into the detection model to obtain a first detection result of the boundary data to be matched and a second detection result of the surface data to be matched; when the first detection result and the second detection result are both qualified, the sealing joint to be detected is qualified.
[0078] Optionally, an alarm and feedback module is also included. When the test result of the sealed joint to be tested is unqualified, the alarm and feedback module sends out a sound signal or a light signal and generates an unqualified report; the unqualified report includes the unqualified type, abnormal location, and severity.
[0079] Optionally, it also includes a data storage and management module, which performs remote backup and cross-regional redundant storage of data through blockchain cloud storage.
[0080] For further introduction to the subway vehicle sealed joint detection system provided in this embodiment, please refer to the first embodiment, and this embodiment will not be described in detail here.
[0081] The subway vehicle sealed joint detection system provided in this embodiment has the same beneficial effects as the above-mentioned subway vehicle sealed joint detection method.
[0082] Embodiment 3
[0083] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the subway vehicle sealing joint detection method are implemented.
[0084] For an introduction to a computer-readable storage medium provided in this embodiment, please refer to Embodiment 1, and this embodiment will not be described in detail here.
[0085] The computer-readable storage medium provided in this embodiment has the same beneficial effects as the above-mentioned subway vehicle sealed joint detection method.
[0086] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
[0087] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0088] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0089] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0090] Obviously, the above embodiments are merely examples for clear explanation and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from these are still within the protection scope of the invention.
Claims
1. A subway vehicle sealing joint detection method, characterized in that: include: Step S1: collecting standard image data of different forms of sealed joints to obtain standard boundary data and standard surface data; collecting image data of the sealed joint to be detected to obtain boundary data to be matched and surface data to be matched; Step S2: preprocessing the standard boundary data, the standard surface data, the boundary data to be matched and the surface data to be matched; Step S3: constructing and training a detection model based on the preprocessed standard boundary data and the preprocessed standard surface data; Step S4: input the preprocessed boundary data to be matched and the preprocessed surface data to be matched into the detection model to obtain a first detection result of the boundary data to be matched and a second detection result of the surface data to be matched; when the first detection result and the second detection result are both qualified, the sealing joint to be detected is qualified.
2. A subway vehicle sealing joint detection method according to claim 1, characterized in that: The method for preprocessing the standard boundary data, the standard surface data, the boundary data to be matched and the surface data to be matched comprises: In MATLAB, first, the median grayscale value of the neighborhood pixels of the data to be processed is calculated; then, the difference between the median grayscale value and the pixel grayscale value of the data to be processed is calculated; then, the filter window is determined based on the difference; finally, the pixels of the data to be processed are subjected to median filtering based on the filter window to remove random noise in the data to be processed.
3. A subway vehicle sealing joint detection method according to claim 1, characterized in that: The method for preprocessing the standard boundary data, the standard surface data, the boundary data to be matched and the surface data to be matched comprises: The first step is to divide the data to be processed into multiple overlapping or non-overlapping local areas, and the shape of the local areas is a regular pattern; the second step is to extract pixels in each of the local areas, calculate the grayscale image histogram based on the extracted pixels, normalize the grayscale image histogram to obtain a cumulative distribution function, equalize the extracted pixels based on the cumulative distribution function, and update the local area based on the equalized pixels; the third step is to merge all the updated local areas into complete image data to balance the illumination impact.
4. A subway vehicle sealing joint detection method according to claim 3, characterized in that: The shape of the local area is a square.
5. A subway vehicle sealing joint detection method according to claim 1, characterized in that: The method for preprocessing the standard boundary data, the standard surface data, the boundary data to be matched and the surface data to be matched comprises: An opening operation is performed on the data to be processed to smooth the data at the edge of the image, and then a closing operation is performed to fill the data in the holes inside the image.
6. A subway vehicle sealing joint detection method according to claim 1, characterized in that: The detection model is a deep convolutional neural network model, which includes an input layer, a convolution layer, a pooling layer, a classification decision layer, a fully connected layer and an output layer.
7. A subway vehicle sealing joint detection method according to claim 1, characterized in that: The standard image data of the sealing joints of different forms and the image data of the sealing joints to be inspected are collected by a CCD camera.
8. A subway vehicle sealing joint detection system, characterized in that: include: An acquisition module collects standard image data of different forms of sealing joints to obtain standard boundary data and standard surface data; Collect image data of the sealed joint to be detected, and obtain boundary data to be matched and surface data to be matched; A preprocessing module, preprocessing the standard boundary data, the standard surface data, the boundary data to be matched and the surface data to be matched; A model building and training module, which builds and trains a detection model based on the preprocessed standard boundary data and the preprocessed standard surface data; The detection module inputs the preprocessed boundary data to be matched and the preprocessed surface data to be matched into the detection model to obtain a first detection result of the boundary data to be matched and a second detection result of the surface data to be matched; when the first detection result and the second detection result are both qualified, the sealing joint to be detected is qualified.
9. A subway vehicle sealing joint detection system according to claim 8, characterized in that: It also includes an alarm and feedback module. When the detection result of the sealed joint to be detected is unqualified, the alarm and feedback module sends out an acoustic signal or a light signal and generates an unqualified report.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the subway vehicle sealed joint detection method according to any one of claims 1 to 7 are implemented.