Image Reconstruction-Based Method and System for Monitoring Structure Settlement

Abstract

This invention discloses a method and system for monitoring settlement of structures based on image reconstruction, applied in the field of intelligent monitoring technology. The method includes: setting a target on the target structure; capturing a reference image of the target and obtaining its initial coordinates; forming a matching image; acquiring multiple measured images and generating a corrected image during settlement monitoring of the target structure; calculating the pixel coordinates of the center point of the large circular image based on the positions of the small and large circular images in the corrected image as measured coordinates; and calculating the settlement value of the target structure based on the measured coordinates and the initial coordinates. This invention obtains a more accurate target pattern by integrating multiple images and combines multiple circular images for point location search, which can more accurately identify the points and calculate more accurate settlement values, effectively reducing the settlement monitoring errors caused by environmental and camera factors.

Description

Technical Field

[0001] This invention relates to the field of intelligent monitoring technology, specifically to a method and system for monitoring the settlement of structures based on image reconstruction. Background Technology

[0002] When using targets for sedimentation monitoring, especially for sub-millimeter level high-precision monitoring, changes in environmental water vapor density and temperature gradients cause light refraction, resulting in systematic displacement errors. At the same time, the radial distortion of the lens increases along the radius, and additional errors will occur at the edge target points.

[0003] In the prior art, various techniques have been provided to reduce the corresponding errors. For example, Chinese Patent Application No. CN201811275287.3 discloses a target surface relative pose measurement method and system based on feature points. The measurement method is as follows: (1) Design 5 black circular target surface feature points, set 4 feature points at the 4 apex corners of the target surface, and set 1 feature point at the center of the target surface; (2) Based on the self-calibration of the feature points, calculate the position and attitude information of the target before and after settlement; (3) Feature point identification and centroid positioning. The measurement system includes multiple image-type settlement monitoring instruments set at equal intervals along the length of the track. The camera lens of each monitoring instrument is aimed at the laser light source in the monitoring instrument in front of or behind it. The output end of the image-type settlement monitoring instrument at the front end of the train's travel direction is connected to the input end of the data acquisition and analysis terminal.

[0004] Chinese patent application number CN202510541965.X discloses a non-contact monocular vision high-precision automatic settlement monitoring method, comprising: First, using the YOLOv10s model for target monitoring to improve the real-time inference speed and recognition accuracy of target identification, and solve the problem of target identification in complex backgrounds; Second, combining digital image processing technology and an edge-based least squares ellipse fitting algorithm to achieve accurate target positioning; Third, based on the camera imaging principle, an improved world coordinate calculation method is introduced to solve for the world coordinates corresponding to pixel coordinates, thereby obtaining three-dimensional world coordinates; Finally, the settlement value of the monitoring area is calculated based on the first frame image, and a deep learning model is used to predict future settlement values, providing early warning for structural safety.

[0005] This shows that although existing technologies can identify targets, they rely on the accuracy of the target pixels. If the target pixels are deformed due to shooting reasons, the detection accuracy will be greatly limited. Summary of the Invention

[0006] In order to at least overcome the above-mentioned shortcomings in the prior art, the purpose of this application is to provide a method and system for monitoring the settlement of structures based on image reconstruction.

[0007] In a first aspect, embodiments of this application provide a method for monitoring structure settlement based on image reconstruction, including:

[0008] A target is set on the target structure; the target includes four small circular images set at the four corners of the target and one large circular image set at the center of the target; the lines connecting the center points of the two sets of small circular images located diagonally intersect at the center point of the large circular image.

[0009] The target is photographed to obtain a reference image of the target, and the pixel coordinates of the center point of the large circular image in the reference image are obtained as the initial coordinates;

[0010] The small and large circular images in the reference image are uniformly divided into at least two images, and the edges are identified to form a matching image;

[0011] When monitoring the settlement of a target structure, multiple measured images of the target are continuously captured, and a corrected image is generated based on the matched image and the multiple measured images.

