Bridge connection sar tomography reference net generation method based on region growing

Abstract

This application relates to a method for generating a bridge-connected SAR tomographic reference network based on region growing. The method includes: processing the connection network with the fewest SPS points using a convex hull function to obtain the minimum bounding polygon of the minimum connection network; defining a search range with any point in the vertex set of the bounding polygon as the center and a connection bridge search radius threshold as the radius; adding SPS points of non-self-connected networks within the search range to a candidate point set; connecting these points with points in the candidate point set using the center point as the candidate point to form candidate connection bridges; updating and judging the candidate connection bridge set based on multiple candidate connection bridges; outputting the final connection bridge set when the candidate point set is empty; merging multiple connection networks connected by the minimum connection network and the connection bridges in the final connection bridge set; and outputting the SAR tomographic reference network when the total number of all connection networks is 1. This method can construct a SAR tomographic reference network that covers the entire scene to the greatest extent.

Description

Technical Field

[0001] This application relates to the field of remote sensing imaging technology, and in particular to a method for generating bridge-connected SAR tomographic reference networks based on region growing. Background Technology

[0002] As an extension of InSAR technology, SAR tomography (TomoSAR) imaging technology also utilizes multi-angle coherent measurements of the same target to form a synthetic aperture in the height direction, and then reconstructs the true SAR imaging 3D scene through height-oriented imaging. Spaceborne SAR tomography systems can further achieve height-dimensional spatial resolution, i.e., 3D imaging, based on the traditional azimuth-range 2D high-resolution imaging detection of spaceborne SAR systems. Due to its height-oriented resolution, TomoSAR can separate overlapping targets in the height direction and extract the 3D spatial position and backscattering information of each scattering element, making it a true 3D imaging technology.

[0003] The atmospheric phase screen error (ASPS) in SAR images acquired by a spaceborne multi-pass SAR system at each pass is closely related to the temperature, humidity, pressure, and total electron content of the ionosphere at each pass. It can be approximated as white noise, varying randomly with the observation time. This leads to reduced coherence of the SAR tomography signal, making it difficult to focus in the altitude direction, thus severely impacting spaceborne SAR tomography 3D imaging. Therefore, atmospheric phase error is a significant error source in spaceborne SAR tomography 3D imaging and must be corrected before imaging. Since ASPS has a certain spatial correlation, it can be assumed that adjacent pixels experience similar ASPS errors. Therefore, phase difference between adjacent pixels can significantly cancel out ASPS errors. This is the basic principle of atmospheric phase correction based on a two-layer network. Currently, the Delaunay triangulation method is mainly used when creating the first reference network of the two-layer network. Delaunay triangulation is a networking method based on the optimal triangle principle, resulting in a low redundancy (approximately 3) in the Delaunay triangulation network. Therefore, Delaunay triangulation is prone to the problem of isolated points and islands of Persistent Scatterers (SPS) during network optimization. Thus, to fully utilize multiple discrete and disconnected networks in the scene, it is necessary to consider connecting them. How to connect them and how to achieve high-quality connections then become a crucial issue. Summary of the Invention

[0004] Therefore, it is necessary to provide a method for generating SAR tomographic reference networks based on region growing, which can construct SAR tomographic reference networks that cover the entire scene to the greatest extent, in order to address the above-mentioned technical problems.

[0005] A method for generating a bridge-connected SAR tomographic reference network based on region growing, the method comprising:

[0006] Acquire multiple remote sensing images within the region of interest; establish and optimize Delaunay triangulation for the multiple remote sensing images to obtain multiple connected networks; the connected networks include multiple SPS points; initialize the connection bridge parameters of the circumscribed polygon vertices of the connected networks to obtain the threshold for the number of connection bridges, the threshold for connection bridge quality, and the threshold for connection bridge search radius of the circumscribed polygon vertices of the connected networks;

[0007] Sort the connections based on the number of SPS points in each connection network, and use the convex hull function to process the connection network with the fewest SPS points to obtain the minimum bounding polygon of the minimum connection network; the minimum bounding polygon includes the vertex set.

[0008] Using any point in the vertex set as the center and the connection bridge search radius threshold as the radius, the search range is defined. SPS points that are not in their own connection network within the search range are included in the candidate point set. The center of the circle is used to connect the points in the candidate point set to form candidate connection bridges.

