Truss arch bridge hanger safety monitoring system and method based on image recognition technology

Abstract

The present invention relates to a method and system for monitoring the safety of a truss arch bridge hanger based on image recognition technology, the method comprising the following steps: providing a marker, and attaching the marker to the hanger; providing a binocular camera, and photographing the marker on the hanger through the binocular camera to obtain video data; extracting corresponding image data from the video data, and identifying the marker in the image data based on image recognition technology to obtain the horizontal displacement of the marker; obtaining a time curve of the marker according to the horizontal displacement of the marker and the corresponding time information; calculating the fundamental frequency of the hanger using the time curve of the marker; and calculating the hanger force according to the fundamental frequency. The method for monitoring the safety of the hanger of the present invention can greatly solve the problem of high cost of hardware equipment. In addition, the measurement accuracy of the binocular camera can collect the tiny vibration of the hanger, thereby obtaining the safety index of the hanger, and the test range of the hanger is wide.

Description

Technical Field

[0001] The present invention relates to the technical field of bridge structure safety, and in particular to a truss arch bridge hanger safety monitoring system and method based on image recognition technology. Background Art

[0002] Truss arch bridges are particularly favored in design selection due to their reasonable force and beautiful appearance. The main and key load-bearing components of truss arch bridges are hangers. Due to the small mass, low damping and high flexibility of the hangers, they are very likely to vibrate under the action of wind, rain and traffic alternating loads, causing fatigue effects and possibly fatigue fractures. Vibration fatigue of the hangers of arch bridges is a common form of local damage to bridge structures. If it is not handled in time and allowed to develop, it will first cause the hanger wire to break, which will lead to the redistribution of internal forces, and may eventually lead to extreme situations where the hangers break one after another, with very serious consequences. Therefore, it is very important to obtain safety indicators such as the force and vibration of the hangers in real time, and then quantitatively evaluate the current state of the structure.

[0003] At present, the monitoring of boom safety indicators is mainly carried out by installing vibration sensors on the boom to obtain the vibration amplitude and vibration frequency, and then solving the boom force based on the relationship between the frequency and the boom force. Although this monitoring method can produce real and reliable monitoring data, due to the limitation of limited spatial resolution, it is usually necessary to install a very dense sensor array, and professional instruments and equipment are required to decode and post-process the signals collected by the sensors. As a result, the hardware cost and installation process of the sensors and corresponding equipment are also very expensive. In addition, once a large number of sensors are installed and operated, a large amount of monitoring data will be collected, and the data processing volume is large. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the prior art and provide a truss arch bridge hanger safety monitoring system and method based on image recognition technology to solve the problems of high cost and large amount of data processing in existing sensor detection.

[0005] The technical solution to achieve the above purpose is:

[0006] The present invention provides a truss arch bridge hanger safety monitoring method based on image recognition technology, comprising the following steps:

[0007] Providing a marking piece, and attaching the marking piece to the suspension rod;

[0008] Providing a binocular camera, and photographing the marking member on the boom by the binocular camera to obtain video data;

[0009] Extracting corresponding image data from the video data, and identifying the marker in the image data based on image recognition technology to obtain the horizontal displacement of the marker;

[0010] Obtaining a time course curve of the marker according to the horizontal displacement of the marker and the corresponding time information;

[0011] Calculating the fundamental frequency of the boom using the time history curve of the marker; and

[0012] The boom force is calculated according to the following formula, and the boom safety monitoring is realized by the boom force.

[0013] T=4WS 2 F 2 / 1000

[0014] In the above formula, T represents the boom force, W represents the linear density of the boom, S represents the length of the boom, and F represents the fundamental frequency of the boom.

