Track traffic fastener model identification system and method

Abstract

The invention discloses a rail transit fastener model identification system and method, and relates to the technical field of fastener overhaul, an initial image of a rail is acquired through an image acquisition module, and an image preprocessing module is used for preprocessing to obtain a reference image; the image analysis module inputs the reference image into a preset model for processing, completes segmentation of the fastener assembly and the steel rail assembly, and obtains segmented images; the positioning module is used for screening the images of part of fastener assemblies and the images of the rail bottom of the steel rail and restoring the images to the reference images; the feature acquisition module is used for acquiring a reference adjustment line based on the segmented image of the rail bottom, acquiring feature parameters of the partial components based on the segmented images of the partial components, and correcting by using the reference adjustment line to obtain reference feature parameters; and the model identification module is used for matching the reference characteristic parameters of the screening component with the data in the constructed database and outputting a corresponding fastener model. According to the invention, efficient identification of the models of the fasteners can be realized, and the identification accuracy is also guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of fastener maintenance, and in particular to a rail transit fastener model identification system and method. Background Art

[0002] In rail transit systems, fasteners are the core components connecting rails and sleepers. Their main functions are to fix the position of rails, maintain track gauge stability, transfer wheel-rail loads to the roadbed, and cushion the vibration impact of train operation. Fastener inspection is to regularly monitor the status of fasteners through technical means. Its significance and role are directly related to the safety of track structure, train operation safety and operational efficiency.

[0003] Traditionally, the dimensional information of each rail fastener is measured manually. Inspectors need to have extensive prior knowledge to distinguish different types of rail fasteners based on the detected dimensional information and determine whether each rail fastener is safe. However, this measurement method consumes a lot of human resources, has low measurement efficiency, and has low measurement accuracy. With the development of 3D point cloud technology, 3D point cloud-related technologies have been used to inspect fasteners. Specifically, the fastener location is found by comparing the detected point cloud image with the prior point cloud image. The 3D point cloud data is then analyzed to obtain the 3D geometric information of the fastener, thereby completing the identification of the fastener model. However, this method has the following disadvantages: the processing of massive 3D point clouds is complex, making it difficult to ensure real-time detection; 3D point cloud data is highly restricted by the environment and equipment, and its point cloud data quality is unstable; when separating individual fasteners from massive point clouds, it is more restricted by complex scenes, and its segmentation efficiency is low; when extracting 3D fastener features from 3D point cloud data, the 3D features are unstable due to the influence of the fastener status, resulting in insufficient robustness.

[0004] In summary, a rail transit fastener model identification system and method are urgently needed to solve the above problems. Summary of the Invention

[0005] In view of this, the present application provides a rail transit fastener model identification system and method to address the deficiencies in the prior art.

[0006] A first aspect of the present application provides a rail transit fastener model identification system, comprising: An image acquisition module, which uses multiple image acquisition devices installed on the rail inspection vehicle to capture initial images of the fastener assembly and the rail assembly; An image preprocessing module, which preprocesses the initial image to obtain a reference image; an image analysis module, inputting the reference image into a preset model for processing, obtaining a segmented image of each component in the fastener assembly and storing the segmented image in a first preset set, and obtaining a segmented image of each component in the rail assembly and storing the segmented image in a second preset set; a positioning module, which selects the segmented images of multiple components in the first preset set and the segmented image of the rail bottom in the second preset set, and positions and restores them to the reference image; a feature acquisition module that acquires a reference adjustment line based on a segmented image of the rail bottom in the reference image; acquires feature parameters of corresponding components based on segmented images of all screening components in the reference image; and modifies the feature parameters of the screening components based on the reference adjustment line to obtain reference feature parameters of the screening components; The model identification module constructs a database for storing fastener size parameters and fastener models, matches the reference feature parameters of all screening components with the size parameters of all fasteners in the database, outputs the corresponding fastener model, and completes the identification of the fastener model.

[0007] In a possible implementation of the first aspect, preprocessing the initial image to obtain a reference image includes: De-noising and geometric correction are performed on the initial image to obtain the reference image.

