CN115078382A - Bridge crack monitoring system based on video image - Google Patents

Abstract

The invention discloses a bridge crack monitoring system based on video images, which comprises a front-end system and a back-end system, wherein the front-end system comprises: the system comprises a camera, a micro-control system, a power supply module and a wireless transmission module, wherein a front-end system is integrally installed on the surface of a measured structure, and a rear-end system comprises: the crack precision calculation system comprises a data storage module and a data display module, wherein the back-end system is used for storing and displaying crack information sent by the front-end system, data are stored in a cloud server and can be displayed on a PC (personal computer) end or a mobile phone app, the front-end system and the back-end system are used for accurately calculating cracks, and the accurate calculation of the cracks is composed of five parts, namely image acquisition, image preprocessing, image segmentation, pixel calibration and width measurement. The device is wireless, low in power consumption and convenient to install, does not damage a structure, is internally provided with a crack edge algorithm, and can realize full-automatic high-precision measurement, an automatic dormancy and awakening function and a crack triggering alarm function.

Description

Bridge Crack Monitoring System Based on Video Image

technical field

The invention relates to the technical field of bridge health monitoring, in particular to a bridge crack monitoring system based on video images.

Background technique

In recent years, my country's transportation industry has developed rapidly, and many bridges have been built all over the country. However, during the construction and use of bridges, due to the cracks in the concrete structure, the bridges have hidden safety hazards and even collapse accidents frequently occur. One of the most common diseases in China, the appearance of cracks not only damages the appearance of the bridge, reduces the stress area of the section, but also affects the anti-penetration performance of the structure, causing the infiltration of moisture and harmful substances, inducing corrosion of steel bars or accelerating the natural environment of concrete. aging, thereby impairing the bearing capacity of the bridge structure and adversely affecting the safety performance of the bridge. Therefore, it is necessary to focus on detection and tracking monitoring of bridge cracks.

The traditional concrete crack detection method is manual detection, which mainly means that the inspectors directly observe the surface cracks with the naked eye or use auxiliary equipment such as telescopes, bridge inspection vehicles, and climbing vehicles to observe the surface cracks of the structure. Although the manual detection method is convenient and flexible to operate, this This method also has problems such as difficulty in detecting high-altitude components, easy fatigue of inspectors, high subjectivity, and serious missed detection and false detection. The traditional concrete crack monitoring technology mainly relies on vibrating wire crack meter or fiber grating crack meter. The sensor is installed on the surface of the concrete structure. When the structure is deformed, the change of the crack is measured by capturing the output signal of the sensor. The disadvantage of the vibrating wire crack meter for measuring cracks is that the sensing range of the sensor is limited, only 10-20cm. Therefore, it is generally installed in The crack position is known, and the time and position of the new crack cannot be sensed; the crack meter is affected by the breaking strain of the steel wire and the optical fiber, and the range is limited.

Fracture will occur after mm, and it is difficult to track the whole process of crack development. The disadvantage of fiber grating crack meter in measuring cracks is that its sensitivity is low, and it is difficult to effectively separate and extract crack information in complex environments such as severe temperature changes and frequent vibrations. As a result, its application in practical engineering is less and less.

In addition, with the advancement of science and technology, some new bridge crack monitoring technologies have been born, such as conductive paint crack monitoring, which forms a conductive film by applying flexible conductive paint on the surface of the structure. When the conductive film is stretched, The contact surface between the conductive particles decreases, and the resistance changes accordingly, and then the monitoring of structural cracks is realized by measuring the resistance change; the crack monitoring of the smart net, which uses enameled copper wire as the sensing and communication material, is used in the structure. Grid coordinates are formed on the surface. When cracks appear on the structure, the development of cracks can be captured by monitoring the fracture of enameled wires in the area; the disadvantages of these two methods are that they will cause structural damage, high initial investment cost, and construction difficulty. , and the input front-end sensors cannot be reused. The bridge crack monitoring method based on machine vision has become a research hotspot in recent years. It captures the image of the measured object and identifies the crack length and width by a computer. This method is easy to install and does not damage the structure. However, the current method of monitoring cracks based on video images is still semi-automatic measurement, which requires manual auxiliary measurement, and cannot realize fully automatic monitoring. Moreover, the accuracy of image recognition cracks is not high, requiring a large number of algorithms, which affects the measurement. efficiency.

Therefore, in view of the technical problem of real-time visual monitoring of cracks in structural characterization and disease, this project plans to develop a type of visual monitoring system for structural cracks. The optical imaging system and photoelectric sensing system are used to record the image of the widest position of the crack regularly, and then use a special The software processes the digitized images, so that useful information of the digitized images of cracks can be obtained, and the width of the measured cracks can be calculated by using algorithms.

