Inverted arch steel bar processing quality detection method and device

Abstract

The application provides an inverted arch steel bar processing quality detection method and device, the method comprising: acquiring a target inverted arch steel bar processing image; inputting the target inverted arch steel bar processing image into a preset steel bar position recognition model, so that the steel bar position recognition model outputs steel bar position information corresponding to the target inverted arch steel bar processing image, and determines a steel bar processing quality detection result corresponding to the target inverted arch steel bar processing image based on the steel bar position information. The application can improve the automation and intelligence of inverted arch steel bar position recognition, and effectively improve the efficiency, accuracy and effectiveness of inverted arch steel bar position recognition, thereby effectively improving the efficiency, accuracy and effectiveness of inverted arch steel bar processing quality detection using the inverted arch steel bar position recognition result.

Description

Technical Field

[0001] This application relates to the field of railway engineering technology, and in particular to a method and device for detecting the processing quality of inverted arch steel bars. Background Technology

[0002] The railway tunnel invert is a reverse arch structure installed at the bottom of a railway tunnel to improve the stress conditions of the superstructure. It forms the foundation of the railway tunnel structure, effectively transferring the ground pressure from above the tunnel through the tunnel sidewalls or the road surface load to the underground, while also resisting the reaction forces from the underlying strata. Therefore, if the construction quality of the railway tunnel invert is not controlled, the entire railway tunnel project is highly susceptible to quality problems. Currently, some railway tunnels constructed using full-face hard rock tunnel boring machines (TBMs) employ prefabrication for their inverts. Prefabricated invert blocks are used as the main structure of the invert and transported directly to the construction site for installation. Therefore, the prefabrication quality of the invert blocks directly affects the quality of tunnel construction. During prefabrication, steel bars need to be pre-processed into steel cages, which are then placed in molds for casting. Therefore, it is necessary to conduct quality inspections on the steel reinforcement used in the prefabricated railway tunnel invert structure to ensure the reliability of the prefabricated railway tunnel invert structure.

[0003] Currently, the main method for inspecting the processing quality of steel reinforcement structures used in precast railway tunnel invert arch structures typically involves manually measuring and marking the positions of steel bars in structures such as steel cages, then manually measuring the quantity, size, and spacing of the steel bars, and finally judging whether the current steel reinforcement structure meets the preset steel reinforcement processing quality requirements based on human experience. However, existing methods for inspecting the processing quality of invert arch steel reinforcement are largely based on visual inspection or the use of measuring tools. Therefore, the inspection results are highly dependent on the inspectors, resulting in low inspection efficiency. Furthermore, due to the high degree of subjectivity in the overall inspection, the accuracy and effectiveness of the inspection results cannot be guaranteed. Summary of the Invention

[0004] In view of this, embodiments of this application provide a method and apparatus for detecting the processing quality of inverted arch reinforcement bars, so as to eliminate or improve one or more defects existing in the prior art.

[0005] One aspect of this application provides a method for inspecting the processing quality of inverted arch reinforcement bars, including:

[0006] Obtain images of the target arch reinforcement processing;

[0007] The target arch rebar processing image is input into a preset rebar position recognition model, so that the rebar position recognition model outputs the rebar position information corresponding to the target arch rebar processing image, and the rebar processing quality inspection result corresponding to the target arch rebar processing image is determined based on the rebar position information.

[0008] In some embodiments of this application, obtaining the target invert arch reinforcement processing image includes:

[0009] Acquire images of the steel reinforcement processing corresponding to the target steel reinforcement structure used for prefabricated railway tunnel invert arch structure blocks;

[0010] The steel reinforcement processing image of the target steel reinforcement structure is preprocessed based on a preset preprocessing method to obtain the corresponding target inverted arch steel reinforcement processing image.

[0011] In some embodiments of this application, it also includes:

[0012] While acquiring images of the steel reinforcement processing of the target steel reinforcement structure used for prefabricated railway tunnel invert arch structure blocks, distance information of the target steel reinforcement structure is also acquired.

[0013] Correspondingly, after the rebar position recognition model outputs the rebar position information corresponding to the target invert rebar processing image, it also includes:

[0014] Based on the rebar position information corresponding to the target arch rebar processing image and the distance information, the target rebar morphology data for detecting the rebar processing quality corresponding to the target rebar structure is determined.

[0015] In some embodiments of this application, it also includes:

[0016] Based on the steel bar defect rule library used to store the correspondence between steel bar shape and processing quality defects, the processing quality of the target steel bar shape data is inspected to obtain the steel bar processing quality inspection result corresponding to the target steel bar structure.

[0017] The steel bar processing quality inspection results are output to visualize the steel bar processing quality inspection results.

[0018] In some embodiments of this application, it also includes:

[0019] Obtain a historical invert arch steel reinforcement processing image dataset, wherein the historical invert arch steel reinforcement processing image dataset includes: steel reinforcement processing images of various steel reinforcement structures for prefabricated railway tunnel invert arch structural blocks stored in the railway engineering management platform, and open-source steel reinforcement processing images collected from outside the railway engineering management platform;

[0020] The historical inverted arch steel bar processing image dataset is expanded, and the historical inverted arch steel bar processing image dataset is preprocessed based on a preset preprocessing method.

