Deformation monitoring system and method based on phase information

Abstract

The invention provides a deformation monitoring system and method based on phase information, and relates to the technical field of engineering surveying, and the system comprises a shooting equipment group which collects a time sequence image of a monitoring area and transmits the time sequence image to a data processing unit, and a plurality of deformation monitoring targets and control targets which are respectively arranged on the surface and the peripheral area of a target deformation body, the data processing unit is used for extracting the center coordinates of the targets according to the time sequence images, correcting the center coordinates of the deformation monitoring targets, correcting the time sequence images after the center coordinates of the deformation monitoring targets are corrected according to the center coordinates of the control targets, and outputting the corrected time sequence images. And calculating the displacement variation of the control target according to the corrected time sequence image through a Moire fringe method, obtaining the correction coordinate of the control target, carrying out fine compensation on the position of the deformation monitoring target in the corrected time sequence image according to the correction coordinate, and determining the deformation monitoring result of the target deformation body. And real-time and high-resolution deformation monitoring is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of engineering surveying, and particularly to a deformation monitoring system and method based on phase information. Background Art

[0002] Deformation is one of the common engineering disasters, widely existing in geological disasters, bridges, tunnels, dams and other important engineering fields. It refers to the phenomenon that rock masses, soil masses or artificial structures undergo overall or local displacements under the action of external forces. In geological disasters, deformation usually appears as slope sliding, forming a slip surface. The part that has undergone displacement is called the sliding mass, and the part that has not undergone displacement is called the sliding bed, and the soft surface is between them. In projects such as bridges, tunnels and dams, deformation appears as structural settlement, crack expansion or local instability, etc. The causes of deformation include both natural factors, such as slope softening caused by rainwater infiltration, soil mass instability caused by groundwater activities, foundation weakening caused by river erosion, and structural vibration caused by earthquake effects; and human factors, such as over-excavation, artificial slope cutting, overloading and improper construction, etc.

[0003] With the development of aerospace remote sensing, Global Navigation Satellite System (GNSS), photogrammetry and computer vision technologies, deformation monitoring technologies have been significantly improved. The current main technologies include Differential Interferometric Synthetic Aperture Radar (DInSAR). This technology has the advantages of wide coverage and high precision, but it is difficult to achieve high-frequency real-time monitoring. Satellite optical remote sensing provides high-resolution images, but it is limited by satellite overpass frequencies and weather conditions, and the real-time performance is insufficient. Unmanned aerial vehicle photogrammetry and Light Detection And Ranging (LiDAR) have the characteristics of strong flexibility and high precision, but they are affected by flight time and weather and are difficult to continuously monitor. GNSS technology can achieve high-precision displacement monitoring, but the monitoring range is limited and it cannot provide high-resolution information over a large area. The efficiency of photogrammetry in large-scale monitoring has certain limitations, especially under the requirement of high-frequency data acquisition, it is easily limited by the equipment operation duration and measurement accuracy. As a technology that has developed rapidly in recent years, computer vision realizes deformation monitoring through template matching, feature point detection and optical flow analysis, with high real-time performance and strong flexibility, but it still has limitations in areas with light changes and low texture. Although these technologies have made remarkable progress in their respective fields, deformation monitoring still faces problems such as high cost, poor real-time performance and limited spatial resolution, and there is an urgent need for a more economical, efficient monitoring solution with real-time performance and high resolution. Summary of the Invention

[0004] The present invention provides a deformation monitoring system and method based on phase information, which are used to solve the defects of high cost, poor real-time performance and limited spatial resolution in deformation monitoring in the prior art, and to achieve more economical, efficient, real-time and high-resolution deformation monitoring.

[0005] The present invention provides a deformation monitoring system based on phase information, including: a group of photographing devices, a plurality of deformation monitoring targets, a plurality of control targets and a data processing unit; The deformation monitoring targets are arranged on the surface of the target deformable body; the control targets are arranged in the surrounding area of the target deformable body and serve as the spatial reference benchmark for the deformation monitoring targets; the target deformable body, the plurality of deformation monitoring targets and the plurality of control targets jointly form a monitoring area; The group of photographing devices includes a group of unmanned aerial vehicles and / or a group of cameras, which are used to collect time-series images of the monitoring area and transmit the time-series images to the data processing unit; the group of unmanned aerial vehicles includes at least one unmanned aerial vehicle, which is arranged directly above the monitoring area; the group of cameras includes at least one camera, which is arranged in the surrounding area of the monitoring area, and the camera scans to cover the monitoring area; The data processing unit is used to extract the center coordinates of the deformation monitoring targets and the control targets according to the time-series images through the maximum cross-correlation algorithm, correct the center coordinates of the deformation monitoring targets in the time-series images through the height compensation algorithm, correct the time-series images after correcting the center coordinates of the deformation monitoring targets according to the center coordinates of the control targets through the similarity transformation algorithm, calculate the displacement change amount of the control targets and obtain the corrected coordinates of the control targets according to the corrected time-series images through the moiré fringe method, finely compensate the positions of the deformation monitoring targets in the corrected time-series images according to the corrected coordinates of the control targets through the similarity transformation algorithm, and finally determine the deformation monitoring result of the target deformable body based on the finely compensated time-series images.

[0006] According to the deformation monitoring system based on phase information provided by the present invention, extracting the center coordinates of the deformation monitoring targets and the control targets according to the time-series images through the maximum cross-correlation algorithm includes: Taking the initial frame image in the time-series images as a template image, and calculating the cross-correlation function between each subsequent frame image in the time-series images and the template image; Based on the maximum correlation value in the cross-correlation function between each subsequent frame image and the template image, determining the center coordinates of the deformation monitoring targets and the control targets in each subsequent frame image.

[0007] A deformation monitoring system based on phase information provided by the present invention corrects the time series images after correcting the central coordinates of the deformation monitoring target according to the central coordinates of the control target through a similarity transformation algorithm, including: Determine a reference coordinate system based on the central coordinates of multiple control targets in the template image and a pre-determined reference control target; the reference control target is used for unified spatial reference of the control targets other than the reference control target among the multiple control targets; Calculate similarity transformation parameters based on the central coordinates of the control targets in the reference coordinate system and the central coordinates of the control targets in the time series images after correcting the central coordinates of the deformation monitoring target; Correct the time series images after correcting the central coordinates of the deformation monitoring target based on the similarity transformation parameters.

[0008] A deformation monitoring system based on phase information provided by the present invention calculates the displacement change amount of the control target and obtains the corrected coordinates of the control target according to the corrected time series images through the Moiré fringe method, including: Extract multiple frames of Moiré fringe images based on the corrected time series images; Perform low-pass filtering and downsampling processing on the multiple frames of Moiré fringe images to obtain multiple frames of downsampled images; Calculate the displacement change amount of the control target and obtain the corrected coordinates of the control target based on the Moiré fringe phases of the multiple frames of downsampled images.

[0009] In a deformation monitoring system based on phase information provided by the present invention, the number of control targets is three, and the three control targets form an L-shaped structure.

[0010] In a deformation monitoring system based on phase information provided by the present invention, the deformation monitoring target and the control target are two-dimensional rectangular planes or three-dimensional cubes; black or colored equally spaced regular dot matrices are arranged on the surfaces of the deformation monitoring target and the control target, and the dot shapes of the equally spaced regular dot matrices are polygons or circles; the printing material of the equally spaced regular dot matrices is a reflection material that is applicable day and night.

[0011] The present invention also provides a deformation monitoring method based on phase information, which is applied to any one of the deformation monitoring systems based on phase information described above, including: Collect time series images of the monitoring area; the monitoring area is jointly composed of a target deformable body, multiple deformation monitoring targets and multiple control targets; Extract the central coordinates of the deformation monitoring target and the control target according to the time series images through the maximum cross-correlation algorithm; The central coordinates of the deformation monitoring target in the time series image are corrected through a height compensation algorithm; The time series image after correcting the central coordinates of the deformed monitoring target is corrected through a similarity transformation algorithm according to the central coordinates of the control target; Through the Moiré fringe method, according to the corrected time series image, the displacement change amount of the control target is calculated and the corrected coordinates of the control target are obtained; Through the similarity transformation algorithm, according to the corrected coordinates of the control target, fine compensation is performed on the position of the deformation monitoring target in the corrected time series image; Based on the finely compensated time series image, the deformation monitoring result of the target deformable body is determined.