SUMMARY OF THE INVENTION

Based on the technical problems existing in the background art, the present invention proposes a bridge crack monitoring system based on video images.

The video image-based bridge crack monitoring system proposed by the present invention includes a front-end system and a back-end system. The front-end system is composed of a camera, a micro-control system, a power supply module and a wireless transmission module. The front-end system is integrated and installed on the structure under test. On the surface, the back-end system is composed of a data storage module and a data display module. The back-end system is used to store and display the crack information sent by the front-end system. The data is stored on the cloud server and can be stored on the PC or The data is displayed on the mobile app. The front-end system and the back-end system are used for accurate calculation of cracks. The accurate calculation of cracks consists of five parts: image acquisition, image preprocessing, image segmentation, pixel calibration and width measurement.

Preferably, the camera is a low-power-consumption camera module, and the camera is assisted with a photographic protective cover and a fill light.

Preferably, the micro-control system is used to periodically issue photographing instructions, process and analyze the acquired images, and send out the crack information obtained after processing.

Preferably, the power supply module is powered by a lithium battery.

Preferably, the wireless transmission module adopts an NB-IoT communication module.

Preferably, the image acquisition: the lighting is supplemented with light, and the high-definition camera lens module directly captures the digital image;

Image preprocessing: process the image at the lowest level, and the processed image is a high-brightness image. The content of preprocessing mainly includes processes such as image grayscale, image enhancement, median filtering, and image binarization. Preprocessing can reduce The information content of the original image, because the general image has redundant information, the preprocessing can also suppress the sudden deformation of the image or enhance some structural features, etc., so as to achieve the purpose of improving the image quality;

Crack edge extraction: The LoG algorithm is used to extract the crack edge in the box girder and the crack at the internal cross section of the tunnel structure. The algorithm can make the extracted crack edge more accurate. The LoG algorithm is jointly proposed by Marr and Hildreth, which is The algorithm formed by combining Gaussian filtering and Laplacian edge detection is called Laplacian Gaussian algorithm. The basic principle of LoG algorithm is: first, use Gaussian-shaped two-dimensional low-pass filtering within a certain range. The filter performs smooth filtering on the image, and then uses the Laplace difference operator to detect the edge of the image;

Pixel calibration: When digital image processing technology is applied to the measurement of crack width, the pixels of the camera must be calibrated, that is, the actual width represented by each pixel in the image is calibrated. The unit of calibration index is mm/pix. In the calibration method, considering the requirements of the measurement accuracy and measurement efficiency of the system, the fixed macro method is used to achieve pixel calibration. parameters, and then calculate the size scale that a single pixel of the image can represent;

Calculation of crack width: Distinguish the upper and lower edges of cracks based on the results of crack boundary recognition, select each point on the upper edge, and use the "minimum distance method" to calculate the width of the target crack.

Preferably, the working process of the front-end system and the back-end system is as follows:

S1 fixes the sensor above the structure under test and points the camera at the target crack;

In S2, the front-end micro-control system sends out photographing instructions at regular intervals, and the camera continuously photographs the structure;

S3, the micro-control system performs image processing to identify the crack width, and sends the crack width and crack picture to the remote server through the wireless transmission module;

After the S4 finishes capturing and transmitting the crack image, the system will automatically enter the sleep state, and will automatically wake up to repeat the work when the next set shooting time arrives;

The S5 finally displays data on the PC or mobile app.

In the present invention, the beneficial effects of the video image-based bridge crack monitoring system are as follows:

1. The traditional concrete crack monitoring technology cannot take into account the crack monitoring accuracy and measurement range at the same time. The present invention uses the Sobel algorithm to remove shadows, contour extraction, adaptive dynamic filtering to suppress noise, ROI selection to lock the crack area, and Yolo network to quickly identify cracks and frame differences. Supported by cutting-edge algorithms such as monitoring crack changes, it can support the simultaneous identification of more than 30 cracks, and the measurement accuracy can reach 0.02mm, which greatly increases the monitoring efficiency of cracks.

2. The previous image recognition crack device needs to transmit the crack image to the back end, and then the image processing server performs the calculation and recognition. The invention detects whether the abnormality is detected by the algorithm deployed at the edge terminal, and the crack condition can be monitored at the terminal. Only when the abnormality occurs, the situation and the picture are reported to the server, and the detailed diagnosis is performed by the powerful computing power of the cloud or manually, which saves the transmission of pictures. The huge traffic also reduces the need for image processing servers.

