CN116168176A - Building geometry and deformation extraction method combining InSAR and laser point cloud - Google Patents
Building geometry and deformation extraction method combining InSAR and laser point cloud Download PDFInfo
- Publication number
- CN116168176A CN116168176A CN202310437935.5A CN202310437935A CN116168176A CN 116168176 A CN116168176 A CN 116168176A CN 202310437935 A CN202310437935 A CN 202310437935A CN 116168176 A CN116168176 A CN 116168176A
- Authority
- CN
- China
- Prior art keywords
- building
- deformation
- point cloud
- model
- dimensional
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000605 extraction Methods 0.000 title claims abstract description 42
- 238000012545 processing Methods 0.000 claims abstract description 52
- 238000004458 analytical method Methods 0.000 claims abstract description 44
- 230000003190 augmentative effect Effects 0.000 claims abstract description 14
- 238000000034 method Methods 0.000 claims description 58
- 238000012795 verification Methods 0.000 claims description 12
- 230000001427 coherent effect Effects 0.000 claims description 11
- 230000005484 gravity Effects 0.000 claims description 8
- 238000004422 calculation algorithm Methods 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 5
- 238000000926 separation method Methods 0.000 claims description 5
- 238000002372 labelling Methods 0.000 claims description 4
- 238000005457 optimization Methods 0.000 claims description 4
- 230000000704 physical effect Effects 0.000 claims description 4
- 239000008186 active pharmaceutical agent Substances 0.000 claims description 3
- 239000004566 building material Substances 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 claims description 3
- 238000004590 computer program Methods 0.000 claims description 2
- 238000003672 processing method Methods 0.000 claims description 2
- 230000004044 response Effects 0.000 abstract description 3
- 230000008569 process Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 7
- 238000012544 monitoring process Methods 0.000 description 7
- 238000011161 development Methods 0.000 description 6
- 230000018109 developmental process Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 6
- 239000004973 liquid crystal related substance Substances 0.000 description 6
- 230000008859 change Effects 0.000 description 5
- 238000005305 interferometry Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 3
- 230000019771 cognition Effects 0.000 description 2
- 239000006185 dispersion Substances 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000011084 recovery Methods 0.000 description 2
- 238000000611 regression analysis Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000001149 cognitive effect Effects 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000000875 corresponding effect Effects 0.000 description 1
- 238000002790 cross-validation Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 230000000873 masking effect Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000002040 relaxant effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/10—Constructive solid geometry [CSG] using solid primitives, e.g. cylinders, cubes
-
- G06T5/70—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
- G06T2207/10044—Radar image
Abstract
The invention discloses a building geometry and deformation extraction method combining InSAR and laser point cloud, which comprises the following steps: performing constraint denoising processing on the SAR image based on the three-dimensional model to obtain a high-precision three-dimensional point cloud; processing a plurality of subspaces obtained by dividing the high-precision three-dimensional points according to the super voxels to obtain building primitives; reducing building primitives to obtain regularized plane primitives, generating an augmented candidate primitive pool, constructing an energy equation, selecting the candidate primitive pool to obtain a subset, and combining the subset to obtain a target building three-dimensional model; and carrying out model analysis on the three-dimensional model of the target building, constructing a temperature deformation model of the building, and verifying based on the temperature deformation model to obtain the temperature deformation and the trend deformation of the building. The invention can rapidly provide high-precision position and deformation information of the building in real time, and plays an important role in safe operation and emergency response of urban buildings.
Description
Technical Field
The invention relates to the technical field of multisource spatial information perception, in particular to a building geometry and deformation extraction method, system, terminal and computer readable storage medium combining InSAR and laser point cloud.
Background
With the rapid development of social economy, the number and scale of urban buildings are gradually increased, and the achievement of great attention is achieved. However, because of the complex geographical environment and dense population, various natural and human factors overlap, which can lead to deformation and damage of urban buildings; therefore, if the fine three-dimensional model and deformation parameters of the building can be quickly extracted, key basic information can be provided for urban safety operation and emergency response, and personal safety and property loss possibly caused by similar events can be effectively reduced.
The time sequence InSAR (Synthetic Aperture Radar Interferometry, radar interferometry) is a high-precision, large-range and low-cost technology for monitoring the deformation of the earth surface, and along with the continuous development of SAR satellites and the continuous development of InSAR technology, the time sequence InSAR (Synthetic Aperture Radar Interferometry, radar interferometry) is one of the most potential technologies for efficiently developing the deformation monitoring of urban buildings, and a certain research result is achieved in the aspect of urban building monitoring. However, in the application of urban complex scenes in high-rise forestation, due to the serious problems of shadow, overlay, decorrelation noise and the like of SAR satellite images, the problems of inaccurate positioning of point targets, aliasing of building temperature deformation and trend deformation and the like still exist in the inversion of the high-precision deformation of the building. With the rapid development of Lidar (Light Detection and Ranging, laser radar), unmanned aerial vehicle photogrammetry and other technologies, the technology for acquiring large-scale high-precision three-dimensional point clouds of urban buildings is mature, and aviation and ground photogrammetry become one of the main means of urban three-dimensional modeling. However, in a complex urban scene, because the density of the space-ground integrated dense matching point cloud is changed severely and the features of the same-name features are different, difficulties such as high-fidelity extraction of geometric primitives of a building, reasoning and recovery of topological relations and the like exist in the three-dimensional modeling of the building.
Therefore, the problem that the high-precision three-dimensional model of the building and the deformation inversion precision of the InSAR building cannot be obtained exists in the prior art, so that reliable three-dimensional position and deformation information of the building cannot be provided for urban safe operation.
Accordingly, the prior art is still in need of improvement and development.
Disclosure of Invention
The invention mainly aims to provide a building geometric and deformation extraction method combining InSAR and laser point cloud, and aims to solve the problem that reliable three-dimensional position and deformation information of a building cannot be provided for urban safe operation in the prior art.
In order to achieve the above object, the present invention provides a method for extracting geometry and deformation of a building combining InSAR and laser point cloud, the method for extracting geometry and deformation of a building combining InSAR and laser point cloud comprising the steps of:
acquiring a time sequence SAR image of a building acquired by an interference synthetic aperture radar and a three-dimensional model of the building acquired by a laser radar, and performing fine denoising processing on a building point target of multiple feature constraints on the SAR image based on the three-dimensional model to obtain a high-precision three-dimensional point cloud;
generating multi-resolution super-voxels according to the multi-scale seed points, dividing the high-precision three-dimensional point cloud based on the super-voxels to obtain a plurality of subspaces, and processing each subspace to obtain building primitives;
Reducing the building primitive to obtain a regularized plane primitive, generating an augmented candidate primitive pool based on the regularized plane primitive, constructing an energy equation, selecting the candidate primitive pool based on the energy equation to obtain a subset, and combining the subset to obtain a target building three-dimensional model;
and carrying out model analysis on the three-dimensional model of the target building to obtain an analysis result, constructing a temperature deformation model of the building based on the analysis result, and carrying out verification processing based on the temperature deformation model to obtain the temperature deformation and trend deformation of the building.