[0012] According to a deformation monitoring method based on phase information provided by the present invention, through the maximum cross-correlation algorithm, according to the time series image, the central coordinates of the deformation monitoring target and the control target are extracted, including: Taking the initial frame image in the time series image as a template image, and calculating the cross-correlation function between each subsequent frame image in the time series image and the template image; Based on the maximum correlation value in the cross-correlation function between each subsequent frame image and the template image, the central coordinates of the deformation monitoring target and the control target in each subsequent frame image are determined.

[0013] According to a deformation monitoring method based on phase information provided by the present invention, through the similarity transformation algorithm, according to the central coordinates of the control target, the time series image after correcting the central coordinates of the deformed monitoring target is corrected, including: Based on the central coordinates of multiple control targets in the template image and a pre-determined reference control target, a reference coordinate system is determined; the reference control target is used for unified spatial reference for the control targets other than the reference control target among the multiple control targets; Based on the central coordinates of the control targets in the reference coordinate system and the central coordinates of the control targets in the time series image after correcting the central coordinates of the deformed monitoring target, similarity transformation parameters are calculated; Based on the similarity transformation parameters, the time series image after correcting the central coordinates of the deformed monitoring target is corrected.

[0014] According to a deformation monitoring method based on phase information provided by the present invention, through the Moiré fringe method, according to the corrected time series image, the displacement change amount of the control target is calculated and the corrected coordinates of the control target are obtained, including: Based on the corrected time series image, multiple frames of Moiré fringe images are extracted; Perform low-pass filtering and downsampling on the multi-frame Moiré fringe images to obtain multi-frame downsampled images; Based on the Moiré fringe phase of the multi-frame downsampled images, calculate the displacement change of the control target and obtain the corrected coordinates of the control target.

[0015] The deformation monitoring system and method based on phase information provided by the present invention jointly form a monitoring area through a target deformable body, a plurality of deformation monitoring targets and a plurality of control targets. The imaging device group acquires time-series images of the monitoring area and transmits the time-series images to the data processing unit. The data processing unit extracts the center coordinates of each target according to the time-series images through the maximum cross-correlation algorithm, corrects the center coordinates of the deformation monitoring targets in the time-series images through the height compensation algorithm, corrects the time-series images after correcting the center coordinates of the deformation monitoring targets according to the center coordinates of the control targets through the similarity transformation algorithm, calculates the displacement change of the control target according to the corrected time-series images through the Moiré fringe method and obtains the corrected coordinates of the control target, and then finely compensates the positions of the deformation monitoring targets in the corrected time-series images according to the corrected coordinates of the control target. Finally, determine the displacement change of the deformation monitoring target according to the finely compensated time-series images to obtain the deformation monitoring result of the target deformable body, and high-precision deformation monitoring of the target deformable body can be realized at low cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 is a schematic structural diagram of the deformation monitoring system based on phase information provided by the present invention.

[0018] Figure 2 is a schematic flow chart of the deformation monitoring method based on phase information provided by the present invention.

[0019] Figure 3 is a schematic structural diagram of the deformation monitoring system provided by the present invention.

[0020] Figure 4 is a schematic diagram of the imaging principle of the reflection marker.

[0021] Figure 5 is a schematic diagram of the relationship between the deformation amount of the reflection marker in the object space and the deformation amount in the image space of the photograph provided by the present invention.

[0022] Figure 6 It is a schematic diagram of the target plane polygon and circular design provided by the present invention.

[0023] Figure 7 It is a schematic diagram of the projection of the height-compensated target target onto the reference plane provided by the present invention.

[0024] Figure 8 It is a schematic diagram of the principle of the sampling Moiré method for precise displacement measurement provided by the present invention.

[0025] Figure 9 It is a schematic diagram of the algorithm flow for calculating the deformation amount provided by the present invention. Detailed implementation manners

[0026] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present invention without creative efforts shall fall within the protection scope of the present invention.

[0027] Figure 1 It is a schematic diagram of the structure of the deformation monitoring system based on phase information provided by the present invention. As Figure 1 shown, the system includes: A shooting device group 100, a plurality of deformation monitoring targets 101, a plurality of control targets 102, and a data processing unit 103; The deformation monitoring target 101 is arranged on the surface of the target deformable body 104; the control target 102 is arranged in the peripheral area of the target deformable body 104 as the spatial reference benchmark for the deformation monitoring target 101; the target deformable body 104, the plurality of deformation monitoring targets 101, and the plurality of control targets 102 together form a monitoring area 105; The shooting device group 100 includes a drone group and / or a camera group, which is used to collect time-series images of the monitoring area 105 and transmit the time-series images to the data processing unit 103; the drone group includes at least one drone, which is arranged directly above the monitoring area; the camera group includes at least one camera, which is arranged in the peripheral area of the monitoring area, and the camera scans to cover the monitoring area; The data processing unit 103 is configured to extract the central coordinates of the deformation monitoring target and the control target from the time series images according to the maximum cross - correlation algorithm, correct the central coordinates of the deformation monitoring target in the time series images through the height compensation algorithm, correct the time series images after correcting the central coordinates of the deformation monitoring target according to the central coordinates of the control target through the similarity transformation algorithm, calculate the displacement change amount of the control target and obtain the corrected coordinates of the control target according to the corrected time series images through the moiré fringe method, finely compensate the position of the deformation monitoring target in the time series images after correction according to the corrected coordinates of the control target, and finally determine the deformation monitoring result of the target deformable body based on the finely compensated time series images.

[0028] Specifically, the target deformable body is a deformable body whose deformation degree needs to be monitored. The deformation monitoring system based on phase information provided by the present invention includes a shooting device group, a plurality of deformation monitoring targets, a plurality of control targets, and a data processing unit, which cooperate with each other to obtain the deformation monitoring result of the target deformable body.

[0029] In the embodiment of the present invention, a plurality of deformation monitoring targets are arranged on the surface of the target deformable body. The deformation monitoring targets have certain geometric shapes and reflection characteristics, and can obtain high - precision image data through the shooting device group. The position change of the deformation monitoring target can reflect the deformation process of the target deformable body. The number of deformation monitoring targets can be determined according to the area of the target deformable body and the complexity of the internal movement.

[0030] In the embodiment of the present invention, a plurality of control targets are arranged in the peripheral area of the target deformable body, mainly used to provide a spatial reference benchmark for the deformation monitoring targets. It can be understood that the position where the control targets are arranged should be stable, so that the positions of the control targets can remain fixed in the images, and the relative positions and movement trajectories of the deformation monitoring targets can be accurately calibrated through the perspectives of the cameras and the drone group.

[0031] Optionally, the number of control targets is three, and the three control targets form an L - shaped structure.

[0032] Specifically, the control targets can be placed in a stable area directly in front of the cameras. Taking the middle control target as the coordinate origin, the two control targets on both sides are respectively arranged in the extension direction of the middle control target, forming an orthogonal structural layout with the middle control target. The three control targets form an "L" - shaped framework.

[0033] Optionally, the deformation monitoring target and the control target are two-dimensional rectangular planes or three-dimensional cubes; the surfaces of the deformation monitoring target and the control target are provided with equally spaced regular dot matrices in black or color, and the dot shapes of the equally spaced regular dot matrices are polygons or circles; the printing material of the equally spaced regular dot matrix is a reflection material that is common day and night.

[0034] The monitoring area is jointly composed of the target deformable body, the deformation monitoring target and the control target. This area is the area that can be covered by the lenses of the shooting device group, and all targets (deformation monitoring targets and control targets) should be arranged within this area to ensure that the shooting device group can completely capture the changes of the deformable body.

[0035] The shooting device group can be composed of a drone group, or the shooting device group can be composed of a camera group, or the shooting device group can be composed of a drone group and a camera group, which is used to obtain the time series images of the monitoring area and transmit the time series images to the data processing unit.

[0036] It can be set that the shooting device group synchronously collects images at a preset time interval, and the time interval can be flexibly adjusted according to the sliding speed and danger level of the target deformable body. When multiple cameras or drones perform three-dimensional space deformation monitoring, photo or video synchronization can be achieved through time tags.