[0009] A candidate bridge set is formed by multiple candidate bridges. Phase difference and spectral estimation algorithms are performed starting from the shortest bridge in the candidate bridge set to obtain the imaging result. The imaging result is then detected according to a pre-set bridge quality threshold. The candidate bridge set is updated using the detection result, and bridges that meet the criteria are added to the candidate bridge set.

[0010] The number of connecting bridges for a single circumscribed polygon vertex is determined based on a preset threshold. When the number of connecting bridges reaches the threshold or the candidate connecting bridge set is empty, the connecting bridge search for the circumscribed polygon vertex ends. The connecting bridge search continues for vertices in the vertex set that have not been searched. After the connecting bridge search for all vertices of the smallest circumscribed polygon is completed, the final connecting bridge set is output.

[0011] The minimum connection network and the connection bridges in the final connection bridge set are merged. When the total number of all connection networks is 1, the SAR tomography reference network is output.

[0012] In one embodiment, the connection bridge parameters of the circumscribed polygon vertices of the network are initialized to obtain a threshold for the number of connection bridges, a threshold for the quality of connection bridges, and a threshold for the search radius of connection bridges at the circumscribed polygon vertices of the network, including:

[0013] The parameters of the circumscribed polygon vertices of the network are initialized to obtain the threshold number of connecting bridges of the circumscribed polygon vertices of the network as B. thres ∈[2,5], the bridge quality threshold is RSR thres ∈[0.6,0.8]; the threshold for the bridge search radius is D. thres ∈[30,50]m.

[0014] In one embodiment, the vertex set of the minimum bounding polygon is Where convhull is the convex hull function. Let x-coordinate vectors be the vectors connecting the points in the network. This is the ordinate vector connecting the points in the network.

[0015] In one embodiment, SPS points that are not part of their own interconnected network within the search range are included in a candidate point set, and candidate connection bridges are formed by connecting the points in the candidate point set with their centers, including:

[0016] SPS points that are not part of their own interconnected network within the search range are added to the candidate point set. Candidate connection bridges are formed by connecting the points in the candidate point set with the center of the circle:

[0017]

[0018] Where, p j For candidate point set P dist The point in the middle, , respectively, are the spatial coordinates of the reference net vertex and the candidate vertex, and D[■] represents the Euclidean distance operator.

[0019] In one embodiment, the imaging results are detected according to a pre-set quality threshold for the connecting bridges, and the candidate connecting bridge set is updated using the detection results. Connecting bridges that meet the criteria are added to the candidate connecting bridge set, including:

[0020] The imaging results are detected based on a pre-set connection bridge quality threshold. If the connection quality of the connection bridge is not less than the pre-set connection bridge quality threshold, the SPS edge corresponding to the connection bridge with a quality not less than the pre-set connection bridge quality threshold is retained. Otherwise, the SPS edge is deleted, and the connection bridge with a quality not less than the pre-set connection bridge quality threshold is added to the candidate connection bridge set.

[0021] In one embodiment, the number of connecting bridges for a single circumscribed polygon vertex is determined based on a preset threshold. When the number of connecting bridges reaches the threshold or the candidate connecting bridge set is empty, the connecting bridge search for the circumscribed polygon vertex ends. The connecting bridge search continues for vertices in the vertex set that have not yet been searched. After completing the connecting bridge search for all vertices of the smallest circumscribed polygon, the final connecting bridge set is output, including:

[0022] The number of connecting bridges for a single circumscribed polygon vertex is determined based on a preset threshold for the number of connecting bridges. The search for connecting bridges for the circumscribed polygon vertex ends when the number of connecting bridges reaches the threshold or the candidate connecting bridge set is empty. The search continues for vertices not yet searched in the vertex set. A search range is defined using any unsearched vertex as the center and a connecting bridge search radius threshold as the radius. SPS points that are not part of their own network within the search range are added to the candidate point set, and these points are connected to points in the candidate point set using the center of the circle to form candidate connecting bridges. A candidate connecting bridge set is formed based on multiple candidate connecting bridges. Phase difference and spectral estimation algorithms are performed starting from the shortest connecting bridge in the candidate connecting bridge set to obtain the imaging result. The imaging result is then tested according to a preset connecting bridge quality threshold, and the candidate connecting bridge set is updated using the test results. Connecting bridges that meet the criteria are added to the candidate connecting bridge set. The search for connecting bridges for a single circumscribed polygon vertex ends when the number of connecting bridges reaches the threshold or the candidate connecting bridge set is empty.