[0008] In a possible implementation manner of the first aspect, the preset model is a SINet model.

[0009] In a possible implementation of the first aspect, inputting the reference image into a preset model for processing includes: The reference images are input into the SINet model in batches for processing, and the images processed by the SINet model are standardized to obtain a segmented image of each component in the fastener assembly and a segmented image of each component in the rail assembly.

[0010] In a possible implementation of the first aspect, screening the segmented images of the plurality of components in the first preset set includes: Acquire segmented images of all components in a single fastener component in the first preset set to obtain a plurality of first segmented images; Based on prior knowledge, all components in a single fastener assembly are divided into symmetrical components denoted as first components and asymmetrical components denoted as second components; Determine the axisymmetric line in all first components, and divide the edge areas on both sides of the axisymmetric line into multiple verification areas of equal size, calculate whether the first segmented images of the verification areas on both sides of the axisymmetric line overlap, and if so, determine that the verification of the group of verification areas is successful; If the ratio between the number of groups of successfully verified corresponding verification areas in a single first component and the number of groups of all verification areas exceeds a set value, the first segmented image corresponding to the first component is used as the first screening result; Construct a standard image library for all asymmetric components, obtain the first segmented image of the second component and traverse the corresponding images in the standard image library, and if the first segmented image of the second component overlaps with an image in the standard image library, use the first segmented image corresponding to the second component as the second screening result; Based on the first screening result and the second screening result, segmented images of a plurality of components are screened in the first preset set.

[0011] In a possible implementation of the first aspect, obtaining the reference adjustment line based on the segmented image of the rail bottom in the reference image includes: Obtaining a segmented image of the rail bottom in the reference image, recorded as a second segmented image; Obtaining an edge region image of the second segmented image close to the fastener component, and determining whether there is an occlusion region in the edge region image; if not, fitting the contour line of the second segmented image close to the fastener component to obtain the baseline adjustment line; and if so, processing the second segmented image to obtain the baseline adjustment line; Processing the second segmented image includes: Taking the top vertices on both sides of the second segmented image as references, determining a plurality of feature points on both sides of the second segmented image at a preset spacing, wherein the feature points on both sides of the second segmented image that are equally spaced from the top vertices constitute a group of feature points, and obtaining a perpendicular line for each group of feature points; It is determined whether the perpendicular lines of all feature points coincide with each other. If so, the perpendicular lines are used as the reference adjustment line. If not, the perpendicular line with the highest coincidence degree is selected as the reference adjustment line.

[0012] In a possible implementation of the first aspect, obtaining characteristic parameters of corresponding components based on segmented images of all screened components in the reference image includes: Obtain any screening component as a first screening component, obtain contour lines on both sides of the segmented image by the first screening component, and perform straight line fitting to obtain a set of reference straight lines related to the first screening component; The distance between the set of reference straight lines is calculated as a characteristic parameter of the first screening component.

[0013] In a possible implementation of the first aspect, the modifying of characteristic parameters of the screening component based on the reference adjustment line includes: Obtaining a set of reference straight lines of the first screening component and corresponding characteristic parameters as first characteristic parameters; Adjusting the corresponding two reference straight lines based on maintaining an equal distance between the reference adjustment line and any reference straight line in the set of reference straight lines; The distance between the adjusted set of reference straight lines is calculated as the reference characteristic parameter of the first screening component.

[0014] In a possible implementation of the first aspect, matching the reference characteristic parameters of all screening components with the size parameters of all fasteners in the database and outputting the corresponding fastener models includes: Matching the reference characteristic parameters of all the screened components with the dimensional parameters of any corresponding fastener component in the database until a match is successful; The fastener models of the fasteners that are successfully matched are output as the fastener models corresponding to all the filtered components.

