CN103323848A - Method and device for extracting height of ground artificial building/structure - Google Patents

Abstract

The invention relates to a method and device for extracting the height of artificial buildings/structures on the ground, wherein the method includes: combining interference image pairs according to the principle of small baseline to process synthetic aperture radar images covering the same area to generate a time series interferogram; The magnitude information of the radar image screens out candidate high-coherence points; the time-series interferogram is processed to obtain the time-series differential interferogram, and the differential interferometric phase of the candidate high-coherence points is extracted; Spatial correlation phase, to obtain the spatial non-correlation phase of the candidate high coherence point; wherein, the spatial non-correlation phase includes the spatial non-correlation DEM correction amount and the residual phase; process the spatial non-correlation phase in the differential interferometric phase of the candidate high coherence point , to obtain the correction value of the uncorrelated DEM in the solution space, and select the final high coherence point from the candidate high coherence points; obtain the height of the artificial building/structure on the ground according to the final high coherence point.

Description

A method and device for extracting the height of artificial buildings/structures on the ground

technical field

The invention relates to the field of synthetic aperture radar images, in particular to a method and device for extracting the height of artificial buildings/structures on the ground by using time series InSAR technology.

Background technique

The height information of artificial buildings/structures is important data for urban planning, early warning and assessment of urban disaster risks, and is also an important basic data for building a 3D digital city. In order to obtain the height information of buildings/structures, total stations, optical stereo photogrammetry or LiDAR technology can be used. The use of total stations cannot quickly complete large-scale height information extraction; optical stereo photogrammetry technology has high requirements for data acquisition and is easily affected by weather conditions; LiDAR technology, as an emerging remote sensing technology, can directly obtain the height information of targets. It is widely used in obtaining 3D digital models of urban areas, but the cost is relatively high. In addition, the use of optical stereo photogrammetry or LiDAR technology often misses objects with a small projected area (such as street lights, electric towers, etc.).

Synthetic Aperture Radar (SAR) technology is a new earth observation technology developed in the 1950s. It emits electromagnetic pulses at regular time intervals on an orbit hundreds of kilometers away from the earth's surface, and uses the movement of the carrier to The echo signals of ground objects are received and recorded at different positions to form high-coverage, high-resolution images. SAR technology belongs to active microwave remote sensing technology, which can not be affected by sunlight, weather, etc., and has the ability of imaging at night and penetrating clouds and fog.

In the 1970s, Synthetic Aperture Radar Interferometry (InSAR) formed an interferogram by using two SAR images of the same area acquired with slightly different spatial perspectives, making full use of the phase information contained in the image. It provides a new idea for obtaining the digital elevation model (DEM, Digital Elevation Model) information of the ground. In 2000, the United States implemented the Shuttle Radar Topography Mapping (SRTM) program. After the main antenna transmits radar waves, it uses the main antenna and a secondary antenna 60 meters away from it to receive the surface echo at the same time. Two SAR images acquired at the same time with a fixed baseline form interference, which realizes the accurate acquisition of the global digital elevation model. Now, SRTMDEM with 90-meter grid spacing is freely available and has been widely used in remote sensing and related fields. The DEM obtained by using InSAR technology reflects the continuous elevation change model of the ground surface, while the height of artificial buildings/structures changes randomly and has spatial discontinuity, so it is difficult to obtain their height information by using traditional InSAR technology.

Contents of the invention

The purpose of the present invention is to address the above problems. The present invention proposes a method and device for extracting the height of artificial buildings/structures on the ground. This technical solution overcomes the shortcomings of traditional techniques and can accurately obtain the height of random changes in artificial buildings/structures. .

In order to achieve the above object, the present invention proposes a method for extracting the height of artificial buildings/structures on the ground, the method comprising:

Combining interferometric image pairs according to the small baseline principle to process synthetic aperture radar images covering the same area to generate time series interferograms;

Screening out candidate high-coherence points according to the amplitude information of the synthetic aperture radar image;

Processing the time series interferogram to obtain a time series differential interferogram, and extracting the differential interferometric phase of the candidate high coherence point;

Filter out the spatially correlated phase in the differential interferometric phase of the candidate high coherence point, and obtain the spatially uncorrelated phase in the differential interferometric phase of the candidate high coherence point; wherein, the spatially uncorrelated phase includes a spatially uncorrelated DEM correction amount and residual phase;

Processing the spatially non-correlated phase in the differential interferometric phase of the candidate high-coherence points, obtaining the correction value of the spatially non-correlated DEM, and selecting the final high-coherence point from the candidate high-coherence points;

The height of the artificial building/structure on the ground is obtained according to the final high coherence point.

Optionally, in an embodiment of the present invention, the step of screening candidate high-coherence points according to the magnitude information of the synthetic aperture radar image includes:

Obtain the amplitude dispersion of the synthetic aperture radar image according to the mean value and standard deviation of the amplitude of the synthetic aperture radar image;

Acquiring candidate high coherence points in the synthetic aperture radar image according to the amplitude dispersion of the synthetic aperture radar image; wherein, the amplitude dispersion of a pixel in the synthetic aperture radar image is greater than a set threshold. Then a pixel in the SAR image is a candidate high coherence point.

Optionally, in an embodiment of the present invention, the step of processing the time-series interferogram to obtain a time-series differential interferogram includes:

obtaining a digital elevation model of the same area covered by the synthetic aperture radar imagery;

According to the digital elevation model, the interferometric phase of the time-series interferogram is subtracted from the interferometric phase caused by the topography to obtain a time-series differential interferogram.

