Dynamic monitoring method and system for vertical rotation closure of arch bridge based on camera measurement

Abstract

The invention relates to the technical field of bridge construction monitoring and computer vision measurement, and particularly discloses an arch bridge vertical rotation closure dynamic monitoring method and system based on camera measurement. According to the method, the image of the target arch rib in the vertical rotation process is obtained through the camera equipment, the spatial position data of the measuring point is calculated, a data basis can be provided for dynamic process analysis, and the dynamic process analysis accuracy is improved by converting the current spatial data of the measuring point into the parameterized curve of the current arch rib. The actual line shape of the arch rib can be visually and globally displayed and compared with the target arch rib; meanwhile, the abnormal deflection in the vertical rotation process can be found in time by estimating the deviation of the arch rib, real-time and accurate monitoring data can be provided for constructors, manual intervention is not needed or little manual intervention is needed in the whole process, personal errors and labor intensity are reduced, and the method has wide industrial popularization prospects.

Description

Technical Field

[0001] This invention relates to the field of bridge construction monitoring and computer vision measurement technology, and in particular to a dynamic monitoring method and system for the vertical rotation and closure of an arch bridge based on camera measurement. Background Technology

[0002] With the rapid advancement of transportation engineering construction, bridge spans are constantly increasing, and long-span arch bridges are being used more and more widely in high-speed and railway engineering. The continuous increase in traffic volume in developed regions has significantly increased the frequency of structural load changes; simultaneously, construction is accelerating its extension into areas with large elevation differences, complex geology, and frequent extreme weather events. The vertical rotation of the arch rib from a horizontal position to an upright position is a critical and high-risk process, but it requires less space, has lower reliance on temporary supports, and shorter high-altitude operation time during the closure section. Given suitable site and organizational conditions, it offers better safety and economy. This necessitates dynamic, continuous, and traceable monitoring of the arch axis shape, vertical rotation angle, and spatial displacement of key measuring points to ensure synchronous hoisting, limit control, and the safety of surrounding structures.

[0003] Currently, the widely used monitoring methods mainly rely on traditional optical measuring instruments such as total stations. These methods typically employ a "point-by-point measurement" mode, which involves periodically or trigger-based measurements at a few pre-deployed discrete measuring points on the arch rib. However, this approach cannot capture the complete deformation trajectory and dynamic response of the arch rib during continuous vertical rotation. The measurement data is discrete and discontinuous, potentially missing critical transient abnormal deformations. Furthermore, the measurement process requires manual aiming and tracking, resulting in low automation, especially during the critical closure stage, where high-frequency real-time data feedback cannot be provided. Similarly, measurement accuracy and reliability significantly decrease at night, in foggy conditions, or under vibration interference.

[0004] Therefore, there is an urgent need for a dynamic monitoring method that can achieve full-process automation, high precision, and overall attitude perception to ensure the safety, accuracy, and efficiency of the vertical rotation and closure construction of the arch bridge. Summary of the Invention

[0005] In order to overcome the above-mentioned technical problems in the existing monitoring of the vertical rotation construction of arch bridges, the present invention provides a dynamic monitoring method and system for the vertical rotation and closure of arch bridges based on video measurement.

[0006] In a first aspect, the present invention provides a dynamic monitoring method for the vertical rotation and closure of an arch bridge based on camera measurement, comprising: Images of the arch rib during its vertical rotation are acquired using a camera device, and the images include targets placed at predetermined positions on the arch rib. Calculate the spatial position of a preset measurement point on the target using the pixel coordinates of the target in the image; By using the spatial positions of preset measuring points, curve fitting is performed to obtain the parametric curve of the arch rib; Based on the parameterized curve, the deviation is estimated by combining it with the model curve of the target state of the arch rib.

[0007] According to a specific implementation method, in the above dynamic monitoring method, the preset position includes key points of the arch rib and fixed reference points, and at least four fixed reference points are selected.

[0008] According to one specific implementation method, the calculation of the spatial position of the preset measuring point in the target in the above dynamic monitoring method specifically includes: The camera device is calibrated using targets at fixed reference points to calculate distortion parameters, obtain the extrinsic parameters of the camera device relative to the global coordinate system, and acquire the first frame image, storing the pixel templates of each target in the first frame image. Extract the subpixel image coordinates of the target in the image, and normalize the subpixel image coordinates using the distortion parameters; Based on the pose estimation of the target plane and combined with the extrinsic parameters, the pose transformation of the target's local coordinate system relative to the global coordinate system is obtained; The coordinates of the preset measurement point on the target in the global coordinate system are obtained based on the pose change, and the three-dimensional displacement of the preset measurement point relative to the initial moment is obtained based on the pixel template, so as to obtain the spatial position of the preset measurement point. Then, the spatial position is rigidly constrained and corrected using a target with a fixed reference point.

[0009] According to one specific implementation, in the above dynamic monitoring method, the key points include the arch foot, suspension point, and arch crown of the target arch rib.

[0010] According to a specific implementation, in the above dynamic monitoring method, when the camera device is a single unit and the field of view is insufficient to cover the vertical rotation path of the arch rib, the camera device adopts a segmented relay acquisition method, so that adjacent fields of view overlap and include a common fixed reference point and a common target, so as to maintain the continuity and consistency of the cross-camera displacement results.