[0021] Each of the steel reinforcement processing images in the historical invert arch steel reinforcement processing image dataset is labeled with the steel reinforcement position to form a corresponding steel reinforcement processing training set;

[0022] The pre-set machine learning model is trained using the aforementioned rebar processing training set to obtain the corresponding rebar position recognition model.

[0023] In some embodiments of this application, the machine learning model includes: a deep residual network based on an attention mechanism.

[0024] In some embodiments of this application, the preprocessing method includes at least one of image light field correction, image white balance processing, image registration, image noise reduction, and image edge enhancement.

[0025] Another aspect of this application provides a device for detecting the processing quality of inverted arch reinforcement bars, comprising:

[0026] The image acquisition module is used to acquire images of the target inverted arch reinforcement processing.

[0027] The model recognition module is used to input the target inverted arch steel bar processing image into a preset steel bar position recognition model, so that the steel bar position recognition model outputs the steel bar position information corresponding to the target inverted arch steel bar processing image, and determines the steel bar processing quality inspection result corresponding to the target inverted arch steel bar processing image based on the steel bar position information.

[0028] Another aspect of this application provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method for detecting the processing quality of inverted arch reinforcement.

[0029] Another aspect of this application provides a computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the aforementioned method for detecting the processing quality of inverted arch reinforcement.

[0030] The method for inspecting the processing quality of inverted arch reinforcement provided in this application involves acquiring a processing image of the target inverted arch reinforcement; inputting the processing image of the target inverted arch reinforcement into a preset reinforcement position recognition model, so that the reinforcement position recognition model outputs reinforcement position information corresponding to the processing image of the target inverted arch reinforcement, and determining the processing quality inspection result of the reinforcement corresponding to the processing image of the target inverted arch reinforcement based on the reinforcement position information. This method can improve the automation and intelligence of inverted arch reinforcement position recognition, save time and labor costs, and effectively improve the efficiency, accuracy and effectiveness of inverted arch reinforcement position recognition. It can also assist in manual inspection of reinforcement processing quality, thereby effectively improving the efficiency, accuracy and effectiveness of using inverted arch reinforcement position recognition results for inverted arch reinforcement processing quality inspection, and ensuring the reliability of inverted arch applications.

[0031] Additional advantages, objectives, and features of this application will be set forth in part in the description which follows, and will in part become apparent to those skilled in the art upon review of the following description, or may be learned by practice of the application. The objectives and other advantages of this application can be realized and obtained by means of the structures specifically pointed out in the specification and drawings.

[0032] Those skilled in the art will understand that the purposes and advantages that can be achieved with this application are not limited to those specifically described above, and that the above and other purposes that this application can achieve will be more clearly understood from the following detailed description. Attached Figure Description

[0033] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, do not constitute a limitation thereof. The components in the drawings are not drawn to scale but are merely for illustrating the principles of this application. For ease of illustration and description of certain parts of this application, corresponding portions in the drawings may be enlarged, i.e., may appear larger relative to other components in an exemplary device actually manufactured according to this application. In the drawings:

[0034] Figure 1 This is a schematic diagram of the overall process of the method for detecting the processing quality of inverted arch reinforcement in one embodiment of this application.

[0035] Figure 2 This is a schematic diagram of a specific process for detecting the processing quality of inverted arch reinforcement in one embodiment of this application.

[0036] Figure 3 This is a schematic diagram illustrating an example of the architecture of ResNet, a deep residual network based on an attention mechanism, in one embodiment of this application.

[0037] Figure 4 This is a schematic diagram of the structure of the inverted arch reinforcement processing quality inspection device in another embodiment of this application.

[0038] Figure 5 This is a schematic diagram illustrating the execution logic of the railway tunnel arch block reinforcement processing quality inspection system provided in the application example of this application.

[0039] Figure 6 This is a flowchart illustrating the method for detecting the processing quality of inverted arch reinforcement bars implemented using the railway tunnel inverted arch reinforcement bar processing quality detection system, as provided in the application example of this application. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the embodiments and accompanying drawings. Here, the illustrative embodiments and their descriptions are used to explain this application, but are not intended to limit it.

[0041] It should also be noted that, in order to avoid obscuring this application with unnecessary details, only the structures and / or processing steps closely related to the solution according to this application are shown in the accompanying drawings, while other details that are not closely related to this application are omitted.

[0042] It should be emphasized that the term "including / comprises" as used herein refers to the presence of a feature, element, step, or component, but does not exclude the presence or addition of one or more other features, elements, steps, or components.

[0043] It should also be noted that, unless otherwise specified, the term "connection" in this article can refer not only to a direct connection, but also to an indirect connection involving an intermediary.

[0044] In the following description, embodiments of the present application will be illustrated with reference to the accompanying drawings. In the drawings, the same reference numerals represent the same or similar parts, or the same or similar steps.

[0045] In order to improve the efficiency of quality inspection of invert arch reinforcement processing and to promptly and accurately identify problems in the reinforcement processing of railway tunnel invert arch blocks, this application considers using machine vision technology to improve the automation and intelligence of invert arch reinforcement position identification, save time and labor costs, and improve the efficiency, accuracy and effectiveness of invert arch reinforcement position identification.

