Mobile tripod type engineering structure displacement detection device and method

Abstract

The embodiments of the present application relate to the technical field of engineering structure displacement detection; in particular to a mobile tripod engineering structure displacement detection device and method. The device comprises a mobile tripod support unit, a holder, a visual core device, an attitude sensor, a processing unit and an output unit; the visual core device collects reference images and real-time images containing multiple measurement point targets, the processing unit performs digital image correlation matching on the image regions corresponding to the multiple measurement point targets, corrects the installation pose deviation according to the attitude information, calculates the horizontal displacement and vertical displacement of the engineering structure based on the corrected displacement change, and outputs the detection result; the method comprises mobile deployment, field of view adjustment, image acquisition, correlation matching, attitude correction, displacement calculation and result output. The device and method can realize multi-point synchronous displacement detection, reduce the requirement for field leveling, and improve the dynamic detection efficiency and measurement stability on the engineering site.

Description

Technical Field

[0001] The embodiments of this application relate to the field of displacement detection technology for engineering structures; in particular, a mobile tripod-type displacement detection device and method for engineering structures. Background Technology

[0002] In existing displacement detection of engineering structures, total station electronic tachometers and other measuring equipment are commonly used to collect coordinates and monitor deformation of objects such as bridges, tunnels, slopes, and buildings. This type of approach typically processes data by plotting, measuring, and uploading data point by point, which takes a long time to complete a single measurement at a single point. This makes it difficult to meet the requirements of simultaneous acquisition of multiple measuring points and high sampling frequency in dynamic detection scenarios.

[0003] Furthermore, existing total station measurement solutions are quite sensitive to installation status and environmental conditions. The equipment usually requires leveling and calibration during installation, and after the equipment is moved, it is often necessary to re-plot points and reset the reference. At the same time, environmental factors such as changes in external temperature, rain, fog, and direct sunlight can easily affect the stability and accuracy of measurements, resulting in weak on-site rapid deployment capabilities and insufficient convenience for local display and remote output of test results. Summary of the Invention

[0004] The embodiments of this application aim to provide a mobile tripod-type displacement detection device and method for engineering structures, so as to at least alleviate the problems of low point-by-point measurement frequency, cumbersome leveling and calibration after mobile deployment, and insufficient detection stability in complex environments in the prior art.

[0005] According to a first aspect of this application, a mobile tripod-type displacement detection device for engineering structures is provided for displacement detection of engineering structures equipped with multiple measuring point targets, including: Mobile tripod support unit; The gimbal is mounted on the mobile tripod support unit; The core vision device is fixed on the gimbal and acquires reference images and real-time images of multiple measurement point targets within the same field of view; An attitude sensor is installed in the vision core device to collect the attitude information of the vision core device; The processing unit is connected to the vision core device and the attitude sensor. It performs digital image correlation matching on the image regions of the multiple measurement point targets in the reference image and the real-time image to synchronously obtain the displacement changes of the multiple measurement point targets in the image. It also corrects the installation posture deviation of the vision core device according to the attitude information and calculates the horizontal and vertical displacements of the engineering structure based on the corrected displacement changes of the multiple measurement point targets in the image. The output unit, connected to the processing unit, is used to output the detection results of the horizontal and vertical displacements.

[0006] The above technical solution enables simultaneous imaging and digital image correlation processing of multiple measurement point targets within the same field of view. Compared with point-by-point measurement, it can improve the efficiency of dynamic displacement detection of multiple measurement points. At the same time, by combining the attitude information obtained by the attitude sensor to correct the installation posture deviation of the core vision equipment, the reliance on manual leveling and calibration can be reduced after the device is moved and deployed, thereby improving the rapid deployment capability on the engineering site.

[0007] In some implementations, the processing unit is connected to the gimbal control unit. The processing unit controls the gimbal rotation based on the distribution information of the multiple measurement point targets, so that the multiple measurement point targets enter the field of view of the vision core device, and initiates displacement detection after positioning is completed. This reduces manual point-by-point searching and improves detection preparation efficiency.

[0008] In some implementations, the processing unit performs sub-pixel localization of the positions of the multiple measurement point targets obtained through digital image correlation matching, in order to further improve the accuracy of displacement calculation.

[0009] In some implementations, the vision core device further includes a temperature control unit for regulating the internal temperature of the vision core device. The vision core device may also include an infrared laser emitting unit and an image enhancement processing module, which performs defogging and anti-exposure processing on the real-time image. This improves imaging and detection stability in complex environments.

[0010] In some implementations, the output unit includes a local display screen and a remote terminal interface that is communicatively connected to the processing unit, for on-site display and remote output, respectively.

[0011] According to a second aspect of this application, a method for displacement detection of engineering structures using a mobile tripod is provided, for detecting the displacement of engineering structures equipped with multiple measuring point targets, including: A mobile tripod support unit is set up, and the core vision device is installed by a gimbal set on the mobile tripod support unit. Adjust the observation direction of the vision core device so that the multiple measurement point targets enter the field of view of the vision core device; Acquire reference images and real-time images of targets containing multiple measurement points within the same field of view; Digital image correlation matching is performed on the image regions of the multiple measurement point targets in the reference image and the real-time image to synchronously obtain the displacement changes of the multiple measurement point targets in the image; The attitude information of the vision core device is collected, and the installation posture deviation of the vision core device is corrected based on the attitude information. The horizontal and vertical displacements of the engineering structure are calculated based on the displacement changes of the multiple measured target points in the image after correction. Output the detection results of the horizontal and vertical displacements.

