Buffer optimization method and system for InSAR deformation monitoring of long-distance linear engineering
By constructing a gradient buffer set and fitting a mapping function, the InSAR buffer range is optimized, solving the problem of improper buffer setting in long-distance linear engineering. This achieves efficient and accurate deformation monitoring, and is applicable to inter-basin water transfer projects such as the South-to-North Water Transfer Project and high-speed railways.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA SOUTH-TO-NORTH WATER DIVERSION GRP MIDDLE LINE CO LTD
- Filing Date
- 2026-03-31
- Publication Date
- 2026-06-19
AI Technical Summary
The existing InSAR technology lacks a systematic quantitative basis for setting buffers in long-distance linear engineering, which makes it difficult to balance deformation inversion accuracy and computational efficiency, resulting in poor adaptability. Furthermore, an excessively large buffer introduces noise, while an excessively small buffer affects unwrapping robustness.
By constructing a set of gradient buffers with the centerline as the axis, and combining the Euclidean distance transformation algorithm, the buffer range is optimized, the mapping function between phase unwrapping time and buffer size is fitted, the objective function is constructed, and the optimal buffer size is adaptively found.
It improves the accuracy of deformation inversion, reduces the consumption of computing resources, enables efficient and accurate monitoring of long-distance linear engineering projects, is applicable to complex geological environments, and enhances robustness and universality.
Smart Images
Figure CN122239059A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of remote sensing monitoring technology, and in particular to an InSAR deformation monitoring buffer optimization method and system for long-distance linear engineering. Background Technology
[0002] Interferometric Synthetic Aperture Radar (InSAR) technology, as a new generation of space-based Earth observation method, has gradually demonstrated its irreplaceable role in the field of surface deformation monitoring due to its millimeter-level monitoring accuracy, all-weather and 24 / 7 observation capabilities, and wide spatial coverage. Especially for large-scale linear water diversion projects, long-distance oil pipelines, high-speed railways, and other long-distance linear infrastructure projects, InSAR technology can provide high spatial density deformation field information, effectively compensating for the shortcomings of traditional manual leveling surveys, such as low point density, long operation cycles, and poor accessibility in harsh environments. This provides a new technical approach for achieving refined and routine safety monitoring of the entire project line.
[0003] In the InSAR deformation extraction process, setting a reasonable spatial buffer for the study area is a crucial and necessary step. The core function of this buffer is to ensure the edge stability of the interferogram during filtering, phase unwrapping, and spatial reference point selection, preventing the propagation of unwrapping errors or deformation field distortion due to missing information in the engineering boundary areas. However, the size of the buffer directly affects the quality and efficiency of data processing: if the buffer is too large, a large number of non-engineering features (such as vegetation cover areas and areas with drastic topographic changes) are included in the calculation range, introducing redundant noise, reducing the deformation inversion accuracy within the engineering area, and significantly increasing computational resource consumption and processing time; if the buffer is too small, the density of high coherence points in the edge areas is insufficient, making it difficult to maintain the spatial continuity of phase unwrapping, increasing the probability of unwrapping errors, and thus weakening the spatial filtering effect, affecting the overall reliability of the deformation field. For long-distance linear engineering projects, with their large spatial span and complex terrain features along the route, the selection of the buffer range becomes a core issue in balancing deformation inversion accuracy and computational efficiency.
[0004] Currently, in engineering InSAR applications, the mainstream processing methods, represented by Small Baseline Set Time Series InSAR (SBAS-InSAR), largely rely on operator experience to determine the buffer range, typically using a fixed empirical value of 1–20 km. This experience-based approach has significant drawbacks: firstly, it lacks systematic quantitative data on the differences in long-distance linear engineering geometry and surface scattering characteristics along the route, blurring the boundary between the engineering and non-engineering areas and resulting in poor adaptability to different engineering scenarios; secondly, it is difficult to achieve an optimal trade-off between accuracy and efficiency—an excessively large buffer introduces numerous non-target area noise points, diluting the effective signal in the engineering area and causing computational redundancy; an excessively small buffer cannot guarantee the spatial connectivity and robustness of the unwrapped network, limiting the robustness of deformation field inversion.
[0005] Therefore, how to scientifically determine the optimal buffer range for InSAR deformation monitoring applicable to long-distance linear engineering while ensuring the accuracy of deformation inversion has become a key technical problem that urgently needs to be solved in the engineering application of radar remote sensing technology. Summary of the Invention
[0006] This invention provides an InSAR deformation monitoring buffer optimization method and system for long-distance linear engineering, which solves the defect of ambiguous definition of InSAR deformation monitoring buffer in the prior art and realizes adaptive optimization of InSAR deformation monitoring buffer.
[0007] In a first aspect, the present invention provides an InSAR deformation monitoring buffer optimization method for long-distance linear engineering, comprising: Acquire temporal SAR imagery data covering long-distance linear engineering projects; Extract the centerline vector data of the long-distance linear project and construct a gradient buffer set with the centerline vector data as the axis; Under the constraints of buffers at each scale in the gradient buffer set, the temporal SAR data is subjected to temporal interferometry, coherence point filtering and phase unwrapping, and a first mapping function between phase unwrapping time and buffer scale is fitted. The phase unwrapping result is subjected to deformation inversion to obtain deformation data, and a second mapping function between the inversion error and the buffer scale is fitted based on the inversion error between the deformation data and the reference point data. Based on the first mapping function and the second mapping function, construct an objective function for buffer scale optimization; Solve the objective function to obtain the optimal solution for the buffer scale.
