A method for detecting overlapping regions in 3D SAR

Abstract

This invention provides a method for detecting overlapping regions in 3D SAR, including collecting and processing image area data; identifying overlapping regions in each image within the area data and merging initial overlapping vector maps; using an overlapping homogeneity coefficient to compare and filter images in the planar dimension to obtain a final overlapping vector map; calculating elevation difference maps by combining processed interferometric DSM data products and point cloud DSM data products within the image area data; calculating slope difference maps by combining processed interferometric DSM data products and point cloud DSM data products within the image area data; and comprehensively verifying the final overlapping vector map, elevation difference map, and slope difference map to obtain the final overlapping region. This invention identifies and verifies overlapping regions in both planar and elevation dimensions, uses an overlapping homogeneity coefficient to filter non-overlapping regions in the 3D SAR DOM data products, and proposes a comprehensive weighting method to calculate the elevation difference between interferometric DSM data products and point cloud DSM data products to determine overlapping regions.

Description

Technical Field

[0001] This invention belongs to the technical field, specifically relating to a method for detecting three-dimensional SAR overlay regions. Background Technology

[0002] Synthetic Aperture Radar (SAR) technology offers unique advantages in topographic mapping applications: it can perform topographic mapping under any weather conditions, unaffected by factors such as clouds, rain, and fog. It can achieve large-scale, one-time mapping of the entire land surface, making it suitable for monitoring and analyzing regional topographic changes. For areas with significant topographic relief, SAR technology can provide high-precision topographic mapping results. This technology is a non-contact remote sensing mapping method, eliminating the need for on-site data collection, thus reducing manpower and material costs and measurement risks.

[0003] During SAR imaging, the emitted radar pulse causes the echoes from closer targets (i.e., the top of tall targets) to arrive first, followed by the echoes from farther targets (i.e., the bottom of tall targets). This results in a top-first, bottom-later image, or an inverted top-and-bottom image, a phenomenon known as overlay. Overlay areas manifest as brightness anomalies or geometric misalignments, commonly found in steep terrain or densely built-up urban areas. This phenomenon not only affects the normal interpretation of images but also poses challenges to applications such as terrain mapping and target recognition based on SAR data.

[0004] Currently, 3D SAR can be used with array 3D imaging radar to acquire radar echo data from multiple orbits in a single flight. 3D imaging algorithms can generate 3D point cloud data, obtain 3D structural information of ground features, and eliminate obscured areas in traditional SAR observation geometry by covering different sides of the target during flight, effectively eliminating the overlapping problem in mountainous areas.

[0005] The data processed by 3D SAR includes interferometric, 3D point cloud, and image data. The data undergoes regional network adjustment, interferometry, point cloud registration and adjustment, point cloud stitching, and image orthorectification. The processing flow is as follows: Regional adjustment is performed on all image data to achieve high-precision geometric orientation and positioning of all images under sparse control conditions, while simultaneously acquiring the geographic coordinates of densified points. Then, interferometric processing is performed on the interferometric data, using densified points obtained from regional network adjustment as reference points for interferometric parameter calibration, generating an interferometric DSM data product. For the 3D point cloud data, denoising, registration, adjustment, and fusion are performed to generate 3D point cloud data, and point cloud DSM data products are generated using point cloud data interpolation. Finally, using the data from the extracted DSM, combined with the regional network adjustment orientation results, orthorectification and regional stitching are performed on the images to generate a DOM data product. The DSM data product includes both interferometric DSM data products and point cloud DSM data products.