Optionally, the method for extracting geometry and deformation of a building by combining InSAR and laser point cloud, wherein the method for obtaining a time sequence SAR image of the building collected by an interferometric synthetic aperture radar and a three-dimensional model of the building collected by the laser radar, performing fine denoising processing on a building point target of multiple feature constraint on the SAR image based on the three-dimensional model, and obtaining a high-precision three-dimensional point cloud specifically comprises:
acquiring a PSI differential interference atlas and an SBAS differential interference atlas of a time sequence synthetic aperture radar image acquired by an interference synthetic aperture radar and semantic information and geometric information of a building acquired by a laser radar;
Performing PS scatterer extraction on the PSI differential interference atlas to obtain a high-coherence structure, performing DS scatterer extraction on the SBAS differential interference atlas to obtain a low-coherence structure, and constructing a three-dimensional scattering scene model of a building foreground-background based on the semantic information;
constructing a scattered signal interpretation model combining coherent information and incoherent information of the synthetic aperture radar, performing scatterer selection recognition on the high-coherence structure and the low-coherence structure based on the scattered signal interpretation model to obtain a recognition result, and performing point target space position matching on the recognition result and the geometric information to obtain a matching result;
and carrying out building point target segmentation on the matching result to obtain a scattering signal and a background signal of the building, and inputting the scattering signal and the background signal into the three-dimensional scattering scene model to obtain the high-precision three-dimensional point cloud of the building.
Optionally, the building geometry and deformation extraction method combining the InSAR and the laser point cloud, wherein the generating a multi-resolution super voxel according to the multi-scale seed point, dividing the high-precision three-dimensional point cloud based on the super voxel to obtain a plurality of subspaces, and processing each subspace to obtain a building primitive specifically includes:
Acquiring an original space-ground integrated point cloud of the building, searching the original space-ground integrated point cloud to obtain a plurality of target adjacent points, and calculating the average distance between the original space-ground integrated point cloud and the plurality of target adjacent points;
dividing the original space-ground integrated point cloud based on the average distance to obtain a plurality of areas, generating multi-resolution super-voxels according to multi-scale seed points, dividing each area based on the super-voxels to obtain a plurality of subspaces, and processing each subspace according to a region growing algorithm to obtain building primitives.
Optionally, the building geometry and deformation extraction method combining the InSAR and the laser point cloud, wherein the reducing the building element to obtain the regularized plane element specifically includes:
respectively calculating a plane equation, a gravity center position and the number of points of the plane primitive, obtaining a normal vector parameter of the plane equation based on the gravity center position and the number of points, and calculating an intersection angle of each plane primitive based on the normal vector parameter;
obtaining mutually coplanar or vertical target plane primitives based on the intersection angle, extracting boundary feature points of the target plane primitives, and processing the boundary feature points according to a least square method to obtain piecewise smooth straight line segments;
And processing the straight line segment according to line-plane projection to obtain a direction angle of the straight line segment, and performing global optimization on the direction angle to obtain the regularized plane primitive.
Optionally, the building geometry and deformation extraction method combining InSAR and laser point cloud, wherein the generating an augmented candidate primitive pool based on the regularized plane primitive, constructing an energy equation, selecting the candidate primitive pool based on the energy equation to obtain a subset, and combining the subset to obtain a target building three-dimensional model specifically includes:
performing element boundary line regularization processing on the regularized plane element to obtain an amplified candidate element pool, and constructing an energy equation based on point characteristics, line characteristics, topological constraint and boundary rules of the candidate element pool;
and calculating a minimum energy solution of the energy equation according to a binary labeling method, selecting the candidate primitive pool based on the minimum energy solution to obtain a plurality of subsets, and combining each subset to obtain the three-dimensional model of the target building.
Optionally, the method for extracting geometry and deformation of a building by combining InSAR and laser point cloud includes performing model analysis on the three-dimensional model of the target building to obtain an analysis result, constructing a temperature deformation model of the building based on the analysis result, and performing verification processing based on the temperature deformation model to obtain temperature deformation and trend deformation of the building, and specifically includes:
Analyzing the temperature deformation space-time evolution characteristics of the three-dimensional model of the target building according to the time sequence interference phase and the time sequence unwrapping phase to obtain an analysis result, and obtaining the temperature deformation characteristics and time sequence deformation information of the building based on the analysis result;
performing coherence weighted least square model fitting on the temperature deformation characteristics and the time series deformation information by combining temperature changes to obtain a temperature deformation model of the building, and performing temperature deformation quantity estimation on the temperature deformation model based on the analysis result to obtain a thermal expansion coefficient of the building;
and carrying out trend deformation and temperature deformation separation on the thermal expansion coefficient according to the mechanical property of the building structure and the physical property of the building material to obtain the temperature deformation and trend deformation of the building.
Optionally, the building geometry and deformation extraction method combining the InSAR and the laser point cloud, wherein the temperature deformation space-time evolution characteristic comprises a propagation direction, spatial distribution and time evolution of the structural temperature deformation.
Optionally, the building geometry and deformation extraction method combining the InSAR and the laser point cloud, wherein the building geometry and deformation extraction system combining the InSAR and the laser point cloud comprises:
The point target fine recognition module is used for acquiring a time sequence SAR image of a building acquired by the interference synthetic aperture radar and a three-dimensional model of the building acquired by the laser radar, and carrying out fine denoising processing on the building point target of multiple feature constraints on the SAR image based on the three-dimensional model to obtain a high-precision three-dimensional point cloud;
the plane primitive extraction module is used for generating multi-resolution super voxels according to the multi-scale seed points, dividing the high-precision three-dimensional point cloud based on the super voxels to obtain a plurality of subspaces, and processing each subspace to obtain a building primitive;
the fine model construction module is used for reducing the building elements to obtain regularized plane elements, generating an augmented candidate element pool based on the regularized plane elements, constructing an energy equation, selecting the candidate element pool based on the energy equation to obtain a subset, and combining the subset to obtain a target building three-dimensional model;
and the high-precision deformation inversion module is used for carrying out model analysis on the three-dimensional model of the target building to obtain an analysis result, constructing a temperature deformation model of the building based on the analysis result, and carrying out verification processing based on the temperature deformation model to obtain the temperature deformation and trend deformation of the building.
In addition, to achieve the above object, the present invention also provides a terminal, wherein the terminal includes: the system comprises a memory, a processor and a program stored on the memory and capable of running on the processor, wherein the program is executed by the processor to realize the steps of the building geometry and deformation extraction method combining InSAR and laser point cloud.
In addition, to achieve the above object, the present invention also provides a computer readable storage medium storing a building geometry and deformation extraction program of a joint InSAR and laser point cloud, which when executed by a processor, implements the steps of the building geometry and deformation extraction method of a joint InSAR and laser point cloud as described above.
In the invention, the SAR image is subjected to constraint denoising treatment based on the three-dimensional model to obtain a high-precision three-dimensional point cloud; processing a plurality of subspaces obtained by dividing the high-precision three-dimensional points according to the super voxels to obtain building primitives; reducing building primitives to obtain regularized plane primitives, generating an augmented candidate primitive pool, constructing an energy equation, selecting the candidate primitive pool to obtain a subset, and combining the subset to obtain a target building three-dimensional model; and carrying out model analysis on the three-dimensional model of the target building, constructing a temperature deformation model of the building, and verifying based on the temperature deformation model to obtain the temperature deformation and the trend deformation of the building. The method can optimize the identification and positioning of the InSAR point targets, effectively improve the efficiency and precision of InSAR point target extraction, assist the temperature deformation modeling of the time sequence InSAR structure, acquire more abundant three-dimensional structure position and deformation information, is crucial for realizing the rapid and high-precision three-dimensional modeling and deformation inversion of the urban building, and has great application potential in the aspects of carrying out the high-precision three-dimensional modeling and deformation inversion on the urban building, improving the deformation generation and evolution rule cognition of the urban building and the like.