[0011] According to one specific implementation, in the above dynamic monitoring method, at least 1 / 6 of the images acquired by the multiple camera devices overlap; wherein the images acquired by the multiple camera devices are fused based on quality weighting.

[0012] According to one specific implementation, in the above dynamic monitoring method, the deviation estimation includes: Calculate the angle between the parameterized curve and the tangent vector to the model curve to obtain the linear deviation; The elevation deviation is obtained by calculating the difference between the height components of the parameterized curve and each preset measuring point in the model curve. The difference between the parameterized curve and the model curve at the closure end in three-dimensional space is calculated to obtain the closure end residual.

[0013] Secondly, the present invention provides a dynamic monitoring system for the vertical rotation and closure of an arch bridge based on camera measurement, comprising: The target is attached to a predetermined position on the target arch rib; Camera equipment is installed on both sides of the target arch rib to acquire images of the target arch rib; A control device is used to control the camera device to acquire images during the vertical rotation of the target arch rib, output deviation estimates, and determine the vertical rotation status of the arch rib based on the deviation estimates, using any of the above-mentioned methods for dynamic monitoring of the vertical rotation and closure of an arch bridge.

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention acquires images of the target arch rib during vertical rotation using a camera device, calculates the spatial position data of the measuring points, and provides a data foundation for dynamic process analysis. By converting the current spatial data of the measuring points into a parametric curve of the current arch rib, the actual shape of the arch rib can be displayed intuitively and globally and compared with the target arch rib. At the same time, this invention can promptly detect abnormal deflection during vertical rotation by estimating the deviation of the arch rib. This invention not only provides construction personnel with real-time and accurate monitoring data, but also requires little or no manual intervention throughout the entire process, reducing human error and labor intensity, and has broad prospects for industry promotion. Attached Figure Description

[0015] Figure 1 A flowchart illustrating a dynamic monitoring method for the vertical rotation and closure of an arch bridge based on camera measurement, provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of a target provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the station layout provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of the station layout provided for another embodiment of the present invention. Detailed Implementation

[0016] The present invention will now be described in further detail with reference to specific embodiments. However, this should not be construed as limiting the scope of the present invention to the following embodiments; all technologies implemented based on the content of the present invention fall within the scope of the present invention.

[0017] Unless otherwise specified, the terms "upper," "lower," "left," "right," "center," "inner," and "outer," etc., used in the description of specific embodiments of the present invention to indicate orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings, or the orientation or positional relationship in which the product / equipment / device is usually placed during use. These terms are merely for the purpose of facilitating the description of the present invention or simplifying the description in specific embodiments, and for enabling those skilled in the art to quickly understand the solution, and do not indicate or imply that a particular device / component / element must have a specific orientation, or be constructed and operated in a specific positional relationship. Therefore, they should not be construed as limitations on the present invention.

[0018] Furthermore, the use of terms such as "horizontal," "vertical," "suspended," "parallel," and "coaxial" does not imply that the corresponding device / component / element must be absolutely horizontal, vertical, suspended, parallel, or coaxial. Slight tilt or deviation is permissible, as long as it does not affect the normal function of the relevant component. For example, "horizontal" simply means that its direction is more horizontal relative to "vertical," not that the structure must be perfectly horizontal; a slight tilt is acceptable. "Coaxial" means that two components are arranged as coaxially as possible, allowing them to move coaxially or approximately coaxially when their relative positions change. Alternatively, it can be simplified to mean that the corresponding device / component / element, when arranged in "horizontal," "vertical," "suspended," "parallel," or "coaxial" directions, can have an error / deviation of ±10% relative to the corresponding direction, more preferably within ±8%, more preferably within ±6%, more preferably within ±5%, and more preferably within ±4%. For example, the deviation in the "coaxial" direction is controlled within 0.2-1mm, preferably within 0.2-0.5mm. As long as the corresponding device / component / element is within the error / deviation range, it can still achieve its function in the solution of the present invention.

[0019] Furthermore, in the description of the embodiments of the present invention, "several", "more than", and "a number of" represent at least two. The number can be any number, such as two, three, four, five, six, seven, eight, or nine, and can even exceed nine.

[0020] Furthermore, in the description of the technical solution of this invention, unless otherwise explicitly specified / limited / restricted, the terms "set up," "install," "connect," "link," "provided with," "laid out," and "arranged" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to connection methods commonly used in the art, such as welding, riveting, bolting, and threaded connections. Such connections can be mechanical, electrical, or communication connections; they can be direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components.

[0021] Most existing technologies focus on monitoring the changes in the three-dimensional coordinates of measuring points, or simply calculating the linear distance deviation from the design position. This method lacks macroscopic control over the overall shape and spatial attitude of the arch rib. For example, even if the coordinate deviations of all measuring points are within the allowable range, the arch rib may still experience unfavorable internal force states due to local torsion or bending. This "attitude deviation" is difficult to effectively identify using traditional point monitoring.

[0022] The ultimate goal of the vertical rotation process is to achieve precise closure. Traditional methods extrapolate the spatial state of the closure joint based on discrete measurement point data, resulting in coarse prediction models and large accumulated errors. They cannot accurately predict key three-dimensional deviations such as misalignment and rotation at the closure joint, often requiring tedious and risky attitude adjustments before closure, severely impacting construction efficiency and safety.

