A linear management method, device and medium for assembled bridge

Abstract

The present application discloses a linear management method, device and medium for an assembled bridge; it relates to the field of bridges and solves the problem that existing detection schemes rely on manual operation. The present application first obtains rich image data by shooting from multiple angles. Then, the image is processed by target detection and ellipse detection algorithms to identify and locate the target. Next, the camera pose and the three-dimensional coordinates of the marker points are determined by a five-point algorithm and triangulation. Finally, the linear result of the bridge is obtained by spatial splicing and linear fitting. The linear result can provide a basis for construction or maintenance during the construction or commissioning of the bridge; the present application only needs to use the image data containing the target to efficiently calculate the linear features of the current state of the bridge. It avoids the cumbersome process of carrying an expensive total station to a permanent observation point for observation every time in the traditional method, and significantly improves the efficiency and convenience of monitoring.

Description

Technical Field

[0001] The present application relates to the field of bridges, and in particular to a linear management method, device and medium for prefabricated bridges. Background Art

[0002] Bridge alignment is an important indicator of bridge structural health. Regular inspection not only ensures accuracy during the construction phase, but also ensures the safety and durability of the bridge during the operation and maintenance phase. With the development of surveying and mapping technologies such as intelligent total stations, electronic levels, global satellite navigation systems, laser scanning, and photogrammetry, emerging bridge measurement methods have developed rapidly.

[0003] Traditional bridge alignment measurement equipment, including total stations and levels, is cumbersome to operate, time-consuming and labor-intensive, and it is difficult to conduct comprehensive measurement and control of all sections, especially when the construction period is tight. However, smart total stations are limited by weather and line of sight, require the use of reflectors, and are complex to operate; electronic levels are not suitable for terrain with large elevation differences, data must be recorded manually, and high requirements are placed on operators.

[0004] It can be seen that how to solve the problem that existing detection solutions rely on manual operation is a technical problem that people in this field need to solve urgently. Summary of the invention

[0005] The purpose of this application is to provide a linear management method, device and medium for a prefabricated bridge to solve the problem that the existing detection scheme relies on manual operation.

[0006] In order to solve the above technical problems, the present application provides a linear management method for an assembled bridge, wherein four fixed targets are installed on one side of the bridge, and a plurality of detachable targets are installed on each box girder, wherein the fixed targets and the detachable targets are targets including circular patterns; the method comprises:

[0007] Take multiple positioning target images of the current bridge at different angles;

[0008] Recognizing the positioning target image based on a target detection algorithm, and classifying and determining fixed target marking points and detachable target marking points;

[0009] Determine the two-dimensional pixel coordinates of the circular centers of the fixed target marking point and the detachable target marking point based on an ellipse detection algorithm;

[0010] Determine the camera pose corresponding to each positioning target image by a five-point algorithm according to the two-dimensional pixel coordinates of at least two positioning target images;

[0011] According to the camera pose of each positioning target image, each marking point is converted to the corresponding camera coordinate system by a triangulation method;

[0012] The overlapping marking points between the positioning target images are spliced ​​and converted to the same world coordinate system to obtain full marking point information including the three-dimensional coordinates of all marking points;

[0013] Linear fitting is performed based on the information of all the marking points to obtain the bridge linear result for managing the current bridge.

[0014] As an optional solution, in the linear management method of the prefabricated bridge, the fixed target and the detachable target are targets including circular patterns and different numbers of rectangular patterns;

[0015] Correspondingly, the target detection algorithm is used to identify the positioning target image, and the fixed target marking points and the detachable target marking points are classified and determined, including:

[0016] Positioning the fixed target and the detachable target according to the circular pattern on the positioning target image based on a target detection algorithm;

[0017] Based on the target detection algorithm, the numbers of the fixed target and the detachable target are determined according to the rectangular pattern on the positioning target image, and the fixed target marking point and the detachable target marking point are determined.

[0018] As an optional solution, in the above-mentioned linear management method of the prefabricated bridge, the two-dimensional pixel coordinates of the circular centers of the fixed target marking point and the detachable target marking point are determined based on the ellipse detection algorithm, including:

[0019] Preprocessing the positioning target image;

[0020] Using ellipse detection technology, circular patterns in the image are identified, and the two-dimensional pixel coordinates of the center of each circular pattern are determined;

[0021] The two-dimensional pixel coordinates are used as the corresponding two-dimensional pixel coordinates of the fixed target marking point and the detachable target marking point.

[0022] As an optional solution, in the above-mentioned linear management method of the prefabricated bridge, the camera pose corresponding to each positioning target image is determined by a five-point algorithm based on the two-dimensional pixel coordinates of at least two positioning target images, including:

[0023] Determining two pairs of positioning target images containing at least five groups of marking points with the same name according to the serial numbers of the detachable targets in each of the positioning target images, wherein the five groups of marking points with the same name are detachable target marking points;

[0024] Determine the relative position between each pair of marker points based on the five-point algorithm and five groups of marker points with the same name;

[0025] The camera pose corresponding to each positioning target image is determined according to the relative pose and the camera intrinsic parameter matrix.

[0026] As an optional solution, in the above-mentioned linear management method of the prefabricated bridge, the converting of each marking point to the corresponding camera coordinate system by a triangulation method according to the camera pose of each positioning target image includes:

[0027] Obtaining the camera pose of each positioning target image, including the rotation matrix and translation vector of each positioning target image;

[0028] For the camera pose corresponding to each positioning target image and the two-dimensional pixel coordinates of the detachable target marking point, the three-dimensional coordinates corresponding to each marking point in the camera coordinate system are calculated by a triangulation method.