[0012] Using the above method, after the device is moved and deployed, digital image correlation matching can be used to solve the synchronous displacement of multiple measurement point targets, and the influence of the device installation posture change can be compensated by attitude information, thereby realizing automated displacement detection for engineering structures. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the overall structure of the mobile tripod-type displacement detection device for engineering structures in the embodiments of this application. Figure 2 This is a schematic diagram illustrating the on-site setup and displacement detection principle of the mobile tripod-type engineering structure displacement detection device in this application embodiment; Figure 3 This is a functional block diagram of the mobile tripod-type displacement detection device for engineering structures in the embodiments of this application; Figure 4 This is a flowchart illustrating the displacement detection method for a mobile tripod-type engineering structure according to an embodiment of this application.

[0014] The system includes: 1. Vision core equipment; 2. Gimbal; 3. Mobile tripod support unit; 4. Attitude sensor; 5. Processing unit; 6. Output unit; 7. Measurement point target; 8. The engineering structure under test; 9. Field of view; 10. Temperature control unit; 11. Infrared laser emitting unit; 12. Image enhancement processing module; 13. Image acquisition module; 14. Digital image correlation matching module; 15. Attitude correction module; and 16. Displacement calculation module. Detailed Implementation

[0015] The technical solution of this application will be further described below with reference to the accompanying drawings and embodiments. It should be noted that the following embodiments are used to illustrate the technical concept and specific implementation of this application; for structural forms, connection methods, and control parameters that are not explicitly defined, equivalent settings can be made according to the detection requirements of the engineering site. The digital image correlation processing in this application represents the technical process of using correlation matching between the grayscale distribution, texture distribution, or target pattern distribution of the reference image and the real-time image to determine the positional change of the measurement point in the image plane. The attitude correction in this application represents the correction of the additional image offset caused by changes in the installation pose of the core vision device using pitch, roll, and azimuth information collected by the attitude sensor, so that the final output displacement result more realistically reflects the deformation of the measured engineering structure itself.

[0016] The first embodiment of this application provides a mobile tripod-type displacement detection device for engineering structures. For example... Figures 1 to 3 As shown, the device includes a mobile tripod support unit 3, a gimbal 2, a vision core device 1, an attitude sensor 4, a processing unit 5, and an output unit 6. Figure 1 The overall structure of the device is shown. Figure 2 The diagram shows the on-site layout of the device relative to the engineering structure 8 being measured, and the field of view 9 formed by the core vision device 1. Figure 3 The diagram illustrates the functional modules within the device used for image acquisition, digital image correlation matching, attitude correction, and displacement calculation. The mobile tripod support unit 3 supports the entire detection device and facilitates its rapid movement and deployment at engineering sites such as bridges, tunnels, slopes, and buildings. In this embodiment, the mobile tripod support unit 3 addresses the issue of rapid positioning of the equipment after it reaches the detection point. After the operator moves the device to a suitable observation position relative to the object being measured, a unified observation field of view capable of simultaneously covering multiple measurement point targets 7 is established through the pan-tilt unit 2 and the core vision device 1. This reduces the need for repeated point searching, plotting, and switching measurement states around a single measurement point, creating a prerequisite for subsequent multi-point synchronous detection.

[0017] exist Figure 3 In the functional module relationship shown, the image acquisition module 13 in the vision core device 1 is used to acquire reference images and real-time images; the digital image correlation matching module 14 in the processing unit 5 is used to perform correlation matching on the image regions corresponding to multiple measurement point targets 7; the attitude correction module 15 is used to correct the installation posture deviation of the vision core device 1 according to the attitude information output by the attitude sensor 4; the displacement calculation module 16 is used to calculate the horizontal and vertical displacements of the engineering structure according to the image displacement changes of multiple measurement point targets 7 after correction; and the output unit 6 is used to display the displacement calculation results on site or output them remotely.

[0018] like Figure 2As shown, multiple measurement point targets 7 are set on the engineering structure 8 under test. After adjustment by the gimbal 2, the vision core device 1 forms a field of view 9 covering multiple measurement point targets 7. In a specific example of bridge deflection detection, multiple measurement point targets 7 can be set up at the bottom of the main beam, the surface of the web, or along the mid-span, quarter-span, and adjacent areas of the supports at the edge of the bridge deck. The measurement point targets 7 can be artificial targets with clear boundaries and stable pattern features, such as high-contrast circular targets, cross targets, or attached marks with characteristic textures, so that the vision core device 1 can still stably identify them at a long distance. After the operator moves the device to the location of the access road under the bridge or the working platform on the side of the bridge, the initial positioning is completed by moving the tripod support unit 3, and then the horizontal rotation angle and pitch angle of the vision core device 1 are adjusted by the gimbal 2 so that multiple measurement point targets 7 in the mid-span area and adjacent areas simultaneously enter the field of view. Compared to the traditional point-by-point measurement method, this embodiment adopts a method of simultaneous imaging of multiple measurement points within the same field of view, thereby providing a foundation for subsequent synchronous displacement calculation based on images.

[0019] like Figure 3 As shown, the image acquisition module 13 can be integrated into the vision core device 1 to acquire reference images and real-time images. The vision core device 1 is used to acquire reference images and real-time images containing multiple measurement point targets 7. The reference image is usually acquired before the detection begins, when the object under test is in its initial stress state or when a reference state is selected; the real-time image is the current image continuously acquired during the process of the object under test being subjected to vehicle loads, construction disturbances, environmental loads, or other external actions. In the bridge scenario, the reference image can be acquired when there is no concentrated vehicle traffic on the bridge surface or when it is under a predetermined reference load state; in the tunnel scenario, the reference image can be acquired after an excavation cycle ends and initial support is completed; in the slope scenario, the reference image can be acquired at the beginning of the monitoring cycle. By first establishing a reference image and then continuously acquiring real-time images, the processing unit 5 can compare the image information at each subsequent moment with the reference state, so that the displacement measurement has a clear reference.

