InSAR intelligent deformation monitoring methods, systems, and storage media focusing on nuclear power plant areas

Abstract

This invention discloses an intelligent InSAR deformation monitoring method, system, and storage medium focusing on nuclear power plant areas, belonging to the field of synthetic aperture radar imaging technology. The method includes: extracting a target area from map data to obtain a range planar map; acquiring SAR images at the start and end time points, cropping the SAR images to obtain a first image; stitching the first image together to obtain a second image; correcting and removing thermal noise from the second image to obtain a first complex image and a second complex image; registering and overlaying the first complex image and the second complex image, and performing phase unwrapping to obtain a terrain deformation map; dividing the range planar map into multiple sub-regions, assigning corresponding attention levels to the sub-regions based on historical deformation data; dividing the terrain deformation map into multiple sub-regions, and analyzing and interpreting the sub-regions sequentially according to their attention levels to obtain deformation information for each sub-region. This invention enables fast and accurate monitoring of terrain deformation.

Description

Technical Field

[0001] This invention belongs to the field of synthetic aperture radar imaging technology, specifically relating to an InSAR intelligent deformation monitoring method, system, and storage medium for focusing on nuclear power plant areas. Background Technology

[0002] As an important energy supply method, the safety and stability of nuclear power plants are key factors in ensuring their normal operation. During the planning, design, construction, operation, and maintenance phases of a nuclear power plant, surface deformation or geological disasters caused by natural factors or human activities may pose potential risks to its safe operation. Therefore, real-time monitoring of the nuclear power plant area is a prerequisite for safe production.

[0003] Currently, deformation monitoring mainly utilizes technologies such as manual instrument measurement, laser scanning measurement, and automated deformation monitoring. For example, Chinese patent document CN110888353B discloses a method for real-time monitoring of surface subsidence in coal mining areas that integrates multiple sensors. This method determines candidate physicochemical sensors based on monitoring indicators, then selects a network topology layout scheme based on on-site surveys; subsequently, it designs a sensor network integration scheme based on the layout scheme to achieve hierarchical integration; finally, it divides the communication methods of the multi-level sensor network and deploys them on-site, thereby achieving real-time monitoring and early warning of surface subsidence in mining areas.

[0004] However, automated monitoring relies on various technologies and software, including sensor technology, data transmission and storage technology, and data processing and analysis software. Technical malfunctions or software update issues can impact the operation of the monitoring system. Furthermore, traditional manual instrument measurement methods use instruments such as theodolites, levels, and total stations for direct measurement. These methods often depend on weather conditions and environmental factors, and their accuracy is significantly affected by human error. For large-scale and long-term monitoring, they suffer from high labor costs and low measurement efficiency. Laser scanning measurement requires direct illumination of the target object's surface and receiving the reflected laser signal. If the target object has obstructions or invisible areas, the laser cannot directly measure it, resulting in inaccurate data acquisition in obstructed areas. Moreover, laser scanning instruments are expensive and have limited measurement range. Therefore, a more accurate terrain monitoring method that is unaffected by human and environmental factors is needed. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides an InSAR intelligent deformation monitoring method, system, and storage medium focusing on nuclear power plant areas, aiming to provide a more accurate solution for monitoring terrain deformation, unaffected by human or environmental factors.

[0006] To achieve the aforementioned objectives, this invention proposes an InSAR intelligent deformation monitoring method focusing on nuclear power plant areas, comprising:

[0007] Extract map data, and extract the target area from the map data to obtain a range plan map;

[0008] Set a start time point and an end time point, acquire multiple SAR images taken at the start time point and the end time point, crop the SAR images, and obtain a first image of the SAR images corresponding to the target area;

[0009] By stitching together the first images taken at the same time point, a second image containing the complete target region is obtained;

[0010] The second image is corrected and thermal noise is removed to obtain a first complex image at the start time point and a second complex image at the end time point;

[0011] The first complex image and the second complex image are registered and superimposed to generate an interferogram. The interferogram is then unwrapped in phase to obtain a terrain deformation map.

[0012] Identify the range planar map, divide the range planar map into multiple sub-regions, record the division method, obtain historical deformation data for each sub-region, and assign a corresponding attention level to the sub-region based on the historical deformation data;

[0013] Based on the aforementioned division method, the terrain deformation map is divided into multiple sub-regions, and the sub-regions are analyzed and interpreted sequentially according to the aforementioned attention level to obtain deformation information for each sub-region. The terrain deformation map and the deformation information are then visualized.

[0014] Further, dividing the range plan view into multiple sub-regions includes the following steps:

[0015] Filter the building areas in the range plan view, and perform contour correction on the building areas to obtain the corrected area;

[0016] Multiple basic shapes are preset, each of which corresponds to a building type. The basic shape corresponding to the correction area is matched to obtain the building type of the correction area. The correction area is expanded so that the expanded correction area is seamlessly connected. The expanded correction area is defined as the sub-region.