[0006] However, existing methods still have the following problems: ① A method and apparatus for evaluating the reconstruction quality of tomographic SAR three-dimensional point clouds (CN 113470002 B): A method for evaluating the reconstruction quality of tomographic SAR three-dimensional point clouds is disclosed, including: normalizing the tomographic SAR three-dimensional point cloud to obtain a normalized tomographic SAR three-dimensional point cloud; performing grayscale quantization on the normalized tomographic SAR three-dimensional point cloud to obtain a grayscale quantized tomographic SAR three-dimensional point cloud; determining a cube neighborhood space of N×N×N point cloud pixels from the grayscale quantized tomographic SAR three-dimensional point cloud; determining the statistical probability distribution P(i,j) of the combination χ(i,j) of the grayscale value i of the point cloud pixel and the grayscale mean j of the cube neighborhood space of the grayscale value i of the point cloud pixel; calculating the three-dimensional entropy of the tomographic SAR three-dimensional point cloud based on the statistical probability distribution P(i,j) of the combination χ(i,j); and evaluating the reconstruction quality of the tomographic SAR three-dimensional point cloud obtained by different reconstruction algorithms based on the value of the three-dimensional entropy. This method evaluates the quality of 3D SAR point cloud reconstruction but does not evaluate the quality of overlapping regions. ② A SAR image overlay detection method, apparatus, device, and storage medium (CN 119780920 A): This discloses a SAR image overlay detection method, apparatus, device, and storage medium. The method includes: acquiring SAR time-series image data, external DEM data, and external surface classification cover data for the current detection area; obtaining a regional amplitude map of the current detection area based on the SAR time-series image data and normalizing the regional amplitude map; obtaining a corresponding interferometric phase map based on the SAR time-series image data and calculating the coherence coefficient; normalizing the external DEM data; then, band-stitching the normalized regional amplitude map, interferometric phase map, coherence coefficient, normalized external DEM data, and external surface classification cover data to generate corresponding data blocks; and inputting the data blocks into a preset overlay recognition model to obtain the overlay recognition result for the current detection area. This method can improve the accuracy of SAR overlay recognition. This method is for detecting SAR image overlay, but it does not perform three-dimensional verification of the overlay area. Summary of the Invention

[0007] This invention provides a method for detecting overlapping regions in three-dimensional SAR, the improvement of which is that the method includes:

[0008] (1) Collect and process data of the shooting area;

[0009] (2) Identify the overlapping regions of each image in the region data and merge the initial overlapping vector maps;

[0010] (3) Use the overlay homogeneity coefficient to compare and filter the final overlay vector image in the planar dimension;

[0011] (4) Calculate the elevation difference map by combining the processed interferometric DSM data product and point cloud DSM data product within the shooting area data;

[0012] (5) Calculate the slope difference map by combining the processed interference DSM data product and point cloud DSM data product within the shooting area data;

[0013] (6) The final overlapping area is obtained by comprehensively verifying the final overlay vector map, elevation difference map and slope difference map.

[0014] Furthermore, step (1) includes collecting raw SAR image data, point cloud data, processed interferometric DSM data products, point cloud DSM data products and DOM data products, digital elevation data and land use cover maps within the shooting area.

[0015] Furthermore,

[0016] (2.1) Construct a geometric model based on the original SAR image data and digital elevation data;

[0017] (2.2) Calculate the slant range of the radar to any point, determine the overlapping area based on the slant range, and generate a vector map of the overlapping area corresponding to each original SAR image;

[0018] (2.3) Merge the overlay vector maps to obtain the initial overlay vector map of the entire shooting area.

[0019] Furthermore, step (3) includes using the overlay homogeneity coefficient to filter the overlay region in the planar dimension based on the original SAR image data, DOM data product and initial overlay vector image, to obtain the final overlay vector image after filtering.

[0020] Furthermore,

[0021] (3.1) Perform corrosion treatment on the overlapping areas;

[0022] (3.2) Preset the overlay homogeneity coefficient;

[0023] (3.3) Calculate and obtain the overlay homogeneity coefficient on the original SAR image data;

[0024] (3.4) Calculate the overlay homogeneity coefficient on the obtained DOM data product;

[0025] (3.5) Calculate the difference in the overlay homogeneity coefficient;

[0026] (3.6) The final overlay vector diagram is obtained by comparing and filtering the overlay homogeneity coefficients based on the initial overlay vector diagram.

[0027] Furthermore, step (4) includes

[0028] (4.1) Select scene points of areal features as reference points based on interferometric DSM data products, point cloud DSM data products and land use cover maps;

[0029] (4.2) Calculate the average elevation and variance of the points;

[0030] (4.3) Calculate the elevation difference and overall variance of the points;

[0031] (4.4) Calculate the overall weight of each point;

[0032] (4.5) Calculate the weighted average of elevation differences;

[0033] (4.6) Calculate the elevation difference map.