[0008] According to the present invention, an InSAR deformation monitoring buffer optimization method for long-range linear engineering is provided, wherein the step of extracting centerline vector data of the long-range linear engineering from the temporal SAR image data and constructing a gradient buffer set with the centerline vector data as the axis includes: Extract the centerline vector data of the long-distance linear project based on the time-series SAR image data; A spatial field centered on the centerline vector data is established using the Euclidean distance transformation algorithm. Select several representative segments in the long-distance linear engineering and construct a gradient buffer set that expands from the inside to the outside for the several representative segments in the spatial field.
[0009] According to the present invention, an InSAR deformation monitoring buffer optimization method for long-distance linear engineering is provided. Under the constraints of buffers at each scale in the gradient buffer set, the method performs temporal interferometry, coherence point selection, and phase unwrapping on the temporal SAR data, and fits a first mapping function between the phase unwrapping time and the buffer scale, including: Temporal interferometry is performed on the range of buffer constraints at each scale of the gradient buffer set in the temporal SAR data to generate interferogram sequences corresponding to each scale buffer. By filtering coherent points in the interferogram sequences corresponding to each scale buffer, the sampling point set corresponding to each scale buffer is obtained. Phase unwrapping is performed on the sampling point sets corresponding to each scale buffer to obtain the phase value of each sampling point, and the phase unwrapping time corresponding to each scale buffer is recorded. Based on the phase unwrapping time corresponding to each scale buffer, a mapping function between phase unwrapping time and buffer scale is fitted.
[0010] According to the InSAR deformation monitoring buffer optimization method for long-distance linear engineering provided by the present invention, after filtering coherent points from the interferogram sequence to obtain the sampling point sets corresponding to each scale buffer, the method further includes: The sampling point density corresponding to each scale buffer is statistically analyzed in real time, and a response curve of the sampling point density as the buffer scale changes is established; wherein, the sampling point density is defined as: the number of sampling points that meet the preset threshold condition within a unit geographical area within the buffer mask range; Marginal benefit constraints are established based on the response curve.
[0011] According to the present invention, an InSAR deformation monitoring buffer optimization method for long-distance linear engineering is provided, wherein the marginal benefit constraint is:
[0012] Among them, the Represents the change in density at sampling points. This represents the change in the buffer zone scale. This represents the preset rate of change threshold.
[0013] According to the present invention, an InSAR deformation monitoring buffer optimization method for long-distance linear engineering is provided, wherein deformation inversion is performed on the phase unwrapping result to obtain deformation data, and a second mapping function between the inversion error and the buffer scale is fitted based on the inversion error between the deformation data and the reference point data, comprising: Deformation inversion is performed on the phase value of each sampling point to obtain the deformation data of each sampling point; The mean square error between the deformation data of each sampling point in each scale buffer and the reference point data is calculated respectively, and a second mapping function between the mean square error and the buffer scale is fitted; wherein, the reference point data is the true value data synchronously observed along the long-distance linear engineering line.
[0014] According to the present invention, an InSAR deformation monitoring buffer optimization method for long-distance linear engineering is provided, wherein constructing an objective function for buffer scale optimization based on a first mapping function and a second mapping function includes: The first mapping function and the second mapping function are standardized respectively to obtain the standardized first mapping function and the standardized second mapping function; The standardized first mapping function and the standardized second mapping function are weighted and summed to obtain the objective function for buffer scaling optimization; wherein the objective function is constrained by the marginal benefit constraint condition.
[0015] Secondly, the present invention also provides an InSAR deformation monitoring buffer optimization system for long-distance linear engineering, comprising: The data acquisition module is used to acquire time-series SAR image data covering long-distance linear engineering projects; A buffer construction module is used to extract the centerline vector data of the long-distance linear project and construct a gradient buffer set with the centerline vector data as the axis. The efficiency measurement module is used to perform temporal interferometry, coherence point filtering, and phase unwrapping on the temporal SAR data under the constraints of buffers at each scale in the gradient buffer set, and to fit a first mapping function between phase unwrapping time and buffer scale. The accuracy measurement module performs deformation inversion on the phase unwrapping result to obtain deformation data, and fits a second mapping function between the inversion error and the buffer scale based on the inversion error between the deformation data and the reference point data. The scaling optimization module is used to construct an objective function for buffer scaling optimization based on the first mapping function and the second mapping function; and to solve the objective function to obtain the optimal solution for the buffer scaling.
[0016] Thirdly, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the InSAR deformation monitoring buffer optimization method for long-distance linear engineering as described above.
[0017] Fourthly, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the InSAR deformation monitoring buffer optimization method for long-distance linear engineering as described above.
[0018] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the InSAR deformation monitoring buffer optimization method for long-distance linear engineering as described above.
[0019] The beneficial effects of the technical solutions provided by some embodiments of the present invention include at least the following: 1) This invention provides an InSAR deformation monitoring buffer optimization method and system for long-distance linear engineering projects. Taking the centerline of the long-distance linear engineering project as a reference, a set of discretized gradient buffers expanding from the inside out is constructed. Under unified InSAR processing parameters, the deformation feature capture capability and unwrapping robustness at different scales are compared. A first mapping function between phase unwrapping time and buffer scale, and a second mapping function between inversion error and buffer scale are fitted to construct the objective function for buffer scale optimization. The optimal solution for the buffer scale is obtained, thereby adaptively identifying the optimal buffer range. This algorithm solves the problem of "scale effect" in linear engineering monitoring from the perspective of algorithm mechanism. It can not only effectively avoid phase unwrapping path breakage and jump error caused by the buffer being too small, but also accurately remove spatial correlation noise introduced by background features far away from the main body of the project, significantly improving the deformation inversion accuracy.