[0046] Machine vision technology has developed rapidly in recent years, with increasingly faster and more accurate target detection. Introducing machine vision technology, with its advantages of high precision, high efficiency, high flexibility, and low cost, into the quality control of steel reinforcement processing in railway tunnel invert blocks can assist manual inspection of steel reinforcement processing quality and improve inspection efficiency. Machine vision technology already has relatively mature target detection models, such as Fast CNN, SSD, RetinaNet, and YOLO algorithms, which have achieved good application value and economic benefits in various industries. This application improves the efficiency and accuracy of steel reinforcement processing quality identification in railway tunnel invert blocks by recognizing the features of steel reinforcement processing images, performing efficient target segmentation and detection, and training models based on deep learning algorithms to perform steel reinforcement quantity counting, size measurement, and spacing measurement.

[0047] The following examples will provide a detailed description.

[0048] Based on this, this application provides a method for detecting the processing quality of inverted arch reinforcing bars, which can be implemented by an inverted arch reinforcing bar processing quality detection device. See [link to relevant documentation]. Figure 1 The method for inspecting the processing quality of the inverted arch reinforcement specifically includes the following:

[0049] Step 100: Obtain the image of the target inverted arch reinforcement processing.

[0050] In one or more embodiments of this application, the invert arch reinforcement processing image refers to the reinforcement processing image corresponding to the reinforcement structure used for prefabricating railway tunnel invert arch structural blocks. The reinforcement processing image refers to image data containing the reinforcement structure, wherein the reinforcement structure refers to a structure composed of multiple reinforcement bars, such as a reinforcement cage. In practical applications, the reinforcement structure can also adopt other structures besides the reinforcement cage.

[0051] Based on this, the target invert arch rebar processing image refers to the rebar processing image that meets the data format requirements and other specifications for inputting the rebar position recognition model mentioned in step 200 below and is to be inspected for the quality of invert arch rebar processing. The historical invert arch rebar processing image dataset refers to the rebar processing image data obtained during the model training phase for training the rebar position recognition model mentioned in step 200 below.

[0052] In step 100, the inverted arch steel bar processing quality inspection device can directly receive target inverted arch steel bar processing image data sent by client devices or image acquisition devices, or it can receive inverted arch steel bar processing image data to be inspected sent by client devices or image acquisition devices, and then process it based on preset image processing rules to obtain the corresponding target inverted arch steel bar processing image data. The specific selection can be made according to parameters such as the data processing performance of the inverted arch steel bar processing quality inspection device.

[0053] Step 200: Input the target inverted arch steel bar processing image into a preset steel bar position recognition model, so that the steel bar position recognition model outputs the steel bar position information corresponding to the target inverted arch steel bar image, and determines the steel bar processing quality inspection result corresponding to the target inverted arch steel bar image based on the steel bar position information.

[0054] It is understood that the rebar location recognition model is an image feature prediction or recognition model, specifically used to output the corresponding rebar location information based on the input image of the inverted arch rebar processing.

[0055] In one or more embodiments of this application, the rebar location information may include: the location points of the two ends of each rebar and the intersection of different rebars in the rebar structure in a preset coordinate system. This coordinate system may be a three-dimensional coordinate system.

[0056] In step 200, after the rebar location recognition model outputs the rebar location information corresponding to the target invert arch rebar image, the invert arch rebar processing quality inspection device can directly send the rebar location information to the client device, display screen, or human-computer interaction screen. This assists the user in calculating the length of each rebar and the spacing between adjacent rebars in the rebar structure based on the rebar location information and the pre-known parameters such as the diameter and density of each rebar in the rebar structure. This data is then summarized into comprehensive morphological data of the invert arch structural block rebar corresponding to the rebar structure. Based on this comprehensive morphological data, the device can determine whether there are processing quality problems in the current rebar structure to complete the invert arch rebar processing quality inspection. Since the rebar location information corresponding to the target invert arch rebar image is directly output by the rebar location recognition model, it saves a significant amount of manpower and time costs previously required for labeling the rebar location information of the rebar structure. It also ensures the accuracy and effectiveness of the rebar location information recognition, thus effectively improving the overall efficiency and effectiveness of the invert arch rebar processing quality inspection.

[0057] In another implementation of step 200, to further improve the efficiency and automation of the inverted arch reinforcement processing quality inspection, after the reinforcement position recognition model outputs the reinforcement position information corresponding to the target inverted arch reinforcement image, the inverted arch reinforcement processing quality inspection device can automatically perform inverted arch reinforcement processing quality inspection on the above-mentioned reinforcement structure based on the reinforcement position information according to the inverted arch reinforcement processing quality inspection logic pre-stored locally, and then directly output the inverted arch reinforcement processing quality inspection result, so as to further save labor and time costs. The specific implementation method is described in detail in the following embodiments.

[0058] As can be seen from the above description, the invert arch reinforcement processing quality inspection method provided in this application embodiment can improve the automation and intelligence of invert arch reinforcement position identification, save time and labor costs, and effectively improve the efficiency, accuracy and effectiveness of invert arch reinforcement position identification. In turn, it can effectively improve the efficiency, accuracy and effectiveness of using invert arch reinforcement position identification results for invert arch reinforcement processing quality inspection, and ensure the reliability of invert arch application.

[0059] To further improve the reliability and effectiveness of the application of images of the target arch reinforcement processing, a method for detecting the processing quality of arch reinforcement provided in this application embodiment is described below. Figure 2 Step 100 in the method for inspecting the processing quality of the inverted arch reinforcement specifically includes the following:

[0060] Step 110: Acquire images of the steel reinforcement processing corresponding to the target steel reinforcement structure used for prefabricated railway tunnel invert arch structure blocks.

