Underforest terrain extraction method and device based on multi-baseline SAR (Synthetic Aperture Radar) data

Abstract

The invention discloses an under-forest terrain extraction method and device based on multi-baseline SAR (Synthetic Aperture Radar) data, and the method comprises the steps: carrying out the signal decomposition of an SAR image data set, carrying out the tomography of a signal obtained through the decomposition through spectrum reconstruction, and obtaining a forest tomography structure section; projecting a main lobe signal in the structural profile to obtain a reconstructed forest three-dimensional point cloud; performing elevation normalization processing on the forest three-dimensional point cloud by using the coarse digital terrestrial model, extracting an elevation normalized ground point cloud in the forest three-dimensional point cloud, and superposing the elevation normalized ground point cloud with the coarse digital terrestrial model to obtain a ground point cloud; and projecting the ground point cloud to the two-dimensional grid to form an under-forest terrain model. By adopting the technical scheme, the three-dimensional point cloud of the forest is reconstructed based on the section of the forest chromatography structure, then the ground point cloud is extracted, and the under-forest terrain is obtained through the ground point cloud, so that the under-forest terrain extraction precision is obviously improved, and the problems of obvious breakpoints and jumping of the under-forest terrain are solved.

Description

Technical Field

[0001] The invention relates to the field of remote sensing technology, and in particular to a method and a device for extracting understory terrain based on multi-baseline SAR data. Background Art

[0002] Understory topography is a forest vertical structure parameter and the focus of forestry surveys, and plays an important role in indicating forest growth potential and timber production. In terms of remote sensing research, there are currently many forms and methods of understory topography remote sensing. In terms of remote sensing types, there are mainly optical remote sensing, LiDAR and synthetic aperture radar (SAR).

[0003] Tomographic SAR is an advanced method that extracts vertical forest structure parameters by analyzing the height distribution differences of multiple baseline observations, thereby providing vertical resolution and three-dimensional detection within the forest. However, in the actual application of this method, the forest tomographic profile is usually screened by setting a threshold to extract the understory terrain information, and the parts that do not meet the threshold are eliminated, which easily eliminates the information representing the real surface, resulting in discontinuous backscatter power distribution in azimuth and range directions within the forest tomographic profile. The understory terrain obtained based on the forest tomographic profile after threshold screening will have obvious breakpoints and jumps. Summary of the invention

[0004] Purpose of the invention: The present invention provides a method and device for extracting understory terrain based on multi-baseline SAR data, aiming to solve the technical problems existing in the prior art that in practical applications of tomographic SAR, the backscatter power distribution of the forest tomographic profile is discontinuous, and obvious breakpoints and jumps will appear in the understory terrain.

[0005] Technical solution: The present invention provides a method for extracting understory terrain based on multi-baseline SAR data, including: acquiring a multi-baseline data set of a target area obtained by collection, synthesizing a SAR image data set, and correcting the interference phase error in the SAR image data set; performing signal decomposition on the SAR image data set, performing tomographic imaging on the decomposed signal using spectrum reconstruction, and obtaining a forest tomographic structural profile; removing sidelobe signals in the structural profile, and retaining mainlobe signals in the structural profile; obtaining a reconstructed three-dimensional point cloud of the forest by projecting the mainlobe signals in the structural profile; generating a coarse digital ground model that represents the undulation of the terrain, performing elevation normalization processing on the three-dimensional point cloud of the forest using the coarse digital ground model, extracting the elevation normalized ground point cloud in the three-dimensional point cloud of the forest using cloth simulation filtering, superimposing the elevation normalized ground point cloud on the coarse digital ground model, and obtaining a ground point cloud in the three-dimensional point cloud of the forest; and projecting the ground point cloud onto a two-dimensional grid to form an understory terrain model.

[0006] Specifically, a P-band multi-baseline data set of the target area collected by the aircraft is obtained.

[0007] Specifically, the phase error, flat ground phase and terrain phase in the SAR image data set are removed.

[0008] Specifically, SKP signal decomposition is performed on the SAR image data set; Capon spectrum reconstruction is used to suppress SAR echo signal noise and improve the resolution of tomography in the altitude direction.

