Underground pipe gallery microcrack detection and measurement method, device, equipment, medium and product

Abstract

The invention discloses an underground pipe gallery microcrack detection and measurement method and device, equipment, a medium and a product, and relates to the technical field of underground pipe gallery body crack detection, and the method comprises the steps: inputting an RGB image of a pipe gallery into a target detection model, and obtaining a plurality of regions containing cracks; carrying out edge detection and contour extraction operation on each crack-containing area to obtain a contour diagram of each crack; obtaining three-dimensional point cloud data of the pipe gallery based on the depth map and the RGB image, and sequentially carrying out point cloud density analysis and local curvature calculation on the three-dimensional point cloud data of the pipe gallery; constructing a three-dimensional point cloud model of the cracks based on the three-dimensional point cloud data of the pipe gallery, the contour map of each crack, the point cloud density analysis result and the curvature of the three-dimensional point cloud data; and processing the three-dimensional point cloud model of the crack through a principal component analysis algorithm to obtain the size of each crack. According to the invention, the accuracy of micro-crack detection and size measurement results can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of underground pipeline corridor crack detection, and in particular to a method, device, equipment, medium and product for detecting and measuring micro-cracks in underground pipeline corridors. Background Art

[0002] Traditional methods for detecting and measuring cracks in pipeline corridors mainly include manual inspections and crack meter measurements. These methods have the advantages of simple operation and low cost, but have disadvantages such as low monitoring efficiency, low data accuracy, and inability to monitor in real time. Traditional methods are also affected by environmental, climate, and human factors, making it difficult to ensure the accuracy and reliability of monitoring data.

[0003] With the advancement of science and technology, modern monitoring technologies have been widely used in pipeline corridor crack monitoring, mainly including digital imaging technology, fiber optic sensing technology, infrared thermal imaging technology, etc. However, these technologies do not produce accurate results when detecting and measuring microcracks. Summary of the Invention

[0004] The purpose of this application is to provide a method, device, equipment, medium and product for detecting and measuring microcracks in underground pipe corridors, which can improve the accuracy of microcrack detection and size measurement results.

[0005] To achieve the above objectives, this application provides the following solutions:

[0006] In a first aspect, the present application provides a method for detecting and measuring microcracks in an underground pipe gallery, comprising:

[0007] Obtain RGB images and depth images of the pipeline gallery;

[0008] Inputting the RGB image into a target detection model to obtain multiple areas containing cracks; the target detection model is obtained by first training a YOLOv8 model using a general crack dataset, and then training the trained YOLOv8 model using an underground pipeline gallery crack dataset;

[0009] Perform edge detection and contour extraction operations on each region containing cracks to obtain a contour map of each crack;

[0010] The 3D point cloud data of the pipeline corridor is obtained based on the depth map and RGB image. The point cloud density analysis and local curvature calculation are performed on the 3D point cloud data of the pipeline corridor in sequence to obtain the point cloud density analysis results and the curvature of the 3D point cloud data;

[0011] The three-dimensional point cloud data of the cracks are obtained based on the three-dimensional point cloud data of the pipeline corridor and the contour map of each crack;

[0012] A 3D point cloud model of the crack is constructed based on the 3D point cloud data of the crack, the point cloud density analysis results and the curvature of the 3D point cloud data;

[0013] The three-dimensional point cloud model of the cracks is processed using the principal component analysis algorithm to obtain the size of each crack.

[0014] In one embodiment, edge detection and contour extraction operations are performed on each region containing cracks to obtain a contour map of each crack, specifically including:

[0015] For any region containing a crack, edge detection is performed on the region to obtain a first image;

[0016] Processing the first image using an adaptive threshold method and a Canny edge detection algorithm to obtain a second image;

[0017] Perform a contour extraction operation on the second image to obtain a contour map of the cracks in the area.

[0018] In one embodiment, the first image is processed using an adaptive threshold method and a Canny edge detection algorithm to obtain a second image, specifically including:

[0019] Processing the first image using an adaptive threshold method to obtain a threshold;

[0020] The first image is processed using a Canny edge detection algorithm based on the threshold to obtain a second image.

[0021] In one embodiment, the RGB image is input into an object detection model to obtain multiple regions containing cracks, specifically including:

[0022] Performing contrast adjustment, noise elimination, and image normalization operations on the RGB image in sequence to obtain a preprocessed image;

[0023] The preprocessed image is input into the target detection model to obtain multiple areas containing cracks.