[0034] Furthermore, step (5) includes

[0035] (5.1) Calculate the slope of the interferometric DSM data product and the point cloud DSM data product based on the interferometric DSM data product and the point cloud DSM data product to obtain the corresponding slope map;

[0036] (5.2) Calculate the slope difference map between the interferometric DSM data product and the point cloud DSM data product.

[0037] Furthermore, step (6) includes performing a comprehensive check on the final overlay vector map based on the elevation difference map and the slope difference map to determine whether it is an overlay area.

[0038] Furthermore, (1) when the average slope difference is less than 25° and the average elevation difference is greater than 150 meters, it is an overlapping area;

[0039] (2) When the average slope difference is greater than 25° and the average elevation difference is greater than 300 meters, it is an overlapping area.

[0040] Beneficial effects:

[0041] This invention identifies and verifies overlapping regions using both planar and elevation dimensions. It employs the overlapping region homogeneity coefficient in 3D SAR DOM data products to filter out non-overlapping regions. Furthermore, it proposes a comprehensive weighting method to calculate the elevation difference between interferometric DSM data products and point cloud DSM data products, effectively addressing the systematic differences between the two DSM data products and enabling more accurate determination of overlapping regions.

[0042] This invention uses elevation and slope differences to comprehensively examine overlapping areas, thereby more accurately identifying overlapping areas.

[0043] This invention uses 3D SAR to perform 3D inspection of overlapping regions, which is unique to this invention.

[0044] It should be understood that the above general description and the following specific embodiments are merely exemplary and illustrative, and do not limit the scope of the claims made in this application. Attached Figure Description

[0045] Figure 1 Here is a flowchart of a three-dimensional SAR overlay region inspection method provided by the present invention;

[0046] It should be understood that the accompanying drawings are not necessarily drawn to scale and present slightly simplified representations of various features illustrating the basic principles of this disclosure. Specific design features of the invention as disclosed herein, including, for example, particular dimensions, orientations, positions, and shapes, will be determined in part by the specifically intended application and usage environment.

[0047] In the figures, throughout the several figures, reference numerals refer to the same or equivalent parts of the invention. Detailed Implementation

[0048] Reference will now be made in detail to various embodiments of the invention, examples of which are illustrated in the accompanying drawings and described below. Although the invention will be described in conjunction with exemplary embodiments thereof, it should be understood that this specification is not intended to limit the invention to those exemplary embodiments. On the other hand, the invention is intended to cover not only the exemplary embodiments thereof, but also various alternatives, modifications, equivalents and other embodiments that may be included within the spirit and scope of the invention as defined by the appended claims.

[0049] Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. The specific structures and functions described in the exemplary embodiments of the present invention are for illustrative purposes only. Embodiments of the present invention can be implemented in various forms, and it should be understood that they should not be construed as limited to the exemplary embodiments described in the exemplary embodiments, but include all modifications, equivalents, or substitutions included within the spirit and scope of the present invention.

[0050] Throughout this specification, the technical terms used are for the purpose of describing various exemplary embodiments only and are not intended to be limiting. It will be further understood that the terms "comprising," "including," "having," etc., when used in the exemplary embodiments, specifically refer to the presence of the stated components, steps, operations, or elements, but do not exclude the presence or addition of one or more other components, steps, operations, or elements.

[0051] Due to the limitations of traditional SAR imaging principles, overlapping regions severely restrict the use of SAR data. However, 3D SAR can effectively solve the problem of overlapping regions, providing a more intuitive and accurate description of the geometry and scattering information of the 3D observation scene. Therefore, the verification of overlapping regions in 3D SAR data is particularly important.

[0052] This invention identifies and verifies overlapping regions from both planar and elevation dimensions. It utilizes the characteristic that the overlapping homogeneity coefficient of overlapping regions increases significantly in 3D SAR DOM data products to filter out non-overlapping regions. It proposes a comprehensive weighting method to calculate the elevation difference between interferometric DSM data products and point cloud DSM data products, which can effectively solve the systematic differences between the two DSM data products and make more accurate judgments on overlapping regions.