[0015] The present invention uses image recognition technology to conduct safety monitoring of the boom force of the boom. Its hardware equipment is mainly a binocular camera with a market price of 30,000 yuan. In the existing sensor detection method, the price of a sensor is about 2,000 yuan. It is necessary to install a sensor on each boom. If a 16-channel demodulator is used, it costs about 100,000 yuan. If 10 booms are installed on the arch bridge, the hardware cost of the existing sensor detection is 120,000 yuan. It can be seen that the boom safety monitoring method of the present invention can greatly solve the problem of high hardware equipment costs. In addition, the measurement accuracy of the binocular camera can collect the tiny vibrations of the boom, thereby obtaining the safety index of the boom, and the test range of the boom is wide. However, in the existing sensor detection method, due to the obvious stiffness effect of the short boom, its vibration amplitude is small, and the sensor cannot measure its vibration acceleration.

[0016] A further improvement of the truss arch bridge hanger safety monitoring method based on image recognition technology of the present invention is that the step of calculating the fundamental frequency of the hanger using the time history curve of the marker comprises:

[0017] Taking a second derivative of the time history curve to obtain the vibration acceleration of the suspension rod;

[0018] The vibration acceleration of the suspension rod is subjected to Fourier transformation to obtain the fundamental frequency of the suspension rod.

[0019] A further improvement of the truss arch bridge hanger safety monitoring method based on image recognition technology of the present invention is that when obtaining the video data of the marker through the binocular camera, the marker is continuously photographed for a certain period of time at intervals to obtain the video data.

[0020] A further improvement of the truss arch bridge hanger safety monitoring method based on image recognition technology of the present invention is that the provided marking member includes a pair of transverse marking strips and a pair of longitudinal marking strips, and the pair of transverse marking strips and the pair of longitudinal marking strips are staggered and connected to form a tic-tac-toe structure;

[0021] When obtaining the time course curve of the marker, the horizontal displacement and the vertical displacement of the tic-tac-toe structure in the image data are measured, and then the time course curve of the marker is obtained in combination with the time information.

[0022] A further improvement of the truss arch bridge suspender safety monitoring method based on image recognition technology of the present invention is that when setting a binocular camera, a corresponding number of suspenders are selected as a measurement group according to the resolution of the binocular camera;

[0023] Measure the distance between two adjacent booms in a measurement group, and take the average value of the measured distances as the installation distance;

[0024] The binocular camera is set on the arch bridge along the midline position of the range where a measurement group is located, and the distance between the binocular camera and the center point of the range where the corresponding measurement group is located is equal to the installation distance. The binocular camera is used to photograph the suspension rods in the measurement group respectively.

[0025] The present invention also provides a truss arch bridge hanger safety monitoring system based on image recognition technology, comprising:

[0026] A marking piece attached to the boom;

[0027] A binocular camera provided on the arch bridge and corresponding to the marking piece on the suspension rod, the binocular camera being used to photograph the marking piece to form video data;

[0028] The processing unit connected to the binocular camera is used to extract corresponding image data from the video data, identify the marker in the image data based on the image recognition technology to obtain the horizontal displacement of the marker, and then calculate the time history curve through the horizontal displacement, and calculate the fundamental frequency of the boom using the time history curve, so as to calculate the boom force through the following formula, and realize the safety monitoring of the boom through the boom force,

[0029] T=4WS 2 F 2 / 1000

[0030] In the above formula, T represents the boom force, W represents the linear density of the boom, S represents the length of the boom, and F represents the fundamental frequency of the boom.

[0031] A further improvement of the truss arch bridge hanger safety monitoring system based on image recognition technology of the present invention is that the processing unit is used to perform a second derivative of the time history curve to obtain the vibration acceleration of the hanger, and is also used to perform Fourier transform on the vibration acceleration of the hanger to obtain the fundamental frequency of the hanger.

[0032] A further improvement of the truss arch bridge hanger safety monitoring system based on image recognition technology of the present invention is that the binocular camera continuously shoots the marking parts for a certain period of time at intervals to form video data.

[0033] A further improvement of the truss arch bridge hanger safety monitoring system based on image recognition technology of the present invention is that the marking member comprises a pair of transverse marking strips and a pair of longitudinal marking strips, and the pair of transverse marking strips and the pair of longitudinal marking strips are staggered and connected to form a tic-tac-toe structure;

[0034] The processing unit is also used to measure the horizontal displacement and vertical displacement of the tic-tac-toe structure in the image data, and then obtain the time course curve of the marking element in combination with the time information.