[0023] In one embodiment, the final number of connection bridges in the connection bridge set does not exceed a preset threshold.

[0024] In one embodiment, multiple connection networks connected by bridges in the minimum connection network and the final connection bridge set are merged. When the total number of all connection networks is 1, a SAR tomographic reference network is output, including:

[0025] The H connection networks connected by the minimum connection network and the final connection bridge set are merged. The H connection networks with fewer SPS points are merged into the largest connection network in the H+1 connection network. The SAR tomography reference network is output when the total number of connection networks is 1 after subtracting the H connection networks from the total number of connection networks.

[0026] In one embodiment, the update method for the largest set of nodes and edges of the connected network is as follows:

[0027]

[0028] Where P represents the set of points, E represents the set of edges, and h represents the connection network number.

[0029] The aforementioned method for generating a bridge-connected SAR tomographic reference network based on region growing, in this application, establishes and optimizes a triangulated network for the remote sensing imaging region. It uses the vertices of the smallest bounding polygon of the minimum connected network as the starting point for growth, connecting multiple connected networks scattered throughout the scene. While ensuring proximity, it selects connecting edges with high imaging quality as bridges between two connected networks, connecting them to neighboring connected networks using these high-quality bridges. This iterative process aims to obtain the maximum bridge-connected reference network, thereby reducing the risk of error propagation between connected networks during overall network calculation. Simultaneously, it connects isolated islands, improving the regional coverage of the reference network. The constructed bridge-connected reference network can cover the entire scene to the maximum extent, thus reducing the existence of isolated islands, and the connecting bridges between the connected networks ensure high-quality imaging. Attached Figure Description

[0030] Figure 1 This is a flowchart illustrating a method for generating a bridge-connected SAR tomographic reference network based on region growing in one embodiment.

[0031] Figure 2 This is an average intensity map of all SAR images in one embodiment;

[0032] Figure 3 This is a schematic diagram of a reference network generated using the original bridge connection method in one embodiment;

[0033] Figure 4 This is a schematic diagram of a novel bridge-type connection reference network established using the method proposed in this application in another embodiment;

[0034] Figure 5 This is a schematic diagram of the maximum reference network obtained based on the original bridge connection method in one embodiment;

[0035] Figure 6 This is a schematic diagram of the maximum reference net obtained based on the method proposed in this application in one embodiment;

[0036] Figure 7 This is an SSPS point height image obtained by solving the original bridge-connected reference network in one embodiment;

[0037] Figure 8 This is an image of the SSPS point height obtained by solving the novel bridge-connected reference network of this application in one embodiment. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0039] In one embodiment, such as Figure 1 As shown, a method for generating a bridge-connected SAR tomographic reference network based on region growing is provided, including the following steps:

[0040] Step 102: Acquire multiple remote sensing images within the region of interest; establish and optimize Delaunay triangulation for the multiple remote sensing images to obtain multiple connected networks; the connected networks include multiple SPS points; initialize the connection bridge parameters of the circumscribed polygon vertices of the connected networks to obtain the threshold for the number of connection bridges, the threshold for connection bridge quality, and the threshold for connection bridge search radius of the circumscribed polygon vertices of the connected networks.

[0041] The threshold for the number of connecting bridges is the upper limit for the number of connecting bridges to each vertex of the polygon. If the threshold is too small, the network structure may become unstable. If the threshold is too large, the network structure may become too redundant, increasing the computation cost. It needs to be set according to specific requirements.

[0042] The formula for calculating the mass of the connecting bridge is:

[0043]

[0044] Wherein, Δg n This represents the differential signal between the endpoints of the connecting bridge. To estimate the obtained differential signal, N is the number of observed images.

[0045] Connector bridge quality threshold RSR thres The general setting is:

[0046] RSR thres ∈[0.6,0.8]

[0047] This value needs to be adjusted according to the actual situation. If it is set too small, it will reduce the accuracy of the overall network solution; if it is set too large, it will be difficult to find connecting bridges, causing the network to become isolated. By setting thresholds for the number of connecting bridges at the circumscribed polygon vertices, the quality of connecting bridges, and the search radius of connecting bridges, it is possible to find the connecting bridges with the best quality.

