A method for three-dimensional reconstruction of deep lunar crust fractures

Abstract

This invention belongs to the field of geophysical potential field separation technology, and particularly relates to a method for three-dimensional reconstruction of deep lunar crustal fractures, comprising: S1, acquiring input data including gravity anomaly data; S2, dividing the gravity anomaly data into multiple local analysis units and establishing a local coordinate system within each local analysis unit; S3, estimating parameters for each local analysis unit; S4, performing multi-scale potential field separation on the local gravity anomaly data of each local analysis unit and extracting fracture candidate boundaries; S5, constructing a pseudo-fracture penalty term and a linear fracture extraction term for the fracture candidate boundaries based on lunar tectonic features; S6, applying inter-layer geometric connectivity constraints to the extracted linear fracture candidate boundaries, merging the fracture candidate boundaries in different interpretation depth layers into a continuous three-dimensional surface of deep lunar crustal fractures. This scheme can significantly improve the spatial positioning accuracy of fractures, the stability of depth interpretation, and the continuity and reliability of three-dimensional geometric morphology.

Description

Technical Field

[0001] This invention belongs to the field of geophysical potential field separation technology, and particularly relates to a three-dimensional reconstruction method for deep fractures in the lunar crust. Background Technology

[0002] Lunar gravity field data is an important source of geophysical information for studying the structure, deep tectonic features and evolution of the lunar crust. With the acquisition of high-resolution gravity field models by lunar exploration missions such as GRAIL, it has become possible to use lunar gravity anomaly data to identify deep structures, explain intracranial inhomogeneities and reconstruct fracture geometry.

[0003] Existing related technologies mainly include the following categories: 1. Identifying tectonic boundaries based on gravity derivatives, gradients, or boundary enhancement attributes, such as the vertical second derivative method, the total horizontal gradient method, or boundary enhancement operators like ThetaMap, which highlight the location of tectonic boundaries by enhancing the response of anomalous gradient bands; 2. Highlighting anomalous features at specific spatial scales through potential field separation, multi-scale filtering, or band-limited frequency band constraints, such as interpolation cutting, wavelet multi-scale decomposition, or matched filtering, to separate field source effects at different scales; 3. Identifying impact basins, ring structures, or other regional geological units by combining information such as topography, gravity, and crust thickness, and revealing the regional tectonic pattern through joint analysis of multi-source data.

[0004] The above methods have certain application value in lunar tectonic research, but when directly used for three-dimensional identification of deep lunar crustal fractures, they still have the following shortcomings: 1. Lunar gravity field data is essentially spherical data. If it is processed uniformly using conventional planar methods over a large spatial area, it is easy to introduce latitude-related scale distortion and geometric distortion, which will affect the accuracy of boundary positioning and the consistency between interpretation results in different regions. This is mainly because the curvature of the lunar sphere causes systematic differences in the actual surface area corresponding to grid cells at different latitudes, resulting in latitude-dependent deviations in the magnitude and direction of gravity gradients calculated directly on the latitude and longitude grid.

[0005] 2. Existing methods often directly correspond a fixed scale, fixed layer cutting parameters, or fixed frequency band to a certain interpretation depth. However, different regions of the Moon have different local spectral characteristics, crustal thickness background, and noise stability. Therefore, the same processing scale may not correspond to the same tectonic layer in different regions, resulting in a lack of stability in the interpretation depth. This is because the correspondence between the cutting radius or frequency band window and the depth in potential field separation is controlled by the local field source spectral characteristics. The burial depth, scale, and density of the field source in different regions vary significantly, and the actual depth corresponding to the same window parameter can differ by several times in different regions.

[0006] 3. The lunar surface is widely characterized by impact basin boundaries, crater boundaries, and mascon anomalies. These structures often form significant ring-shaped boundaries or boundary-like responses in a gravitational field, making them prone to misidentification as linear fractures or fracture boundaries during fracture identification, resulting in strong false fracture interference. The gradient band amplitude of such ring-shaped structures is often comparable to that of linear fractures, exhibiting continuous high-amplitude responses in gradient enhancement or boundary identification attributes. Conventional threshold segmentation methods struggle to effectively distinguish between the two without prior constraints.