[0020] 2) This invention uses the centerline of the water diversion project as a reference to construct a set of discretized gradient buffers that expand from the inside out. This is equivalent to establishing a multi-scale observation space from the local engineering structure to the distant background environment, realizing multi-scale gradient space modeling, and providing a sample basis for finding the signal-to-noise ratio balance point.
[0021] 3) This invention constructs an objective function by real-time statistical analysis of the changing trends of inversion error (representing accuracy) and phase unwrapping time (representing computational overhead) with the buffer scale. Combined with a marginal effect constraint based on coherence point density (representing data quality), it automatically determines the optimal signal-to-noise ratio balance point for InSAR deformation monitoring of long-distance linear engineering projects, achieving adaptive optimization of the buffer range for InSAR deformation monitoring. Compared to the traditional approach of setting buffers based on experience, this invention achieves adaptive optimization decision-making based on a dual-parameter constraint of "accuracy-efficiency," improving adaptability to specific linear engineering projects.
[0022] 4) By accurately locating the optimal buffer scale, this invention enables the automatic elimination of invalid pixel calculations in redundant background areas during subsequent InSAR deformation monitoring, significantly reducing the matrix dimension and computational overhead in the singular value decomposition and spatial filtering processes. While ensuring millimeter-level inversion accuracy, it achieves on-demand allocation of computing resources, significantly improves computational efficiency and resource optimization rate, and greatly enhances the technical feasibility of high-frequency and rapid security early warning for ultra-long-distance linear engineering projects.
[0023] 5) Based on empirical evidence from multiple sections and geological backgrounds in long-distance linear engineering, this invention introduces external high-precision benchmark data for closed-loop verification, and determines the optimal buffer range for millimeter-level deformation monitoring in such large-scale long-distance linear engineering projects, providing core technical indicators to support the standardized monitoring procedures for subsequent large-scale long-distance linear engineering projects.
[0024] 6) The optimization mechanism proposed in this invention has strong adaptive characteristics. This mechanism is not only applicable to inter-basin water transfer projects such as the South-to-North Water Transfer Project, but can also be extended to the deformation monitoring field of linear projects such as long-distance oil pipelines and high-speed railways. It enhances the robustness and universality of deformation monitoring and inversion in complex geological environments and has significant industrial application breadth. Attached Figure Description
[0025] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0026] Figure 1This is a flowchart illustrating an InSAR deformation monitoring buffer optimization method for long-distance linear engineering provided by the present invention. Figure 2 This is an example diagram of two representative canal sections and their gradient buffer sets of the South-to-North Water Diversion Project provided by the present invention; wherein, (a) is a representative canal section (MRP) and its gradient buffer range, with red dots representing benchmarks, and (b) is another representative canal section and its gradient buffer range, with red dots also representing benchmarks; Figure 3 This is a schematic diagram of the structure of an InSAR deformation monitoring buffer optimization system for long-distance linear engineering provided by the present invention; Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation
[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.
[0028] Please see Figure 1 , Figure 1 A flowchart illustrating an InSAR deformation monitoring buffer optimization method for long-distance linear engineering, provided as an embodiment of the present invention, includes: S101. Acquire time-series SAR image data covering long-distance linear engineering projects; S102. Extract the centerline vector data of the long-distance linear project and construct a gradient buffer set with the centerline vector data as the axis. S103. Under the constraints of each scale buffer in the gradient buffer set, perform time-series interferometric solution, coherence point screening and phase unwrapping on the time-series SAR data, and fit the first mapping function between the phase unwrapping time and the buffer scale. S104. Perform deformation inversion on the phase unwrapping result to obtain deformation data, and fit a second mapping function between the inversion error and the buffer scale based on the inversion error between the deformation data and the reference point data. S105. Construct an objective function for buffer scale optimization based on the first mapping function and the second mapping function; S106. Solve the objective function to obtain the optimal solution for the buffer scale.
[0029] This invention provides an InSAR deformation monitoring buffer optimization method and system for long-distance linear engineering projects. Using the centerline of the linear engineering project as a reference, a set of discretized gradient buffers expanding outwards is constructed, establishing a multi-scale observation space from the local engineering structure to the distant background environment. Then, under unified InSAR processing parameters, the deformation feature capture capability and unwrapping robustness at different scales are compared. A first mapping function between phase unwrapping time and buffer scale, and a second mapping function between inversion error and buffer scale are fitted to construct the objective function for buffer scale optimization. The optimal solution for the buffer scale is found by finding the signal-to-noise ratio balance point, thereby adaptively identifying the optimal buffer range. This algorithm solves the "scale effect" problem in linear engineering monitoring from an algorithmic perspective. It not only effectively avoids phase unwrapping path breaks and jump errors caused by excessively small buffers, but also accurately removes spatially related noise introduced by background features far from the main engineering body, significantly improving deformation inversion accuracy.