[0048] Step 104: Sort the connections according to the number of SPS points of each connection network, and use the convex hull function to process the connection network with the fewest SPS points to obtain the minimum bounding polygon of the minimum connection network; the minimum bounding polygon includes the vertex set.

[0049] Sort each network according to the number of SPS points it contains. Assuming there are Q networks, the resulting sorted sequence is as follows:

[0050] S Net ={S1>S2>…>S Q};

[0051] Sort the connections by the number of SPS points in each network from highest to lowest, and select the network with the smallest number of SPS points for each bridge connection. The smallest network S is obtained from the sorting. Q The minimum bounding polygon of the small connected network is obtained using the convex hull function, and thus the vertex set of the bounding polygon is obtained:

[0052]

[0053] Where convhull is the convex hull function. Let x-coordinate vectors be the vectors connecting the points in the network. This is the ordinate vector connecting the points in the network.

[0054] Step 106: Determine the search range with any point in the vertex set as the center and the connection bridge search radius threshold as the radius. Include the SPS points that are not in their own connection network within the search range into the candidate point set, and connect them with the points in the candidate point set with the center of the circle to form candidate connection bridges.

[0055] A point p in the vertex set of the circumscribed polygon i Using the center as the search radius threshold D for the connecting bridge thres Define the search range for the radius, and add the SPS points that are not self-connected within this range to the candidate point set P. dist Then, this vertex is connected to points in the candidate point set to form candidate connection bridges.

[0056] Using any point in the circumscribed polygon vertex set as the center, a bridge search is performed. If no suitable bridge is found around the vertex, the search jumps to the next vertex. After traversing each vertex in the circumscribed polygon vertex set, multiple bridges corresponding to the circumscribed polygon vertex set can be obtained.

[0057] Step 108: Based on multiple candidate bridges forming a candidate bridge set, phase difference and spectral estimation algorithms are performed starting from the shortest bridge in the candidate bridge set to obtain the imaging result. The imaging result is then inspected according to a pre-set bridge quality threshold, and the candidate bridge set is updated using the inspection results. Bridges meeting the criteria are added to the candidate bridge set. The number of bridges connecting a single circumscribed polygon vertex is determined according to a pre-set bridge quantity threshold. When the number of bridges reaches the threshold or the candidate bridge set is empty, the bridge search for the circumscribed polygon vertex ends. The bridge search continues for vertices not yet searched in the vertex set. After completing the bridge search for all vertices of the smallest circumscribed polygon, the final bridge set is output.

[0058] A candidate bridge set is constructed based on multiple candidate bridges. These bridges are then sorted by length in ascending order. Phase difference and spectral estimation algorithms are performed starting with the shortest bridge to obtain the imaging result, which represents the imaging quality. Bridges with imaging quality not less than a pre-set quality threshold are added to the updated bridge set. Simultaneously, the total number of connections formed by each polygon vertex must not exceed the threshold.

[0059] B total ={B total}∪{B k},|B k |≤B thres .

[0060] Then, determine whether there are still circumscribed polygon vertices. If the minimum circumscribed polygon vertex set is not empty, continue to repeat steps 106-108 until the minimum circumscribed polygon vertex set is empty, and then output the final connection bridge set.

[0061] Step 110: Merge the multiple connection networks connected by the minimum connection network and the connection bridges in the final connection bridge set. When the total number of all connection networks is 1, output the SAR tomography reference network.

[0062] Merge the minimum connection network and the H connection networks connected to it by the bridges in its final bridge set. The merging rule is: merge the H connection networks with fewer SPS points into the largest connection network among these H+1 connection networks. The point set P and edge set E of the largest connection network are then updated as follows:

[0063]

[0064] The total number of all connected networks is now updated as follows:

[0065] Num = QH

[0066] If Num≠1, continue repeating steps 104 to 110 until the total number of connected networks becomes 1, and the maximum number of connected networks output is the SAR tomography reference network.

[0067] In the aforementioned method for generating a bridged SAR tomographic reference network based on region growing, this application establishes and optimizes a triangulated network for the remote sensing imaging region. The minimum bounding polygon vertex of the minimum connected network is considered as the starting point for growth. Multiple connected networks scattered throughout the scene are connected. While ensuring proximity, connecting edges with high imaging quality are selected as bridges between two connected networks. These high-quality bridges are then used to connect the network to neighboring connected networks. This process is iterated to obtain the maximum bridged reference network, thereby reducing the risk of error propagation between connected networks during overall network calculation. Simultaneously, connecting isolated islands improves the regional coverage of the reference network. The constructed bridged reference network can cover the entire scene to the maximum extent, thus reducing the existence of isolated islands. Furthermore, the connecting bridges between connected networks ensure high-quality imaging.