[0017] Furthermore, before assigning the attention level, the following steps are also included:

[0018] Set a monitoring duration, acquire deformation data of the sub-region within multiple consecutive monitoring durations, and extract first data, second data and third data from them, wherein the first data is the deformation data with the smallest value, the second data is the deformation data with the largest value, and the third data is the deformation data of the last monitoring duration;

[0019] Obtain the cumulative deformation at the end of the last monitoring period, and expand the first data, the second data, and the third data by a predetermined multiple to obtain the first calibration value, the second calibration value, and the third calibration value. Add the cumulative deformation to the first calibration value, the second calibration value, and the third calibration value to obtain the first predicted value, the second predicted value, and the third predicted value.

[0020] Establish a coordinate system and mark the starting point, the first prediction point, the second prediction point, and the third prediction point in the coordinate system. The horizontal coordinates of the starting point, the first prediction point, the second prediction point, and the third prediction point are respectively the third data, the first data, the second data, and the third data, and the vertical coordinates are respectively the cumulative deformation, the first predicted value, the second predicted value, and the third predicted value.

[0021] By connecting the starting point, the first prediction point, the second prediction point, and the third prediction point, an evaluation profile is obtained.

[0022] Furthermore, assigning the attention level to the sub-region includes the following steps:

[0023] Mark the first hazard value and the second hazard value on the Y-axis of the coordinate system. Generate a first boundary line and a second boundary line parallel to the X-axis based on the first hazard value and the second hazard value. Obtain the first area S1 of the evaluation contour below the first boundary line and the second area S2 above the second boundary line. Calculate the concern value α based on the first formula, which is: Wherein, ε1 and ε2 are the first preset value and the second preset value, respectively;

[0024] A corresponding adjustment weight is set for each of the building types. Based on the building types included in the sub-region, the attention value is adjusted using the corresponding adjustment weight to obtain the importance value of the sub-region. Multiple grade ranges are set, and based on the grade range in which the importance value is located, the corresponding attention level is assigned to the sub-region.

[0025] Furthermore, the contour correction of the building area includes the following steps:

[0026] Locate the vertices in the building area, connect the vertices to obtain a first boundary line, define the direction perpendicular to the first boundary line as a first direction and a second direction, move the first boundary line along the first direction by a preset distance to obtain a first adjustment line, move the first boundary line along the second direction by the preset distance to obtain a second adjustment line, and calculate the evaluation values ​​of the first boundary line, the first adjustment line, and the second adjustment line.

[0027] Set a first limit and a second limit, and repeatedly move the first boundary line until the first limit and the second limit are reached. Define the boundary line corresponding to the maximum evaluation value during the movement as the first correction line.

[0028] Locate the midpoint of the first correction line, rotate the first correction line with the midpoint as the center and a preset angle as the step size, calculate the evaluation value of the first correction line after each rotation, and after one rotation, stop the first correction line at the position with the maximum evaluation value.

[0029] The first correction lines with an included angle greater than the first threshold and adjacent to each other are merged to obtain the second correction line. The first correction line and the second correction line are then trimmed or extended to correct the outline of the building area into a closed shape.

[0030] Further, calculating the evaluation value includes the following steps:

[0031] Multiple observation points are located on the boundary line. A judgment region is generated with each observation point as the center. Pixels in the judgment region are matched pairwise based on the straight lines containing the first and second directions to obtain N pixel pairs. A reasonable value β for each observation point is calculated using a second formula, which is: Where, d n1 and d n2 p represents the distances between the first pixel and the observation point in the nth pixel pair, and the distances between the second pixel and the observation point, respectively. n1 and p n2 These are the pixel values ​​of the first and second pixels in the nth pixel pair, respectively;

[0032] The evaluation value γ of the first boundary line is calculated using a third formula, which is: Where M is the number of observation points in the boundary line, α m The reasonable value is the m-th observation point in the first boundary line.

[0033] Furthermore, based on the number and curvature of the vertices in the modified region as features, the basic shape is matched using the Hausdorff distance as a criterion.

[0034] Further, correcting the second image includes the following steps:

[0035] Acquire orbital data, establish an orbital correction model based on the orbital data, and correct the orbital error of the second image based on the orbital correction model.