[0061] Step 120: Preprocess the steel reinforcement processing image of the target steel reinforcement structure based on the preset preprocessing method to obtain the corresponding target invert arch steel reinforcement processing image.

[0062] To further improve the automation level of invert arch reinforcement processing quality inspection, in an embodiment of this application, a method for invert arch reinforcement processing quality inspection is provided, see [link to relevant documentation]. Figure 2 The method for detecting the processing quality of inverted arch reinforcement also includes the following specific content before step 120:

[0063] Step 111: While acquiring images of the steel reinforcement processing of the target steel reinforcement structure used for prefabricated railway tunnel arch structure blocks, acquire distance information of the target steel reinforcement structure.

[0064] In step 111, the inverted arch steel bar processing quality inspection device can receive steel bar processing images and distance information for the target steel bar structure sent by a dedicated image acquisition device for railway steel bar processing inspection.

[0065] It is understandable that a dedicated image acquisition device for railway rebar processing and inspection is pre-installed. This dedicated image acquisition device for railway rebar processing and inspection can simultaneously acquire image information and distance information. This dedicated image acquisition device for railway rebar processing and inspection is equipped with a depth camera (such as the Intel RealSense series D435i depth camera), and the depth camera captures images.

[0066] Correspondingly, after step 200 in the method for inspecting the processing quality of inverted arch reinforcement, the following content is also included:

[0067] Step 300: Based on the rebar position information corresponding to the target arch rebar processing image and the distance information, determine the target rebar morphology data corresponding to the target rebar structure for detecting the rebar processing quality.

[0068] In step 300, the rebar position recognition model is used to mark the rebar positions in the image. Then, the invert arch rebar processing quality inspection device, combined with distance information, calculates the number, spacing, and length of the rebars, forming comprehensive morphological data of the invert arch structural block rebars. A Yolov5s target detection network is used to identify the target rebars and count their quantity. The color image stream obtained from the depth camera is aligned with the depth stream, so that each pixel in the color image corresponds to a depth value, which is used as the z-coordinate. Then, the x and y coordinates of that pixel are obtained through camera intrinsic parameters. The obtained x, y, and z coordinates are the three-dimensional coordinates of that pixel in the camera coordinate system. Using this method, the three-dimensional coordinates of the center points of the two measured rebars are calculated, and the distance between these two points can be calculated using the Pythagorean theorem, which is the rebar spacing. Simultaneously, the three-dimensional coordinates of the two end points of the rebar are identified through target detection, and the distance between the two end points is calculated in the same way to obtain the rebar length.

[0069] To further improve the automation level of invert arch reinforcement processing quality inspection, in an embodiment of this application, a method for invert arch reinforcement processing quality inspection is provided, see [link to relevant documentation]. Figure 2 The method for detecting the processing quality of inverted arch reinforcement also includes the following steps prior to step 300:

[0070] Step 400: Based on the rebar defect rule library used to store the correspondence between rebar morphology and processing quality defects, perform processing quality inspection on the target rebar morphology data to obtain the rebar processing quality inspection result corresponding to the target rebar structure.

[0071] Specifically, the rebar defect rule library, also known as the tunnel invert arch structural block rebar defect rule library, is pre-established to store the correspondence between rebar morphology data and processing quality defect data. In step 400, the invert arch rebar processing quality detection device can input the target rebar morphology data into the rule library for analysis, output defect status judgment, and mark defect locations. The defect rule library can use a structured data table, including: image information data (image number, image storage location, production site number, annotation personnel number), rebar morphology data (quantity, length, thickness), and defect annotation data (defect type, defect severity). The images in this defect rule library are structural blocks that have been manually identified as defective. Each defect type (such as rebar deformation) has a large inter-class distance in its corresponding morphological feature data. By calculating the nearest spatial distance from the input morphological feature data, the corresponding defect type can be obtained.

[0072] In one or more embodiments of this application, the target steel reinforcement morphology data may also be referred to as the comprehensive morphology data of the steel reinforcement in the inverted arch structure block.

[0073] Step 500: Output the steel bar processing quality inspection results to visualize the steel bar processing quality inspection results.

[0074] In step 500, the inverted arch steel bar processing quality inspection device can send the steel bar processing quality inspection results to the display and interaction tool, display the defect detection results in a visual form and generate a defect report. The inspection results can be manually reviewed and adjusted, and the adjusted results can be fed back to the training set in real time to dynamically improve the recognition accuracy.

[0075] To improve the reliability and effectiveness of the rebar location identification model, in an embodiment of this application, a method for detecting the processing quality of inverted arch rebar is provided (see...). Figure 2 The method for detecting the processing quality of inverted arch reinforcement also includes the following specific content before step 200:

[0076] Step 010: Obtain a historical invert arch reinforcement processing image dataset, wherein the historical invert arch reinforcement processing image dataset includes: reinforcement processing images of various reinforcement structures for prefabricated railway tunnel invert arch structural blocks stored in the railway engineering management platform, and open-source reinforcement processing images collected from outside the railway engineering management platform.