[0029] As an optional solution, in the above-mentioned linear management method of the prefabricated bridge, the overlapping marking points between the positioning target images are spliced ​​and converted to the same world coordinate system to obtain full marking point information including the three-dimensional coordinates of all marking points, including:

[0030] Determining the detachable target marking points of the overlapping parts according to the serial numbers of the detachable targets in each positioning target image;

[0031] Mapping the camera coordinate system of each positioning target image into a unified world coordinate system;

[0032] Coordinate splicing is performed according to the overlapping detachable target marking points to form splicing coordinate point information with continuous and non-repetitive coordinate points;

[0033] According to the spliced ​​coordinate point information, the full marking point information including the three-dimensional coordinates of all marking points is obtained.

[0034] As an optional solution, in the above-mentioned method for managing the linear shape of a prefabricated bridge, performing linear fitting according to the information of all marking points to obtain the linear shape result of the bridge for managing the current bridge includes:

[0035] Get the preset design line shape;

[0036] Perform linear fitting based on the information of all the marking points to obtain the bridge line shape result;

[0037] The deviation analysis result is obtained by comparing the bridge alignment result with the preset design alignment.

[0038] In order to solve the above technical problems, the present application also provides a linear management device for an assembled bridge, wherein four fixed targets are installed on one side of the bridge, and a plurality of detachable targets are installed on each box girder, wherein the fixed targets and the detachable targets are targets including circular patterns; the device comprises:

[0039] A shooting module is used to shoot multiple positioning target images of the current bridge at different angles;

[0040] A recognition module, used to recognize the positioning target image based on a target detection algorithm, and classify and determine fixed target marking points and detachable target marking points;

[0041] A positioning module, used to determine the two-dimensional pixel coordinates of the circular centers of the fixed target marking point and the detachable target marking point based on an ellipse detection algorithm;

[0042] An analysis module, used to determine the camera pose corresponding to each positioning target image through a five-point algorithm according to the two-dimensional pixel coordinates of at least two positioning target images;

[0043] A conversion module, used to convert each marker point to a corresponding camera coordinate system by a triangulation method according to the camera pose of each positioning target image;

[0044] A splicing module, used for splicing the overlapping marking points between the positioning target images and converting them into the same world coordinate system to obtain full marking point information including the three-dimensional coordinates of all marking points;

[0045] The fitting analysis module is used to perform linear fitting based on the information of all marking points to obtain the bridge linear result and manage the current bridge.

[0046] In order to solve the above technical problems, the present application also provides a linear management device for a prefabricated bridge, comprising:

[0047] Memory for storing computer programs;

[0048] A processor is used to implement the steps of the above-mentioned linear management method of the prefabricated bridge when executing the computer program.

[0049] In order to solve the above technical problems, the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned linear management method of the prefabricated bridge are implemented.

[0050] The linear management method of the prefabricated bridge provided in the present application has four fixed targets installed on one side of the bridge and multiple detachable targets installed on each box girder, wherein the fixed targets and the detachable targets are targets containing circular patterns; the method includes: shooting multiple positioning target images at different angles of the current bridge; identifying the positioning target images based on the target detection algorithm, and classifying and determining the fixed target marking points and the detachable target marking points; determining the two-dimensional pixel coordinates of the circular centers of the fixed target marking points and the detachable target marking points based on the ellipse detection algorithm; determining the camera pose corresponding to each positioning target image through the five-point algorithm according to the two-dimensional pixel coordinates of at least two positioning target images; converting each marking point to the corresponding camera coordinate system through the triangulation method according to the camera pose of each positioning target image; splicing the overlapping marking points between the positioning target images and converting them to the same world coordinate system to obtain the full marking point information containing the three-dimensional coordinates of all marking points; performing linear fitting based on the full marking point information to obtain the bridge linear result to manage the current bridge. First, the present application obtains rich image data through multi-angle shooting. Then, the image is processed through target detection and ellipse detection algorithms to identify and locate the target. Next, the camera pose and the three-dimensional coordinates of the marker points are determined using a five-point algorithm and triangulation. Finally, the linear result of the bridge is obtained through spatial stitching and linear fitting. The linear result can provide a basis for construction or maintenance during the construction or commissioning of the bridge. This application only needs to use the image data containing the target to efficiently calculate the linear features of the current state of the bridge. It avoids the tedious process of carrying an expensive total station to a permanent observation point for observation every time in the traditional method, significantly improving the efficiency and convenience of monitoring. The high degree of automation reduces manual operations, labor intensity and human errors.

[0051] In addition, the present application also provides a device and a medium, which correspond to the linear management method of the above-mentioned prefabricated bridge and have the same effect as above. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to 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 paying any creative work.

[0053] Figure 1 A flowchart of a linear management method for a prefabricated bridge is provided for an embodiment of the present application;

[0054] Figure 2 A schematic diagram of the arrangement of a detachable target on a box beam provided in an embodiment of the present application;

[0055] Figure 3A schematic diagram of the installation of a detachable target provided in an embodiment of the present application;

[0056] Figure 4 A schematic diagram of a target image provided in an embodiment of the present application;

[0057] Figure 5 A schematic diagram of the epipolar geometric relationship of points of the same name provided in an embodiment of the present application;

[0058] Figure 6 A schematic diagram of the same coordinates in binocular vision principle space provided in an embodiment of the present application;

[0059] Figure 7 A schematic diagram of image stitching provided in an embodiment of the present application;

[0060] Figure 8 A structural diagram of a linear management device for a prefabricated bridge provided in an embodiment of the present application;

[0061] Fig. 9 A structural diagram of another linear management device for a prefabricated bridge provided in an embodiment of the present application.

[0062] Reference numerals:

[0063] 11-box beam; 12-detachable target. DETAILED DESCRIPTION

[0064] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0065] The core of this application is to provide a linear management method, device and medium for a prefabricated bridge.