[0030] This invention can be applied to deformation monitoring in inter-basin water transfer projects such as the South-to-North Water Diversion Project, long-distance oil pipelines, and long-distance linear projects such as high-speed railways. It has good robustness and universality for deformation monitoring and inversion in complex geological environments.
[0031] This embodiment mainly uses the South-to-North Water Diversion Project as an example to illustrate the InSAR deformation monitoring buffer optimization method for long-distance linear projects proposed in this invention.
[0032] In step S101 above, time-series SAR image data covering long-distance linear engineering projects are acquired from remote sensing satellites.
[0033] For example, temporal SAR imagery data from the Sentinel-1A satellite in interferometric wide swath (IW) mode was collected. The data coverage includes four ascending orbits numbered 11, 40, 113, and 142, ensuring complete space monitoring of the 1,432-kilometer-long South-to-North Water Diversion Project. The spatial resolution of this SAR imagery data is 5 meters in the range direction and 20 meters in the azimuth direction, with a temporal resolution of 12 days.
[0034] For example, in addition to Sentinel-1 satellite data, other satellite platforms that provide SAR data can also be used, such as TerraSAR satellite, ALOS satellite, Gaofen-3 satellite, etc.
[0035] In step S102 above, based on the centerline of the extracted long-distance linear engineering, buffers of different gradient scales are constructed to form a gradient buffer set.
[0036] In some possible embodiments, the step of extracting the centerline vector data of the long-distance linear engineering and constructing a gradient buffer set with the centerline vector data as the axis includes: Extract the centerline vector data of the long-distance linear project based on the time-series SAR image data; A spatial field centered on the centerline vector data is established using the Euclidean distance transformation algorithm. Select several representative segments in the long-distance linear engineering and construct a gradient buffer set that expands from the inside to the outside for the several representative segments in the spatial field.
[0037] Specifically, taking the South-to-North Water Diversion Project as an example, based on the time-series SAR image data obtained in step S101, the centerline vector data of the South-to-North Water Diversion Project is extracted, and a spatial field with the centerline of the project as the axis is established using the Euclidean distance transformation algorithm. Several representative canal sections are selected from the South-to-North Water Diversion Project, and a gradient buffer set expanding from the inside to the outside is constructed.
[0038] like Figure 2 The image shown is an example diagram of two representative sections of the South-to-North Water Diversion Project's central route and their gradient buffer sets. Figure 2 In diagram (a), a representative canal segment (MRP) and its gradient buffer range are shown, with red dots representing benchmarks. In diagram (b), another representative canal segment and its gradient buffer range are shown, with red dots also representing benchmarks. In both (a) and (b), the gradient buffer set is... D ={ d 1 = 1 ,d 2 = 3 ,d 3 = 5 ,d 4 = 7 ,d 5 = 10 ,d 6 = 15}, of which, d i This represents the distance from the centerline of a representative canal section to the boundary of the buffer zone (unit: km ), i =1,2,..,6, representing 6 different gradient scale numbers.
[0039] For example, gradient buffer sets can also be constructed using incremental step sizes. For instance, a fixed step size can be set, and gradient buffer sets can be constructed through gradient increments or other methods, as long as the differences in deformation extraction results caused by different buffer scales can be distinguished.
[0040] This invention uses the centerline of the water diversion project as a reference to construct a set of discretized gradient buffers that expand from the inside out. This is equivalent to establishing a multi-scale observation space from the local engineering structure to the distant background environment, realizing multi-scale gradient space modeling, and providing a sample basis for finding the signal-to-noise ratio balance point.
[0041] Next, based on the existing temporal InSAR method, this invention proposes a buffer adaptive optimization mechanism. Under the same InSAR processing parameters (such as the same coherence threshold and amplitude deviation index), the deformation feature capture capability and unwrapping robustness under different scale buffer ranges are compared, thereby adaptively finding the optimal buffer range.
[0042] In theory, any time-series InSAR method for surface deformation inversion can serve as the basic model for the buffer adaptive optimization mechanism of this invention, such as Small Baseline Subset Interferometric Syntheitc Aperture Radar (SBAS-InSAR), Permanent Scatterer Interferometric Syntheitc Aperture Radar (PS-InSAR), and Distributed Scatterer Interferometric Syntheitc Aperture Radar (DS-InSAR).
[0043] In this embodiment, the buffer adaptive optimization mechanism of the present invention is explained using SBAS-InSAR technology as the basic model.
[0044] Small baseline ensemble aperture radar interferometry constructs a small baseline interferometric set and uses singular value decomposition (SVD) to solve for the deformation time series. The buffer zone is typically set empirically to be 1–20 km. The basic procedure of this method is as follows: (1) Image acquisition and preprocessing: N SAR images of the study area are acquired and registered according to the preset time baseline and spatial baseline threshold.
[0045] (2) Interference pair generation and differential processing: M interference pairs are generated according to the combination conditions, and differential interferometry is performed to remove the flat phase and the terrain phase, and generate a differential interferogram.
[0046] (3) Coherence point screening: Using the coherence coefficient threshold or amplitude deviation index, identify sampling points with high coherence, namely permanent scatterers (PS).
[0047] (4) Phase unwrapping: Three-dimensional or two-dimensional phase unwrapping is performed on the selected PS points.
[0048] (5) Deformation rate and residual topography inversion: Establish the observation equation, use the singular value decomposition (SVD) method to solve the matrix equation of the multi-source small baseline set, and invert the deformation rate and residual topography error of each PS point in the observation time series.