[0012] The pixel coordinates of the center point of the large circular image are calculated based on the positions of the small and large circular images in the corrected image and used as the measured coordinates.

[0013] The settlement value of the target structure is calculated based on the measured coordinates and the initial coordinates.

[0014] In one possible implementation, forming a matching image includes:

[0015] Edge recognition is performed on the first circular image to identify the first edge pixels of the first circular image; the first circular image includes the small circular image and the large circular image;

[0016] At the center point of the first circular image, the first circular image is horizontally and vertically cut to form four uniform first segmented images;

[0017] The first edge pixel corresponding to each first segmented image is used as the matching image corresponding to that first segmented image.

[0018] In one possible implementation, generating the corrected image based on the matched image and multiple measured images includes:

[0019] Edge recognition is performed on each of the measured images to identify the second edge pixels of each second circular image; the second circular image includes the small circular image and the large circular image;

[0020] At the center point of each second circular image, the second circular image is horizontally and vertically cut to form four uniform second segmented images;

[0021] Identify the similarity between the second edge pixel of each second segmented image and the matching image at the corresponding position, and select the second segmented image with the highest similarity as the selected image at that position;

[0022] All selected images are stitched together to form the corrected image.

[0023] In one possible implementation, the calculation of the measured coordinates includes:

[0024] Ellipse fitting is performed on all the small circular images and the large circular images in the corrected image, and the pixel coordinates of the center point of the ellipse are generated;

[0025] The pixel coordinates of the intersection point of the lines connecting the center points of the ellipses of the two sets of small circular images located diagonally opposite each other are calculated as the first center point coordinates, and the pixel coordinates of the center point of the ellipse of the large circular image are obtained as the second center point coordinates.

[0026] The measured coordinates are obtained by taking a weighted average of the coordinates of the first center point and the coordinates of the second center point.

[0027] In one possible implementation, the weights for the weighted average are obtained by:

[0028] Multiple reference images of the target are captured by photographing the target, and the pixel coordinates of the center points of the small and large circular images in all the reference images are obtained.

[0029] The pixel coordinates of the intersection of the lines connecting the center points of two sets of diagonally opposite small circular images are calculated based on the pixel coordinates of the center points of all the small circular images as the first pixel coordinates, and the pixel coordinates of the center points of all the large circular images are calculated as the second pixel coordinates.

[0030] The covariance of all the first pixel coordinates is calculated to form the first covariance, and the covariance of all the second pixel coordinates is calculated to form the second covariance;

[0031] The weights are formed by normalizing the reciprocals of the first and second covariances.

[0032] In one possible implementation, calculating the settlement value of the target structure based on the measured coordinates and the initial coordinates includes:

[0033] The horizontal scaling factor in the horizontal direction and the vertical scaling factor in the vertical direction are calculated based on the ratio between the physical length and the pixel length in the reference image.

[0034] Calculate the horizontal pixel coordinate difference in the horizontal direction and the vertical pixel coordinate difference in the vertical direction based on the difference between the measured coordinates and the initial coordinates;

[0035] Multiply the horizontal pixel coordinate difference by the horizontal scaling factor to obtain the horizontal displacement, and multiply the vertical scaling factor by the vertical pixel coordinate difference to obtain the vertical displacement;

[0036] The horizontal and vertical displacements are used as the settlement values.

[0037] Secondly, this application also provides a structure settlement monitoring system based on image reconstruction, including:

[0038] A building unit is configured to set a target on a target structure; the target includes four small circular images set at the four corners of the target and a large circular image set at the center of the target; the lines connecting the center points of the two sets of diagonally opposite small circular images intersect at the center point of the large circular image.

[0039] The initial unit is configured to capture a reference image of the target and obtain the pixel coordinates of the center point of the large circular image in the reference image as the initial coordinates;

[0040] The splitting unit is configured to uniformly split both the small circular image and the large circular image in the reference image into at least two images and identify edges to form a matching image;

[0041] The correction unit is configured to continuously capture multiple measured images of the target when monitoring the settlement of the target structure, and generate a correction image based on the matching image and the multiple measured images.