[0066] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0067] In order to solve the above problems, the embodiment of the present application provides a linear management method for an assembled bridge, wherein four fixed targets are installed on one side of the bridge, and a plurality of detachable targets 12 are installed on each box girder, wherein the fixed targets and the detachable targets 12 are targets including circular patterns; Figure 1 A flowchart of a linear management method for an assembled bridge is provided for an embodiment of the present application, such as Figure 1 As shown, the method includes:

[0068] S11: Take multiple positioning target images of the current bridge at different angles;

[0069] S12: Identify the positioning target image based on the target detection algorithm, and classify and determine the fixed target marking points and the detachable target marking points;

[0070] S13: determining the two-dimensional pixel coordinates of the circular centers of the fixed target marking point and the detachable target marking point based on an ellipse detection algorithm;

[0071] S14: determining the camera pose corresponding to each positioning target image by a five-point algorithm according to the two-dimensional pixel coordinates of at least two positioning target images;

[0072] S15: According to the camera pose of each positioning target image, each marking point is converted to the corresponding camera coordinate system by a triangulation method;

[0073] S16: splicing the overlapping marking points between the positioning target images and converting them into the same world coordinate system to obtain full marking point information including the three-dimensional coordinates of all marking points;

[0074] S17: Perform linear fitting based on all the marking point information to obtain the bridge line shape result to manage the current bridge.

[0075] This embodiment is applicable to newly built or in-service prefabricated bridges, especially in the scenarios where the bridge line shape needs to be accurately measured and monitored during the bridge construction and maintenance phases.

[0076] In this embodiment, step S11 uses a drone or a fixed camera to photograph the bridge from different angles to obtain multi-view images including fixed and detachable targets 12. These images will be used for subsequent measurement and analysis.

[0077] Figure 2 A schematic diagram of the arrangement of a detachable target on a box beam provided in an embodiment of the present application is shown in FIG. Figure 2 As shown, when manufacturing the assembled box girder, the detachable target is fixed around the box girder as an assembled component. Figure 3 A schematic diagram of the installation of a detachable target provided in an embodiment of the present application is shown in FIG. Figure 3 As shown, a mounting opening is provided on the box beam 11, and a detachable target 12 is movably mounted on the mounting opening. If damaged, it can be conveniently replaced. Specifically, the detachable target 12 is a polycarbonate plate that is sun-proof and rain-proof.

[0078] Step S12 identifies the positioning target image based on the target detection algorithm, and classifies and determines the fixed target marking points and the detachable target marking points. The target detection algorithm, such as the improved YOLOv8-UMT model (the YOLOv8 model is a deep learning model for target detection, and is one of the versions of the Yolo (You Only Look Once) series of models. YOLOv8 is introduced on the basis of Unified Multi-modal Transformers (UMT), and the modal feature fusion part of UMT is optimized), classifies the detachable target in the image and determines the marking points. The algorithm improves the detection accuracy of small targets through a bi-directional feature pyramid network (BiFPN) and a normalized Gaussian Wasserstein distance (NWD).

[0079] YOLOv8 is designed to be fast, accurate, and easy to use. However, the original YOLOv8 model has difficulties in accurately detecting small target points in high-resolution drone images. On the one hand, drone images cover a large area, causing the target points to appear very small. The standard feature pyramid in YOLOv8 cannot effectively detect these small target points because their multi-scale feature representation is limited. On the other hand, the intersection-over-union (IoU) loss function used by YOLOv8 is very sensitive to small changes in position information, resulting in unreliable detection results. Repeated detections of one location may produce different results. By applying the Bidirectional Feature Pyramid Network (BiFPN) and A Normalized Gaussian Wasserstein Distance (NWD) on the framework of YOLOv8. BiFPN combines context enhancement and feature refinement to enhance the ability to detect small objects. Using NWD instead of IoU as a new metric to measure the similarity between bounding boxes of small objects solves the sensitivity problem of IoU and improves the reliability of small target detection.

[0080] By classifying and identifying the fixed targets in the image, we can determine which ones are fixed on one side of the bridge and which ones are detachable targets installed on the box girder. The detachable targets are used for subsequent bridge line shape analysis.

[0081] Step S13 determines the center point of the circular pattern on the target by an ellipse detection algorithm as the two-dimensional pixel coordinate of the marking point. When shooting from the side, the circle in the target will become an ellipse. The existing Hough transform method and RANSAC method (Random Sample Consensus) may miss or misdetect when processing images with complex backgrounds. In order to overcome these shortcomings, this embodiment proposes an improved ellipse detection algorithm, named Adaptive Multi-Angle Ellipse Detection (AMAEllipse Detection). The algorithm of this embodiment significantly improves the accuracy and robustness of ellipse detection by introducing multi-angle coverage verification and adaptive optimization methods. Specifically, 1) input an image and convert the image into a grayscale image to simplify subsequent processing steps. Then, an edge detection algorithm (Canny algorithm) is used to extract edge points in the image; 2) a set of ellipse candidate sets is preliminarily generated from the extracted edge points. Through the multi-angle coverage verification method, the integrity of each candidate ellipse is evaluated, and those candidate ellipses that do not meet the integrity requirements are eliminated; 3) the normal vector of the edge point on each candidate ellipse is calculated. By checking the consistency of the normal vector, the noise points and false detection points are identified and removed; 4) According to the results of the normal vector consistency check, the parameters of the ellipse are dynamically adjusted. Through the iterative optimization method, the ellipse parameters are gradually converged to the optimal value, so as to accurately fit the ellipse shape; 5) Based on the optimized ellipse parameters, the detected ellipse is output. These results are displayed in the form of images or data. Specifically, the fixed pixel coordinates are the coordinates of the center of the circular pattern on the detachable target image on the positioning target image.

[0082] Through precise recognition and mapping, high-precision conversion from image pixel coordinates to physical space coordinates is ensured. The entire recognition and mapping process is automated, reducing manual operations and improving efficiency.

[0083] Step S14 determines the camera pose corresponding to each positioning target image through a five-point algorithm according to the two-dimensional pixel coordinates of at least two positioning target images; and calculates the rotation and translation parameters of the camera through a five-point algorithm using identifiable non-collinear marker point pairs in at least two images.

