Earth surface deformation hidden danger point automatic identification method based on equal deformation line

Abstract

The invention discloses a ground surface deformation hidden danger point automatic identification method based on an equal deformation line. The method specifically comprises the following steps of 1, presetting a minimum equal deformation number threshold value included in the equal deformation line; 2, generating an equal deformation line with a numerical difference of S in the earth surface deformation rate diagram; 3, calculating the area of a region surrounded by each equal deformation line and the number of the equal deformation lines included in each equal deformation line; step 4, determining candidate hidden danger points based on the area of the area surrounded by each isomorphic line and the number of the isomorphic lines included in each isomorphic line; 5, if other candidate hidden danger points exist in the candidate hidden danger points, selecting an area surrounded by the candidate hidden danger points on the outermost layer as candidate deformation hidden danger points; and 6, calculating the maximum deformation rate in the candidate deformation hidden danger points, and determining a final deformation hidden danger point based on the maximum deformation rate.

Description

Technical Field

[0001] This invention belongs to the field of remote sensing geological disaster prevention and control, and in particular relates to an automatic identification method for potential surface deformation points based on isomorphic lines. Background Technology

[0002] Surface deformation is a localized disruption of surface stability caused by factors such as precipitation, earthquakes, engineering construction, underground resource extraction, or underground space utilization. It is one of the important characteristics of landslides and ground subsidence before they evolve into disasters. Conducting surface deformation hazard identification based on deformation monitoring data can provide a scientific basis for decision-making in the prevention of landslides and ground subsidence disasters.

[0003] Traditional methods for monitoring surface deformation typically involve deploying leveling monitoring points, Global Navigation Satellite System (GNSS) monitoring points, or underground inclinometers. While these methods offer high accuracy and fast sampling rates, they require on-site equipment installation, have limited monitoring range, and low spatial resolution. High and steep mountainous areas, often inaccessible to personnel, often fail to receive effective monitoring, potentially leading to unforeseen disasters. Optical imagery-based methods for surface deformation identification and monitoring utilize the shape, texture, and spectral characteristics of deformation, employing visual interpretation, change detection, machine learning, and object-oriented classification to achieve remote sensing detection. However, these methods are susceptible to cloud cover and fog. Digital elevation models (DEMs) and orthophotos generated using airborne radar (LIDAR) can extract detailed geological environmental information about disaster-prone areas. These methods offer advantages such as flexible operation, the ability to penetrate vegetation, and lack of geographical limitations. However, their deformation monitoring accuracy is only at the centimeter level or even lower, and identifying large-scale, slow-moving deformation presents technical challenges. Interferometric Synthetic Aperture Radar (InSAR) technology, by interferometrically processing the phase information of two Synthetic Aperture Radar (SAR) images captured at different times, can obtain minute surface deformations. It offers advantages such as remote sensing detection, minimal impact from clouds and rain, high spatial resolution, wide coverage, and high monitoring accuracy, and has been widely applied in the field of surface deformation monitoring. Early identification of potential deformation hazards using InSAR deformation monitoring data relied primarily on expert experience, visually interpreting the fringe patterns of the average deformation rate map of coherent targets—a labor-intensive and inefficient process. In recent years, with the launch and operation of new-generation SAR satellites, ideal orbital control has ensured abundant observational data across the global landmass, leading to a continuous stream of large-scale, wide-area InSAR surface deformation monitoring results. Given such a large monitoring scope, manually interpreting deformation hazard points would be extremely labor-intensive and impractical. The method of identifying deformation hazard points by setting a threshold for the deformation rate of coherent targets only identifies areas near the deformation center with a deformation rate greater than the threshold as hazard areas. It misses areas outside the deformation center with a deformation rate less than the threshold, resulting in fragmented and incomplete deformation zones. In recent years, deep learning has gradually become an important tool for automatic deformation hazard identification due to its powerful image learning, analysis, and prediction capabilities. However, the significant variations in the magnitude and morphology of surface deformation pose a considerable challenge to manually delineating training samples, resulting in a current lack of deep learning training sample sets for deformation hazard point identification based on InSAR deformation data, leading to an imbalanced sample problem. Summary of the Invention

[0004] Purpose of the invention: In order to solve the problems existing in the prior art, the present invention provides an automatic identification method for potential surface deformation points based on isotropic lines.