[0036] This invention also provides an intelligent InSAR deformation monitoring system focusing on nuclear power plant areas. This system is used to implement the aforementioned intelligent InSAR deformation monitoring method focusing on nuclear power plant areas. The system includes:

[0037] The database module is used to extract map data, extract target areas from the map data, and obtain a range planar map;

[0038] The processing module includes a set start time point and an end time point. The processing module acquires multiple SAR images taken at the start time point and the end time point, crops the SAR images to obtain a first image corresponding to the target area, stitches together the first images at the same time point to obtain a second image containing the complete target area, corrects and removes thermal noise from the second image to obtain a first complex image at the start time point and a second complex image at the end time point, registers and superimposes the first complex image and the second complex image to generate an interferogram, and performs phase unwrapping on the interferogram to obtain a terrain deformation map.

[0039] The identification module is used to identify the range planar map, divide the range planar map into multiple sub-regions, record the division method, obtain historical deformation data of each sub-region, and assign a corresponding attention level to the sub-region based on the historical deformation data;

[0040] The interpretation module divides the terrain deformation map into multiple sub-regions based on the division method, and analyzes and interprets the sub-regions sequentially according to the attention level to obtain the deformation information of each sub-region, and then visualizes the terrain deformation map and the deformation information.

[0041] The present invention also discloses a computer storage medium, characterized in that the computer storage medium stores program instructions, wherein the program instructions, when executed, control the device where the computer storage medium is located to execute the method described above.

[0042] Compared with the prior art, the beneficial effects of the present invention are at least as follows:

[0043] This invention allows for remote sensing monitoring of the Earth's surface using radar sensors mounted on satellites or aircraft, eliminating the need for on-site contact with the monitored objects. This avoids interference and risks to the area surrounding nuclear power plants and reduces human error. InSAR technology can cover a large area of ​​the Earth's surface; a single monitoring session can encompass the entire area surrounding a nuclear power plant, providing comprehensive monitoring data that helps in the overall assessment of geological and surface changes in the region. InSAR technology is unaffected by weather conditions and lighting, enabling monitoring of target areas at any time and under any weather conditions, ensuring continuous monitoring and data acquisition of the area surrounding nuclear power plants.

[0044] This invention analyzes sub-regions one by one, prioritizing the analysis of sub-regions with higher importance, thereby obtaining deformation information of important areas as quickly as possible. In addition, this method not only ensures analysis time but also has low requirements for computing power. Attached Figure Description

[0045] Figure 1 This is a flowchart illustrating the steps of the InSAR intelligent deformation monitoring method for nuclear power plant areas described in this invention.

[0046] Figure 2 A schematic diagram illustrating the principle of the evaluation profile described in this invention;

[0047] Figure 3 This is a schematic diagram illustrating the process of contour correction of a building area according to the present invention.

[0048] Figure 4 This is a schematic diagram illustrating the division of the judgment region according to the present invention;

[0049] Figure 5 This is a schematic diagram of the InSAR intelligent deformation monitoring system for nuclear power plant areas, which is the focus of this invention. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0051] It is understood that the terms "first," "second," etc., used in this application may be used herein to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are used only to distinguish one element from another. For example, without departing from the scope of this application, a first script may be referred to as a second script, and similarly, a second script may be referred to as a first script.

[0052] like Figure 1As shown, the InSAR intelligent deformation monitoring method focusing on nuclear power plant areas includes:

[0053] Step S1: Extract map data, extract the target area from the map data, and obtain a range planar map.

[0054] Specifically, the first step is to obtain optical map data including nuclear power plants from the database, then delineate the target area from it to obtain a planar map that only includes the nuclear power plant area. Subsequently, the SAR image can be cropped based on the planar map.

[0055] Step S2: Set the start time point and the end time point, acquire multiple SAR images taken at the start time point and the end time point, crop the SAR images, and obtain the first image of the target area in the SAR images.

[0056] This embodiment acquires multiple images of the target area from the Sentinel-1 satellite, setting the start time to March 1st and the end time to March 15th. Multiple SAR images taken on March 1st are then acquired. The SAR images are then compared with the area plan to determine which parts of the target area are included in the SAR images. For example, if SAR image 1 includes the left half of the target area and SAR image 2 includes the right half, then the left half of SAR image 1 and the right half of SAR image 2 are cropped out as the first image. The image processing method at the end time point is the same as at the start time point, and will not be described here.

[0057] Step S3: Stitch together the first images at the same time point to obtain a second image containing the complete target area.

[0058] For example, the two first images cropped from the starting time point respectively include the left and right halves of the target area, so they need to be stitched together to generate a second image that includes the complete target area.

[0059] Step S4: Correct and remove thermal noise from the second image to obtain the first complex image at the start time point and the second complex image at the end time point.

[0060] Before correction, this embodiment first acquires precise orbital information and corrects the radar satellite's orbit based on this information to ensure high precision and accuracy of the acquired interferometric image. Thermal noise removal reduces the thermal noise generated by the satellite itself, improving the accuracy of the radar backscatter signal. Additionally, multi-view and filtering methods can be used to remove errors and further enhance the accuracy and reliability of the image. Finally, a first and second complex image are obtained.