[0084] The five-point algorithm is applied to solve the basic matrix and orient the camera position corresponding to each image. The five-point algorithm is a method to obtain the essential matrix between two images when the camera intrinsic parameters are known and the five groups of corresponding points between two images are known, and then decompose the corresponding rotation matrix and translation vector.

[0085] Step S15 converts each marking point to the corresponding camera coordinate system through the triangulation method according to the camera pose of each positioning target image. Specifically, according to the camera pose, the coordinates of the circular center points of these targets in their respective independent camera coordinate systems are obtained through the triangulation principle.

[0086] Step S16 transforms the coordinates of all target circular midpoints into the same world coordinate system through coordinate conversion and sequence dot splicing based on marker points, forming a complete three-dimensional coordinate set, including the coordinate set of all detachable target marker points in the same world coordinate system.

[0087] The linear fitting algorithm, such as the least squares method, is used to analyze the information of all the marking points to obtain the linear result of the bridge.

[0088] The entire control process starts with image acquisition, and the target is accurately identified and located through automated target detection and ellipse detection algorithms. The five-point algorithm is combined with the triangulation method to achieve accurate estimation of the camera pose and calculation of the three-dimensional coordinates of the marker points. Through intelligent image stitching technology, a unified world coordinate system is constructed, providing the necessary data basis for linear fitting. Using drones for image acquisition can cover the hard-to-reach areas of the bridge and provide more comprehensive bridge status information. Through computer vision technology, automated detection and analysis of the bridge line shape is achieved, which helps to promptly detect structural deviations and potential problems.

[0089] According to the linear management method of the prefabricated bridge provided by the present embodiment, four fixed targets are installed on one side of the bridge, and multiple removable targets 12 are installed on each box girder, wherein the fixed targets and the removable targets 12 are targets containing circular patterns; the method comprises: shooting multiple positioning target images at different angles of the current bridge; identifying the positioning target images based on the target detection algorithm, and classifying and determining the fixed target marking points and the removable target marking points; determining the two-dimensional pixel coordinates of the circular centers of the fixed target marking points and the removable target marking points based on the ellipse detection algorithm; determining the camera pose corresponding to each positioning target image through the five-point algorithm according to the two-dimensional pixel coordinates of at least two positioning target images; converting each marking point to the corresponding camera coordinate system through the triangulation method according to the camera pose of each positioning target image; splicing the overlapping marking points between the positioning target images and converting them to the same world coordinate system to obtain the full marking point information containing the three-dimensional coordinates of all marking points; performing linear fitting based on the full marking point information to obtain the bridge linear result to manage the current bridge. First, the present application obtains rich image data through multi-angle shooting. Then, the image is processed through target detection and ellipse detection algorithms to identify and locate the target. Next, the camera pose and the three-dimensional coordinates of the marker points are determined using a five-point algorithm and triangulation. Finally, the linear result of the bridge is obtained through spatial stitching and linear fitting. The linear result can provide a basis for construction or maintenance during the construction or commissioning of the bridge. This application only needs to use the image data containing the target to efficiently calculate the linear features of the current state of the bridge. It avoids the tedious process of carrying an expensive total station to a permanent observation point for observation every time in the traditional method, significantly improving the efficiency and convenience of monitoring. The high degree of automation reduces manual operations, labor intensity and human errors.

[0090] According to the above embodiment, in another specific embodiment, the fixed target and the removable target 12 are targets including circular patterns and different numbers of rectangular patterns;

[0091] Correspondingly, the positioning target image is identified based on the target detection algorithm, and the fixed target marking points and the detachable target marking points are classified and determined, including:

[0092] Based on the target detection algorithm, the fixed target and the detachable target are located according to the circular pattern on the positioning target image;

[0093] Based on the target detection algorithm, the numbers of the fixed target and the detachable target are determined according to the rectangular pattern on the positioning target image, and the fixed target marking points and the detachable target marking points are determined.

[0094] The removable target is made of polycarbonate sheet, designed to be easy to install and remove, adapt to outdoor environment, and has sun and rain protection.

[0095] Figure 4 A target image schematic diagram provided in an embodiment of the present application is as follows: Figure 4 As shown, the circular pattern on the target is used to determine the center position, and the rectangular pattern is used for numbering and classification for easy identification and tracking.

[0096] The acquired positioning target image is preprocessed, including grayscale, filtering and denoising, edge enhancement, etc., to improve the image quality and enhance the recognizability of the target pattern.

[0097] Use deep learning models such as YOLOv8-UMT to classify and identify removable targets in images. The algorithm is pre-trained and fine-tuned to accurately identify rectangular and circular patterns in images.

[0098] The target detection algorithm numbers and classifies the removable targets according to the rectangular pattern. The rectangular pattern on each target forms a unique identifier, which helps to distinguish different fixed targets and removable targets for subsequent data processing and analysis.

[0099] The circular pattern is used to initially locate the position of the target, and the rectangular pattern is used to further distinguish and number it to ensure that each target can be accurately identified. According to the number and arrangement of the rectangular patterns, a unique number is assigned to each fixed target and removable target, which increases the identifiability and flexibility of the system, especially when the targets are densely packed or partially blocked. These center positions serve as markers on the image, providing key information for three-dimensional coordinate solution.

[0100] Through the above steps, the recognition accuracy of the target and the robustness of the system are improved by combining circular and rectangular patterns. Different numbers of rectangular patterns provide a unique identification for each target, simplifying the classification and management of targets. Combining the information of circular and rectangular patterns can more accurately estimate the camera pose and the three-dimensional coordinates of the target.

[0101] In a specific embodiment, in addition to circles and rectangles, other geometric shapes, such as triangles or stars, may be considered to be added to further improve the recognition capability and flexibility of the system.