[0035] A further improvement of the truss arch bridge hanger safety monitoring system based on image recognition technology of the present invention is that the binocular camera is arranged in front of multiple hangers to take pictures of each hanger separately. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 The present invention is a flow chart of a truss arch bridge hanger safety monitoring method based on image recognition technology.

[0037] Figure 2 It is a schematic diagram of the structure of an arch bridge monitored by the truss arch bridge hanger safety monitoring system and method based on image recognition technology of the present invention.

[0038] Figure 3 It is a structural schematic diagram of the identification parts in the truss arch bridge hanger safety monitoring system and method based on image recognition technology of the present invention. DETAILED DESCRIPTION

[0039] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0040] See also Figure 1The present invention provides a truss arch bridge suspender safety monitoring system and method based on image recognition technology. By collecting high-frequency video with a binocular camera and extracting images for recognition, high-resolution collection of spatial information can be achieved, and the spatial information of the entire field of view is encoded in a non-contact manner, avoiding the high cost and complex installation process of contact sensors. The present invention uses image recognition technology to truly realize the automation and intelligence of the monitoring process, ensuring the stability and accuracy of the results. The truss arch bridge suspender safety monitoring system and method based on image recognition technology of the present invention are described below in conjunction with the accompanying drawings.

[0041] The present invention provides a truss arch bridge suspender safety monitoring system based on image recognition technology, such as Figure 2 As shown, it is used to monitor the suspension rod force of the suspension rod 11 on the arch bridge 10, combined with Figure 3 As shown, the safety monitoring system includes a marker 21, a binocular camera and a processing unit, wherein the marker 21 is attached to the suspension rod 11, the binocular camera is arranged on the arch bridge 10 and corresponds to the marker on the suspension rod 11, and the binocular camera is used to shoot the marker to form video data; the processing unit is connected to the binocular camera, and is used to receive the video data formed by the binocular camera. The processing unit is used to extract corresponding image data from the video data, and identify the marker in the image data based on the image recognition technology to obtain the horizontal displacement of the marker, and then calculate the time history curve through the horizontal displacement, and calculate the fundamental frequency of the suspension rod by using the time history curve, so as to calculate the suspension rod force through the following formula, and realize the safety monitoring of the suspension rod through the suspension rod force:

[0042] T=4WS 2 F 2 / 1000

[0043] In the above formula, T represents the boom force, W represents the linear density of the boom, S represents the length of the boom, and F represents the fundamental frequency of the boom. The length of the boom is the distance between the upper and lower anchor clamps of the boom, which can be obtained by measurement. The linear density of the boom is calculated from the length and density of the boom.

[0044] like Figure 2 As shown, a truss is installed on the arch bridge 10, and the arched part of the truss is connected to the bridge deck 12 through a hanger 11. A plurality of hangers 11 are arranged at equal intervals. The hangers play a vital role in the stability of the truss. Therefore, the hanger force of the hanger is an important monitoring indicator for the safety of the arch bridge structure. The safety monitoring system of the present invention collects video data of the hanger, extracts image data, and then obtains the horizontal displacement of the hanger based on image recognition technology. Since the vibration load on the hanger is mostly reflected in the horizontal displacement of the hanger, the fundamental frequency of the hanger is calculated by monitoring the horizontal displacement of the hanger, which can more accurately reflect the hanger force and provide a guarantee for the structural safety of the arch bridge.

[0045] Compared with the traditional means of obtaining boom force indicators based on sensors, the present invention monitors boom force based on image recognition technology. Images are a dense perception means with a wider source, lower acquisition cost, and more intuitive information reflection, which has obvious advantages. In addition, the monitoring system of the present invention reduces the dependence on subjective experience and consciousness in manual detection, thereby truly realizing the automation and intelligence of the monitoring process and ensuring the stability and accuracy of the results. The safety monitoring system of the present invention can greatly reduce the cost of hardware equipment and simplify the hardware installation program. In addition, the lengths of the booms are different. The length of the booms set at the smaller arch of the truss is smaller. The short boom has obvious stiffness effect and small vibration amplitude. Traditional sensors cannot monitor its vibration acceleration, so it is impossible to measure its boom force. The binocular camera of the present invention has high measurement accuracy and can collect the tiny vibration of the boom. Even a short boom can monitor its boom force, thereby obtaining the safety index of the boom, and the test range of the boom is wide.