[0077] In step 010, the inverted arch steel bar processing quality inspection device can rely on the existing steel bar processing image data in the railway engineering management platform, collect open source steel bar processing images as a supplement, and form a historical inverted arch steel bar processing image dataset containing various steel bar processing images. This historical inverted arch steel bar processing image dataset can be stored in a railway steel bar processing database.

[0078] In one or more embodiments of the application examples in this application, the railway engineering management platform refers to a unified and open engineering information platform and application.

[0079] Step 020: Expand the historical inverted arch rebar processing image dataset, and preprocess the historical inverted arch rebar processing image dataset based on a preset preprocessing method.

[0080] In step 020, the inverted arch steel bar processing quality detection device can expand the database by means of translation, flipping, rotation, cutting, scaling and other methods to improve the training effect.

[0081] Step 030: Label the rebar positions in each of the rebar processing images in the historical invert arch rebar processing image dataset to form a corresponding rebar processing training set.

[0082] In step 030, an interactive rebar parameter annotation tool can be constructed to manually annotate key parameters such as rebar location, forming a railway rebar processing training set for training the rebar detection model. The workflow of the interactive rebar parameter annotation tool can include: 1. Loading an image of the rebar to be annotated from an image library, segmenting the image into several regions and displaying the boundaries; 2. Annotating all rebar regions in the image; 3. Annotating the morphological parameters of the rebar regions in the image; 4. Confirming whether there are defects in the rebar in the image, and if so, specifying the defect type; 5. Storing the results with defects in the defect library and the results without defects in the training set database, repeating step 1.

[0083] Step 040: Use the steel bar processing training set to train a preset machine learning model to obtain the corresponding steel bar position recognition model.

[0084] To further improve the reliability and effectiveness of the rebar location identification model, the machine learning model in the rebar processing quality detection method for inverted arches provided in this application embodiment includes: a deep residual network ResNet based on an attention mechanism.

[0085] Specifically, see Figure 3 This study employs a deep residual network, ResNet, with an attention mechanism as the overall model framework. The initial network parameters are taken from ImageNet, a large visualization database used for visual object recognition software research. Transfer learning is performed using a railway steel rebar processing training set to reduce data requirements and training time. The network is based on ResNet-101 (the suffix number represents the network depth), with improvements made to the residual block structure. The positions of downsampling and mean pooling operations within the residual blocks are adjusted to ensure traversability and feature reuse in the feature extraction process. The network model introduces an efficient attention mechanism module, ECA, using global mean pooling to effectively reduce the number of parameters while maintaining network performance. The network uses two residual modules and two attention modules as a unit, with ReLU as the activation function, SGDM as the optimizer, and Dropout with a probability of 0.4 in the fully connected layers.

[0086] exist Figure 3In this code, "ConV(1×1)" represents a convolutional layer with a kernel size of 1; "AvgPool(2×2's=2)" represents an average pooling layer with a kernel size of 2 and a stride of 2; and "ConV(3×3's=2)" represents a convolutional layer with a kernel size of 3 and a stride of 2. "H" represents the height of the feature map or image; "W" represents the width of the feature map or image; "C" represents the number of channels in the feature map or image; "1×1×C" indicates that the feature map or image has a size where both width and height are 1; "ConV(1×k)" represents a convolutional layer with a kernel size of 1×k; "δ" indicates that the standard Sigmoid activation function is used in this layer; and "x" indicates that the input data and the calculated feature data are convolved.

[0087] To further improve the accuracy of the quality inspection of inverted arch reinforcement processing, in an embodiment of the inverted arch reinforcement processing quality inspection method provided in this application, the preprocessing method mentioned in steps 020 and 120 of the inverted arch reinforcement processing quality inspection method specifically includes at least one of the following: image light field correction, image white balance processing, image registration, image noise reduction, and image edge enhancement.

[0088] In a preferred embodiment, the preprocessing may include: sequentially performing image light field correction, image white balance processing, image registration, image noise reduction, and image edge enhancement.

[0089] Specifically, the LinkNet light field correction method can be used for light field correction and white balance, the SURF algorithm can be used for image registration and correction, and a modular autoencoder method can be used for image denoising and edge enhancement, etc. Among them, the SURF algorithm is a fast and robust feature extraction registration algorithm proposed based on the SIFT algorithm.

[0090] From a software perspective, this application also provides a device for performing all or part of the inverted arch reinforcement processing quality inspection method, see [link to relevant documentation]. Figure 4 The aforementioned inverted arch reinforcement processing quality inspection device specifically includes the following components:

[0091] Image acquisition module 10 is used to acquire images of the target inverted arch reinforcement processing.

[0092] The model recognition module 20 is used to input the target inverted arch steel bar processing image into a preset steel bar position recognition model, so that the steel bar position recognition model outputs the steel bar position information corresponding to the target inverted arch steel bar image, so as to determine the steel bar processing quality inspection result corresponding to the target inverted arch steel bar image based on the steel bar position information.

[0093] The embodiments of the inverted arch steel bar processing quality inspection device provided in this application can be used to execute the processing flow of the inverted arch steel bar processing quality inspection method in the above embodiments. Its functions will not be repeated here, but can be referred to the detailed description of the inverted arch steel bar processing quality inspection method embodiments above.

[0094] The part of the inverted arch rebar processing quality inspection device that performs the inverted arch rebar processing quality inspection can be executed on a server. Alternatively, in another practical application, all operations can be completed on the client device. The choice can be made based on the processing capacity of the client device and the limitations of the user's usage scenario. This application does not impose any limitations on this. If all operations are completed on the client device, the client device may further include a processor for the specific processing of the inverted arch rebar processing quality inspection.

