Automatic Monitoring and Early Warning Method for the Deformation of Offshore Guide Dikes Based on Remote Sensing Images

Abstract

The present invention relates to the field of ocean engineering technology, specifically an automatic monitoring and early warning method for the deformation of offshore breakwaters based on remote sensing images, including: obtaining remote sensing images and standard images of offshore breakwaters, and forming a remote sensing image set by arranging the remote sensing images according to the acquisition time and position; dividing the remote sensing image set into several image pairs, performing image registration on the basis of the positions of multiple image pairs and the standard image on the offshore breakwater to obtain target remote sensing images; performing image differencing on the target remote sensing images at different time points to obtain deformation features, and determining deformation regions related to the distribution of each deformation feature; combining the deformation regions according to the distribution of the deformation features, determining the deformation trend direction corresponding to the deformation regions, analyzing the influence of the deformation amount and deformation rate in the deformation trend direction to obtain the influence path corresponding to each deformation region; determining the alarm type corresponding to the offshore breakwater based on the obtained influence path; improving the accuracy and efficiency of offshore breakwater deformation recognition.

Description

Technical Field

[0001] The present invention relates to the technical field of ocean engineering, specifically an automatic monitoring and early warning method for the deformation of offshore breakwaters based on remote sensing images. Background Art

[0002] In ocean engineering, as an important infrastructure, the safety of offshore breakwaters is directly related to multiple aspects such as ocean transportation and ocean resource development. However, due to the complexity of the ocean environment and the particularity of the breakwater structure, traditional deformation monitoring methods often struggle to achieve real-time monitoring and comprehensive coverage. In recent years, with the rapid development of remote sensing technology, using remote sensing images for offshore breakwater deformation monitoring has gradually become a new trend. Remote sensing images have the advantages of wide coverage, high update frequency, and large amount of information, enabling real-time monitoring and comprehensive coverage of offshore breakwaters. Therefore, based on the background art, the current solution combines remote sensing image technology and automated algorithms to achieve automatic monitoring and early warning of offshore breakwater deformation, aiming to improve the monitoring efficiency and accuracy and provide strong guarantee for the safe operation of ocean engineering.

[0003] For example, Chinese Patent Publication No. CN118730007A discloses a method and device for automatic multi-point deformation monitoring of the dam surface based on acoustic wave recognition. Among them, the method for automatic multi-point deformation monitoring of the dam surface includes: installing acoustic wave receiving devices at three fixed points on the mountain bodies on both sides of the dam that are not collinear; driving an unmanned vehicle on the dam surface, installing an acoustic wave transmitter on the unmanned vehicle, and continuously emitting acoustic wave signals with a specific frequency into the air; the acoustic wave receiving devices at the fixed points of the dam calculate the distance between the receiving device and the vehicle transmitter point according to the time difference between the emission and arrival of the received acoustic wave signals; calculate the GNSS coordinates at the vehicle transmitter point according to the three-point positioning method and the GNSS coordinates of the fixed points; the vehicle regularly patrols according to the set route and calculates the immediate coordinates of the set points, and compares them with the historical coordinates to obtain the displacement change of the fixed points on the dam surface.

[0004] For example, Chinese Patent Publication No. CN117433480A discloses a high-precision position real-time dynamic monitoring and early warning system and method based on remote sensing monitoring, including: a sensing module for obtaining monitoring data of the monitoring object; a communication module for transmitting the monitoring data to a safety monitoring cloud platform; a safety monitoring cloud platform for analyzing the monitoring data, calculating the position and displacement information of the monitoring station, and providing real-time millimeter-level deformation data for engineering monitoring; a risk analysis and early warning module for statistically analyzing the monitoring data based on the safety monitoring cloud platform and outputting risk assessment and early warning information externally through a combined early warning matrix.

[0005] In the prior art, methods such as acoustic wave reflection, flow detection, and pressure detection are described for treating offshore structures such as guide dikes. However, the distribution and change trend of the deformation amounts generated on the guide dike when the guide dike deforms are ignored, resulting in poor effect of risk identification of the guide dike and inability to make corresponding treatments in time according to the deformation trend of the guide dike. Summary of the Invention

[0006] To solve the above technical problems, the technical solution adopted by the present invention is: an automatic monitoring and early warning method for deformation of an offshore guide dike based on remote sensing images, including: S1, obtaining remote sensing images and standard images of the offshore guide dike, and forming a remote sensing image set according to the acquisition time and position of the remote sensing images; the remote sensing image set includes satellite remote sensing images, unmanned aerial vehicle remote sensing images, and SAR images.

[0007] S2, dividing the remote sensing image set into several image pairs, and performing image registration according to the positions of the multiple image pairs and the standard image on the offshore guide dike to obtain the registered target remote sensing images.

[0008] S3, performing image differencing on the target remote sensing images at different time points, analyzing the deformation characteristics related to the surface deformation of the offshore guide dike, and determining the deformation regions related to the distribution of each deformation characteristic.

[0009] S4, combining the deformation regions according to the distribution of the deformation characteristics, determining the deformation trend direction corresponding to the deformation regions, and performing impact analysis on the deformation amounts and deformation rates in the deformation trend direction to obtain the impact paths corresponding to each deformation region.