[0053] This invention proposes a method for detecting overlapping regions in three-dimensional SAR, the method comprising:

[0054] 1. Collect and process data from the shooting area;

[0055] Collect and organize raw SAR image data and point cloud data captured by 3D SAR, and process them into interferometric DSM data products, point cloud DSM data products, and DOM data products.

[0056] Collect and organize reference data for the shooting area, including digital elevation data (DEM) and land use cover maps (the maps should include information on cities, forests, farmland, water bodies, bare land, etc.).

[0057] The original image includes metadata information, specifically flight conditions such as sensor position and attitude information, which is used for subsequent identification of overlapping areas.

[0058] 2. Identify the overlapping regions of each image in the region data and merge the initial overlapping vector maps;

[0059] 2.1 For the raw SAR image data taken from each angle, a geometric model is constructed based on the collected flight conditions and the known digital elevation data (DEM) of the area, according to the coverage area, radar location, and terrain height.

[0060] 2.2 On each radar beam scanning distance line, the slant range of the radar to any point is calculated. The overlapping area is determined based on the change of slant range, thereby generating a vector map of the overlapping area corresponding to each original SAR image. Each record corresponds to one overlapping area.

[0061] 2.3 Each original SAR image corresponds to a vector map of the overlay region. These vector maps are merged to obtain the initial overlay vector map of the entire captured area. A lookup table of the original SAR image data corresponding to each overlay region needs to be recorded for subsequent analysis.

[0062] 3. Use the overlay homogeneity coefficient to compare and filter the final overlay vector image in the planar dimension;

[0063] 3.1 On the initial overlay vector image, for each overlay region, erode (shrink inward) by 3 pixels. If the diameter or width of the region is less than 3 pixels, no erosion operation is performed.

[0064] The generated eroded overlay vector image is used to eliminate the effects of edge recognition errors.

[0065] 3.2 Preset Layover Homogeneity Coefficient (LHC)

[0066]

[0067] in, It is the mean of the image. It is the standard deviation of the image.

[0068] 3.3 Select each record of the eroded overlay vector map in sequence, and obtain the corresponding original SAR image for the overlay area based on the comparison table between the overlay area and the original SAR image data; calculate the overlay homogeneity coefficient on the original SAR image corresponding to the overlay area. .

[0069] 3.4 The DOM data product is generated by interferometric analysis of multiple original SAR images. Compared to the original SAR images, some areas may have resolved the overlay phenomenon. Each record of the eroded overlay vector image is selected sequentially, and the overlay homogeneity coefficient corresponding to that area is calculated on the DOM data product. .

[0070] 3.5 Calculate the difference in the overlay homogeneity coefficient for each overlay region.

[0071]

[0072] 3.6 The homogeneity coefficient of the overlay is compared and screened based on the initial overlay vector diagram. The overlay area is a mixture of different scattering bodies (such as building tops and hillsides), and the standard deviation ( The superposition coefficient (LHC) will increase significantly, while the superposition homogeneity coefficient will decrease.

[0073] If the overlay problem is resolved in the DOM data product, the data should be more uniform, thus lowering the standard deviation and increasing the LHC. > .

[0074] According to the specific thresholds that 3D SAR should have in different applications The final overlay vector image is obtained by filtering the overlay regions. If a certain overlay region... If the LHC in the region does not rise effectively, then the region is considered not to be a masking region and needs to be removed from the masking vector map to obtain the final masking vector map.

[0075] 4. Calculate the elevation difference map by combining the processed interferometric DSM data product and point cloud DSM data product within the captured area data;

[0076] 4.1 On the land use cover map, select scene points of areal features as reference points based on interferometric DSM data products and point cloud DSM data products, such as water bodies, farmland, bare land, etc. The area is relatively flat, so select points with little elevation undulation in the surrounding area as reference points.

[0077] 4.2 Calculate the average and variance of the elevations of the points. Randomly select M points within the region. For each point with coordinates (PointX, PointY), use this point as the center and select a D×D area as the calculation region for that point. Assume that the number of points in this area is n. First, calculate the elevation value of each point. The average elevation of this region in the interferometric DSM data product is ,

[0078]

[0079] Elevation variance within the calculation area

[0080]

[0081] Similarly, calculate the average elevation on the point cloud DSM data product. and elevation variance .