[0015] A second aspect of the present application provides a method for identifying a rail transit fastener model, comprising: Using multiple image acquisition devices installed on the rail inspection vehicle, an initial image of the fastener assembly and the rail assembly is acquired; Preprocessing the initial image to obtain a reference image; Inputting the reference image into a preset model for processing, obtaining a segmented image of each component in the fastener assembly and storing it in a first preset set, and obtaining a segmented image of each component in the rail assembly and storing it in a second preset set; Filtering the segmented images of multiple components in the first preset set and filtering the segmented images of the rail bottom in the second preset set, and positioning and restoring both to the reference image; Based on the segmented image of the rail bottom in the reference image, a reference adjustment line is obtained; based on the segmented images of all screening components in the reference image, characteristic parameters of corresponding components are obtained, and based on the reference adjustment line, the characteristic parameters of the screening components are corrected to obtain reference characteristic parameters of the screening components; A database is constructed to store the corresponding fastener size parameters and fastener models, the reference feature parameters of all screening components are matched with the size parameters of all fasteners in the database, the corresponding fastener models are output, and the fastener model identification is completed.

[0016] The beneficial effects are as follows: the present invention discloses a rail transit fastener model recognition system and method, wherein the image acquisition module collects the initial images of the fastener assembly and the rail assembly, and uses the image preprocessing module to preprocess and obtain a reference image; the image analysis module inputs the reference image into a preset model for processing, completes the segmentation of each fastener assembly and each rail assembly, and obtains a corresponding segmented image; the positioning module filters the images of some components in the segmented image of the fastener assembly and filters the images of the rail in the segmented image of the rail assembly, and both are positioned and restored to the reference image; the feature acquisition module obtains a reference adjustment line based on the segmented image of the rail bottom in the reference image, and then obtains the feature parameters of the screening assembly based on the segmented image of the screening assembly in the reference image, and at the same time uses the reference adjustment line to correct the feature parameters to obtain the reference feature parameters; the model recognition module uses the reference feature parameters of the screening assembly to match the data in the database storing the fastener size parameters and the fastener model, outputs the corresponding fastener model, and completes the fastener model recognition. The present invention can not only realize the efficient recognition of the fastener model, but also ensure the recognition accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without any creative work.

[0018] Figure 1 This is a schematic diagram of the composition of a rail transit fastener model identification system provided in an embodiment of the present application; Figure 2 This is an example diagram of a feature acquisition module in a rail transit fastener model identification system provided in an embodiment of the present application; Figure 3 This is a flow chart of a method for identifying rail transit fastener models provided in an embodiment of the present application. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0020] In this application, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or apparatus comprising the element.

[0021] Example 1 In the existing technology, there has been a method of using 3D point cloud related technologies to detect fasteners. Specifically, the fastener position is found by comparing the detection point cloud image with the prior point cloud image, and then the 3D geometric information of the fastener is obtained by analyzing the 3D point cloud data, thereby completing the identification of the fastener model. However, this method has the following disadvantages: the processing of massive 3D point clouds is very complex, which makes it difficult to ensure the real-time detection; the 3D point cloud data is highly restricted by the environment and equipment, and its point cloud data quality is unstable; when separating individual fasteners from massive point clouds, it is more restricted by complex scenes and its segmentation efficiency is low; when extracting the 3D features of fasteners from 3D point cloud data, the 3D features will be unstable due to the influence of the fastener status, resulting in insufficient robustness.

[0022] Therefore, this application provides a rail transit fastener model identification system, such as Figure 1 As shown, including: An image acquisition module, which uses multiple image acquisition devices installed on the rail inspection vehicle to capture initial images of the fastener assembly and the rail assembly; An image preprocessing module, which preprocesses the initial image to obtain a reference image; an image analysis module, inputting the reference image into a preset model for processing, obtaining a segmented image of each component in the fastener assembly and storing the segmented image in a first preset set, and obtaining a segmented image of each component in the rail assembly and storing the segmented image in a second preset set; a positioning module, which selects the segmented images of multiple components in the first preset set and the segmented image of the rail bottom in the second preset set, and positions and restores them to the reference image; a feature acquisition module that acquires a reference adjustment line based on a segmented image of the rail bottom in the reference image; acquires feature parameters of corresponding components based on segmented images of all screening components in the reference image; and modifies the feature parameters of the screening components based on the reference adjustment line to obtain reference feature parameters of the screening components; The model identification module constructs a database for storing fastener size parameters and fastener models, matches the reference feature parameters of all screening components with the size parameters of all fasteners in the database, outputs the corresponding fastener model, and completes the identification of the fastener model.