Optionally, in an embodiment of the present invention, the differential interferometric phases of the candidate high coherence points include interferometric phases caused by surface deformation, interferometric phases caused by atmospheric inhomogeneity, error phases caused by inaccurate orbital data, spatial non-correlation Phase and decoherence noise caused by DEM corrections; among them, the interference phase caused by surface deformation, the interference phase caused by atmospheric inhomogeneity, the error phase caused by inaccurate orbital data, and the phase and loss caused by spatially non-correlated DEM corrections Coherent noise includes spatially correlated phase and spatially uncorrelated phase.

Optionally, in an embodiment of the present invention, the step of filtering out the spatial correlation phase in the differential interferometric phase of the candidate high coherence point further includes:

According to the respective characteristics of the interferometric phase caused by the surface deformation, the interferometric phase caused by atmospheric inhomogeneity, the error phase caused by the imprecise orbital data, the phase caused by the spatial non-correlated DEM correction and the decoherence noise, combined with the synthetic aperture radar Adaptive bandpass filtering is performed on the resolution of the image to filter out the spatial correlation phase in the differential interferometric phase of the candidate high coherence points.

Optionally, in an embodiment of the present invention, the step of processing the spatially uncorrelated phase in the differential interferometric phase of the candidate high coherence point includes:

Obtain a temporal coherence factor according to the interferometric phases of the candidate high coherence points in the N time-series interferograms, the estimation of the spatial correlation items of the differential interferometric phases of the candidate high coherence points, and the residual phase;

Determine whether the pixel is the final high-coherence point according to the temporal coherence factor; wherein, when the temporal coherence factor is greater than a threshold, the pixel is the final high-coherence point;

The spatial non-correlation DEM correction value of the final high coherence point is obtained through a space search method.

Optionally, in an embodiment of the present invention, the step of obtaining the height of ground artificial buildings/structures according to the final high coherence point includes:

Using geocoding to obtain the plane position information of the building/structure, and then combining the synthetic aperture radar image to determine the spatial position of the building/structure;

According to the spatial topological relationship, the high coherence points of the building/structure are classified, and the high coherence points on the top of the building/structure are identified, and the spatially non-correlated DEM correction amount of the high coherence point on the top of the building/structure is is the height of the building/structure.

In order to achieve the above object, the present invention also proposes a device for extracting the height of artificial buildings/structures on the ground, which device includes:

The time series interferogram acquisition unit is used to combine the interference image pairs to process the synthetic aperture radar images covering the same area according to the small baseline principle to generate a time series interferogram;

A candidate high-coherence point acquisition unit, configured to filter out candidate high-coherence points according to the amplitude information of the synthetic aperture radar image;

A time-series interferogram processing unit, configured to process the time-series interferogram to obtain a time-series difference interferogram, and extract the interferometric phase of the candidate high-coherence point;

A candidate high-coherence point processing unit, configured to filter out a spatially correlated phase in the differential interferometric phase of the candidate high-coherence point, and obtain a spatially uncorrelated phase in the differential interferometric phase of the candidate high-coherence point; wherein, the spatial Non-correlated phase includes spatially non-correlated DEM corrections and residual phase;

The final high-coherence point determination unit is used to process the spatially non-correlated phase in the differential interferometric phase of the candidate high-coherence points, obtain and solve the spatially non-correlated DEM correction value, and select the final high-coherence point from the candidate High coherence points;

The ground artificial building/structure height acquiring unit is configured to acquire the ground artificial building/structure height according to the final high coherence point.

Optionally, in an embodiment of the present invention, the candidate high coherence point acquisition unit includes:

The amplitude dispersion acquisition module is used to obtain the amplitude dispersion of the synthetic aperture radar image according to the mean value and standard deviation of the amplitude of the synthetic aperture radar image;

The module for judging candidate high coherence points is used to obtain candidate high coherence points in the synthetic aperture radar image according to the amplitude dispersion of the synthetic aperture radar image; wherein, the amplitude dispersion of a pixel in the synthetic aperture radar image is greater than a set Set the threshold. Then this point is a candidate high coherence point.

Optionally, in an embodiment of the present invention, the time series interferogram processing unit includes:

Establishing a digital elevation model module for obtaining a digital elevation model of the same area covered by the synthetic aperture radar image;

The module for obtaining a time-series differential interferogram is used to obtain a time-series differential interferogram by subtracting an interference phase caused by topography from the interferometric phase of the time-series interferogram according to the digital elevation model.

Optionally, in an embodiment of the present invention, the differential interferometric phases of the candidate high-coherence points acquired by the time series interferogram processing unit include interferometric phases caused by surface deformation, interferometric phases caused by atmospheric inhomogeneity, and inaccurate orbital data. The phase error caused by DEM correction and decoherence noise; among them, the interference phase caused by surface deformation, the interference phase caused by atmospheric inhomogeneity, the error phase caused by inaccurate orbital data, and the phase caused by DEM correction and decoherent noise both include spatially correlated phase and spatially uncorrelated phase.

Optionally, in an embodiment of the present invention, the candidate high coherence point processing unit is further configured to use the interferometric phase caused by the surface deformation, the interferometric phase caused by atmospheric inhomogeneity, the error phase caused by inaccurate orbit data, The respective characteristics of the phase and decoherence noise caused by the DEM correction amount are combined with the resolution of the synthetic aperture radar image to perform adaptive bandpass filtering to filter out the spatial correlation phase in the differential interferometric phase of the candidate high coherence point.