[0009] Specifically, the size of the sliding window is set, and the maximum power position corresponding to the maximum backscatter coefficient in the sliding window is determined in the structural section; at the maximum power position, the pixels corresponding to the backscatter signals in the upper and lower ranges of the window are extracted through the sliding window; the sidelobe signals are eliminated and the mainlobe signals are extracted pixel by pixel for the structural section in the azimuth or distance direction.

[0010] Specifically, in azimuth or distance, according to the elevation information corresponding to the pixels of the backscattered signal in the structural section, the pixels are projected to the corresponding height to reconstruct the three-dimensional point cloud of the forest.

[0011] Specifically, the forest 3D point cloud is gridded in the horizontal direction, and the lowest point in each grid is used as a reference point to obtain the corresponding coarse digital ground model.

[0012] Specifically, for each target point in the forest 3D point cloud, the elevation difference between the target point elevation value and the elevation value of the corresponding point in the rough digital ground model is calculated, and the elevation difference is used as the elevation value of the target point, thereby obtaining the forest 3D point cloud after elevation normalization.

[0013] Specifically, the elevation information of the ground point cloud is interpolated and then projected onto a two-dimensional grid to form an understory terrain model.

[0014] The present invention also provides an understory terrain extraction device based on multi-baseline SAR data, comprising an image data synthesis unit, a tomography unit, a signal extraction unit, a point cloud reconstruction unit, a ground point cloud extraction unit and an understory terrain model generation unit, wherein: the image data synthesis unit is used to obtain the multi-baseline data set of the target area obtained by acquisition, synthesize the SAR image data set, and correct the interference phase error in the SAR image data set; the tomography unit is used to perform signal decomposition on the SAR image data set, perform tomography on the decomposed signal by using spectrum reconstruction, and obtain the forest tomography. The structure section; the signal extraction unit is used to remove the side lobe signal in the structure section and retain the main lobe signal in the structure section; the point cloud reconstruction unit is used to obtain a reconstructed forest three-dimensional point cloud by projecting the main lobe signal in the structure section; the ground point cloud extraction unit is used to generate a coarse digital ground model that represents the terrain undulation, use the coarse digital ground model to perform elevation normalization on the forest three-dimensional point cloud, use cloth simulation filtering to extract the elevation normalized ground point cloud in the forest three-dimensional point cloud, and superimpose the elevation normalized ground points on the coarse digital ground model to obtain the ground point cloud in the forest three-dimensional point cloud; The understory terrain model generating unit is used to project the ground point cloud onto a two-dimensional grid to form an understory terrain model.

[0015] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages: reconstructing the forest three-dimensional point cloud based on the forest tomography structure profile, then extracting the ground point cloud therein, and obtaining the underforest terrain through the ground point cloud, which significantly improves the accuracy of underforest terrain extraction and overcomes the problem of obvious breakpoints and jumps in the underforest terrain. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of the process of extracting understory terrain based on multi-baseline SAR data provided by the present invention; Figure 2 and Figure 3 They are respectively the RGB false color image and the SLC intensity image in the SAR image data set provided by the present invention; Figure 4 , Figure 5 and Figure 6 The interferometric phases after removing the phase error, removing the flat ground phase and removing the terrain phase of the SAR image data set provided by the present invention respectively; Figure 7 A schematic diagram of a forest tomography structure cross section provided by the present invention, wherein Figure 7 (a) and Figure 7 (b) are schematic cross-sectional views of the forest tomography structure in the azimuth and distance directions provided by the present invention; Figure 8 and Fig. 9Schematic diagrams of the forest tomography structure sections for retaining sidelobe signals and eliminating sidelobe signals provided by the present invention respectively; Fig.10 A schematic diagram of a reconstructed three-dimensional point cloud of a forest provided by the present invention; Fig.11 A schematic diagram of the principle of cloth simulation filtering provided by the present invention; Fig.12 and Fig.13 They are ground points and non-ground points extracted from the forest three-dimensional point cloud by the MCSF algorithm provided by the present invention respectively; Fig.14 The forest understory terrain model is obtained by applying the method provided by the present invention. DETAILED DESCRIPTION

[0017] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0018] See also Figure 1 , which is a flow chart of the method for extracting understory terrain based on multi-baseline SAR data provided by the present invention.