[0024] In one embodiment, the size of the crack includes: the maximum length, average width, and maximum depth of the crack. The size of each crack is obtained by processing the three-dimensional point cloud model of the crack using a principal component analysis algorithm, specifically including:

[0025] The three-dimensional point cloud model of the crack is processed by the principal component analysis algorithm to obtain the main direction axis, secondary direction axis and normal direction axis;

[0026] Project the three-dimensional point cloud model of the crack onto the main direction axis and calculate the maximum length of each crack;

[0027] Project the three-dimensional point cloud model of the crack onto the secondary direction axis and calculate the average width of each crack;

[0028] The three-dimensional point cloud model of the crack is projected onto the normal axis, and the maximum depth of each crack is calculated.

[0029] In one embodiment, obtaining three-dimensional point cloud data of the pipeline corridor based on the depth map and the RGB image specifically includes:

[0030] Obtain the initial 3D point cloud data of the pipeline corridor based on the depth map and RGB image;

[0031] The initial three-dimensional point cloud data of the pipeline corridor is optimized to obtain the three-dimensional point cloud data of the pipeline corridor.

[0032] In a second aspect, the present application provides an underground pipe gallery micro-crack detection and measurement device, comprising:

[0033] Input module, used to obtain RGB images and depth images of the pipeline corridor;

[0034] A crack detection module is configured to input the RGB image into a target detection model to obtain multiple regions containing cracks, and perform edge detection and contour extraction on each region containing cracks to obtain a contour map of each crack; the target detection model is obtained by first training a YOLOv8 model using a general crack dataset, and then training the trained YOLOv8 model using an underground pipeline corridor crack dataset;

[0035] The three-dimensional measurement module is used to obtain the three-dimensional point cloud data of the pipeline corridor based on the depth map and RGB image, perform point cloud density analysis and local curvature calculation on the three-dimensional point cloud data of the pipeline corridor in sequence, and obtain the point cloud density analysis results and the curvature of the three-dimensional point cloud data; obtain the three-dimensional point cloud data of the cracks based on the three-dimensional point cloud data of the pipeline corridor and the contour map of each crack; construct a three-dimensional point cloud model of the cracks based on the three-dimensional point cloud data of the cracks, the point cloud density analysis results and the curvature of the three-dimensional point cloud data; and process the three-dimensional point cloud model of the cracks using the principal component analysis algorithm to obtain the size of each crack.

[0036] In a third aspect, the present application provides a computer device comprising: 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 any of the above-mentioned methods for detecting and measuring microcracks in underground pipelines.

[0037] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned methods for detecting and measuring microcracks in underground pipe galleries.

[0038] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the underground pipeline corridor microcrack detection and measurement method described in any one of the above items.

[0039] According to the specific embodiments provided in this application, this application has the following technical effects:

[0040] The present application provides a method, apparatus, equipment, medium, and product for detecting and measuring microcracks in underground pipe corridors. A YOLOv8 model is first trained using a general crack dataset, and then the trained YOLOv8 model is trained using the underground pipe corridor crack dataset to obtain a target detection model. This can improve the target detection model's accuracy in detecting cracks, enhance the ability to express microcrack features through point cloud density analysis and local curvature calculation, and thus improve the accuracy of microcrack detection and size measurement results. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] 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. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0042] Figure 1 A schematic flow chart of a method for detecting and measuring microcracks in an underground pipe gallery provided in one embodiment of the present application;

[0043] Figure 2 This is a structural block diagram of an underground pipe gallery micro-crack detection and measurement device provided in one embodiment of the present application;

[0044] Figure 3 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0045] 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.

[0046] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0047] In an exemplary embodiment, a method for detecting and measuring micro cracks in an underground pipe gallery is provided. Figure 1 As shown, the process includes the following steps 201 to 207. In which:

[0048] Step 201: Acquire an RGB image and a depth image of the pipe gallery. Specifically, the RGB image and the depth image may be acquired using a RealSense camera.

[0049] Step 202: Input the RGB image into an object detection model to obtain multiple crack-containing regions. The object detection model is obtained by first training a YOLOv8 model using a general crack dataset and then training the trained YOLOv8 model using an underground pipeline gallery crack dataset. The general crack dataset includes various crack images. The underground pipeline gallery crack dataset contains images of pipeline gallery cracks from multiple angles and scales, enhancing the model's ability to learn crack features.