[0046] In a specific embodiment of the present invention, the processing unit includes a receiving module and an image extraction module. The receiving module is communicatively connected to the binocular camera and is used to receive video data captured by the binocular camera; the image extraction module is connected to the receiving module, and the image extraction module is used to extract the image from the video data and extract the image at a 2 frame rate to form image data.

[0047] Furthermore, the processing unit also includes an image recognition module connected to the image extraction module and a calculation module connected to the image recognition module. The image recognition module receives the image data formed by the image extraction module, recognizes each image in the image data in chronological order, and recognizes the specific position of the identification member in the image; the calculation module receives the specific position of the identification member in each image recognized by the image recognition module, which is equivalent to obtaining the position change information of the identification member within a period of time, and can also calculate the time course curve of the identification member. Preferably, the calculation module calculates the horizontal displacement of the identification member corresponding to each time point according to the actual size of the identification member and the specific position of the identification member, and calculates the time course curve of the identification member using the horizontal displacement and time information.

[0048] In a specific embodiment of the present invention, the processing unit is used to perform a second derivative of the time history curve to obtain the vibration acceleration of the suspension rod, and is also used to perform Fourier transform on the vibration acceleration of the suspension rod to obtain the fundamental frequency of the suspension rod.

[0049] Preferably, the calculation module performs a derivation of the time history curve to obtain a velocity, which is the velocity of the suspension rod, and records and stores the velocity of the suspension rod. The calculation module performs a derivation of the velocity of the suspension rod to obtain an acceleration, which is the vibration acceleration of the suspension rod, and records and stores the vibration acceleration of the suspension rod. The calculation module performs a Fourier transform on the vibration acceleration of the suspension rod to obtain a vibration frequency, which is the fundamental frequency of the suspension rod vibration.

[0050] In a specific embodiment of the present invention, the binocular camera continuously shoots the marking part for a certain period of time at intervals to form video data. Preferably, the binocular camera continuously captures 1 minute of boom vibration video at 60fps every 10 minutes, and the monitoring time of each boom is 1 hour each time, so 6 minutes of boom vibration video is obtained.

[0051] In a specific embodiment of the present invention, Figure 3 As shown, the mark 21 includes a pair of transverse marking bars 211 and a pair of longitudinal marking bars 212, which are staggered and connected to form a tic-tac-toe structure; the processing unit is also used to measure the horizontal displacement and vertical displacement of the tic-tac-toe structure in the image data, and then combine the time information to obtain the time curve of the marking part.

[0052] The vertical displacement of the pair of transverse marking strips 211 is the vertical displacement of the boom, and the transverse displacement of the pair of longitudinal marking strips 212 is the horizontal displacement of the boom. The processing unit of the present invention considers both the horizontal displacement and the vertical displacement of the boom during image recognition, which can improve the monitoring accuracy of the boom force and improve safety.

[0053] Preferably, when measuring the horizontal displacement and the vertical displacement, the corresponding horizontal displacement and the vertical displacement are obtained by measuring the displacement of the pixel points.

[0054] Furthermore, a center mark point 213 is provided at the center of the tic-tac-toe structure, and the tic-tac-toe structure is calibrated using the center mark point 213. When shooting a video, the binocular camera can be aligned with the center mark point 213. When performing image recognition, the center mark point 213 is used for image alignment to measure the vertical displacement of a pair of horizontal mark bars 211 and the horizontal displacement of a pair of vertical mark bars 212.

[0055] In a specific embodiment of the present invention, a binocular camera is disposed in front of a plurality of booms to photograph each boom separately.