3. According to taking pictures twice a day and running the machine for 1 minute each time, the average current is about 200uA. When using a large-capacity battery (20000mAH), it can provide 10 years of use, and it is easy to replace, which meets the long-term health monitoring of bridges. Crack observation requirements.

4. By adding a power supply module and a wireless transmission module, the function of long-term online monitoring is realized; at the same time, through the development of the front-end micro-control system, the relevant algorithms based on crack image recognition can be integrated into the system, and the crack width can be calculated by edge computing. The result data is directly sent back to the back end to realize real-time observation of crack change trend and abnormal alarm.

In the invention, the auxiliary support is installed on the surface of the structure, the miniature camera is fixed on the support, and the MCU (micro-processing unit) periodically sends out shooting instructions to shoot the target crack, and then analyzes the crack width through a preprocessing algorithm. The crack width and photos are sent to the cloud platform. This crack monitoring method can achieve: wireless equipment, low power consumption, easy installation, no damage to structures, built-in crack edge algorithm, automatic high-precision measurement, automatic sleep and wake-up functions and Cracks trigger alarm function.

Description of drawings

Fig. 1 is the composition diagram of the bridge crack monitoring system based on video image proposed by the present invention;

2 is a schematic diagram of accurate calculation of cracks in the video image-based bridge crack monitoring system proposed by the present invention;

Fig. 3 is the minimum distance method crack width calculation schematic diagram of the video image-based bridge crack monitoring system proposed by the present invention;

FIG. 4 is a software diagram of the back-end system of the video image-based bridge crack monitoring system proposed by the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments.

Referring to Figures 1-4, the video image-based bridge crack monitoring system includes a front-end system and a back-end system. The front-end system consists of a camera, a micro-control system, a power supply module and a wireless transmission module. The front-end system is integrated and installed in the tested On the surface of the structure, the back-end system consists of a data storage module and a data display module. The back-end system is used to store and display the crack information sent by the front-end system. The data is stored on the cloud server and can be stored on a PC. The data is displayed on the terminal or on the mobile phone app. The front-end system and the back-end system are used for accurate calculation of cracks. The accurate calculation of cracks consists of five parts: image acquisition, image preprocessing, image segmentation, pixel calibration and width measurement.

In the present invention, the camera is a low-power-consumption camera module, and the camera is assisted with a photographic protective cover and a fill light.

In the present invention, the micro-control system is used for regularly issuing photographing instructions, processing and analyzing the acquired images, and sending out the crack information obtained after processing.

In the present invention, the power supply module is powered by a lithium battery.

In the present invention, the wireless transmission module adopts an NB-IoT communication module.

In the present invention, the image acquisition: the light is supplemented by the lighting equipment, and the high-definition camera lens module directly captures the digital image;

Image preprocessing: process the image at the lowest level, and the processed image is a high-brightness image. The content of preprocessing mainly includes processes such as image grayscale, image enhancement, median filtering, and image binarization. Preprocessing can reduce The information content of the original image, because the general image has redundant information, the preprocessing can also achieve the purpose of improving the image quality by suppressing the sudden deformation of the image or enhancing some structural features;

Crack edge extraction: The LoG algorithm is used to extract the crack edge in the box girder and the crack at the internal cross section of the tunnel structure. The algorithm can make the extracted crack edge more accurate. The LoG algorithm is jointly proposed by Marr and Hildreth, which is The algorithm formed by combining Gaussian filtering and Laplacian edge detection is called Laplacian Gaussian algorithm. The basic principle of LoG algorithm is: first, use Gaussian-shaped two-dimensional low-pass filtering within a certain range. The filter performs smooth filtering on the image, and then uses the Laplace difference operator to detect the edge of the image;

Pixel calibration: When digital image processing technology is applied to the measurement of crack width, the pixels of the camera must be calibrated, that is, the actual width represented by each pixel in the image is calibrated. The unit of calibration index is mm/pix. In the calibration method, considering the requirements of the measurement accuracy and measurement efficiency of the system, the fixed macro method is used to achieve pixel calibration. parameters, and then calculate the size scale that a single pixel of the image can represent;

Calculation of crack width: Distinguish the upper and lower edges of cracks based on the results of crack boundary recognition, select each point on the upper edge, and use the "minimum distance method" to calculate the width of the target crack.