[0042] The calculation unit is configured to calculate the pixel coordinates of the center point of the large circular image as measured coordinates based on the positions of the small circular image and the large circular image in the corrected image.

[0043] A settlement unit is configured to calculate the settlement value of the target structure based on the measured coordinates and the initial coordinates.

[0044] In one possible implementation, the splitting unit is further configured as follows:

[0045] Edge recognition is performed on the first circular image to identify the first edge pixels of the first circular image; the first circular image includes the small circular image and the large circular image.

[0046] At the center point of the first circular image, the first circular image is horizontally and vertically cut to form four uniform first segmented images;

[0047] The first edge pixel corresponding to each first segmented image is used as the matching image corresponding to that first segmented image.

[0048] In one possible implementation, the correction unit is further configured as follows:

[0049] Edge recognition is performed on each of the measured images to identify the second edge pixels of each second circular image; the second circular image includes the small circular image and the large circular image;

[0050] At the center point of each second circular image, the second circular image is horizontally and vertically cut to form four uniform second segmented images;

[0051] Identify the similarity between the second edge pixel of each second segmented image and the matching image at the corresponding position, and select the second segmented image with the highest similarity as the selected image at that position;

[0052] All selected images are stitched together to form the corrected image.

[0053] In one possible implementation, the computing unit is further configured as follows:

[0054] Ellipse fitting is performed on all the small circular images and the large circular images in the corrected image, and the pixel coordinates of the center point of the ellipse are generated;

[0055] The pixel coordinates of the intersection point of the lines connecting the center points of the ellipses of the two sets of small circular images located diagonally opposite each other are calculated as the first center point coordinates, and the pixel coordinates of the center point of the ellipse of the large circular image are obtained as the second center point coordinates.

[0056] The measured coordinates are obtained by taking a weighted average of the coordinates of the first center point and the coordinates of the second center point.

[0057] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0058] This invention relates to a method and system for monitoring settlement of structures based on image reconstruction. By integrating multiple images, a more accurate target pattern is obtained. By combining multiple circular images for point location search, the point location can be identified more accurately, thereby calculating a more accurate settlement value. This effectively reduces the settlement value monitoring error caused by environmental and camera factors. Attached Figure Description

[0059] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0060] Figure 1 This is a schematic diagram of the method steps in an embodiment of this application;

[0061] Figure 2 This is a schematic diagram of the target in this application;

[0062] Figure 3 This is a schematic diagram of the target structure and target of this application;

[0063] Figure 4 This is a schematic diagram of the target cutting in this application. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the accompanying drawings in this application are for illustrative and descriptive purposes only and are not intended to limit the scope of protection of this application. Furthermore, it should be understood that the schematic drawings are not drawn to scale. The flowcharts used in this application illustrate operations implemented according to some embodiments of this application. It should be understood that the operations in the flowcharts may not be implemented in sequence, and steps without logical contextual relationships may be reversed or implemented simultaneously. In addition, those skilled in the art, guided by the content of this application, may add one or more other operations to the flowcharts, or remove one or more operations from the flowcharts.

[0065] Furthermore, the described embodiments are merely some, not all, of the embodiments of this application. The components of the embodiments of this application described and illustrated herein can typically be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0066] Please refer to the following: Figure 1 This is a schematic flowchart of the structure settlement monitoring method based on image reconstruction provided in an embodiment of the present invention. Further, the structure settlement monitoring method based on image reconstruction may specifically include the contents described in steps S1-S6.

[0067] S1: Set a target on the target structure; the target includes four small circular images set at the four corners of the target and one large circular image set at the center of the target; the lines connecting the center points of the two sets of small circular images located diagonally intersect at the center point of the large circular image.

[0068] S2: Take a picture of the target to obtain a reference image of the target, and obtain the pixel coordinates of the center point of the large circular image in the reference image as the initial coordinates;

[0069] S3: Divide the small circular image and the large circular image in the reference image into at least two images and identify the edges to form a matching image;

[0070] S4: When monitoring the settlement of the target structure, the target is continuously photographed to obtain multiple measured images of the target, and a corrected image is generated based on the matching image and the multiple measured images.