[0007] 4. For candidate boundary results obtained from different interpretation depth layers, existing technologies mostly adopt layer-by-layer interpretation or simple superposition methods, which are prone to problems such as inter-layer jumps, misconnections, and boundary fragmentation. It is difficult to stably extract continuous three-dimensional curved surfaces of deep lunar crust fractures. This is because the potential field separation results of different depth layers are independent and lack inter-layer geometric continuity constraints. Moreover, the identification results of each layer are affected by local noise and interference from nearby anomalies, causing the response positions of the same fracture at different depths to drift laterally and undergo morphological abrupt changes.

[0008] The aforementioned issues are intertwined in the three-dimensional identification of deep lunar fractures: spherical distortion affects the accuracy of regional positioning, unstable depth calibration disrupts the consistency of interpretation across different regions, pseudo-fractures in ring structures are highly confused with real fractures in terms of boundary properties, and interlayer fragmentation makes three-dimensional morphological reconstruction difficult. Existing technologies have shortcomings in addressing each of these issues individually, and a systematic solution to address all four issues simultaneously has not yet been reported.

[0009] Therefore, it is urgent to propose a three-dimensional reconstruction method for deep lunar crustal fractures to adapt to the characteristics of lunar spherical gravity field data, establish a more reasonable interpretation of depth calibration, suppress lunar-specific ring structure artifacts, and achieve stable merging of candidate boundaries of multiple depth layers into continuous three-dimensional fracture surfaces. Summary of the Invention

[0010] In view of this, the present invention aims to provide a method for three-dimensional reconstruction of deep lunar crust fractures to solve the problems of spherical distortion of lunar gravity data, unstable depth calibration, interference from pseudo-fractures in ring structures, and discontinuity of interlayer fractures, so as to achieve high-fidelity three-dimensional reconstruction of deep lunar crust fractures.

[0011] To achieve the above objectives, the technical solution created by this invention is implemented as follows: A method for three-dimensional reconstruction of deep lunar crustal fractures includes the following steps: S1. Obtain input data, including gravity anomaly data; S2. Divide the gravity anomaly data into multiple local analysis units and establish a local coordinate system within each local analysis unit; S3. Perform parameter estimation for each local analysis unit, including at least calculating local spectral characteristic parameters, lunar crust thickness prior parameters, and extensional stability index; and perform regional adaptive calibration based on the correspondence between processing scale parameters and interpretation depth based on local spectral characteristic parameters, lunar crust thickness prior parameters, and extensional stability index. S4. Based on the processing scale parameters after regional adaptive calibration, multi-scale potential field separation is performed on the local gravity anomaly data of each local analysis unit to obtain local gravity anomaly field data corresponding to different interpretation depth layers, and fracture candidate boundaries are extracted based on the local gravity anomaly field data. S5. For fracture candidate boundaries, construct pseudo-fracture penalty terms and linear fracture extraction terms based on lunar tectonic features to suppress fracture candidate boundaries and extract linear fracture candidate boundaries. S6. Apply interlayer geometric connectivity constraints to the extracted linear fracture candidate boundaries, and merge the fracture candidate boundaries in different interpretation depth layers into a continuous three-dimensional surface of deep lunar crust fractures.

[0012] Furthermore, the local analysis units in S2 are constructed using any one of the following methods: fixed latitude and longitude window division, equal area block division, or sliding window division with overlapping areas; overlapping areas are set between adjacent local analysis units; the method for establishing a local coordinate system in S2 is as follows: using the center point of the local analysis unit as the reference point, a local tangent plane coordinate system or a local equal area coordinate system is established, and the gravity anomaly data represented by spherical coordinates is converted into gravity anomaly data represented by local planar coordinates.

[0013] Furthermore, the extensional stability index in S3 is constructed based on the differences between local anomaly field results at different processing scales. These differences include at least one of the following: mean square error between results at adjacent processing scales, gradient volatility, noise amplification, and boundary position drift.