[0023] Furthermore, most existing early warning systems are based on judging the exceedance of single-point displacement, lacking a comprehensive evaluation index that can reflect the overall abnormal state of the structure. This early warning mechanism has low sensitivity and poor foresight, often only triggering when the problem is already quite obvious, failing to provide sufficient buffer time for construction decisions.

[0024] This invention elevates the monitoring dimension from "points" to "lines" by fitting discrete measuring points into continuous parametric curves, thereby achieving real-time, continuous, and visual monitoring of the overall arch rib alignment. Furthermore, it addresses the weakness of existing technologies in monitoring the spatial attitude (such as rotation angles and torsional deformation) of the arch rib. By introducing the "tangent vector angle" as a core parameter for deviation estimation, this invention can sensitively capture the deflection and torsional deformation of the arch rib control section during vertical rotation, which is impossible to effectively achieve using traditional coordinate difference comparison methods. Further, this invention provides the most direct and comprehensive closure deviation data (including longitudinal, lateral, and vertical deviations) for construction control by directly calculating the "three-dimensional difference vector" between the parametric curve and the model curve at the closure end, greatly improving the accuracy of closure prediction and the scientific nature of adjustment commands. Moreover, this invention integrates camera measurement, automatic image processing, curve fitting, and intelligent deviation estimation to construct a fully automatic closed-loop monitoring system. Based on comprehensive deviation analysis results (linear deviation, height difference, and three-dimensional vector), it achieves advanced and intelligent early warning, transforming passive response into proactive intervention.

[0025] The technical solution provided by the present invention will be further described in detail below with reference to specific embodiments.

[0026] This invention provides a dynamic monitoring system for the vertical rotation and closure of an arch bridge based on camera measurement, comprising: The target is attached to a predetermined position on the target arch rib; Camera equipment is installed on both sides of the target arch rib to acquire images of the target arch rib; The control device is used to employ a camera-based dynamic monitoring method for the vertical rotation and closure of an arch bridge. It controls the camera device to acquire images during the vertical rotation of the target arch rib, outputs a deviation estimate, and determines the vertical rotation status of the arch rib based on the deviation estimate.

[0027] The following section provides a detailed description of a dynamic monitoring method for the vertical rotation and closure of an arch bridge based on camera measurement, using specific implementation methods.

[0028] For details, please refer to Figure 1 This document illustrates a flowchart of a dynamic monitoring method for the vertical rotation and closure of an arch bridge based on camera measurement, provided by an embodiment of the present invention. Taking an arch rib that has rotated vertically from a horizontal position to a position awaiting closure as the object, the entire process of vertical rotation on one side is monitored by a single-lens camera. The method includes the following steps: S1: Acquire images of the arch rib during its vertical rotation using a camera device, the images containing targets placed at preset positions on the arch rib.

[0029] Specifically, the preset locations include key points and fixed reference points of the arch rib, with at least four fixed reference points selected. Among them, the key points include the arch foot, suspension point, and arch crown of the target arch rib.

[0030] Before the data collection begins in this step, targets are set up at the preset locations. The targets can be reflective markers, and a set of fixed points is formed by selecting several reference points located on the fixed reference point. At the same time, camera equipment is set up in a place with good visibility so that each camera equipment can monitor the targets at key points and fixed reference points.

[0031] S2: Calculate the spatial position of the preset measurement point in the target using the pixel coordinates of the target in the image.

[0032] Specifically, the camera device is calibrated using targets at fixed reference points, distortion parameters are calculated, the external parameters of the camera device relative to the global coordinate system are obtained, the first frame image is acquired, and the pixel templates of each target in the first frame image are stored. Extract the subpixel image coordinates of the target in the image, and normalize the subpixel image coordinates using the distortion parameters; Based on the pose estimation of the target plane and combined with the extrinsic parameters, the pose transformation of the target's local coordinate system relative to the global coordinate system is obtained; The coordinates of the preset measurement point on the target in the global coordinate system are obtained based on the pose change, and the three-dimensional displacement of the preset measurement point relative to the initial moment is obtained based on the pixel template, so as to obtain the spatial position of the preset measurement point. Then, the spatial position is rigidly constrained and corrected using a target with a fixed reference point.

[0033] It should be noted that, in the description of the embodiments of the present invention, the preset measurement points are points selected based on the target, generally the four corner points and the midpoint on the target.

[0034] S3: Use the spatial location of preset measuring points to perform curve fitting to obtain the parametric curve of the arch rib.

[0035] S4: Based on the parameterized curve, the deviation is estimated by combining the model curve of the target state of the arch rib.

[0036] Specifically, this step can be based on the known arch axis coefficient and target arch rib model to perform curve fitting and deviation estimation on the spatial position of the preset measuring points during the vertical rotation process, and output the linear deviation, elevation deviation and closure end residual results.

[0037] Furthermore, the spatial positions of preset measuring points at various times are used to fit the changes of the arch axis over time, and real-time early warning and monitoring are carried out based on the joint criteria composed of linear threshold, elevation threshold and closure end threshold. When each component does not exceed the preset threshold, it is determined that the vertical rotation is in place and the confirmation is completed.