[0071] S5: Calculate the pixel coordinates of the center point of the large circular image based on the positions of the small and large circular images in the corrected image, and use them as the measured coordinates.

[0072] S6: Calculate the settlement value of the target structure based on the measured coordinates and the initial coordinates.

[0073] In the implementation of this application, a target completely different from the prior art is used; please refer to the specific form for details. Figure 2 The target is a square target, consisting of four small circular images A, B, C, and D, and one large circular image M. The four small circular images are positioned at the four corners of the target, and the large circular image is positioned at the center. For a 100cm x 100cm target, the diameter of the large circular image is typically around 50cm, and the diameter of the small circular images is typically around 30cm. Furthermore, the lines connecting the center points of any two sets of diagonally opposite small circular images intersect at the center point of the large circular image. In other words, the target can simultaneously possess two sets of parameters for verifying the center point: the intersection point and the center point of the large circular image. (Example) Figure 2 In the target, the line connecting the small circular images A and D intersects the line connecting the small circular images C and B at point I. On this target, point I is the center of the large circular image M.

[0074] In implementing this embodiment, it is necessary to photograph the target using an imaging device and generate corresponding initial data. It should be understood that the imaging device should maintain a fixed shooting point, whether photographing a reference image or a measured image. Please refer to [link to relevant documentation]. Figure 3 Examples of capturing reference images and actual test images are shown. The captured images need to be preprocessed using image recognition to identify the target before further center point identification can be performed. The reference images should be captured under ideal conditions in the actual shooting environment.

[0075] For example, the pixel coordinates are calculated from the bounding box coordinates (normalized center point, width, and height) output by YOLOv5, and the image corresponding to the target is cropped from the original image. During preprocessing, the color image is first converted to grayscale using the Luminance algorithm (gray=0.299R+0.587G+0.114B) while retaining brightness information. Then, contrast-limited adaptive histogram equalization (CLAHE) is used to enhance local image contrast and highlight edge information. Homomorphic filtering is then used to separate the incident and reflected components of the image, suppressing noise and enhancing details. Next, the image is converted to a black and white binary image by dynamically setting the threshold based on the p-parameter method, simplifying feature extraction. It should be understood that most targets are laid out in black and white, so the binarized image will not lose much information. Finally, residual bounding box noise is removed through logical NOT operation, edge detection, and morphological operations (opening / closing operations) to complete the image preprocessing.

[0076] In this embodiment, all coordinate data are pixel coordinates, which will not be described further below. The order of determining the initial coordinates and edge recognition can be interchanged. That is, edge recognition can be performed first and then the initial coordinates can be calculated based on the results of edge recognition; or the initial coordinates can be determined by manual marking first and then edge recognition can be performed. The specific process is not limited in this embodiment.

[0077] In this embodiment, to more accurately detect small and large circular images during the recognition process, templates for detecting small and large circular images need to be constructed first. This involves uniformly splitting both the small and large circular images in the reference image into at least two images, and identifying the edges as matching images. Generally, the circular image is divided into four equal parts according to horizontal and vertical directions. It should be understood that multiple matching images exist, each corresponding to an image split from a specific part.

[0078] In this embodiment of the application, during actual settlement monitoring, multiple measured images need to be continuously captured and equally divided. This allows for the identification of the most matching image block at each corresponding location through image matching, which is then stitched together to form a corrected image. The measured coordinates can be calculated by analyzing the center points of the small and medium-sized circular images and the large circular images in the corrected image. The settlement value is then calculated based on the difference between the measured coordinates and the initial coordinates.

[0079] In one possible implementation, forming a matching image includes:

[0080] Edge recognition is performed on the first circular image to identify the first edge pixels of the first circular image; the first circular image includes the small circular image and the large circular image.