[0050] Step 203: performing edge detection and contour extraction operations on each region containing cracks to obtain a contour map of each crack.

[0051] Step 204: Obtain three-dimensional point cloud data of the pipeline corridor based on the depth map and the RGB image, perform point cloud density analysis and local curvature calculation on the three-dimensional point cloud data of the pipeline corridor in sequence, and obtain point cloud density analysis results and curvature of the three-dimensional point cloud data.

[0052] Step 205: Obtain three-dimensional point cloud data of the cracks based on the three-dimensional point cloud data of the pipeline corridor and the contour map of each crack.

[0053] Step 206: constructing a three-dimensional point cloud model of the crack based on the three-dimensional point cloud data of the crack, the point cloud density analysis result, and the curvature of the three-dimensional point cloud data.

[0054] Step 207: Process the three-dimensional point cloud model of the cracks using a principal component analysis algorithm to obtain the size of each crack.

[0055] In practical applications, the object detection model is developed by training a pre-set neural network based on transfer learning. Transfer learning enhances the model's generalization capabilities and enables recognition. The pre-set neural network is a YOLOv8 model, which includes a CSP backbone network (Backbone), a Feature Pyramid Network (FPN), and an SPPF spatial feature enhancement module, effectively preventing information loss in deep networks.

[0056] The specific training process is to first complete the initial training on a general crack dataset. However, the model at this time is not very effective in identifying cracks in underground pipeline corridors. Then, the underground pipeline corridor crack dataset is used to fine-tune the model after initial training. This application uses the underground pipeline corridor crack dataset to fine-tune the model while retaining the model's underlying feature extraction capabilities. This significantly improves the model's recognition ability for specific scenarios, thereby obtaining a model with better performance in the task of identifying cracks in underground pipeline corridors. This model is more effective in identifying cracks in underground pipeline corridors. Through the transfer learning strategy, the model can quickly adapt to the crack detection task on a smaller professional dataset (underground pipeline corridor crack dataset), significantly improving detection accuracy.

[0057] In another exemplary embodiment of the present application, edge detection and contour extraction operations are performed on each region containing cracks to obtain a contour map of each crack, specifically including:

[0058] For any region containing cracks, edge detection is performed on the region to obtain a first image.

[0059] The first image is processed using an adaptive threshold method and a Canny edge detection algorithm to obtain a second image.

[0060] Perform a contour extraction operation on the second image to obtain a contour map of the cracks in the area.

[0061] In another exemplary embodiment of the present application, the first image is processed using an adaptive threshold method and a Canny edge detection algorithm to obtain a second image, specifically including:

[0062] The first image is processed using an adaptive threshold method to obtain a threshold.

[0063] The first image is processed based on the threshold using a Canny edge detection algorithm to obtain a second image. The precise edge contour of the crack is extracted using the Canny edge detection algorithm.

[0064] In another exemplary embodiment of the present application, after performing edge detection and contour extraction operations on each area containing cracks to obtain the contour map of each crack, it also includes: performing morphological processing on the extracted crack contour image. Specifically, isolated noise points are first removed, and then the broken crack edges are connected through the expansion operation. If necessary, the crack contour structure can be further smoothed and improved by combining the opening operation or the closing operation. After morphological processing, the crack contour is more continuous and complete, which is conducive to the subsequent accurate drawing and annotation of the detected crack area on the original image. The present application further processes each area within the bounding box detected by the target detection model, applies the edge detection algorithm to extract the crack contour in the area of ​​interest, enhances the crack contour through morphological processing, and draws the detected crack contour on the original image.

[0065] In another exemplary embodiment of the present application, the adaptive threshold method is the OTSU algorithm. In Canny edge detection, there is another important operation step: dual-threshold hysteresis threshold processing, a high threshold TH and a low threshold TL, with a ratio of 2:1 to 3:1. This brings a major disadvantage of Canny edge detection, which cannot better adapt to the crack edges of underground pipeline corridors. Therefore, the present application uses the OTSU algorithm to process the image, and uses the grayscale features of the image to calculate the optimal threshold to maximize the inter-class variance between the target and background components. First, the OTSU algorithm extracts the image histogram, calculates the cumulative sum of the grayscale values, and calculates the mean of the grayscale values. Then, when the grayscale level in the image is less than the k value, the probability is recorded as p1, and when the grayscale level is greater than the k value, the probability is recorded as p2. Finally, the class variance is calculated, and k is the optimal threshold when the class variance takes the maximum value. After obtaining the optimal threshold k, the image is binarized to separate the target object from the background.