[0049] (6) Atmospheric correction and orbital error smoothing: The residual phase is processed by a spatiotemporal filter (low-pass / high-pass filter) to separate the atmospheric delay phase and orbital noise, and finally obtain a high-precision deformation sequence result.
[0050] The embodiments of this invention use SBAS-InSAR technology as the basic model, embedding a buffer adaptive optimization mechanism into this basic model to form an optimized SBAS-InSAR algorithm, achieving buffer adaptive optimization, such as... Figure 1 As shown. Steps S103 to S106 correspond to the InSAR deformation monitoring buffer optimization method based on the optimized SBAS-InSAR algorithm of the present invention.
[0051] In step S103 above, for each gradient scale in the gradient buffer set... d i The timing interferometry solution, coherence point selection, and phase unwrapping are executed sequentially.
[0052] In some possible embodiments, under the constraints of buffers at each scale in the gradient buffer set, the temporal SAR data undergoes temporal interferometry, coherence point filtering, and phase unwrapping, and a first mapping function between phase unwrapping time and buffer scale is fitted, including: S103-1. Perform time-series interferometric calculation on the range of buffer constraints at each scale of the gradient buffer set in the time-series SAR data to generate interferogram sequences corresponding to each scale buffer. S103-2. Perform coherence point filtering on the interferogram sequences corresponding to each scale buffer to obtain the sampling point set corresponding to each scale buffer. S103-3. Perform phase unwrapping on the sampling point set corresponding to each scale buffer to obtain the phase value of each sampling point, and record the phase unwrapping time corresponding to each scale buffer. S103-4. Based on the phase unwrapping time corresponding to each scale buffer, fit the first mapping function between the phase unwrapping time and the buffer scale.
[0053] Specifically, in step S103-1 above, firstly, N+1 SAR images covering the same area are registered. One scene is selected as the master image, and all other secondary images are registered with the master image, for example, a quadratic polynomial can be used for registration.
[0054] Then, interferometric pair matching is performed. The interferometric pair matching follows the principle of free pairwise combination; from N+1 SAR images, this can generate... For interferometric pairs, following the Small Baseline Set (SBAS) method, a spatiotemporal baseline threshold is set for interferometric pair matching. Only interferometric pairs that meet the spatiotemporal baseline threshold requirements are selected. Assuming there are M interferometric pairs, then... .
[0055] Finally, differential interferometry is performed, which involves multiplying the two SAR images in each interferometric pair by conjugate to generate a differential interferogram, resulting in the interferogram sequence corresponding to each scale buffer.
[0056] In step S103-2 above, to accurately obtain the minute deformation information of the monitored object, sampling points that are not easily affected by time decoherence are selected, namely permanent scatterers (PS). Based on the strong reflection characteristics of PS points, pixels with high amplitude values can be selected as PS candidate points. Then, according to a preset amplitude dispersion index threshold, PS candidate points with an amplitude dispersion index (ADI) higher than the preset threshold are further screened. The amplitude dispersion index is expressed as:
[0057] in, The standard deviation of the time series amplitude. This represents the average amplitude over time.
[0058] Subsequently, for the gradient buffer set D constructed in step S102, the PS points in each scale buffer of the gradient buffer set D are extracted to obtain the PS point set corresponding to each scale buffer, forming a PS point set under different spatial scale constraints.
[0059] In some possible embodiments, after filtering coherent points from the interferogram sequence to obtain the sampling point sets corresponding to each scale buffer, the method further includes: The sampling point density corresponding to each scale buffer is statistically analyzed in real time, and the response curve of the sampling point density as the buffer scale changes is established. Marginal benefit constraints are established based on the response curve.
[0060] Specifically, the sampling point density within each scale buffer is calculated in real time. P densityBased on the sampling point density corresponding to each scale buffer, the sampling point density can be fitted. P density With buffer scale d The response curves between these curves are used to quantitatively analyze the minimum boundary of deformation detection.
[0061] Sampling point density can P density Defined as: at a specific scale d i Within the buffer mask area, the number of PS points that meet the preset amplitude deviation index threshold condition per unit geographical area is calculated using the following basic formula:
[0062] in, N PS This represents the total number of sample points retained after filtering within the buffer zone; A buffer This represents the geographic area of the buffer zone at this scale, in units of... km 2 .
[0063] Furthermore, based on the sampling point density P density With buffer scale d The response curves between can be constructed P density Follow d Marginal benefits of value change , Represented as:
[0064] Among them, the Represents the change in density at sampling points. This represents the change in the buffer scale.
[0065] when When the buffer range falls below a certain threshold, it indicates that increasing the buffer range contributes very little to the PS point density; the corresponding buffer range is the minimum constraint condition. Therefore, the marginal benefit constraint condition can be:
[0066] in, This represents the preset rate of change threshold.
[0067] Since large-scale water diversion projects are mostly man-made structures with strong scattering characteristics, their PS point density exhibits a significant peak around the project structures, while rapidly attenuating further away from the project area. Therefore, this spatial distribution characteristic provides a physical basis for adaptively determining the optimal buffer scale, ensuring that subsequent interferometric processing both fully preserves the phase information of the project structure and effectively suppresses background interference from distant locations.
[0068] In step S103-3 above, the interference phase in the interferogram is the principal value of the wrapped phase distributed between [-π, π), which differs from the true phase value by 2. k π k Integer ambiguities (where the integer value is an integer) need to be determined through phase unwrapping. k This allows for the determination of accurate deformation information.