[0081] At the center point of the first circular image, the first circular image is horizontally and vertically cut to form four uniform first segmented images;

[0082] The first edge pixel corresponding to each first segmented image is used as the matching image corresponding to that first segmented image.

[0083] When implementing the embodiments of this application, please refer to Figure 4 The process of constructing the matching image involves dividing the image into four equal parts, with the red edges representing the identified edges. Since the actual coordinates detected during settlement recognition are x and y coordinates, corresponding to horizontal and vertical displacements, this four-part division method ensures that the control points at the ends have fixed coordinate values ​​in one direction, thus improving detection accuracy.

[0084] In one possible implementation, generating the corrected image based on the matched image and multiple measured images includes:

[0085] Edge recognition is performed on each of the measured images to identify the second edge pixels of each second circular image; the second circular image includes the small circular image and the large circular image;

[0086] At the center point of each second circular image, the second circular image is horizontally and vertically cut to form four uniform second segmented images;

[0087] Identify the similarity between the second edge pixel of each second segmented image and the matching image at the corresponding position, and select the second segmented image with the highest similarity as the selected image at that position;

[0088] All selected images are stitched together to form the corrected image.

[0089] In implementing this application, a specific method for generating a corrected image is provided. This involves edge recognition of each measured image to identify the second edge pixels. The recognition process can employ the methods described in the previous embodiments, and this application does not impose further limitations. Then, each second circular image is uniformly divided to form multiple second segmented images. The similarity between the second edge pixels corresponding to the second segmented images and the first edge pixels in the matching images can be calculated. The selected image with the highest similarity in each region is then stitched together to form the corrected image. The similarity can be calculated using Euclidean distance or Pearson correlation coefficient. Similarity calculation between curves is a mature existing technology, and this application does not impose further limitations. This method ensures that the different circular images in the final corrected image are all in a shape closest to a circle, thereby improving the accuracy of subsequent recognition.

[0090] In one possible implementation, the calculation of the measured coordinates includes:

[0091] Ellipse fitting is performed on all the small circular images and the large circular images in the corrected image, and the pixel coordinates of the center point of the ellipse are generated;

[0092] The pixel coordinates of the intersection point of the lines connecting the center points of the ellipses of the two sets of small circular images located diagonally opposite each other are calculated as the first center point coordinates, and the pixel coordinates of the center point of the ellipse of the large circular image are obtained as the second center point coordinates.

[0093] The measured coordinates are obtained by taking a weighted average of the coordinates of the first center point and the coordinates of the second center point.

[0094] In implementing this embodiment, it is necessary to perform ellipse fitting on both small and large circular images, and calculate the coordinates of the center point of the fitted ellipse. For example, the equation for the fitted ellipse is as follows:

[0095]

[0096] The fitting process uses the least squares method. It should be understood that the fitted data consists entirely of edge pixel data. The center point coordinates can then be calculated using the following formula:

[0097]

[0098] In the formula, A, B, C, D, E, and F are the coefficients of the elliptic equation, and x... c Let y be the horizontal coordinate of the center point of the circle. c Let y be the y-coordinate of the center point of the circle.

[0099] In this embodiment of the application, after calculating the center point coordinates of each circular image, the first center point coordinates can be calculated based on the center point coordinates of the small circular image, and the center point coordinates of the large circular image can be used as the second center point coordinates. Under the condition of no error, the two coordinates should coincide. However, when considering the error, the two coordinates need to be weighted and summed to improve the accuracy of the final monitoring.

[0100] In one possible implementation, the weights for the weighted average are obtained by:

[0101] Multiple reference images of the target are captured by photographing the target, and the pixel coordinates of the center points of the small and large circular images in all the reference images are obtained.

[0102] The pixel coordinates of the intersection of the lines connecting the center points of two sets of diagonally opposite small circular images are calculated based on the pixel coordinates of the center points of all the small circular images as the first pixel coordinates, and the pixel coordinates of the center points of all the large circular images are calculated as the second pixel coordinates.

[0103] The covariance of all the first pixel coordinates is calculated to form the first covariance, and the covariance of all the second pixel coordinates is calculated to form the second covariance;

[0104] The weights are formed by normalizing the reciprocals of the first and second covariances.