Drawings
FIG. 1 is a flow chart of a preferred embodiment of a method for building geometry and deformation extraction combining InSAR with laser point cloud in the present invention;
FIG. 2 is a schematic diagram of Lidar point cloud-assisted point target refinement identification in an embodiment of the invention;
FIG. 3 is a schematic diagram of a fine reconstruction process of a hollow integrated dense matching point cloud building in accordance with an embodiment of the present invention;
FIG. 4 is a schematic diagram of a time series InSAR temperature deformation modeling process based on a high-precision three-dimensional model in an embodiment of the present invention;
FIG. 5 is a schematic overall flow diagram of a building geometry and deformation extraction method combining InSAR and laser point cloud in the present disclosure;
FIG. 6 is a schematic diagram of a preferred embodiment of a building geometry and deformation extraction system combining InSAR with laser point cloud in accordance with the present invention;
FIG. 7 is a schematic diagram of the operating environment of a preferred embodiment of the terminal of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more clear and clear, the present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
It should be noted that, if directional indications (such as up, down, left, right, front, and rear … …) are included in the embodiments of the present invention, the directional indications are merely used to explain the relative positional relationship, movement conditions, etc. between the components in a specific posture (as shown in the drawings), and if the specific posture is changed, the directional indications are correspondingly changed.
In addition, if there is a description of "first", "second", etc. in the embodiments of the present invention, the description of "first", "second", etc. is for descriptive purposes only and is not to be construed as indicating or implying a relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include at least one such feature. In addition, the technical solutions of the embodiments may be combined with each other, but it is necessary to base that the technical solutions can be realized by those skilled in the art, and when the technical solutions are contradictory or cannot be realized, the combination of the technical solutions should be considered to be absent and not within the scope of protection claimed in the present invention.
The method for extracting the geometry and the deformation of the building combining the InSAR and the laser point cloud according to the preferred embodiment of the invention, as shown in fig. 1, comprises the following steps:
s10, acquiring a time sequence SAR image of a building acquired by an interference synthetic aperture radar and a three-dimensional model of the building acquired by a laser radar, and performing fine denoising processing on a building point target of multiple feature constraints on the SAR image based on the three-dimensional model to obtain a high-precision three-dimensional point cloud.
Specifically, after point cloud data of an air-ground platform are acquired, the point cloud data are spliced, the point cloud distribution conditions of different surfaces of each site building are analyzed integrally, and a splicing surface is created according to the concentration of the point cloud; in order to avoid the problem of low accuracy of remote point cloud data of each site, before data processing, the point cloud data acquired by each site are classified according to distance, and point clouds except 100m from the site are deleted, so that the accuracy of overall data is improved, a final high-accuracy three-dimensional point cloud of a building is generated, and effective space reference is provided for the identification and matching of InSAR point targets. Because the urban high-rise building stands and the buildings are dense, scattering signals of the buildings in SAR images are easily interfered by shadows, overlay masks, background noise and the like, the coherence is low, and dense and high-precision building point targets are difficult to extract by applying a conventional InSAR method; as shown in fig. 2, the method researches radar foreground-background scattering three-dimensional scene modeling based on three-dimensional point cloud by introducing high-precision three-dimensional geometric information of Lidar point cloud, and provides powerful basis for accurately simulating radar scattering process of a building and accurately matching and dividing point targets of the building; firstly, constructing a scattered signal interpretation model combined by SAR coherent/incoherent information, adopting a point selection strategy of mutually supplementing SAR coherent/incoherent information, and jointly extracting a permanent coherent and a partial coherent scatterer of a building as a candidate point; for a high-coherence structure, taking a backward scattering signal statistic value (such as average amplitude, amplitude dispersion and the like) of a time sequence SAR image as an incoherent index, taking the time coherence of a target as a coherent index, extracting a point target candidate set which keeps strong and stable backward scattering for a long time, carrying out supplementary observation on SAR coherent/incoherent information, and maximizing the number of detectable permanent coherent scatterers; whereas for low coherence structures, short-time empty baseline interferograms with spectral filtering minimize the effect of decoherence effects so that partially coherent scatterers can also be identified. Meanwhile, considering that stronger vibration possibly occurs in super high-rise buildings, taking the partial coherence coefficient of a small baseline interference pair as a coherence index, and properly relaxing model parameters, such as setting a loose amplitude dispersion threshold value, noise error tolerance and the like as incoherent indexes, and extracting as many point targets as possible on the building structure by combining SAR coherence/incoherent information.
Combining the point target candidate sets acquired by the two different interference data sets, and improving the extraction density and coverage integrity of the building point targets in the complex urban scene; for overlapping points in the candidate set, project is to sum phases of two interference atlases, and a signal-to-noise ratio of each interference atlas is used as a weight value to calculate a weighted average value of the phases of the overlapping points; the position of a point target obtained by the conventional InSAR method is often difficult to accurately identify under the influence of SAR image resolution and side-view imaging geometry, and great challenges are brought to fine denoising of the building point target; in order to improve the efficiency of cognitive interpretation of building structures in SAR images in complex urban scenes, the method introduces high-precision three-dimensional geometric information of Lidar point clouds, assists in identifying accurate positions of building structures, explores imaging characteristics and scattering signal types at different positions of the building, constructs a foreground-background scattering model of the building in the complex urban scenes, analyzes the visibility of the urban building in the SAR images, and accurately inverts the three-dimensional spatial distribution of InSAR point targets.
Performing feature level (such as spatial distribution, coherence, scattering category and the like) description on scattering signals and background signals of urban buildings, performing space-time feature matching on three-dimensional geometric information of candidate InSAR point targets and Lidar point clouds, obtaining spatial position differences of the building scattering signals and the background signals, accurately dividing the building scattering signals and the background signals based on high-precision geometric information, and realizing adjacent scatterer clustering based on real space positions; on the basis, the semantic information of the Lidar point cloud is combined, a building foreground-background three-dimensional scattering scene model is constructed, scattering signal characteristics and differences of building scattering signals and background signals are distinguished, separation of elevation-background point targets is achieved, noise point targets which are not on a building structure are removed by combining multiple scattering characteristic differences, and recognition fineness of InSAR building point targets in complex urban scenes is improved.
And S20, generating multi-resolution super-voxels according to the multi-scale seed points, dividing the high-precision three-dimensional point cloud based on the super-voxels to obtain a plurality of subspaces, and processing each subspace to obtain a building primitive.