[0037] The drone group includes at least one drone, which is set directly above the monitoring area. The drone provides an aerial shooting perspective and can efficiently cover the entire monitoring area. The images collected by the drone can obtain the spatial information of the deformation area in real time, especially having advantages for the monitoring of large-scale or irregular areas.

[0038] The camera group includes at least one camera, which is arranged in the peripheral area of the monitoring area. The camera provides a multi-angle ground perspective and can scan and cover the monitoring area, and can provide information in different directions and depths, which helps to perform precise three-dimensional reconstruction of the target area. It can be understood that the camera in the present invention can also be understood as a camera, a video camera, etc., and the present invention does not limit this.

[0039] It can be understood that the shooting device group can include one camera, or multiple cameras, or one drone, or multiple drones, or one camera plus one drone, or one camera plus multiple drones, etc.

[0040] The data processing unit is the core part, which is responsible for processing the time series images transmitted by the shooting device group, and performing deformation monitoring analysis through the maximum cross-correlation algorithm, height compensation algorithm, similarity transformation algorithm and moiré fringe method, and finally outputting the deformation monitoring results.

[0041] Among them, the maximum cross-correlation algorithm is used to extract the center coordinate trajectories of each target (including deformation monitoring targets and control targets). The maximum cross-correlation algorithm can be used to compare images in a continuous time series to extract the position changes of the targets in the images. The position of the target in each frame of the image is obtained through cross-correlation calculation with the template image, so as to obtain the center coordinate trajectory of the target.

[0042] Through the height compensation algorithm, the deviation of the image coordinate transformation caused by the height difference (vertical displacement) of the deformation monitoring target relative to the reference plane can be corrected. Adjust the position of the deformation monitoring target in the image coordinate system, eliminate the geometric distortion caused by the vertical displacement, ensure the accuracy of the coordinates of the deformation monitoring target in the image, and thus improve the accuracy of the similarity transformation matrix when correcting the coordinates of the deformation monitoring target.

[0043] In the process of correcting the center coordinates of the deformation monitoring target in the time series image through the height compensation algorithm, the projected center coordinates of the deformation monitoring target on the reference plane after correction can be determined based on the center coordinates of multiple control targets in the time series image and the height difference (vertical displacement) of the deformation monitoring target relative to the reference plane.

[0044] Specifically, the height difference (vertical displacement) of the target object relative to the reference plane will cause the projection position of the target point in the image to shift. In order to eliminate the image coordinate shift caused by the height difference, the coordinates of the target point are height-compensated to make it more accurate when calculating the similarity transformation. The compensated coordinates can be calculated by the following formula:

[0045] Among them, is the height difference, is the distance from the camera to the reference plane, is the coordinate of the target point in the image, is the coordinate of the target point in the image. is the coordinate of the reference point in the image, is the coordinate of the reference point in the image.

[0046] The similarity transformation algorithm is used to correct the time series image to eliminate the errors caused by the movement of the shooting device (such as rotation, scaling, translation). The change in the center position of the control target in each frame of the image can be calculated through the similarity transformation algorithm, that is, based on the center coordinates of the control target, compensation is made using the geometric transformation (rotation, translation, scaling) of the image, so as to obtain the accurate target trajectory in the unified coordinate system.

[0047] The Moiré fringe method can be used to calculate the displacement change of the control target based on the corrected time-series images and correct the coordinates of the control target. Through the Moiré fringe method, the fringe information of the control target can be extracted from the images, the phase difference can be calculated, and thus the displacement change of the control target can be deduced. The deformed object will cause changes in the fringe pattern, and the displacement can be accurately calculated using the phase difference of the fringes, which is especially suitable for precise displacement measurement.

[0048] Through the Moiré fringe method, the displacement change of the control target can be calculated based on the corrected time-series images and the coordinates of the control target can be corrected to obtain the corrected coordinates of the control target. Then, using the similarity transformation algorithm, the position of the deformation monitoring target in the corrected time-series images (i.e., the time-series images processed by the first similarity transformation) can be finely compensated according to the corrected coordinates of the control target. Finally, based on the finely compensated time-series images and combined with the displacement change of the baseline between the deformation monitoring target and the control target, the displacement change of the deformation monitoring target can be calculated.

[0049] Finally, the data processing unit can determine the deformation monitoring result based on the calculated displacement change of the deformation monitoring target. By comprehensively analyzing the displacement change of the deformation monitoring target, the data processing unit can output the deformation monitoring result of the target deformed body in real time. Combining the spatial position and deformation process of the deformation monitoring target, the overall deformation situation of the target deformed body can be monitored, analyzed, and predicted.

[0050] The deformation monitoring system based on phase information provided by the present invention forms a monitoring area through the target deformed body, multiple deformation monitoring targets, and multiple control targets. The imaging device group acquires the time-series images of the monitoring area and transmits the time-series images to the data processing unit. The data processing unit extracts the center coordinates of each target from the time-series images through the maximum cross-correlation algorithm, corrects the center coordinates of the deformation monitoring targets in the time-series images through the height compensation algorithm, corrects the time-series images after correcting the center coordinates of the deformation monitoring targets according to the center coordinates of the control targets through the similarity transformation algorithm, calculates the displacement change of the control target based on the corrected time-series images through the Moiré fringe method and obtains the corrected coordinates of the control target, then finely compensates the position of the deformation monitoring target in the corrected time-series images according to the corrected coordinates of the control target, and finally calculates the displacement change of the deformation monitoring target based on the finely compensated time-series images to obtain the deformation monitoring result of the target deformed body, which can achieve high-precision deformation monitoring of the target deformed body at low cost.

[0051] According to a deformation monitoring system based on phase information provided by the present invention, the center coordinates of the deformation monitoring target and the control target are extracted from the time-series images through the maximum cross-correlation algorithm, including: Take the initial frame image in the time - series image as the template image, and calculate the cross - correlation function between each subsequent frame image in the time - series image and the template image; Based on the maximum correlation value in the cross - correlation function between each subsequent frame image and the template image, determine the central coordinates of the deformation monitoring target and the control target in each subsequent frame image.

[0052] Specifically, the continuous time - series images contain the image features of the deformation monitoring target and the control target. These images are arranged in chronological order, and each frame image corresponds to the deformation monitoring target and the control target images at different time points. By analyzing these images, the movement trajectories of the deformation monitoring target and the control target at different time points can be traced.

[0053] In the process of extracting the central coordinates of the deformation monitoring target and the control target from the time - series images, an initial frame image can be selected as the template image , usually this image contains obvious feature points or stripe patterns for subsequent image matching. The template image should have sufficient identification features to ensure the accuracy of the matching process.

[0054] Through the maximum cross - correlation algorithm, subsequent frame images in the time - series image can be used as input images , and compared with the template image to calculate the cross - correlation function . The cross - correlation function is used to measure the similarity between the template image and the input image, and the positions of the centers of the deformation monitoring target and the control target are determined by finding the position of the maximum correlation value.

[0055] The cross - correlation function can be calculated by Fourier transform, and the specific steps are as follows: For the given template image and the input image , perform two - dimensional discrete Fourier transform to obtain and respectively:

[0056] The cross - correlation function is calculated by inverse Fourier transform:

[0057] where is the complex conjugate of , and and are the dimensions of the image.

[0058] To simplify the optimization objective, facilitate calculation and find the optimal solution, find the cross - correlation function Maximized displacement For each frame of image, define the objective function The formula is as follows:

[0059] Calculate the constant factor :

[0060] The displacement parameter determined by maximizing cross - correlation and , can accurately record the central coordinates of the deformation monitoring target and the control target in each frame of image. The obtained central coordinates of the deformation monitoring target and the control target can be expressed as where represents a certain specific marker, represents the video frame number.

[0061] According to a deformation monitoring system based on phase information provided by the present invention, the time - series images after correcting the central coordinates of the deformation monitoring target are corrected according to the central coordinates of the control target through a similarity transformation algorithm, including: Based on the central coordinates of multiple control targets in the template image and a pre - determined reference control target, determine the reference coordinate system; the reference control target is used for unified spatial reference of the control targets other than the reference control target among the multiple control targets; Based on the central coordinates of the control targets in the reference coordinate system and the central coordinates of the control targets in the time - series images after correcting the central coordinates of the deformation monitoring target, calculate the similarity transformation parameters; Based on the similarity transformation parameters, correct the time - series images after correcting the central coordinates of the deformation monitoring target.