[0105] In the implementation of this application embodiment, the weighted weight calculation is based on statistical analysis, which can be performed simultaneously with the initial coordinate calculation. This requires obtaining the first pixel coordinates and second pixel coordinates of multiple reference images, and performing statistical analysis on the first pixel coordinates and second pixel coordinates respectively to calculate the first covariance and the second covariance. The covariance characterizes the dispersion of the two-dimensional data of pixel coordinates. The higher the dispersion, the lower the reliability of the data. Therefore, the reciprocal of the first covariance and the reciprocal of the second covariance can be normalized to form the corresponding weight value to improve the detection accuracy.

[0106] In one possible implementation, calculating the settlement value of the target structure based on the measured coordinates and the initial coordinates includes:

[0107] The horizontal scaling factor in the horizontal direction and the vertical scaling factor in the vertical direction are calculated based on the ratio between the physical length and the pixel length in the reference image.

[0108] Calculate the horizontal pixel coordinate difference in the horizontal direction and the vertical pixel coordinate difference in the vertical direction based on the difference between the measured coordinates and the initial coordinates;

[0109] Multiply the horizontal pixel coordinate difference by the horizontal scaling factor to obtain the horizontal displacement, and multiply the vertical scaling factor by the vertical pixel coordinate difference to obtain the vertical displacement;

[0110] The horizontal and vertical displacements are used as the settlement values.

[0111] When implementing the embodiments of this application, please refer to Figure 2 The vertical and horizontal scale factors characterize the proportional relationship between physical size and pixel size, and are calculated using the following formula:

[0112]

[0113] In the formula, S X S is the horizontal scaling factor. Y D is the vertical scaling factor. AB Let d be the physical distance between the center points of circles A and B. AB D is the pixel distance between the center points of circles A and B; CD Let d be the physical distance between the center points of circles A and B. CD D is the pixel distance between the center points of circles A and B; AC Let d be the physical distance between the center points of circles A and B. AC D is the pixel distance between the center points of circles A and B; BD Let d be the physical distance between the center points of circles A and B. BD The pixel distance between the center points of circles A and B.

[0114] When calculating horizontal and vertical displacements, the following formula is used:

[0115]

[0116] In the formula, X D For horizontal displacement, Y D For vertical displacement, x t The horizontal value of the measured coordinates, y t x0 represents the vertical value of the measured coordinates, x0 represents the horizontal value of the initial coordinates, and y0 represents the vertical value of the initial coordinates.

[0117] Based on the same inventive concept, this application also provides a structure settlement monitoring system based on image reconstruction, comprising:

[0118] A building unit is configured to set a target on a target structure; the target includes four small circular images set at the four corners of the target and a large circular image set at the center of the target; the lines connecting the center points of the two sets of diagonally opposite small circular images intersect at the center point of the large circular image.

[0119] The initial unit is configured to capture a reference image of the target and obtain the pixel coordinates of the center point of the large circular image in the reference image as the initial coordinates;

[0120] The splitting unit is configured to uniformly split both the small circular image and the large circular image in the reference image into at least two images and identify edges to form a matching image;

[0121] The correction unit is configured to continuously capture multiple measured images of the target when monitoring the settlement of the target structure, and generate a correction image based on the matching image and the multiple measured images.

[0122] The calculation unit is configured to calculate the pixel coordinates of the center point of the large circular image as measured coordinates based on the positions of the small circular image and the large circular image in the corrected image.

[0123] A settlement unit is configured to calculate the settlement value of the target structure based on the measured coordinates and the initial coordinates.

[0124] In one possible implementation, the splitting unit is further configured as follows:

[0125] Edge recognition is performed on the first circular image to identify the first edge pixels of the first circular image; the first circular image includes the small circular image and the large circular image;

[0126] At the center point of the first circular image, the first circular image is horizontally and vertically cut to form four uniform first segmented images;

[0127] The first edge pixel corresponding to each first segmented image is used as the matching image corresponding to that first segmented image.