[0038] The deviation estimation includes: Calculate the angle between the parameterized curve and the tangent vector to the model curve to obtain the linear deviation; The elevation deviation is obtained by calculating the difference between the height components of the parameterized curve and each preset measuring point in the model curve. The difference between the parameterized curve and the model curve at the closure end in three-dimensional space is calculated to obtain the closure end residual.

[0039] In an optional implementation, step one involves examining the site environment and selecting a location with good visibility, no obstructions, and the ability to observe the entire side of the arch rib to be vertically rotated. Cameras can be used as measuring stations, with one monocular camera positioned on each side of the arch rib for measurement. Two monocular industrial cameras are located on platforms on either side of the arch rib, arranged close to the longitudinal direction of the bridge to minimize the angle between the camera's optical axis and the bridge axis, ensuring the cameras can monitor the entire arch rib on that side.

[0040] Specifically, each camera should have a complete field of view covering all key points on the upper and lower chords of its side, as well as multiple fixed reference points. Before setting up the targets, it is sufficient to have complete coverage of the area where the arch rib and fixed reference points are located on its side. In particular, the industrial camera used for monocular cameras should preferably be a high-resolution camera. The higher the resolution, the better the monitoring effect. The camera pixel resolution should be greater than 24 million pixels, the lens should preferably be 50mm, and the sampling frequency should be less than 10Hz. The specific camera pixel resolution, lens focal length and exposure parameters can be selected according to the viewing distance, lighting conditions and on-site environment. In particular, since the vertical rotation of the arch rib does not last long, the camera station does not need to be fixed for a long time. It is advisable to use a total station tripod and fix it to the station with bolts, which facilitates the adjustment and relocation of the actual station position.

[0041] Step two, place 2×2 four-quadrant square reflective markers on the upper and lower chords of the arch rib, such as... Figure 2 As shown, the color topology is white in the upper left / lower right and black in the lower left / upper right; a small notch or mark can be set in one corner to eliminate directional ambiguity. The target pasting positions include, but are not limited to, several measuring points evenly distributed on the arch foot, hanging point, arch top, and arch rib, arranged symmetrically along the upstream and downstream; Specifically, the side length of a single target should be 40-60cm, and the four squares that make up a single target should be of equal size. If the target is too small, it will be difficult for the camera to extract the target. If the target is too large, it will be difficult to set up or affect the on-site construction. Under the premise of meeting the on-site implementation conditions, the larger the target, the better. In particular, the target placement locations must include the arch foot, hanging point, arch crown, and other key locations of the arch ribs. The remaining space should be evenly distributed with targets at reasonable intervals of 10 to 50 meters. Under the condition of meeting the on-site implementation requirements, the more targets are placed, the more accurately the arch rib condition can be reflected. In particular, the targets on the upper and lower chords of the arch rib, as well as on the upstream and downstream sides, should be arranged as symmetrically as possible; In particular, the reflective markings should be affixed as close as possible to the direction of the monitoring camera; Specifically, within the field of view of the measuring camera, at least four fixed locations are selected to affix reflective markers as fixing points. These markers are affixed to fixed reference points that do not participate in vertical rotation, and these four fixing points should avoid being coplanar as much as possible, for image stabilization and external parameter verification.

[0042] Step 3: After the targets are pasted, the camera is fixed at the measurement station location, and the measurement station is set up. The camera's focal length, aperture, and other parameters are adjusted and locked to ensure that each target in the camera's view is clearly imaged and has distinct boundaries. Please refer to [the relevant documentation / reference]. Figure 3 The diagram illustrates the station layout provided in an embodiment of the present invention. The camera is fixed using a tripod and connected to the host computer via a wired connection. The host computer supplies power to the camera and can control parameters such as camera shutter frequency and exposure time. In particular, the host can operate normally when connected to a 220V standard voltage power supply. It is advisable to use a fixed line and power supply to ensure a stable supply. Specifically, the host has a built-in network connection function to send data to the monitoring platform and perform time alignment; In particular, the main unit has a built-in Windows operating system for easy operation and control, and is equipped with camera photo storage function and displacement calculation software.

[0043] Step 4: Before the vertical rotation begins, determine the initial three-dimensional coordinates of all cameras and targets, establish a global coordinate system using a fixed point set, complete the camera intrinsic parameter calibration, calculate the distortion parameters, obtain the camera extrinsic parameters relative to the global coordinate system, acquire the first frame image, and store the pixel templates of the four corner points of each target. In particular, the camera mount can be geometrically calibrated in one go to ensure that the measurement coordinates accurately reflect the spatial position of the target and the camera; Preferably, a total station should be used to determine the initial three-dimensional coordinates of all cameras and targets; Preferably, the pixel is first subjected to distortion correction before entering the intrinsic parameter mapping. The formula for pixel distortion removal and normalization is:

[0044] in This is the intrinsic parameter matrix. It is in homogeneous form of pixel coordinates. These are the image points after distortion correction and normalization. Preferably, the extrinsic parameters of the camera pose relative to the global coordinate system are:

[0045] in, The extrinsic parameters of the camera relative to the global coordinate system. For rotation from the global coordinate system to the camera coordinate system, This is the translation from the global coordinate system to the camera coordinate system, and ; In particular, Solve using the following formula:

[0046] Where s is the scale factor, Let be the homogeneous form of the three-dimensional coordinates of a point in the global coordinate system.

[0047] In particular, camera intrinsic parameter calibration and distortion correction can be completed in advance in a laboratory or other stable environment.