[0023] This embodiment provides a rail transit fastener model identification system, specifically comprising: Image acquisition module: Multiple image acquisition devices installed on the track inspection vehicle capture initial images of the fastener assembly and rail assembly. The specific image acquisition equipment consists of a light source, an industrial high-speed 3D camera, and a fill light. When capturing images, the light source and fill light are used for lighting. Four high-speed industrial cameras capture initial images of both sides of the two rails (image resolution is 3200*1024). It should be noted that fasteners in rail transit are intermediate parts connecting the rails and sleepers. They mainly include components such as clamping parts, pads, and gauge adjustment parts. The specific components vary depending on the sleeper type. The rail assembly mainly includes the rail body, as well as rail connection components and rail anti-climbing components. In this embodiment, the image acquisition device mainly captures images of the rail bottom and some rail connection components.

[0024] The image preprocessing module preprocesses the initial image to obtain a reference image; the specific preprocessing methods are denoising and geometric correction processing. Denoising can be processed by Gaussian filtering, median filtering, adaptive filtering, etc., which is not specifically limited in this embodiment; and geometric correction processing is to eliminate the image geometric distortion caused by lens distortion or installation deviation, unify the image coordinate system, and can use a distortion correction algorithm or a perspective transformation algorithm, which is not specifically limited in this embodiment.

[0025] The image analysis module feeds the reference image into the SINet model for processing. The SINet model is a deep learning-based neural network model widely used in image recognition, segmentation, and enhancement. The core concept for image segmentation is to simulate visual perception mechanisms, combine attention mechanisms with multi-scale feature fusion, accurately locate the target area, and generate pixel-level segmentation masks. After the reference image is processed by the SINet model, segmented images of each component in the fastener assembly are obtained and stored in a first preset set, and segmented images of each component in the rail assembly are stored in a second preset set. Furthermore, this embodiment also standardizes all segmented images. Its core goal is to eliminate irrelevant interference factors in the image, improve data quality and consistency, and lay the foundation for subsequent analysis, modeling, or application.

[0026] The positioning module filters the segmented images of multiple components in the first preset set and the segmented images of the rail bottom in the preset set, and locates and restores them to the reference image. In the prior art, the corresponding size parameters are obtained by processing the massive three-dimensional point clouds of the fastener components to achieve the identification of the fastener model. However, due to the limitations of the acquisition environment or acquisition equipment, as well as the lack of adjustment reference, the extraction of the three-dimensional features of the fasteners from the three-dimensional point cloud data is not stable, resulting in the subsequent distortion of the fastener size parameters, affecting the identification of the fastener model. Therefore, this embodiment sets a positioning module, and when obtaining the characteristic parameters of the fastener components, first filters the segmented images of the components to ensure the quality and integrity of the subsequent processed images, and filters the segmented images of the rail bottom for subsequent adjustment reference. The reason why the rail body is used as the adjustment reference is that in the rail transit system, the rail is the core load-bearing component of the track structure and the direct guide structure of the train form, and its geometric shape and position have a high degree of stability and accuracy. In addition, with regard to the screening of the segmented images of the components, the screening logic is as follows: In rail transit, most fastener components are symmetrically arranged, but there are also a few asymmetric components. Therefore, this embodiment first distinguishes all components based on prior knowledge, and records the symmetrical components as the first component, and the asymmetric components (some special fastening parts, such as special-shaped spring bars and eccentric buckle plates) as the second component. Determine the axisymmetric line in the first component, and divide the edge areas on both sides of the axisymmetric line into multiple verification areas of equal size. For example, the edge areas on both sides of the axisymmetric line of the gauge baffle are divided into 10 mxm rectangular areas. If the corresponding verification areas on both sides overlap, then the verification of this group of verification areas is determined to be successful. If there are 9 groups of successfully verified corresponding verification areas in a single first component, the ratio between the number of groups of successful verification areas and the number of groups of all verification areas is 0.9, while the set value is 0.8. Therefore, the first segmented image of the first component is used as the first screening result. Since the width or length parameters of the components are calculated when the fastener component size parameters are subsequently used to identify the fastener model, the above screening can ensure that there will be no distortion or large errors when the corresponding component feature parameters are subsequently obtained. As for the asymmetric component, i.e., the second component, since it accounts for a relatively small proportion, this embodiment constructs a standard image library of all asymmetric components, and uses the segmented image of the second component and the corresponding image in the standard image library to traverse. If they overlap, the segmented image corresponding to the second component is used as the second screening result; Finally, the first screening result and the second screening result are combined to screen the segmented images of multiple components in the first preset set, such as the segmented image of the gauge plate and the segmented image of the iron plate.