[0061] Step S5: Register and overlay the first complex image and the second complex image to generate an interferogram. Perform phase unwrapping on the interferogram to obtain a terrain deformation map.

[0062] Specifically, the phase difference between the first and second complex images is obtained through registration and overlay to generate an interferometric image, which reflects the changes in surface topography at the start and end time points. However, in some cases, abrupt phase changes may occur, leading to discontinuities that can interfere with accurate deformation measurements. Therefore, phase unwrapping is necessary to obtain accurate deformation information; this implementation uses the minimum flow method for phase unwrapping to obtain a topographic deformation map.

[0063] Step S6: Identify the extent planar map, divide the extent planar map into multiple sub-regions, record the division method, obtain the historical deformation data of each sub-region, and assign the corresponding attention level to the sub-region based on the historical deformation data.

[0064] Specifically, based on the building types included in the scope plan, the scope plan is divided into multiple sub-areas, such as office areas, residential areas, reactor buildings, and fuel buildings. Then, based on the historical deformation data of each sub-area, each sub-area is assigned a corresponding level of attention. In this embodiment, there are attention level 1, attention level 2, and attention level 3, with attention level 1 being the most important and attention level 2 being the next most important.

[0065] Step S7: Divide the terrain deformation map into multiple sub-regions based on the division method, and analyze and interpret the sub-regions in sequence according to the level of attention to obtain the deformation information of each sub-region. Then, visualize the terrain deformation map and deformation information.

[0066] Similarly, by dividing the terrain deformation map into multiple sub-regions using a planar mapping approach, priority is given to processing sub-regions of interest level 1 during analysis and interpretation. Analyzing and interpreting the entire terrain deformation map would be time-consuming with limited computing power, while increasing computing power would increase equipment costs. Therefore, this invention, while analyzing each sub-region individually, prioritizes the analysis of more important sub-regions, thus obtaining deformation information for key areas as quickly as possible. Furthermore, this method not only ensures efficient analysis but also has lower computational requirements. Finally, the deformation is visualized using different colors to represent different deformation levels, allowing even those without specialized knowledge to understand the deformation situation.

[0067] This invention allows for remote sensing monitoring of the Earth's surface using radar sensors mounted on satellites or aircraft, eliminating the need for on-site contact with the monitored objects. This avoids interference and risks to the area surrounding nuclear power plants and reduces human error. InSAR technology can cover a large area of ​​the Earth's surface; a single monitoring session can encompass the entire area surrounding a nuclear power plant, providing comprehensive monitoring data that helps in the overall assessment of geological and surface changes in the region. InSAR technology is unaffected by weather conditions and lighting, enabling monitoring of target areas at any time and under any weather conditions, ensuring continuous monitoring and data acquisition of the area surrounding nuclear power plants.

[0068] This invention analyzes sub-regions one by one, prioritizing the analysis of sub-regions with higher importance, thereby obtaining deformation information of important areas as quickly as possible. In addition, this method not only ensures analysis time but also has low requirements for computing power.

[0069] Of particular note is that the above-mentioned technical solutions can quickly and accurately monitor terrain deformation.

[0070] This embodiment divides the area plan into multiple sub-regions, including the following steps:

[0071] Filter the building areas in the plan view, correct the outlines of the building areas, and obtain the corrected area;

[0072] Multiple basic shapes are preset, each corresponding to a building type. The basic shape corresponding to the correction area is matched to obtain the building type of the correction area. The correction area is expanded to make the expanded correction area seamlessly connected, and the expanded correction area is defined as a sub-region.

[0073] This embodiment extracts building regions from a range planar map through spectral semantic segmentation. Specifically, it acquires the spectral data of the range planar map, which includes energy reflection information in different bands. Spectral semantic segmentation infers the semantic category of each pixel based on the energy reflection information, such as building or road. Then, pixels that are classified and located close to each other are merged into pixel blocks, which are the building regions. Here, recognition errors may occur, such as a rectangular building being identified as an irregular shape. Therefore, it is necessary to correct the outline of the building region. The specific correction method will be introduced later.

[0074] In nuclear power plants, different building areas have specific shapes. For example, a building formed by combining circles and rectangles is generally a combination of the containment vessel, turbine power plant, fuel handling plant, and auxiliary equipment plant. A separate rectangular area is an office area or accommodation area. Therefore, the shape of the building area determines its corresponding building type. In addition, due to the presence of roads, the extracted building areas are not adjacent to each other, that is, there will be gaps between two sub-areas. Therefore, this embodiment also expands the building area to make the various sub-areas closely connected. Here, when expanding, only the plant type building area is expanded.