[0048] Step 5: Just before the vertical rotation begins, the control host issues a synchronous measurement command, and all cameras begin to continuously capture images to start the measurement. Each camera synchronously acquires an image sequence containing preset measurement points and fixed point sets, extracts the sub-pixel image coordinates of the four corner points of the target, and normalizes the image points. Specifically, target position detection and corner sub-pixel extraction are performed on each frame of the image in chronological order to obtain the pixel coordinates of the four corners of each target. ; in, Refers to horizontal pixels, with positive pixels pointing to the right; Refers to vertical pixels, with downward pixels being positive.

[0049] Specifically, after obtaining the pixel coordinates, normalize them using the formula from step four:

[0050] Step 6: Based on the pose estimation of the target plane, obtain the pose transformation of the target's local coordinate system relative to the global coordinate system. In this embodiment, the homography decomposition method is used; The steps of using the homography decomposition method include: Step 6-1: Estimate the homography matrix; does a homography matrix exist? The following relationship exists:

[0051] in, The homogeneous coordinates of the target plane points in the target local coordinate system are obtained using DLT. And perform scale normalization; Step 6-2: Determine the pose transformation of the target relative to the camera coordinate system. The following relationship exists:

[0052] make for The List, Then we have: , , ,

[0053] right By performing orthogonalization, we obtain , Therefore:

[0054] Step 6-3: Determine the pose transformation of the target's local coordinate system relative to the global coordinate system. :

[0055] Step 7: Calculate the coordinates of the preset measuring point in the global coordinate system and its three-dimensional displacement relative to the initial time.

[0056] Specifically, the local coordinates of the preset measurement points on the target are... Coordinates transformed to the global coordinate system The formula is:

[0057] Specifically, with reference time Based on the reference, the corresponding three-dimensional displacement is:

[0058] Step 8: Use a fixed point set to perform rigid body constraint correction on the preset measurement points to compensate for the overall drift caused by camera support micro-motion, temperature drift and timing error. Specifically, the local coordinates of the preset measurement points on the target are... Coordinates transformed to the global coordinate system The formula is:

[0059] Specifically, with reference time Based on the reference, the corresponding three-dimensional displacement is:

[0060] Specifically, the overall rigid body drift of the scene in each frame is estimated based on a fixed set of points. Then the corrected coordinates and displacements are:

[0061]

[0062] Step nine: The host computer transmits the calculated global three-dimensional displacement data of each preset measuring point to the monitoring platform via a wireless network. The monitoring platform can choose whether to download the locally stored time-series images. After receiving the displacement data from the host computer, the monitoring platform performs curve fitting and deviation estimation on the spatial position of the preset measuring points at each time point based on the known arch axis coefficient and the target arch rib model, and outputs the linear deviation, elevation deviation, and closure end residual results.

[0063] Specifically, the process of outputting the results includes: Step 9-1: Use cubic splines to perform three-dimensional weighted least squares fitting on the series of measuring points on the arch rib at time t to obtain the parametric curve of the current arch rib. ,in As a mileage sequence, according to the order of the projections of the measuring points on the arch rib onto the plane, each point is projected onto the plane projection curve of the target arch axis, and the target mileage of a certain projection point is taken as... ; Step 9-2: Define the target arch rib model curve as follows Its tangent vector is The current fitted curve is The tangent vector is ; Step 9-3: Calculate the alignment deviation. At each control section, the alignment deviation is defined as the angle between the two tangent vectors:

[0064] Step 9-4: Calculate the elevation deviation. At each control section, compare the height component (vertical Z component) of the current position with the target position:

[0065] Step 9-5: Calculate the residual at the closure end. At this point, obtain the three-dimensional difference vector between the current endpoint and the target endpoint:

[0066] Step 10: Use the spatial positions of the preset measuring points at each time point to fit the change of the arch axis over time, and monitor it in real time based on the joint criteria composed of the linear threshold, elevation threshold and closure end threshold. When each component does not exceed the preset threshold, determine that the vertical rotation is in place and complete the confirmation.

[0067] In particular, the linear threshold, elevation threshold, and closure end threshold are given by a comprehensive analysis of the design and all three-dimensional measurement data of the cameras and targets before the measurement begins; Under this typical working condition, this embodiment can acquire the three-dimensional displacement and overall axis during the vertical rotation of the arch rib using a non-contact, single-camera setup. The camera resolution and stability meet the process control requirements. The system is robust to the effects of changes in illumination, local occlusion, and moderate-amplitude vibrations. The data processing link is lightweight, enabling real-time updates and monitoring. Actual implementation parameters can be flexibly determined within a reasonable range based on the on-site line of sight and construction conditions without affecting the essence of the method.

[0068] The method provided in the embodiments of the present invention will be described in detail below with reference to specific implementation methods.

[0069] Taking an arch rib that has been rotated from a vertical position to a position awaiting closure as an example, a combined monitoring scheme using multiple monocular cameras was adopted. The main process is consistent with the above, but specific design and calculations were implemented to address the issues of a single monocular camera not being able to cover one side of the arch rib and the fluctuation of the measurement platform.