[0102] According to the above embodiment, in another specific embodiment, in the above-mentioned linear management method of the prefabricated bridge, the two-dimensional pixel coordinates of the circular centers of the fixed target marking points and the detachable target marking points are determined based on the ellipse detection algorithm, including:

[0103] Preprocessing the positioning target image;

[0104] Using ellipse detection technology, circular patterns in the image are identified and the two-dimensional pixel coordinates of the center of each circular pattern are determined;

[0105] The two-dimensional pixel coordinates are used as the two-dimensional pixel coordinates of the corresponding fixed target marking points and the detachable target marking points.

[0106] This embodiment uses image processing technology to improve image quality and enhance the accuracy of subsequent ellipse detection. The preprocessing steps may include denoising, grayscale, contrast enhancement, edge enhancement, etc.

[0107] Use advanced image processing algorithms such as Hough transform, least squares fitting or deep learning-based ellipse detection models to identify circular patterns in images. Ellipse detection not only identifies circles, but also handles partially occluded or deformed circular patterns.

[0108] Once the circular pattern is identified, the algorithm calculates and determines the center position of each circle in the form of 2D pixel coordinates. These coordinates are the key features of the fixed target and the removable target markers. The determined 2D pixel coordinates are directly used as the target markers for subsequent target recognition and pose estimation steps.

[0109] Improved target identification accuracy through pre-processing and advanced ellipse detection technology. Accurately identify targets even in less than ideal image conditions. Automated image processing and coordinate extraction reduces the need for manual operation.

[0110] According to the above embodiment, in another specific embodiment, in the above method for managing the linearity of the prefabricated bridge, the camera pose corresponding to each positioning target image is determined by a five-point algorithm according to the two-dimensional pixel coordinates of at least two positioning target images, including:

[0111] Determine two pairs of positioning target images containing at least five groups of marking points with the same name according to the numbers of the detachable targets in each positioning target image, wherein the five groups of marking points with the same name are detachable target marking points;

[0112] Determine the relative position between each pair of marker points based on the five-point algorithm and five groups of marker points with the same name;

[0113] The camera pose corresponding to each positioning target image is determined according to the relative pose and the camera intrinsic parameter matrix.

[0114] From multiple positioning target images, identify and select at least five groups of markers with the same name according to the number of the detachable target. These points are consistent in different images and can be used as the basis for matching. Use the five-point algorithm, which is a method that uses at least five matching point pairs to calculate the relative pose between two images. The algorithm can calculate the relative rotation and translation. Through the five-point algorithm, the relative pose between each pair of markers with the same name is determined, including rotation and translation parameters, which reflects the spatial relationship between the two images. Combined with the known camera intrinsic parameter matrix, the relative pose is converted to the absolute camera pose of each image. The camera intrinsic parameters include focal length, principal point coordinates, etc., which are crucial for accurate pose calculation. Finally, the pose of the camera corresponding to each positioning target image in the world coordinate system is determined to provide accurate spatial information for subsequent 3D reconstruction and line management.

[0115] Figure 5 A schematic diagram of the epipolar geometric relationship of a point with the same name provided in an embodiment of the present application, such as Figure 5 As shown, the line connecting the center positions of the two cameras As the baseline, there is a plane bundle with the baseline as the axis, which intersects the two image planes, and there is a plane , let the camera center , , the same-named image points on the two images , and space point M are all in this plane, see Figure 5 Therefore, the points with the same name on the graph (set as and ) can be connected to each other through the basic matrix F, satisfying the following relationship:

[0116] (1)

[0117] The camera can be calibrated to obtain the intrinsic parameter matrix K. At this time, the basic matrix F degenerates into the essential matrix E. The relationship between the three is as follows:

[0118] (2)

[0119] For two images with satisfactory overlap, the five-point algorithm uses five baselines provided by five groups of points with the same name.

[0120] By solving the above equations simultaneously, we can estimate the essential matrix E, which determines the camera position and attitude during shooting. Therefore, by solving all overlapping images obtained from different camera positions, we can get the relative camera positions between each other, and thus determine the three-dimensional space coordinates of the camera.

[0121] After obtaining the spatial position of the camera, the coordinates of a point in space can be calculated through the positions of image points in two or more images and the shooting positions of the images.

[0122] Figure 6 A schematic diagram of the same coordinates of a binocular vision principle space provided in an embodiment of the present application, such as Figure 6 As shown in , according to the binocular vision principle, the three-dimensional coordinates of a point can be determined by taking photos from two different angles. Assume that for a point P in space, photos are taken from two angles, such as Figure 6 As shown, , The pixel coordinates of the left and right image points are respectively. The left camera coordinate system and the world coordinate system Completely overlapped, the camera coordinate system on the right is ,in They are respectively the center of light.

[0123] The camera imaging model can be expressed as:

[0124] (3)

[0125] In the above formula is the camera intrinsic parameter matrix, which can be obtained through camera calibration; and is the focal length of the camera in the horizontal and vertical directions. x and C y are the coordinates of the principal point of the image, usually the center of the image. and are the rotation matrix and translation vector from the world coordinate system (WCS) to the camera coordinate system. 11 、r 12 、r 13 : This line represents the direction of the x-axis of the camera coordinate system in the world coordinate system. That is, when a point moves along the x-axis in the camera coordinate system, its change in the world coordinate system. 21 、r 22 、r 23 : This line represents the direction of the y-axis of the camera coordinate system in the world coordinate system. That is, when a point moves along the y-axis in the camera coordinate system, its change in the world coordinate system. 31 、r 32 、r 33 : This line indicates the direction of the z-axis of the camera coordinate system in the world coordinate system. x Indicates the translation of an object or coordinate system on the x-axis. y Indicates the translation of an object or coordinate system on the y-axis. zRepresents the translation of an object or coordinate system on the z-axis. That is, when a point moves along the z-axis in the camera coordinate system, its change in the world coordinate system.