[0019] In the embodiment of the present invention, a multi-baseline data set of the target area is acquired, synthesized to obtain a SAR image data set, and the interference phase error in the SAR image data set is corrected.

[0020] See also Figure 2 and Figure 3 , which are respectively the RGB false color image and the single look complex data SLC (Single Look Comple) intensity image of the SAR image dataset provided by the present invention.

[0021] In the embodiment of the present invention, a P-band multi-baseline data set of a target area collected by an aircraft is acquired.

[0022] In specific implementation, the SAR radar system can be installed on an airplane or a drone, and the radar image of the ground target (target forest area) can be obtained through the movement of the aircraft. During the propagation and reflection process of the radar signal, its phase or amplitude information changes due to various factors, thereby losing the original coherence (the phase relationship between the signals becomes unstable or cannot be maintained consistently). Multi-baseline refers to the use of multiple radar antennas or multiple observation positions (usually through multiple flights or multiple satellites) to obtain multiple images of the same target area, and process them through the interference relationship between these images. Therefore, in SAR image processing, decoherence and interference phase errors will lead to a decrease in image quality and reduce the accuracy of interferometric measurement.

[0023] In specific implementation, the P-band dataset refers to radar remote sensing data collected in the P-band (frequency range is about 300 MHz to 1 GHz, wavelength is about 30 cm to 1 m). P-band radar data has the characteristics of strong penetration ability and can penetrate the vegetation canopy, which is suitable for monitoring forest biomass and understory terrain.

[0024] In the specific implementation, in the embodiment of the present invention, the flight altitude of the carrier platform is about 4000 meters, the flight speed is about 80m / s, the re-orbit flight interval is about 30 minutes, the average spatial baseline is about 60 meters, and a total of 6 images are acquired in this flight mission. Figure 2 The RGB false color image of the main image after Pauli basis decomposition, Figure 3 It is the single look complex data SLC (Single LookComple) intensity map.

[0025] See also Figure 4 , Figure 5 and Figure 6 The interference phases are respectively the phase error, the flat ground phase and the terrain phase removed from the SAR image data set provided by the present invention.

[0026] In the embodiment of the present invention, the phase error, flat ground phase and terrain phase in the SAR image data set are removed.

[0027] In a specific implementation, data preprocessing is performed on the obtained SAR image data set to correct the interference error caused by external factors in the airborne SAR system, so as to obtain a high-precision spatial baseline and a corrected interference phase. In the embodiment of the present invention, the corrected interference phase is shown in the figure, where: Figure 4 is the interference phase after removing the phase error, Figure 5 is the interference phase after removing the flat-earth phase, Figure 6 is the interference phase after removing the terrain phase.

[0028] In the embodiment of the present invention, the SAR image data set is subjected to signal decomposition, and spectrum reconstruction is used to perform tomography on the decomposed signal to obtain a forest tomographic structure profile.

[0029] See also Figure 7 (a) is a schematic diagram of the azimuthal forest tomography structure section provided by the present invention, see Figure 7 (b) is a schematic diagram of the distance-oriented forest tomography structure profile provided by the present invention.

[0030] In the embodiment of the present invention, SKP signal decomposition is performed on the SAR image data set; Capon spectrum reconstruction is used to suppress SAR echo signal noise and improve the resolution of tomography in the height direction.

[0031] In specific implementation, SKP (Kronecker Product Sum) decomposition is an effective signal processing method that can decompose complex scattering mechanisms into multiple simple scattering components. Through SKP decomposition, the scattering characteristics of the forest canopy and the ground can be extracted, thus providing basic data for the reconstruction of the vertical structure of the forest.

[0032] In a specific implementation, in an embodiment of the present invention, the airborne radar system uses an HH polarization channel, which means that the radar system transmits electromagnetic waves in a horizontal direction and receives echo signals in a horizontal direction. The microwave signal of the HH polarization channel is more sensitive to the scattering characteristics of the forest surface and is suitable for monitoring the terrain under the forest.