[0005] Technical solution: This invention provides an automatic identification method for potential surface deformation points based on isotropic lines, specifically including the following steps:

[0006] Step 1: Preset a threshold N for the minimum number of uniformly deformable lines contained within a uniformly deformable line. th Based on N th Set the step size S;

[0007] Step 2: Generate isodeformation lines with a numerical difference of S in the surface deformation rate map;

[0008] Step 3: Calculate the area enclosed by each isomorphic line and the number of isomorphic lines contained within each isomorphic line;

[0009] Step 4: Determine candidate potential hazard points based on the area enclosed by each isomorphic line and the number of isomorphic lines contained within each isomorphic line;

[0010] Step 5: If there are other candidate potential hazards among the candidate potential hazards, then select the area surrounded by the outermost candidate potential hazards as the candidate deformation potential hazard point;

[0011] Step 6: Calculate the maximum deformation rate among the candidate deformation hazard points, and determine the final deformation hazard point based on the maximum deformation rate.

[0012] Furthermore, in step one, the step size S is calculated using the following formula:

[0013] S = V th / (N th +2)

[0014] Among them, V th N is the preset deformation rate threshold. th This is the preset threshold for the number of isomorphic lines.

[0015] Furthermore, the method also includes removing interference lines between steps two and three. Specifically, the average distance between two adjacent lines is calculated. If the average distance between two adjacent lines is greater than or equal to a preset threshold, one line is randomly deleted between the two adjacent lines until the average distance between two adjacent lines is greater than the preset threshold.

[0016] Furthermore, step four specifically involves the following two conditions being met: Candidate potential hazards must satisfy the following conditions:

[0017] (1) The area of ​​the region enclosed by the isomorphic lines of the candidate hidden danger point is greater than the preset minimum area threshold and less than the preset maximum area threshold.

[0018] (2) The number of isomorphic lines contained within the isomorphic lines of the candidate hazard point is greater than or equal to N. th .

[0019] Furthermore, the minimum area threshold and the maximum area threshold are determined based on the crustal deformation monitoring specifications based on InSAR technology and after considering redundancy.

[0020] Furthermore, in step six, candidate deformation hazard points with a maximum deformation rate greater than or equal to the deformation rate threshold are selected as the final deformation hazard points.

[0021] Furthermore, the deformation rate threshold is determined based on the crustal deformation monitoring specifications based on InSAR technology and after considering redundancy.

[0022] A computer device includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to implement the automatic identification method for surface deformation hazard points based on isotropic lines.

[0023] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the automatic identification method for surface deformation hazard points based on isotropic lines.

[0024] Beneficial Effects: This invention utilizes surface deformation rate data acquired through InSAR technology to generate isomorphic lines. By analyzing the characteristics of these isomorphic lines, and using deformation area, deformation magnitude, and deformation morphology as criteria for identifying potential deformation points, it automatically identifies potential surface deformation points, delineates their boundaries, and provides information such as area and deformation magnitude. This invention has a clear underlying mechanism, requires no training samples, can suppress noise interference, requires few and easily determined input parameters, and can automatically, comprehensively, and completely identify potential deformation points, delineate their boundaries, and provide information such as area and deformation magnitude, thus improving the efficiency and accuracy of potential deformation point identification. Attached Figure Description

[0025] Figure 1 This is a schematic diagram of geological disaster patterns;

[0026] Figure 2 This is a flowchart of the method of the present invention;

[0027] Figure 3 This is a deformation rate diagram of the simulated region in this invention;

[0028] Figure 4 This is a schematic diagram of the isomorphic lines of the simulated region;

[0029] Figure 5 This is a comparison chart of the recognition results of the method of the present invention and the threshold segmentation method;

[0030] Figure 6 Figure 1 is a schematic diagram of the Xining test data study area, where Figure (a) is a panoramic view of the study area; Figure (b) is a distribution map of GNSS stations in the Nanshan landslide monitoring area.