[0068] In one embodiment, the connection bridge parameters of the circumscribed polygon vertices of the network are initialized to obtain a threshold for the number of connection bridges, a threshold for the quality of connection bridges, and a threshold for the search radius of connection bridges at the circumscribed polygon vertices of the network, including:

[0069] The parameters of the circumscribed polygon vertices of the network are initialized to obtain the threshold number of connecting bridges of the circumscribed polygon vertices of the network as B. thres ∈[2,5], the bridge quality threshold is RSR thres ∈[0.6,0.8]; the threshold for the bridge search radius is D. thres ∈[30,50]m.

[0070] In one embodiment, the vertex set of the minimum bounding polygon is Where convhull is the convex hull function. Let x-coordinate vectors be the vectors connecting the points in the network. This is the ordinate vector connecting the points in the network.

[0071] In one embodiment, SPS points that are not part of their own interconnected network within the search range are included in a candidate point set, and candidate connection bridges are formed by connecting the points in the candidate point set with their centers, including:

[0072] SPS points that are not part of their own interconnected network within the search range are added to the candidate point set. Candidate connection bridges are formed by connecting the points in the candidate point set with the center of the circle:

[0073]

[0074] Where, p j For candidate point set P dist The point in the middle, , respectively, are the spatial coordinates of the reference net vertex and the candidate vertex, and D[■] represents the Euclidean distance operator.

[0075] In one embodiment, the imaging results are detected according to a pre-set quality threshold for the connecting bridges, and the candidate connecting bridge set is updated using the detection results. Connecting bridges that meet the criteria are added to the candidate connecting bridge set, including:

[0076] The imaging results are detected based on a pre-set connection bridge quality threshold. If the connection quality of the connection bridge is not less than the pre-set connection bridge quality threshold, the SPS edge corresponding to the connection bridge with a quality not less than the pre-set connection bridge quality threshold is retained. Otherwise, the SPS edge is deleted, and the connection bridge with a quality not less than the pre-set connection bridge quality threshold is added to the candidate connection bridge set.

[0077] In one embodiment, the number of connecting bridges for a single circumscribed polygon vertex is determined based on a preset threshold. When the number of connecting bridges reaches the threshold or the candidate connecting bridge set is empty, the connecting bridge search for the circumscribed polygon vertex ends. The connecting bridge search continues for vertices in the vertex set that have not yet been searched. After completing the connecting bridge search for all vertices of the smallest circumscribed polygon, the final connecting bridge set is output, including:

[0078] The number of connecting bridges for a single circumscribed polygon vertex is determined based on a preset threshold for the number of connecting bridges. The search for connecting bridges for the circumscribed polygon vertex ends when the number of connecting bridges reaches the threshold or the candidate connecting bridge set is empty. The search continues for vertices not yet searched in the vertex set. A search range is defined using any unsearched vertex as the center and a connecting bridge search radius threshold as the radius. SPS points that are not part of their own network within the search range are added to the candidate point set, and these points are connected to points in the candidate point set using the center of the circle to form candidate connecting bridges. A candidate connecting bridge set is formed based on multiple candidate connecting bridges. Phase difference and spectral estimation algorithms are performed starting from the shortest connecting bridge in the candidate connecting bridge set to obtain the imaging result. The imaging result is then tested according to a preset connecting bridge quality threshold, and the candidate connecting bridge set is updated using the test results. Connecting bridges that meet the criteria are added to the candidate connecting bridge set. The search for connecting bridges for a single circumscribed polygon vertex ends when the number of connecting bridges reaches the threshold or the candidate connecting bridge set is empty.

[0079] In one embodiment, the final number of connection bridges in the connection bridge set does not exceed a preset threshold.

[0080] In one embodiment, multiple connection networks connected by bridges in the minimum connection network and the final connection bridge set are merged. When the total number of all connection networks is 1, a SAR tomographic reference network is output, including:

[0081] The H connection networks connected by the minimum connection network and the final connection bridge set are merged. The H connection networks with fewer SPS points are merged into the largest connection network in the H+1 connection network. The SAR tomography reference network is output when the total number of connection networks is 1 after subtracting the H connection networks from the total number of connection networks.