[0128] In one possible implementation, the correction unit is further configured as follows:

[0129] Edge recognition is performed on each of the measured images to identify the second edge pixels of each second circular image; the second circular image includes the small circular image and the large circular image;

[0130] At the center point of each second circular image, the second circular image is horizontally and vertically cut to form four uniform second segmented images;

[0131] Identify the similarity between the second edge pixel of each second segmented image and the matching image at the corresponding position, and select the second segmented image with the highest similarity as the selected image at that position;

[0132] All selected images are stitched together to form the corrected image.

[0133] In one possible implementation, the computing unit is further configured as follows:

[0134] Ellipse fitting is performed on all the small circular images and the large circular images in the corrected image, and the pixel coordinates of the center point of the ellipse are generated;

[0135] The pixel coordinates of the intersection point of the lines connecting the center points of the ellipses of the two sets of small circular images located diagonally opposite each other are calculated as the first center point coordinates, and the pixel coordinates of the center point of the ellipse of the large circular image are obtained as the second center point coordinates.

[0136] The measured coordinates are obtained by taking a weighted average of the coordinates of the first center point and the coordinates of the second center point.

[0137] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0138] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices or units, or may be electrical, mechanical or other forms of connection.

[0139] The units described as separate components may or may not be physically separate. As will be apparent to those skilled in the art, the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0140] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0141] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or grid device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0142] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for monitoring structure settlement based on image reconstruction, characterized in that, include: A target is set on the target structure; the target includes four small circular images set at the four corners of the target and one large circular image set at the center of the target; the lines connecting the center points of the two sets of small circular images located diagonally intersect at the center point of the large circular image. The target is photographed to obtain a reference image of the target, and the pixel coordinates of the center point of the large circular image in the reference image are obtained as the initial coordinates; The small and large circular images in the reference image are uniformly divided into at least two images, and the edges are identified to form a matching image; When monitoring the settlement of a target structure, multiple measured images of the target are continuously captured, and a corrected image is generated based on the matched image and the multiple measured images. The pixel coordinates of the center point of the large circular image are calculated based on the positions of the small and large circular images in the corrected image and used as the measured coordinates. The settlement value of the target structure is calculated based on the measured coordinates and the initial coordinates.

2. The method for monitoring structure settlement based on image reconstruction according to claim 1, characterized in that, Forming a matching image includes: Edge recognition is performed on the first circular image to identify the first edge pixels of the first circular image; the first circular image includes the small circular image and the large circular image; At the center point of the first circular image, the first circular image is horizontally and vertically cut to form four uniform first segmented images; The first edge pixel corresponding to each first segmented image is used as the matching image corresponding to that first segmented image.

3. The method for monitoring structure settlement based on image reconstruction according to claim 2, characterized in that, Generating a corrected image based on the matched image and multiple measured images includes: Edge recognition is performed on each of the measured images to identify the second edge pixels of each second circular image; the second circular image includes the small circular image and the large circular image; At the center point of each second circular image, the second circular image is horizontally and vertically cut to form four uniform second segmented images; Identify the similarity between the second edge pixel of each second segmented image and the matching image at the corresponding position, and select the second segmented image with the highest similarity as the selected image at that position; All selected images are stitched together to form the corrected image.

4. The method for monitoring structure settlement based on image reconstruction according to claim 1, characterized in that, The calculation of the measured coordinates includes: Ellipse fitting is performed on all the small circular images and the large circular images in the corrected image, and the pixel coordinates of the center point of the ellipse are generated; The pixel coordinates of the intersection point of the lines connecting the center points of the ellipses of the two sets of small circular images located diagonally opposite each other are calculated as the first center point coordinates, and the pixel coordinates of the center point of the ellipse of the large circular image are obtained as the second center point coordinates. The measured coordinates are obtained by taking a weighted average of the coordinates of the first center point and the coordinates of the second center point.