[0027] Feature acquisition module, please refer to Figure 2 First, the baseline adjustment line is obtained using the segmented image of the rail bottom in the baseline image. Then, based on the segmented images of all the filtered components in the baseline image, the characteristic parameters of the corresponding components are obtained. The characteristic parameters of the filtered components are corrected using the baseline adjustment line to obtain the baseline characteristic parameters of the filtered components. The processing logic for obtaining the baseline adjustment line using the segmented image of the rail bottom is as follows: Although the rails are the core load-bearing components of the track structure and the direct guiding structure of the train, their geometric shape and position have high stability and accuracy. However, in long-term daily operation, the bottom area of ​​the rails is easily blocked by stones and other debris. Therefore, this embodiment first obtains a segmented image of the bottom of the rail and records it as the second segmented image, then obtains the edge area image and detects whether there is an occlusion area. The detection algorithm can use a target detection model (such as YOLO, Faster R-CNN), and the identification of occlusion areas in this field is a relatively mature technology, which will not be elaborated in this embodiment. If there is no occlusion area in the edge area image of the rail bottom, the contour line of the second segmented image close to the fastener assembly side is obtained and fitted to obtain the reference adjustment line, such as Figure 2The center line Lr is the reference adjustment line. If there is an occlusion area in the rail bottom edge area image, a certain error may occur if the contour line is still fitted. Therefore, this embodiment first determines the two vertices at the top of the second segmented image, and then uses the two vertices as a reference to determine multiple feature points (such as 6, covering the collected rail bottom area as much as possible) at equal intervals on both sides. The feature points with equal intervals are grouped together, and then the perpendicular line of each group of feature points is obtained. In theory, it is the center line of the second segmented image (rail bottom area). If all the perpendicular lines coincide, they can be directly used as the reference adjustment line. If they do not completely coincide, the perpendicular line with the highest coincidence degree is selected as the reference adjustment line. If 5 groups of perpendicular lines coincide, this can minimize the influence of the occlusion area on the determination of the reference adjustment line.

[0028] Based on the segmented images of all filtered components, the processing logic for obtaining the feature parameters of the corresponding components is as follows: Obtain the contour lines on both sides of the segmented image of any screening component and perform straight line fitting to obtain a set of reference straight lines; calculate and process the distance between the reference straight lines, that is, obtain the image resolution, and obtain the actual size parameters of the component as the characteristic parameters of the screening component, such as Figure 2 The characteristic parameter of the middle gauge baffle is the fitting line L g With L t The distance b between them, the characteristic parameter of the insulating block is the fitting straight line L j With L jr The distance a between them.

[0029] The logic for correcting the characteristic parameters of the screening components through the baseline adjustment line is as follows: A set of reference straight lines of the first screening component and corresponding characteristic parameters are obtained and recorded as first characteristic parameters; based on the fact that the distance between the reference adjustment line and any reference straight line in the set of reference straight lines remains equal (that is, the two reference straight lines remain parallel to the reference adjustment line), the corresponding two reference straight lines are adjusted; and the distance between the adjusted set of reference straight lines is calculated as the reference characteristic parameter of the first screening component.

[0030] The model identification module constructs a database for storing fastener size parameters and fastener models, matches the reference feature parameters of all screening components with the size parameters of all fasteners in the database, outputs the corresponding fastener model, and completes the identification of the fastener model.