[0033] In the specific implementation, the spectrum reconstruction method is used to suppress the SAR echo signal noise and improve the resolution of tomography in the vertical direction. For example, the scattering characteristics of the forest are complex, and the SAR echo signal noise is relatively serious; and due to the limitations of data acquisition hardware, environment and other conditions, the spatial resolution of the original SAR signal is difficult to meet the requirements for extracting forest structural parameters. Therefore, selecting a suitable spectrum analysis method is of great significance for extracting the vertical structural parameters of the forest. Figure 7 (a) and Figure 7 (b) (HH polarization channel, cross-section position: 150 in azimuth and 150 in distance).

[0034] In the embodiment of the present invention, the side lobe signal in the structural section is removed, and the main lobe signal in the structural section is retained.

[0035] See also Figure 8 and Fig. 9 , which is a schematic diagram of a forest tomography structure profile for retaining sidelobe signals and eliminating sidelobe signals provided by the present invention.

[0036] In an embodiment of the present invention, a sliding window method is used to extract the main lobe signal, and the process includes: setting the size of the sliding window, determining the maximum power position corresponding to the maximum backscatter coefficient in the sliding window in the structural section; at the maximum power position, extracting the pixels corresponding to the backscatter signal in the upper and lower ranges of the window through the sliding window; and eliminating the sidelobe signal and extracting the main lobe signal on the structural section pixel by pixel in the azimuth or distance direction.

[0037] In practice, the main lobe is the area where the signal energy is most concentrated, while the side lobes are the energy distribution in other directions outside the main lobe. In SAR imaging, these side lobes capture the scattered energy around the target and form stray signals in the image. In the vertical structure profile of the forest, the side lobe signal may mask the weak scattered signals inside the forest (such as small branches, vegetation canopy, etc.), thereby affecting the accurate analysis of the forest structure. For example, the side lobes of strong scattering points (such as tree trunks) may extend to adjacent vegetation areas, causing the signals in these areas to be misjudged.

[0038] In a specific implementation, in an embodiment of the present invention, when extracting the main lobe signal from the forest tomography profile under HH polarization, the upper filter window size is set to 6, and the lower filter window size is set to 3. The extracted main lobe signal is the most complete, and the side lobe signal forest profile (HH polarization) is eliminated. Figure 5 As shown, Figure 8 In order to preserve the tomographic profile of the sidelobe signal, Fig. 9 Tomographic profile for eliminating sidelobe signals.

[0039] In a specific implementation, the backscatter coefficient is the reflection intensity after the radar signal interacts with the target, and is usually used to characterize the scattering characteristics of the target. The maximum power position corresponding to the maximum backscatter coefficient is determined because, in the tomographic profile of the forest, the stronger the HH polarization backscatter power, the closer its corresponding elevation distribution is to the surface, which is conducive to extracting the terrain part.

[0040] In a specific implementation, in each sliding window, the maximum value and position of the backscatter coefficient are found, which usually corresponds to the main lobe signal of the target. With the maximum power position as the center, the valid pixels in the upper and lower ranges of the window are retained, and the backscatter coefficients of these pixels usually contain useful information of the target.

[0041] In the embodiment of the present invention, a reconstructed three-dimensional point cloud of the forest is obtained by projecting the main lobe signal in the structural section.

[0042] See also Fig.10 , which is a schematic diagram of the reconstructed forest three-dimensional point cloud provided by the present invention.

[0043] In specific implementation, in the existing scheme, a pre-set threshold is used to screen pixels of the forest tomography profile to extract terrain information, and the understory terrain is directly generated based on the processed forest tomography profile. The problem is that the forest tomography profile is processed by the threshold, that is, the part that does not meet the threshold is directly eliminated. However, for example, the reflectivity of the surface and near-surface may be very low after attenuation due to the influence of vegetation. If the threshold is directly used to cut off, then this part of the information representing the real surface will be eliminated, which will lead to discontinuous backscatter power distribution in azimuth and range directions within the forest tomography profile, and the obtained understory terrain will have obvious breakpoints and jumps.