[0031] Figure 7 This is the result of the deformation hazard identification in the Xining test experiment. Detailed Implementation

[0032] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0033] Surface deformation is a localized disruption of surface stability. Table 1 shows some typical deformation field characteristics given in the "Specification for Monitoring Crustal Deformation Based on InSAR Technology (GB / T 44146-2024)". As can be seen from Table 1, surface deformation is characterized by large magnitude but limited area. Furthermore, such as... Figure 1 As shown, surface deformation also exhibits three-dimensional morphological characteristics such as funnel-shaped (ground subsidence or collapse), tongue-shaped (mudslide), and chair-shaped (landslide).

[0034] Table 1

[0035]

[0036] Isomorphic lines can characterize the level of deformation, calculate the area of ​​the enclosed region, and display the three-dimensional morphology, effectively reflecting the characteristics of surface deformation. Therefore, this invention proposes to generate isomorphic lines using surface deformation rate data obtained through InSAR technology. By analyzing the isomorphic lines, areas enclosed by isomorphic lines with deformation rates greater than a certain threshold, areas within a certain range, and exhibiting a certain three-dimensional morphology (i.e., containing a certain number of layers (lines) of isomorphic lines within the range) are identified as potential hazard points. The method flow is as follows: Figure 2 As shown. The surface deformation rate data acquired by InSAR technology is preprocessed to remove non-values; the absolute value of the deformation rate is taken to reduce algorithm complexity. Given the area threshold, deformation rate threshold, and threshold for the number of layers containing isomorphic lines, the isomorphic line step size is calculated according to equation (1):

[0037] S = V th / (N th +2)

[0038] Among them, V th N is the preset deformation rate threshold. th The preset threshold for the number of equal deformation line layers (N in this embodiment)th =2), generate isomorphic lines with a step size S, calculate the average spacing between two adjacent isomorphic lines, and remove isomorphic lines with a spacing less than x pixels (x is a preset pixel threshold, x = 1 in this embodiment), until the average spacing of all isomorphic lines is greater than or equal to x pixels. Calculate the area of ​​the region enclosed by each isomorphic line and the number of isomorphic lines contained within the isomorphic line; when the area enclosed by an isomorphic line is greater than the minimum area threshold and less than the maximum area threshold, and the number of isomorphic line layers contained within it is greater than or equal to the threshold N for the number of isomorphic line layers. th If the deformation lines are in the specified range, the area enclosed by these deformation lines is identified as a candidate deformation hazard point. The inclusion relationship of the deformation lines among the candidate deformation hazard points is analyzed, and the outermost layer of candidate deformation hazard points is selected, with the maximum deformation rate within its range calculated. Based on the deformation rate threshold, if the maximum deformation rate within the outermost layer of candidate deformation hazard points exceeds the threshold, the area is determined to be a deformation hazard point. The location, area, and deformation rate of the hazard point are output.

[0039] This invention removes isomorphic lines with a spacing of less than one pixel and sets a threshold for the number of isomorphic line layers, requiring that candidate deformation hazard points contain multiple layers of isomorphic lines. This ensures that candidate deformation hazard points have a certain three-dimensional shape and avoids the influence of multiple consecutive noise points (such as atmospheric delay or orbital error).

[0040] An embodiment of the present invention: utilizing as Figure 3 The simulated data shown illustrates the calculation process of the invention method and analyzes its advantages over the threshold segmentation method. Figure 3 As can be seen, the simulation data includes four deformation zones: A, B, C, and D. The statistical information and simulation scenarios for each zone are shown in Table 2.

[0041] Table 2

[0042]

[0043] The simulated deformation rate data were analyzed using the threshold parameters listed in Table 3, and the results are as follows: Figure 4 And Table 4. Figure 4 The white lines represent isomorphic lines with an average distance of less than one pixel, the yellow lines represent isomorphic lines with an average distance of one pixel or more, and the red lines represent potential deformation points. Figure 4As can be seen, area A meets all the threshold conditions, and the area enclosed by the red line is identified as a potential hazard point. Furthermore, the hazard point information (Table 4) is basically consistent with the true values ​​listed in Table 2. Area B has an area smaller than the minimum area threshold; area C has a relatively low overall deformation rate, resulting in fewer generated isomorphic lines, and therefore does not meet the threshold condition for the number of isomorphic line layers. Thus, these two areas were not identified as potential hazard points. Since an interpolation algorithm is used to generate isomorphic lines, multiple isomorphic lines were generated at the edge of area D. By calculating the average distance between adjacent isomorphic lines, isomorphic lines with an average distance less than one pixel were removed, ultimately retaining only the outermost yellow isomorphic line. This causes area D to not meet the threshold condition for the number of isomorphic line layers. Therefore, areas like area D, which do not possess deformation morphology characteristics, were not identified as potential deformation hazard points.