[0075] This embodiment includes the following steps before assigning a level of attention:

[0076] Set the monitoring duration, obtain deformation data of the sub-region within multiple consecutive monitoring durations, and extract the first data, second data, and third data from them. The first data is the deformation data with the smallest value, the second data is the deformation data with the largest value, and the third data is the deformation data of the last monitoring duration.

[0077] In this embodiment, the monitoring period is set to 15 days. After dividing the sub-regions, deformation data for each sub-region is obtained for days 1-15, days 16-30, etc., over the past 150 days. Finally, 10 data points are obtained. The first, second, and third data points are extracted from these 10 data points. The third data point is the deformation data between days 136 and 150, which is the deformation data for the last monitoring period.

[0078] Obtain the cumulative deformation at the end of the last monitoring period. Multiply the first, second, and third data by a predetermined factor to obtain the first, second, and third calibration values. Add the cumulative deformation to the first, second, and third calibration values ​​to obtain the first, second, and third predicted values.

[0079] In this embodiment, the cumulative deformation is the sum of deformation from day 1 to day 150, for example, 20 mm. Assuming the first data point is 0 mm, the second is 5 mm, and the third is 2 mm, multiplying each by a factor of 10 yields a first calibration value of 0 mm, a second calibration value of 50 mm, and a third calibration value of 20 mm. Adding these three values ​​to the cumulative deformation yields a first predicted value of 20 mm, a second predicted value of 70 mm, and a third predicted value of 40 mm.

[0080] Establish a coordinate system and mark the starting point, the first prediction point, the second prediction point, and the third prediction point in the coordinate system. The horizontal coordinates of the starting point, the first prediction point, the second prediction point, and the third prediction point are the third data, the first data, the second data, and the third data, respectively, and the vertical coordinates are the cumulative deformation, the first predicted value, the second predicted value, and the third predicted value, respectively.

[0081] Connect the starting point, the first prediction point, the second prediction point, and the third prediction point to obtain the evaluation profile.

[0082] In the coordinate system, the horizontal axis corresponds to the deformation observed during the monitoring period, and the vertical axis corresponds to the cumulative deformation. According to the above scheme, the coordinates of the starting point are (2, 20), the coordinates of the first predicted point are (0, 20), the coordinates of the second predicted point are (5, 70), and the coordinates of the third predicted point are (2, 40). Figure 2 In the example, the starting point is A, the first prediction point is B1, the second prediction point is B2, and the third prediction point is B3. Connecting these points sequentially will yield the evaluation profile.

[0083] This embodiment assigns a corresponding attention level to a sub-region through the following steps:

[0084] Mark the first and second hazard values ​​on the Y-axis of the coordinate system. Generate a first and second boundary line parallel to the X-axis based on the first and second hazard values. Obtain the first area S1 below the first boundary line and the second area S2 above the second boundary line of the evaluation profile. Calculate the concern value α based on the first formula, which is: Wherein, ε1 and ε2 are the first preset value and the second preset value, respectively.

[0085] Refer again Figure 2 If the first hazard value is less than the second hazard value, the first boundary line generated based on the first hazard value is C1, and the second boundary line generated based on the second hazard value is C2. The area below the first boundary line is D1, and the area above the first boundary line is D2. The first area S1 of region D1 and the second area S2 of region D2 are calculated and then substituted into the first formula to calculate the value of concern. Setting the first and second preset values ​​here is to avoid the first or second area being 0. It can be seen that the larger the second area, the larger the calculated value of concern; the larger the first area, the smaller the calculated value of concern. The evaluation profile is equivalent to a prediction of future deformation. Therefore, if a large deformation exceeding the second hazard value suddenly occurs during a certain monitoring period in the past, the value of concern will be large. If the largest deformation does not exceed the first hazard value, the value of concern will be small.

[0086] Set corresponding adjustment weights for each building type. Based on the building types included in the sub-region, adjust the attention value using the corresponding adjustment weights to obtain the importance value of the sub-region. Set multiple grade ranges and assign the corresponding attention level to the sub-region based on the grade range where the importance value is located.

[0087] Here, adjustment weights are set according to the importance of building types. For example, if the importance of office areas is lower than that of factory areas, then the weight of office areas is set to 0.2 and the weight of factory areas is set to 0.4. Then, the adjustment weights corresponding to the attention values ​​calculated for each sub-area are multiplied together to obtain the importance value. A classification range is also set here. For example, the classification range for attention level 1 is an importance value greater than or equal to 2. If the calculated importance value of a sub-area is 3, then that sub-area is set to attention level 1.