[0056] Preferably, combined Figure 2 As shown, the booms 11 are divided into multiple groups, and the division is carried out according to the arrangement order of the booms 11, and every 4 booms can be divided into a group. A binocular camera is set up in front of a group of booms, and each boom in the group is photographed.

[0057] Take the No. 1 to No. 4 booms as an example to illustrate the installation of the binocular camera. Measure the distance between two adjacent booms from No. 1 to No. 4 booms, and take the average of all the distances to get the average distance. Set the binocular camera on the center line of the No. 1 to No. 4 booms, and the distance between the binocular camera and the middle point of the No. 1 to No. 4 booms is equal to the average distance. Then the binocular camera can collect videos of the No. 1 to No. 4 booms at this position.

[0058] The beneficial effects of the truss arch bridge hanger safety monitoring system based on image recognition technology of the present invention are:

[0059] The present invention can not only realize stable tracking of time-series images through image feature matching and random consistency methods to ensure that the three-dimensional vibration monitoring accuracy reaches 0.5mm, but also can be used through networking to quickly build a multi-point synchronous observation system.

[0060] The safety monitoring system of the present invention has the advantages of low equipment cost and convenient construction. When using the traditional sensor test boom, each boom requires a sensor, and the price of each sensor is about 2,000 yuan. It needs to be equipped with a demodulator, and the price of a 16-channel demodulator is 100,000 yuan. The purchase, construction, design, and installation of sensors and demodulation equipment all require certain costs. The main equipment of the present invention is a binocular camera, which costs about 30,000 yuan. Compared with the wiring and installation of the sensor, the binocular camera of the present invention only needs to be powered on for shooting, and the marking piece attached to the boom can be a sticker, which is low in cost and easy to attach.

[0061] The safety monitoring system of the present invention has a wide test range. The short suspension rods on the arch bridge have obvious stiffness effects, and the vibration acceleration of the suspension rods cannot be obtained through sensors. However, the frame rate of the binocular camera of the present invention is 60fps, which is accurate enough to collect the tiny vibrations of the short suspension rods and obtain the safety indicators of all suspension rods.

[0062] The safety monitoring system of the present invention has comprehensive test indicators. When using a sensor to detect the vibration acceleration of the suspension rod, it is impossible to calculate the vibration velocity and displacement response, but the image recognition of the present invention obtains the vibration displacement time history curve of the suspension rod, and the corresponding velocity and acceleration can be obtained by derivation.

[0063] The safety monitoring system of the present invention can reduce the influence of external human factors, and the data is safe and reliable. The sensor detection method often causes power outages and network disconnections due to the complex conditions of the bridge site, resulting in data loss. The test time of the present invention is short, and all test data can be safely and reliably collected and saved.

[0064] The safety monitoring system of the present invention is easy to operate. Sensor detection requires filtering, elimination and filling due to data packet loss and other problems, while the image acquisition and data analysis of the present invention can be automatically realized by using a packaged program, which is easy to operate.

[0065] The image recognition-based monitoring system for the safety index of the suspender of the present invention has a simple structure, low cost, high accuracy, wide test range and strong applicability. It not only obtains the vibration acceleration, vibration velocity and amplitude of the suspender to evaluate the safety of the internal steel wire of the suspender, but also obtains the apparent image and video of the suspender, which is convenient for the bridge management department to track the apparent changes of the suspender and check the apparent defects in time.

[0066] The present invention also provides a truss arch bridge hanger safety monitoring method based on image recognition technology, and the safety monitoring method is described below.

[0067] like Figure 1 As shown, the truss arch bridge hanger safety monitoring method based on image recognition technology of the present invention comprises the following steps:

[0068] Execute step S101, provide a marking piece, and attach the marking piece to the suspension rod; then execute step S102;

[0069] Execute step S102, provide a binocular camera, and use the binocular camera to shoot the marking part on the boom to obtain video data; then execute step S103;

[0070] Execute step S103, extract corresponding image data from the video data, identify the marker in the image data based on image recognition technology to obtain the horizontal displacement of the marker; then execute step S104;

[0071] Execute step S104 to obtain a time course curve of the marking member according to the horizontal displacement of the marking member and the corresponding time information; then execute step S105;

[0072] Execute step S105, calculate the fundamental frequency of the boom using the time history curve of the marker; then execute step S106;

[0073] Execute step S106, calculate the boom force according to the following formula, and implement boom safety monitoring through the boom force.