[0031] In the prior art, fastener size parameters are obtained and then identified by processing massive 3D point clouds. However, due to limitations in the acquisition environment or equipment, as well as a lack of an adjustment reference, the extraction of 3D fastener features from the 3D point cloud data is unstable, resulting in subsequent distortion of the fastener size parameters and affecting the identification of the fastener model. However, this embodiment constructs an image acquisition module, an image preprocessing module, an image analysis module, a positioning module, a feature acquisition module, and a model recognition module. Using image vision, it not only filters the images of the fasteners to ensure that the subsequent acquisition of the fastener feature parameters is as accurate as possible, but also uses a segmented image of the rail bottom to determine a reference adjustment line. After the feature parameters of the corresponding fasteners are obtained using the segmented images of the screened fasteners, they are corrected using the reference adjustment line to obtain reference feature parameters. Finally, the reference feature parameters of the fasteners (which must be restored to actual size after obtaining the image resolution) are matched with the fastener parameters stored in a database to output the corresponding fastener model and complete the identification of the fastener model.

[0032] In some embodiments, preprocessing the initial image to obtain a reference image includes: De-noising and geometric correction are performed on the initial image to obtain the reference image.

[0033] In some embodiments, the preset model is a SINet model.

[0034] In some embodiments, inputting the reference image into a preset model for processing includes: The reference images are input into the SINet model in batches for processing, and the images processed by the SINet model are standardized to obtain a segmented image of each component in the fastener assembly and a segmented image of each component in the rail assembly.

[0035] In some embodiments, screening the segmented images of the plurality of components in the first preset set includes: Acquire segmented images of all components in a single fastener component in the first preset set to obtain a plurality of first segmented images; Based on prior knowledge, all components in a single fastener assembly are divided into symmetrical components denoted as first components and asymmetrical components denoted as second components; Determine the axisymmetric line in all first components, and divide the edge areas on both sides of the axisymmetric line into multiple verification areas of equal size, calculate whether the first segmented images of the verification areas on both sides of the axisymmetric line overlap, and if so, determine that the verification of the group of verification areas is successful; If the ratio between the number of groups of successfully verified corresponding verification areas in a single first component and the number of groups of all verification areas exceeds a set value, the first segmented image corresponding to the first component is used as the first screening result; Construct a standard image library for all asymmetric components, obtain the first segmented image of the second component and traverse the corresponding images in the standard image library, and if the first segmented image of the second component overlaps with an image in the standard image library, use the first segmented image corresponding to the second component as the second screening result; Based on the first screening result and the second screening result, segmented images of a plurality of components are screened in the first preset set.

[0036] In some embodiments, obtaining the reference adjustment line based on the segmented image of the rail bottom in the reference image includes: Obtaining a segmented image of the rail bottom in the reference image, recorded as a second segmented image; Obtaining an edge region image of the second segmented image close to the fastener component, and determining whether there is an occlusion region in the edge region image; if not, fitting the contour line of the second segmented image close to the fastener component to obtain the baseline adjustment line; and if so, processing the second segmented image to obtain the baseline adjustment line; Processing the second segmented image includes: Taking the top vertices on both sides of the second segmented image as references, determining a plurality of feature points on both sides of the second segmented image at a preset spacing, wherein the feature points on both sides of the second segmented image that are equally spaced from the top vertices constitute a group of feature points, and obtaining a perpendicular line for each group of feature points; It is determined whether the perpendicular lines of all feature points coincide with each other. If so, the perpendicular lines are used as the reference adjustment line. If not, the perpendicular line with the highest coincidence degree is selected as the reference adjustment line.

[0037] In some embodiments, obtaining characteristic parameters of corresponding components based on the segmented images of all the screened components in the reference image includes: Obtain any screening component as a first screening component, obtain contour lines on both sides of the segmented image by the first screening component, and perform straight line fitting to obtain a set of reference straight lines related to the first screening component; The distance between the set of reference straight lines is calculated as a characteristic parameter of the first screening component.