Optionally, in an embodiment of the present invention, the final high coherence point determination unit includes:

The temporal coherence factor acquisition module is used to obtain the temporal coherence according to the interference phase of the candidate high coherence point in the N time series interferograms, the estimation of the differential interferometric phase space correlation item of the candidate high coherence point and the residual phase factor;

A final high-coherence point determination module, configured to determine whether the candidate high-coherence point is the final high-coherence point according to the temporal coherence factor; wherein, when the temporal coherence factor is greater than a threshold, the candidate high-coherence point The coherent point is the final high coherent point;

A space search module, configured to obtain the spatially uncorrelated DEM correction value of the final highly coherent point through a space search method.

Optionally, in an embodiment of the present invention, the acquisition unit for the height of artificial buildings/structures on the ground includes:

A building/structure spatial location information acquisition module, used to obtain the building/structure plane location information by using geocoding, and then combine the synthetic aperture radar image to determine the building/structure spatial location;

The height acquisition module of the building/structure is used to classify the high coherence points of the building/structure according to the spatial topology relationship, identify the high coherence point on the top of the building/structure, and the height of the top of the building/structure The spatially uncorrelated DEM correction amount of the coherent point is the height of the building/structure.

The above technical scheme has the following beneficial effects:

(1) This technical solution can also obtain its historical deformation information while extracting the height of ground objects;

(2) This technical solution can extract the height information of some ground objects with small projection area (such as street lamps, electric towers, etc.), which are often missed by optical stereo photogrammetry or LiDAR technology due to their small projection area these features;

(3) This technical solution is not affected by the weather, and has the advantages of all-day, all-weather, and wide coverage;

(4) This technical solution can obtain the digital surface model (Digital Surface Model, DSM) of the area after accurately obtaining the height information of the artificial building/structure.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

Fig. 1 is a kind of flow chart of the method for extracting ground artificial building/structure height that the present invention proposes;

Fig. 2 is a kind of device structural drawing that extracts the height of artificial building/structure on the ground that the present invention proposes;

Fig. 3 is a structure diagram of a candidate high coherence point acquisition unit in a device for extracting the height of ground artificial buildings/structures proposed by the present invention;

Fig. 4 is a structural diagram of a time-series interferogram processing unit in a device for extracting the height of ground artificial buildings/structures proposed by the present invention;

Fig. 5 is a final high coherence point determination unit structure diagram in a device for extracting the height of ground artificial buildings/structures proposed by the present invention;

Fig. 6 is a structure diagram of the acquisition unit for the height of ground artificial buildings/structures in a device for extracting the height of artificial buildings/structures proposed by the present invention;

Fig. 7 is the flow chart of the embodiment of the technical solution of the present invention;

FIG. 8 is a high-coherence point target extracted through time series InSAR analysis according to an embodiment of the present invention;

FIG. 9 is a height map of all highly coherent point targets extracted based on time series InSAR analysis according to an embodiment of the present invention;

Fig. 10 is the inversion result of the building height near the Beijing-Tianjin line highway passing the Yongding Xinhe Bridge in the region obtained by the embodiment of the present invention;

Fig. 11 is the inversion result of the height of buildings in the Ruijingjiayuan community along the Tianjin Metro Line 1 in the region obtained by the embodiment of the present invention;

Fig. 12 is the distribution of 9 local regions selected for accuracy verification in the region according to the embodiment of the present invention;

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Apparently, the described embodiments are only some of the embodiments of the present invention, but not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

In the 1980s, researchers discovered that the differential synthetic aperture radar interferometry (Differential Synthetic Aperture Radar Interferometry, DInSAR) of repeated orbits can be used, that is, two SAR images of the same area acquired at different times form interference, and then from the interference The phase generated by the terrain is subtracted from the phase to obtain deformation information about the surface. In order to overcome the limitation that DInSAR technology is susceptible to factors such as time decoherence, space decoherence, and atmospheric interference, researchers have developed time series InSAR technology, which analyzes point targets that can maintain stable scattering characteristics in time series, In this way, some limiting factors of DInSAR technology are overcome. Afterwards, based on the idea of stable scatterers, researchers successively proposed a variety of time series InSAR technical analysis methods, such as: small baseline subsets (SBAS), coherent target (CT) and based on nonlinear DInSAR analysis method for long time series of deformation models. Since urban areas have dense natural point targets, this technology has been gradually applied in engineering applications in urban settlement monitoring.

The time series InSAR technology can not only be applied to the deformation monitoring of the target, but also accurately estimate the height of the target according to the phase of the separated terrain error. Most of the point targets in urban areas are distributed on the tops and corners of buildings, bridges, street lights, transmission towers, communication towers, etc. Terrain errors on these point targets contain both spatially correlated and spatially uncorrelated terms. Among them, the space-related items mainly reflect the systematic error of the digital elevation model (DEM) used, while the space-independent items mainly represent the height of the point target itself relative to the ground, such as the height of buildings. Therefore, in addition to detecting the deformation information of ground targets, time series InSAR technology can also be used to study the height information of ground targets. Moreover, the spatial resolution of synthetic aperture radar satellites is getting higher and higher, which makes the SAR images more detailed in describing the ground objects, and also greatly increases the density of the extracted point targets. Therefore, the height information of artificial buildings/structures can be extracted accurately by using the time series InSAR technology based on high-resolution SAR images.