[0069] In embodiments of the present invention, a spatiotemporal three-dimensional unwrapping method can be used for unwrapping, that is, two-dimensional space plus a time dimension. Specifically, taking advantage of the temporal continuity of PS points, for each scale buffer, the temporal phase difference between PS points of temporally adjacent interference pairs is calculated, and then the iterative least squares method is applied to the temporal phase difference to perform two-dimensional spatial unwrapping to obtain the phase value of each sampling point.
[0070] In steps S103-4 above, to facilitate subsequent adaptive optimization of the buffer, phase unwrapping is performed on the sampling points corresponding to each scale buffer, and the phase unwrapping time corresponding to each scale buffer is recorded, thereby fitting the first mapping function between the phase unwrapping time and the buffer scale.
[0071] Phase unwrapping is a core step in the temporal InSAR deformation extraction process, determining the final deformation extraction accuracy and also being the most computationally time-consuming step. This invention performs phase unwrapping on the sampling point sets corresponding to each scale buffer of the gradient buffer set D constructed in step S102, fits the first mapping function between the phase unwrapping time and the buffer scale, and analyzes the phase unwrapping time. T cost Follow d The purpose of studying the changing trend of the growth slope caused by the increase in value is to find the balance point between "accuracy improvement / time cost".
[0072] In step S104 above, the phase values of each PS point obtained from phase unwrapping in step S103 are subjected to deformation inversion, and the reference point data synchronously observed along the engineering line are introduced as the true value. For each gradient... d i Calculate the inversion error between the InSAR inversion result and the reference point, and fit a second mapping function between the inversion error and the buffer scale.
[0073] In some possible embodiments, performing deformation inversion on the phase unwrapping result to obtain deformation data, and fitting a second mapping function between the inversion error and the buffer scale based on the inversion error between the deformation data and the reference point data, includes: S104-1. Perform deformation inversion on the phase value of each sampling point to obtain the deformation data of each sampling point; S104-2. Calculate the mean square error between the deformation data of each sampling point in each scale buffer and the reference point data, and fit the second mapping function between the mean square error and the buffer scale; wherein, the reference point data is the true value data synchronously observed along the long-distance linear engineering line.
[0074] Specifically, in step S104-1 above, atmospheric correction and orbital error smoothing are performed on the phase value obtained from phase unwrapping in step S103. Atmospheric delay phase and orbital noise are separated by a spatiotemporal filter to obtain high-precision deformation data.
[0075] Since the phase value obtained after phase unwrapping in step S103 is still the total phase including deformation phase, atmospheric delay phase, orbital error phase, terrain residual phase, and noise phase, it can be expressed as:
[0076] in, This represents the phase value after phase unwrapping at each sampling point. For deformation phase, For atmospheric delayed phase, For orbital error phase, For terrain residual phase, This is the noise phase.
[0077] To extract the deformation phase more accurately, it is necessary to further remove the atmospheric delay phase, orbital error phase, terrain residual phase, and noise phase.
[0078] Among them, the terrain residual phase originates from the residual terrain phase caused by inaccuracies in the external DEM. It has a linear relationship with the vertical baseline. Therefore, the terrain residual phase... The correction is obtained by estimating the correction at each PS point through least-squares regression of the unwrapped phase and vertical baseline. Orbital errors exhibit a large-scale trend surface with gentle spatial variations, typically fitted with low-order polynomials (first or second order). Therefore, the orbital error phase... It can be subtracted by fitting a linear function. Atmospheric delayed phase is spatially correlated and temporally random (or varies at low frequencies); therefore, atmospheric delayed phase... It can be extracted using spatial low-pass filtering and temporal high-pass filtering.
[0079] After deducting the above error terms, the remaining phase is mainly caused by deformation. Assuming the deformation rate is linear between two adjacent time intervals, i.e., the deformation is piecewise linear over the complete observation period, the above equation can be expressed as:
[0080] in, Let M be the unwrapped interferometric phases, V be the K piecewise linear deformation rates to be solved, and B be a coefficient matrix of size M×K. When all interferograms are in the same subset, B is an invertible matrix, and the deformation rate is solved using the least squares method. If constrained by the spatiotemporal baseline, there are discontinuous subsets of interferograms, then B is rank deficient. In this case, singular value decomposition is performed on B to obtain the deformation rate at each sampling point. Integrating the deformation rate over each time interval yields the deformation time series for the complete observation period.
[0081] It is understandable that the deformation data for each sampling point includes the deformation rate and the deformation time series.
[0082] In step S104-2 above, based on the deformation data of each sampling point obtained in step S104-1, the benchmark data of synchronous observation along the engineering line is introduced as the true value data, for each gradient. d i The mean square error between the deformation data of each sampling point in the corresponding buffer and the reference point data is calculated respectively. Based on the mean square error corresponding to each buffer scale, a second mapping function between the mean square error and the buffer scale is fitted.
[0083] Taking the South-to-North Water Diversion Project as an example, this invention, based on empirical evidence from multiple sections and geological backgrounds within the South-to-North Water Diversion Project, introduces external high-precision benchmark data to participate in closed-loop verification, and determines the optimal buffer threshold for millimeter-level deformation monitoring in such large-scale, long-distance linear projects, providing core technical indicators to support the standardized monitoring procedures for subsequent large-scale water diversion projects.