[0066] In another exemplary embodiment of the present application, the RGB image is input into the target detection model to obtain multiple regions containing cracks, specifically including:

[0067] The RGB image is sequentially subjected to contrast adjustment, noise removal, and image normalization to obtain a preprocessed image. Specifically, contrast adjustment is first achieved by using CLAHE (Contrast Limited Adaptive Histogram Equalization) to enhance image contrast. A bilateral filter is then used to reduce noise while preserving edge detail to achieve noise removal. Finally, image normalization is performed to convert the grayscale image into a binary image for subsequent contour extraction, achieving image normalization.

[0068] The preprocessed image is input into the target detection model to obtain multiple areas containing cracks.

[0069] In another exemplary embodiment of the present application, obtaining three-dimensional point cloud data of a pipeline corridor based on a depth map and an RGB image specifically includes:

[0070] The initial three-dimensional point cloud data of the pipeline corridor is obtained based on the depth map and RGB image. Specifically, the depth image is preprocessed, including depth map conversion, unit conversion and noise filtering, to remove outliers (such as invalid depth, extreme values) and noise points to ensure the accuracy and reliability of the depth data. Then, the RGB image and the preprocessed depth image are aligned to ensure that the RGB image and the depth image are spatially aligned (that is, the same pixel corresponds to the same spatial position in the two images). The aligned image is then scaled and channel-matched. Then, for each pixel on the channel-matched image, its two-dimensional coordinates (u, v), depth value D, and corresponding RGB value (R, G, B) are obtained. Using the camera intrinsic parameters, (u, v, D) is converted into three-dimensional spatial coordinates (X, Y, Z). Based on (X, Y, Z, R, G, B), the colored initial point cloud data is generated, i.e., the initial three-dimensional point cloud data of the pipeline corridor. This step processes the RGB image and depth information, and maps the depth information back to three-dimensional space (combining the two-dimensional pixel coordinates with the corresponding depth values ​​and converting them into three-dimensional space coordinates to generate initial point cloud data, and combining the depth information to convert the 2D edge points into 3D point cloud data).

[0071] The initial three-dimensional point cloud data of the utility corridor is optimized to obtain the three-dimensional point cloud data of the utility corridor. Specifically, the initial three-dimensional point cloud data is optimized, including statistical filtering to remove outliers, voxel downsampling to reduce data volume, and voxel center calculation to improve accuracy, thereby finally obtaining the three-dimensional point cloud data of the utility corridor.

[0072] In another exemplary embodiment of the present application, point cloud density analysis and local curvature calculation are performed on the three-dimensional point cloud data of the pipeline corridor in sequence to obtain point cloud density analysis results and curvature of the three-dimensional point cloud data, specifically:

[0073] Perform point cloud density analysis on the three-dimensional point cloud data of the pipeline corridor, analyze the distribution density of the point cloud in space, and identify areas with sparse or abnormally dense point clouds. These areas often correspond to important features such as microcracks or crack intersections. Obtain point cloud density distribution maps or density feature parameters to provide a basis for subsequent feature enhancement and crack identification.

[0074] The local curvature calculation is performed on the point cloud data that has undergone point cloud density analysis. The local curvature of each point in the point cloud is calculated to highlight the tiny structural features such as crack edges and corners, further enhance the expression ability of microcracks, obtain the curvature value distribution of each point, and form a curvature feature map. Based on the curvature feature map, the crack identification results can be more refined. Density analysis can assist in locating the microcrack area, and curvature calculation further highlights the tiny structural features.

[0075] In another exemplary embodiment of the present application, three-dimensional point cloud data of the cracks are obtained based on the three-dimensional point cloud data of the pipeline corridor and the contour map of each crack. Specifically, the coordinates of each crack in the map are obtained based on the contour map of each crack to obtain a coordinate set, and the three-dimensional point cloud data corresponding to each coordinate in the coordinate set is obtained in the three-dimensional point cloud data of the pipeline corridor to obtain the three-dimensional point cloud data of the cracks.