[0014] Furthermore, in S4, for the overlapping areas between adjacent local analysis units, the fracture candidate boundaries are stitched together using a weighted fusion method. The fracture candidate boundaries are extracted based on at least one of the following attributes: vertical second derivative zero-crossing attribute, horizontal gradient attribute, structural coherence attribute, and continuity attribute between adjacent interpretation depth layers. The above attributes are used to extract fracture candidate boundaries, rather than being directly used as the final fracture surface determination result.

[0015] Furthermore, the pseudo-fracture penalty term in S5 is constructed jointly based on at least two features from the axis ratio constraint term, ellipse fit, angle coverage, closure, cumulative rotation angle, and concentric loop response.

[0016] Furthermore, the pseudo-fracture penalty term in S5 is expressed by the following formula: ; in, Indicates the shaft ratio constraint term; Indicates the ellipse fit degree; Indicates angular coverage; Indicates the degree of closure; Indicates the cumulative turning angle; Indicates a concentric ring response; , , , This is the weight value.

[0017] Furthermore, the candidate boundaries for fracture-like structures in S5 are selected using a comprehensive score, which is expressed by the following formula: ; in, This represents the overall fracture candidate score; Indicates linear score; This indicates a circular penalty term.

[0018] Furthermore, the interlayer geometric connectivity constraints in S6 include at least two of the following: spatial displacement constraints between adjacent interpretation depth layers, orientation change constraints between adjacent interpretation depth layers, dip angle change constraints between adjacent interpretation depth layers, and minimum connected domain area constraints.

[0019] When candidate boundaries of adjacent interpretation depth layers satisfy the following formula, they are merged into adjacent layer segments of the same continuous fracture surface: ; in, Spatial offset between adjacent candidate boundaries of depth layers; This refers to the change in orientation between adjacent candidate boundaries of depth layers. The change in dip angle between adjacent candidate boundaries of depth layers is used to explain the amount of dip angle variation. This is a preset threshold.

[0020] Furthermore, in S3, the region-adaptive calibration of the correspondence between the processing scale parameter and the interpretation depth must satisfy the following formula: ; in, For the first One explanation depth parameter; For the first One processing scale parameter; It is the central wavelength or dominant wavelength of the local spectrum; The prior parameter for the lunar crust thickness corresponding to the local analysis unit; To extend the stability index; These are latitude-related correction parameters.

[0021] Compared with the prior art, the present invention can achieve the following beneficial effects: 1. Improve the regional adaptability of fracture identification under spherical gravity field conditions: This scheme divides the lunar spherical gravity field into multiple local analysis units and establishes a local coordinate system in each local analysis unit for regional processing. This avoids the scale distortion problem caused by large-scale uniform planar processing and helps to improve the consistency and comparability of gravity boundary identification results in different regions.

[0022] 2. Improve the stability of the correspondence between processing scale and interpretation depth: This scheme combines local spectral features, lunar crust thickness priors, and continuation stability to perform regional adaptive calibration of the relationship between processing scale and interpretation depth. It no longer assumes that the same scale in different regions corresponds to the same depth, thereby improving the stability and rationality of interpretation depth layers in different regions.

[0023] 3. Effectively reduce false fracture misjudgments caused by lunar-specific ring structures: This scheme explicitly considers the ring gravity response caused by impact basin ring boundaries, crater boundaries, and mascon anomalies as the source of artifacts in fracture identification. Through a comprehensive scoring mechanism, it effectively reduces false fracture identification in the context of lunar-specific structures and improves the reliability of fracture interpretation.

[0024] 4. Improve the continuity and stability of the three-dimensional fracture surface in deep fractures: This scheme applies spatial displacement, orientation change, dip angle change and connectivity constraints to the candidate boundaries of different interpretation depth layers, so that the multi-layer candidate boundaries can be stably merged into a continuous three-dimensional fracture surface, avoiding the inter-layer jump, misconnection and boundary fragmentation problems common in traditional layer-by-layer interpretation.