[0062] Specifically, the time - series images are corrected through the similarity transformation algorithm, aiming to eliminate the motion errors (such as rotation, scaling, translation, etc.) that may be caused by the camera or other devices during the shooting process, thereby improving the accuracy of deformation monitoring. The similarity transformation algorithm can adjust the target coordinates in the image to make them consistent with the reference coordinate system and compensate for the motion effects of the device.

[0063] The following similarity transformation formula can be used to compensate the marked center position:

[0064] where, is the scaling factor, is the rotation angle, are the translation components in the x and y directions respectively. is the point position in the transformed coordinate system; It is the point position in the original coordinate system.

[0065] First, a reference coordinate system needs to be selected as a benchmark. In the embodiments of the present invention, the reference coordinate system can be determined according to the central coordinates of multiple control targets in the template image and a pre-determined reference control target. The pre-determined reference control target is used for unified spatial reference for all control targets except this reference control target among multiple control targets, so as to construct a complete monitoring target network.

[0066] In some embodiments, the number of control targets is three or more, and these control targets form an L-shaped structure. Then, the control target at the intersection of the L-shaped structure can be set as the reference control target, used as the unified spatial reference for other control targets, construct a complete monitoring target network, and determine the reference coordinate system.

[0067] For the central coordinates of the control targets in each frame of image, the similarity transformation parameters (i.e., scaling factor, rotation angle, and translation component) can be calculated by comparing the position of the control target in the current frame of image and the position of the target in the reference coordinate system.

[0068] For example, the rotation angle can be obtained by calculating the angle difference between the coordinates of the control target in the current frame of image and the coordinates of the target in the reference coordinate system ; the scaling factor can be calculated by comparing the actual size of the control target in the current frame of image with the size of the target in the reference coordinate system ; the translation component can be obtained by calculating the gap between the center of the control target in the reference coordinate system and the current frame of image .

[0069] Then, through the similarity transformation parameters calculated above, the coordinates in the original image can be subjected to similarity transformation to obtain the corrected coordinates, thus completing the correction process of the time series image after correcting the central coordinates of the deformed monitoring target.

[0070] Through these steps, the error caused by the movement of the device can be effectively eliminated, a more accurate target displacement trajectory can be obtained, and thus the accuracy of deformation monitoring can be improved.

[0071] According to a deformation monitoring system based on phase information provided by the present invention, the displacement change amount of the control target is calculated according to the corrected time series image by the Moiré fringe method and the corrected coordinates of the control target are obtained, including: Based on the corrected time series image, multiple frames of Moiré fringe images are extracted; The multiple frames of Moiré fringe images are subjected to low-pass filtering and downsampling processing to obtain multiple frames of downsampled images; Based on the Moiré fringe phase of the multiple frames of downsampled images, the displacement change amount of the control target is calculated and the corrected coordinates of the control target are obtained.

[0072] Specifically, on the target surface, Moiré fringes are generated by optical interference phenomena, and deformation will cause changes in the fringes. In order to further improve the high-precision sub-millimeter displacement accuracy in the plane, the Moiré fringe method is adopted in the embodiments of the present invention. This method extracts the phase difference of the Moiré fringes before and after deformation through low-pass filtering, downsampling, and phase analysis, and calculates the plane displacement.

[0073] In the original Moiré fringe image, the intensity distribution of the Moiré fringes can be expressed by the following formula:

[0074] where is the amplitude of the fringe intensity, is the background intensity, is the fringe period (in pixels), is the initial phase. In the experiment, the fringe period is usually the same in the x and y directions, that is, a unified period parameter.

[0075] After extracting multiple frames of Moiré fringe images, the original Moiré fringe image can be subjected to low-pass filtering and downsampling processing. The original Moiré fringe image (represented by its intensity distribution ) may contain noise and high-frequency interference, such as uneven illumination, complex background, etc.

[0076] To extract the main components of the fringes, first, the Moiré fringe image is processed by low-pass filtering to remove high-frequency noise and retain the low-frequency main information of the fringes, obtaining the low-pass filtered Moiré fringe image, whose intensity distribution can be represented by .

[0077]

[0078] where and represent the Fourier transform and the inverse Fourier transform respectively, is the transfer function of the low-pass filter. By using a Gaussian low-pass filter:

[0079] In the formula, is the cut-off frequency.

[0080] The low-pass filtered Moiré fringe image (represented by its intensity distribution ) enters the downsampling process. The purpose of downsampling is to obtain the phase distribution using multiple sampling points, and the specific formula is:

[0081] is the integer sampling points, is the phase distribution of the Moiré fringe.

[0082] After obtaining the downsampled image, the phase distribution of the Moiré fringe can be extracted by using the phase-shifting technique through the downsampled image. The specific calculation formula is:

[0083] Calculation of the phase difference and the planar displacement. In the images before and after the target deformation, the phase difference of the Moiré fringe is defined as:

[0084] is the phase distribution of the deformed Moiré fringe, is the phase distribution of the Moiré fringe before deformation.

[0085] Finally, the displacement change amount in the plane can be directly calculated through the phase difference :

[0086] High-precision measurement of planar displacement can be achieved by sampling the Moiré fringe method, which can provide an accuracy of 1 / 100 pixel and 1 / 1000 fringe period. The high-frequency noise of the fringe image can be removed by low-pass filtering to extract the fringe information. By using multiple frames of downsampled fringe images, the periodic phase information can be extracted, and the displacement change amount of the fringe before and after deformation can be calculated through the phase difference formula, and then the correction coordinates of the control target can be obtained. It should be noted that if the target is three-dimensional, or the deformation of the deformable body has a height component, etc., the three-dimensional displacement of the target can be calculated by combining the spatial position of the target (such as through the measurement of cameras and drones) :

[0087] , and are the displacements along the three directions of X, Y, and Z respectively, which are calculated by this algorithm.

[0088] The deformation monitoring method based on phase information provided by the present invention will be described below. The deformation monitoring method based on phase information described below can be mutually corresponding and referred to the deformation monitoring system based on phase information described above.

[0089] Figure 2 is the flow schematic diagram of the deformation monitoring method based on phase information provided by the present invention. As Figure 2 shown, this method is applied to any deformation monitoring system based on phase information as described in the previous system embodiment, and includes the following steps: Step 200: Collect time - series images of the monitoring area; the monitoring area is jointly composed of a target deformable body, multiple deformation monitoring targets, and multiple control targets.

[0090] Step 201: According to the time - series images, extract the central coordinates of the deformation monitoring targets and control targets through the maximum cross - correlation algorithm.

[0091] Step 202: Correct the central coordinates of the deformation monitoring targets in the time - series images through the height compensation algorithm.

[0092] Step 203: According to the central coordinates of the control targets, correct the time - series images after correcting the central coordinates of the deformation monitoring targets through the similarity transformation algorithm.

[0093] Step 204: According to the corrected time - series images, calculate the displacement change of the control targets and obtain the corrected coordinates of the control targets through the moiré fringe method.

[0094] Step 205: According to the corrected coordinates of the control targets, perform fine compensation on the positions of the deformation monitoring targets in the corrected time - series images through the similarity transformation algorithm.

[0095] Step 206: Based on the time - series images after fine compensation, determine the deformation monitoring results of the target deformable body.

[0096] Specifically, the target deformable body is the deformable body whose deformation degree needs to be monitored. The deformation monitoring system based on phase information provided by the present invention includes a group of photographing devices, multiple deformation monitoring targets, multiple control targets, and a data processing unit, which cooperate with each other to obtain the deformation monitoring results of the target deformable body.

[0097] In the embodiments of the present invention, multiple deformation monitoring targets are provided, and the multiple deformation monitoring targets are arranged on the surface of the target deformable body. The deformation monitoring targets have certain geometric shapes and reflection characteristics, and can obtain high - precision image data through the group of photographing devices. The position change of the deformation monitoring targets can reflect the deformation process of the target deformable body. The number of deformation monitoring targets can be determined according to the area of the target deformable body and the complexity of the internal movement.

[0098] In the embodiments of the present invention, multiple control targets are provided, and the multiple control targets are set in the peripheral area of the target deformable body, mainly used to provide a spatial reference benchmark for the deformation monitoring targets. It can be understood that the positions where the control targets are set should be stable, so that the positions of the control targets can remain fixed in the images, and the relative positions and movement trajectories of the deformation monitoring targets can be accurately calibrated through the perspectives of the cameras and the unmanned aerial vehicle group.