[0074] T=4WS 2 F 2 / 1000

[0075] In the above formula, T represents the boom force, W represents the linear density of the boom, S represents the length of the boom, and F represents the fundamental frequency of the boom.

[0076] The boom safety monitoring method of the present invention can greatly solve the problem of high hardware equipment costs. In addition, the measurement accuracy of the binocular camera can collect the tiny vibration of the boom, thereby obtaining the safety index of the boom, and the test range of the boom is wide. However, the existing sensor detection method cannot measure its vibration acceleration due to the obvious stiffness effect of the short boom and its small vibration amplitude.

[0077] In a specific embodiment of the present invention, the step of calculating the fundamental frequency of the boom using the time history curve of the marker comprises:

[0078] Take the second derivative of the time history curve to obtain the vibration acceleration of the hanger;

[0079] The vibration acceleration of the suspension rod is transformed by Fourier to obtain the fundamental frequency of the suspension rod.

[0080] Furthermore, the time history curve is differentiated once to obtain the velocity of the suspension rod, and the velocity of the suspension rod is differentiated again to obtain the vibration acceleration of the suspension rod.

[0081] In a specific embodiment of the present invention, when obtaining the video data of the marker by the binocular camera, the marker is continuously photographed for a certain period of time at intervals to obtain the video data. Preferably, the binocular camera is used to continuously collect 1 minute of boom vibration video at 60fps every 10 minutes, and each boom is monitored for 1 hour each time, so that 6 minutes of boom vibration video is obtained.

[0082] In a specific embodiment of the present invention, the provided marking member includes a pair of transverse marking strips and a pair of longitudinal marking strips, wherein the pair of transverse marking strips and the pair of longitudinal marking strips are staggered and connected to form a tic-tac-toe structure;

[0083] When obtaining the time course curve of the marker, the horizontal displacement and vertical displacement of the tic-tac-toe structure in the image data are measured, and then the time course curve of the marker is obtained in combination with the time information.

[0084] In a specific embodiment of the present invention, when setting the binocular camera, a corresponding number of booms are selected as a measurement group according to the resolution of the binocular camera;

[0085] Measure the distance between two adjacent booms in a measurement group, and take the average value of the measured distances as the installation distance;

[0086] A binocular camera is set on the arch bridge along the midline of a measurement group, and the distance between the binocular camera and the center point of the corresponding measurement group is equal to the installation distance. The binocular camera is used to shoot the booms in the measurement group respectively.

[0087] The present invention is described in detail above in conjunction with the embodiments of the accompanying drawings. A person skilled in the art can make various variations of the present invention according to the above description. Therefore, certain details in the embodiments should not constitute a limitation of the present invention, and the scope of protection of the present invention shall be defined by the scope of the attached claims.