[0070] A dynamic monitoring method for the vertical rotation and closure of an arch bridge based on camera measurement includes the following steps: S1: Install reflective markers at preset locations such as key points and fixed reference points on the arch rib, and select several reference points located on the fixed reference points to form a fixed point set; set up camera stations in places with good visibility so that each camera can monitor the key points and fixed reference points. S2: Before the measurement begins, determine the initial three-dimensional coordinates of all cameras and targets, establish a global coordinate system using a fixed point set, complete the camera intrinsic parameter calibration, calculate the distortion parameters, obtain the camera extrinsic parameters relative to the global coordinate system, acquire the first frame image, and store the pixel templates of the four corner points of each target. S3: After the measurement begins, each camera synchronously acquires an image sequence containing key points and fixed point sets, extracts the sub-pixel image coordinates of the four corner points of the target, and normalizes the image points. S4: During the vertical rotation, the pose transformation of the target's local coordinate system relative to the global coordinate system is obtained based on the pose estimation of the target plane. S5: Calculate the coordinates of the preset measurement points in the global coordinate system and their three-dimensional displacement relative to the initial time. Use a fixed point set to perform rigid body constraint correction on the preset measurement points to compensate for the overall drift caused by camera support micro-motion, temperature drift and timing error. S6: When the field of view of a single camera is insufficient to cover the vertical turning path, a multi-camera segmented relay acquisition method is adopted to make adjacent fields of view overlap and include common fixed points and common target features. The measurement results of the relay segments are seamlessly fused based on quality weighting to maintain the continuity and consistency of cross-camera displacement results. S7: Based on the known arch axis coefficient and target arch rib model, curve fitting and deviation estimation are performed on the spatial position of the preset measuring points at each time, and the results of linear deviation, elevation deviation and closure end residual are output. S8: The spatial position of the preset measuring point at each time moment is used to fit the change of the arch axis over time, and real-time warning and visualization are performed based on the joint criterion composed of the linear threshold, the elevation threshold and the closure end threshold. When each component does not exceed the preset threshold, it is determined that the vertical rotation is in place and the confirmation is completed.

[0071] In an optional implementation, step one involves examining the site environment and, based on the physical parameters, geometric parameters, and actual working conditions of the arch rib to be vertically rotated, selecting a location with good visibility and no obstructions to set up camera stations. One or more monocular cameras are arranged on each side of the arch rib to be vertically rotated for measurement.

[0072] In particular, if a single camera cannot observe the entire side of the arch rib to be vertically rotated, multiple single cameras can be combined for measurement. The fields of view of adjacent cameras have an overlap of no less than 1 / 6 of the screen, and the overlap area contains several common fixed targets. This allows the combined camera's field of view to observe the entire side of the arch rib. The field of view of each camera can completely cover the upper and lower chord measurement points and fixed points of its segment, realizing segmented relay and continuous tracking. The overall effect is equivalent to "large field of view" monitoring.

[0073] Step two: Install 2×2 four-quadrant square reflective markers at the upper and lower chords of the arch rib and at the fixing points, such as... Figure 2 As shown, the color topology is white in the top left / bottom right and black in the bottom left / top right; a small notch or mark can be set in one corner to eliminate directional ambiguity.

[0074] In particular, when the target is prone to falling off due to environmental factors such as rain and fog, transparent tape can be added to the non-corner positions of the target to stabilize it and ensure its stability. Specifically, the target placement locations must include the arch foot, hanging points, arch crown, and other key arch rib locations. The remaining space should be evenly distributed with targets at intervals of 10-50 meters. Where on-site implementation conditions are met, additional targets can be placed at the arch crown and arch foot locations. Specifically, within the overlapping area of ​​the camera's field of view, at least three fixed locations are selected to affix reflective markers as fixing points. At least four fixed locations are selected throughout the entire camera's field of view to affix reflective markers to rigid components that do not participate in vertical rotation. These four fixing points should avoid being coplanar as much as possible, for image stabilization and external parameter verification.

[0075] Step 3: After the targets are pasted, the camera is fixed at the station location, completing the station setup. The camera's focal length, aperture, and other parameters are adjusted and locked to ensure clear images and distinct boundaries for each target within the camera's view. Please refer to [link / reference needed]. Figure 4The diagram illustrates a station setup according to another embodiment of the present invention. The camera is fixed using a tripod and connected to the main unit via a wired connection. The camera is powered by the main unit, which can control parameters such as camera shutter frequency and exposure time. Specifically, the main unit can operate normally when connected to a 220V standard voltage power supply. If there is no fixed power supply, a high-capacity mobile power supply can be connected to power the measurement. In particular, fiber optic connections can be used between different hosts to ensure synchronized triggering between different cameras, which is more accurate than network time alignment. In particular, if the vertical rotation time is long and there are too many images stored locally, the host must ensure that it has enough storage space, no less than 500G of storage space, or the local storage data should be retrieved and cleared every once in a while to ensure that the host operating system will not affect the calculation speed due to insufficient storage space. In particular, necessary protective measures should be taken for the cameras, main units, and other equipment at the measurement station, such as sunshades and protective enclosures, to protect the station from interference during the measurement process.

[0076] Step 4: Before the vertical rotation begins, determine the initial three-dimensional coordinates of all cameras and targets, establish a global coordinate system using a fixed point set, complete the camera intrinsic parameter calibration, calculate the distortion parameters, obtain the camera extrinsic parameters relative to the global coordinate system, acquire the first frame image, and store the pixel templates of the four corner points of each target. In particular, the measurement and calibration steps should be performed as close as possible to the moment of vertical rotation to ensure that the initial state of the arch rib measured is consistent with the initial state of the arch rib before vertical rotation.