In the present invention, the working process of the front-end system and the back-end system is as follows:

S1 fixes the sensor above the structure under test and points the camera at the target crack;

In S2, the front-end micro-control system sends out photographing instructions at regular intervals, and the camera continuously photographs the structure;

S3, the micro-control system performs image processing to identify the crack width, and sends the crack width and crack picture to the remote server through the wireless transmission module;

After the S4 finishes capturing and transmitting the crack image, the system will automatically enter the sleep state, and will automatically wake up to repeat the work when the next set shooting time arrives;

The S5 finally displays data on the PC or mobile app.

In the present invention, the sensor is fixed above the measured structure, and the camera is aimed at the target crack; the front-end micro-control system sends out photographing instructions at regular intervals, and the camera continuously shoots the structure; and then the micro-control system performs image processing to identify cracks The crack width and crack picture are sent to the remote server through the wireless transmission module; after the crack image is captured and transmitted, the system will automatically enter the sleep state, and will automatically wake up when the next set photo time arrives to repeat the work; finally Display data on PC or mobile app.

The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited to this. The equivalent replacement or change of the inventive concept thereof shall be included within the protection scope of the present invention.

Claims (7)

1. based on the bridge crack monitoring system of video image, it is characterized in that, comprise front-end system and back-end system, described front-end system is made up of: camera, micro-control system, power supply module and wireless transmission module, front-end system is integrated and installed in the tested On the surface of the structure, the back-end system consists of a data storage module and a data display module. The back-end system is used to store and display the crack information sent by the front-end system. The data is stored on the cloud server and can be stored on a PC. The data is displayed on the terminal or on the mobile phone app. The front-end system and the back-end system are used for accurate calculation of cracks. The accurate calculation of cracks consists of five parts: image acquisition, image preprocessing, image segmentation, pixel calibration and width measurement. 2 . The bridge crack monitoring system based on video images according to claim 1 , wherein the camera is a low-power-consumption camera module, and the camera is assisted with a camera protection cover and a fill light. 3 . 3. the bridge crack monitoring system based on video image according to claim 1, is characterized in that, described micro-control system is used for regularly sending out photographing instruction, the image acquired is processed and analyzed, and the crack obtained after processing is information is sent. 4 . The bridge crack monitoring system based on video images according to claim 1 , wherein the power supply module is powered by a lithium battery. 5 . 5. The bridge crack monitoring system based on video image according to claim 1, is characterized in that, described wireless transmission module adopts NB-IoT communication module. 6. The bridge crack monitoring system based on video image according to claim 1, is characterized in that, described image acquisition: fill light by illuminating equipment, high-definition camera lens module directly shoots into digital image; Image preprocessing: process the image at the lowest level, and the processed image is a high-brightness image. The content of preprocessing mainly includes processes such as image grayscale, image enhancement, median filtering, and image binarization. Preprocessing can reduce The information content of the original image, because the general image has redundant information, the preprocessing can also achieve the purpose of improving the image quality by suppressing the sudden deformation of the image or enhancing some structural features; Crack edge extraction: The LoG algorithm is used to extract the crack edge in the box girder and the crack at the internal cross section of the tunnel structure. The algorithm can make the extracted crack edge more accurate. The LoG algorithm is jointly proposed by Marr and Hildreth, which is The algorithm formed by combining Gaussian filtering and Laplacian edge detection is called Laplacian Gaussian algorithm. The basic principle of LoG algorithm is: first, use Gaussian-shaped two-dimensional low-pass filtering within a certain range. The filter performs smooth filtering on the image, and then uses the Laplace difference operator to detect the edge of the image; Pixel calibration: When digital image processing technology is applied to the measurement of crack width, the pixels of the camera must be calibrated, that is, the actual width represented by each pixel in the image is calibrated. The unit of calibration index is mm/pix. In the calibration method, considering the requirements of the measurement accuracy and measurement efficiency of the system, the fixed macro method is used to achieve pixel calibration. parameters, and then calculate the size scale that a single pixel of the image can represent; Calculation of crack width: Distinguish the upper and lower edges of cracks based on the results of crack boundary recognition, select each point on the upper edge, and use the "minimum distance method" to calculate the width of the target crack. 7. the bridge crack monitoring system based on video image according to claim 1, is characterized in that, the process flow that described front-end system and back-end system work is as follows: S1 fixes the sensor above the structure under test and points the camera at the target crack; In S2, the front-end micro-control system sends out photographing instructions at regular intervals, and the camera continuously photographs the structure; S3, the micro-control system performs image processing to identify the crack width, and sends the crack width and crack picture to the remote server through the wireless transmission module; After the S4 finishes capturing and transmitting the crack image, the system will automatically enter the sleep state, and will automatically wake up to repeat the work when the next set shooting time arrives; The S5 finally displays data on the PC or mobile app.

Images