[0095] The aforementioned client device may have a communication module (i.e., a communication unit) that can communicate with a remote server to achieve data transmission. The server may include a server on the task scheduling center side; in other implementation scenarios, it may also include a server on an intermediate platform, such as a server on a third-party server platform that has a communication link with the task scheduling center server. The server may include a single computer device, a server cluster consisting of multiple servers, or a distributed server structure.

[0096] The server and the client device can communicate using any suitable network protocol, including those not yet developed as of the date of this application. Such network protocols may include, for example, TCP / IP, UDP / IP, HTTP, HTTPS, etc. Furthermore, such network protocols may also include RPC (Remote Procedure Call Protocol) and REST (Representational State Transfer Protocol) protocols used on top of the aforementioned protocols.

[0097] As can be seen from the above description, the invert arch reinforcement processing quality inspection device provided in this application embodiment can improve the automation and intelligence of invert arch reinforcement position identification, save time and labor costs, and effectively improve the efficiency, accuracy and effectiveness of invert arch reinforcement position identification. In turn, it can effectively improve the efficiency, accuracy and effectiveness of using the invert arch reinforcement position identification results to inspect the invert arch reinforcement processing quality, and ensure the reliability of invert arch application.

[0098] To further illustrate this solution, this application also provides a specific application example of a method for detecting the processing quality of inverted arch reinforcement bars, specifically involving the process of detecting the processing quality of reinforcement bars in inverted arch structural blocks in railway tunnels, improving the efficiency of detecting the processing quality of inverted arch reinforcement bars, and timely and accurately identifying problems in the processing of reinforcement bars in inverted arch structural blocks in railway tunnels.

[0099] This application provides an application example of a railway tunnel invert arch reinforcement processing quality inspection system, used to implement the aforementioned invert arch reinforcement processing quality inspection method. (See also...) Figure 5 The process includes dataset collection and creation, rebar recognition model training, rebar detection and parameter identification, and result display and feedback. Dataset collection and creation includes: a railway rebar processing database, an interactive rebar parameter annotation tool, a railway rebar processing training set, and a rebar processing image preprocessing module. Rebar detection and parameter identification includes: comprehensive morphological data of inverted arch block rebars and a defect rule library for tunnel inverted arch block rebars. Result display and feedback includes: defect reports and manual review and adjustment.

[0100] Specifically, by integrating algorithms such as machine vision, image processing, and deep learning, the existing steel bar processing image data of the railway engineering management platform is expanded to form a railway steel bar processing training set. Multiple image processing methods are integrated to form an image preprocessing module. A general framework for a steel bar location recognition model is established, and the model is trained and optimized using the railway steel bar processing training set to form a steel bar location recognition model. On-site photography is conducted using instruments equipped with imaging and laser ranging capabilities. After image information preprocessing, it is input into the steel bar location recognition model to generate steel bar location information. Combined with ranging information, the spacing, density, and length of the steel bars are calculated to form comprehensive morphological data of the invert arch structure block steel bars. A defect rule library for tunnel invert arch steel block reinforcement is established. The comprehensive morphological data of the steel bars is input into the rule library for analysis, and defect status judgments and defect location markings are output. The recognition results are displayed and a defect report is generated for manual review. If there are inaccuracies in the recognition, manual adjustments can be made and feedback can be sent back to the railway steel bar processing training set.

[0101] The application example provided in this application presents a method for inspecting the processing quality of inverted arch reinforcement bars, implemented using a railway tunnel inverted arch reinforcement bar processing quality inspection system. This method includes processes such as data acquisition and creation, reinforcement bar location identification model training, reinforcement bar detection and parameter identification, and result display and feedback. See also... Figure 6 The specific method for detecting the processing quality of inverted arch reinforcement bars using the railway tunnel inverted arch reinforcement bar processing quality detection system is described below:

[0102] (1) Dataset Collection and Creation: Based on the existing steel bar processing image data in the railway engineering management platform, open-source steel bar processing images were collected as a supplement to form a railway steel bar processing database. The database was expanded using translation, flipping, rotation, cropping, scaling, and other methods to improve the training effect. An interactive steel bar parameter annotation tool was constructed to manually annotate key parameters such as steel bar position, forming a railway steel bar processing training set for training the steel bar detection model. Based on the image characteristics in the training set, the LinkNet method was used for light field correction and white balance, the SURF method was used for image registration and correction, and the modular autoencoder method was used for image noise reduction and edge enhancement, which were integrated to form a steel bar processing image preprocessing functional module.

[0103] (2) Training of the rebar location recognition model: A deep residual network with an attention mechanism is adopted as the overall framework of the model. The initial parameters of the network are taken from ImageNet, and transfer learning is performed using the railway rebar processing training set to reduce the amount of data required and the training time. The residual block structure in the model is improved to ensure that the feature information extraction process satisfies traversal and has the effect of feature reuse. The network model introduces an efficient attention mechanism module ECA and uses global mean pooling to ensure network performance while effectively reducing the number of parameters.