[0020] like Figure 3As shown, the processing unit 5 may include a digital image correlation matching module 14, an attitude correction module 15, and a displacement calculation module 16. The digital image correlation matching module 14 performs correlation matching on the image regions of multiple measurement point targets 7. The attitude correction module 15 corrects installation pose deviations based on attitude information. The displacement calculation module 16 converts the corrected image displacement changes into the actual displacement of the engineering structure. In this embodiment, the processing unit 5 performs digital image correlation matching on the image regions of multiple measurement point targets 7 in a reference image and a real-time image. Specifically, firstly, the corresponding target image region is extracted around each measurement point target 7 in the reference image to form a reference sub-region used for subsequent matching. Then, a search region is established in the real-time image for the possible displacement range of each reference sub-region. Next, the correlation degree is calculated within the search region based on grayscale distribution similarity, texture similarity, or target contour similarity to determine the position that best matches the reference sub-region. Finally, the change in the best-matched position relative to the reference position is used as the displacement change of the measurement point target 7 in the image plane. Since multiple measurement point targets 7 are located in the same field of view, the processing unit 5 can perform the above processing in parallel for multiple reference sub-regions, thereby obtaining the displacement change results of multiple measurement points at the same time. Therefore, the digital image correlation processing adopted in this embodiment uses a method of synchronously solving the displacement of multiple measurement points based on the same image or at the same time, which can fundamentally alleviate the frequency limitation problem caused by point-by-point measurement.

[0021] like Figure 3As shown, the attitude sensor 4 is connected to the attitude correction module 15, and the processing unit 5 performs installation posture deviation correction based on the attitude information output by the attitude sensor 4. To ensure that the digital image correlation matching results accurately reflect the deformation of the engineering structure, this embodiment includes an attitude sensor 4 in the vision core device 1. The attitude sensor 4 can collect real-time information on the pitch, roll, and azimuth changes of the vision core device 1 during the detection process. At the engineering site, although the mobile tripod support unit 3 can be quickly deployed, it is inevitably affected by factors such as uneven ground, changes in tripod force, wind load disturbance, and personnel operation, causing slight changes in the vision core device 1 relative to the initial reference attitude. After attitude compensation, the change in the position of the measurement point in the real-time image mainly reflects the imaging change caused by the actual displacement of the measured object, suppressing the additional image shift caused by the device's own attitude change, thereby improving the accuracy of the displacement results. In this embodiment, attitude information is collected by the attitude sensor 4, and the processing unit 5 establishes a correction relationship between the attitude information and the image shift. First, the overall image drift caused by the installation posture deviation is corrected, and then the corrected image data is used for displacement calculation. In this way, even if the device is moved and deployed between different detection points, or if the device posture drifts slightly during the same detection process, the reliance on manual precision leveling can still be reduced, and the output results can more stably reflect the deformation of the engineering structure itself.

[0022] After obtaining the image displacement changes of multiple measurement point targets 7 after correction, the processing unit 5 can further convert the displacement changes in the image plane into the actual displacement data of the measured engineering structure based on the preset imaging parameters, pixel scale conversion relationship, observation distance information, or target scale information of the device. In this application, the focus is on outputting the horizontal and vertical displacements of the engineering structure. For bridge deflection detection, vertical displacement usually corresponds to the deflection or rebound of the bridge under load; for slope surface displacement detection, horizontal displacement usually corresponds to the lateral displacement component of the slope along the slope surface or perpendicular to the slope surface; for building facade monitoring, vertical displacement can be used to characterize the settlement trend, and horizontal displacement can be used to characterize the lateral displacement trend of the wall. Through a unified image acquisition, correlation matching, attitude correction, and displacement conversion process, this embodiment enables the same device to be applicable to the displacement detection needs of multiple types of engineering structures.

[0023] Output unit 6 is connected to processing unit 5 and is used to output detection results. In a typical bridge dynamic monitoring example, output unit 6 can display the current displacement value, displacement change curve, timestamp, and alarm threshold status of multiple measurement point targets 7 on the built-in display screen of the vision core device 1 in real time. Simultaneously, it can also send data to the field control room, tablet terminal, or management platform via a remote terminal interface for simultaneous viewing by technicians. This technical solution addresses situations where on-site measurement and judgment are frequently required, such as observing whether local bridge deformation exceeds preset limits during construction hoisting, observing abnormal convergence changes near temporary supports during tunnel excavation, and quickly confirming whether monitoring points have experienced sudden displacement after a rainstorm on a slope. If detection results can only be exported afterward, it is difficult to support timely on-site decision-making; this embodiment combines local and remote output to enable on-site judgment and remote viewing of detection results.

[0024] The device in this embodiment solves the technical problems better than traditional solutions by forming a complete collaborative relationship between multiple components. The mobile tripod support unit 3 solves the problem of rapid positioning and deployment; the gimbal 2 solves the problem of multiple measurement points entering a unified field of view; the vision core device 1 solves the problem of synchronous acquisition; digital image correlation matching solves the problem of synchronous displacement calculation for multiple measurement points; the attitude sensor 4 and its calibration process solve the problem of installation posture deviation under mobile deployment conditions; and the output unit 6 solves the problem of immediate use of detection results. Because these links are interconnected, this embodiment can complete multi-point displacement detection on the engineering site with high efficiency, less prior preparation, and strong environmental adaptability, and output data results that can be directly used for acceptance testing, load testing, and long-term monitoring.

[0025] The second embodiment of this application further explains the principles of automatic point finding and automatic measurement initiation of the gimbal based on the first embodiment. In the first embodiment, multiple measurement point targets 7 are included in a unified field of view at the engineering site, which is a prerequisite for the subsequent synchronous detection of digital images. However, in practical applications, the distribution of measurement points for different engineering objects often varies greatly. For example, in a bridge scenario, measurement points may be scattered along the length of the beam; in a tunnel scenario, measurement points may be arranged along the lining contour or around the tunnel face; in a slope scenario, measurement points may be arranged in layers at the top, middle, and bottom of the slope. If the operator relies entirely on visual observation and manually adjusts the gimbal 2 repeatedly, the operation time will increase significantly, and inconsistencies in the field of view of different measurements are likely to occur during repeated operations. Therefore, this embodiment focuses on explaining the workflow of automatic point finding, automatic positioning, and automatic measurement initiation.