[0099] The monitoring area is jointly composed of the target deformable body, deformation monitoring targets, and control targets. This area is the area that can be covered by the lens of the camera group. All targets (deformation monitoring targets and control targets) should be arranged within this area to ensure that the camera group can completely capture the changes of the deformable body.

[0100] First, the deformation monitoring system based on phase information needs to obtain the time-series images of the monitoring area. Images can be synchronously acquired at a preset time interval, which can be flexibly adjusted according to the sliding speed and risk level of the target deformable body. When performing three-dimensional space deformation monitoring through multiple cameras or drones, photo or video synchronization can be achieved through time tags.

[0101] After obtaining the time-series images, the time-series images can be processed. Through the maximum cross-correlation algorithm, height compensation algorithm, similarity transformation algorithm, and moiré fringe method, deformation monitoring analysis is carried out, and finally the deformation monitoring results are output.

[0102] Among them, the maximum cross-correlation algorithm is used to extract the center coordinate trajectories of each target (including deformation monitoring targets and control targets). The maximum cross-correlation algorithm can be used to compare the images of consecutive time series, and extract the position changes of the targets in the images. The position of the target in each frame of the image is obtained through the cross-correlation calculation with the template image, so as to obtain the center coordinate trajectory of the target.

[0103] Through the height compensation algorithm, the image coordinate transformation deviation caused by the height difference (vertical displacement) of the deformation monitoring target relative to the reference plane can be corrected. Adjust the position of the deformation monitoring target in the image coordinate system, eliminate the geometric distortion caused by the vertical displacement, ensure the accuracy of the coordinates of the deformation monitoring target in the image, and thus improve the accuracy of the similarity transformation matrix when correcting the coordinates of the deformation monitoring target.

[0104] During the process of correcting the center coordinates of the deformation monitoring targets in the time-series images through the height compensation algorithm, based on the center coordinates of multiple control targets in the time-series images and the height difference (vertical displacement) of the deformation monitoring targets relative to the reference plane, the projected center coordinates of the corrected deformation monitoring targets on the reference plane can be determined.

[0105] Specifically, the height difference (vertical displacement) of the target object relative to the reference plane will cause the projection position of the target point in the image to shift. In order to eliminate the image coordinate shift caused by the height difference, the coordinates of the target point are height-compensated to make it more accurate when calculating the similarity transformation. The compensated coordinates can be calculated by the following formula:

[0106] where is the height difference, is the distance from the camera to the reference plane, are the coordinates of the target point in the image, are the coordinates of the target point in the image. are the coordinates of the reference point in the image, are the coordinates of the reference point in the image.

[0107] The similarity transformation algorithm is used to correct the time - series images, eliminating the errors caused by the movement of the shooting device (such as rotation, scaling, translation). The change in the position of the center of the control target in each frame of the image can be calculated through the similarity transformation algorithm, that is, compensation is performed using the geometric transformation (rotation, translation, scaling) of the image according to the center coordinates of the control target, so as to obtain the accurate target trajectory under the unified coordinate system.

[0108] The Moiré fringe method can be used to calculate the displacement change of the control target based on the corrected time - series images and correct the coordinates of the control target. Through the Moiré fringe method, the fringe information of the control target can be extracted from the image, the phase difference can be calculated, and thus the displacement change of the control target can be deduced. The deformed object will cause changes in the fringe pattern, and the displacement can be accurately calculated using the phase difference of the fringes, which is especially suitable for precise displacement measurement.

[0109] Through the Moiré fringe method, the displacement change of the control target can be calculated based on the corrected time - series images and the coordinates of the control target can be corrected to obtain the corrected coordinates of the control target. Then, the similarity transformation algorithm is used to finely compensate the position of the deformation monitoring target in the corrected time - series images (i.e., the time - series images processed by the first similarity transformation) according to the corrected coordinates of the control target. Finally, based on the finely compensated time - series images, combined with the displacement change of the baseline between the deformation monitoring target and the control target, the displacement change of the deformation monitoring target can be calculated.

[0110] Finally, based on the calculated displacement change of the deformation monitoring target, the deformation monitoring result can be determined. By comprehensively analyzing the displacement change of the deformation monitoring target, the deformation monitoring result of the target deformable body can be output in real - time. Combining the spatial position and deformation process of the deformation monitoring target, the overall deformation situation of the target deformable body can be monitored, analyzed and predicted.

[0111] The deformation monitoring method based on phase information provided by the present invention forms a monitoring area jointly by a target deformable body, a plurality of deformation monitoring targets and a plurality of control targets. An image acquisition device group acquires time-series images of the monitoring area and transmits the time-series images to a data processing unit. The data processing unit extracts the center coordinates of each target through a maximum cross-correlation algorithm according to the time-series images, corrects the center coordinates of the deformation monitoring targets in the time-series images through a height compensation algorithm, corrects the time-series images after correcting the center coordinates of the deformation monitoring targets according to the center coordinates of the control targets through a similarity transformation algorithm, calculates the displacement change amount of the control targets and obtains the corrected coordinates of the control targets through a Moiré fringe method according to the corrected time-series images, then finely compensates the positions of the deformation monitoring targets in the time-series images after correction according to the corrected coordinates of the control targets to obtain the finely compensated time-series images, and finally calculates the displacement change amount of the deformation monitoring targets according to the finely compensated time-series images to obtain the deformation monitoring result of the target deformable body, and high-precision deformation monitoring of the target deformable body can be realized at low cost.

[0112] A deformation monitoring method based on phase information provided by the present invention extracts the center coordinates of deformation monitoring targets and control targets according to time-series images through a maximum cross-correlation algorithm, including: Taking the initial frame image in the time-series images as a template image, and calculating the cross-correlation function between each subsequent frame image in the time-series images and the template image; Based on the maximum correlation value in the cross-correlation function between each subsequent frame image and the template image, determining the center coordinates of the deformation monitoring targets and control targets in each subsequent frame image.

[0113] Specifically, consecutive time-series images contain the image features of deformation monitoring targets and control targets. These images are arranged in chronological order, and each frame image corresponds to the images of deformation monitoring targets and control targets at different time points. By analyzing these images, the movement trajectories of deformation monitoring targets and control targets at different time points can be traced.

[0114] During the process of extracting the center coordinates of deformation monitoring targets and control targets according to time-series images, the initial frame image can be selected as the template image , usually this image contains obvious feature points or stripe patterns for subsequent image matching. The template image should have sufficient identification features to ensure the accuracy of the matching process.

[0115] Through the maximum cross-correlation algorithm, the subsequent frame images in the time-series images can be used as input images , and compared with the template image to calculate the cross-correlation function . The cross - correlation function is used to measure the similarity between the template image and the input image, and determines the positions of the deformation monitoring target and the control target center by finding the position of the maximum correlation value.

[0116] Cross - correlation function can be calculated through Fourier transform, and the specific steps are as follows: For a given template image and the input image , perform two - dimensional discrete Fourier transform to obtain and respectively:

[0117] Cross - correlation function is calculated through inverse Fourier transform:

[0118] where is 's complex conjugate, and are the dimensions of the image.

[0119] To simplify the optimization objective, facilitate calculation and find the optimal solution, find the displacement that maximizes the cross - correlation function . For each frame of image, define the objective function The formula is as follows:

[0120] Calculate the constant factor :

[0121] The displacement parameters and determined by cross - correlation maximization can accurately record the center coordinates of the deformation monitoring target and the control target in each frame of image. The obtained center coordinates of the deformation monitoring target and the control target can be expressed as where represents a specific marker, represents the video frame number.

[0122] According to a deformation monitoring method based on phase information provided by the present invention, through a similarity transformation algorithm, according to the center coordinates of the control target, the time - series images after correcting the center coordinates of the deformation monitoring target are corrected, including: Determine a reference coordinate system based on the central coordinates of multiple control targets in a template image and a pre-determined reference control target; the reference control target is used for unified spatial reference of control targets other than the reference control target among the multiple control targets; Calculate similarity transformation parameters based on the central coordinates of the control targets in the reference coordinate system and the central coordinates of the control targets in the time-series images after correcting the central coordinates of the deformation monitoring targets; Correct the time-series images after correcting the central coordinates of the deformation monitoring targets based on the similarity transformation parameters.