[0088] This embodiment includes the following steps for contour correction of the building area:

[0089] Locate the vertices in the building area, connect the vertices to obtain the first boundary line, define the direction perpendicular to the first boundary line as the first direction and the second direction, move the first boundary line along the first direction by a preset distance to obtain the first adjustment line, move the first boundary line along the second direction by a preset distance to obtain the second adjustment line, and calculate the evaluation values ​​of the first boundary line, the first adjustment line and the second adjustment line.

[0090] Set a first boundary and a second boundary, and repeatedly move the first boundary line until the first boundary and the second boundary are reached. Define the boundary line corresponding to the maximum evaluation value during the movement as the first correction line.

[0091] Vertices in a building area can be determined by curvature. For example, a threshold can be set, and points with curvature greater than the threshold can be defined as vertices. The method for calculating curvature is existing technology and will not be elaborated here. Alternatively, vertices can be determined by calculating whether the angle between two contour lines of the building area is greater than a preset angle; see reference. Figure 3 Assuming we have obtained vertices E1 and E2, we connect vertices E1 and E2 to obtain the first boundary line L1. We then obtain the first direction R1 and the second direction R2 perpendicular to the first boundary line L1. We move the first boundary line L1 along the first direction R1 to obtain the first adjustment line L2. We then move the first boundary line L1 along the second direction R2 to obtain the first adjustment line L3. Finally, we calculate the evaluation values ​​of the first boundary line, the first adjustment line, and the second adjustment line. The specific calculation method for the evaluation values ​​will be introduced later.

[0092] To limit the adjustment range of the first boundary line, this embodiment sets a first limit and a second limit. Based on the first and second adjustment lines, the line continues to move outward along the first direction R1 and the second direction R2, and the evaluation value of the adjusted line after the movement is recalculated. This process is repeated until the first and second limits are reached. Finally, the boundary line or adjustment line corresponding to the maximum evaluation value is defined as the first correction line. This method ensures that the first boundary line is in the correct position.

[0093] Locate the midpoint of the first correction line, rotate the first correction line with the midpoint as the center and the preset angle as the step size, calculate the evaluation value of the first correction line after each rotation, and after one rotation, stop the first correction line at the position with the maximum evaluation value.

[0094] For example, if the first boundary line L1 is finally used as the first correction line, then the center E3 of the first correction line is located. Then, the first correction line is rotated around E3 at a preset angle of 5°, and the evaluation value of the first correction line is calculated after each rotation. Finally, the first correction line is adjusted to the position with the maximum evaluation value. This method ensures that the first boundary line is at the correct angle.

[0095] The first correction lines with an angle greater than the first threshold and adjacent lines are merged to obtain the second correction line. The first correction line and the second correction line are then trimmed or extended to correct the outline of the building area into a closed shape.

[0096] Continue to refer to Figure 3 For example, after adjusting all the boundary lines, there is an angle between the two first correction lines at point F, but the angle is greater than the first threshold, indicating that the two first correction lines are very likely to belong to a straight line. Therefore, the two are merged into a second correction line. Finally, the first and second correction lines are connected to each other by extension and trimming to form a closed figure, such as... Figure 3 The building area K1 is modified to the modified area K2.

[0097] The calculation of the evaluation value in this embodiment includes the following steps:

[0098] Multiple observation points are located on the boundary line. A judgment region is generated with each observation point as the center. Pixels in the judgment region are matched pairwise based on the straight lines containing the first and second directions to obtain N pixel pairs. The reasonable value β for each observation point is calculated using the second formula, which is: Where, d n1 and d n2 p represents the distances between the first pixel and the observation point in the nth pixel pair, and the distances between the second pixel and the observation point, respectively. n1 and p n2 These are the pixel values ​​of the first and second pixels in the nth pixel pair, respectively.

[0099] The evaluation value γ of the first boundary line is calculated using the third formula, which is: Where M is the number of observation points in the boundary line, α m This is the reasonable value for the m-th observation point in the first boundary line.

[0100] Reference Figure 3 Here, vertices E1 and E2, and the midpoint E3 are set as observation points. The example given is the observation point generated from the midpoint E3. (Refer to...) Figure 4 If the pixel position corresponding to the midpoint E3 is H0, then a judgment area is generated with H0 as the center. The judgment area includes 6 pixels around H0, corresponding to H1 to H6. Then, the first direction R1 and the second direction R2 are obtained. Since the first direction R1 and the second direction R2 are on the same straight line, the two pixels that are collinear in this direction are matched. Based on this, the matching result of this embodiment has 3 pixel pairs, where pixel pair 1 is H1 and H6, pixel pair 2 is H2 and H4, and pixel pair 3 is H3 and H5. Finally, the pixel values ​​of pixels H1 to H6 are obtained and substituted into the second formula to calculate the reasonable value of the midpoint E3. In the second formula, the closer the two pixels in a pixel pair are to the observation point, the greater their influence on the observation point. Therefore, the closer the two pixels are to the observation point, the smaller dn1 and dn2 are, and the larger the first half of the second formula is. If the observation point is the actual outline point of the building, then the difference between the pixel values ​​on both sides should be large. Therefore, by calculating through the second formula, the probability of a point being an edge point can be determined. The larger the reasonable value, the greater the probability that the observation point is an edge point. Finally, the reasonable values ​​of all observation points on a boundary line are added together and the average value is calculated to obtain the evaluation value of a boundary line.