[0044] In the present invention, a forest three-dimensional point cloud is reconstructed based on the forest tomography structure profile, and then the ground point cloud is extracted from the forest three-dimensional point cloud, and the terrain under the forest is obtained based on the ground point cloud, which avoids the disadvantages of using thresholds to extract terrain information in the existing scheme and significantly improves the accuracy of under-forest terrain extraction. The reasons why the present invention can produce better technical effects are, firstly, to retain and utilize the forest tomography structure profile information to the maximum extent, that is, to include the information representing the real surface of the part cut off by the threshold, and secondly, to convert the retained profile information into point cloud information, utilizing the advantages of point cloud ground point filtering, and effectively combining the two.

[0045] In an embodiment of the present invention, the step of reconstructing a three-dimensional point cloud of a forest includes: projecting pixels to corresponding heights according to elevation information corresponding to pixels of backscattered signals in a structural section, azimuthally or distance-wise, to reconstruct a three-dimensional point cloud of the forest.

[0046] In a specific implementation, in the embodiment of the present invention, the image size is 359 (distance) × 269 (azimuth). Fig. 9 The section corresponding to the distance direction of 100 is selected, where the X-axis (ranging from 1 to 269) represents the pixel in the azimuth direction corresponding to the distance direction of 100, and the height corresponding to the Y-axis is the height of this pixel projected onto the 3D point cloud. For example, the point (100, 120, 1650) represents the distance direction of 100, and the height corresponding to the pixel in the azimuth direction of 120 is 1650 meters. Then, each pixel in the distance direction of 100 is saved in this format, and then the entire scene image is converted into a 3D point cloud in each distance direction. Finally, the complete forest 3D point cloud space is reconstructed, and the same is true for each azimuth direction.

[0047] In an embodiment of the present invention, a coarse digital ground model representing the undulation of the terrain is generated, and after the coarse digital ground model is used to perform elevation normalization processing on the forest three-dimensional point cloud, the elevation normalized ground point cloud in the forest three-dimensional point cloud is extracted using cloth simulation filtering, and the elevation normalized ground point cloud is superimposed (for example, added) on the coarse digital ground model to obtain the ground point cloud in the forest three-dimensional point cloud.

[0048] In a specific implementation, the MCSF (Modified Cloth Simulation Filter) provided by the present invention is an improvement on the cloth simulation filter algorithm CSF (Cloth Simulation Filter), and is combined with the digital terrain model DTM (Digital Terrain Model) to achieve higher classification accuracy even in areas with large terrain undulations.

[0049] In the embodiment of the present invention, the three-dimensional point cloud of the forest is gridded in the horizontal direction, and the lowest point in each grid is used as a reference point to obtain a corresponding rough digital ground model.

[0050] In specific implementation, there are multiple schemes for generating a rough digital terrain model. In the embodiment of the present invention, it can also be obtained through the PTDF (progressive TIN densification filtering) algorithm, selecting the lowest point and four corner points in the X×X square grid as initial seed points to construct an initial TIN model; then iterate the TIN model, and in each iteration process, use the pre-set threshold parameters (mainly including the maximum angle θ, the maximum distance d and the maximum terrain slope s) to judge the "potential points (unclassified points)" point by point; finally, generate a DTM based on the detected ground points. This method can improve the accuracy of the rough DTM in order to retain the main terrain change features, while the CSF algorithm has a better filtering effect in the area of ​​detailed terrain changes, so it is not advisable to excessively pursue the accuracy of the rough DTM.

[0051] In an embodiment of the present invention, a coarse digital ground model is used to perform elevation normalization processing on a forest three-dimensional point cloud. The processing process includes: for each target point in the forest three-dimensional point cloud, calculating the elevation difference between the elevation value of the target point and the elevation value of the corresponding point in the coarse digital ground model, and using the elevation difference as the elevation value of the target point, thereby obtaining an elevation-normalized forest three-dimensional point cloud.