[0123] Specifically, correct the time-series images through a similarity transformation algorithm, aiming to eliminate motion errors (such as rotation, scaling, translation, etc.) that may be caused by the camera or other devices during the shooting process, thereby improving the accuracy of deformation monitoring. The similarity transformation algorithm can adjust the target coordinates in the image to make them consistent with the reference coordinate system and compensate for the motion effects of the devices.

[0124] The following similarity transformation formula can be used to mark the central position for compensation:

[0125] Among them, is the scaling factor, is the rotation angle, are the translation components in the x and y directions respectively. is the point position in the transformed coordinate system; is the point position in the original coordinate system.

[0126] First, a reference coordinate system needs to be selected as the benchmark. In the embodiments of the present invention, a reference coordinate system can be determined according to the central coordinates of multiple control targets in a template image and a pre-determined reference control target. The pre-determined reference control target is used for unified spatial reference of all control targets other than the reference control target among the multiple control targets, thereby constructing a complete monitoring target network.

[0127] In some embodiments, the number of control targets is three or more, and these control targets form an L-shaped structure. Then, the control target at the intersection of the L-shaped structure can be set as the reference control target for unified spatial reference of other control targets, construct a complete monitoring target network, and determine the reference coordinate system.

[0128] For the central coordinates of the control targets in each frame of the image, the similarity transformation parameters (i.e., the scaling factor, rotation angle, and translation components) can be calculated by comparing the position of the control target in the current frame of the image with the position of the target in the reference coordinate system.

[0129] For example, the rotation angle can be obtained by calculating the angular difference between the coordinates of the control target in the current frame image and the coordinates of the target in the reference coordinate system. The scaling factor can be calculated by comparing the actual size of the control target in the current frame image with the size of the target in the reference coordinate system. The translation component can be obtained by calculating the difference between the center of the control target in the reference coordinate system and the current frame image. .

[0130] Then, through the similarity transformation parameters calculated above, the coordinates in the original image can be subjected to similarity transformation to obtain the corrected coordinates, thereby completing the correction process of the time series image after correcting the central coordinates of the deformation monitoring target.

[0131] Through these steps, the errors caused by equipment movement can be effectively eliminated, a more accurate target displacement trajectory can be obtained, and thus the accuracy of deformation monitoring can be improved.

[0132] According to a deformation monitoring method based on phase information provided by the present invention, through the Moiré fringe method, based on the corrected time series image, the displacement change amount of the control target is calculated and the corrected coordinates of the control target are obtained, including: Based on the corrected time series image, multiple frames of Moiré fringe images are extracted; The multiple frames of Moiré fringe images are subjected to low-pass filtering and downsampling processing to obtain multiple frames of downsampled images; Based on the Moiré fringe phase of the multiple frames of downsampled images, the displacement change amount of the control target is calculated and the corrected coordinates of the control target are obtained.

[0133] Specifically, on the surface of the target, Moiré fringes are generated by optical interference phenomena, and deformation will cause changes in the fringes. In order to further improve the high-precision sub-millimeter displacement accuracy in the plane, the Moiré fringe method is adopted in the embodiments of the present invention. This method extracts the phase difference of the Moiré fringes before and after deformation through low-pass filtering, downsampling and phase analysis, and calculates the plane displacement.

[0134] In the original Moiré fringe image, the intensity distribution of the Moiré fringes can be expressed by the following formula:

[0135] where is the amplitude of the fringe intensity, is the background intensity, is the fringe period (in pixels), is the initial phase. In the experiment, the fringe period is usually the same in the x and y directions, that is, a unified period parameter.

[0136] After extracting multiple frames of moiré fringe images, the original moiré fringe images can be subjected to low-pass filtering and downsampling. The original moiré fringe images (represented by their intensity distributions may contain noise and high-frequency interference, such as uneven illumination, complex backgrounds, etc.

[0137] To extract the main components of the fringes, the moiré fringe images are first processed by low-pass filtering to remove high-frequency noise and retain the main low-frequency information of the fringes, obtaining the low-pass filtered moiré fringe images, whose intensity distributions can be represented by where

[0138]

[0139] where and represent the Fourier transform and the inverse Fourier transform respectively, and is the transfer function of the low-pass filter. By using a Gaussian low-pass filter:

[0140] where is the cut-off frequency.

[0141] The moiré fringe images after low-pass filtering (represented by their intensity distributions enter the downsampling process. The purpose of downsampling is to obtain the phase distribution using multiple sampling points, and the specific formula is:

[0142] is the number of integer sampling points, and is the phase distribution of the moiré fringes.

[0143] After obtaining the downsampled images, the phase distribution of the moiré fringes can be extracted using the phase-shifting technique through the downsampled images, and the specific calculation formula is:

[0144] Calculation of the phase difference and the in-plane displacement. In the images before and after the target deformation, the phase difference of the moiré fringes is defined as:

[0145] is the phase distribution of the moiré fringes after deformation, and is the phase distribution of the moiré fringes before deformation.

[0146] Finally, the change in the in-plane displacement can be directly calculated through the phase difference:

[0147] High-precision measurement of planar displacement can be achieved by sampling the Moiré fringe method, providing an accuracy of 1 / 100 pixel and 1 / 1000 fringe period. High-frequency noise in the fringe image can be removed by low-pass filtering to extract fringe information. By using multi-frame downsampled fringe images, periodic phase information is extracted, and the displacement change of the fringes before and after deformation is calculated through the phase difference formula, thereby obtaining the correction coordinates of the control target.

[0148] It should be noted that if the target is three-dimensional, or the deformation of the deformable body has a height component, etc., the three-dimensional displacement of the target can be calculated by combining the spatial position of the target (such as through the measurement of cameras and drones). :

[0149] , and are the displacements along the three directions of X, Y, and Z respectively, calculated by this algorithm.

[0150] The deformation monitoring system and method based on phase information provided by the present invention are further explained below through examples in specific application scenarios.

[0151] In this embodiment, technologies such as target tracking, similarity transformation, and Moiré fringes are adopted to construct a comprehensive algorithm for high-precision three-dimensional deformation monitoring, achieving high-precision real-time monitoring of deformation at low cost. The following combines the attached Figures 3 - 9 to detail the specific implementation manner of this embodiment.

[0152] I. System design: Figure 3 is a schematic structural diagram of the deformation monitoring system provided by the present invention. As Figure 3 shown, the deformation monitoring system of this embodiment mainly consists of the following parts (where 110 represents the main direction of landslide movement): 1. Camera The camera is a high-resolution monitoring camera or a high-definition camera, with panchromatic or color imaging. The shooting range of the camera includes all control points and targets for deformation monitoring, such as Figure 3The shooting range of the medium and high-resolution monitoring camera 1 (310) is between 311 and 313, and the shooting range of the high-resolution monitoring camera 2 (320) is between 321 and 323; there are no less than two control targets for single-camera monitoring points, and no less than three control targets for double-camera monitoring points; the main optical axes of the double cameras are arranged approximately orthogonally. For example, the main optical axis 312 of the high-resolution monitoring camera 1 (310) in the figure is approximately orthogonal to the main optical axis 322 of the high-resolution monitoring camera 2 (320), and the camera focal length is infinity. When two or more cameras perform three-dimensional deformation monitoring, photo or video synchronization is achieved through time tags, and the image data is transmitted to the control center in real time.

[0153] In this embodiment, high-resolution dual monitoring cameras are used. The high-resolution monitoring cameras are respectively arranged in the peripheral area of the deformed body, and the main optical axes are arranged approximately orthogonally, such as Figure 3 the orthogonal main optical axis (510). The camera selects a 24mm equivalent focal length lens, and the imaging unit has a 4k resolution, which can achieve high-precision image data acquisition.

[0154] 2. UAV monitoring module The UAV adopts a lightweight design and has the ability to take high-definition images. The UAV hovers above the measurement area and covers the entire monitoring area through orthophoto image acquisition, such as Figure 3 the acquisition range of the UAV (410) in the figure can be represented by 411, 412, and 413. The UAV can independently perform two-dimensional displacement monitoring or cooperate with ground cameras to achieve integrated air-ground three-dimensional monitoring.