[0026] In a specific implementation process, the operator can first preview the target area at low magnification using the vision core device 1. The processing unit 5 identifies the approximate location of multiple measurement point targets 7 based on the target features, edge features, or preset positional relationships in the preview image. When the system determines that some targets are located at the edge of the current field of view or have not yet fully entered the field of view, the processing unit 5 outputs control commands to the gimbal 2, driving the gimbal 2 to make small angle adjustments in the horizontal and pitch directions until all targets to be measured enter the set effective field of view. If the system detects that the local target imaging size is too small, the edge sharpness is insufficient, or the distance between adjacent targets does not meet the preset processing requirements, it can further instruct the gimbal 2 and the vision core device 1 to perform fine-tuning, such as fine-tuning the lens focal length, changing the shooting magnification, or aligning with the center of the visual axis, in order to improve the stability of subsequent related matching.

[0027] Taking tunnel lining deformation monitoring as an example, construction sites are typically poorly lit and have limited space, making it unsuitable for prolonged manual point-finding operations once the equipment is in place. In this embodiment, after the operator places the device at a suitable distance from the lining wall, the system can perform a preliminary scan to identify multiple measurement point targets on the lining surface. When the top target, side wall target, and near-bottom target are all identified as observable, the system records the current gimbal attitude parameters as the observation attitude. If the top target deviates too close to the upper boundary of the field of view, the system automatically lowers the pitch angle and re-verifies whether all targets are still within the effective area. If the side wall target is too close to the edge of the image, the system automatically fine-tunes the horizontal rotation angle. Through this method of controlling the gimbal attitude based on imaging results, the device can form a stable observation image suitable for digital image correlation processing in a short time.

[0028] The technical problem solved by automatic measurement startup lies in the fact that the quality of digital image correlation matching is closely related to the quality of the reference image. If the reference image is directly acquired before the gimbal has stabilized, before the field of view covers all key targets, or when the image is still jittery, the subsequent detection accuracy will be affected. Therefore, in this embodiment, after completing automatic point finding, the processing unit 5 can perform a stability judgment on the current image, such as the target boundary change in several consecutive frames being less than a preset threshold, the attitude fluctuation output by the attitude sensor being within the allowable range, and the image brightness being within the effective range. When the above conditions are met, the system automatically acquires the reference image and switches to real-time detection mode. This completes the process of first automatically finding the point, then automatically confirming the imaging stability, and finally automatically starting the measurement, which can avoid the fluctuation of reference image quality caused by inconsistent manual operation rhythm, thereby improving the consistency of subsequent displacement calculation.

[0029] In bridge load testing scenarios, automated point location and automatic start-up technologies are applicable because bridge static or dynamic load tests typically have well-defined test windows, with key time points for vehicle arrival, loading start, and unloading end. If manual location of targets and decision-making for image acquisition are still required before the test begins, critical load phases may be missed. This embodiment allows operators to complete equipment setup before loading, and the system automatically incorporates all key targets into the field of view and establishes reference images. Real-time images are continuously acquired during vehicle entry and exit, thus fully recording displacement changes at locations such as mid-span and quarter-span points. Therefore, automatic point location by the pan-tilt unit is a crucial pre-test guarantee in the multi-point synchronous detection chain.

[0030] The third embodiment of this application further explains the sub-pixel localization process in digital image correlation processing, based on the first embodiment. Digital image correlation matching can first determine the optimal matching position of multiple measurement point targets in the reference image and the real-time image at the pixel level. However, in engineering displacement detection, especially in scenarios such as bridge deflection, initial changes in building settlement, or monitoring of small slope displacements, the actual physical displacement corresponding to a single frame image may be much smaller than the actual size represented by a single pixel. If only the pixel-level matching position is used as the final result, when the actual displacement is small, it is easy to exhibit jumps in measurement values, large quantization errors, or step-like changes in results across multiple frames. Therefore, this embodiment further performs sub-pixel localization based on pixel-level correlation matching to improve the resolution capability of displacement solving.

[0031] Specifically, the processing unit first determines the optimal pixel-level matching position of a target measurement point in the real-time image based on the correlation matching results of the digital image. Then, it constructs a smaller local fitting region around this position and continuously fits the correlation distribution, grayscale change trend, or target edge position to obtain a more refined displacement position between adjacent pixels. In this process, pixel-level matching gives the approximate pixel position of the displacement, while sub-pixel localization further gives the specific offset of the displacement within that pixel. Through this processing, even if the target measurement point 7 in the real-time image undergoes only a very small imaging displacement change, the processing unit can still obtain a smoother and more continuous displacement time series.

[0032] Taking the deflection change of a bridge during the slow entry of a vehicle as an example, before the vehicle reaches the mid-span, the beam displacement is often in a gradual accumulation stage, and the displacement increment between individual sampling moments may be very small. If only pixel-level matching is used, the detection result of the mid-span target may remain unchanged for several frames, and then suddenly jump to the displacement corresponding to a single pixel, which is not conducive to truly reflecting the gradual stress process of the beam. After introducing sub-pixel localization, the processing unit can resolve more subtle displacement increments between consecutive frames, making the mid-span displacement curve show a more continuous change over time that conforms to the actual stress state. This has a positive effect on assessing structural stiffness, observing the rebound characteristics during loading and unloading, and identifying abnormal vibration responses.

[0033] Taking building facade deformation monitoring as an example, in the early stages of long-term monitoring, the displacement changes of the measurement point targets 7 set near the edges of walls or window openings are usually small. Without sufficient subpixel resolution, early deformation trends may be masked by noise. By performing subpixel localization after digital image correlation matching, the system can become more sensitive to minute displacements, thereby detecting settlement differences, local tilting, or subtle deformation signs before crack development earlier. Therefore, subpixel localization represents a refinement step in digital image correlation processing, used to address the problem that pixel-level resolution is insufficient to support high-precision displacement monitoring.