As shown in FIG. 1 , it is one of the flow charts of a method for extracting the height of ground artificial buildings/structures proposed by the present invention. The method includes:

Step 101): Combining the interferometric image pairs according to the small baseline principle to process the synthetic aperture radar images covering the same area to generate a time series interferogram;

Step 102): Screening out candidate high-coherence points according to the magnitude information of the SAR image;

Step 103): Process the time series interferogram to obtain a time series differential interferogram, and extract the differential interferometric phase of the candidate high coherence point;

Step 104): Filter out the spatially correlated phase in the differential interferometric phase of the candidate high coherence point, and obtain the spatially uncorrelated phase in the differential interferometric phase of the candidate high coherence point; wherein, the spatially uncorrelated phase includes spatial Uncorrelated DEM corrections and residual phases;

Step 105): Processing the spatially non-correlated phase in the differential interferometric phase of the candidate high-coherence points, obtaining the correction value of the spatially non-correlated DEM, and selecting the final high-coherence point from the candidate high-coherence points;

Step 106): Acquiring the height of the artificial building/structure on the ground according to the final high coherence point.

The above step 101 and step 102 can be arranged in an indeterminate order, and can still solve the problems existing in the prior art.

Preferably, said step 102 further includes:

Step 1021): According to the mean value and standard deviation of the amplitude of the synthetic aperture radar image, obtain the amplitude dispersion of the synthetic aperture radar image;

Step 1022): Obtain candidate high-coherence points in the SAR image according to the amplitude dispersion of the SAR image; wherein, the amplitude dispersion of a pixel in the SAR image is greater than a set threshold. Then a pixel in the SAR image is a candidate high coherence point.

Furthermore, the step 103 further includes:

Step 1031): Obtaining a digital elevation model of the same area covered by the SAR image;

Step 1032): According to the digital elevation model, the interferometric phase of the time-series interferogram is subtracted from the interferometric phase caused by the topography to obtain a time-series differential interferogram.

Specifically, the differential interferometric phases of candidate high-coherence points in the time series interferogram include interferometric phases caused by surface deformation, interferometric phases caused by atmospheric inhomogeneity, error phases caused by inaccurate orbital data, phases caused by DEM corrections, and Decoherence noise; wherein, the interferometric phase caused by surface deformation, the interferometric phase caused by atmospheric inhomogeneity, the error phase caused by inaccurate orbital data, the phase caused by DEM correction and the decoherent noise all include spatially correlated phase and spatially incoherent relative phase.

Optionally, the step 104 further includes: according to the interference phase caused by the surface deformation, the interference phase caused by atmospheric inhomogeneity, the error phase caused by the inaccurate orbit data, the phase caused by the DEM correction amount and the decoherence noise respectively characteristics, combined with the resolution of the synthetic aperture radar image to perform adaptive band-pass filtering to filter out the spatial correlation phase in the differential interferometric phase of the candidate high coherence point.

Preferably, the step 105) includes:

Step 1051): Obtain the temporal coherence factor according to the interferometric phases of the candidate high coherence points in the N time series interferograms, the estimation of the spatial correlation items of the differential interferometric phases of the candidate high coherence points, and the residual phase;

Step 1052): Determine whether the pixel is the final high-coherence point according to the temporal coherence factor; wherein, when the temporal coherence factor is greater than a threshold, the pixel is the final high-coherence point;

Step 1053): Obtain the spatially uncorrelated DEM correction value of the final highly coherent point through a space search method.

Preferably, the step 106 specifically includes:

Step 1061): Using geocoding to obtain the plane position information of the building/structure, and then combining the synthetic aperture radar image to determine the spatial position of the building/structure;

Step 1062): Classify the high coherence points of the building/structure according to the spatial topological relationship, identify the high coherence points on the top of the building/structure, and the spatial non-correlation of the high coherence points on the top of the building/structure The DEM correction amount is the height of the building/structure.

As shown in FIG. 2 , it is a structural diagram of a device for extracting the height of ground artificial buildings/structures proposed by the present invention. The unit includes:

The time series interferogram acquisition unit 201 is used to combine the interference image pairs according to the small baseline principle to process the synthetic aperture radar images covering the same area to generate a time series interferogram;

A candidate high coherence point acquisition unit 202, configured to screen out candidate high coherence points according to the magnitude information of the synthetic aperture radar image;

A time-series interferogram processing unit 203, configured to process the time-series interferogram to obtain a time-series differential interferogram, and extract the differential interferometric phase of the candidate high coherence point;

The candidate high coherence point processing unit 204 is configured to filter out the spatially correlated phase in the differential interferometric phase of the candidate high coherence point, and obtain the spatially uncorrelated phase in the differential interferometric phase of the candidate high coherent point; wherein, the The spatially uncorrelated phase includes the spatially uncorrelated DEM correction and residual phase;

The final high-coherence point determination unit 205 is configured to process the spatially non-correlated phase in the differential interferometric phase of the candidate high-coherence points, obtain and solve the spatially non-correlated DEM correction value, and select from the candidate high-coherence points final high coherence point;

The ground artificial building/structure height obtaining unit 206 is configured to obtain the ground artificial building/structure height according to the final high coherence point.