[0104] (3) Reinforcing bar inspection and parameter identification: A dedicated image acquisition device for railway reinforcing bar processing and inspection is set up, which can simultaneously acquire image information and distance information. After preprocessing the image information, the reinforcing bar positions in the image are marked using a reinforcing bar position recognition model. Combined with the distance measurement information, the spacing, density, and length of the reinforcing bars are calculated to form comprehensive morphological data of the reinforcing bars in the tunnel invert arch structure block. A rule library for reinforcing bar defects in the tunnel invert arch structure block is established. The comprehensive morphological data of the reinforcing bars is input into the rule library for analysis, and the defect status judgment and defect location are output.

[0105] (4) Results display and feedback: Construct a tool for displaying and interacting with the detection results of steel bar processing defects. Display the defect detection results in a visual form and generate defect reports. The detection results can be manually reviewed and adjusted. The adjusted results can be fed back to the training set in real time to dynamically improve the recognition accuracy.

[0106] This application proposes a quality inspection system for the rebar processing of railway tunnel invert blocks. By integrating algorithms such as machine vision, image processing, and deep learning, it expands the existing rebar processing image data of the railway engineering management platform to form a railway rebar processing training set and establishes an image preprocessing module. A general framework for a rebar position recognition model is established, and the model is trained and optimized using the railway rebar processing training set to form a rebar position recognition model. On-site photography is conducted using instruments equipped with imaging and laser ranging. After image preprocessing, the images are input into the rebar position recognition model to generate rebar position information. Combined with ranging information, the rebar spacing, density, and length are calculated to form comprehensive morphological data of the invert block rebar. A defect rule library for tunnel invert rebar is established. The comprehensive morphological data of the rebar is input into the rule library for analysis, and the system outputs defect judgments and marks defect locations. The recognition results are displayed and a defect report is generated for manual review. If there are inaccuracies in the recognition, manual adjustments can be made and feedback can be sent back to the railway rebar processing training set.

[0107] Compared with existing technologies, the beneficial effects of the application examples in this application are:

[0108] This application proposes a method for inspecting the processing quality of inverted arch reinforcement bars. Employing machine vision, neural networks, and other technologies, the method intelligently identifies the processing status of reinforcement bars in railway tunnel inverted arch structural blocks captured on-site. It extracts parameters such as the quantity, spacing, and dimensions of the reinforcement bars and identifies defects in these parameters. This enables automatic assessment of the processing quality of the railway tunnel inverted arch structural blocks' reinforcement bars, allowing for timely detection of quality problems during processing. Corrective measures can then be taken and optimized before concrete pouring, thereby improving the prefabrication quality of the tunnel inverted arch structural blocks.

[0109] This application also provides an electronic device, which may include a processor, a memory, a receiver, and a transmitter. The processor is used to execute the inverted arch reinforcement processing quality inspection method mentioned in the above embodiments. The processor and memory can be connected via a bus or other means, taking a bus connection as an example. The receiver can be connected to the processor and memory via wired or wireless means. The electronic device can receive real-time motion data from sensors in the wireless multimedia sensor network and receive raw video sequences from the video acquisition device.

[0110] The processor can be a central processing unit (CPU). The processor can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or combinations of the above types of chips.

[0111] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the inverted arch rebar processing quality inspection method in the embodiments of this application. The processor executes various functional applications and data processing by running the non-transitory software programs, instructions, and modules stored in the memory, thereby implementing the inverted arch rebar processing quality inspection method in the above method embodiments.

[0112] The memory may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created by the processor, etc. Furthermore, the memory may include high-speed random access memory and non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, the memory may optionally include memory remotely located relative to the processor, which can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0113] The one or more modules are stored in the memory, and when executed by the processor, the method for detecting the processing quality of the inverted arch reinforcement in the embodiment is executed.

[0114] In some embodiments of this application, the user equipment may include a processor, a memory, and a transceiver unit. The transceiver unit may include a receiver and a transmitter. The processor, memory, receiver, and transmitter may be connected via a bus system. The memory is used to store computer instructions, and the processor is used to execute the computer instructions stored in the memory to control the transceiver unit to send and receive signals.

[0115] As one implementation method, the functions of the receiver and transmitter in this application can be implemented by transceiver circuits or dedicated transceiver chips, and the processor can be implemented by dedicated processing chips, processing circuits or general-purpose chips.

[0116] As another implementation approach, the server provided in this application embodiment can be implemented using a general-purpose computer. That is, the program code implementing the processor, receiver, and transmitter functions is stored in memory, and the general-purpose processor implements the processor, receiver, and transmitter functions by executing the code in memory.

[0117] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the aforementioned method for detecting the processing quality of inverted arch reinforcement. The computer-readable storage medium can be a tangible storage medium, such as random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, floppy disks, hard disks, removable storage disks, CD-ROMs, or any other form of storage medium known in the art.

[0118] Those skilled in the art will understand that the exemplary components, systems, and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software, or a combination of both. Whether implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can 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. When implemented in hardware, it can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. The programs or code segments can be stored in a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave.

[0119] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0120] In this application, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or in place of features of other embodiments.