[0034] Furthermore, there is a synergistic relationship between subpixel localization and the pattern design of the measurement point target. If the target boundary is clear, the texture is stable, and the contrast is high, the peak position of the correlation matching result is easier to determine accurately, and the subpixel fitting process is more stable. Conversely, if the target pattern is blurry, the boundary is severely reflective, or it is partially occluded, the fitting result may fluctuate more. Therefore, in this embodiment, it is preferable to make the measurement point target 7 have sufficient image features to work with digital image correlation and subpixel localization to improve detection accuracy. In this process, there is a sequential relationship between target design and algorithm processing: the more stable the front-end imaging quality, the more reliable the back-end subpixel solution, and the more realistically the output displacement result can characterize the deformation of the structure.

[0035] The fourth embodiment of this application further illustrates the imaging stability enhancement scheme under complex environments, based on the first embodiment. Displacement monitoring at engineering sites is typically conducted in complex environments. Bridge monitoring may face direct sunlight, backlighting, vehicle headlight interference, and temperature variations; tunnel monitoring may face moisture, dust, and localized fogging; and slope monitoring may face wind, rain, frost, fog, and significant diurnal temperature variations. These environmental factors directly affect the imaging quality of the core vision device. Since digital image correlation processing highly depends on the comparability between the reference image and the real-time image, without stabilization of the imaging link, even with a well-designed subsequent algorithm, excessive fluctuations in the quality of the original image may affect the accuracy of displacement measurement. This embodiment focuses on the roles of the temperature control unit 10, the infrared laser emission unit 11, and the image enhancement processing module 12 in this link.

[0036] The temperature control unit 10 is used to regulate the internal temperature of the vision core device 1. Vision measurement devices typically include image sensors, lens assemblies, processing circuits, and storage circuits. These components may experience thermal drift, changes in imaging parameters, or unstable operating states in environments with large temperature variations. For example, in a bridge nighttime monitoring scenario, the device may initially be in a low-temperature environment, then gradually heat up due to continuous operation; under intense sunlight during the day, the rising shell temperature may further conduct to the internal optical components. If the internal temperature fluctuates significantly, the imaging state at the time of reference image establishment may deviate from the imaging state during subsequent real-time image acquisition, affecting the stability of digital image correlation matching. By regulating the internal temperature through the temperature control unit 10, the vision core device 1 can maintain relatively stable imaging conditions over a longer monitoring period, thereby reducing the indirect impact of temperature changes on displacement measurement.

[0037] like Figure 3 As shown, the infrared laser emitting unit 11 and the image enhancement processing module 12 primarily address complex visual environment issues. In some long-distance bridge inspection scenarios, direct sunlight at low angles may cause overexposure in some target areas, resulting in local brightness saturation of the image, making it difficult to stably extract grayscale features from digital image correlation. In foggy environments such as tunnels or slopes, the overall image contrast may decrease, and the target edges may become blurred, affecting matching accuracy. To address these issues, the infrared laser emitting unit 11 can provide auxiliary illumination or auxiliary feature enhancement for the target area, while the image enhancement processing module 12 can perform fog-penetrating processing and anti-exposure processing on the real-time image. The core function of fog-penetrating processing is to improve the local contrast and edge sharpness of the image in foggy scenes, allowing the target contour weakened by fog to regain sufficient relevant matching features. The core function of anti-exposure processing is to suppress brightness saturation in strong light areas and enhance effective details in weak texture areas, thereby maintaining the consistency of features available for matching between the reference image and the real-time image.

[0038] Taking a post-rain slope inspection scenario as an example, after the monitoring personnel move the device to the observation point at the foot of the slope, they find that there is obvious water mist in the air. The contrast of the directly acquired real-time image is low, and the edges of some slope targets are not clear. Without processing, the peak value of digital image correlation matching will not be prominent enough, and the displacement results of multiple measuring points are prone to fluctuation. In this embodiment, the image enhancement processing module can first perform defogging processing on the real-time image to improve the distinguishability between the target and the background, and then input the enhanced image into the correlation matching process. At the same time, if the brightness of some targets is too high due to the influence of sunlight reflection, the overly bright areas can be further suppressed through anti-exposure processing. Therefore, even if the original environment is not ideal, the real-time image after image enhancement can still maintain a sufficiently high comparability with the reference image, thereby ensuring that displacement detection can continue, without having to be forced to stop when the environment is not ideal, as is the case with traditional precision optical measurement.

[0039] It should be noted that the environmental enhancement scheme in this embodiment is used to ensure the quality of the input image upon which digital image correlation processing depends. That is, there is a sequential relationship between the temperature control unit 10, the infrared laser emission unit 11, and the image enhancement processing module 12 and the processing unit 5: the former improves imaging stability and feature visibility, while the latter completes the synchronous displacement calculation of multiple measurement points based on a stable image. Through this linked design, the applicability of the device in this application in complex engineering environments is expanded.

[0040] The fifth embodiment of this application further describes the anomaly tracing and security extension functions based on the first embodiment. Displacement detection devices are often located on bridges, slopes, and buildings without close, continuous monitoring, especially during long-term or nighttime monitoring. These devices may face challenges such as human approach, accidental contact, external obstruction, and interference from abnormal environments. Based solely on the displacement data, it is sometimes difficult to distinguish whether the anomaly in the detection result originates from a sudden change in the actual displacement of the measured object or from external interference with the device. Therefore, in some implementations, the core vision device can also integrate a surveillance camera, and the device can also include a human body sensing module, an audible and visual alarm module, and an alarm sending module to improve the traceability of anomalies while ensuring detection functionality.