[0155] 3. Deformation monitoring targets and control targets The deformation monitoring targets are arranged on the surface of the deformed body, and the number is determined according to the deformation characteristics of the deformed body and the required spatial resolution. The monitoring board adopts a two-dimensional plane design, and the targets adopt a two-dimensional rectangle or three-dimensional cube design. Regular dot matrices with equal spacing in black or color are set on the surface. The dot shape is polygonal or circular, and the printing material of the dots can adopt a reflective material that is applicable day and night. The deformation monitoring targets and control targets adopt a single-sided or double-sided design in two dimensions and a three-dimensional layout in three dimensions, and are directly drawn on the surface of the monitoring object. The same pattern design is used for each surface. All deformation monitoring targets are arranged within the overlapping shooting areas of cameras, cameras, and UAVs to ensure complete coverage of the monitoring area and high-precision displacement measurement.

[0156] In this embodiment, multiple three-dimensional deformation monitoring targets are arranged on the surface of the deformed body, such as Figure 3 the deformation monitoring target 1 (210), deformation monitoring target 2 (211), deformation monitoring target 3 (212), and deformation monitoring target 4 (213) in the figure. Regular dot matrices with equal spacing in black or color are set on the surface. The dot shape is polygonal or circular, and the printing material of the dots can adopt a reflective material that is applicable day and night. The target spacing is set to 100mm.

[0157] The control target is arranged in the area outside the deformable body, used to correct the imaging deviation between the camera and the deformation monitoring target, and provide a reliable reference for deformation calculation.

[0158] The control target can be placed in the stable area directly in front of the camera. Taking the middle control target as the coordinate origin, the control targets on both sides are respectively arranged in the extension direction of the middle control target, forming an orthogonal structural layout with the middle control target. The three control targets form an "L"-shaped frame, such as Figure 3 the control target 1 (220), control target 2 (221) and control target 3 (222) in

[0159] Each camera, camera head and drone can achieve precise correction of the deformation monitoring trajectory through three control targets, and at the same time ensure that each device can independently complete the monitoring of displacement changes in the plane. The collaborative work of dual cameras or dual camera heads can achieve comprehensive monitoring of three-dimensional displacements.

[0160] Figure 4 is a schematic diagram of the imaging principle of the reflection marker provided by the present invention, Figure 5 is a schematic diagram of the relationship between the deformation of the reflection marker in the object space and the deformation in the image space of the photo provided by the present invention, Figure 6 is a schematic diagram of the plane polygon and circular design of the target provided by the present invention, Figure 7 is a schematic diagram of the projection of the target with height compensation onto the reference plane provided by the present invention. According to Figures 4 - 7 it can be seen that calibration can be performed according to the control targets in the images obtained by the high-resolution dual monitoring cameras and the drone monitoring module, and deformation data of the deformable body can be obtained according to the deformation monitoring targets.

[0161] 4. Control Center The control of the camera, the storage of photos and the deformation calculation are all completed on the computer in the control center. The control center is communicatively connected to the camera and the drone through the Internet of Things, and is mainly responsible for sending acquisition instructions, receiving image data, calculating the three-dimensional displacement of the monitoring points, and generating real-time deformation information.

[0162] II. Implementation Steps: The specific usage process of the deformation monitoring system in this embodiment is as follows: 1. Equipment Deployment 1.1 Camera Deployment Two cameras are respectively arranged in the peripheral area of the deformable body, and the camera focal lengths are adjusted to cover all monitoring points.

[0163] Ensure that the main optical axes of the two cameras are approximately perpendicular.

[0164] 1.2 Deformation Monitoring Target Deployment Adopt the insertion or embedding fixation method, and the number of deformation monitoring targets is determined according to the area of the deformed body and the complexity of internal movement.

[0165] 1.3 UAV-assisted layout The UAV hovers above the measurement area to ensure that the orthophoto covers all monitoring areas.

[0166] 2. Image acquisition 2.1 Acquisition strategy The camera and the UAV synchronously acquire images at a preset time interval, which is flexibly adjusted according to the sliding speed and danger level of the deformed body.

[0167] Rapid deformation: It is recommended to set a 1-second time interval or directly shoot a video to capture continuous dynamic changes.

[0168] Slow deformation: It can be extended to 5 - 10 seconds or shoot videos regularly.

[0169] 2.2 Data management The acquired images are named according to the time tag and device tag and are transmitted to the control center in real time.

[0170] 3. Calculation of deformation amount The control center processes the images through the embedded algorithm to calculate the three-dimensional displacement of the deformation monitoring points.

[0171] 3.1 Target tracking Using the maximum cross-correlation technology, extract the pixel trajectory of the monitoring point in the time-series images to obtain the pixel change amount.

[0172] 3.2 Height compensation and correction Through the height compensation algorithm, correct the projection center coordinates of the deformation monitoring target on the reference plane.

[0173] 3.3 Similarity transformation correction Correct the rotation, scaling, and translation of the image caused by camera movement or perspective change to ensure data accuracy.

[0174] 3.4 Moiré fringe technology-assisted correction Figure 8 It is a schematic diagram of the principle of the sampling Moiré method for precise displacement measurement provided by the present invention. As Figure 8 shown, the displacement change amount of the control mark can be obtained by extracting the phase information of the control mark. Among them, before deformation, in the original Moiré fringe image, the intensity distribution of the Moiré fringe is , and after low-pass filtering, it is obtained , and then the phase distribution (Moiré phase) of the Moiré fringe is obtained ; correspondingly, after deformation, in the original Moiré fringe image, the intensity distribution of the Moiré fringe is , which is obtained through low-pass filtering processing , and then the phase distribution of the Moiré fringes (Moiré phase) is obtained ; the phase difference of the Moiré fringes in the images before and after deformation is obtained , and then the displacement change amount in the plane is calculated .

[0175] 3.5 X and Y Displacement Calculation Through re-performing similarity transformation for ultra-fine image compensation, the displacements of the deformation monitoring target in the X and Y directions are finally accurately calculated.

[0176] 3.6 Three-dimensional Displacement Solution Based on the above results, through multi-view observation of two cameras, the three-dimensional displacement amount of the monitoring point is calculated :[[]]END]]

[0177] 4. Result Output 4.1 Real-time Monitoring Information The control center generates deformation time series data and displays the deformation process in real time in the form of a curve graph, a heat map or a numerical table.

[0178] 4.2 Accuracy Verification At a monitoring distance of 35 meters, based on a camera with a focal length of 24 mm, an image with a resolution of 4k and a marker spacing of 100 mm, the deformation monitoring accuracy of this system can reach the sub-millimeter level.

[0179] III. Effect Verification: This embodiment shows a high-precision monitoring effect in the experiment.

[0180] Three-dimensional Deformation Amount Monitoring: Experimental verification shows that this embodiment can accurately monitor the displacement changes of the deformation monitoring target on the surface of the deformed object in the X, Y, and Z directions.

[0181] Comparison with Traditional Methods: Compared with the monitoring method based on stereophotogrammetry, the calculation error of this embodiment is significantly reduced and the monitoring accuracy is higher. Combining technologies such as target tracking, similarity transformation and Moiré fringes, this embodiment can still maintain a high accuracy under complex lighting conditions and on the surface of complex deformed objects.

[0182] Supplementary Explanation: Adaptability and Expandability: This embodiment can flexibly adjust the camera focal length, the target point spacing and the monitoring point layout scheme according to the size of the deformation area and the monitoring accuracy. It is applicable to various deformation types and different deformation complexity scenarios.

[0183] Multiple configuration support: This embodiment supports single or dual cameras, single or dual cameras and independent monitoring by drones, as well as collaborative operations between drones and cameras or cameras.

[0184] Compared with the existing deformation monitoring method based on stereophotogrammetry technology, this embodiment has the following beneficial effects: This embodiment provides a deformation monitoring system and method based on phase information, including: deformation monitoring targets arranged on the surface of the deformed body and control targets arranged in the surrounding area of the deformation; the control targets are used for calibration and reference, and the targets are two-dimensional rectangular planes or three-dimensional cubes, with regularly spaced dot matrices of black or colored polygons or circles on the surface, and the printing material of the dots can use reflective materials that are common day and night; the system also includes a high-resolution monitoring camera, a high-definition camera, and a drone hovering directly above the monitoring area.