Specifically, accurate extraction of point cloud building plane primitives is the basis of refined three-dimensional modeling, while existing point cloud plane primitive extraction algorithms can extract building primitives from homogeneous point clouds acquired by a single platform, but cannot be applied to space-ground integrated dense matching point clouds with severe density and noise variation and extremely large primitive geometric size differences. In order to improve the accuracy and the integrity of automatic extraction of building primitives, the invention performs space division on the point cloud with density change by constructing adaptive resolution super voxels, thereby reducing the interference of the density change on primitive extraction; building element region growth is carried out according to the local density distribution characteristics, and the extraction accuracy and the integrity of building elements are improved; in order to eliminate the obstruction of space-ground integrated point cloud density mutation on building primitive extraction, the original space-ground integrated point cloud is divided into sub-areas with uniform density through point-by-point domain analysis, and then multi-scale area growth is carried out to obtain self-adaptive multi-resolution super-voxels. Firstly, searching k nearest points of each original space-ground integrated point cloud, and calculating average distances d between the original space-ground integrated point cloud and the k points, wherein the average distances d can reflect local densities of the point clouds; then, taking the value of the average distance d as a characteristic, dividing the original space-ground integrated point cloud into a plurality of areas with similar densities by using a k-means clustering method to obtain areas with similar densities; and then generating multi-resolution super voxels by utilizing multi-scale seed points, dividing each region with similar density into a plurality of subspaces with high homogeneity and non-overlapping each other, and extracting continuous plane regions in the point cloud as building primitives by using a region growing algorithm.
And S30, restoring the building primitive to obtain a regularized plane primitive, generating an augmented candidate primitive pool based on the regularized plane primitive, constructing an energy equation, selecting the candidate primitive pool based on the energy equation to obtain a subset, and combining the subset to obtain the target building three-dimensional model.
Specifically, after a certain number of building elements are extracted, the boundaries of the building elements with redundant roughness and sharp characteristic degradation are required to be abstracted, so that boundary line characteristics with concise expression and high regularity are obtained; when the existing primitive boundary simplification and regularization algorithm processes the space-ground integrated dense matching point cloud with extremely large primitive geometric size, sampling density and noise level difference, the regularity and fidelity are difficult to balance, and the global rule relation between the building surface characteristics and the line characteristics is not considered, so that a multi-level building primitive boundary regularization method is adopted in the embodiment of the invention, as shown in fig. 3; firstly, regularizing primitive planes of a building, and recovering the parallel (coplanar) and vertical relation between the primitive planes; then, locally considering the initial position of the boundary feature point and the normal vector thereof, and enabling the boundary point to move along the normal vector through least square constraint to realize noise suppression, so as to generate a piecewise smooth straight line segment; and finally, all primitive boundary straight-line segments of the same building are connected through line-plane projection, global optimization is carried out on the direction angles of the straight-line segments, and the fidelity and the global regularity of the boundary lines of the building are improved.
Further, there is often a clear parallel, coplanar or perpendicular relationship between different planar primitives of the same building, however, the point cloud data is affected by noise, which breaks this regularity between primitives; therefore, in the same building, the approximately parallel or vertical primitive point clouds are subjected to projection processing, and the regular relation among primitive planes is restored; first, the plane equation of the point cloud contained in each primitive is calculated:
A i x +B i y+C i z+Di=0; wherein, the liquid crystal display device comprises a liquid crystal display device,A i ,B i ,C i ,D i are all the firstiPlane equation parameters for the individual primitives; then, the center of gravity position of each planar primitive is calculated（GX i ,GY i ,GZ i ）Sum of pointsNiObtaining a normal vector parameter of the plane equation based on the gravity center position and the point number; calculating intersections between planes by normal vector parameters in plane equationsCells that are angular and tend to be 0 ° or tend to be 90 ° within a certain threshold range are considered to be parallel (coplanar) or perpendicular to each other; finally, according to the direction angle of the primitive plane primitive, the weighted average direction angle of the primitive group is obtained by taking the number of primitive points as the weight, and the primitive point clouds of the same group are subjected to projection conversion according to the direction angle, so that the regularization of the primitive plane of the building is realized.
Further, after the building elements and boundary characteristics thereof are obtained, the topological relation among the building elements needs to be restored to obtain a complete three-dimensional building model; the existing primitive topology recovery and model generation algorithm only aims at the point cloud generated by a single aviation platform, and cannot effectively process the space-ground integrated dense matching point cloud with a large number of primitives, complex topological relations and data loss among detail primitives; therefore, in the embodiment of the invention, the intelligent augmented candidate primitive pool is generated by combining the adjacent relation between the primitive surface-line characteristics of the building and the regularized relation between primitive boundary lines and between the boundary lines and primitive planes, so as to make up for possible data loss; then constructing an energy equation, and selecting a subset from the candidate primitive pool to form a building three-dimensional model with high geometric precision, good regularity and correct topological relation. In order to select the primitive clusters with optimal geometry and topology from the primitive pool to form a three-dimensional model of a building, an energy equation is established E，E=E pts +E lines +E topo +E regu Comprehensively considering the fitting property of candidate polygons to point cloud dataE pts ) Fitting of candidate polygon boundary to primitive boundary line characteristicsE lines ) Topological constraint relation between polygon surface and lineE topo ) And regularity of polygonal boundary lineE regu ) These four factors; for the point characteristics, the average distance from the point clouds to the primitive plane and the projection coverage proportion are counted as fitness measures by extracting the original point clouds contained in a certain buffer range of each candidate primitive; for line characteristics, the superposition distance between candidate element intersecting lines and the same-direction adjacent building boundary characteristic lines is calculatedA vertical distance from the straight line; for topology correctness, the topology conflict among the selected plane primitives is avoided by explicitly restricting the number of primitives connected by the same intersection line; for boundary regularity, the directions of the element intersection lines are calculated, the element intersection lines are grouped according to the mutually perpendicular relation, the accumulated length of each group is counted, and the occurrence of redundant directions is punished; and finally, calculating the minimum energy solution of the energy equation by a binary labeling method to obtain the three-dimensional building model with good point-line characteristic fitting, high regularity and correct topological relation.
And S40, carrying out model analysis on the three-dimensional model of the target building to obtain an analysis result, constructing a temperature deformation model of the building based on the analysis result, and carrying out verification processing based on the temperature deformation model to obtain the temperature deformation and trend deformation of the building.
Specifically, in the embodiment of the invention, a time sequence InSAR temperature deformation modeling method considering a fine three-dimensional structure model of a building is provided, and the high-precision three-dimensional model of the building not only comprises geometric information but also comprises semantic information; modeling and analyzing the temperature deformation of the building by adopting a least square method of coherence constraint, wherein deformation signals are obtained by time sequence InSAR technology space-time filtering estimation, and modeling and analyzing the residual deformation phase based on temperature data driving are carried out after other phases are separated without depending on a linear deformation model hypothesis of the structure; the method mainly comprises three stages: a model analysis stage, a model construction stage, and a model verification stage, as shown in fig. 4; in the model analysis stage, mainly analyzing a time sequence differential interference pattern and a unwrapped phase pattern, starting from the development direction and the intensive degree change condition of interference fringes, qualitatively judging the correlation relation between structural variables and temperature change, and by introducing a high-precision three-dimensional model of a building, matching InSAR deformation results to a specific three-dimensional space of the building structure, exploring the space-time evolution characteristics of temperature deformation at different positions of the building and the differences thereof, mainly comprising the propagation direction, the space distribution, the time evolution and the like of the temperature deformation of the structure, and providing an effective data basis for constructing a temperature deformation model of the building structure in the three-dimensional space.