As shown in FIG. 3 , it is a structure diagram of a candidate high coherence point acquisition unit in a device for extracting the height of ground artificial buildings/structures proposed by the present invention. The candidate high coherence point acquisition unit 202 includes:

The amplitude dispersion acquisition module 2021 is used to acquire the amplitude dispersion of the synthetic aperture radar image according to the mean value and standard deviation of the amplitude of the synthetic aperture radar image;

The judging candidate high coherence point module 2022 is used to obtain candidate high coherence points in the synthetic aperture radar image according to the amplitude dispersion of the synthetic aperture radar image; wherein, the amplitude dispersion of one pixel in the synthetic aperture radar image is greater than one Set the threshold. Then this point is a candidate high coherence point.

As shown in FIG. 4 , it is a structural diagram of a time-series interferogram processing unit in a device for extracting the height of artificial buildings/structures on the ground proposed by the present invention. The time series interferogram processing unit 203 includes:

Establish a digital elevation model module 2031 for obtaining the digital elevation model of the same area covered by the synthetic aperture radar image;

Obtaining a time-series differential interferogram module 2032, configured to obtain a time-series differential interferogram by subtracting an interference phase caused by topography from the interferometric phase of the time-series interferogram according to the digital elevation model.

The differential interferometric phases of candidate high-coherence points acquired by the time series interferogram processing unit 203 include interferometric phases caused by surface deformation, interferometric phases caused by atmospheric inhomogeneity, error phases caused by inaccurate orbital data, and phases caused by DEM corrections. and decoherence noise; wherein, the interferometric phase caused by surface deformation, the interferometric phase caused by atmospheric inhomogeneity, the error phase caused by inaccurate orbital data, the phase caused by DEM correction and the decoherent noise all include spatially correlated phase and space non-correlated phase.

Furthermore, the candidate high coherence point processing unit 204 is further used to calculate the interferometric phase caused by the surface deformation, the interferometric phase caused by atmospheric inhomogeneity, the error phase caused by inaccurate orbital data, and the phase caused by the DEM correction amount. The respective characteristics of the decoherence noise are combined with the resolution of the synthetic aperture radar image to perform adaptive bandpass filtering to filter out the spatial correlation phase in the differential interferometric phase of the candidate high coherence point.

As shown in FIG. 5 , it is a final high-coherence point determination unit structure diagram in a device for extracting the height of ground artificial buildings/structures proposed by the present invention. The final high coherence point determination unit 205 includes:

The temporal coherence factor acquisition module 2051 is used to obtain the temporal state according to the interferometric phases of the candidate high coherence points in the N time series interferograms, the estimation of the differential interferometric phase spatial correlation items of the candidate high coherence points, and the residual phase coherence factor;

The final high coherence point determination module 2052 is configured to determine whether the candidate high coherence point is the final high coherence point according to the temporal coherence factor; wherein, when the temporal coherence factor is greater than a threshold, the candidate The high coherence point is the final high coherence point;

The spatial search module 2053 is configured to obtain the spatially uncorrelated DEM correction value of the final highly coherent point through a spatial search method.

As shown in FIG. 6 , it is a structure diagram of the acquisition unit for the height of ground artificial buildings/structures in a device for extracting the height of artificial buildings/structures on the ground proposed by the present invention. The ground artificial building/structure height acquisition unit 206 includes:

The building/structure spatial location information acquisition module 2061 is used to obtain the building/structure plane location information by using geocoding, and then combine the synthetic aperture radar image to determine the building/structure spatial location;

The building/structure height acquisition module 2062 is used to classify the high coherence points of the building/structure according to the spatial topology relationship, identify the high coherence point on the top of the building/structure, and the top of the building/structure The spatially uncorrelated DEM correction amount of the highly coherent point is the height of the building/structure.

Example:

As shown in Fig. 7, it is a flowchart of an embodiment of the technical solution of the present invention. The technical solution steps include:

Step A: Combine the interference image pairs according to the small baseline principle to generate a differential interferogram.

可表示为：Acquire more than 10 scenes of high-resolution SAR images covering the same area, according to the principle of small baseline interference pair selection, that is, the spatial baseline of the two SAR images is less than the set threshold, and the time interval of acquisition is less than the set threshold. All the images of the two scenes satisfying these conditions are formed into an interferogram. Then use the existing DEM (usually you can use the free SRTM DEM) to remove the topographic phase in each interferogram to obtain a differential interferogram, then the differential interferometric phase of the xth pixel on the ith differential interferogram

Can be expressed as:

In the formula, φ _D,x,i is the interference phase caused by surface deformation; φ _A,x,i is the interference phase caused by atmospheric inhomogeneity; Δφ _S,x,i is the error phase caused by inaccurate orbit data; Δφ _{θ ,x,i} is the phase caused by the DEM error (that is, the DEM correction Δh _x ); φ _N,x,i is the decoherence noise.

与空间非相关项

空间相关项反映DEM的系统误差

空间非相关相位

与空间非相关DEM改正量

线性相关，反映的是像素x相对于参考DEM的高度。因此，公式（1）可改写为：Within a certain spatial distance, φ _D,x,i , φ _A,x,i and Δφ _S,x,i have strong spatial correlation characteristics, and φ _N,x,i behaves as random noise. Δφ _θ,x,i caused by the DEM correction Δh _x can be divided into spatially dependent terms

unrelated to space

Spatial related items Reflect the systematic error of DEM

spatially uncorrelated phase

Corrections for space-independent DEM

Linear correlation, reflecting the height of pixel x relative to the reference DEM. Therefore, formula (1) can be rewritten as:

表示的空间相关项。即差分干涉图中每个像素的差分干涉相位均为空间相关相位和空间非相关相位组成。In the formula,

express space-related items. That is, the differential interferometric phase of each pixel in the differential interferogram is composed of a spatially correlated phase and a spatially uncorrelated phase.