[0082] In one embodiment, the update method for the largest set of nodes and edges of the connected network is as follows:

[0083]

[0084] Where P represents the set of points, E represents the set of edges, and h represents the connection network number.

[0085] In a specific embodiment, such as Figure 2 As shown, Figure 2 This is the average intensity map of all SAR images, with the horizontal axis representing the azimuth and the vertical axis representing the range. This map is based on measured data obtained from the TerraSAR-X satellite, using 27 X-band Staring images in observation mode, with azimuth and range resolutions of 0.24m and 0.59m, respectively. Figure 3 This is a reference network generated using the existing bridging method. White lines represent connecting bridges, and the other colored lines represent sub-connecting networks. The strategy of this bridging method is to find connecting bridges between the largest sub-network and all other connecting networks, and to use all connecting bridges that meet the distance constraints. This results in overly complex and redundant connecting bridges, and some sub-networks may be too far from the largest sub-network to be connected by bridges, thus reducing the coverage of the largest reference network.

[0086] Figure 4 This application presents a novel bridge-type connection reference network established using the method proposed in this application. Compared with existing methods, this application connects each subnet to one or more of its nearest sub-connection networks. The connection bridge structure of this reference network is simple, with low redundancy, and the reference network generated based on the method proposed in this invention can basically utilize the existing sub-connection networks to cover the entire scene as much as possible.

[0087] Figure 5 This is the largest reference mesh obtained based on the original bridge connection method. It is clear that the sub-connected meshes located in the lower left and lower right corners of the scene are not connected, resulting in these areas not being covered.

[0088] Figure 6 This is the maximum reference network obtained based on the method proposed in this application. Most subnets are connected together, making the maximum reference network as close as possible to the sum of all subnets, thus maximizing coverage.

[0089] Figure 7This image shows the SSPS point heights obtained using the existing bridge-connected reference network. There are significant issues with the inversion heights of the building in the upper right corner and other buildings, and the heights of buildings in areas previously not covered by the network cannot be obtained.

[0090] Figure 8 This image shows the SSPS point heights calculated using the novel bridge-connected reference network of this application. The heights of most buildings in the scene correspond to actual heights, and the SSPS point coverage within the scene is also the widest possible.

[0091] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process 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 executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0092] 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.

[0093] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for generating a bridge-connected SAR tomographic reference network based on region growing, characterized in that, The method includes: Acquire multiple remote sensing images within the region of interest; establish and optimize a Delaunay triangulation network for the multiple remote sensing images to obtain multiple connected networks; each connected network includes multiple SPS points; initialize the connection bridge parameters of the circumscribed polygon vertices of the connected network to obtain the threshold values ​​for the number of connection bridges, the quality of connection bridges, and the search radius of connection bridges for the circumscribed polygon vertices of the connected network. Sort the connections based on the number of SPS points in each connection network, and use the convex hull function to process the connection network with the fewest SPS points to obtain the minimum bounding polygon of the minimum connection network; the minimum bounding polygon includes the vertex set. Using any point in the vertex set as the center and the connection bridge search radius threshold as the radius, the search range is defined. SPS points that are not in their own connection network within the search range are included in the candidate point set. The center of the circle is used to connect with the points in the candidate point set to form candidate connection bridges. A candidate bridge set is formed by multiple candidate bridges. Phase difference and spectral estimation algorithms are performed starting from the shortest bridge in the candidate bridge set to obtain the imaging result. The imaging result is then detected according to a pre-set bridge quality threshold. The candidate bridge set is updated using the detection result, and bridges that meet the criteria are added to the candidate bridge set. The number of connecting bridges for a single circumscribed polygon vertex is determined based on a preset threshold for the number of connecting bridges. When the number of connecting bridges reaches the threshold or the candidate connecting bridge set is empty, the connecting bridge search for the circumscribed polygon vertex ends. The connecting bridge search continues for vertices in the vertex set that have not been searched. After the connecting bridge search for all vertices of the smallest circumscribed polygon is completed, the final connecting bridge set is output. The minimum connection network and the multiple connection networks connected by the connection bridges in the final connection bridge set are merged, and the SAR tomography reference network is output when the total number of all connection networks is 1.