In the model construction stage, assuming that the temperature on the building structure is uniformly distributed in a three-dimensional space, quantitatively establishing a correlation model between time series deformation and temperature change of the structure by using a least square regression analysis method, and solving model parameters which minimize an error function; however, this conventional least squares method is very sensitive to outliers in the dataset, which may cause bias in the phase model estimation; thus, in embodiments of the present invention this problem is alleviated by introducing the coherence of interference pairs into the estimation of the temperature deformation, i.e. the observations obtained for interference pairs with higher coherence are considered more reliable, and the maximum coherence coefficient of the interferogram is used to determine the weights of the corresponding time series observations, i.e.:
wherein->
Is the number of SAR images,/->
Indicating measured values,/->
Representing the estimated value fitted by the model, +.>
The weight value of the main image is 1 for the maximum coherence coefficient of the j interference pair. According to the previous large-scale research, the conclusion that the temperature deformation of the structure is positively correlated with the temperature difference and the temperature strain caused by the larger temperature difference is more severe is obtained; therefore, regression analysis is performed on the temperature deformation assuming that the temperature deformation model is:
Wherein, the liquid crystal display device comprises a liquid crystal display device,
for temperature change->
The temperature deformation quantity which occurs along the temperature deformation propagation direction is a constant; based on the temperature deformation model, more useful information about the temperature deformation of the structure can be explored, and the method is not limited in the invention; first, when->
When equal to zero, assume that the temperature at the time of main image acquisition is +.>
The reference temperature of the structure can be calculated>
I.e. the temperature at which no temperature deformation occurs; furthermore, the thermal expansion coefficient of the structure +.>
(/ DEGC 2) can also be calculated, and the calculation formula is as follows:
In the model verification stage, qualitatively comparing the spatial distribution of the model with the spatial distribution of the temperature deformation predicted based on the structural mechanics principle; quantitatively, comparing the thermal expansion coefficient estimated by the temperature deformation model with the actual physical properties of the material, or performing cross validation on the temperature deformation models estimated by different SAR data sets, and separating the temperature deformation models from the overall deformation after estimating the temperature deformation, so as to obtain the temperature deformation and trend deformation of the building; according to the time sequence InSAR temperature deformation modeling method based on the high-precision three-dimensional model, the high-precision three-dimensional model of the building is introduced into deformation phase resolving, the distribution and evolution rules of interference fringes on different structures of the building are accurately extracted, the building trend deformation and the temperature deformation are separated, and the three-dimensional space fineness of deformation inversion is improved.
Further, the conventional deformation monitoring product only analyzes the whole deformation of the building, and is difficult to realize detailed deformation inversion and temperature deformation separation of different positions of the building, so that great challenges are brought to practical application of urban building deformation monitoring; according to the invention, lidar high-precision three-dimensional information is integrated into time sequence InSAR building deformation inversion, and by acquiring and integrating high-precision three-dimensional point clouds of a surface building, the method not only can optimize the identification and positioning of InSAR point targets and effectively improve the efficiency and precision of InSAR point target extraction, but also can assist in time sequence InSAR structure temperature deformation modeling, acquire real-time and high-precision building three-dimensional position and deformation information products which are more valuable to users, and send the obtained products to related departments and management personnel or end users, thereby better serving three-dimensional modeling and safety monitoring application of urban buildings, having the characteristics of low monitoring cost, high efficiency, high precision and the like, and having great application potential in the aspects of carrying out high-precision three-dimensional modeling and deformation inversion on urban buildings, improving deformation generation and evolution cognition on urban buildings and the like.
Further, as shown in fig. 5, the overall flow diagram of the building geometry and deformation extraction method combining the InSAR and the laser point cloud specifically includes: aiming at the problems of small density and low precision of InSAR point targets of urban buildings in complex scenes, which lead to overlapping and masking of SAR signals, the invention realizes the accurate identification of the InSAR point targets through the accurate identification of the InSAR point targets assisted by Lidar point clouds; the second difficulty is that the density and the precision of the point cloud are greatly changed due to large geometric size difference of building elements, and the fine reconstruction of the point cloud building is realized through the dense matching of the space and the ground, and the obtained high-precision three-dimensional model of the building is used as a data basis; the third difficulty is that deformation phase estimation deviation of the building is large, so that a deformation mechanism is complex, and InSAR high-precision deformation inversion of the building is realized through InSAR temperature deformation modeling based on a high-precision three-dimensional model.
Further, as shown in fig. 6, the present invention further provides a system for extracting geometry and deformation of a building by combining the InSAR and the laser point cloud based on the method for extracting geometry and deformation of a building by combining the InSAR and the laser point cloud, where the system for extracting geometry and deformation of a building by combining the InSAR and the laser point cloud includes:
The point target fine recognition module 51 is used for acquiring a time sequence SAR image of a building acquired by the interference synthetic aperture radar and a three-dimensional model of the building acquired by the laser radar, and performing multi-feature constraint building point target fine denoising processing on the SAR image based on the three-dimensional model to obtain a high-precision three-dimensional point cloud;
the plane primitive extraction module 52 is configured to generate a multi-resolution super voxel according to a multi-scale seed point, divide the high-precision three-dimensional point cloud based on the super voxel to obtain a plurality of subspaces, and process each subspace to obtain a building primitive;
the fine model construction module 53 is configured to restore the building element to obtain a regularized plane element, generate an augmented candidate element pool based on the regularized plane element, construct an energy equation, select the candidate element pool based on the energy equation to obtain a subset, and combine the subset to obtain a target building three-dimensional model;
the high-precision deformation inversion module 54 is configured to perform model analysis on the three-dimensional model of the target building to obtain an analysis result, construct a temperature deformation model of the building based on the analysis result, and perform verification processing based on the temperature deformation model to obtain temperature deformation and trend deformation of the building.
Further, as shown in fig. 7, the invention further provides a terminal based on the building geometry and deformation extraction method combining the InSAR and the laser point cloud, which comprises a processor 10, a memory 20 and a display 30. Fig. 7 shows only some of the components of the terminal, but it should be understood that not all of the illustrated components are required to be implemented and that more or fewer components may alternatively be implemented.
The memory 20 may in some embodiments be an internal storage unit of the terminal, such as a hard disk or a memory of the terminal. The memory 20 may in other embodiments also be an external storage device of the terminal, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card) or the like, which are provided on the terminal. Further, the memory 20 may also include both an internal storage unit and an external storage device of the terminal. The memory 20 is used for storing application software installed in the terminal and various data, such as program codes of the installation terminal. The memory 20 may also be used to temporarily store data that has been output or is to be output. In an embodiment, the memory 20 stores a building geometry and deformation extraction program 40 of the combined InSAR and laser point cloud, and the building geometry and deformation extraction program 40 of the combined InSAR and laser point cloud may be executed by the processor 10, so as to implement a building geometry and deformation extraction method of the combined InSAR and laser point cloud in the present application.
The processor 10 may in some embodiments be a central processing unit (Central Processing Unit, CPU), microprocessor or other data processing chip for executing program code or processing data stored in the memory 20, for example, for performing the building information processing method of combining InSAR with lidar, etc.
The display 30 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch, or the like in some embodiments. The display 30 is used for displaying information at the terminal and for displaying a visual user interface. The components 10-30 of the terminal communicate with each other via a system bus.