[0025] 5. More suitable for interpretation of the lunar deep structure and three-dimensional geometric analysis: This scheme can not only improve the accuracy and continuity of fracture boundary identification, but also further output the spatial geometric parameters of the fracture surface, providing a more stable and targeted technical means for interpretation of the lunar deep structure, analysis of the tectonic pattern and related planetary geophysical research. Attached Figure Description

[0026] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings: Figure 1 This is a schematic diagram of the overall process of the three-dimensional reconstruction method in an embodiment of the present invention; Figure 2 This is a schematic diagram of the division of local analysis units in an embodiment of the present invention; Figure 3This is a schematic diagram of region adaptive calibration in an embodiment of the present invention. Detailed Implementation

[0027] To make the purpose, technical solution, and advantages of this invention clearer, the following description is provided in conjunction with the appendix. Figure 1-3 The present invention will be further described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are only for explaining the present invention and do not constitute a limitation thereof.

[0028] A method for three-dimensional reconstruction of deep lunar crustal fractures includes the following steps: S1. Acquire input data, including gravity anomaly data, and preferably combine it with auxiliary information such as lunar topography data, lunar crust thickness model, impact basin boundary catalog or crater catalog. The gravity anomaly data is lunar spherical Bouguer gravity anomaly data, which is Bouguer gravity anomaly data generated based on the GRAIL lunar gravity field model. The lunar topography data is preferably LOLA topography model data. Unify all types of data under the same spherical reference frame and grid resolution, and set a buffer around the target area to reduce the impact of boundary truncation on subsequent processing results.

[0029] S2. Divide the gravity anomaly data into multiple local analysis units, and establish a local coordinate system in each local analysis unit to obtain the corresponding local gravity anomaly data.

[0030] The local analysis unit is constructed using any one of the following methods: fixed latitude and longitude window division, equal area block division, or sliding window division with overlapping regions. An overlapping region is set between adjacent local analysis units to perform weighted fusion of the reconstruction results in the overlapping region. In this embodiment, the sliding window construction method with overlapping regions is preferred.

[0031] The method for establishing a local coordinate system is as follows: taking the center point of the local analysis unit as the reference point, establish a local tangent plane coordinate system or a local equal area coordinate system, and convert the gravity anomaly data represented by spherical coordinates into gravity anomaly data represented by local planar coordinates.

[0032] S3. Perform parameter estimation for each local analysis unit, at least calculate the local spectral characteristic parameters, the prior parameters of lunar crust thickness, and the extension stability index. Based on the local spectral characteristic parameters, the prior parameters of lunar crust thickness, and the extension stability index, perform regional adaptive calibration on the correspondence between the processing scale parameters and the interpretation depth to obtain the interpretation depth layer corresponding to each local analysis unit. This step does not simply use a fixed frequency band or a fixed scale to correspond to a uniform depth, but rather performs regional correction of the interpretation depth according to the characteristics of the local analysis unit.

[0033] Local spectral characteristic parameters can be obtained through local spectral analysis, preferably the local spectral center wavelength or dominant wavelength, denoted as . The prior parameter for the lunar crust thickness corresponding to the local analysis unit is denoted as... It can be calculated from the lunar crust thickness model data of the corresponding region; the extensional stability index is marked as It is constructed based on the differences between local anomaly field results at different processing scales. The differences include at least one of the following: mean square error between results at adjacent processing scales, gradient volatility, noise amplification rate, and boundary position drift; at the same time, latitude-related correction parameters are introduced. .

[0034] get , , , After obtaining the parameters, the correspondence between the processing scale parameters and the interpretation depth is adaptively calibrated in a region, which must satisfy the following formula: ;in, For the first One explanation depth parameter; For the first Each processing scale parameter.

[0035] This calibration relationship enables a region-adaptive correspondence between the processing scale and interpretation depth in different local analysis units, rather than uniformly adopting a fixed scale corresponding to a fixed depth.