[0038] In some embodiments, modifying characteristic parameters of the screening component based on the baseline adjustment line includes: Obtaining a set of reference straight lines of the first screening component and corresponding characteristic parameters as first characteristic parameters; Adjusting the corresponding two reference straight lines based on maintaining an equal distance between the reference adjustment line and any reference straight line in the set of reference straight lines; The distance between the adjusted set of reference straight lines is calculated as the reference characteristic parameter of the first screening component.

[0039] In some embodiments, matching the reference characteristic parameters of all screening components with the dimensional parameters of all fasteners in the database and outputting the corresponding fastener models includes: Matching the reference characteristic parameters of all the screened components with the dimensional parameters of any corresponding fastener component in the database until a match is successful; The fastener models of the fasteners that are successfully matched are output as the fastener models corresponding to all the filtered components.

[0040] Example 2 Based on the rail transit fastener model identification system provided in the first embodiment of the present application, the second embodiment of the present application also provides a rail transit fastener model identification method, such as Figure 3 Shown, including: Using multiple image acquisition devices installed on the rail inspection vehicle, an initial image of the fastener assembly and the rail assembly is acquired; Preprocessing the initial image to obtain a reference image; Inputting the reference image into a preset model for processing, obtaining a segmented image of each component in the fastener assembly and storing it in a first preset set, and obtaining a segmented image of each component in the rail assembly and storing it in a second preset set; Filtering the segmented images of multiple components in the first preset set and filtering the segmented images of the rail bottom in the second preset set, and positioning and restoring both to the reference image; Based on the segmented image of the rail bottom in the reference image, a reference adjustment line is obtained; based on the segmented images of all screening components in the reference image, characteristic parameters of corresponding components are obtained, and based on the reference adjustment line, the characteristic parameters of the screening components are corrected to obtain reference characteristic parameters of the screening components; A database is constructed to store the corresponding fastener size parameters and fastener models, the reference feature parameters of all screening components are matched with the size parameters of all fasteners in the database, the corresponding fastener models are output, and the fastener model identification is completed.

[0041] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computing software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0042] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0043] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A rail transit fastener model identification system, characterized in that: include: An image acquisition module, which uses multiple image acquisition devices installed on the rail inspection vehicle to capture initial images of the fastener assembly and the rail assembly; An image preprocessing module, which preprocesses the initial image to obtain a reference image; an image analysis module, inputting the reference image into a preset model for processing, obtaining a segmented image of each component in the fastener assembly and storing the segmented image in a first preset set, and obtaining a segmented image of each component in the rail assembly and storing the segmented image in a second preset set; a positioning module, which selects the segmented images of multiple components in the first preset set and the segmented image of the rail bottom in the second preset set, and positions and restores them to the reference image; a feature acquisition module that acquires a reference adjustment line based on a segmented image of the rail bottom in the reference image; acquires feature parameters of corresponding components based on segmented images of all screening components in the reference image; and modifies the feature parameters of the screening components based on the reference adjustment line to obtain reference feature parameters of the screening components; The model identification module constructs a database for storing fastener size parameters and fastener models, matches the reference feature parameters of all screening components with the size parameters of all fasteners in the database, outputs the corresponding fastener model, and completes the identification of the fastener model.

2. A rail transit fastener model identification system according to claim 1, characterized in that: Preprocessing the initial image to obtain a reference image includes: De-noising and geometric correction are performed on the initial image to obtain the reference image.

3. A rail transit fastener model identification system according to claim 1, characterized in that: The preset model is the SINet model.

4. A rail transit fastener model identification system according to claim 3, characterized in that: Inputting the reference image into a preset model for processing includes: The reference images are input into the SINet model in batches for processing, and the images processed by the SINet model are standardized to obtain a segmented image of each component in the fastener assembly and a segmented image of each component in the rail assembly.