[0126] The spatial coordinates of the target center in WCS; is the pixel coordinate of the target center in the camera coordinate system.

[0127] Left camera coordinate system: With the world coordinate system: If they coincide, the rotation matrix of the left camera coordinate system relative to the world coordinate system is the unit matrix, and the translation vector is 0.

[0128] Substituting R=I and T=0 into equation (3), we can obtain the projection model of the left camera:

[0129] (4)

[0130] For the right camera, the relationship between the rotation matrix relative to the world coordinate system and the translation vector converted to the left and right camera coordinate systems can be obtained through camera calibration. Therefore, the projection model of the right camera can be rewritten as:

[0131] (5)

[0132] t x Indicates the translation of an object or coordinate system on the x-axis. y Indicates the translation of an object or coordinate system on the y-axis. z Indicates the translation of an object or coordinate system on the z-axis. , Respectively represent the pixel coordinates of the projection model of the left camera; , They represent the pixel coordinates of the projection model of the right camera respectively, and s is the scale factor.

[0133] Once the stereo vision system is calibrated, the coordinates of the target point in the WCS can be calculated by combining equations (4) and (5): There are four sets of independent equations to solve the three unknowns, which can be achieved by the least squares method.

[0134] According to the above embodiment, in another specific embodiment, according to the camera pose of each positioning target image, converting each marking point to the corresponding camera coordinate system by a triangulation method includes:

[0135] Obtain the camera pose of each positioning target image, including the rotation matrix and translation vector of each positioning target image;

[0136] For the camera pose corresponding to each positioning target image and the two-dimensional pixel coordinates of the detachable target marker point, the three-dimensional coordinates of each marker point in the camera coordinate system are calculated by triangulation method.

[0137] Determine the camera pose for each positioning target image, which includes the rotation matrix and translation vector. For each image, use the known camera pose (rotation matrix and translation vector) and the 2D pixel coordinates to calculate the 3D coordinates of each marker point through triangulation. Triangulation is a method of determining the position of a point in space using 2D image information from multiple viewpoints. In this process, images captured by cameras at least two different positions are used to determine the 3D coordinates of the marker point. Combine the camera intrinsic parameters (such as focal length, principal point coordinates) and the camera pose to convert the 2D pixel coordinates into 3D coordinates in the camera coordinate system.

[0138] This embodiment achieves high-precision coordinate recovery from a two-dimensional image to a three-dimensional space through precise triangulation.

[0139] According to the above embodiment, in another specific embodiment, in the above-mentioned linear management method of the prefabricated bridge, the overlapping marking points between the positioning target images are spliced ​​and converted to the same world coordinate system to obtain full marking point information including the three-dimensional coordinates of all marking points, including:

[0140] Determine the detachable target marking points of the overlapping parts according to the serial numbers of the detachable targets in each positioning target image;

[0141] Map the camera coordinate system of each positioning target image into a unified world coordinate system;

[0142] Coordinate splicing is performed according to the overlapping detachable target marking points to form splicing coordinate point information with continuous and non-repetitive coordinate points;

[0143] According to the splicing coordinate point information, the full marking point information including the three-dimensional coordinates of all marking points is obtained.

[0144] Analyze each positioning target image and identify the marking points of the overlapping parts according to the numbers of the detachable targets. Figure 7 A schematic diagram of image stitching provided in an embodiment of the present application is shown in FIG. Figure 7 As shown, the marker points belonging to the overlapping area in each positioning target image are identified and marked. These marker points are visible in different images and are key references in the stitching process. Using the overlapping marker points, the relative spatial relationship between each two positioning target images is determined. This usually involves calculating the geometric transformation parameters between the images.

[0145] The marker points in each camera coordinate system are mapped to a unified world coordinate system through the known camera pose (rotation matrix and translation vector). Coordinate stitching is performed using overlapping detachable target marker points to ensure that the formed three-dimensional coordinate point set is continuous and has no duplication.

[0146] Using the known physical coordinates, perform 3D reconstruction and solution on the first positioning target image to obtain the 3D coordinates of the marker points in the image in the world coordinate system. Repeat the 3D reconstruction and solution steps for each positioning target image to obtain the coordinates of the marker points in each image in their own independent camera coordinate system.

[0147] By calculating the rotation vector and translation vector, the coordinates of all marker points in the camera coordinate system are transformed into a unified world coordinate system. This step ensures that all image data are in the same reference frame.

[0148] Seamlessly stitch the image areas corresponding to the marker points converted to the same coordinate system. This step may involve image cropping, interpolation, fusion and other technologies to ensure that the stitched image is visually coherent.

[0149] Based on the coordinate point information obtained by stitching, the full marker point information containing the 3D coordinates of all marker points is constructed to provide comprehensive data support for the linear analysis of the bridge. Accurate 3D reconstruction and coordinate transformation ensure the alignment between images and reduce stitching errors. Optimized image fusion technology improves the visual quality of the stitched image.

[0150] According to the above embodiment, in another specific embodiment, in the above method for managing the linear shape of the prefabricated bridge, linear fitting is performed according to all the marking point information to obtain the linear shape result of the bridge to manage the current bridge, including:

[0151] Get the preset design line shape;

[0152] Perform linear fitting based on all the marking point information to obtain the bridge alignment result;

[0153] By comparing the bridge alignment results with the preset design alignment, the deviation analysis results are obtained.

[0154] This embodiment obtains the design alignment data of the bridge from the design document or the computer-aided design (CAD) model, which usually includes the expected geometric shape and size of the bridge. The actual alignment of the bridge is obtained by linear fitting using the full marking point information (i.e., the three-dimensional coordinates of all marking points) through mathematical methods (such as the least squares method). The bridge alignment result obtained by fitting is compared with the design alignment, and the difference between the two is analyzed to obtain the deviation analysis result.