[0077] Step 5: Just before the vertical rotation begins, the control host issues a synchronous measurement command, and all cameras begin to continuously capture images to start the measurement. Each camera synchronously acquires an image sequence containing key points and fixed point sets, extracts the sub-pixel image coordinates of the four corner points of the target, and normalizes the image points. Step 6: Based on the pose estimation of the planar target, obtain the pose transformation of the target's local coordinate system relative to the global coordinate system. In this embodiment, the PnP algorithm is used to solve the problem; The steps for solving using the PnP algorithm include: Step Six-One: Collect and Construct 3D-2D Pairs ; Step 6-2: Use EPNP to calculate initial values ​​for extrinsic parameters: Calculate the initial values ​​for all 3D points. Expressed using an affine combination of four virtual control points, the "control points in the camera coordinate system" are represented. Using the position of "in" as the unknown, a system of linear equations is constructed and solved using the pinhole projection relationship. A single rigid body registration is then performed to obtain the result. , .

[0078] The projection relationship is as follows:

[0079] The initial values ​​of the target's pose transformation relative to the camera coordinate system are obtained:

[0080] Step 6-3: Refine using Gauss-Newton reprojection Starting with the initial value of EPNP, minimize the reprojection error of all corner points to obtain the refined result. , ,Right now

[0081] Step 6-4: Determine the pose transformation of the target's local coordinate system relative to the global coordinate system. :

[0082] Step 7: Calculate the coordinates of the preset measurement points in the global coordinate system and their three-dimensional displacement relative to the initial time. Use a fixed point set to perform rigid body constraint correction on the preset measurement points to compensate for the overall drift caused by camera support micro-motion, temperature drift and timing errors. Step 8: When the field of view of a single camera is insufficient to cover the vertical rotation path and a multi-camera segmented relay acquisition method is adopted, in order to maintain the continuity and consistency of the cross-camera displacement results, the measurement results of the relay segment are seamlessly fused based on quality weighting, while ensuring that adjacent fields of view overlap and include common fixed points and common target features. In particular, within the overlapping area of ​​adjacent cameras, for common-view measurement points, in order to ensure continuity and consistency across camera positions, the quality-weighted fusion result is used as the actual calculated position of the measurement point. The weight is estimated by quality indicators such as the corner strength of the measurement point to ensure continuous monitoring of the target.

[0083] Specifically, the process of seamlessly integrating the measurement results of the stress section includes: Step 8-1: Calculate the mass weight for each common-view preset measurement point at both camera positions. , The weights are calculated by combining factors such as corner strength and reprojection error, so that "more reliable" camera positions contribute more weight during fusion. Step 8-2: To ensure a smooth transition of displacement of the measuring points within the overlapping area, a relay coefficient is defined for the points in the overlapping area, gradually transitioning from 1 to 0 based on their spatial location. Multiply the relay coefficient by the mass weight and then normalize:

[0084] Step 8-3: Calculate the fusion displacement and fusion position: Fusion location:

[0085] refer to Momentary fusion displacement:

[0086] in, The calibrated measuring point in camera A at time t coordinates The calibrated measuring point in camera B at time t The coordinates.

[0087] Step nine: The host computer sends the calculated global three-dimensional displacement data of each target to the cloud platform via a wireless network. After receiving the displacement data from the host computer, the cloud platform performs curve fitting and deviation estimation on the spatial position of the preset measuring points at each time point based on the known arch axis coefficient and target arch rib model, and outputs the linear deviation, elevation deviation and closure end residual results; it also generates a new arch axis and arch bridge model and displays it on the webpage.

[0088] Step 10: Use the spatial positions of the preset measuring points at each time point to fit the change of the arch axis over time, and perform real-time monitoring and visualization based on the joint criteria composed of the linear threshold, elevation threshold and closure end threshold. When each component does not exceed the preset threshold, it is determined that the vertical rotation is in place and the confirmation is completed.

[0089] In particular, to verify the reliability of the measured values, inclinometers and total stations can be temporarily set up at several key sections for cross-checking. The difference between the two is usually kept within an acceptable range. If there is a deviation, the external parameters can be slightly re-estimated through fixed points to eliminate the influence.

[0090] Under this typical working condition, this embodiment, with multiple monocular cameras monitoring in segments and maintaining approximately 1 / 6 overlap in the field of view, achieved continuous and seamless displacement tracking and arch axis updates throughout the entire process of the system's vertical rotation from the arch rib's vertical state to the closure state. Rigid body correction involving fixed reference points effectively eliminated spurious drift caused by camera support micro-motion and temperature drift, while weighted fusion of the overlapping area ensured natural connection of target displacement measurements at the junction of adjacent cameras. It exhibited good robustness to local occlusion, changes in illumination, and moderate-amplitude vibrations. The processing results were dynamically displayed on a webpage, and specific parameters could be adjusted appropriately according to on-site needs without affecting the substantive implementation of the method.