[0036] S4. Based on the processing scale parameters after regional adaptive calibration, multi-scale potential field separation is performed on the local gravity anomaly data of each local analysis unit to obtain local gravity anomaly field data corresponding to different interpretation depth layers, and fracture candidate boundaries are extracted based on the local gravity anomaly field data.

[0037] Multi-scale potential field separation is achieved using one or more combinations of interpolation-cutting potential field separation methods, local filtering separation methods, and multi-scale nonlinear filtering methods. The gravity anomaly data in the local analysis unit is represented as follows: ; in: This refers to gravity anomaly data within a local analysis unit. For the first Each interpretation depth corresponds to a regional field; For the first Each interpretation depth corresponds to a local anomaly field.

[0038] For the overlapping areas between adjacent local analysis units, the fracture candidate boundary is stitched together using a weighted fusion method. The fracture candidate boundary is extracted based on at least one of the following attributes: vertical second derivative zero-crossing attribute, horizontal gradient attribute, structural coherence attribute, and continuity attribute between adjacent interpretation depth layers. The above attributes are used to extract fracture candidate boundaries, but are not directly used as the final fracture surface determination result.

[0039] S5. For fracture candidate boundaries, construct pseudo-fracture penalty terms and linear fracture extraction terms based on lunar tectonic features to suppress fracture candidate boundaries and extract linear fracture candidate boundaries. The pseudo-fracture penalty terms are jointly constructed based on at least two features from axis ratio constraint terms, ellipse fitting degree, angle coverage, closure degree, cumulative rotation angle, and concentric ring response.

[0040] The pseudo-fracture penalty term is expressed by the following formula: ; in, This indicates the axis ratio constraint term, used to avoid misjudging extremely narrow linear boundaries as loop structures; It represents the ellipse fit, indicating the degree of similarity between the zero line and the boundary of a circle or ellipse; It indicates the angular coverage, used to determine whether the zero-value line covers a relatively complete annular area; This indicates the degree of closure, used to measure whether the zero-value line is closed at both ends; It represents the cumulative turning angle, used to determine whether it has continuous turning characteristics with a ring-shaped boundary; This indicates a concentric ring response, used to describe whether there are multiple ring-shaped zero-value lines with similar centers in the region; , , , The weight values ​​are for the interval [0,1]. If values ​​can be obtained It can also be adjusted according to the specific circumstances.

[0041] Candidate linear fracture boundaries are extracted using a comprehensive score. The expression for the linear score is as follows: ; in, Indicates the first The linearity of the edge reflects whether the candidate boundary as a whole is a thin, elongated line; Indicates directional stability and is used to measure whether the boundary orientation is concentrated; The three factors represent the continuity of the boundary and are used to evaluate whether there are obvious breaks, jumps, or discontinuities in the candidate boundary; the linear break feature score of the candidate boundary is obtained by weighting the three factors. , , These are weighting coefficients. ,and: .

[0042] Candidate boundaries for fracture types are selected using a comprehensive score, which is expressed by the following formula: ; in, This represents the overall fracture candidate score; Indicates linear score; This indicates a circular penalty term.

[0043] In this formula, when The higher the value, the more linearly broken the candidate boundary becomes; while when... When the value is larger, the boundary also has obvious ring-like features, which weakens the final fracture candidate score. This can effectively reduce the interference of impact craters, basins and multi-ring structures on the fracture identification results while retaining the linear fracture boundary. When the ring-like interference penalty term is greater than a certain threshold, such boundaries are judged as ring-like interference and removed from the fracture candidate results. When the comprehensive fracture candidate score exceeds the fracture discrimination threshold, it indicates that the boundary has obvious linear fracture features and the ring-like interference is weak, so it is retained as a linear fracture candidate boundary.

[0044] First, determine the cyclic penalty term for each candidate boundary. Does it exceed the set threshold? :

[0045] If this condition is met, it indicates that the candidate boundary has strong ring-like characteristics and may be related to impact craters, basin boundaries, or multi-ring structures. Therefore, such boundaries are identified as ring-like interferences and removed from the candidate fracture results.