[0101] This embodiment uses the number and curvature of the vertices in the modified region as features and the Hausdorff distance as the criterion for matching the basic shape.

[0102] This embodiment corrects the second image by including the following steps:

[0103] Acquire orbital data, establish an orbital correction model based on the orbital data, and correct the orbital error of the second image based on the orbital correction model.

[0104] like Figure 5 As shown, the present invention also provides an InSAR deformation intelligent monitoring system focusing on nuclear power plant areas. This system is used to implement the aforementioned InSAR deformation intelligent monitoring method focusing on nuclear power plant areas. The system includes:

[0105] The database module stores map data. Map data is extracted from the database module, and the target area is extracted from the map data to obtain a planar map of the area.

[0106] The processing module has a set start time point and end time point. The remote sensing satellite takes SAR images at regular intervals. The processing module acquires multiple SAR images taken at the start time point and end time point, crops the SAR images to obtain the first image of the target area in the SAR images, stitches the first images at the same time point to obtain the second image containing the complete target area, corrects and removes thermal noise from the second image to obtain the first complex image at the start time point and the second complex image at the end time point. The processing module registers and superimposes the first complex image and the second complex image to generate an interferogram, and performs phase unwrapping on the interferogram to obtain the terrain deformation map.

[0107] The identification module is used to identify the extent planar map, divide the extent planar map into multiple sub-regions, record the division method, obtain historical deformation data for each sub-region, and assign a corresponding attention level to the sub-region based on the historical deformation data.

[0108] The interpretation module divides the terrain deformation map into multiple sub-regions based on the division method, and analyzes and interprets the sub-regions in sequence according to the level of attention to obtain the deformation information of each sub-region, and visualizes the terrain deformation map and deformation information.

[0109] The present invention also discloses a computer storage medium, characterized in that the computer storage medium stores program instructions, wherein the program instructions, when running, control the device where the computer storage medium is located to execute the above-described method.

[0110] It should be understood that although the steps in the flowcharts of the various embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the various embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0111] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0112] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this patent should be determined by the appended claims.

[0113] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for intelligent InSAR deformation monitoring focusing on nuclear power plant areas, characterized in that, include: Extract map data, and extract the target area from the map data to obtain a range plan map; Set a start time point and an end time point, acquire multiple SAR images taken at the start time point and the end time point, crop the SAR images, and obtain a first image of the SAR images corresponding to the target area; By stitching together the first images taken at the same time point, a second image containing the complete target region is obtained; The second image is corrected and thermal noise is removed to obtain a first complex image at the start time point and a second complex image at the end time point; The first complex image and the second complex image are registered and superimposed to generate an interferogram. The interferogram is then unwrapped in phase to obtain a terrain deformation map. Identify the range planar map, divide the range planar map into multiple sub-regions, record the division method, obtain historical deformation data for each sub-region, and assign a corresponding attention level to the sub-region based on the historical deformation data; Based on the aforementioned division method, the terrain deformation map is divided into multiple sub-regions, and the sub-regions are analyzed and interpreted sequentially according to the aforementioned attention level to obtain deformation information for each sub-region. The terrain deformation map and the deformation information are then visualized. Dividing the range plan into multiple sub-regions includes the following steps: Filter the building areas in the range plan view, and perform contour correction on the building areas to obtain the corrected area; Multiple basic shapes are preset, each of which corresponds to a building type. The basic shape corresponding to the correction area is matched to obtain the building type of the correction area. The correction area is expanded so that the expanded correction area is seamlessly connected. The expanded correction area is defined as the sub-region. Before assigning the attention level, the following steps are also included: Set a monitoring duration, acquire deformation data of the sub-region within multiple consecutive monitoring durations, and extract first data, second data and third data from them, wherein the first data is the deformation data with the smallest value, the second data is the deformation data with the largest value, and the third data is the deformation data of the last monitoring duration; Obtain the cumulative deformation at the end of the last monitoring period, and expand the first data, the second data, and the third data by a predetermined multiple to obtain the first calibration value, the second calibration value, and the third calibration value. Add the cumulative deformation to the first calibration value, the second calibration value, and the third calibration value to obtain the first predicted value, the second predicted value, and the third predicted value. Establish a coordinate system and mark the starting point, the first prediction point, the second prediction point, and the third prediction point in the coordinate system. The horizontal coordinates of the starting point, the first prediction point, the second prediction point, and the third prediction point are respectively the third data, the first data, the second data, and the third data, and the vertical coordinates are respectively the cumulative deformation, the first predicted value, the second predicted value, and the third predicted value. By connecting the starting point, the first prediction point, the second prediction point, and the third prediction point, an evaluation profile is obtained.