[0091] Based on the above technical solutions, this invention proposes a dynamic monitoring method for the vertical rotation and closure of arch bridges based on camera measurement. Camera calibration and alignment are completed within the global construction coordinate system established by a total station. Combining sub-pixel corner point extraction and two types of pose solving, real-time calculation and visualization of three-dimensional displacement and arch axis along multiple measurement points throughout the vertical rotation process are achieved. Considering the characteristics of vertical rotation organization, a multi-camera relay with overlapping adjacent fields of view and a quality-weighted fusion of "shared fixed points + shared target features" are adopted. Rigid body constraint correction at fixed points suppresses station micro-motion and environmental drift, ensuring continuous consistency and coordinate uniformity across camera positions. At the results end, based on the known arch axis coefficient and target arch rib model, online output of alignment deviation, elevation deviation, and closure end residuals is provided, along with threshold warnings to support decisions on vertical rotation and precise closure. The system has a small footprint, allows for rapid expansion or replacement of camera positions, and is easy to maintain. It possesses good robustness and millimeter-level displacement monitoring capabilities, balancing safety and economy, and is suitable for dynamic monitoring and process control throughout the entire vertical rotation construction process of arch bridges.

[0092] Furthermore, this invention acquires images of the target arch rib during the vertical rotation process using a camera device, calculates the spatial position data of the measuring points, and provides a data foundation for dynamic process analysis. By converting the current spatial data of the measuring points into a parameterized curve of the current arch rib, the actual shape of the arch rib can be displayed intuitively and globally and compared with the target arch rib. At the same time, this invention can promptly detect abnormal deflection during the vertical rotation process by estimating the deviation of the arch rib. This invention not only provides construction personnel with real-time and accurate monitoring data, but also requires little or no manual intervention throughout the entire process, reducing human error and labor intensity, and has broad prospects for industry promotion.

[0093] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A dynamic monitoring method for the vertical rotation and closure of an arch bridge based on camera measurement, characterized in that, include: Images of the arch rib during its vertical rotation are acquired using a camera device, and the images include targets placed at predetermined positions on the arch rib. Calculate the spatial position of a preset measurement point on the target using the pixel coordinates of the target in the image; By using the spatial positions of preset measuring points, curve fitting is performed to obtain the parametric curve of the arch rib; Based on the parameterized curve, the deviation is estimated by combining it with the model curve of the target state of the arch rib.

2. The method for dynamic monitoring of the vertical rotation and closure of an arch bridge based on camera measurement according to claim 1, characterized in that, The preset positions include key points of the arch rib and fixed reference points, and at least four fixed reference points are selected.

3. The method for dynamic monitoring of the vertical rotation and closure of an arch bridge based on camera measurement according to claim 2, characterized in that, Calculating the spatial location of preset measurement points in the target includes: The camera device is calibrated using targets at fixed reference points to calculate distortion parameters, obtain the extrinsic parameters of the camera device relative to the global coordinate system, and acquire the first frame image, storing the pixel templates of each target in the first frame image. Extract the subpixel image coordinates of the target in the image, and normalize the subpixel image coordinates using the distortion parameters; Based on the pose estimation of the target plane and combined with the extrinsic parameters, the pose transformation of the target's local coordinate system relative to the global coordinate system is obtained; The coordinates of the preset measurement point on the target in the global coordinate system are obtained based on the pose change, and the three-dimensional displacement of the preset measurement point relative to the initial moment is obtained based on the pixel template, so as to obtain the spatial position of the preset measurement point. Then, the spatial position is rigidly constrained and corrected using a target with a fixed reference point.

4. The method for dynamic monitoring of the vertical rotation and closure of an arch bridge based on camera measurement according to claim 2, characterized in that, The key points include the arch foot, suspension point, and arch crown of the target arch rib.

5. A dynamic monitoring method for the vertical rotation and closure of an arch bridge based on camera measurement according to claim 2, characterized in that, Given that the camera device is a single unit and the field of view is insufficient to cover the vertical rotation path of the arch rib, the camera device adopts a segmented relay acquisition method, so that adjacent fields of view overlap and include a common fixed reference point and a common target, in order to maintain the continuity and consistency of the cross-camera displacement results.

6. The method for dynamic monitoring of the vertical rotation and closure of an arch bridge based on camera measurement according to claim 5, characterized in that, At least 1 / 6 of the images acquired by the multiple camera devices overlap; wherein the images acquired by the multiple camera devices are fused based on quality weighting.

7. The method for dynamic monitoring of the vertical rotation and closure of an arch bridge based on camera measurement according to claim 1, characterized in that, The deviation estimation includes: Calculate the angle between the parameterized curve and the tangent vector to the model curve to obtain the linear deviation; The elevation deviation is obtained by calculating the difference between the height components of the parameterized curve and each preset measuring point in the model curve. The difference between the parameterized curve and the model curve at the closure end in three-dimensional space is calculated to obtain the closure end residual.

8. A dynamic monitoring system for the vertical rotation and closure of an arch bridge based on camera measurement, characterized in that, include: The target is attached to a predetermined position on the target arch rib; Camera equipment is installed on both sides of the target arch rib to acquire images of the target arch rib; A control device is used to employ a dynamic monitoring method for the vertical rotation and closure of an arch bridge based on camera measurement as described in any one of claims 1 to 7, to control the camera device to acquire images during the vertical rotation of the target arch rib, output deviation estimates, and determine the vertical rotation status of the arch rib based on the deviation estimates.

Images