Step B: Screen highly correlated candidate points based on amplitude information.

The algorithm based on the amplitude information of high-resolution SAR images is fast and simple, so by setting a loose threshold, as many high-correlation candidate points as possible can be selected. These highly correlated candidate points can be refined using phase information in subsequent steps. The amplitude dispersion D _A of high resolution SAR image is defined as:

D _A ≈σ _A /μ _A (3)

In the formula, σ _A and μ _A represent the mean and standard deviation of the amplitude of the high-resolution SAR image, respectively.

Step C: Separating the spatially correlated phases.

According to the characteristics of each phase in formula (1), the size of the spatial grid is set in combination with the resolution of the SAR image, and in each grid, an adaptive bandpass filter is used to filter out the phases in formula (1) according to formula (4). spatial correlation phase

表示带通滤波得到的空间相关项

的估值，空间相关项残余相位δ_x,i很小，可以与φ_N,x,i并入为残差相位Δφ_res,x,i。并且

与空间非相关DEM改正量

的关系记为：In the formula,

Indicates the spatial correlation term obtained by bandpass filtering

The estimation of the spatial correlation term residual phase δ _x,i is very small, and can be combined with φ _N,x,i as the residual phase Δφ _res,x,i . and

Corrections for space-independent DEM

The relationship is recorded as:

$\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\mathrm{<span\; class="notranslate">\; \&phi;</span>}}_{\mathrm{<span\; class="notranslate">\; \&theta;</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}^{\mathrm{<span\; class="notranslate">\; u</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; B</span>}}_{<span\; class="notranslate">\; \&perp;</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; x</span>}}^{\mathrm{<span\; class="notranslate">\; u</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$

In the formula, the coefficient a is determined by the slant distance, incident angle and radar wavelength of the point, and is a known quantity. B _⊥,x,i is the vertical baseline distance of the xth pixel in the i-th interferogram.

Step D: Use the space search method to solve the correction value of the spatially non-correlated DEM, and determine the final high-coherence point.

For high-coherence point targets, it needs to maintain phase stability in time series, that is, the absolute value of the residual phase Δφ _res,x,i needs not to exceed π, while maintaining stability in the time series interferogram. Therefore, taking pixel x as an example, pixel x is a candidate high coherence point, and the final high coherence point target can be identified and extracted by analyzing the residual phase of pixel x. The temporal coherence factor γ _x is defined as an index for pixel x to maintain phase stability on N time series interferograms, and the temporal coherence factor γ _x is obtained according to formula (6);

与空间非相关DEM改正量

线性相关，因此可以利用空间搜索法估计出

同时使得时态相干因子γ_x取得最大值。当γ_x大于设定的阈值时，即可认为像素x为最终的高相干点目标，同时也获得了该高相干点目标对应的DEM空间非相关改正量

的最优解。Since the residual phase Δφ _res,x,i is mainly dominated by noise, γ _x can reflect the degree of noise pollution of the pixel x, and the larger the γ _x , the higher the quality of the pixel, so γ _x can be defined as the judgment pixel x Criteria for whether the target is a highly coherent point. Due to the winding of the phase, the formula (6) is an indeterminate equation, and there is no definite solution. But when |Δφ _res,x,i |<π, the space search method can be used to solve it. consider at the same time

Corrections for space-independent DEM

linear correlation, so the spatial search method can be used to estimate

At the same time, the temporal coherence factor γ _x is maximized. When γ _x is greater than the set threshold, the pixel x can be considered as the final high-coherence point target, and the DEM spatial non-correlation correction corresponding to the high-coherence point target is also obtained

the optimal solution of .

Step E: Classify the spatial position of the highly coherent point target, and extract the height of the building.

反映的即是这些高相干点目标相对于参考DEM的高度。另外，得益于高分辨率影像数据使得高相干点目标密度大大增加，建/构筑物上细节结构可以被充分识别出来。通过时间序列InSAR技术得到各高相干点目标的空间非相关DEM改正量

后，利用地理编码获取其平面位置信息，再结合高分辨率光学影像确定空间位置。根据空间拓扑关系对高相干点目标进行分类，识别出位于建筑顶部的高相干点目标，该处高相干点目标对应的空间非相关DEM改正量

即为人工建/构筑物的高度提取。In urban areas, high-coherence point targets are mainly distributed on man-made buildings/structures. The time series InSAR technology is the most commonly used external DEM data such as SRTM DEM, ASTER GDEM, etc. to remove terrain. Most of them cannot reflect the height of buildings/structures in urban areas, and the elevation of a high-coherence point target on the reference DEM is not the point target. actual elevation. In fact, the spatially uncorrelated DEM corrections

What is reflected is the height of these highly coherent point targets relative to the reference DEM. In addition, thanks to the high-resolution image data, the density of high-coherence point targets is greatly increased, and the detailed structures on buildings/structures can be fully identified. The spatially uncorrelated DEM corrections of each highly coherent point target are obtained by time series InSAR technology

After that, geocoding is used to obtain its planar location information, and then combined with high-resolution optical images to determine its spatial location. Classify high-coherence point targets according to the spatial topological relationship, identify high-coherence point targets located on the top of the building, and the spatially non-correlated DEM corrections corresponding to the high-coherence point targets

It is the height extraction of artificial buildings/structures.