[0041] In a long-term building tilt monitoring scenario, the device may be deployed near residential buildings or inside construction fences, where there is significant pedestrian traffic. If the displacement curve suddenly jumps at a certain moment, it is difficult to determine from the data alone whether it is due to a sudden deformation of the building itself, someone touching the tripod, obstructing the lens, or temporarily moving the equipment. In this embodiment, when the human body sensing module detects someone approaching the preset range of the device, the system can trigger the audible and visual alarm module to issue an on-site warning, control the monitoring camera to record the current scene, and simultaneously send a prompt message to the management platform or management personnel terminal through the alarm sending module. Therefore, even if abnormal fluctuations are subsequently found in the data, the source of the anomaly can be identified by cross-checking the on-site images and alarm records for the corresponding time period.

[0042] In bridge construction monitoring scenarios, equipment is typically deployed on access roads, bridge-side maintenance platforms, or temporary construction areas. The passage of construction machinery, material handling by personnel, or nighttime inspections can all disturb the equipment. Without an anomaly tracking mechanism, if monitoring results are abnormal, technicians often need to redeploy the equipment and re-establish reference images, affecting work continuity. This embodiment uses surveillance camera recording, human body sensing triggering, and alarm sending. It can be used for theft prevention and to assist in determining the cause of changes in the device's attitude, thereby prompting operators to re-establish reference images or update the attitude baseline when necessary. Therefore, the security and tracking functions in this embodiment are an extension of improving the long-term reliability of on-site engineering applications.

[0043] The sixth embodiment of this application provides a method for detecting the displacement of a mobile tripod-type engineering structure. For example... Figure 4 As shown, this method corresponds to the aforementioned device embodiment and can form a process link based on the functions available at the device level. It is used to perform displacement detection on engineering structures with multiple measurement point targets, including: S1. Deploy a mobile tripod support unit, and install the vision core device by means of a gimbal set on the mobile tripod support unit; S2. Adjust the observation direction of the vision core device so that the multiple measurement point targets enter the field of view of the vision core device; S3. Acquire reference images and real-time images of multiple measurement point targets within the same field of view; S4. Perform digital image correlation matching on the image regions of the multiple measurement point targets in the reference image and the real-time image to synchronously obtain the displacement changes of the multiple measurement point targets in the image; S5. Collect the attitude information of the vision core device, and correct the installation posture deviation of the vision core device according to the attitude information; S6. Calculate the horizontal and vertical displacements of the engineering structure based on the displacement changes of the multiple measured target points in the image after correction; S7. Output the detection results of the horizontal and vertical displacements.

[0044] Compared with the traditional point-by-point measurement method, the method in this embodiment adopts the approach of "deploying equipment - establishing a unified field of view - acquiring reference images - continuously acquiring real-time images - synchronous correlation matching of multiple measurement points - attitude compensation - unified output results", thereby ensuring the synchronous detection capability of multiple measurement points in the process.

[0045] Combination Figure 1 As shown, in step S1, a movable tripod support unit 3 is deployed, and the vision core device 1 is installed via a gimbal 2 mounted on the movable tripod support unit 3. The key to this step is selecting a location that facilitates simultaneous entry of multiple targets into the field of view while ensuring stable support of the device, based on the engineering object, observation distance, distribution of measurement points, and surrounding obstructions. For example, in bridge inspection, priority can be given to observation locations that cover the main measurement points, such as underpasses, riverbank platforms, or bridge-side maintenance walkways; in slope inspection, priority can be given to stable areas at the slope toe or observation platforms that do not obstruct key parts of the slope; in building inspection, open spaces facing the target facade with a good viewing angle can be selected. By selecting a suitable location, the subsequent gimbal adjustment range can be reduced from the outset, and the stability of the image's geometric relationship can be improved.

[0046] In step S2, the observation direction of the vision core device 1 is adjusted so that multiple measurement point targets 7 enter the field of view of the vision core device 1. This step can be completed manually or automatically by the processing unit 5 controlling the pan-tilt unit 2. In practice, a preview image can be acquired first to identify the approximate position of each measurement point target 7, and then the pan-tilt unit 2 can be controlled to rotate and tilt to place multiple targets in suitable positions in the image. Here, "suitable position" means that all key targets are visible, and that the size, spacing, and edge sharpness of each target in the image are sufficient to meet the requirements of subsequent digital image processing. If some targets are located too close to the edge of the field of view, they are easily moved out of the image due to slight equipment vibration during subsequent detection. Therefore, it is preferable to place the key targets in the middle or upper middle area of ​​the field of view.

[0047] In step S3, reference images and real-time images of the target 7 containing multiple measurement points are acquired. Reference images are typically acquired once at the start of the detection, but can also be acquired again after a long monitoring period, significant changes in environmental conditions, or after redeployment of the equipment. Real-time images are acquired continuously at a preset sampling frequency. To ensure the quality of the reference images, it is preferable to establish them when the gimbal 2 is stable, the attitude sensor 4 outputs stable data, and the imaging is clear and unobstructed. Taking bridge dynamic load tests as an example, reference images can be acquired before the loaded vehicle enters; taking tunnel convergence monitoring as an example, reference images can be acquired after the current round of excavation and temporary support is completed and the surrounding rock condition is temporarily stable. By appropriately selecting the timing of reference image acquisition, subsequent real-time displacement results can more accurately reflect changes relative to the initial state.

[0048] In step S4, digital image correlation matching is performed on the image regions of multiple measurement point targets 7 in the reference image and the real-time image to obtain the displacement changes of the multiple measurement point targets 7 in the image. This step may further include: extracting the reference sub-regions corresponding to each of the multiple targets in the reference image; determining the search range for each reference sub-region in the real-time image; calculating the correlation between the reference sub-region and the candidate positions within the search range; selecting the position with the best correlation as the matching result; and forming the image displacement changes of multiple measurement points based on the matching results. Since multiple targets can participate in the processing synchronously in the same frame of real-time image, the system can obtain the displacement information of multiple measurement points within one sampling period. Compared with the traditional point-by-point plotting scheme, this step shortens the total time consumption of multi-point measurement and facilitates the observation of the dynamic response of the structure.