[0185] Figure 9 It is a schematic diagram of the algorithm flow for calculating the deformation amount provided by the present invention. As Figure 9 shown, the images or videos recorded in the monitoring area are collected through cameras, cameras, and drones, and a comprehensive algorithm is formed by combining technologies such as target tracking, height compensation, similarity transformation correction, and Moiré fringe phase analysis. Through marker center tracking, pixel-level accuracy is achieved, and through rough image compensation using height compensation and similarity transformation, pixel-level accuracy is achieved; the reference marker displacement uses the sampling Moiré method to achieve 1 / 100 pixel-level accuracy; through ultra-fine image compensation using similarity transformation, 1 / 100 pixel-level accuracy is achieved; finally, based on the calculation of the reference line displacement, the displacement data of the deformation monitoring points in the X, Y, and Z directions can be accurately calculated, and the three-dimensional deformation information (landslide sliding displacement data) of the deformed body can be obtained in real time. This embodiment supports multiple monitoring configurations, including single high-resolution cameras or cameras, dual high-resolution cameras or dual cameras, single drones, and combinations of drones, cameras, and cameras, and can achieve high-precision deformation monitoring at low cost.

[0186] In addition, this embodiment can also monitor the deformation process of stable feature points on the deformed body. These feature points are similar to reflection marks and are suitable for environments with good light conditions, improving the spatial resolution of deformation monitoring. Through the comprehensive algorithm and the layout of the deformation monitoring targets, this embodiment can not only achieve the monitoring of the high-precision three-dimensional deformation of the deformed body, but also comprehensively cover the deformation process, providing strong technical support for the real-time monitoring and early warning of engineering disasters.

[0187] The system embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0188] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0189] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments or equivalently replace some of the technical features. And these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A deformation monitoring system based on phase information, characterized in that Including: A group of photographing devices, a plurality of deformation monitoring targets, a plurality of control targets, and a data processing unit; The deformation monitoring targets are arranged on the surface of the target deformable body; The control targets are arranged in the peripheral area of the target deformable body and serve as the spatial reference benchmark for the deformation monitoring targets; The target deformable body, the plurality of deformation monitoring targets, and the plurality of control targets jointly form a monitoring area; The group of photographing devices includes a group of unmanned aerial vehicles and / or a group of cameras, which are used to collect time-series images of the monitoring area and transmit the time-series images to the data processing unit; the group of unmanned aerial vehicles includes at least one unmanned aerial vehicle, which is arranged directly above the monitoring area; the group of cameras includes at least one camera, which is arranged in the peripheral area of the monitoring area, and the camera scans to cover the monitoring area; The data processing unit is used to extract the center coordinates of the deformation monitoring targets and the control targets according to the time-series images through the maximum cross-correlation algorithm, correct the center coordinates of the deformation monitoring targets in the time-series images through the height compensation algorithm, correct the time-series images after correcting the center coordinates of the deformation monitoring targets according to the center coordinates of the control targets through the similarity transformation algorithm, calculate the displacement change amount of the control targets and obtain the corrected coordinates of the control targets according to the corrected time-series images through the moiré fringe method, finely compensate the positions of the deformation monitoring targets in the corrected time-series images according to the corrected coordinates of the control targets through the similarity transformation algorithm, and finally determine the deformation monitoring result of the target deformable body based on the finely compensated time-series images.

2. The deformation monitoring system based on phase information according to claim 1, wherein Extracting the center coordinates of the deformation monitoring targets and the control targets according to the time-series images through the maximum cross-correlation algorithm includes: Taking the initial frame image in the time-series images as the template image, and calculating the cross-correlation function between each subsequent frame image in the time-series images and the template image; Based on the maximum correlation value in the cross-correlation function between each subsequent frame image and the template image, determining the center coordinates of the deformation monitoring targets and the control targets in each subsequent frame image.

3. The deformation monitoring system based on phase information according to claim 1 or 2, characterized in that, Correcting the time-series images after correcting the center coordinates of the deformation monitoring targets according to the center coordinates of the control targets through the similarity transformation algorithm includes: Determining a reference coordinate system based on the center coordinates of a plurality of control targets in the template image and a pre-determined reference control target; the reference control target is used for unified spatial reference of the control targets other than the reference control target among the plurality of control targets; Calculating similarity transformation parameters based on the center coordinates of the control targets in the reference coordinate system and the center coordinates of the control targets in the time-series images after correcting the center coordinates of the deformation monitoring targets; Based on the similarity transformation parameters, correcting the time-series images after correcting the center coordinates of the deformation monitoring targets.

4. The deformation monitoring system based on phase information according to claim 3, wherein Calculating the displacement change amount of the control targets and obtaining the corrected coordinates of the control targets according to the corrected time-series images through the moiré fringe method includes: Extract multiple frames of moiré fringe images based on the corrected time-series images; Perform low-pass filtering and downsampling on the multiple frames of moiré fringe images to obtain multiple frames of downsampled images; Calculate the displacement change of the control target based on the moiré fringe phase of the multiple frames of downsampled images and obtain the corrected coordinates of the control target.

5. The deformation monitoring system based on phase information according to claim 1, wherein The number of the control targets is three, and the three control targets form an L-shaped structure.

6. The deformation monitoring system based on phase information according to claim 1 or 5, characterized in that, The deformation monitoring target and the control target are two-dimensional rectangular planes or three-dimensional cubes; black or colored equally-spaced regular dot matrices are arranged on the surfaces of the deformation monitoring target and the control target, and the dot shapes of the equally-spaced regular dot matrices are polygons or circles; the printing material of the equally-spaced regular dot matrices is a reflection material that is applicable day and night.

7. A deformation monitoring method based on phase information, characterized in that, Applied to the deformation monitoring system based on phase information according to any one of claims 1 to 6, comprising: Collect time-series images of the monitoring area; the monitoring area is jointly composed of a target deformable body, multiple deformation monitoring targets and multiple control targets; Extract the central coordinates of the deformation monitoring targets and the control targets according to the time-series images through the maximum cross-correlation algorithm; Correct the central coordinates of the deformation monitoring targets in the time-series images through the height compensation algorithm; Correct the time-series images after correcting the central coordinates of the deformation monitoring targets according to the central coordinates of the control targets through the similarity transformation algorithm; Calculate the displacement change of the control target according to the corrected time-series images through the moiré fringe method and obtain the corrected coordinates of the control target; Perform fine compensation on the positions of the deformation monitoring targets in the corrected time-series images according to the corrected coordinates of the control targets through the similarity transformation algorithm; Determine the deformation monitoring result of the target deformable body based on the time-series images after fine compensation.

8. The deformation monitoring method based on phase information according to claim 7, wherein Extract the central coordinates of the deformation monitoring targets and the control targets according to the time-series images through the maximum cross-correlation algorithm, comprising: Use the initial frame image in the time-series images as the template image, and calculate the cross-correlation function between each subsequent frame image in the time-series images and the template image; Determine the central coordinates of the deformation monitoring targets and the control targets in each subsequent frame image based on the maximum correlation value in the cross-correlation function between each subsequent frame image and the template image.

9. The deformation monitoring method based on phase information according to claim 7 or 8, characterized in that Correct the time-series images after correcting the central coordinates of the deformation monitoring targets according to the central coordinates of the control targets through the similarity transformation algorithm, comprising: Determine a reference coordinate system based on the central coordinates of multiple control targets in the template image and a pre-determined reference control target; the reference control target is used for unified spatial reference for the control targets other than the reference control target among the multiple control targets; Calculate the similarity transformation parameters based on the central coordinates of the control targets in the reference coordinate system and the central coordinates of the control targets in the time-series images after correcting the central coordinates of the deformation monitoring targets; Correct the time-series images after correcting the central coordinates of the deformation monitoring targets based on the similarity transformation parameters.

10. The deformation monitoring method based on phase information according to claim 9, wherein By the Moiré fringe method, based on the corrected time series images, calculating the displacement change amount of the control target and obtaining the corrected coordinates of the control target, including: Extracting multiple frames of Moiré fringe images based on the corrected time series images; Performing low-pass filtering and downsampling processing on the multiple frames of Moiré fringe images to obtain multiple frames of downsampled images; Calculating the displacement change amount of the control target and obtaining the corrected coordinates of the control target based on the Moiré fringe phases of the multiple frames of downsampled images.

Images