[0082] 4.3 Calculate the elevation difference and comprehensive variance of the point location. Subtract the elevation values ​​of the two regions from the interferometric DSM data product and the point cloud DSM data product to obtain the elevation difference value corresponding to the point location.

[0083]

[0084] The overall variance at this point is

[0085]

[0086] 4.4 Calculate the overall weight of each point. The formula for calculating the overall weight of each point is as follows:

[0087]

[0088] 4.5 Calculate the weighted average of elevation differences. After determining the weight of each point, calculate the weighted average of the elevation differences of the points.

[0089]

[0090] Defined as a reference value for elevation difference. This is equivalent to the systematic error caused by imaging errors, shooting methods, etc., between the two DSMs. It serves as a reference value for subsequent use.

[0091] The core idea behind determining weights through variance is to assign higher weights to indicators with smaller variances (indicating more reliable data). Weights are inversely proportional to variance; the smaller the variance, the larger the weight.

[0092] 4.6 Calculate the elevation difference map and calculate the elevation difference value at each point.

[0093]

[0094] Based on the above formula, an elevation difference map was obtained on the entire interferometric DSM data product, in meters.

[0095] 5. Calculate the slope difference map by combining the processed interferometric DSM data product and point cloud DSM data product within the captured area data;

[0096] 5.1 Based on the interferometric DSM data product and the point cloud DSM data product, calculate the slope of each DSM data product. The slope is calculated using the finite difference method. Select a pixel within the DSM data product and choose a 3×3 window around it. Use the third-order inverse distance squared weighted difference algorithm to calculate the slope of the center pixel. The specific calculation formula is as follows:

[0097]

[0098] in, This indicates the rate of change of elevation of the window in the x-direction. This indicates the rate of change of elevation of the window in the y-direction.

[0099] Slope values ​​are not calculated for edge points and blank areas.

[0100] The slope of the interferometric DSM data product and the point cloud DSM data product were calculated using the method described above, resulting in slope maps corresponding to the two DSMs. and .

[0101] 5.2 Calculate the slope difference map between the interferometric DSM data product and the point cloud DSM data product;

[0102]

[0103] 6. The final overlapping area is obtained by comprehensively verifying the final overlay vector map, elevation difference map, and slope difference map.

[0104] On the final overlay vector map obtained above, a comprehensive check is performed based on the elevation difference map and the slope difference map to determine whether it is an overlay area; and the average elevation difference and the average slope difference are calculated for each overlay area.

[0105] The criteria for determining the overlapping area are as follows:

[0106] 1. When the average slope difference is less than 25° and the average elevation difference is greater than 150 meters, it is considered an overlapping area.

[0107] 2. When the average slope difference is greater than 25° and the average elevation difference is greater than 300 meters, it is considered an overlapping area.

[0108] In addition to the conditions mentioned above, the determination of whether it is an overlapping area should be made based on the actual situation.

[0109] In the above technical solution, the criteria for determining overlapping areas are based on the terrain classification in GB / T 13990-1992 "Specifications for Aerial Photogrammetry of 1:5000, 1:10000 Topographic Maps". Terrain categories are determined by the ground slope angle and elevation difference within the map sheet area. When the elevation difference and ground slope angle contradict each other, the ground slope angle takes precedence. A comprehensive verification of overlapping areas is then performed based on the above data. Specific terrain categories are shown in Table 1 below:

[0110] The unit is meters.

[0111] Terrain categories ground tilt angle Elevation difference flat land Below 2° <20 hilly areas 2°-6° 20-150 mountainous areas 6°-25° 150-300 High mountain Above 25° >300

[0112] Table 1

[0113] In summary, the present invention provides significant economic and social benefits for the identification and verification of overlapping regions from a three-dimensional perspective of plane and elevation, and is conducive to the widespread use of subsequent SAR data.