In order to better illustrate the effectiveness and superiority of the technical solution of the present invention, the following comparative analysis of the embodiment of the present invention after applying the above-mentioned technical solution is as follows: The high-coherence point targets in Beichen District in the northwest of the city are actually 6.6km long and 8.5km wide, covering an area of about 60km ² . Finally, 446,176 high-coherence points were detected with a density of about 7,500 high-coherence points/km ² . As shown in FIG. 9 , it is a height map of all highly coherent point targets extracted based on time series InSAR analysis in the embodiment of the present invention, and the unit is m. As shown in FIG. 10 , it is a local area obtained by the embodiment of the present invention, and the height inversion results of some buildings near the Yongding Xinhe Bridge on the Beijing-Tianjin Line are displayed. As shown in FIG. 11 , it is a display of the height inversion results of some buildings near Ruijingjiayuan Community along Tianjin Metro Line 1 in a local area obtained by the embodiment of the present invention. As shown in FIG. 12 , it is a schematic diagram of the distribution of nine local areas selected for accuracy verification in the embodiment area of the embodiment of the present invention. In order to verify the accuracy of the inversion results of the embodiment of the invention, the height data of 22 ground objects were selected as the true value to verify the accuracy of the results of the InSAR inversion of building heights in the region of the embodiment by using the mirror-free total station, as shown in Table 1. As shown, the error standard deviation of the inversion results is 2.1m, and the mean is 0.8m. The results show that the accuracy of the building height inversion result of the embodiment of the present invention is relatively high.

Those of ordinary skill in the art can understand that all or part of the steps in the methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the program can be executed when executed , including all or part of the steps above, the storage medium, such as: ROM/RAM, magnetic disk, optical disk, etc.

Table 1 Unit/m

The specific embodiments described above have further described the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above descriptions are only specific embodiments of the present invention and are not intended to limit the scope of the present invention. Protection scope, within the spirit and principles of the present invention, any modification, equivalent replacement, improvement, etc., shall be included in the protection scope of the present invention.

Claims (14)