2. The method according to claim 1, characterized in that, Assigning the attention level to the sub-region includes the following steps: Mark a first danger value and a second danger value on the Y-axis of the coordinate system. Generate a first boundary line and a second boundary line parallel to the X-axis based on the first danger value and the second danger value. Obtain the first area of ​​the evaluation profile below the first boundary line. The second area above the second boundary line The attention value is calculated based on the first formula. The first formula is: ,in, and These are the first preset value and the second preset value, respectively. A corresponding adjustment weight is set for each of the building types. Based on the building types included in the sub-region, the attention value is adjusted using the corresponding adjustment weight to obtain the importance value of the sub-region. Multiple grade ranges are set, and based on the grade range in which the importance value is located, the corresponding attention level is assigned to the sub-region.

3. The method according to claim 1, characterized in that, The contour correction of the building area includes the following steps: Locate the vertices in the building area, connect the vertices to obtain a first boundary line, define the direction perpendicular to the first boundary line as a first direction and a second direction, move the first boundary line along the first direction by a preset distance to obtain a first adjustment line, move the first boundary line along the second direction by the preset distance to obtain a second adjustment line, and calculate the evaluation values ​​of the first boundary line, the first adjustment line, and the second adjustment line. Set a first limit and a second limit, and repeatedly move the first boundary line until the first limit and the second limit are reached. Define the boundary line corresponding to the maximum evaluation value during the movement as the first correction line. Locate the midpoint of the first correction line, rotate the first correction line with the midpoint as the center and a preset angle as the step size, calculate the evaluation value of the first correction line after each rotation, and after one rotation, stop the first correction line at the position with the maximum evaluation value. The first correction lines with an included angle greater than the first threshold and adjacent to each other are merged to obtain the second correction line. The first correction line and the second correction line are then trimmed or extended to correct the outline of the building area into a closed shape.

4. The method according to claim 3, characterized in that, Calculating the evaluation value includes the following steps: Multiple observation points are located on the boundary line. A judgment region is generated with each observation point as the center. Pixels in the judgment region are matched pairwise based on the straight lines containing the first and second directions to obtain N pixel pairs. A reasonable value for each observation point is calculated using a second formula. The second formula is: ,in, and These are the distances between the first pixel and the observation point in the nth pixel pair, and the distance between the second pixel and the observation point, respectively. and These are the pixel values ​​of the first and second pixels in the nth pixel pair, respectively; The evaluation value of the first boundary line is calculated using the third formula. The third formula is: Where M is the number of observation points in the boundary line. The reasonable value is the m-th observation point in the first boundary line.

5. The method according to claim 1, characterized in that, The basic shape is matched based on the number and curvature of the vertices in the modified region as features, using the Hausdorff distance as a criterion.

6. The method according to claim 1, characterized in that, Correcting the second image includes the following steps: Acquire orbital data, establish an orbital correction model based on the orbital data, and correct the orbital error of the second image based on the orbital correction model.

7. An InSAR intelligent deformation monitoring system focusing on nuclear power plant areas, used to implement the method as described in any one of claims 1-6, characterized in that, include: The database module is used to extract map data, extract target areas from the map data, and obtain a range planar map; The processing module includes a set start time point and an end time point. The processing module acquires multiple SAR images taken at the start time point and the end time point, crops the SAR images to obtain a first image corresponding to the target area, stitches together the first images at the same time point to obtain a second image containing the complete target area, corrects and removes thermal noise from the second image to obtain a first complex image at the start time point and a second complex image at the end time point, registers and superimposes the first complex image and the second complex image to generate an interferogram, and performs phase unwrapping on the interferogram to obtain a terrain deformation map. The identification module is used to identify the range planar map, divide the range planar map into multiple sub-regions, record the division method, obtain historical deformation data of each sub-region, and assign a corresponding attention level to the sub-region based on the historical deformation data; The interpretation module divides the terrain deformation map into multiple sub-regions based on the division method, and analyzes and interprets the sub-regions in sequence according to the attention level to obtain the deformation information of each sub-region, and visualizes the terrain deformation map and the deformation information.

8. A computer storage medium, characterized in that, The computer storage medium stores program instructions, wherein when the program instructions are executed, they control the device where the computer storage medium is located to perform the method described in any one of claims 1-6.

Images