[0046] Right now: Eliminate: Impact crater / basin annular zero-value line.

[0047] If this condition is not met, that is: If the result is positive, it indicates that the candidate boundary does not have obvious ring-shaped interference characteristics and can proceed to the next step of fracture candidate discrimination.

[0048] Then, determine its comprehensive fracture candidate score. Does it exceed the fracture detection threshold? :

[0049] If this condition is met, it indicates that the boundary has obvious linear fracture characteristics and the ring interference is weak, so it is retained as a candidate boundary for linear fracture.

[0050] Right now: Retain: Linear fracture candidate boundaries.

[0051] If this condition is not met, that is: This indicates that although the boundary is not a typical ring-shaped disturbance, there is insufficient evidence of linear breakage. It may belong to a noisy boundary, a weak anomaly boundary, a discontinuous boundary, or a candidate boundary that needs further verification by combining imagery, topographic, and geological data. Therefore, it is classified as: candidate boundary to be verified / low confidence.

[0052] S6. Apply interlayer geometric connectivity constraints to the extracted linear fracture candidate boundaries, and merge the fracture candidate boundaries in different interpretation depth layers into a continuous three-dimensional surface of deep lunar crust fractures. The interlayer geometric connectivity constraints include at least two of the following: spatial displacement constraints between adjacent interpretation depth layers, strike variation constraints between adjacent interpretation depth layers, dip variation constraints between adjacent interpretation depth layers, and minimum connected domain area constraints. Through these constraints, the candidate boundaries in different interpretation depth layers are merged into a continuous three-dimensional surface of deep lunar crust fractures, and the surface can be further partially fitted to obtain orientation parameters such as the strike, dip, and dip angle of the fractures.

[0053] When candidate boundaries of adjacent interpretation depth layers satisfy the following formula, they are merged into adjacent layer segments of the same continuous fracture surface: ; in, Spatial offset between adjacent candidate boundaries of depth layers; This refers to the change in orientation between adjacent candidate boundaries of depth layers. The change in dip angle between adjacent candidate boundaries of depth layers is used to explain the amount of dip angle variation. This is a preset threshold.

[0054] For overlapping areas between adjacent local analysis units, a weighted fusion method is used to stitch the merged results together to reduce the block edge effect and improve the spatial continuity of the fracture surface. After the above steps, a continuous three-dimensional fracture surface of the deep lunar crust in the target study area can be obtained.

[0055] In this embodiment, after obtaining the continuous fracture surface, the fracture surface can be further fitted locally to obtain attitude parameters such as the fracture strike, dip direction, and dip angle. The locally fitted plane is represented as follows: ; Its normal vector n is: ; Local tilt angle Represented as: ; in, A, B, C, D These are the coefficients of the fitted plane. , , These are the spatial coordinates of the grid nodes.

[0056] Using the above method, geometric attitude information of different parts of the fracture surface can be obtained. After processing the target study area using the method in this embodiment, a regional adaptive correspondence of interpretation depth can be formed at the local analysis unit scale, and the interference of pseudo-fractures caused by impact basin ring boundaries, crater boundaries and mascon anomalies can be effectively reduced. Compared with the layer-by-layer interpretation based solely on the zero cross of the vertical second derivative or the results of fixed-scale layer shearing, this embodiment can more stably preserve the linear fracture boundary and merge candidate boundaries of multiple depth layers to form a continuous three-dimensional fracture surface.

[0057] The following is an appendix Figure 1-3 This embodiment will be further described as follows: Figure 1 This is a schematic diagram of the overall process of the three-dimensional reconstruction method in an embodiment of the present invention.

[0058] Figure 2 This is a schematic diagram of the division of local analysis units in an embodiment of the present invention. Figure 2 The diagram shows three local analysis units. First, the study area or working area is determined in the spherical Bouguer gravity anomaly data. Then, multiple local analysis units are constructed within the study area. The local analysis units can be constructed by methods such as fixed latitude and longitude window division, equal area block division, or sliding window division with overlapping areas. In this embodiment, the sliding window method with overlapping areas is preferred.

[0059] An overlapping region of a certain width is set between adjacent local analysis units to weightedly fuse the reconstruction results obtained by different local units within the overlapping region, thereby reducing the discontinuity at the boundaries of the local window. For any local analysis unit, a local coordinate system is established with its center point as the reference point. For example, units A, B, and C are respectively set with... , and Establish a local coordinate system for the origin. , and Subsequently, the gravity anomaly data represented in the original spherical coordinates were... Gravity anomaly data converted to local planar coordinates This allows us to obtain the local gravity anomaly data corresponding to each local analysis unit. .

[0060] Figure 3This is a schematic diagram of regional adaptive calibration in an embodiment of the present invention. Taking two local analysis units as an example, as shown in the figure, the left side is local analysis unit A, and the right side is local analysis unit B. Each local analysis unit determines a set of local parameters based on the gravity anomaly data characteristics within that unit. These parameters collectively determine the processing scale parameters. Explanation of depth parameters The conversion relationship between them, where, For the first A processing scale parameter can be understood as a certain processing scale, analysis radius, filtering scale, or extension scale. For the first An interpretation depth parameter represents the interpretation depth corresponding to this scale.

[0061] The formula is as follows: ;in, It is the central wavelength or dominant wavelength of the local spectrum; The prior parameter for the lunar crust thickness corresponding to the local analysis unit; To extend the stability index; These are latitude-related correction parameters.

[0062] Figure 3 In the local analysis unit A, the parameter values ​​are: ; Therefore, within unit A, The corresponding interpretation depth parameter is: ; The diagram illustrates: ; This indicates that within local analysis unit A, different processing scales are converted into the corresponding interpretation depths within that unit.

[0063] In local analysis unit B, since the local gravity anomaly characteristics, background field, tectonic conditions, or spectral characteristics may differ, the parameter values ​​become:

[0064] Therefore, within local analysis unit B, The corresponding interpretation depth parameter is:

[0065] The diagram illustrates:

[0066] at the same time, This indicates that even if the same processing scale is used in two local analysis units. Because the local parameters are different, the final interpretation depth is also different.

[0067] In other words:

[0068]

[0069] because:

[0070] so:

[0071] This indicates that the scale-depth transformation relationship varies with the local analytical unit, rather than remaining fixed throughout the entire study area.

[0072] Therefore, even when the same processing scale parameter is used in different local analysis units, Because the feature parameters of each local unit are different, their corresponding interpretation depth parameters are also different. It may also be different; that is, the relationship between scale and depth is not a globally uniform one-to-one correspondence, but rather has local adaptability.

[0073] This embodiment also provides a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the steps of the above method. The storage medium includes, but is not limited to, USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, optical disks, etc.

[0074] This embodiment also provides an electronic device, including a memory and a processor. The memory stores a computer program, and the processor implements the steps of the above method when it calls the computer program.

[0075] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this invention disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this invention can be achieved, and this is not limited herein.

[0076] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for three-dimensional reconstruction of deep lunar crustal fractures, characterized in that, Includes the following steps: S1. Obtain input data, including gravity anomaly data; S2. Divide the gravity anomaly data into multiple local analysis units and establish a local coordinate system within each local analysis unit; S3. Perform parameter estimation for each local analysis unit, including at least calculating local spectral characteristic parameters, lunar crust thickness prior parameters, and extensional stability index; and perform regional adaptive calibration based on the correspondence between processing scale parameters and interpretation depth based on local spectral characteristic parameters, lunar crust thickness prior parameters, and extensional stability index. S4. Based on the processing scale parameters after regional adaptive calibration, multi-scale potential field separation is performed on the local gravity anomaly data of each local analysis unit to obtain local gravity anomaly field data corresponding to different interpretation depth layers, and fracture candidate boundaries are extracted based on the local gravity anomaly field data. S5. For fracture candidate boundaries, construct pseudo-fracture penalty terms and linear fracture extraction terms based on lunar tectonic features to suppress fracture candidate boundaries and extract linear fracture candidate boundaries. S6. Apply interlayer geometric connectivity constraints to the extracted linear fracture candidate boundaries, and merge the fracture candidate boundaries in different interpretation depth layers into a continuous three-dimensional surface of deep lunar crust fractures.