5. The method for monitoring structure settlement based on image reconstruction according to claim 4, characterized in that, The weights for the weighted average are obtained by: Multiple reference images of the target are captured by photographing the target, and the pixel coordinates of the center points of the small and large circular images in all the reference images are obtained. The pixel coordinates of the intersection of the lines connecting the center points of two sets of diagonally opposite small circular images are calculated based on the pixel coordinates of the center points of all the small circular images as the first pixel coordinates, and the pixel coordinates of the center points of all the large circular images are calculated as the second pixel coordinates. The covariance of all the first pixel coordinates is calculated to form the first covariance, and the covariance of all the second pixel coordinates is calculated to form the second covariance; The weights are formed by normalizing the reciprocals of the first and second covariances.

6. The method for monitoring structure settlement based on image reconstruction according to claim 1, characterized in that, Calculating the settlement value of the target structure based on the measured coordinates and the initial coordinates includes: The horizontal scaling factor in the horizontal direction and the vertical scaling factor in the vertical direction are calculated based on the ratio between the physical length and the pixel length in the reference image. Calculate the horizontal pixel coordinate difference in the horizontal direction and the vertical pixel coordinate difference in the vertical direction based on the difference between the measured coordinates and the initial coordinates; Multiply the horizontal pixel coordinate difference by the horizontal scaling factor to obtain the horizontal displacement, and multiply the vertical scaling factor by the vertical pixel coordinate difference to obtain the vertical displacement; The horizontal and vertical displacements are used as the settlement values.

7. A structure settlement monitoring system based on image reconstruction, characterized in that, include: A building unit is configured to set a target on a target structure; the target includes four small circular images set at the four corners of the target and a large circular image set at the center of the target; the lines connecting the center points of the two sets of diagonally opposite small circular images intersect at the center point of the large circular image. The initial unit is configured to capture a reference image of the target and obtain the pixel coordinates of the center point of the large circular image in the reference image as the initial coordinates; The splitting unit is configured to uniformly split both the small circular image and the large circular image in the reference image into at least two images and identify edges to form a matching image; The correction unit is configured to continuously capture multiple measured images of the target when monitoring the settlement of the target structure, and generate a correction image based on the matching image and the multiple measured images. The calculation unit is configured to calculate the pixel coordinates of the center point of the large circular image as measured coordinates based on the positions of the small circular image and the large circular image in the corrected image. A settlement unit is configured to calculate the settlement value of the target structure based on the measured coordinates and the initial coordinates.

8. The structure settlement monitoring system based on image reconstruction according to claim 7, characterized in that, The splitting unit is also configured to: Edge recognition is performed on the first circular image to identify the first edge pixels of the first circular image; the first circular image includes the small circular image and the large circular image. At the center point of the first circular image, the first circular image is horizontally and vertically cut to form four uniform first segmented images; The first edge pixel corresponding to each first segmented image is used as the matching image corresponding to that first segmented image.

9. The structure settlement monitoring system based on image reconstruction according to claim 8, characterized in that, The correction unit is also configured to: Edge recognition is performed on each of the measured images to identify the second edge pixels of each second circular image; the second circular image includes the small circular image and the large circular image; At the center point of each second circular image, the second circular image is horizontally and vertically cut to form four uniform second segmented images; Identify the similarity between the second edge pixel of each second segmented image and the matching image at the corresponding position, and select the second segmented image with the highest similarity as the selected image at that position; All selected images are stitched together to form the corrected image.

10. The structure settlement monitoring system based on image reconstruction according to claim 7, characterized in that, The computing unit is further configured to: Ellipse fitting is performed on all the small circular images and the large circular images in the corrected image, and the pixel coordinates of the center point of the ellipse are generated; The pixel coordinates of the intersection point of the lines connecting the center points of the ellipses of the two sets of small circular images located diagonally opposite each other are calculated as the first center point coordinates, and the pixel coordinates of the center point of the ellipse of the large circular image are obtained as the second center point coordinates. The measured coordinates are obtained by taking a weighted average of the coordinates of the first center point and the coordinates of the second center point.

Images