[0010] S5, determining the alarm type corresponding to the offshore guide dike based on the obtained impact paths.

[0011] The beneficial effects of the present invention are as follows: First, by obtaining satellite remote sensing images, unmanned aerial vehicle remote sensing images, and SAR images, after determining the specific edge conditions on the guide dike, the SAR images are used to determine the relative deformation amounts generated, and relative deformation regions are set according to the deformation distribution, so that when identifying deformation characteristics, it can be judged whether the current deformation is caused by erosion or other forms of deformation such as accumulation, which is convenient for subsequent verification of the specific conditions on the guide dike and provides a basis for subsequent decision-making and treatment of the guide dike.

[0012] Second, the present invention identifies the trend directions of the deformation amounts on each deformation region, and projects and transforms these identified trend directions with the guide dike edge to obtain the deformation trend of the current guide dike at the edge. Finally, the deformation trend direction of the guide dike is identified according to the relative center trajectory. It can be understood that after identifying multiple small deformation features, the deformation features are combined into the main trend of the deformation of the guide dike, which is convenient for subsequent personnel to make decisions based on this main trend. At the same time, for the identification of the relative deformation trends in the macro and local aspects, the deformation trend of the guide dike can be more comprehensively displayed, improving the accuracy of subsequent early warning of the deformation trend.

[0013] Third, the present invention processes the deformation path on the current guide dike, and can find out the main path of the current guide dike during deformation and the points with prominent deformation on this path, which is convenient for subsequent personnel to judge the structural stability of the guide dike based on these points. At the same time, matching the points on these paths with relevant cases and treatment measures in the database can further assist personnel in making decisions and help reduce safety accidents of the guide dike structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The present invention will be further described below with reference to the drawings and embodiments.

[0015] Figure 1 is a flowchart of an automatic monitoring and early warning method for deformation of a marine guide dike based on remote sensing images.

[0016] Figure 2 is a flowchart of step S2 of the automatic monitoring and early warning method for deformation of a marine guide dike based on remote sensing images.

[0017] Figure 3 is a flowchart of step S3 of the automatic monitoring and early warning method for deformation of a marine guide dike based on remote sensing images.

[0018] Figure 4 is a flowchart of step S4 of the automatic monitoring and early warning method for deformation of a marine guide dike based on remote sensing images.

[0019] Figure 5 is a flowchart of step S5 of the automatic monitoring and early warning method for deformation of a marine guide dike based on remote sensing images. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] The embodiments of the present invention will be described in detail below. The following described embodiments are exemplary and are only used to explain the present invention and should not be construed as a limitation to the present invention. For those not specified in the embodiments regarding specific technologies or conditions, they shall be carried out according to the technologies or conditions described in the literature in the field or according to the product specifications.

[0021] It should be noted that a guide dike is a dike built on one or both sides of the channel in the estuary bar area, which is a water conservancy facility used to direct the flow of water, scour sediment, and increase or maintain the water depth of the approach channel. When the guide dike deforms, it generally means that after the guide dike itself is scoured by seawater, the rock mass or rock structure at the corresponding position is damaged or dust accumulates. When using remote sensing images to identify relevant guide dikes, relevant deformations occur, that is, under the action of natural forces such as tides and waves, the water-facing side of the guide dike will be strongly scoured. Long-term scouring will lead to the erosion of the guide dike foundation and the destruction of the structure; and the upstream side of the guide dike may be severely silted up due to sediment interception, while the downstream side may be scoured due to insufficient sediment supply. This unbalanced sediment siltation and scouring will also cause the deformation of the guide dike; ultimately, all these contents will be identified to determine the main influencing factors on the guide dike during the process of guiding the water flow, so as to accurately identify the hydrological information generated by soil erosion or water flow orientation in the corresponding area. By monitoring the changes of the guide dike, potential safety hazards can be discovered in time to ensure the safe navigation of ships in the channel. At the same time, by monitoring the changes of the guide dike, the degree of its impact on the marine environment can be evaluated, and necessary measures can be taken to reduce the adverse effects to ensure the long-term stable operation of the guide dike.

[0022] Refer to Figure 1 , An automatic monitoring and early warning method for the deformation of offshore guide dikes based on remote sensing images, including: S1, obtaining remote sensing images and standard images of offshore guide dikes, and forming a remote sensing image set according to the acquisition time and location of the remote sensing images; the remote sensing image set includes satellite remote sensing images, unmanned aerial vehicle remote sensing images, and SAR images.

[0023] S2, dividing the remote sensing image set into several image pairs, and performing image registration according to the positions of multiple image pairs and the standard image on the offshore guide dike to obtain the registered target remote sensing images.

[0024] S3, performing image differencing on the target remote sensing images at different time points, analyzing the deformation characteristics related to the surface deformation of the offshore guide dike, and determining the deformation regions related to the distribution of each deformation characteristic.

[0025] S4, combining the deformation regions according to the distribution of the deformation characteristics, determining the deformation trend direction corresponding to the deformation regions, and performing impact analysis on the deformation amount and deformation rate in the deformation trend direction to obtain the impact paths corresponding to each deformation region.