1. A method for extracting the height of artificial buildings/structures on the ground, characterized in that the method comprises: Combining interferometric image pairs according to the small baseline principle to process synthetic aperture radar images covering the same area to generate time series interferograms; Screening out candidate high-coherence points according to the amplitude information of the synthetic aperture radar image; Processing the time series interferogram to obtain a time series differential interferogram, and extracting the differential interferometric phase of the candidate high coherence point; Filter out the spatially correlated phase in the differential interferometric phase of the candidate high coherence point, and obtain the spatially uncorrelated phase in the differential interferometric phase of the candidate high coherence point; wherein, the spatially uncorrelated phase includes a spatially uncorrelated DEM correction amount and residual phase; Processing the spatially non-correlated phase in the differential interferometric phase of the candidate high-coherence points, obtaining the correction value of the spatially non-correlated DEM, and selecting the final high-coherence point from the candidate high-coherence points; The height of the artificial building/structure on the ground is obtained according to the final high coherence point. 2. The method according to claim 1, wherein the step of screening out candidate high coherence points according to the amplitude information of the synthetic aperture radar image comprises: Obtain the amplitude dispersion of the synthetic aperture radar image according to the mean value and standard deviation of the amplitude of the synthetic aperture radar image; Acquiring candidate high coherence points in the synthetic aperture radar image according to the amplitude dispersion of the synthetic aperture radar image; wherein, the amplitude dispersion of a pixel in the synthetic aperture radar image is greater than a set threshold. Then a pixel in the SAR image is a candidate high coherence point. 3. The method according to claim 1, wherein the step of processing the time-series interferogram to obtain a time-series differential interferogram comprises: obtaining a digital elevation model of the same area covered by the synthetic aperture radar imagery; According to the digital elevation model, the interferometric phase of the time-series interferogram is subtracted from the interferometric phase caused by the topography to obtain a time-series differential interferogram. 4. The method according to claim 1, wherein the differential interferometric phases of the candidate high coherence points include interferometric phases caused by surface deformation, interferometric phases caused by atmospheric inhomogeneity, error phases caused by inaccurate orbital data, Phase and decoherence noise caused by DEM corrections; among them, the interference phase caused by surface deformation, the interference phase caused by atmospheric inhomogeneity, the error phase caused by inaccurate orbital data, and the phase and loss caused by spatially non-correlated DEM corrections Coherent noise includes spatially correlated phase and spatially uncorrelated phase. 5. The method according to claim 1, wherein the step of filtering out the spatial correlation phase in the differential interferometric phase of the candidate high coherence point further comprises: According to the respective characteristics of the interference phase caused by the surface deformation, the interference phase caused by the uneven atmosphere, the error phase caused by the inaccurate orbit data, the phase caused by the DEM correction amount and the decoherence noise, combined with the resolution of the synthetic aperture radar image Adaptive bandpass filtering is performed at the rate to filter out the spatial correlation phase in the differential interferometric phase of the candidate high coherence point. 6. The method according to claim 1, wherein the step of processing the spatially uncorrelated phase in the differential interferometric phase of the candidate high coherence point comprises: Obtain a temporal coherence factor according to the interferometric phases of the candidate high coherence points in the N time-series interferograms, the estimation of the spatial correlation items of the differential interferometric phases of the candidate high coherence points, and the residual phase; Determine whether the pixel is the final high-coherence point according to the temporal coherence factor; wherein, when the temporal coherence factor is greater than a threshold, the pixel is the final high-coherence point; The spatial non-correlation DEM correction value of the final high coherence point is obtained through a space search method. 7. The method according to claim 1, wherein the step of obtaining the height of ground man-made buildings/structures according to the final high coherence point comprises: Using geocoding to obtain the plane position information of the building/structure, and then combining the synthetic aperture radar image to determine the spatial position of the building/structure; According to the spatial topological relationship, the high coherence points of the building/structure are classified, and the high coherence points on the top of the building/structure are identified, and the spatially non-correlated DEM correction amount of the high coherence point on the top of the building/structure is is the height of the building/structure. 8. A device for extracting the height of artificial buildings/structures on the ground, characterized in that the device comprises: The time series interferogram acquisition unit is used to combine the interference image pairs to process the synthetic aperture radar images covering the same area according to the small baseline principle to generate a time series interferogram; A candidate high-coherence point acquisition unit, configured to filter out candidate high-coherence points according to the amplitude information of the synthetic aperture radar image; A time-series interferogram processing unit, configured to process the time-series interferogram to obtain a time-series difference interferogram, and extract the interferometric phase of the candidate high-coherence point; A candidate high-coherence point processing unit, configured to filter out a spatially correlated phase in the differential interferometric phase of the candidate high-coherence point, and obtain a spatially uncorrelated phase in the differential interferometric phase of the candidate high-coherence point; wherein, the spatial Non-correlated phase includes spatially non-correlated DEM corrections and residual phase; The final high-coherence point determination unit is used to process the spatially non-correlated phase in the differential interferometric phase of the candidate high-coherence points, obtain and solve the spatially non-correlated DEM correction value, and select the final high-coherence point from the candidate High coherence point; The ground artificial building/structure height acquiring unit is configured to acquire the ground artificial building/structure height according to the final high coherence point. 9. The device according to claim 8, wherein the candidate high coherence point acquiring unit comprises: The amplitude dispersion acquisition module is used to obtain the amplitude dispersion of the synthetic aperture radar image according to the mean value and standard deviation of the amplitude of the synthetic aperture radar image; The module for judging candidate high coherence points is used to obtain candidate high coherence points in the synthetic aperture radar image according to the amplitude dispersion of the synthetic aperture radar image; wherein, the amplitude dispersion of a pixel in the synthetic aperture radar image is greater than a set Set the threshold. Then this point is a candidate high coherence point. 10. The device according to claim 8, wherein the time series interferogram processing unit comprises: Establishing a digital elevation model module for obtaining a digital elevation model of the same area covered by the synthetic aperture radar image; The module for obtaining a time-series differential interferogram is used to obtain a time-series differential interferogram by subtracting an interference phase caused by topography from the interferometric phase of the time-series interferogram according to the digital elevation model. 11. The device according to claim 8, wherein the differential interferometric phases of the candidate high-coherence points acquired by the time series interferogram processing unit include interferometric phases caused by surface deformation, interferometric phases caused by atmospheric inhomogeneity, orbital Error phase caused by inaccurate data, phase and decoherence noise caused by spatially non-correlated DEM corrections; among them, the interference phase caused by surface deformation, the interference phase caused by atmospheric inhomogeneity, the error phase caused by inaccurate orbital data, Both phase and decoherence noise caused by DEM corrections include spatially correlated phase and spatially noncorrelated phase. 12. The device according to claim 8, wherein the candidate high-coherence point processing unit is further used for the interference phase caused by the surface deformation, the interference phase caused by atmospheric inhomogeneity, and the inaccurate orbit data The respective characteristics of the error phase, the phase caused by the spatial non-correlated DEM correction, and the decoherence noise, combined with the resolution of the synthetic aperture radar image, perform adaptive bandpass filtering, and filter out the differential interferometric phase of the candidate high coherence point The spatial correlation phase of . 13. The device according to claim 8, wherein the final high coherence point determining unit comprises: The temporal coherence factor acquisition module is used to obtain the temporal coherence according to the interference phase of the candidate high coherence point in the N time series interferograms, the estimation of the differential interferometric phase space correlation item of the candidate high coherence point and the residual phase factor; A final high-coherence point determination module, configured to determine whether the candidate high-coherence point is the final high-coherence point according to the temporal coherence factor; wherein, when the temporal coherence factor is greater than a threshold, the candidate high-coherence point The coherent point is the final high coherent point; A space search module, configured to obtain the spatially uncorrelated DEM correction value of the final highly coherent point through a space search method. 14. The device according to claim 8, wherein said ground man-made building/structure height acquiring unit comprises: A building/structure spatial location information acquisition module, used to obtain the building/structure plane location information by using geocoding, and then determine the building/structure spatial location in combination with synthetic aperture radar images; The height acquisition module of the building/structure is used to classify the high coherence points of the building/structure according to the spatial topology relationship, identify the high coherence point on the top of the building/structure, and the height of the top of the building/structure The spatially uncorrelated DEM correction amount of the coherent point is the height of the building/structure.

Images