[0049] In step S5, the attitude information of the vision core device 1 is acquired, and the installation posture deviation of the vision core device 1 is corrected based on the attitude information. The direct reason for this step is that moving the tripod support unit 3 enables the device to be deployed quickly, but it may also cause attitude differences in different detection positions, different ground conditions, and different weather conditions. If attitude compensation is not performed, the image displacement change obtained in step four may be mixed with background offset caused by the device's own pitch, roll, or turn changes. This step first estimates the overall offset trend of the current image relative to the reference attitude based on the output of the attitude sensor 4, and then the processing unit 5 removes or corrects this offset from the image displacement change, so that the retained displacement result is closer to the actual deformation of the measured engineering structure itself. Through this processing, the method of this embodiment allows the device to reduce the leveling requirements through data compensation after it is quickly in place.

[0050] In step S6, based on the displacement changes of multiple calibrated measurement point targets 7 in the image, the horizontal and vertical displacements of the engineering structure are calculated. This step can be performed in conjunction with the device's calibration parameters, target scale information, observation geometry, or pre-set conversion factors. For bridge monitoring, the vertical displacement curves of the mid-span measurement point, quarter-span measurement point, and adjacent support measurement points can be output separately to assess deflection distribution; for slope monitoring, the horizontal and vertical displacements of the slope top, middle, and toe measurement points can be output separately to analyze deformation propagation trends; for building monitoring, the displacement changes of measurement points on different floors or facades can be output to assist in judging uneven settlement or structural lateral displacement. Since the displacement results of multiple measurement points are obtained under the same time reference, further analysis of inter-measurement differences, deformation compatibility analysis, and anomaly identification can be performed.

[0051] In step S7, the detection results of horizontal and vertical displacements are output. The output may include numerical values, curves, threshold prompts, and anomaly alarms. In a slope emergency inspection example, managers are more concerned about whether certain key measuring points experience a sudden increase in displacement within a short period. Therefore, the system can compare the displacement changes of each key measuring point with preset thresholds, and when the threshold is exceeded, a prompt is simultaneously displayed on the local screen and a remote terminal. In a bridge load test example, technicians may be more concerned about the displacement curve throughout the loading process. Therefore, the system can plot the curve on the local screen in real time and synchronously send the data to the test recording terminal. Thus, the output step in this method embodiment is used to convert the detection results generated in the preceding steps into data that can be used for on-site judgment and subsequent analysis.

[0052] In some implementations, the above method also includes sub-pixel localization of the correlation matching results of the digital image. This process can be performed after step four and before step six. Processing unit 5 continuously fits the local region around the pixel-level optimal matching position to obtain a finer displacement offset within the pixel. In this way, even if the displacement change between adjacent sampling times is small, a more continuous and smoother displacement sequence can be output in scenarios such as small vibrations of bridges, early settlement of buildings, or slow creep of slopes. Without this processing, the detection result may remain unchanged for several frames and then jump abruptly, which is not conducive to reflecting the real deformation process.

[0053] In some implementations, the above method further includes regulating the internal temperature of the vision core device 1 and performing defogging and anti-exposure processing on the real-time images before acquisition. This extended step can be applied to complex environmental scenarios such as open-air bridges, slopes, and low-light tunnels. The reason for this is that digital image correlation matching can only stably obtain correlation results with prominent peaks when sufficient feature comparability is maintained between the real-time image and the reference image; temperature control, defogging, and anti-exposure processing are precisely to maintain this comparability. After these preprocessing steps, the correlation matching success rate and stability in step four are usually higher, which in turn helps to make the displacement calculation results in step six more reliable.

[0054] In some implementations, the above methods also include abnormal scene recording and alarm linkage. When the system detects that someone is approaching the device, the device is subjected to abnormal contact, a large area of ​​obstruction appears in the image, or a sudden change in displacement results that is inconsistent with adjacent measuring points within a short period of time, the monitoring camera can simultaneously record the scene and send a prompt to the management platform through the alarm sending module. In this way, in subsequent data analysis, technicians can compare the period of abnormal displacement with the scene footage to determine whether the anomaly originates from structural deformation or external interference, thereby improving the reliability of the monitoring conclusions.

[0055] Furthermore, to illustrate the applicability of the method and apparatus of this application to different engineering applications, three comprehensive application examples are given below.

[0056] Firstly, in the dynamic load testing of bridges, multiple measurement point targets are first set up at key locations on the beam. The device is then placed in a stable area under or on the side of the bridge to complete automatic point finding and reference image establishment. When the test vehicle enters the bridge deck, the system continuously acquires real-time images, performs digital image correlation matching and sub-pixel positioning on multiple measurement points simultaneously, and outputs displacement curves of the mid-span and adjacent measurement points in combination with attitude correction, thereby reflecting the stress response process of the bridge.

[0057] Secondly, in tunnel convergence monitoring, multiple targets are deployed at the arch crown, arch waist, and sidewalls. After the device is deployed in a stable area behind the tunnel face, the pan-tilt unit is automatically adjusted to bring all targets into the field of view. Subsequently, displacement data of each measuring point is continuously output for analysis of the deformation development of the surrounding rock and support structure.

[0058] Third, during post-rain slope inspections, targets are distributed at the top, middle, and bottom of the slope. The image enhancement processing module improves the real-time image quality under foggy conditions, and digital image correlation processing is used to output displacement results at multiple measurement points to help determine whether there is a risk of sudden increase in local deformation.