[0076] In another exemplary embodiment of the present application, a three-dimensional point cloud model of the crack is constructed based on the three-dimensional point cloud data of the crack, the point cloud density analysis results and the curvature of the three-dimensional point cloud data. Specifically, three-dimensional feature extraction of the crack is performed, three-dimensional geometric features of the crack are extracted based on the curvature of the three-dimensional point cloud data and the point cloud density analysis results, the three-dimensional characteristics of the crack are analyzed, a three-dimensional point cloud model is constructed, and crack surface reconstruction and three-dimensional visualization are achieved.

[0077] In another exemplary embodiment of the present application, the dimensions of the cracks include: the maximum length, average width, and maximum depth of the cracks. The dimensions of each crack are obtained by processing the three-dimensional point cloud model of the cracks using a principal component analysis algorithm, specifically including:

[0078] The three-dimensional point cloud model of the crack is processed by the principal component analysis algorithm to obtain the main direction axis (the main extension direction of the crack), the secondary direction axis (the width direction of the crack) and the normal direction axis (the depth direction of the crack).

[0079] The 3D point cloud model is projected onto the main direction axis and the maximum length of each crack is calculated.

[0080] The 3D point cloud model is projected onto the secondary direction axis, and the average width of each crack is calculated. Specifically, the 3D point cloud model is projected onto the secondary direction axis, and after removing outliers through statistical analysis, the average width of the crack is calculated.

[0081] The 3D point cloud model is projected onto the normal axis to calculate the maximum depth of each crack.

[0082] This application combines RGB images and depth images, and realizes cross-modal fusion of image features and depth information through feature alignment, scale normalization and channel matching, thereby completing the final crack detection and size measurement, effectively preventing the problem of information loss in deep networks.

[0083] When training the YOLOv8 model, this application completes initial training using a general crack dataset. While retaining the model's underlying feature extraction capabilities, the underground pipeline gallery crack dataset is used to fine-tune the model to achieve knowledge transfer. Through the transfer learning strategy, the model can quickly adapt to crack detection tasks on smaller professional datasets (underground pipeline gallery crack datasets), significantly improving detection accuracy.

[0084] Based on the same inventive concept, the embodiments of the present application also provide an underground pipe gallery microcrack detection and measurement device for implementing the above-mentioned underground pipe gallery microcrack detection and measurement method. The implementation solution provided by this device is similar to the implementation solution described in the above-mentioned method. Therefore, the specific limitations of one or more embodiments of the underground pipe gallery microcrack detection and measurement device provided below can be found in the above-mentioned limitations of the underground pipe gallery microcrack detection and measurement method, and will not be repeated here.

[0085] In an exemplary embodiment, Figure 2 As shown, a device for detecting and measuring micro cracks in an underground pipe gallery is provided, comprising:

[0086] The input module is used to obtain the RGB image and depth image of the pipeline corridor.

[0087] The crack detection module is used to input the RGB image into a target detection model to obtain multiple areas containing cracks, and perform edge detection and contour extraction operations on each area containing cracks to obtain a contour map of each crack; the target detection model is obtained by first training a YOLOv8 model using a general crack dataset, and then training the trained YOLOv8 model using an underground pipeline corridor crack dataset.

[0088] The three-dimensional measurement module is used to obtain the three-dimensional point cloud data of the pipeline corridor based on the depth map and RGB image, perform point cloud density analysis and local curvature calculation on the three-dimensional point cloud data of the pipeline corridor in sequence, and obtain the point cloud density analysis results and the curvature of the three-dimensional point cloud data; obtain the three-dimensional point cloud data of the cracks based on the three-dimensional point cloud data of the pipeline corridor and the contour map of each crack; construct a three-dimensional point cloud model of the cracks based on the three-dimensional point cloud data of the cracks, the point cloud density analysis results and the curvature of the three-dimensional point cloud data; and process the three-dimensional point cloud model of the cracks using the principal component analysis algorithm to obtain the size of each crack.

[0089] As an optional implementation, the underground pipe gallery micro-crack detection and measurement device further includes: an image preprocessing module for performing contrast adjustment, noise elimination, and image normalization operations on the RGB image in sequence.

[0090] As an optional implementation, the crack detection module includes: a target detection unit, an edge detection unit, an adaptive threshold method unit, a Canny edge detection unit, and a contour extraction and analysis unit.