[0026] S5, determining the alarm type corresponding to the offshore guide dike based on the obtained impact paths.

[0027] In an embodiment of the present invention, multiple images are used to illustrate the deformation and specific conditions of a marine breakwater; satellite remote sensing images directly display the macroscopic scene of the marine breakwater, and the images are acquired with a relatively large acquisition period; unmanned aerial vehicle (UAV) remote sensing images collect images of the marine breakwater using a camera or other sensors along the flight path of the UAV, and high-resolution images or video materials are collected in real time; SAR images use microwave imaging technology to generate images by receiving signals reflected back from the ground. When there are minor deformations on the earth's surface, it will cause changes in the path lengths of two SAR images, which are reflected in the phase represented by the images. Then, by analyzing the phase difference evolution, the amount of deformation occurring on the ground can be obtained, thereby knowing the deformation and change trend of the marine breakwater; standard images represent the marine breakwater under normal processing and are used for differencing with the current remote sensing images.

[0028] When composing the remote sensing image set, the marine breakwater to be collected is corresponded according to the positions corresponding to the current satellite remote sensing images, UAV remote sensing images, and SAR images, so as to obtain an image set of the marine breakwater composed of multiple images. This image set will perform image enhancement and be combined according to the acquisition time and path to ensure that the corresponding images can correspond to each other during image calibration, so as to obtain the multi-source data information in the current image.

[0029] At the same time, since the content represented by SAR images is different from that of satellite remote sensing images and UAV remote sensing images, satellite remote sensing images and UAV remote sensing images show the color and texture of the marine breakwater, mostly in the form of color images, while SAR images present grayscale images, reflecting the characteristics of the reflection of radar waves by objects on the marine breakwater. This leads to the grayscale processing of satellite remote sensing images and UAV remote sensing images during image calibration. After obtaining the grayscale histograms, the content on the satellite remote sensing images, UAV remote sensing images, and SAR images is subjected to image calibration to ensure the consistency and comparability of the current target remote sensing image in showing the marine breakwater.

[0030] The set standard images will be set as the corresponding standard images in the forms of satellite remote sensing images, UAV remote sensing images, and SAR images in sequence. These standard images can be the image data collected in the past or the images processed from the data collected in the past, and these images are used to compare with the data obtained currently to complete the deformation analysis of the marine breakwater.

[0031] In an embodiment of the present invention, in step S2, alignment is mainly performed on the SAR images, satellite remote sensing images, and optical images corresponding to unmanned aerial vehicle (UAV) remote sensing images existing in the remote sensing image set to determine that the information after image comparison can assist in identifying a sufficient number of deformed regions. After comparing the spatial consistency of these images, a target remote sensing image of multiple image combinations with spatial consistency is obtained. The target remote sensing image includes satellite remote sensing images, UAV remote sensing images, and SAR images with consistent position representations.

[0032] When the remote sensing image set is divided into several image pairs, the remote sensing image set is divided according to the acquisition time. For example, images acquired at the same time are paired two by two as an image pair, and this image pair represents an image pair of adjacent times. Dividing into multiple image pairs is to ensure that the acquired images can cover the guide dike and the spatial ranges shown by the images are consistent without other incorrect images.

[0033] As Figure 2 shown, the implementation method of step S2 includes: S21, obtaining the acquisition time of the remote sensing image set, taking adjacent frame images of the remote sensing image set within adjacent acquisition times as an image pair to obtain several image pairs. For adjacent frame images, it means adjacent in the analogical sense when continuously acquiring images at a certain position under the corresponding acquisition time. For example, the UAV remote sensing images in the remote sensing image set are represented as video-like data during acquisition, and image frames are extracted from this video-like data. If they are satellite remote sensing images and SAR images, they are adjacent in the analogical sense. Generally, images with close times are selected as image pairs. The content represented by such image pairs is almost the same and the difference is small, which is convenient for image calibration with the standard image.

[0034] S22, preprocess each image pair. The preprocessing includes geometric correction, radiometric normalization, etc. Extract feature points from the preprocessed image pairs and the standard image, and perform the nearest neighbor algorithm on the feature points to obtain the feature point pairs of the matching image pairs and the standard image. For geometric correction, it is a method of changing the original coordinates of each image pair using affine coefficients, which is described in the prior art and will not be elaborated here. As for radiometric normalization, it is used to eliminate the difference in illumination conditions in the satellite remote sensing images and UAV remote sensing images in the image pairs. The original values of the pixel points in the satellite remote sensing images and UAV remote sensing images are subtracted by the average value of the pixel points and then divided by the standard deviation of the pixel points for normalization processing.

[0035] After that, the nearest neighbor algorithm is implemented by calculating the distance between pixel points of the extracted feature points, that is, calculating the distance between the pixel points corresponding to the feature points in the image pairs and the standard image to determine whether the corresponding positions in the images are the same, and retaining the points with high confidence as the implementation of the matching method at this time, finally obtaining multiple feature point pairs.

[0036] S23. After transforming the feature point pairs of the matched image pairs and the standard images, the image pairs that meet the mean square error are output as the target remote sensing images.