[0059] Through the above embodiments, the technical solution proposed in this application, focusing on the technical objective of multi-point dynamic displacement detection of engineering structures, provides a complete link design for on-site deployment, field of view establishment, image acquisition, digital image correlation solving, attitude compensation, environment enhancement, and result output. In particular, there is a clear functional synergy between digital image correlation processing and attitude correction: the former is responsible for converting image changes of multiple measurement point targets within the same field of view into displacement information, while the latter is responsible for eliminating the additional effects caused by the device's own attitude changes; the two work together to enable the device to achieve synchronous displacement detection of multiple measurement points even under mobile deployment conditions.

[0060] Those skilled in the art will understand that, without departing from the core concept of this application, the pattern form of the measurement point target, the lens parameters of the core vision device, the driving method of the gimbal, the specific type of the attitude sensor, and the specific implementation program of digital image correlation matching can all be equivalently replaced according to the application scenario. For example, in long-distance bridge detection, a longer focal length imaging component can be selected, while in close-range building monitoring, a wider field of view imaging component can be selected; in scenarios with fewer and more concentrated measurement points, the search area can be set smaller to improve processing efficiency; in scenarios with longer monitoring cycles, reference images can be reconstructed periodically according to environmental changes. However, regardless of the specific implementation adopted, as long as it is still based on the correlation matching of multiple measurement point reference images and real-time images within the same field of view, combined with attitude information to correct the installation posture deviation, and outputting the displacement result of the engineering structure, it falls under the technical concept disclosed in the embodiments of this application.

[0061] The above embodiments are merely illustrative examples of the specific working principle of the technical solution of this application and do not constitute a limitation on the scope of protection of this application. Equivalent substitutions, modifications and changes made to the above embodiments without departing from the core inventive concept of this application shall all fall within the scope of disclosure of this application. The scope of protection of this application shall be determined by the technical solution defined in the claims.

Claims

1. A mobile tripod-type displacement detection device for engineering structures, used for displacement detection of engineering structures with multiple measuring point targets, characterized in that, include: Mobile tripod support unit; The gimbal is mounted on the mobile tripod support unit; The core vision device is fixed on the gimbal and acquires reference images and real-time images of multiple measurement point targets within the same field of view; An attitude sensor is installed in the vision core device to collect the attitude information of the vision core device; The processing unit is connected to the vision core device and the attitude sensor. It performs digital image correlation matching on the image regions of the multiple measurement point targets in the reference image and the real-time image to synchronously obtain the displacement changes of the multiple measurement point targets in the image. It also corrects the installation posture deviation of the vision core device according to the attitude information and calculates the horizontal and vertical displacements of the engineering structure based on the corrected displacement changes of the multiple measurement point targets in the image. The output unit is connected to the processing unit and outputs the detection results of the horizontal and vertical displacements.

2. The mobile tripod-type displacement detection device for engineering structures as described in claim 1, characterized in that, The processing unit is connected to the gimbal control. The processing unit controls the gimbal to rotate according to the distribution information of the multiple measurement point targets, so that the multiple measurement point targets enter the field of view of the vision core device, and starts displacement detection after the positioning is completed.

3. The mobile tripod-type displacement detection device for engineering structures as described in claim 1, characterized in that, The processing unit also performs subpixel localization on the positions of the multiple measurement point targets obtained through the digital image correlation matching, so as to determine the displacement changes of each measurement point target.

4. The mobile tripod-type displacement detection device for engineering structures as described in claim 1, characterized in that, The vision core device also includes a temperature control unit, which is used to regulate the internal temperature of the vision core device.

5. The mobile tripod-type displacement detection device for engineering structures as described in claim 1, characterized in that, The core vision device also includes an infrared laser emitting unit and an image enhancement processing module, which is used to perform fog removal and anti-exposure processing on the real-time image.

6. The mobile tripod-type displacement detection device for engineering structures as described in claim 1, characterized in that, The output unit includes a local display screen mounted on the vision core device and a remote terminal interface that is communicatively connected to the processing unit, for on-site display and remote output, respectively.

7. A mobile tripod-based displacement detection method for engineering structures, used for displacement detection of engineering structures with multiple measuring point targets, characterized in that, include: A mobile tripod support unit is set up, and the core vision device is installed by a gimbal set on the mobile tripod support unit; Adjust the observation direction of the vision core device so that the multiple measurement point targets enter the field of view of the vision core device; Acquire reference images and real-time images of targets containing multiple measurement points within the same field of view; Digital image correlation matching is performed on the image regions of the multiple measurement point targets in the reference image and the real-time image to synchronously obtain the displacement changes of the multiple measurement point targets in the image; The attitude information of the vision core device is collected, and the installation posture deviation of the vision core device is corrected based on the attitude information. The horizontal and vertical displacements of the engineering structure are calculated based on the displacement changes of the multiple measured target points in the image after correction. Output the detection results of the horizontal and vertical displacements.

8. The displacement detection method for a mobile tripod-type engineering structure as described in claim 7, characterized in that, Adjusting the observation direction of the vision core device includes: The gimbal is automatically rotated according to the distribution information of the multiple measurement point targets, so that the multiple measurement point targets enter the field of view of the vision core device; Displacement detection is automatically initiated after positioning is completed.

9. The displacement detection method for a mobile tripod-type engineering structure as described in claim 7, characterized in that, The step of performing digital image correlation matching on the image regions of the multiple measurement point targets in the reference image and the real-time image includes: Extract the image regions corresponding to each of the multiple measurement point targets from the reference image; In the real-time image, each of the image regions is respectively matched for relevant information; Subpixel localization is performed on the positions obtained from the relevant matching to determine the displacement changes of the target at each measurement point.

10. The displacement detection method for a mobile tripod-type engineering structure as described in claim 7, characterized in that, The method further includes: Before acquiring the real-time images, the internal temperature of the vision core device is adjusted; The real-time image is subjected to defogging and anti-exposure processing; The detection results are displayed on a local screen and sent to a remote terminal.

Images