[0155] The comparison results are used to evaluate deviations during bridge construction or use to determine whether they are within an acceptable error range.

[0156] Select appropriate mathematical models and algorithms for linear fitting. Common methods include least squares, polynomial fitting, or other nonlinear fitting techniques. Using the fitting algorithm, determine the linear shape of the bridge based on the 3D coordinates of the marker points. This is usually represented as a series of line segments connecting the marker points, reflecting the actual shape of the bridge. Analyze the fitting results to determine whether they meet the expected accuracy requirements. Evaluate the deviation of the fitted curve from the actual bridge design linear shape. Calculate the deviation between the fitted linear shape and the design linear shape, which can provide important information for the construction accuracy or health monitoring of the bridge.

[0157] Preferably, the bridge alignment results are visualized in a graphical form to facilitate observation and analysis of the actual bridge alignment. Linear fitting provides an accurate method to measure and analyze bridge alignment. By fitting three-dimensional coordinates, the actual shape and status of the bridge can be fully understood.

[0158] Compare the three-dimensional coordinates of the actual line shape with the designed line shape and analyze the differences between the two. This may involve point-to-line distance calculations, curve fitting metrics, etc. Quantify the comparative analysis results and calculate deviation values, such as maximum deviation, average deviation, deviation distribution, etc. Display the deviation analysis results in the form of graphics or charts, such as deviation curves, deviation histograms, etc., to intuitively reflect the deviation situation. Generate a detailed deviation report based on the deviation analysis results. The report should include key deviation data, possible cause analysis, and recommended adjustment measures.

[0159] In the above embodiments, the linear management method of the prefabricated bridge is described in detail, and the present application also provides an embodiment corresponding to the linear management device of the prefabricated bridge. It should be noted that the present application describes the embodiments of the device part from two perspectives, one is based on the functional module perspective, and the other is based on the hardware perspective.

[0160] From the perspective of functional modules, Figure 8 A structural diagram of a linear management device for an assembled bridge provided in an embodiment of the present application, such as Figure 8 As shown, a linear management device for a prefabricated bridge comprises:

[0161] A shooting module 21 is used to shoot multiple positioning target images of the current bridge at different angles;

[0162] The recognition module 22 is used to recognize the positioning target image based on the target detection algorithm, and classify and determine the fixed target marking points and the detachable target marking points;

[0163] A positioning module 23, for determining the two-dimensional pixel coordinates of the circular centers of the fixed target marking points and the detachable target marking points based on an ellipse detection algorithm;

[0164] An analysis module 24 is used to determine the camera pose corresponding to each positioning target image by a five-point algorithm according to the two-dimensional pixel coordinates of at least two positioning target images;

[0165] The conversion module 25 is used to convert each marking point to a corresponding camera coordinate system by a triangulation method according to the camera pose of each positioning target image;

[0166] A stitching module 26 is used to stitch the overlapping marking points between the positioning target images and convert them into the same world coordinate system to obtain full marking point information including the three-dimensional coordinates of all marking points;

[0167] The fitting analysis module 27 is used to perform linear fitting based on all the marking point information to obtain the bridge line shape result and manage the current bridge.

[0168] Since the embodiments of the apparatus part correspond to the embodiments of the method part, please refer to the description of the embodiments of the method part for the embodiments of the apparatus part, which will not be repeated here.

[0169] Fig. 9 A structural diagram of another linear management device for an assembled bridge provided in an embodiment of the present application, such as Fig. 9 As shown, the linear management device for the prefabricated bridge includes: a memory 30 for storing a computer program;

[0170] The processor 31 is used to implement the steps of the method for obtaining user operation habit information as described in the above embodiment (linear management method for prefabricated bridges) when executing a computer program.

[0171] The linear management device for the prefabricated bridge provided in this embodiment may include but is not limited to a smart phone, a tablet computer, a laptop computer or a desktop computer.

[0172] Among them, the processor 31 may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor 31 may be implemented in at least one hardware form of a digital signal processor (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 31 may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a central processing unit (CPU); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor 31 may be integrated with a graphics processing unit (GPU), and the GPU is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor 31 may also include an artificial intelligence (AI) processor, which is used to process computing operations related to machine learning.

[0173] The memory 30 may include one or more computer-readable storage media, which may be non-transitory. The memory 30 may also include a high-speed random access memory, and a non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In this embodiment, the memory 30 is at least used to store the following computer program 301, wherein, after the computer program is loaded and executed by the processor 31, it can implement the relevant steps of the linear management method of the prefabricated bridge disclosed in any of the aforementioned embodiments. In addition, the resources stored in the memory 30 may also include an operating system 302 and data 303, etc., and the storage method may be temporary storage or permanent storage. Among them, the operating system 302 may include Windows, Unix, Linux, etc. Data 303 may include, but is not limited to, data involved in implementing the linear management method of the prefabricated bridge, etc.

[0174] In some embodiments, the linear management device of the prefabricated bridge may also include a display screen 32 , an input and output interface 33 , a communication interface 34 , a power supply 35 , and a communication bus 36 .

[0175] Those skilled in the art will understand that Fig. 9 The structure shown in the figure does not constitute a limitation on the linear management device of the prefabricated bridge, and may include more or fewer components than shown in the figure.

[0176] The linear management device for a prefabricated bridge provided in an embodiment of the present application includes a memory and a processor. When the processor executes a program stored in the memory, it can implement the following method: linear management method for a prefabricated bridge.

[0177] Finally, the present application also provides an embodiment corresponding to a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps described in the embodiment of the linear management method for prefabricated bridges are implemented.

[0178] It is understandable that if the method in the above embodiment is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0179] The computer-readable storage medium provided in this embodiment stores a computer program thereon. When a processor executes the program, the following method can be implemented: a linear management method for a prefabricated bridge.

[0180] The linear management method, device and medium of the prefabricated bridge provided by the present application are introduced in detail above. The various embodiments in the specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same and similar parts between the embodiments can be referenced to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the scope of protection of the claims of the present application.