[0091] The target detection unit is used to input the RGB image into a target detection model to obtain multiple areas containing cracks.

[0092] The edge detection unit is used to perform edge detection on any region containing cracks to obtain a first image.

[0093] The adaptive threshold method unit is used to process the first image using the adaptive threshold method to obtain a threshold.

[0094] A Canny edge detection unit is configured to process the first image using a Canny edge detection algorithm based on the threshold to obtain a second image.

[0095] The contour extraction and analysis unit is used to perform a contour extraction operation on the second image to obtain a contour map of the cracks in the area.

[0096] As an optional implementation, the three-dimensional measurement module includes: a point cloud generation and processing unit, a target three-dimensional feature extraction unit and a crack size calculation unit.

[0097] The point cloud generation and processing unit is used to obtain initial three-dimensional point cloud data of the pipeline corridor based on the depth map and the RGB image; optimize the initial three-dimensional point cloud data of the pipeline corridor to obtain three-dimensional point cloud data of the pipeline corridor; and perform point cloud density analysis and local curvature calculation on the three-dimensional point cloud data of the pipeline corridor in sequence to obtain point cloud density analysis results and the curvature of the three-dimensional point cloud data.

[0098] The target three-dimensional feature extraction unit is used to obtain three-dimensional point cloud data of the crack based on the three-dimensional point cloud data of the pipeline corridor and the contour map of each crack; and to construct a three-dimensional point cloud model of the crack based on the three-dimensional point cloud data of the crack, the point cloud density analysis results and the curvature of the three-dimensional point cloud data.

[0099] The crack size calculation unit is used to process the three-dimensional point cloud model of the cracks through a principal component analysis algorithm to obtain the size of each crack.

[0100] The 3D measurement module is used to analyze and measure the three-dimensional spatial features of cracks. It specifically includes depth data processing, reading and processing depth images, creating custom depth frame objects to process depth maps of different formats, and applying depth filtering algorithms to remove outliers and noise points. The point cloud generation and processing unit generates 3D point cloud data based on RGB images and depth maps, uses the Open3D library to process and analyze point cloud data, and implements point cloud optimization, denoising, and smoothing.

[0101] This application inputs the RGB image into the CSP backbone network of the target detection model for feature extraction, ultimately obtaining a multi-scale feature map. This process then enters the feature processing process, which first processes the acquired multi-scale feature map through the C2f residual module, then uses the SPPF spatial feature enhancement module to enhance spatial features, and then implements multi-scale feature fusion through the Feature Pyramid Network (FPN). After feature processing is completed, the extracted contour is fused with the previously aligned depth information, and finally the depth value of the detected target is read for subsequent 3D measurement.

[0102] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O) and a communication interface. The processor, memory and input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store underground pipeline corridor microcrack detection and measurement data. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for detecting and measuring underground pipeline corridor microcracks is implemented.

[0103] Those skilled in the art will understand that Figure 3 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present application and does not constitute a limitation on the computer device to which the solution of the present application is applied. A specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor implements the above-mentioned method embodiments when executing the computer program.

[0104] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, which implements the above-mentioned method embodiments when executed by a processor.

[0105] In an exemplary embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the above method embodiments are implemented.

[0106] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0107] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, database or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory may include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0108] The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processors involved in the various embodiments provided herein may include, but are not limited to, general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic units, data processing logic units based on quantum computing, and the like.

[0109] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0110] This document uses specific examples to illustrate the principles and implementation methods of this application. The description of the above examples is only intended to help understand the method and core concept of this application. At the same time, for those skilled in the art, based on the concept of this application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting this application.

Claims

1. A method for detecting and measuring microcracks in an underground pipe gallery, characterized in that: The underground pipe gallery microcrack detection and measurement method includes: Obtain RGB images and depth images of the pipeline gallery; Inputting the RGB image into a target detection model to obtain multiple areas containing cracks; the target detection model is obtained by first training a YOLOv8 model using a general crack dataset, and then training the trained YOLOv8 model using an underground pipeline gallery crack dataset; Perform edge detection and contour extraction operations on each region containing cracks to obtain a contour map of each crack; The 3D point cloud data of the pipeline corridor is obtained based on the depth map and RGB image. The point cloud density analysis and local curvature calculation are performed on the 3D point cloud data of the pipeline corridor in sequence to obtain the point cloud density analysis results and the curvature of the 3D point cloud data; The three-dimensional point cloud data of the cracks are obtained based on the three-dimensional point cloud data of the pipeline corridor and the contour map of each crack; A 3D point cloud model of the crack is constructed based on the 3D point cloud data of the crack, the point cloud density analysis results and the curvature of the 3D point cloud data; The three-dimensional point cloud model of the cracks is processed using the principal component analysis algorithm to obtain the size of each crack.