[0037] The transformation here refers to methods such as affine transformation and projection transformation. At this time, it does not limit the processing methods of the matched images and the standard images, but only limits the image pairs that meet the mean square error. It is required that the RMSE of satellite remote sensing images and UAV remote sensing images is < 1 pixel, and the RMSE of SAR images is < 0.2 pixel; the content that is not restricted is described in the prior art, so no further explanation will be given here.

[0038] After image registration is completed, it is also necessary to verify the distribution of these images on the offshore breakwater to determine whether the position after image calibration is the same as the distribution of the offshore breakwater.

[0039] When outputting the image pairs that meet the mean square error as the target remote sensing images, the following implementation methods are also included: sequentially obtain the first image region and the second image region according to the images corresponding to multiple image pairs. The first image region represents the region represented by the initial frame of the remote sensing image set in the image pair, and the second image region represents the region represented by the intermediate frame of the remote sensing image set in the image pair. This initial frame represents the image with the earliest time in the corresponding data when aligning the images for a certain region after forming the image pairs according to time adjacency; the intermediate frame represents the region represented by the intermediate time in these image pairs.

[0040] Detect the similarity between the first image region and the second image region, and judge the registration degree of the target remote sensing image according to the similarity between the first image region and the second image region. Here, the image pairs within a short time interval are compared, that is, to evaluate whether the relevant images are aligned after image registration at this time. The similarity at this time is calculated in the form of the average pixel distance or the mean square error to calculate the similarity between the first image region and the second image region.

[0041] When the similarity between the first image region and the second image region is greater than the similarity threshold, and the registration degree of the target remote sensing image meets the mean square error, the corresponding target remote sensing image is output. For the similarity threshold, the average value of the similarities between the first image region and the second image region that are normally used as the target remote sensing images in the historical data can be adopted, so as to obtain the image set suitable for differential processing.

[0042] Geographical features such as offshore breakwaters may change over a long period of time due to natural factors (such as tides and wind waves) or human activities (such as construction and maintenance). By comparing the similarity between the initial frame and the intermediate frame, not only can the registration effect be verified, but these changes can also be indirectly detected. If the similarity decreases significantly, it may mean that some parts of the offshore breakwater have changed significantly during the time interval between image acquisitions. This comparison also emphasizes the temporal continuity of the currently acquired remote sensing image set to determine whether the images at corresponding positions can maintain a large similarity in time after calibration at spatial positions, so as to reduce the influence of natural factors, acquisition time, etc. on the images.

[0043] In one embodiment of the present invention, when performing image differencing in step S3, it is necessary to identify the actual volume and area of the breakwater in the case of image differencing to verify the deformation generated at this time and the specific situation of the current breakwater. At the same time, due to the differences between satellite remote sensing images, unmanned aerial vehicle (UAV) remote sensing images, and SAR images in the target remote sensing images, different-dimensional differencing processing needs to be performed, and then the features in the differenced regions of these images are superimposed to display the specific situation of the current target remote sensing image. For example, satellite remote sensing images can identify changes in the large-scale coverage of offshore breakwaters, UAV remote sensing images can identify local deformations or damages, and SAR images can identify large-scale deformations. At this time, the features of image differencing will sequentially identify whether there are obvious deformations on the main body of the offshore breakwater from the coverage situation where the three exist, and then use SAR images to identify the deformation situation in the large-scale range. Finally, use UAV remote sensing images to identify the local deformations and changes to determine the overall relative deformation trend, and finally obtain the deformation region to be identified, or find the associations and similarities between the deformation regions.

[0044] As Figure 3 shown, the implementation method of step S3 includes: S31, performing image differencing on the target remote sensing images at different time points to respectively obtain a morphological differenced image and an intensity differenced image.

[0045] During image differencing, satellite remote sensing images and UAV remote sensing images perform differencing on pixel values to obtain texture features regarding the texture and morphological changes of the breakwater, so as to determine the differences in texture features and identify whether there are relevant morphological changes in the breakwater. Then, when using SAR images to identify whether there are deformations in these regions with morphological changes, the phase difference form can be used to identify the deformation of the deformation region, and then the features are superimposed to obtain the features required for differencing.

[0046] The morphological difference image represents the morphological changes of the guide dike in satellite remote sensing images and UAV remote sensing images, described by pixel value differences. The intensity difference image is the description of the intensity value difference of SAR images using the standard deviation of pixel points at different time points.

[0047] S32. Extract the edge features of the guide dike in the morphological difference image to identify the morphological change areas on the surface of the guide dike.

[0048] When extracting edge features, perform Canny detection on the edge of the guide dike, then identify the parts with edge differences detected, and determine the differences existing in this area, such as edge breaks or areas with differences caused by deformation. These areas will be marked in the form of pixel differences, and the morphological change areas will be displayed using an image.

[0049] S33. According to the intensity difference image, identify the deformation amount and deformation rate of the guide dike, and generate the deformation field of the intensity difference image according to the deformation amount and deformation rate.

[0050] The deformation amount of the intensity difference image can be calculated by obtaining the phase difference and radar wavelength of the intensity difference image. For example, the deformation amount is expressed as: ; where represents the deformation amount, represents the radar wavelength, represents pi, represents the phase difference. The phase difference is obtained in the form of differential interference and phase unwrapping after removing the topographic phase to obtain the phase difference of the guide dike phase. This phase difference will represent the subtle deformation of the guide dike at different time points, and then the deformation amount can be directly calculated.