2. The method according to claim 1, characterized in that, The connection bridge parameters of the circumscribed polygon vertices of the network are initialized to obtain the connection bridge quantity threshold, connection bridge quality threshold, and connection bridge search radius threshold of the circumscribed polygon vertices of the network, including: The parameters of the circumscribed polygon vertices of the connection network are initialized to obtain the threshold number of connection bridges of the circumscribed polygon vertices of the connection network. The quality threshold of the connecting bridge is The threshold for the search radius of the connecting bridge is: .

3. The method according to claim 1, characterized in that, The vertex set of the minimum bounding polygon is Where convhull is the convex hull function. Let x-coordinate vectors be the vectors connecting the points in the network. This is the ordinate vector connecting the points in the network.

4. The method according to claim 2, characterized in that, SPS points that are not part of their own interconnected network within the search range are added to the candidate point set. Candidate connection bridges are formed by connecting these points with points in the candidate point set, centered on a circle. These bridges include: SPS points that are not part of their own interconnected network within the search range are added to the candidate point set. Candidate connection bridges are formed by connecting the points in the candidate point set with the center of the circle: in, For candidate point set The point in the middle, , These are the spatial coordinates of the reference network vertex and the candidate vertex, respectively. This represents the Euclidean distance operator.

5. The method according to claim 4, characterized in that, The imaging results are detected based on a pre-set quality threshold for the connecting bridges, and the candidate connecting bridge set is updated using the detection results. Connecting bridges that meet the criteria are added to the candidate connecting bridge set, including: The imaging results are detected according to a pre-set connection bridge quality threshold. If the connection quality of the connection bridge is not less than the pre-set connection bridge quality threshold, the SPS edge corresponding to the connection bridge with a quality not less than the pre-set connection bridge quality threshold is retained; otherwise, the SPS edge is deleted, and the connection bridge with a quality not less than the pre-set connection bridge quality threshold is added to the candidate connection bridge set.

6. The method according to claim 1, characterized in that, The number of connecting bridges for a single circumscribed polygon vertex is determined based on a preset threshold for the number of connecting bridges. When the number of connecting bridges reaches the threshold or the candidate connecting bridge set is empty, the connecting bridge search for the circumscribed polygon vertex ends. Continue searching for connecting bridges on the vertices not yet searched in the vertex set. After completing the connecting bridge search for all vertices of the minimum bounding polygon, output the final connecting bridge set, including: The number of connecting bridges for a single circumscribed polygon vertex is determined based on a preset threshold for the number of connecting bridges. When the number of connecting bridges reaches the threshold or the candidate connecting bridge set is empty, the connecting bridge search for the circumscribed polygon vertex ends. The search continues for connecting bridges for vertices not yet searched in the vertex set. A search range is defined with any unsearched vertex in the vertex set as the center and a connecting bridge search radius threshold as the radius. SPS points that are not part of their own interconnected network within the search range are added to the candidate point set, and the center of each point is connected to a point in the candidate point set to form a candidate connecting bridge. A candidate connecting bridge set is formed based on multiple candidate connecting bridges. Phase difference and spectral estimation algorithms are performed starting from the shortest connecting bridge in the candidate connecting bridge set to obtain the imaging result. The imaging result is then tested according to a preset connecting bridge quality threshold, and the candidate connecting bridge set is updated using the test results. Connecting bridges that meet the criteria are added to the candidate connecting bridge set. The number of connecting bridges for a single circumscribed polygon vertex is determined based on a preset threshold for the number of connecting bridges. When the number of connecting bridges reaches the threshold or the candidate connecting bridge set is empty, the connecting bridge search for the circumscribed polygon vertex ends. After completing the connecting bridge search for all vertices of the smallest circumscribed polygon, the final connecting bridge set is output.

7. The method according to claim 1, characterized in that, The minimum connection network and the multiple connection networks connected by the connection bridges in the final connection bridge set are merged. When the total number of all connection networks is 1, a SAR tomographic reference network is output, including: Connecting the minimum connection network and the connection bridges in the final connection bridge set H The connection networks are merged, and those with fewer SPS points are merged. H A connection network was incorporated In the largest connection network, subtract the total number of connections. H A total of 1 connection network is used to output the SAR tomography reference network.

8. The method according to claim 7, characterized in that, The update method for the vertex set and edge set of the largest connected network is as follows: in, P represents a point set, E Represents an edge set. Indicates the connection number.

Images