[0044] Table 3

[0045]

[0046]

[0047] Table 4

[0048]

[0049] The identification results of the method proposed in this invention are compared with those of the threshold segmentation method. The deformation rate threshold in the threshold segmentation method is the same as that in the method of this invention, set at 10 mm / yr. The results are as follows: Figure 5 As shown. Figure 5 The red line represents the recognition result of the method of this invention, and the black line represents the recognition result of the threshold segmentation method. Figure 5 It can be seen that the threshold segmentation method only identifies areas in region A with a deformation rate greater than the threshold as potential hazards, omitting small deformation areas around region A, and its identification range is less complete than that of the method of this invention. The threshold segmentation method also identifies regions B and D as potential hazards, but during manual visual interpretation, areas with relatively high deformation rates but small areas (such as region B) are usually identified as noise; and areas with relatively high deformation rates but no deformation morphological characteristics (such as region D) are identified as untangling, atmospheric, or orbital errors. Tests demonstrate that the method of this invention is superior to the threshold segmentation method in maintaining the overall morphological characteristics of the deformation area and suppressing the effects of noise and untangling errors.

[0050] This embodiment is tested using actual data from Xining City, such as... Figure 6As shown in Figure (a) (the red box indicates the Nanshan landslide monitoring area), Xining City is located in the northeastern part of Qinghai Province, situated in the Huangshui Valley area between the Qinghai-Tibet Plateau and the Longxi Loess Plateau. The terrain is high in the southeast and low in the northwest, surrounded by mountains. High terraces and hills are distributed along both sides of the urban area of ​​Xining City. Geological tectonic movements are active, loess is widely distributed, and the slope rock and soil structure has poor stability, making it prone to geological disasters such as landslides, debris flows, and landslides. Stacking technology was used to process 73 C-band, wide-swath interferometric, VV-polarized Sentinel-1A satellite deorbiting data covering Xining City, Qinghai Province, from January 7, 2018 to June 13, 2020, to obtain a surface deformation rate map with a spatial resolution of 15m × 18m. The Xining City Nanshan landslide automated monitoring project (site locations are shown in the figure) was employed. Figure 6 (b) The GNSS data collected was used to verify the accuracy of the InSAR deformation rate, showing that it has high accuracy and reliability.

[0051] The threshold values ​​for each item in this invention are determined according to the "Specification for Crustal Deformation Monitoring Based on InSAR Technology (GB / T 44146-2024)" with a certain degree of redundancy. The SAR data used to obtain the surface deformation rate in Xining City spans approximately 2.43 years (January 7, 2018 to June 13, 2020). Referring to Table 1, the typical landslide precursor deformation is greater than 50 mm, which is equivalent to a deformation rate greater than 20.58 mm / yr (50 mm / 2.43 years). Combining this with Table 1, the typical ground subsidence deformation rate is greater than 20 mm / yr, and considering a certain degree of redundancy, a deformation rate threshold of 15 mm / yr is taken. Referring to the spatial distribution range of typical deformation fields in Table 1, a maximum area threshold of 100 km² is taken. 2 Minimum area threshold: 0.1km 2 The larger the threshold for the number of layers of isomorphic lines, the smaller the step size of the isomorphic lines calculated according to equation (1), and the greater the computational load of the isomorphic lines. Taking a threshold of 2 layers of isomorphic lines can ensure that the deformation zone has a certain shape and the computational load is moderate. Based on the above analysis, the threshold parameters of the Xining test experiment in this embodiment are shown in Table 5.