[0051] That is, step S33 further includes: performing interference processing on intensity difference images at multiple different time points to generate an interferogram corresponding to the intensity difference image, removing the topographic phase from the part with phase difference in the interferogram to generate multiple interference pairs, and sequentially identifying the deformation amount and deformation rate of the guide dike deformation according to the phase difference on the interference pairs.

[0052] At this time, the topographic phase will be processed according to the different phase values generated by the radar mapping to the sea water, and the generated topographic phase will be removed. Then, an algorithm such as SNAPHU is used to restore the phase to a continuous phase; the places with phase differences will be regarded as interference pairs, and then the phase difference and the corresponding deformation amount can be calculated.

[0053] For the deformation rate, it needs to be set according to the number of interference pairs used at this time. For example ; where represents the deformation rate, Represents the time span, usually expressed in years; Represents the phase difference of the i-th interference pair, Represents the number of interference pairs, and the value range of i is from 1 to N.

[0054] After that, the positions corresponding to the deformation amount and deformation rate will be obtained and marked on the guide dike, so as to obtain a deformation field that can represent the spatial distribution of the guide dike.

[0055] S34. Superimpose the morphological change area and the deformation field to obtain the superimposed deformation characteristics.

[0056] Register the morphological change area and the deformation field according to the points on the image. After determining that the positions of the expressions are the same, fuse the information on the morphological change area with the information on the deformation field at the corresponding positions. For example, in the form of weighted superposition, superimpose the weights of the texture existing at the corresponding positions and the weights of the deformation occurring, to illustrate what other forms of situations exist when deformation occurs on the guide dike, whether it is the deformation caused by the corrosion of the guide dike, or whether the guide dike itself has produced subtle deformations while remaining basically intact. After combining these information, it is the deformation characteristic required currently.

[0057] In the subsequent processing of the deformation area, the neural network will use the deformation characteristics obtained at this time to predict or segment the deformation characteristics, in order to obtain deformation areas with different deformation degrees, such as areas with fast deformation rate, large deformation, etc.

[0058] The implementation methods for determining the deformation areas related to the distribution of each deformation characteristic also include the following: taking the position where any deformation characteristic is located as the central position, comparing the differences between adjacent deformation characteristics at the central position. The differences between adjacent deformation characteristics are calculated based on the deformation amount. The differences between adjacent deformation characteristics can identify whether there are significant changes in the deformation characteristics within the local range, and whether the change of the deformation amount is continuous within the local area or there are phenomena of jumping and interruption. The value calculated at this time is to represent these existing phenomena with data and then perform clustering to obtain areas with relatively consistent deformation and areas with local anomalies.

[0059] Cluster the deformation characteristics according to the offset of the differences between adjacent deformation characteristics in the horizontal direction, and combine them into multiple deformation areas.

[0060] At this time, clustering is to take the values of the differences between adjacent deformation characteristics to identify whether the area represented by the deformation characteristic has undergone overall lateral displacement, local distortion deformation, and stretching or compression deformation. Cluster or analyze these values to find the points with the same pattern and combine them into the required deformation area.

[0061] When calculating the difference between adjacent deformation features, the Euclidean distance metric between the deformation features is used for identification. For example, the implementation method of clustering the deformation features and combining them into multiple deformation regions includes: converting the deformation features into horizontal displacement vectors and corresponding spatial coordinates, calculating the spatial distances between the deformation features, and determining whether there are adjacent deformation features that satisfy spatial adjacency at the central position. When the spatial adjacency is satisfied and the difference between the adjacent deformation features is less than the deformation difference threshold, it is considered that the adjacent deformation features and the central position are in the same region; then, all the deformation features that satisfy spatial adjacency and the deformation difference threshold are recursively processed to obtain the deformation region; for spatial adjacency, it is determined whether the distance between the adjacent deformation features is less than the preset adjacency distance. At this time, the preset adjacency distance is set according to the influence of the seawall erosion by the sea waves. For example, for severely eroded areas, it is set to a few centimeters, and for less severely eroded areas, the identified deformation features are enlarged to within a few meters to determine the identification and processing of relatively local and larger regions; at the same time, the deformation difference threshold is the average value of the differences between adjacent deformation features in historical data to identify obvious and relatively average regions; finally, the regions with consistent changes are combined. For points where the difference between adjacent deformation features is greater than or equal to the deformation difference threshold, when this situation occurs, it indicates that there is an interruption or points of different deformations adjacent to the current deformation feature. At this time, a new central position will be reselected for clustering. Finally, all the deformation features will be combined together in the form of spatial adjacency and smaller differences, which is convenient for subsequent analysis of the distribution of deformation features on different deformation regions.

[0062] In an embodiment of the present invention, step S4 mainly identifies the deformation trend direction according to the deformation amount existing in the deformation region. After fitting the points that generate deformation, the direction and corresponding value of the generated deformation can be known. Then, by combining the directions on all the deformation regions, a trend line can be fitted to represent the deformation trend direction of the currently identified seawall.