[0084] In step S105 above, an objective function for buffer scale optimization is constructed based on the first mapping function between phase unwrapping time and buffer scale, and the second mapping function between mean square error and buffer scale.
[0085] In some possible embodiments, constructing the objective function for buffer scaling optimization based on the first mapping function and the second mapping function includes: S105-1. Standardize the first mapping function and the second mapping function respectively to obtain a standardized first mapping function and a standardized second mapping function; S105-2. The standardized first mapping function and the standardized second mapping function are weighted and summed to obtain the objective function for buffer scaling optimization; wherein the objective function is constrained by the marginal benefit constraint condition.
[0086] Specifically, in step S105-1 above, the Max-Min method is used to standardize the first mapping function and the second mapping function. Let the first mapping function between phase unwrapping time and buffer scale be expressed as... The second mapping function between mean squared error and buffer scale is expressed as: The expressions for standardizing the first mapping function and the second mapping function are as follows:
[0087]
[0088] in, They represent the first mapping function respectively. The maximum and minimum values within the range of values for the buffer scale d; They represent the second mapping function respectively. The maximum and minimum values within the range of values for the buffer scale d; The first mapping function represents the standardization. This is the standardized second mapping function.
[0089] In step S105-1 above, to quantify the relationship between accuracy and efficiency, an objective function is established using a standardized first mapping function and a standardized second mapping function.
[0090] Define the objective function :
[0091] in, The weight representing precision can be set as follows: ; The weight for efficiency can be set to In the objective function d The scope is subject to Constraints, i.e. .
[0092] In step S105-1 above, the objective function is minimized. The optimal solution for the buffer scale is obtained by solving the problem, and it serves as the optimal buffer range for InSAR deformation monitoring.
[0093] Unlike traditional methods that rely on experience to set buffer zones, this invention proposes an adaptive optimization decision-making mechanism based on a dual-parameter constraint of "accuracy-efficiency" to achieve dynamic optimization. This mechanism constructs an objective function using the real-time statistical trends of inversion error (representing accuracy) and computation time (representing computational cost) with spatial scale *d*. Based on the overlap, slope variation, and saturation characteristics of these two curves, combined with the marginal effect constraint based on coherence point density (representing data quality), the optimal signal-to-noise ratio balance point for long-distance linear engineering InSAR deformation monitoring is automatically determined by solving for the minimum value of the objective function, thus optimizing the buffer range for InSAR deformation monitoring.
[0094] Furthermore, by accurately locating the optimal buffer scale, the calculation of invalid pixels in redundant background areas can be automatically eliminated when InSAR deformation monitoring is performed based on the optimal buffer scale. This significantly reduces the matrix dimension and computational overhead in the singular value decomposition (SVD) and spatial filtering processes. While ensuring millimeter-level inversion accuracy, it enables on-demand allocation of computing resources, significantly improves the computational efficiency and resource optimization rate of InSAR deformation monitoring, and greatly enhances the technical feasibility of high-frequency and rapid safety early warning for ultra-long-distance linear engineering projects.
[0095] In summary, this invention proposes an adaptive method for determining the optimal InSAR deformation monitoring buffer. By constructing a multi-scale spatial observation model, the influence characteristics of spatial geometric constraints on phase unwrapping and deformation inversion accuracy are quantitatively analyzed, thereby determining the optimal observation boundary.
[0096] It is understood that the embodiments of the present invention are based on the South-to-North Water Diversion Project, and the technical solution is also applicable to various long-distance linear projects such as railways, highways, and transportation pipelines.
[0097] Please see Figure 3 , Figure 3 A schematic diagram of an InSAR deformation monitoring buffer optimization system for long-distance linear engineering, provided as an embodiment of the present invention, is shown. The system includes: The data acquisition module 310 is used to acquire time-series SAR image data covering long-distance linear engineering projects; The buffer construction module 320 is used to extract the centerline vector data of the long-distance linear project and construct a gradient buffer set with the centerline vector data as the axis. The efficiency measurement module 330 is used to perform temporal interferometry, coherence point filtering, and phase unwrapping on the temporal SAR data under the constraints of buffers of each scale in the gradient buffer set, and to fit a first mapping function between phase unwrapping time and buffer scale. The accuracy measurement module 340 performs deformation inversion on the phase unwrapping result to obtain deformation data, and fits a second mapping function between the inversion error and the buffer scale based on the inversion error between the deformation data and the reference point data. The scaling optimization module 350 is used to construct an objective function for buffer scaling optimization based on the first mapping function and the second mapping function; and solve the objective function to obtain the optimal solution for the buffer scaling.
[0098] The InSAR deformation monitoring buffer optimization system for long-distance linear engineering described above and the InSAR deformation monitoring buffer optimization method for long-distance linear engineering described above can be referred to in correspondence.
[0099] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4 As shown, the electronic device may include a processor 410, a communication interface 420, a memory 430, and a communication bus 440. The processor 410, communication interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions stored in the memory 430 to execute the InSAR deformation monitoring buffer optimization method for long-distance linear engineering provided in the above-described method embodiments.
[0100] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0101] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute an InSAR deformation monitoring buffer optimization method for long-distance linear engineering provided by the above-described method embodiments.
[0102] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform an InSAR deformation monitoring buffer optimization method for long-distance linear engineering provided by the methods described above.
[0103] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.
[0104] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.