[0052] Table 5

[0053] parameter Value Including the threshold of the number of isomorphic line layers 2 Deformation rate threshold 15mm / yr Maximum area threshold <![CDATA[100km 2 ]]> Minimum area threshold <![CDATA[0.1km 2 ]]>

[0054] Using the parameters listed in Table 5, the deformation rate data of Xining were analyzed using the method of this invention, and the identification results are as follows: Figure 7 As shown in Table 6, detailed information on some potential deformation points is provided.

[0055] Table 6

[0056]

[0057] Depend on Figure 7As can be seen, the method proposed in this invention correctly identified all 43 potential deformation points. Among them, point No. 1 has the largest area, reaching 1.79 km². 2 The deformation rate reached -80.95 mm / yr; the area of ​​the hazard point No. 43 was the smallest, reaching 0.10 km². 2 The deformation rate reached -26.89 mm / yr. Figure 7 Regions A, B, and C in the upper right corner were not identified as potential hazards because their maximum deformation rate was less than the deformation rate threshold, and region D was not identified because its area was less than the area threshold. While reducing the deformation rate and area thresholds can enhance the method's ability to identify small deformations, it can also generalize the results, rendering the hazard identification process meaningless. As deformation continues to develop, the magnitude and area of ​​deformation gradually increase, and these previously undetected small deformations will be identified in subsequent observations.

[0058] It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the present invention will not describe the various possible combinations separately.

Claims

1. An automatic identification method for potential surface deformation hazards based on isomorphic lines, characterized in that, Specifically, the steps include the following: Step 1: Preset a threshold N for the minimum number of uniformly deformable lines contained within a uniformly deformable line. th Based on N th Set the step size S; Step 2: Generate isodeformation lines with a numerical difference of S in the surface deformation rate map; Step 3: Calculate the area enclosed by each isomorphic line and the number of isomorphic lines contained within each isomorphic line; Step 4: Determine candidate potential hazard points based on the area enclosed by each isomorphic line and the number of isomorphic lines contained within each isomorphic line; Step 5: If there are other candidate potential hazards among the candidate potential hazards, then select the area surrounded by the outermost candidate potential hazards as the candidate deformation potential hazard point; Step 6: Calculate the maximum deformation rate among the candidate deformation hazard points, and determine the final deformation hazard point based on the maximum deformation rate.

2. The method for automatic identification of surface deformation hazard points based on isomorphic lines according to claim 1, characterized in that, In step one, the step size S is calculated using the following formula: SV th / (N th +2) Among them, V th N is the preset deformation rate threshold. th This is the preset threshold for the number of isomorphic lines.

3. The method for automatic identification of surface deformation hazard points based on isomorphic lines according to claim 1, characterized in that, The method further includes removing interference lines between steps two and three. Specifically, it calculates the average distance between two adjacent lines. If the average distance between two adjacent lines is greater than or equal to a preset threshold, it randomly deletes one line between the two adjacent lines until the average distance between two adjacent lines is greater than the preset threshold.

4. The method for automatic identification of surface deformation hazard points based on isomorphic lines according to claim 1, characterized in that, Step four specifically involves: the candidate potential hazard points meeting the following two conditions: (1) The area of ​​the region enclosed by the isomorphic lines of the candidate hidden danger point is greater than the preset minimum area threshold and less than the preset maximum area threshold. (2) The number of isomorphic lines contained within the isomorphic lines of the candidate hazard point is greater than or equal to N. th .

5. The method for automatic identification of surface deformation hazard points based on isomorphic lines according to claim 4, characterized in that, The minimum area threshold and the maximum area threshold are determined based on the crustal deformation monitoring specifications based on InSAR technology and after considering redundancy.

6. The method for automatic identification of surface deformation hazard points based on isomorphic lines according to claim 1, characterized in that, In step six, candidate deformation hazard points with a maximum deformation rate greater than or equal to the deformation rate threshold are selected as the final deformation hazard points.

7. The method for automatic identification of surface deformation hazard points based on isomorphic lines according to claim 6, characterized in that, The deformation rate threshold is determined based on the crustal deformation monitoring standard based on InSAR technology and after considering redundancy.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, When the processor executes the computer program, it implements the automatic identification method for potential surface deformation hazards based on isotropic lines as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the automatic identification method for potential surface deformation hazards based on isotropic lines as described in any one of claims 1 to 7.

Images