In one embodiment, the following steps are implemented when the processor 10 executes the building geometry and deformation extraction program 40 of the joint InSAR and laser point cloud in the memory 20:
acquiring a time sequence SAR image of a building acquired by an interference synthetic aperture radar and a three-dimensional model of the building acquired by a laser radar, and performing fine denoising processing on a building point target of multiple feature constraints on the SAR image based on the three-dimensional model to obtain a high-precision three-dimensional point cloud;
Generating multi-resolution super-voxels according to the multi-scale seed points, dividing the high-precision three-dimensional point cloud based on the super-voxels to obtain a plurality of subspaces, and processing each subspace to obtain building primitives;
reducing the building primitive to obtain a regularized plane primitive, generating an augmented candidate primitive pool based on the regularized plane primitive, constructing an energy equation, selecting the candidate primitive pool based on the energy equation to obtain a subset, and combining the subset to obtain a target building three-dimensional model;
and carrying out model analysis on the three-dimensional model of the target building to obtain an analysis result, constructing a temperature deformation model of the building based on the analysis result, and carrying out verification processing based on the temperature deformation model to obtain the temperature deformation and trend deformation of the building.
The method for acquiring the time sequence SAR image of the building acquired by the interference synthetic aperture radar and the three-dimensional model of the building acquired by the laser radar, performing fine denoising processing on the building point target of multiple feature constraints on the SAR image based on the three-dimensional model to obtain high-precision three-dimensional point cloud, and specifically comprises the following steps:
Acquiring a PSI differential interference atlas and an SBAS differential interference atlas of a time sequence synthetic aperture radar image acquired by an interference synthetic aperture radar and semantic information and geometric information of a building acquired by a laser radar;
performing PS scatterer extraction on the PSI differential interference atlas to obtain a high-coherence structure, performing DS scatterer extraction on the SBAS differential interference atlas to obtain a low-coherence structure, and constructing a three-dimensional scattering scene model of a building foreground-background based on the semantic information;
constructing a scattered signal interpretation model combining coherent information and incoherent information of the synthetic aperture radar, performing scatterer selection recognition on the high-coherence structure and the low-coherence structure based on the scattered signal interpretation model to obtain a recognition result, and performing point target space position matching on the recognition result and the geometric information to obtain a matching result;
and carrying out building point target segmentation on the matching result to obtain a scattering signal and a background signal of the building, and inputting the scattering signal and the background signal into the three-dimensional scattering scene model to obtain the high-precision three-dimensional point cloud of the building.
Generating a multi-resolution super-voxel according to the multi-scale seed point, dividing the high-precision three-dimensional point cloud based on the super-voxel to obtain a plurality of subspaces, and processing each subspace to obtain a building primitive, wherein the method specifically comprises the following steps of:
Acquiring an original space-ground integrated point cloud of the building, searching the original space-ground integrated point cloud to obtain a plurality of target adjacent points, and calculating the average distance between the original space-ground integrated point cloud and the plurality of target adjacent points;
dividing the original space-ground integrated point cloud based on the average distance to obtain a plurality of areas, generating multi-resolution super-voxels according to multi-scale seed points, dividing each area based on the super-voxels to obtain a plurality of subspaces, and processing each subspace according to a region growing algorithm to obtain building primitives.
The method for restoring the building element to obtain the regularized plane element specifically comprises the following steps:
respectively calculating a plane equation, a gravity center position and the number of points of the plane primitive, obtaining a normal vector parameter of the plane equation based on the gravity center position and the number of points, and calculating an intersection angle of each plane primitive based on the normal vector parameter;
obtaining mutually coplanar or vertical target plane primitives based on the intersection angle, extracting boundary feature points of the target plane primitives, and processing the boundary feature points according to a least square method to obtain piecewise smooth straight line segments;
And processing the straight line segment according to line-plane projection to obtain a direction angle of the straight line segment, and performing global optimization on the direction angle to obtain the regularized plane primitive.
The method specifically comprises the steps of generating an augmented candidate primitive pool based on the regularized plane primitive, constructing an energy equation, selecting the candidate primitive pool based on the energy equation to obtain a subset, and combining the subset to obtain a target building three-dimensional model, wherein the method specifically comprises the following steps:
performing element boundary line regularization processing on the regularized plane element to obtain an amplified candidate element pool, and constructing an energy equation based on point characteristics, line characteristics, topological constraint and boundary rules of the candidate element pool;
and calculating a minimum energy solution of the energy equation according to a binary labeling method, selecting the candidate primitive pool based on the minimum energy solution to obtain a plurality of subsets, and combining each subset to obtain the three-dimensional model of the target building.
The method comprises the steps of carrying out model analysis on the three-dimensional model of the target building to obtain an analysis result, constructing a temperature deformation model of the building based on the analysis result, and carrying out verification processing based on the temperature deformation model to obtain the temperature deformation and trend deformation of the building, wherein the method specifically comprises the following steps of:
Analyzing the temperature deformation space-time evolution characteristics of the three-dimensional model of the target building according to the time sequence interference phase and the time sequence unwrapping phase to obtain an analysis result, and obtaining the temperature deformation characteristics and time sequence deformation information of the building based on the analysis result;
performing coherence weighted least square model fitting on the temperature deformation characteristics and the time series deformation information by combining temperature changes to obtain a temperature deformation model of the building, and performing temperature deformation quantity estimation on the temperature deformation model based on the analysis result to obtain a thermal expansion coefficient of the building;
and carrying out trend deformation and temperature deformation separation on the thermal expansion coefficient according to the mechanical property of the building structure and the physical property of the building material to obtain the temperature deformation and trend deformation of the building.
The temperature deformation space-time evolution characteristics comprise the propagation direction, the spatial distribution and the time evolution of the temperature deformation of the structure.
The present invention also provides a computer readable storage medium storing a building geometry and deformation extraction program of a joint InSAR and laser point cloud, which when executed by a processor, implements the steps of the building geometry and deformation extraction method of a joint InSAR and laser point cloud as described above.
In summary, the method for extracting geometry and deformation of a building combining InSAR and laser point cloud provided by the invention comprises the following steps: performing constraint denoising processing on the SAR image based on the three-dimensional model to obtain a high-precision three-dimensional point cloud; processing a plurality of subspaces obtained by dividing the high-precision three-dimensional points according to the super voxels to obtain building primitives; reducing building primitives to obtain regularized plane primitives, generating an augmented candidate primitive pool, constructing an energy equation, selecting the candidate primitive pool to obtain a subset, and combining the subset to obtain a target building three-dimensional model; and carrying out model analysis on the three-dimensional model of the target building, constructing a temperature deformation model of the building, and verifying based on the temperature deformation model to obtain the temperature deformation and the trend deformation of the building. The invention can rapidly provide high-precision position and deformation information of the building in real time, and plays an important role in safe operation and emergency response of urban buildings.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
Of course, those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by a computer program for instructing relevant hardware (e.g., processor, controller, etc.), the program may be stored on a computer readable storage medium, and the program may include the above described methods when executed. The computer readable storage medium may be a memory, a magnetic disk, an optical disk, etc.
It is to be understood that the invention is not limited in its application to the examples described above, but is capable of modification and variation in light of the above teachings by those skilled in the art, and that all such modifications and variations are intended to be included within the scope of the appended claims.
Claims (10)
1. The method for extracting the geometry and the deformation of the building by combining the InSAR and the laser point cloud is characterized by comprising the following steps of:
acquiring a time sequence SAR image of a building acquired by an interference synthetic aperture radar and a three-dimensional model of the building acquired by a laser radar, and performing fine denoising processing on a building point target of multiple feature constraints on the SAR image based on the three-dimensional model to obtain a high-precision three-dimensional point cloud;
Generating multi-resolution super-voxels according to the multi-scale seed points, dividing the high-precision three-dimensional point cloud based on the super-voxels to obtain a plurality of subspaces, and processing each subspace to obtain building primitives;
reducing the building primitive to obtain a regularized plane primitive, generating an augmented candidate primitive pool based on the regularized plane primitive, constructing an energy equation, selecting the candidate primitive pool based on the energy equation to obtain a subset, and combining the subset to obtain a target building three-dimensional model;
and carrying out model analysis on the three-dimensional model of the target building to obtain an analysis result, constructing a temperature deformation model of the building based on the analysis result, and carrying out verification processing based on the temperature deformation model to obtain the temperature deformation and trend deformation of the building.