2. The method for three-dimensional reconstruction of deep lunar crustal fractures according to claim 1, characterized in that, The local analysis units in S2 are constructed using any one of the following methods: fixed latitude and longitude window division, equal area block division, or sliding window division with overlapping areas. Overlapping areas are set between adjacent local analysis units. The method for establishing a local coordinate system in S2 is as follows: using the center point of the local analysis unit as the reference point, a local tangent plane coordinate system or a local equal area coordinate system is established, and the gravity anomaly data represented by spherical coordinates is converted into gravity anomaly data represented by local planar coordinates.

3. The method for three-dimensional reconstruction of deep lunar crustal fractures according to claim 1, characterized in that, The extensional stability index in S3 is constructed based on the differences between local anomaly field results at different processing scales. The differences include at least one of the following: mean square error between results at adjacent processing scales, gradient volatility, noise amplification, and boundary position drift.

4. The method for three-dimensional reconstruction of deep lunar crustal fractures according to claim 1, characterized in that, For the overlapping areas between adjacent local analysis units in S4, the fracture candidate boundaries are stitched together using a weighted fusion method. The fracture candidate boundaries are extracted based on at least one of the following attributes: vertical second derivative zero-crossing attribute, horizontal gradient attribute, structural coherence attribute, and continuity attribute between adjacent interpretation depth layers. The above attributes are used to extract fracture candidate boundaries, but are not directly used as the final fracture surface determination result.

5. The method for three-dimensional reconstruction of deep lunar crustal fractures according to claim 1, characterized in that, The pseudo-fracture penalty term in S5 is constructed jointly based on at least two features from the axis ratio constraint term, ellipse fit, angle coverage, closure, cumulative rotation angle, and concentric loop response.

6. The method for three-dimensional reconstruction of deep lunar crustal fractures according to claim 5, characterized in that, The pseudo-fracture penalty term in S5 is expressed by the following formula: ； in, Indicates the shaft ratio constraint term; Indicates the ellipse fit degree; Indicates angular coverage; Indicates the degree of closure; Indicates the cumulative turning angle; Indicates a concentric ring response; , , , This represents the weight value.

7. The method for three-dimensional reconstruction of deep lunar crustal fractures according to claim 1, characterized in that, The candidate boundaries for fracture-like structures in S5 are selected using a comprehensive score, which is expressed by the following formula: ； in, This represents the overall fracture candidate score; Indicates linear score; This indicates a circular penalty term.

8. The method for three-dimensional reconstruction of deep lunar crustal fractures according to claim 1, characterized in that, The interlayer geometric connectivity constraints in S6 include at least two of the following: spatial displacement constraints between adjacent interpretation depth layers, orientation change constraints between adjacent interpretation depth layers, dip angle change constraints between adjacent interpretation depth layers, and minimum connected domain area constraints.

9. The method for three-dimensional reconstruction of deep lunar crustal fractures according to claim 8, characterized in that, When candidate boundaries of adjacent interpretation depth layers satisfy the following formula, they are merged into adjacent layer segments of the same continuous fracture surface: ； in, Spatial offset between adjacent candidate boundaries of depth layers; This refers to the change in orientation between adjacent candidate boundaries of depth layers. The change in dip angle between adjacent candidate boundaries of depth layers is used to explain the amount of dip angle variation. This is a preset threshold.

10. The method for three-dimensional reconstruction of deep lunar crustal fractures according to claim 1, characterized in that, In S3, the region-adaptive calibration of the correspondence between processing scale parameters and interpretation depth must satisfy the following formula: ； in, For the first One explanation depth parameter; For the first One processing scale parameter; It is the central wavelength or dominant wavelength of the local spectrum; The prior parameter for the lunar crust thickness corresponding to the local analysis unit; To extend the stability index; These are latitude-related correction parameters.

Images