[0063] For the influence path, it represents the path by which the occurrence of the deformation amount is affected by the environment, or in the case of global distribution, the overall change path of the deformation amount, as well as the weights between multiple nodes on this path, to obtain this influence path that expresses the correlation between different deformation regions.

[0064] Such as Figure 4As shown in the figure, the implementation of step S4 includes: S41, identifying the amount of deformation and the deformation rate on each deformation region to determine the trend direction of each deformation region; the trend direction of each deformation region will identify the deformation generated in this deformation region and represent it in the form of a vector. For example, an arrow is used to represent the direction corresponding to the position before and after deformation, and the length of the arrow represents the magnitude of the amount of deformation, so as to obtain the trend direction existing on each deformation region. It should be noted that these trend directions represent the direction in which the points with displacement in this region move, rather than representing the average trend of the entire region; even though each deformation region is obtained by aggregation through spatial continuity and similar deformation, there are still some points whose deformation trend directions are different from those of other points within the deformation region. At this time, the trend directions generated by all points are displayed to facilitate the subsequent combination of these points.

[0065] S42, fitting the trend directions of each deformation region along the edge of the guide dike, identifying the trend directions at the left and right edges of the guide dike, and using the trend directions at the left and right edges of the guide dike and the central trajectory during deformation of the guide dike to obtain the deformation trend direction corresponding to the deformation region.

[0066] For the central trajectory, it represents the center line of all points that generate deformation when deformation occurs. The coordinates of the center lines at all time points are fitted by the method of weighted averaging to obtain the currently used central trajectory.

[0067] When fitting the trend directions of each deformation region along the edge of the guide dike, mainly connect the points represented by these trend direction arrows along the edge of the guide dike under normal conditions to determine the trend that can be shown under the currently generated amount of deformation. The obtained main trend direction is used as the deformation trend direction.

[0068] When obtaining the deformation trend direction corresponding to the deformation region, the processing method for each deformation region can be to identify the deformation points on each deformation region, convert the amount of deformation of the deformation points into displacement vectors, that is, represent the relative displacement of the deformation points on the horizontal coordinate X-axis and Y-axis according to the displacement generated by the deformation points. Then, construct a covariance matrix for all deformation points within the deformation region, and then solve the eigenvectors of the covariance matrix. Use the arctangent function to represent the ratio of the corresponding data of the eigenvectors of the covariance matrix on the Y-axis and X-axis, and the trend direction corresponding to each deformation region can be obtained. This trend direction represents an angle value.

[0069] After that, project this trend direction onto the left and right edges of the guide dike, and use the projected edge points to calculate the projected angle. The calculation method of the projected angle value is the same as the calculation method of the trend direction corresponding to the deformation region, and multiple adjacent edge points are used for joint calculation; then, the angles of all edge points on the left and right edges are weighted and averaged to obtain the angle values corresponding to the left and right edges.

[0070] Then calculate the weighted average of the corresponding angles of each center point on the center line at adjacent time points as the center trajectory at this time. Finally, weighted sum the angle values of the center trajectory with the angle values corresponding to the left and right edges to obtain an overall angle value, which represents the deformation trend direction corresponding to the deformation area.

[0071] After obtaining the deformation trend direction, the main direction of the current guide dike during deformation can be known, and the deformation amount and deformation rate in this direction are analyzed to find out the main factors that mainly affect the deformation of the guide dike under the influence of multiple points.

[0072] Therefore, the implementation method of obtaining the influence paths corresponding to each deformation area in step S4 further includes: using the Moran index to extract the relevant points corresponding to the deformation amount and deformation rate in the deformation trend direction, and taking the shortest path of the relevant points as the output influence path.

[0073] For the Moran index, after setting weights for the positions corresponding to the deformation amount and deformation rate in the deformation trend direction, calculate their spatial correlation. The weight at each position is set according to the ratio of its deformation amount to the total deformation amount. When using the Moran index for calculation, the deformation amount and deformation rate are calculated separately; for example, compare the deformation amount of a single position with the average value of all deformation amounts to obtain the Moran index of the deformation amount. At this time, the calculation of the Moran index is for all positions with deformations existing in the deformation trend direction, and calculate the deformation amount and deformation rate at these positions; then calculate the deformation rate of a single position and compare it with the average value of all deformation rates to obtain the Moran index of the deformation rate. It should be noted that the weight used for the deformation rate is the ratio of its deformation amount to the total deformation amount; then add the Moran indices calculated for the deformation amount and deformation rate, and judge whether the added value is greater than the preset value. The preset value will adopt the average value of the Moran indices calculated from the deformation amount and deformation rate measured in the areas with large gradient changes, large settlement deformations, and surface cracks in the historical data; when it is greater than the preset value, set the position corresponding to the deformation amount and deformation rate as the relevant point. Finally, calculate the relevant points in the way of the shortest path, use the Dijkstra algorithm to connect the relevant points. The weight of each relevant point is the weight used when calculating its Moran index, and the value of each relevant point is the sum of the Moran indices calculated for its deformation amount and deformation rate. Then, according to the distance and weight of each relevant point in the spatial position, obtain the shortest path, and take the shortest path obtained at this time as the subsequent output influence path.