2. The method for extracting geometry and deformation of a building by combining InSAR and laser point cloud according to claim 1, wherein the method for obtaining a time sequence SAR image of a building collected by an interferometric synthetic aperture radar and a three-dimensional model of a building collected by a laser radar, performing fine denoising processing on a building point target of multiple feature constraints on the SAR image based on the three-dimensional model, and obtaining a high-precision three-dimensional point cloud specifically comprises:
Acquiring a PSI differential interference atlas and an SBAS differential interference atlas of a time sequence synthetic aperture radar image acquired by an interference synthetic aperture radar and semantic information and geometric information of a building acquired by a laser radar;
performing PS scatterer extraction on the PSI differential interference atlas to obtain a high-coherence structure, performing DS scatterer extraction on the SBAS differential interference atlas to obtain a low-coherence structure, and constructing a three-dimensional scattering scene model of a building foreground-background based on the semantic information;
constructing a scattered signal interpretation model combining coherent information and incoherent information of the synthetic aperture radar, performing scatterer selection recognition on the high-coherence structure and the low-coherence structure based on the scattered signal interpretation model to obtain a recognition result, and performing point target space position matching on the recognition result and the geometric information to obtain a matching result;
and carrying out building point target segmentation on the matching result to obtain a scattering signal and a background signal of the building, and inputting the scattering signal and the background signal into the three-dimensional scattering scene model to obtain the high-precision three-dimensional point cloud of the building.
3. The method for extracting geometry and deformation of a building by combining InSAR and laser point cloud according to claim 1, wherein generating multi-resolution super-voxels according to multi-scale seed points, dividing the high-precision three-dimensional point cloud based on the super-voxels to obtain a plurality of subspaces, and processing each subspace to obtain building primitives, specifically comprising:
Acquiring an original space-ground integrated point cloud of the building, searching the original space-ground integrated point cloud to obtain a plurality of target adjacent points, and calculating the average distance between the original space-ground integrated point cloud and the plurality of target adjacent points;
dividing the original space-ground integrated point cloud based on the average distance to obtain a plurality of areas, generating multi-resolution super-voxels according to multi-scale seed points, dividing each area based on the super-voxels to obtain a plurality of subspaces, and processing each subspace according to a region growing algorithm to obtain building primitives.
4. The method for extracting geometry and deformation of a building by combining InSAR and laser point cloud according to claim 1, wherein the reducing the building element to obtain a regularized plane element specifically comprises:
respectively calculating a plane equation, a gravity center position and the number of points of the plane primitive, obtaining a normal vector parameter of the plane equation based on the gravity center position and the number of points, and calculating an intersection angle of each plane primitive based on the normal vector parameter;
obtaining mutually coplanar or vertical target plane primitives based on the intersection angle, extracting boundary feature points of the target plane primitives, and processing the boundary feature points according to a least square method to obtain piecewise smooth straight line segments;
And processing the straight line segment according to line-plane projection to obtain a direction angle of the straight line segment, and performing global optimization on the direction angle to obtain the regularized plane primitive.
5. The method for extracting geometry and deformation of a building by combining InSAR and laser point cloud according to claim 1, wherein the generating an augmented candidate primitive pool based on the regularized plane primitive and constructing an energy equation, selecting the candidate primitive pool based on the energy equation to obtain a subset, and combining the subset to obtain a three-dimensional model of the target building specifically comprises:
performing element boundary line regularization processing on the regularized plane element to obtain an amplified candidate element pool, and constructing an energy equation based on point characteristics, line characteristics, topological constraint and boundary rules of the candidate element pool;
and calculating a minimum energy solution of the energy equation according to a binary labeling method, selecting the candidate primitive pool based on the minimum energy solution to obtain a plurality of subsets, and combining each subset to obtain the three-dimensional model of the target building.
6. The method for extracting geometry and deformation of a building by combining InSAR and laser point cloud according to claim 1, wherein the method for analyzing the three-dimensional model of the target building to obtain an analysis result, constructing a temperature deformation model of the building based on the analysis result, and performing verification processing based on the temperature deformation model to obtain temperature deformation and trend deformation of the building specifically comprises:
Analyzing the temperature deformation space-time evolution characteristics of the three-dimensional model of the target building according to the time sequence interference phase and the time sequence unwrapping phase to obtain an analysis result, and obtaining the temperature deformation characteristics and time sequence deformation information of the building based on the analysis result;
performing coherence weighted least square model fitting on the temperature deformation characteristics and the time series deformation information by combining temperature changes to obtain a temperature deformation model of the building, and performing temperature deformation quantity estimation on the temperature deformation model based on the analysis result to obtain a thermal expansion coefficient of the building;
and carrying out trend deformation and temperature deformation separation on the thermal expansion coefficient according to the mechanical property of the building structure and the physical property of the building material to obtain the temperature deformation and trend deformation of the building.
7. The method for extracting geometry and deformation of a building by combining InSAR and laser point cloud according to claim 6, wherein the temperature deformation space-time evolution features comprise propagation direction, spatial distribution and time evolution of structural temperature deformation.
8. A building geometry and deformation extraction system combining InSAR and laser point cloud, characterized in that the building geometry and deformation extraction system combining InSAR and laser point cloud comprises:
The point target fine recognition module is used for acquiring a time sequence SAR image of a building acquired by the interference synthetic aperture radar and a three-dimensional model of the building acquired by the laser radar, and carrying out fine denoising processing on the building point target of multiple feature constraints on the SAR image based on the three-dimensional model to obtain a high-precision three-dimensional point cloud;
the plane primitive extraction module is used for generating multi-resolution super voxels according to the multi-scale seed points, dividing the high-precision three-dimensional point cloud based on the super voxels to obtain a plurality of subspaces, and processing each subspace to obtain a building primitive;
the fine model construction module is used for reducing the building elements to obtain regularized plane elements, generating an augmented candidate element pool based on the regularized plane elements, constructing an energy equation, selecting the candidate element pool based on the energy equation to obtain a subset, and combining the subset to obtain a target building three-dimensional model;
and the high-precision deformation inversion module is used for carrying out model analysis on the three-dimensional model of the target building to obtain an analysis result, constructing a temperature deformation model of the building based on the analysis result, and carrying out verification processing based on the temperature deformation model to obtain the temperature deformation and trend deformation of the building.
9. A terminal comprising a memory, a processor and a program stored on the memory and executable on the processor, which when executed by the processor performs the steps of the building geometry and distortion extraction method of combining InSAR with a laser point cloud according to any of claims 1-7.