5. A rail transit fastener model identification system according to claim 1, characterized in that: Screening the segmented images of the plurality of components in the first preset set includes: Acquire segmented images of all components in a single fastener component in the first preset set to obtain a plurality of first segmented images; Based on prior knowledge, all components in a single fastener assembly are divided into symmetrical components denoted as first components and asymmetrical components denoted as second components; Determine the axisymmetric line in all first components, and divide the edge areas on both sides of the axisymmetric line into multiple verification areas of equal size, calculate whether the first segmented images of the verification areas on both sides of the axisymmetric line overlap, and if so, determine that the verification of the group of verification areas is successful; If the ratio between the number of groups of successfully verified corresponding verification areas in a single first component and the number of groups of all verification areas exceeds a set value, the first segmented image corresponding to the first component is used as the first screening result; Construct a standard image library for all asymmetric components, obtain the first segmented image of the second component and traverse the corresponding images in the standard image library, and if the first segmented image of the second component overlaps with an image in the standard image library, use the first segmented image corresponding to the second component as the second screening result; Based on the first screening result and the second screening result, segmented images of a plurality of components are screened in the first preset set.

6. A rail transit fastener model identification system according to claim 1, characterized in that: The step of obtaining the reference adjustment line based on the segmented image of the rail bottom in the reference image includes: Obtaining a segmented image of the rail bottom in the reference image, recorded as a second segmented image; Obtaining an edge region image of the second segmented image close to the fastener component, and determining whether there is an occlusion region in the edge region image; if not, fitting the contour line of the second segmented image close to the fastener component to obtain the baseline adjustment line; and if so, processing the second segmented image to obtain the baseline adjustment line; Processing the second segmented image includes: Taking the top vertices on both sides of the second segmented image as references, determining a plurality of feature points on both sides of the second segmented image at a preset spacing, wherein the feature points on both sides of the second segmented image that are equally spaced from the top vertices constitute a group of feature points, and obtaining a perpendicular line for each group of feature points; It is determined whether the perpendicular lines of all feature points coincide with each other. If so, the perpendicular lines are used as the reference adjustment line. If not, the perpendicular line with the highest coincidence degree is selected as the reference adjustment line.

7. A rail transit fastener model identification system according to claim 6, characterized in that: Based on the segmented images of all the screened components in the reference image, obtaining characteristic parameters of the corresponding components includes: Obtain any screening component as a first screening component, obtain contour lines on both sides of the segmented image by the first screening component, and perform straight line fitting to obtain a set of reference straight lines related to the first screening component; The distance between the set of reference straight lines is calculated as a characteristic parameter of the first screening component.

8. A rail transit fastener model identification system according to claim 7, characterized in that: Based on the reference adjustment line, the characteristic parameters of the screening component are modified including: Obtaining a set of reference straight lines of the first screening component and corresponding characteristic parameters as first characteristic parameters; Adjusting the corresponding two reference straight lines based on maintaining an equal distance between the reference adjustment line and any reference straight line in the set of reference straight lines; The distance between the adjusted set of reference straight lines is calculated as the reference characteristic parameter of the first screening component.

9. The rail transit fastener model identification system according to claim 1, characterized in that: Match the reference characteristic parameters of all screened components with the dimensional parameters of all fasteners in the database, and output the corresponding fastener models including: Matching the reference characteristic parameters of all the screened components with the dimensional parameters of any corresponding fastener component in the database until a match is successful; The fastener models of the fasteners that are successfully matched are output as the fastener models corresponding to all the filtered components.

10. A rail transit fastener model identification method, characterized in that: include: Using multiple image acquisition devices installed on the rail inspection vehicle, an initial image of the fastener assembly and the rail assembly is acquired; Preprocessing the initial image to obtain a reference image; Inputting the reference image into a preset model for processing, obtaining a segmented image of each component in the fastener assembly and storing it in a first preset set, and obtaining a segmented image of each component in the rail assembly and storing it in a second preset set; Filtering the segmented images of multiple components in the first preset set and filtering the segmented images of the rail bottom in the second preset set, and positioning and restoring both to the reference image; Acquiring a reference adjustment line based on a segmented image of the rail bottom in the reference image; Based on the segmented images of all screening components in the reference image, characteristic parameters of corresponding components are obtained, and based on the reference adjustment line, the characteristic parameters of the screening components are corrected to obtain reference characteristic parameters of the screening components; A database is constructed to store the corresponding fastener size parameters and fastener models, the reference feature parameters of all screening components are matched with the size parameters of all fasteners in the database, the corresponding fastener models are output, and the fastener model identification is completed.

Images