[0074] The shortest path obtained at this time will represent the path connected by the areas with obvious deformations in the deformation trend direction, which will represent the part with obvious displacement on the guide dike and how this part generates relevant displacements.

[0075] In an embodiment of the present invention, step S5 mainly performs rule matching on the obtained influence path, taking the deformation amount, deformation rate, path direction on the influence path, and the position corresponding to the influence path as input features, and then matching the type that needs to be alarmed at this time to determine whether subsequent processing is required for the current guide dike.

[0076] As Figure 5 shown, the implementation manner of step S5 includes: S51, separating the features of the influence path according to the deformation amount and deformation rate on the influence path, and performing rule matching between the separated features and a preset database; when separating the features, mainly dividing according to the values of the deformation amount and deformation rate to separate the parts with high deformation amount and low deformation amount for rule matching. When performing rule matching, match these feature values with the features set in the preset database and calculate their similarity. The similarity can be in the forms of Pearson correlation coefficient, cosine similarity, confidence level, and support degree, etc. After calculating the similarity of these separated features, consider the part with a similarity greater than 0.6 as a completed match, that is, when calculating the current Pearson correlation coefficient, cosine similarity, confidence level, support degree, etc., under the parameter conditions of specific requirements, it is required to calculate a value exceeding 0.6 to consider that there are features in the preset database that can match the current influence path.

[0077] S52, if there are matching features, take the alarm type represented by this feature as the subtype of the current influence path. After obtaining the matching features, determine the connection relationship of the subtypes according to the distribution of each subtype on the influence path, and execute the path operations of this alarm type in sequence according to the connection relationship of the subtypes.

[0078] For the representation of the connection relationship of subtypes in the guide dike deformation scenario, connect the alarm types in the order of their appearance according to the appearance time of the alarm types corresponding to the guide dike deformation, and execute the operations that should be performed after the appearance of the alarm types in sequence. If there are repeated operations, delete the repeated parts.

[0079] In S53, if no matching feature exists, the deformation trend direction corresponding to the influence path and the position corresponding to the influence path are used as the matching objects. The alarm type with the maximum similarity value after matching is used as the alarm type corresponding to the current influence path, and the operation event corresponding to this alarm type is executed. When no matching feature can be found, the angle corresponding to the change trend direction and the position where deformation occurs on the influence path are calculated with the content in the preset database to find the similar parts. At this time, the calculation method is the same as that in the matching calculation, and it can be represented in forms such as Pearson correlation coefficient, cosine similarity, confidence level, and support degree. Then, the alarm type with the maximum value is selected. This alarm type will contain data related to the deformation trend direction corresponding to the influence path and the position corresponding to the influence path in the preset database to obtain a calculated similarity. After that, the current guide dike is alarmed or other processing methods are carried out according to the operation event recorded on this alarm type.

[0080] For example, there is a monitoring system for a marine guide dike that has collected the deformation amount and deformation rate data for each month in the past year. In a specific month, abnormal deformation was detected on a section of the guide dike about 50 meters long, and the deformation amount in the central part was significantly higher than that at both ends. After analysis, the deformation amount in the central part of this section of the guide dike reached 10 cm / month, while that at both ends was only 2 cm / month.

[0081] At this time, according to the deformation amount and deformation rate of the influence path, the guide dike is divided into three parts: both ends are used as low-deformation areas, and the middle part is used as a high-deformation area.

[0082] By calculating the similarity with the historical events recorded in the preset database, it is found that the deformation mode in the middle part is closest to the "local structural failure" alarm type in the historical records, and its Pearson correlation coefficient is 0.75.

[0083] After confirming it as "local structural failure", the system automatically starts an emergency repair program and increases the real-time monitoring frequency of this area.

[0084] If no directly matching feature is found, then the deformation trend direction of this area will be further analyzed (for example, whether it shows inclination or sinking along a certain specific direction), and the position information will be combined to find the closest historical case, and then the handling of the relevant situation in the historical case will be used to deal with the handling method of the marine guide dike at this time.

[0085] Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention, and still be covered by the protection scope of the present invention.

Claims

1. An automatic monitoring and early warning method for offshore embankment deformation based on remote sensing images, characterized in that: include: S1, obtaining remote sensing images and standard images of the offshore dike, and composing the remote sensing images into a remote sensing image set according to the time and location of acquisition; the remote sensing image set includes satellite remote sensing images, UAV remote sensing images and SAR images; S2, dividing the remote sensing image set into a plurality of image pairs, performing image registration according to the positions of the plurality of image pairs and the standard image at the offshore guide embankment, and obtaining a registered target remote sensing image; S3, performing image difference on the target remote sensing images at different time points, analyzing deformation features related to the surface deformation of the offshore dike, and determining deformation areas related to the distribution of each deformation feature; When performing image differentiation, the satellite remote sensing image and the UAV remote sensing image are differentiated in pixel values ​​to obtain the texture features of the guide bank regarding texture and morphological changes; the intensity difference image is a SAR image that uses the standard deviation of pixels at different time points to describe the intensity value difference; S4, identifying the deformation amount and deformation rate on each deformation area, and determining the trend direction of each deformation area; fitting the trend direction of each deformation area according to the edge of the guide bank, identifying the trend direction of the guide bank at the left and right edges, using the trend direction of the guide bank at the left and right edges and the center trajectory of the guide bank during deformation, obtaining the deformation trend direction corresponding to the deformation area, performing influence analysis on the deformation amount and deformation rate in the deformation trend direction, and obtaining the influence path corresponding to each deformation area; S5: Determine the alarm type corresponding to the offshore dike based on the acquired impact path.