[0114] The foregoing description of specific exemplary embodiments of the invention has been presented for purposes of illustration and description. It is not intended to exclude or limit the invention to the precise forms disclosed, and it will be apparent that many modifications and alterations are possible in light of the foregoing teachings. Exemplary embodiments were chosen and described to explain certain principles of the invention and their practical application, so that others skilled in the art can make or utilize various exemplary embodiments of the invention, and their various alternatives and modifications. The purpose is that the scope of the invention will be defined by the appended claims and their equivalents.

[0115] It is understood that the above embodiments are merely exemplary implementations used to illustrate the principles of the present invention, and the present invention is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A method for detecting overlapping regions in three-dimensional SAR, characterized in that, The method includes: (1) Collect and process data from the shooting area; (2) Identify the overlapping regions of each image in the region data and merge the initial overlapping vector maps; (3) Use the overlay homogeneity coefficient to compare and filter the final overlay vector image in the planar dimension; Among them, the overlay homogeneity coefficient LHC: ； It is the mean of the image. It is the standard deviation of the image; Each record of the eroded overlay vector map is selected sequentially, and the corresponding original SAR image is obtained according to the comparison table between the overlay area and the original SAR image data. The overlay homogeneity coefficient on the original SAR image corresponding to the overlay area is then calculated. ; For each record in the eroded overlay vector image, calculate the overlay homogeneity coefficient for that region on the DOM data product. ; Calculate the difference in the hyperbolicity coefficient for each hyperbolic region. ; ； The final overlay vector diagram is obtained by comparing and filtering the overlay homogeneity coefficients based on the initial overlay vector diagram. If a certain overlapping area If the area is not an overlay area, it is considered to be removed from the overlay vector image and the final overlay vector image is obtained by filtering. (4) Calculate the elevation difference map by combining the processed interferometric DSM data product and point cloud DSM data product within the shooting area data; Based on the interferometric DSM data product, the point cloud DSM data product, and the land use cover map, select scene points with relatively small elevation fluctuations in the surrounding area as reference points; Calculate the average elevation and variance of the reference point on both the interferometric DSM data product and the point cloud DSM data product. Subtract the average elevation values ​​from the interferometric DSM data product and the point cloud DSM data product to obtain the elevation difference value corresponding to the reference point. Add the variances of the interferometric DSM data product and the point cloud DSM data product to obtain the comprehensive variance corresponding to the reference point. Calculate the comprehensive weight of each reference point using the following formula: ； After determining the weight of each point, the weighted average of the elevation differences between the points is calculated. The formula for calculating the weighted average of the elevation differences is as follows: ； An elevation difference map is obtained by calculating the elevation difference value at each point. The formula for calculating the elevation difference value at each point is as follows: ； (5) Calculate the slope difference map by combining the processed interference DSM data product and point cloud DSM data product within the shooting area data; (6) The final overlapping area is obtained by comprehensively verifying the final overlay vector map, elevation difference map and slope difference map; On the final overlay vector map, a comprehensive check is performed based on the elevation difference map and the slope difference map. If the average slope difference is less than 25° and the average elevation difference is greater than 150 meters, it is considered an overlay area; if the average slope difference is greater than 25° and the average elevation difference is greater than 300 meters, it is considered an overlay area.

2. The method for detecting three-dimensional SAR overlay regions according to claim 1, characterized in that, Step (1) includes collecting raw SAR image data, point cloud data, processed interferometric DSM data products, point cloud DSM data products and DOM data products, digital elevation data and land use cover map within the shooting area.

3. The method for detecting three-dimensional SAR overlay regions according to claim 1, characterized in that, Step (2) includes (2.1) Construct a geometric model based on the original SAR image data and digital elevation data; (2.2) Calculate the slant range of the radar to any point, determine the overlapping area based on the slant range, and generate a vector map of the overlapping area corresponding to each original SAR image; (2.3) Merge the overlay vector maps to obtain the initial overlay vector map of the entire shooting area.

4. The method for detecting three-dimensional SAR overlay regions according to claim 1, characterized in that, Step (5) includes (5.1) Calculate the slope of the interferometric DSM data product and the point cloud DSM data product based on the interferometric DSM data product and the point cloud DSM data product to obtain the corresponding slope map; (5.2) Calculate the slope difference map between the interferometric DSM data product and the point cloud DSM data product.

Images