Claims

1. A truss arch bridge hanger safety monitoring method based on image recognition technology, characterized in that: The steps include: Providing a marking piece, and attaching the marking piece to the suspension rod; Providing a binocular camera, and photographing the marking member on the boom by the binocular camera to obtain video data; Extracting corresponding image data from the video data, and identifying the marker in the image data based on image recognition technology to obtain the horizontal displacement of the marker; Obtaining a time course curve of the marker according to the horizontal displacement of the marker and the corresponding time information; Calculating the fundamental frequency of the boom using the time history curve of the marker; as well as The boom force is calculated according to the following formula, and the boom safety monitoring is realized by the boom force. T =4 WS 2 F 2 / 1000In the above formula, T represents the boom force, W represents the linear density of the boom, S Indicates the length of the boom, F represents the fundamental frequency of the boom; The step of calculating the fundamental frequency of the boom using the time history curve of the marker comprises: Taking a second derivative of the time history curve to obtain the vibration acceleration of the suspension rod; Performing Fourier transformation on the vibration acceleration of the suspension rod to obtain the fundamental frequency of the suspension rod; When obtaining the video data of the marking part through the binocular camera, the marking part is continuously photographed for a certain period of time at intervals to obtain the video data; The provided marking piece includes a pair of transverse marking strips and a pair of longitudinal marking strips, wherein the pair of transverse marking strips and the pair of longitudinal marking strips are staggered and connected to form a tic-tac-toe structure; When obtaining the time course curve of the marker, the horizontal displacement and the vertical displacement of the tic-tac-toe structure in the image data are measured, and then the time course curve of the marker is obtained in combination with the time information; When setting the binocular camera, a corresponding number of booms are selected as a measurement group according to the resolution of the binocular camera; Measure the distance between two adjacent booms in a measurement group, and take the average value of the measured distances as the installation distance; The binocular camera is installed on the arch bridge along the midline position of a measurement group, and the distance between the binocular camera and the center point of the corresponding measurement group is equal to the installation distance, and the binocular camera is used to photograph the suspension rods in the measurement group respectively; When measuring horizontal displacement and vertical displacement, the corresponding horizontal displacement and vertical displacement are obtained by measuring the displacement of the pixel points; A center mark point is set at the center of the tic-tac-toe structure, and the tic-tac-toe structure is calibrated using the center mark point. When shooting a video, the binocular camera can be aligned with the center mark point for shooting. When performing image recognition, the center mark point is used for image alignment to measure the vertical displacement of a pair of horizontal mark bars and the horizontal displacement of a pair of vertical mark bars.

2. A truss arch bridge suspender safety monitoring system based on image recognition technology, characterized in that: include: A marking piece attached to the boom; A binocular camera provided on the arch bridge and corresponding to the marking piece on the suspension rod, wherein the binocular camera is used to photograph the marking piece to form video data; The processing unit connected to the binocular camera is used to extract corresponding image data from the video data, identify the marker in the image data based on the image recognition technology to obtain the horizontal displacement of the marker, and then calculate the time history curve through the horizontal displacement, and calculate the fundamental frequency of the boom using the time history curve, so as to calculate the boom force through the following formula, and realize the safety monitoring of the boom through the boom force, T =4 WS 2 F 2 / 1000 In the above formula, T represents the boom force, W represents the linear density of the boom, S Indicates the length of the boom, F represents the fundamental frequency of the boom; The processing unit is used to perform a second derivative of the time history curve to obtain the vibration acceleration of the suspension rod, and is also used to perform Fourier transform on the vibration acceleration of the suspension rod to obtain the fundamental frequency of the suspension rod; The binocular camera continuously shoots the marking element at intervals for a certain period of time, thereby forming video data; The marking member comprises a pair of transverse marking strips and a pair of longitudinal marking strips, wherein the pair of transverse marking strips and the pair of longitudinal marking strips are staggered and connected to form a tic-tac-toe structure; The processing unit is also used to measure the horizontal displacement and vertical displacement of the tic-tac-toe structure in the image data, and then obtain the time course curve of the marking element in combination with the time information; The binocular camera is arranged in front of a plurality of booms to photograph each boom respectively; When setting the binocular camera, a corresponding number of booms are selected as a measurement group according to the resolution of the binocular camera; Measure the distance between two adjacent booms in a measurement group, and take the average value of the measured distances as the installation distance; The binocular camera is installed on the arch bridge along the midline position of a measurement group, and the distance between the binocular camera and the center point of the corresponding measurement group is equal to the installation distance, and the binocular camera is used to photograph the suspension rods in the measurement group respectively; When measuring horizontal displacement and vertical displacement, the corresponding horizontal displacement and vertical displacement are obtained by measuring the displacement of the pixel points; A center mark point is set at the center of the tic-tac-toe structure, and the tic-tac-toe structure is calibrated using the center mark point. When shooting a video, the binocular camera can be aligned with the center mark point for shooting. When performing image recognition, the center mark point is used for image alignment to measure the vertical displacement of a pair of horizontal mark bars and the horizontal displacement of a pair of vertical mark bars.

Images