[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for optimizing InSAR deformation monitoring buffers for long-distance linear engineering projects, characterized in that, include: Acquire temporal SAR imagery data covering long-distance linear engineering projects; Extract the centerline vector data of the long-distance linear project and construct a gradient buffer set with the centerline vector data as the axis; Under the constraints of buffers at each scale in the gradient buffer set, the temporal SAR data is subjected to temporal interferometry, coherence point filtering and phase unwrapping, and a first mapping function between phase unwrapping time and buffer scale is fitted. The phase unwrapping result is subjected to deformation inversion to obtain deformation data, and a second mapping function between the inversion error and the buffer scale is fitted based on the inversion error between the deformation data and the reference point data. Based on the first mapping function and the second mapping function, construct an objective function for buffer scale optimization; Solve the objective function to obtain the optimal solution for the buffer scale.
2. The InSAR deformation monitoring buffer optimization method for long-distance linear engineering as described in claim 1, characterized in that, The step of extracting the centerline vector data of the long-distance linear project and constructing a gradient buffer set with the centerline vector data as the axis includes: Extract the centerline vector data of the long-distance linear project based on the time-series SAR image data; A spatial field centered on the centerline vector data is established using the Euclidean distance transformation algorithm. Select several representative segments in the long-distance linear engineering and construct a gradient buffer set that expands from the inside to the outside for the several representative segments in the spatial field.
3. The InSAR deformation monitoring buffer optimization method for long-distance linear engineering as described in claim 1, characterized in that, Under the constraints of buffers at each scale in the gradient buffer set, the temporal SAR data undergoes temporal interferometry, coherence point filtering, and phase unwrapping. A first mapping function between phase unwrapping time and buffer scale is then fitted, including: Temporal interferometry is performed on the range of buffer constraints at each scale of the gradient buffer set in the temporal SAR data to generate interferogram sequences corresponding to each scale buffer. By filtering coherent points in the interferogram sequences corresponding to each scale buffer, the sampling point set corresponding to each scale buffer is obtained. Phase unwrapping is performed on the sampling point sets corresponding to each scale buffer to obtain the phase value of each sampling point, and the phase unwrapping time corresponding to each scale buffer is recorded. Based on the phase unwrapping time corresponding to each scale buffer, a first mapping function between the phase unwrapping time and the buffer scale is fitted.
4. The InSAR deformation monitoring buffer optimization method for long-distance linear engineering as described in claim 3, characterized in that, After filtering coherent points from the interferogram sequence to obtain the sampling point sets corresponding to each scale buffer, the method further includes: The sampling point density corresponding to each scale buffer is statistically analyzed in real time, and a response curve of the sampling point density as the buffer scale changes is established; wherein, the sampling point density is defined as: the number of sampling points that meet the preset threshold condition within a unit geographical area within the buffer mask range; Marginal benefit constraints are established based on the response curve.
5. The InSAR deformation monitoring buffer optimization method for long-distance linear engineering as described in claim 4, characterized in that, The marginal benefit constraint is as follows: Among them, the Represents the change in density at sampling points. This represents the change in the buffer zone scale. This represents the preset rate of change threshold.
6. The InSAR deformation monitoring buffer optimization method for long-distance linear engineering as described in claim 3, characterized in that, The deformation inversion of the phase unwrapping result yields deformation data, and based on the inversion error between the deformation data and the reference point data, a second mapping function between the inversion error and the buffer scale is fitted, including: Deformation inversion is performed on the phase value of each sampling point to obtain the deformation data of each sampling point; The mean square error between the deformation data of each sampling point in each scale buffer and the reference point data is calculated respectively, and a second mapping function between the mean square error and the buffer scale is fitted; wherein, the reference point data is the true value data synchronously observed along the long-distance linear engineering line.
7. The InSAR deformation monitoring buffer optimization method for long-distance linear engineering as described in claim 5, characterized in that, The step of constructing an objective function for buffer scaling optimization based on the first mapping function and the second mapping function includes: The first mapping function and the second mapping function are standardized respectively to obtain the standardized first mapping function and the standardized second mapping function; The standardized first mapping function and the standardized second mapping function are weighted and summed to obtain the objective function for buffer scaling optimization; wherein the objective function is constrained by the marginal benefit constraint.
8. An InSAR deformation monitoring buffer optimization system for long-distance linear engineering, characterized in that, include: The data acquisition module is used to acquire time-series SAR image data covering long-distance linear engineering projects; A buffer construction module is used to extract the centerline vector data of the long-distance linear project and construct a gradient buffer set with the centerline vector data as the axis. The efficiency measurement module is used to perform temporal interferometry, coherence point filtering, and phase unwrapping on the temporal SAR data under the constraints of buffers of each scale in the gradient buffer set, and to fit a first mapping function between phase unwrapping time and buffer scale. The accuracy measurement module performs deformation inversion on the phase unwrapping result to obtain deformation data, and fits a second mapping function between the inversion error and the buffer scale based on the inversion error between the deformation data and the reference point data. The scaling optimization module is used to construct an objective function for buffer scaling optimization based on the first mapping function and the second mapping function; and to solve the objective function to obtain the optimal solution for the buffer scaling.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the InSAR deformation monitoring buffer optimization method for long-distance linear engineering as described in any one of claims 1 to 7.
10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the InSAR deformation monitoring buffer optimization method for long-distance linear engineering as described in any one of claims 1 to 7.