[0181] It should also be noted that, in this specification, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device including the element.

Claims

1. A linear management method for an assembled bridge, characterized in that: Four fixed targets are installed on one side of the bridge, and a plurality of detachable targets are installed on each box girder, wherein the fixed targets and the detachable targets are targets including circular patterns; the method comprises: Take multiple positioning target images of the current bridge at different angles; Recognizing the positioning target image based on a target detection algorithm, and classifying and determining fixed target marking points and detachable target marking points; Determine the two-dimensional pixel coordinates of the circular centers of the fixed target marking point and the detachable target marking point based on an ellipse detection algorithm; Determine the camera pose corresponding to each positioning target image by a five-point algorithm according to the two-dimensional pixel coordinates of at least two positioning target images; According to the camera pose of each positioning target image, each marking point is converted to the corresponding camera coordinate system by a triangulation method; The overlapping marking points between the positioning target images are spliced ​​and converted to the same world coordinate system to obtain full marking point information including the three-dimensional coordinates of all marking points; Linear fitting is performed based on the information of all the marking points to obtain the bridge linear result for managing the current bridge.

2. The linear management method of a prefabricated bridge according to claim 1, characterized in that: The fixed target and the detachable target are targets including circular patterns and rectangular patterns of different numbers; Correspondingly, the target detection algorithm is used to identify the positioning target image, and the fixed target marking points and the detachable target marking points are classified and determined, including: Positioning the fixed target and the detachable target according to the circular pattern on the positioning target image based on a target detection algorithm; Based on the target detection algorithm, the numbers of the fixed target and the detachable target are determined according to the rectangular pattern on the positioning target image, and the fixed target marking point and the detachable target marking point are determined.

3. The linear management method of a prefabricated bridge according to claim 1, characterized in that: The method of determining the two-dimensional pixel coordinates of the circular centers of the fixed target marking point and the detachable target marking point based on an ellipse detection algorithm includes: Preprocessing the positioning target image; Using ellipse detection technology, circular patterns in the image are identified, and the two-dimensional pixel coordinates of the center of each circular pattern are determined; The two-dimensional pixel coordinates are used as the corresponding two-dimensional pixel coordinates of the fixed target marking point and the detachable target marking point.

4. The linear management method of the prefabricated bridge according to claim 3 is characterized in that: The step of determining the camera pose corresponding to each positioning target image by a five-point algorithm according to the two-dimensional pixel coordinates of at least two positioning target images comprises: Determining two pairs of positioning target images containing at least five groups of marking points with the same name according to the serial numbers of the detachable targets in each of the positioning target images, wherein the five groups of marking points with the same name are detachable target marking points; Determine the relative position between each pair of marker points according to the five-point algorithm and the five groups of marker points with the same name; The camera pose corresponding to each positioning target image is determined according to the relative pose and the camera intrinsic parameter matrix.

5. The linear management method of a prefabricated bridge according to claim 4, characterized in that: The converting each marker point to a corresponding camera coordinate system by a triangulation method according to the camera pose of each positioning target image comprises: Obtaining the camera pose of each positioning target image, including the rotation matrix and translation vector of each positioning target image; For the camera pose corresponding to each positioning target image and the two-dimensional pixel coordinates of the detachable target marking point, the three-dimensional coordinates corresponding to each marking point in the camera coordinate system are calculated by a triangulation method.

6. The linear management method of a prefabricated bridge according to claim 5, characterized in that: The overlapping marking points between the positioning target images are spliced ​​and converted to the same world coordinate system to obtain full marking point information including the three-dimensional coordinates of all marking points, including: Determining the detachable target marking points of the overlapping parts according to the serial numbers of the detachable targets in each positioning target image; Mapping the camera coordinate system of each positioning target image into a unified world coordinate system; Coordinate splicing is performed according to the overlapping detachable target marking points to form splicing coordinate point information with continuous and non-repetitive coordinate points; According to the spliced ​​coordinate point information, the full marking point information including the three-dimensional coordinates of all marking points is obtained.

7. The linear management method of a prefabricated bridge according to claim 1, characterized in that: The linear fitting is performed according to the information of all the marking points to obtain the bridge line shape result to manage the current bridge, including: Get the preset design line shape; Perform linear fitting based on the information of all the marking points to obtain the bridge line shape result; The deviation analysis result is obtained by comparing the bridge alignment result with the preset design alignment.

8. A linear management device for a prefabricated bridge, characterized in that: Four fixed targets are installed on one side of the bridge, and a plurality of detachable targets are installed on each box girder, wherein the fixed targets and the detachable targets are targets containing circular patterns; the device comprises: A shooting module is used to shoot multiple positioning target images of the current bridge at different angles; A recognition module, used to recognize the positioning target image based on a target detection algorithm, and classify and determine fixed target marking points and detachable target marking points; A positioning module, used to determine the two-dimensional pixel coordinates of the circular centers of the fixed target marking point and the detachable target marking point based on an ellipse detection algorithm; An analysis module, used to determine the camera pose corresponding to each positioning target image through a five-point algorithm according to the two-dimensional pixel coordinates of at least two positioning target images; A conversion module, used to convert each marker point to a corresponding camera coordinate system by a triangulation method according to the camera pose of each positioning target image; A splicing module, used for splicing the overlapping marking points between the positioning target images and converting them into the same world coordinate system to obtain full marking point information including the three-dimensional coordinates of all marking points; The fitting analysis module is used to perform linear fitting based on the information of all marking points to obtain the bridge linear result and manage the current bridge.

9. A linear management device for a prefabricated bridge, characterized in that: include: Memory for storing computer programs; A processor, for implementing the steps of the linear management method of a prefabricated bridge as described in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the linear management method of the prefabricated bridge as described in any one of claims 1 to 7 are implemented.

Images