2. The method for automatic monitoring and early warning of offshore embankment deformation based on remote sensing images according to claim 1 is characterized in that: The implementation of step S2 includes: S21, obtaining the acquisition time of the remote sensing image set, taking adjacent frame images of the remote sensing image set within adjacent acquisition times as an image pair, and obtaining a plurality of image pairs; S22, preprocessing each image pair, extracting feature points from the preprocessed image pair and the standard image, and performing a nearest neighbor algorithm on the feature points to obtain feature point pairs of the matching image pair and the standard image; S23, after transforming the feature point pairs of the matching image pair and the standard image, the image pair satisfying the mean square error is output as the target remote sensing image.

3. The method for automatic monitoring and early warning of offshore embankment deformation based on remote sensing images according to claim 2 is characterized in that: When the image pair that satisfies the mean square error is output as the target remote sensing image, the following implementation methods are also included: According to the images corresponding to the plurality of image pairs, a first image region and a second image region are sequentially acquired, wherein the first image region represents the region represented by the initial frame of the remote sensing image set in the image pair, and the second image region represents the region represented by the intermediate frame of the remote sensing image set in the image pair; Detecting the similarity between the first image region and the second image region, and determining the registration degree of the target remote sensing image according to the similarity between the first image region and the second image region; When the similarity between the first image area and the second image area is greater than the similarity threshold, and the registration degree of the target remote sensing image satisfies the mean square error, the corresponding target remote sensing image is output.

4. The method for automatic monitoring and early warning of offshore embankment deformation based on remote sensing images according to claim 1 is characterized in that: The implementation of step S3 includes: S31, performing image difference on the target remote sensing images at different time points to obtain a morphological difference image and an intensity difference image respectively; S32, extracting edge features of the guide bank edge in the morphological difference image to identify the morphological change area on the guide bank surface; S33, identifying the deformation amount and deformation rate of the guide bank according to the intensity difference image, and generating a deformation field of the intensity difference image according to the deformation amount and deformation rate; S34, superimposing the morphological change region and the deformation field to obtain a superimposed deformation feature.

5. The method for automatic monitoring and early warning of offshore embankment deformation based on remote sensing images according to claim 4 is characterized in that: Step S33 also includes: interfering the intensity difference images at multiple different time points to generate an interference pattern corresponding to the intensity difference image, removing the terrain phase from the part of the interference pattern that produces a phase difference, and generating multiple interference pairs, and identifying the deformation amount and deformation rate of the guide embankment in turn according to the phase difference on the interference pairs.

6. The method for automatic monitoring and early warning of offshore embankment deformation based on remote sensing images according to claim 1 is characterized in that: The implementation method of determining the deformation area related to the distribution of each deformation feature also includes the following: Taking the position of any deformation feature as the center position, the difference between adjacent deformation features at the center position is compared; According to the horizontal offset of the difference between adjacent deformation features, the deformation features are clustered and combined into multiple deformation regions.

7. The method for automatic monitoring and early warning of offshore embankment deformation based on remote sensing images according to claim 6 is characterized in that: The implementation methods of clustering deformation features and combining them into multiple deformation regions include: The deformation features are converted into horizontal displacement vectors and corresponding spatial coordinates, the spatial distance between each deformation feature is calculated, and it is determined whether there are adjacent deformation features that meet the spatial adjacency at the center position. When the spatial adjacency is met and the difference between adjacent deformation features is less than the deformation difference threshold, the adjacent deformation features and the center position are considered to be in the same area; then all deformation features that meet the spatial adjacency and deformation difference thresholds are recursively calculated to obtain the deformation area.

8. The method for automatic monitoring and early warning of offshore embankment deformation based on remote sensing images according to claim 1 is characterized in that: The implementation method of obtaining the influence path corresponding to each deformation area in step S4 also includes: The Moran index is used to extract the relevant points corresponding to the deformation amount and deformation rate in the deformation trend direction, and the shortest path of the relevant points is used as the output influence path.

9. The method for automatic monitoring and early warning of offshore embankment deformation based on remote sensing images according to claim 1 is characterized in that: The implementation of step S5 includes: S51, separating the features of the influencing path according to the deformation amount and deformation rate on the influencing path, and matching the separated features with the preset database by rules; S52, if there is a matching feature, the alarm type represented by the feature is used as a subtype of the current influencing path, the subtypes of the current influencing path are distributed according to the distribution of each subtype on the influencing path, the connection relationship of the subtypes is determined, and the path operation of the alarm type is performed in sequence according to the connection relationship of the subtypes; S53, if there is no matching feature, the deformation trend direction and the position corresponding to the influencing path are taken as matching objects, the alarm type with the largest similarity value after matching is taken as the alarm type corresponding to the current influencing path, and the operation event corresponding to the alarm type is executed.

Images