2. The underground pipe gallery microcrack detection and measurement method according to claim 1 is characterized in that: Perform edge detection and contour extraction on each crack-containing region to obtain the contour map of each crack, including: For any region containing a crack, edge detection is performed on the region to obtain a first image; Processing the first image using an adaptive threshold method and a Canny edge detection algorithm to obtain a second image; Perform a contour extraction operation on the second image to obtain a contour map of the cracks in the area.

3. The underground pipe gallery micro-crack detection and measurement method according to claim 2 is characterized in that: The first image is processed using an adaptive threshold method and a Canny edge detection algorithm to obtain a second image, specifically including: Processing the first image using an adaptive threshold method to obtain a threshold; The first image is processed using a Canny edge detection algorithm based on the threshold to obtain a second image.

4. The underground pipe gallery microcrack detection and measurement method according to claim 1 is characterized in that: The RGB image is input into the target detection model to obtain multiple areas containing cracks, specifically including: Performing contrast adjustment, noise elimination, and image normalization operations on the RGB image in sequence to obtain a preprocessed image; The preprocessed image is input into the target detection model to obtain multiple areas containing cracks.

5. The underground pipe gallery micro-crack detection and measurement method according to claim 1 is characterized in that: The dimensions of the cracks include: maximum length, average width, and maximum depth. The principal component analysis algorithm is used to process the three-dimensional point cloud model of the cracks to obtain the dimensions of each crack, including: The three-dimensional point cloud model of the crack is processed by the principal component analysis algorithm to obtain the main direction axis, secondary direction axis and normal direction axis; Project the three-dimensional point cloud model of the crack onto the main direction axis and calculate the maximum length of each crack; Project the three-dimensional point cloud model of the crack onto the secondary direction axis and calculate the average width of each crack; The three-dimensional point cloud model of the crack is projected onto the normal axis, and the maximum depth of each crack is calculated.

6. The underground pipe gallery micro-crack detection and measurement method according to claim 1 is characterized in that: The three-dimensional point cloud data of the pipeline corridor is obtained based on the depth map and RGB image, including: Obtain the initial 3D point cloud data of the pipeline corridor based on the depth map and RGB image; The initial three-dimensional point cloud data of the pipeline corridor is optimized to obtain the three-dimensional point cloud data of the pipeline corridor.

7. An underground pipe gallery micro-crack detection and measurement device, characterized in that: The underground pipe gallery micro-crack detection and measurement device includes: Input module, used to obtain RGB images and depth images of the pipeline corridor; A crack detection module is configured to input the RGB image into a target detection model to obtain multiple regions containing cracks, and perform edge detection and contour extraction on each region containing cracks to obtain a contour map of each crack; the target detection model is obtained by first training a YOLOv8 model using a general crack dataset, and then training the trained YOLOv8 model using an underground pipeline corridor crack dataset; The three-dimensional measurement module is used to obtain the three-dimensional point cloud data of the pipeline corridor based on the depth map and RGB image, perform point cloud density analysis and local curvature calculation on the three-dimensional point cloud data of the pipeline corridor in sequence, and obtain the point cloud density analysis results and the curvature of the three-dimensional point cloud data; obtain the three-dimensional point cloud data of the cracks based on the three-dimensional point cloud data of the pipeline corridor and the contour map of each crack; construct a three-dimensional point cloud model of the cracks based on the three-dimensional point cloud data of the cracks, the point cloud density analysis results and the curvature of the three-dimensional point cloud data; and process the three-dimensional point cloud model of the cracks using the principal component analysis algorithm to obtain the size of each crack.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and runnable on the processor, characterized in that the processor executes the computer program to implement the underground pipeline corridor microcrack detection and measurement method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the underground pipe gallery microcrack detection and measurement method described in any one of claims 1 to 6 is implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the underground pipe gallery microcrack detection and measurement method described in any one of claims 1 to 6 is implemented.

Images