[0121] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to the embodiments of this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for detecting the processing quality of inverted arch steel bars, characterized in that, The method comprises the following steps: Collect the steel bar processing image corresponding to the target steel bar structure of the prefabricated railway tunnel inverted arch structure block; at the same time, collect the distance information of the target steel bar structure; Preprocess the steel bar processing image of the target steel bar structure based on a preset preprocessing method to obtain a corresponding target inverted arch steel bar processing image; Input the target inverted arch steel bar processing image into a preset steel bar position recognition model to make the steel bar position recognition model output the steel bar position information corresponding to the target inverted arch steel bar processing image; Determine the target steel bar form data corresponding to the target steel bar structure for detecting the steel bar processing quality according to the steel bar position information corresponding to the target inverted arch steel bar processing image and the distance information, wherein the target steel bar form data includes the number of steel bars, the steel bar spacing and the length of steel bars; the number of steel bars is obtained by using a Yolov5s target detection network for identification; the color image stream obtained by the depth camera is aligned with the depth stream, so that each pixel on the color image corresponds to a depth value as the z coordinate, and then the x coordinate and y coordinate of the pixel are obtained through the camera internal parameter to obtain the three-dimensional coordinates of the pixel in the camera coordinate system; the three-dimensional coordinates of the center points of the two measured steel bars are obtained by using this method, and then the distance between the two steel bar center points is obtained by using the Pythagorean theorem, that is, the steel bar spacing; Detect the processing quality of the target steel bar form data according to the steel bar defect rule library for storing the corresponding relationship between the steel bar form and the processing quality defect to obtain the steel bar processing quality detection result corresponding to the target steel bar structure; wherein the steel bar defect rule library is pre-established to store the corresponding relationship between the steel bar form data and the processing quality defect data, wherein the steel bar defect rule library is realized by using a structured data table, and the content of the structured data table includes image information data, steel bar form data and defect annotation data; the image information data includes image number, image storage location, production work point number and annotation personnel number; the defect annotation data includes defect type and defect degree; and the steel bar processing quality detection result is outputted to visually display the steel bar processing quality detection result.

2. The method for detecting the quality of the inverted arch steel bar processing according to claim 1, characterized in that, Further comprising: Obtain a historical inverted arch steel bar processing image dataset, wherein the historical inverted arch steel bar processing image dataset includes steel bar processing images of various steel bar structures of the prefabricated railway tunnel inverted arch structure block stored in a railway engineering management platform, and open source steel bar processing images collected from outside the railway engineering management platform; Perform expansion processing on the historical inverted arch steel bar processing image dataset, and preprocess the historical inverted arch steel bar processing image dataset based on a preset preprocessing method; Respectively perform steel bar position annotation on each steel bar processing image in the historical inverted arch steel bar processing image dataset to form a corresponding steel bar processing training set; Train a preset machine learning model by using the steel bar processing training set to obtain a corresponding steel bar position recognition model.

3. The method for detecting the quality of the inverted arch steel bar processing according to claim 2, characterized in that, The machine learning model comprises a deep residual network based on an attention mechanism.

4. The method for detecting the quality of the inverted arch steel bar processing according to claim 1 or 2, characterized in that, The preprocessing mode comprises at least one of image light field correction, image white balance processing, image registration, image noise reduction and image edge enhancement.

5. A device for detecting the quality of processing of inverted arch steel reinforcement, characterized in that, Comprise: The image acquisition module is used for collecting a steel bar processing image corresponding to a target steel bar structure of a prefabricated railway tunnel inverted arch structure block; simultaneously, distance information of the target steel bar structure is collected; The target steel bar processing image of the target steel bar structure is preprocessed based on a preset preprocessing mode, to obtain a corresponding target inverted arch steel bar processing image; The model identification module is used for inputting the target inverted arch steel bar processing image into a preset steel bar position identification model, so that the steel bar position identification model outputs steel bar position information corresponding to the target inverted arch steel bar processing image; The inverted arch steel bar processing quality detection device is further used for executing the following content: According to the steel bar position information corresponding to the target inverted arch steel bar processing image and the distance information, target steel bar shape data corresponding to the target steel bar structure for detecting steel bar processing quality is determined, wherein the target steel bar shape data comprises a steel bar quantity, a steel bar spacing and a steel bar length; the steel bar quantity is obtained by using a Yolov5s target detection network for identification; a color image stream obtained by a depth camera is aligned with a depth stream, so that each pixel on the color image corresponds to a depth value as a z coordinate, and then x and y coordinates of the pixel are obtained through camera internal parameters, to obtain three-dimensional coordinates of the pixel in a camera coordinate system; three-dimensional coordinates of two steel bar center points are obtained by using this method, and then the distance between the two steel bar center points is obtained by using the Pythagorean theorem, that is, the steel bar spacing; According to a steel bar defect rule library for storing a corresponding relationship between steel bar shape and processing quality defects, the target steel bar shape data is subjected to processing quality detection, to obtain steel bar processing quality detection results corresponding to the target steel bar structure; wherein the steel bar defect rule library is pre-established and is used for storing a corresponding relationship between steel bar shape data and processing quality defect data, wherein the steel bar defect rule library is realized by using a structured data table, and content of the structured data table comprises image information data, steel bar shape data and defect annotation data; the image information data comprises an image number, an image storage position, a production work point number and an annotation personnel number; the defect annotation data comprises a defect type and a defect degree; The steel bar processing quality detection results are outputted to visually display the steel bar processing quality detection results.

6. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to realize the inverted arch steel bar processing quality detection method according to any one of claims 1 to 4.

7. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the inverted arch steel bar processing quality detection method according to any one of claims 1 to 4.

Images