10. A computer-readable storage medium having stored thereon a computer program, the computer-readable storage medium having stored thereon a building geometry and deformation extraction program of a joint InSAR and laser point cloud, which when executed by a processor, implements the steps of the building three-dimensional information dynamic processing method of the joint InSAR and laser radar according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310437935.5A CN116168176B (en) | 2023-04-23 | 2023-04-23 | Building geometry and deformation extraction method combining InSAR and laser point cloud |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310437935.5A CN116168176B (en) | 2023-04-23 | 2023-04-23 | Building geometry and deformation extraction method combining InSAR and laser point cloud |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116168176A true CN116168176A (en) | 2023-05-26 |
| CN116168176B CN116168176B (en) | 2023-07-18 |
Family
ID=86413445
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310437935.5A Active CN116168176B (en) | 2023-04-23 | 2023-04-23 | Building geometry and deformation extraction method combining InSAR and laser point cloud |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116168176B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116908789A (en) * | 2023-09-13 | 2023-10-20 | 长江空间信息技术工程有限公司(武汉) | Foundation synthetic aperture radar interferometry building elevation deformation information extraction method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5988862A (en) * | 1996-04-24 | 1999-11-23 | Cyra Technologies, Inc. | Integrated system for quickly and accurately imaging and modeling three dimensional objects |
| US6420698B1 (en) * | 1997-04-24 | 2002-07-16 | Cyra Technologies, Inc. | Integrated system for quickly and accurately imaging and modeling three-dimensional objects |
| CN107545602A (en) * | 2017-08-04 | 2018-01-05 | 长江空间信息技术工程有限公司(武汉) | Building Modeling method under spatial topotaxy constraint based on LiDAR point cloud |
| CN108627834A (en) * | 2018-06-07 | 2018-10-09 | 北京城建勘测设计研究院有限责任公司 | A kind of subway road structure monitoring method and device based on ground InSAR |
| US20190026400A1 (en) * | 2017-07-18 | 2019-01-24 | Fuscoe Engineering, Inc. | Three-dimensional modeling from point cloud data migration |
| CN111815776A (en) * | 2020-02-04 | 2020-10-23 | 山东水利技师学院 | Three-dimensional building fine geometric reconstruction method integrating airborne and vehicle-mounted three-dimensional laser point clouds and streetscape images |
| WO2023028774A1 (en) * | 2021-08-30 | 2023-03-09 | 华为技术有限公司 | Lidar calibration method and apparatus, and storage medium |
-
2023
- 2023-04-23 CN CN202310437935.5A patent/CN116168176B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5988862A (en) * | 1996-04-24 | 1999-11-23 | Cyra Technologies, Inc. | Integrated system for quickly and accurately imaging and modeling three dimensional objects |
| US6420698B1 (en) * | 1997-04-24 | 2002-07-16 | Cyra Technologies, Inc. | Integrated system for quickly and accurately imaging and modeling three-dimensional objects |
| US20190026400A1 (en) * | 2017-07-18 | 2019-01-24 | Fuscoe Engineering, Inc. | Three-dimensional modeling from point cloud data migration |
| CN107545602A (en) * | 2017-08-04 | 2018-01-05 | 长江空间信息技术工程有限公司(武汉) | Building Modeling method under spatial topotaxy constraint based on LiDAR point cloud |
| CN108627834A (en) * | 2018-06-07 | 2018-10-09 | 北京城建勘测设计研究院有限责任公司 | A kind of subway road structure monitoring method and device based on ground InSAR |
| CN111815776A (en) * | 2020-02-04 | 2020-10-23 | 山东水利技师学院 | Three-dimensional building fine geometric reconstruction method integrating airborne and vehicle-mounted three-dimensional laser point clouds and streetscape images |
| WO2023028774A1 (en) * | 2021-08-30 | 2023-03-09 | 华为技术有限公司 | Lidar calibration method and apparatus, and storage medium |
Non-Patent Citations (4)
| Title |
|---|
| SINA MONTAZERI 等: "Three-Dimensional Deformation Monitoring of Urban Infrastructure by Tomographic SAR Using Multitrack TerraSAR-X Data Stacks", 《IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING》, vol. 54, no. 12, pages 6868 - 6878, XP011624353, DOI: 10.1109/TGRS.2016.2585741 * |
| 秦晓琼: "时间序列D-InSAR城市基础设施精细形变测量研究", 《中国博士学位论文全文数据库基础科学辑》, no. 6, pages 008 - 63 * |
| 肖娇琳 等: "三维激光扫描建筑物变形信息提取方法研究", 《测绘与空间地理信息》, vol. 45, no. 6, pages 206 - 209 * |
| 郑翔天 等: "点云辅助GB-InSAR影像与地形数据应急变形监测方法", 《武汉大学学报(信息科学版)》, vol. 47, no. 7, pages 1081 - 1092 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116908789A (en) * | 2023-09-13 | 2023-10-20 | 长江空间信息技术工程有限公司(武汉) | Foundation synthetic aperture radar interferometry building elevation deformation information extraction method |
| CN116908789B (en) * | 2023-09-13 | 2023-12-05 | 长江空间信息技术工程有限公司(武汉) | Foundation synthetic aperture radar interferometry building elevation deformation information extraction method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116168176B (en) | 2023-07-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shahzad et al. | Buildings detection in VHR SAR images using fully convolution neural networks | |
| Li et al. | Stacking ensemble of deep learning methods for landslide susceptibility mapping in the Three Gorges Reservoir area, China | |
| CN116168176B (en) | Building geometry and deformation extraction method combining InSAR and laser point cloud | |
| Schmitt et al. | Reconstruction of individual trees from multi-aspect TomoSAR data | |
| Méneroux et al. | Traffic signal detection from in-vehicle GPS speed profiles using functional data analysis and machine learning | |
| Danescu et al. | A particle-based solution for modeling and tracking dynamic digital elevation maps | |
| Eroglu et al. | A terrain referenced UAV localization algorithm using binary search method | |
| Xiu et al. | 3D semantic segmentation for high-resolution aerial survey derived point clouds using deep learning | |
| Long et al. | Object detection research of SAR image using improved faster region-based convolutional neural network | |
| Liu et al. | A novel entropy-based method to quantify forest canopy structural complexity from multiplatform lidar point clouds | |
| Jiangui et al. | A method for main road extraction from airborne LiDAR data in urban area | |
| Gézero et al. | Automated road curb break lines extraction from mobile lidar point clouds | |
| MohanRajan et al. | Fuzzy Swin transformer for Land Use/Land Cover change detection using LISS-III Satellite data | |
| Minh et al. | Interferometric phase linking: Algorithm, application, and perspective | |
| Nefeslioglu et al. | Integration of an InSAR and ANN for sinkhole susceptibility mapping: A case study from Kirikkale-Delice (Turkey) | |
| Crosby et al. | Object identification in 3D flash lidar images | |
| CN113409446B (en) | Blind person assisted vision processing method and device | |
| Kumar et al. | An efficient method for road tracking from satellite images using hybrid multi-kernel partial least square analysis and particle filter | |
| Lv et al. | Marsnet: Automated rock segmentation with transformers for tianwen-1 mission | |
| Wang et al. | 3D building reconstruction from LiDAR data | |
| Chen et al. | 3D Object detector: A multiscale region proposal network based on autonomous driving | |
| Guo et al. | Construction of 3D landscape indexes based on oblique photogrammetry and its application for islands | |
| Wang et al. | InSAR Phase Unwrapping Algorithm Based on Deep GAN | |
| Fernández et al. | Dynamics of the Burlan and Pomacochas Lakes Using SAR Data in GEE, Machine Learning Classifiers, and Regression Methods | |
| Sharafzadeh et al. | 3D modelling of urban area using synthetic aperture radar (SAR) |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |