Camera chain bridge deflection measurement method and measurement device considering atmospheric disturbance

Abstract

The application discloses a camera chain bridge deflection measurement method considering atmospheric disturbance, proposes to optimize the traditional collimation disturbance elimination structure from three aspects of dimension reduction, i.e., only paying attention to the camera attitude related to the measured displacement, considering the system servo attitude and the inclination angle sensing constraint, establishes a dimension reduction collimation disturbance elimination and bridge deflection measurement model under the coupling action of atmospheric disturbance and camera servo, and finally forms a camera chain bridge deflection online measurement method considering atmospheric disturbance image point drift. The measurement device is provided. The application provides important technical support for intelligent diagnosis and operation and maintenance of bridge structures, further improves the precision in long-distance measurement, only one infrared lamp is needed for the marker point of each measuring station in the application, the engineering implementation difficulty and the measurement cost are reduced, the data extraction time of response is shortened, and the cost is reduced.

Description

Technical Field

[0001] This invention relates to a method for monitoring the health of bridges, and more particularly to a method and apparatus for measuring the deflection of camera-chain bridges that takes into account atmospheric disturbances. Background Technology

[0002] To ensure the safe operation and reliability of bridges, since the 1990s, both domestic and international researchers have been developing and conducting theoretical research on bridge structural health monitoring systems. Deflection monitoring is one of the main components of bridge structural health monitoring. Among these methods, photogrammetry has the advantages of simultaneous multi-point monitoring across the entire field, real-time monitoring of both dynamic and static conditions, and precise positioning of measuring points. Therefore, it has been widely used in bridge deflection monitoring.

[0003] However, photogrammetry is susceptible to complex environmental influences. Atmospheric disturbance, a phenomenon where the constant state of the atmosphere is disrupted or altered by dynamic or thermal forces, has been proven to cause accuracy loss in precise optical measurements such as morphology, displacement, and velocity. Currently, the problem of atmospheric disturbance remains unresolved in the research and application of camera chains. In the field of precision photogrammetry, methods such as image restoration, signal filtering, flow field reduction, and camera system design are commonly used for disturbance reduction; however, these methods all have limitations in practical applications. Specifically, image restoration methods, which do not involve optical principles, have limited effectiveness in reducing disturbance for the quantitative deflection measurement to be conducted in this study; signal filtering cannot accurately separate atmospheric disturbance signals from numerous possible interference signals; the key to flow field reduction correction methods is establishing a relatively stable atmospheric model, but this method is not suitable for near-surface atmospheric environments with uneven refractive index distribution and rapid disturbance changes.

[0004] Zhang Jiaming et al. proposed a low-frequency disturbance reduction method based on the design of structured light (SL). This method, which further improves measurement accuracy through structured light design, is commonly used in deformation monitoring of large civil structures. Zhang Jiaming et al. applied the method of adding optical path constraints through structured light design to the study of atmospheric low-frequency disturbance reduction. Based on the assumptions that the optical axes of the two cameras are horizontal, parallel, and stationary, they analyzed the pose constraints of the imaging model and the pose change caused by the actual optical path deflection. They found that the errors in the measurements of the two cameras can cancel each other out. Therefore, by optimizing the calculated position and attitude parameters, the goal of reducing airflow disturbance errors can be achieved. However, this method still has some shortcomings in engineering applications. On the one hand, the required array of multi-point targets makes system design and initial calibration more complex, and its engineering applicability needs to be improved. Furthermore, this method assumes that the camera and target are stationary, which has limitations. On the other hand, this method is based on pose recognition theory, and the commonly used pose recognition method is the iterative n-point pose solving algorithm PnP (perspective-n-point). This method itself has limitations, including long processing time and the possibility of getting trapped in local optima. Most importantly, the angular amplitude of a typical camera following the bridge structure's rotation is generally at the arcminute or arcsecond level, while the phase attitude angle calculation accuracy based on the PnP method is only around 0.02° in ideal indoor environments. This means that in long-distance measurement applications in complex environments, its attitude angle resolution cannot meet practical requirements. Summary of the Invention

[0005] Objective of the Invention: To address the aforementioned problems, the objective of this invention is to provide a camera chain bridge deflection measurement method that considers atmospheric disturbances. This method establishes an online bridge deflection measurement method with good engineering applicability while meeting the operating characteristics of a servo camera chain system. In particular, it corrects for image point drift caused by atmospheric disturbances, resulting in more accurate bridge deflection measurement results under complex atmospheric conditions and higher precision for long-distance measurements using the camera chain. A measurement device for this method is also provided.

[0006] Technical solution: A method for measuring the deflection of a camera chain bridge considering atmospheric disturbances, comprising the following steps:

[0007] S1: The camera chain and the high-precision tilt sensor mounted on each camera are ready;

[0008] S2: Determine the object distance d from the cooperative target to its monitored camera, and calculate the squareness coefficient k:

[0009]

[0010] Where f is the lens focal length and l is the pixel size of the camera target surface;

[0011] S3: The tilt sensor tracks the amount of rotational attitude change of each camera, including the pitch angle θ and the rotation angle β;

[0012] S4: The camera acquires a computational image. Using the initial computational image as a reference, a coarse computational window M×N for the target is determined in the initial reference image Frame_0 and remains unchanged. The center pixel coordinates (x, y, y) of the cooperative target in the image sequence Frame_j are determined using the gray-scale centroid method. j ,y j ):

[0013]

[0014] Where g(x,y) represents the gray level at pixel coordinates (x,y); from (x j ,y j Determine the vertical displacement h of the cooperative target in the acquired image at each image acquisition time. j :

[0015] h j =y j -y0 (3)

[0016] S5: Determine the pixel distance from the target center to the image center in the initial image frame Frame_0:

[0017]

[0018] Among them, (x c ,y c The value ) represents the pixel coordinates of the image center. Here, the parameter r is only calculated in the initial image and is used in subsequent images. It is a known constant.

[0019] S6: Substitute the above parameters into the eye-to-eye monitoring model that considers changes in camera angle and attitude, as well as atmospheric disturbances:

[0020]

[0021] Where c(i) represents the i-th camera station, and M(i) represents the cooperative marker point on the i-th camera station; ΔY M ΔYC represents the actual vertical displacement of the cooperative marker itself; ΔYC represents the actual vertical displacement of the camera. This represents the imaging deviation of the cooperative marker point M(i+1) in the y-direction of camera C(i) caused by atmospheric disturbance. Similarly, α represents the imaging deviation of the cooperative marker point M(i) in the y-direction of the camera C(i+1) caused by atmospheric disturbance; α represents the angle between the line connecting the target image point position and the image center point and the horizontal axis.

[0022] The constraints of this equation are as follows:

[0023] a. Based on existing research findings

[0024] b. The vertical displacement of the marker point and the fixed camera station is consistent, and the vertical displacement of the dual-head camera at each station is also consistent, that is, ΔY c(i) =ΔY M(i) ;

[0025] S7: Solve the equation system (5) using principal component analysis to obtain ΔY, which is the vertical displacement of the structure where the station is located, i.e., the deflection.

[0026] Furthermore, in step S6, ΔY M This is the vertical displacement of the structural position of the cooperative marker point. For a stable marker point installed at the end of the camera chain, this value is zero.

[0027] Furthermore, in step S6, ΔY C This refers to the vertical displacement of the bridge structure where the camera is located.

[0028] Furthermore, in step S6, the imaging offset of atmospheric disturbance is set to one dimension, that is, the cooperative marker point corresponding to Δy related to the vertical displacement is a single point.

[0029] This method's monitoring model considers imaging shifts caused by atmospheric disturbances, a factor not currently addressed in camera chain monitoring models. Furthermore, it improves upon existing line-of-sight imaging de-interference methods, which consider six degrees of freedom and use a multi-point planar target as the cooperation marker. This reduces the consideration of imaging shifts due to atmospheric disturbances to one dimension, specifically Δy, related to vertical displacement, with a single-point cooperation marker. This simplifies the model and reduces engineering implementation difficulty and measurement costs. In the monitoring model, the camera's rotational attitude parameters are not calculated using equations but are directly read from angle sensors mounted on the camera, making the process faster and increasing the accuracy of long-distance measurements.

[0030] A measuring device using the above-mentioned camera chain bridge deflection measurement method considering atmospheric disturbance includes multiple camera detection systems installed at intervals along the extension direction of the bridge to form a camera chain. Each camera detection system includes a camera, an tilt sensor, a marker point, and a wireless network card. The camera has an embedded CPU and online image processing software. There are two cameras, which are fixed back-to-back in the transverse direction to form a dual-head reverse image measurement system. A marker point is installed in the middle of the upper surface, and a tilt sensor is installed in the middle of the lower surface. The wireless network card is installed on one of the cameras. The camera and the tilt sensor are respectively connected to a remote control terminal via the wireless network card. Each dual-head reverse image measurement system observes the marker points on the two adjacent camera detection systems.

[0031] Furthermore, the markers are infrared light targets, and a camera detection system is set up at each of the four or eight equal divisions of the total length of the bridge.

[0032] The dual-head reverse image measurement system is used to track marker point images. The marker point can be simultaneously observed by cameras on the two dual-head reverse image measurement systems. A high-precision tilt sensor is used to track the rotation angle of the structure as it moves along with the bridge structure. During installation, it is necessary to ensure that one axis is parallel to the optical axis of the dual-head reverse image measurement system as much as possible. Each station has an embedded CPU and online image processing software, which can extract the pixel displacement of the target in the image. The wireless network card transmits the extracted displacement information to the remote control terminal. The terminal computer further analyzes the vertical displacement of the cooperative marker point (i.e., the position of the bridge structure where the station is located), i.e., the deflection. The data can be viewed and retrieved in real time at the remote control terminal.

[0033] Beneficial effects: Compared with existing technologies, the advantages of this invention are: it further considers atmospheric disturbance issues, exhibits better resistance to atmospheric disturbances in long-distance measurements, and has higher long-distance measurement accuracy and better engineering feasibility. This invention uses a high-precision tilting instrument to track the camera's rotation attitude, further improving accuracy in long-distance measurements. Each station marker only requires one infrared lamp, reducing engineering implementation difficulty and measurement costs, shortening data extraction time, and lowering overall costs. Attached Figure Description

[0034] Figure 1 This is a flowchart of the measurement method of the present invention;

[0035] Figure 2 This is a schematic diagram of the measuring device structure of the present invention;

[0036] Figure 3 This is a schematic diagram of the monitoring model construction process of the present invention, wherein,

[0037] (a) represents the attitude change parameters of the camera as it changes with the structure, which are considered in this method; ① represents the camera C. i The follower attitude variables, ② are the camera C i+1 The follow-up attitude variables;

[0038] (b) represents the image point shift in the image formed by the camera on the opposite structure due to atmospheric disturbances, as considered in this method; ① represents the image point shift of the camera C caused by atmospheric disturbances. i The imaging shift, ② is the camera C caused by atmospheric disturbance. i+1 Imaging offset;

[0039] (c) is a schematic diagram of a bridge deflection measurement system. Detailed Implementation

[0040] The present invention will be further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention.

[0041] A method for measuring the deflection of a camera chain bridge that takes into account atmospheric disturbances, such as Figure 1 As shown, it includes the following steps:

[0042] S1: The camera chain and the high-precision tilt sensor mounted on each camera are ready;

[0043] S2: Determine the object distance d from the cooperative target to its monitored camera, such as... Figure 3 As shown in (a), calculate the Fangda coefficient k:

[0044]

[0045] Where f is the lens focal length and l is the pixel size of the camera target surface;

[0046] S3: The tilt sensor tracks the changes in the rotational attitude of each camera, including the pitch angle θ and the rotation angle β, such as... Figure 3 As shown in (a), the atmospheric disturbance is denoted as A;

[0047] S4: The camera acquires a computational image. Using the initial computational image as a reference, a coarse computational window M×N for the target is determined in the initial reference image Frame_0 and remains unchanged. The center pixel coordinates (x, y, y) of the cooperative target in the image sequence Frame_j are determined using the gray-scale centroid method. j ,y j ):

[0048]

[0049] Where g(x,y) represents the gray level at pixel coordinates (x,y); from (x j ,y jDetermine the vertical displacement h of the cooperative target in the acquired image at each image acquisition time. j :

[0050] h j =y j -y0 (3)

[0051] S5: Determine the pixel distance from the target center to the image center in the initial image frame Frame_0:

[0052]

[0053] Among them, (x c ,y c The value ) represents the pixel coordinates of the image center. Here, the parameter r is only calculated in the initial image and is used in subsequent images. It is a known constant.

[0054] S6: Substitute the above parameters into the eye-to-eye monitoring model that considers changes in camera angle and attitude, as well as atmospheric disturbances:

[0055]

[0056] Where c(i) represents the i-th camera station, and M(i) represents the cooperative marker on the i-th camera station; ΔY M This represents the actual vertical displacement of the cooperative marker itself, that is, the vertical displacement of the structural position where the cooperative marker is located. Therefore, for markers installed on structural supports at the end of the camera chain, this value is zero; ΔY C This indicates the actual vertical displacement of the camera, which is the vertical displacement of the bridge structure where the camera is located. This represents the imaging deviation of the cooperative marker point M(i+1) in the y-direction of camera C(i) caused by atmospheric disturbance. Similarly, This represents the imaging deviation of the cooperative marker point M(i) in the y-direction within camera C(i+1) caused by atmospheric disturbance, where α represents the angle between the line connecting the target image point and the image center point and the horizontal axis. Figure 3 As shown in (b). The constraints of this equation are as follows:

[0057] a. Based on existing research findings

[0058] b. The vertical displacement of the marker point and the fixed camera station is consistent, and the vertical displacement of the dual-head camera at each station is also consistent, that is, ΔY c(i) =ΔY M(i) ;

[0059] S7: Solve the equation system (5) using principal component analysis to obtain ΔY, which is the vertical displacement of the structure where the station is located, i.e., the deflection.

[0060] This bridge deflection measurement method can effectively reduce measurement errors caused by atmospheric disturbances and improve the accuracy of camera chains when operating at long distances.

[0061] The measuring device using the above-mentioned camera chain bridge deflection measurement method that takes atmospheric disturbance into account, such as... Figure 2 , Figure 3 As shown in (c), a camera detection system comprises multiple cameras installed at intervals along the bridge's extension direction to form a camera chain. Each camera detection system includes a camera 1, an tilt sensor 2, marker points 3, and a wireless network card 4. Camera 1 has an embedded CPU and online image processing software. Two cameras 1 are fixed back-to-back in the transverse direction to form a dual-head reverse image measurement system. A marker point 3 is installed in the middle of the upper surface, and the tilt sensor 2 is installed in the middle of the lower surface. The marker point 3 is an infrared target. The spacing between adjacent image measurement systems can be determined according to the bridge's measurement point layout requirements and the focal length of the lens used. It can be one camera detection system at each of the 4 to 12 equal divisions of the bridge's total length, often eight or four divisions. For long-span bridges, the spacing can be appropriately increased, or cameras 1 equipped with telephoto lenses, such as 100mm or 200mm fixed-focus lenses, can be selected. A wireless network card 4 is installed on one of the cameras 1. Camera 1 and tilt sensor 2 are connected to the remote control terminal 5 via the wireless network card 4. Each dual-head reverse image measurement system observes the marker points 3 on the two adjacent camera detection systems. The optical axis formed by the multiple cameras 1 on the camera chain is aligned with the direction of the bridge axis.

Claims

1. A method for measuring the deflection of a camera chain bridge considering atmospheric disturbance, characterized in that... Includes the following steps: S1: The camera chain and the high-precision tilt sensor mounted on each camera are ready; S2: Determine the object distance d from the cooperative target to its monitored camera, and calculate the magnification factor k: Where f is the lens focal length and l is the pixel size of the camera target surface; S3: The tilt sensor tracks the amount of rotational attitude change of each camera, including the pitch angle θ and the rotation angle β; S4: The camera acquires a computational image. Using the initial computational image as a reference, a coarse computational window M×N for the target is determined in the initial reference image Frame_0 and remains unchanged. The center pixel coordinates (x, y, y) of the cooperative target in the image sequence Frame_j are determined using the gray-scale centroid method. j ,y j ): Where g(x,y) represents the gray level at pixel coordinates (x,y); from (x j ,y j Determine the vertical displacement h of the cooperative target in the acquired image at each image acquisition time. j : h j =y j -y0 (3) S5: Determine the pixel distance from the target center to the image center in the initial image frame Frame_0: Among them, (x c ,y c The value ) represents the pixel coordinates of the image center. Here, the parameter r is only calculated in the initial image and is used in subsequent images. It is a known constant. S6: Substitute the above parameters into the eye-to-eye monitoring model that considers changes in camera angle and attitude, as well as atmospheric disturbances: Where c(i) represents the i-th camera station, and M(i) represents the cooperative marker point on the i-th camera station; ΔY M ΔY represents the actual vertical displacement of the cooperation marker itself. c This indicates the actual vertical displacement of the camera; This represents the imaging deviation of the cooperative marker point M(i+1) in the y-direction of camera C(i) caused by atmospheric disturbance. Similarly, α represents the imaging deviation of the cooperative marker point M(i) in the y-direction of the camera C(i+1) caused by atmospheric disturbance; α represents the angle between the line connecting the target image point position and the image center point and the horizontal axis. The constraints of this equation are as follows: a. Based on existing research findings b. The vertical displacement of the marker point and the fixed camera station is consistent, and the vertical displacement of the dual-head camera at each station is also consistent, that is, ΔY c(i) =ΔY M(i) ; S7: Solve the equation system (5) using principal component analysis to obtain ΔY, which is the vertical displacement of the structure where the station is located, i.e., the deflection.

2. The method for measuring the deflection of a camera chain bridge considering atmospheric disturbance as described in claim 1, characterized in that: In step S6, ΔY M This is the vertical displacement of the structural position of the cooperative marker point. For a stable marker point installed at the end of the camera chain, this value is zero.

3. The method for measuring the deflection of a camera chain bridge considering atmospheric disturbance as described in claim 1, characterized in that: In step S6, ΔY C This refers to the vertical displacement of the bridge structure where the camera is located.

4. The method for measuring the deflection of a camera chain bridge considering atmospheric disturbance as described in claim 1, characterized in that: In step S6, the imaging offset of atmospheric disturbance is set to one dimension, that is, the cooperative marker point corresponding to Δy related to vertical displacement is a single point.

5. A measuring device using the camera chain bridge deflection measurement method considering atmospheric disturbance as described in any one of claims 1 to 4, characterized in that: The system comprises multiple camera detection systems installed at intervals along the bridge's extension direction to form a camera chain. Each system includes a camera, tilt sensor, marker point, and wireless network card. The camera has an embedded CPU and online image processing software. Two cameras are fixed back-to-back in the transverse direction to form a dual-head reverse image measurement system. A marker point is installed in the middle of the upper surface, and a tilt sensor is installed in the middle of the lower surface. The wireless network card is installed on one of the cameras. The camera and tilt sensor are connected to a remote control terminal via the wireless network card. Each dual-head reverse image measurement system observes the marker points on the two adjacent camera detection systems.

6. The measuring device for measuring the deflection of a camera chain bridge considering atmospheric disturbance as described in claim 5, characterized in that: The markers are infrared light targets, and a camera detection system is set up at each of the four or eight equal divisions of the total length of the bridge.

Images