[0052] See also Fig.11 , which is a schematic diagram of the principle of cloth simulation filtering provided by the present invention In the specific implementation, the cloth simulation filtering process includes: (1) cloth state initialization: This process starts with the inversion of the point cloud, followed by a uniform cloth simulation covering the data set; (2) particle movement: particles in the cloth will be displaced in the vertical direction under the influence of gravity; (3) collision detection: under the influence of gravity, some particles will reach below the surface when they are displaced, and the particle position will deviate from the actual terrain; (4) displacement calculation driven by internal force: This step mainly considers the internal driving factors between particles in the cloth; (5) calculation of the distance moved between adjacent cloth points due to elevation difference. If two adjacent nodes are both movable points and have different height values, the two points will move in opposite directions at the same distance in the vertical direction. If one of the two points is immovable, the other point will be moved. If the two points have the same elevation, no displacement will occur. In these cases, the distance moved is called the corrected displacement, from which the position of the particle can be calculated. In the embodiment of the present invention, the parameters of the bag filter simulation algorithm can be set as follows: the classification threshold of the ground point is 0.5m, the cloth resolution is 0.1m, the maximum number of iterations is 500 times, and the filtering scene is set to flat terrain (because the elevation normalization is performed). Fig.12 , which are ground points extracted by the MCSF algorithm, see Fig.13 , which are non-ground points extracted by the MCSF algorithm.

[0053] In the embodiment of the present invention, the elevation information of the ground point cloud is interpolated and then projected onto a two-dimensional grid to form an understory terrain model.

[0054] See also Fig.14 , which is an understory terrain model obtained by applying the method provided by the present invention.

[0055] In the specific implementation, there may still be a small number of missing elevation values ​​after the ground points are separated by the MCSF algorithm. The missing terrain values ​​(elevation information) can be completed through the interpolation algorithm, and then the interpolated ground point cloud data can be projected onto a two-dimensional grid to form an understory terrain model.

[0056] In the specific implementation, the generation process of the forest terrain model includes: (1) grid division: divide the entire area into grids of a certain size, each grid represents an area; (2) point cloud projection: project each point in the ground point cloud to the grid where it is located, and the grid where the point is located can be determined by calculating the minimum distance from the point to the grid boundary; (3) form the forest terrain model: in each grid, the elevation information of the point cloud, such as the highest point, the lowest point, the average elevation, etc., is statistically analyzed to form the forest terrain model.

[0057] The present invention also provides an understory terrain extraction device based on multi-baseline SAR data, comprising an image data synthesis unit, a tomography unit, a signal extraction unit, a point cloud reconstruction unit, a ground point cloud extraction unit and an understory terrain model generation unit, wherein: the image data synthesis unit is used to obtain the multi-baseline data set of the target area obtained by acquisition, synthesize the SAR image data set, and correct the interference phase error in the SAR image data set; the tomography unit is used to perform signal decomposition on the SAR image data set, perform tomography on the decomposed signal by using spectrum reconstruction, and obtain the forest tomography structure profile; the signal extraction unit is used to remove The sidelobe signal in the structural section is obtained by retaining the mainlobe signal in the structural section; the point cloud reconstruction unit is used to obtain a reconstructed forest three-dimensional point cloud by projecting the mainlobe signal in the structural section; the ground point cloud extraction unit is used to generate a coarse digital ground model that represents the terrain undulation, and after using the coarse digital ground model to perform elevation normalization on the forest three-dimensional point cloud, the elevation normalized ground point cloud in the forest three-dimensional point cloud is extracted using cloth simulation filtering, and the elevation normalized ground points are superimposed on the coarse digital ground model to obtain the ground point cloud in the forest three-dimensional point cloud; the understory terrain model generation unit is used to project the ground point cloud onto a two-dimensional grid to form an understory terrain model.

[0058] In a specific implementation, the device for extracting understory terrain based on multi-baseline SAR data provided by the present invention includes an execution unit for executing a method, step or function, and the methods, steps or functions executed by the execution unit can refer to the method for extracting understory terrain based on multi-baseline SAR data provided by the present invention.

Claims

1. A method for extracting understory terrain based on multi-baseline SAR data, characterized in that: include: Acquire the multi-baseline data set of the target area, synthesize it to obtain the SAR image data set, and correct the interference phase error in the SAR image data set; The SAR image data set is decomposed, and the decomposed signal is tomographically imaged using spectrum reconstruction to obtain the forest tomographic structure profile; Remove the side lobe signal in the structural section and retain the main lobe signal in the structural section; The reconstructed three-dimensional point cloud of the forest is obtained by projecting the main lobe signal in the structural section; Generate a coarse digital ground model that represents the undulation of the terrain, use the coarse digital ground model to perform elevation normalization on the forest 3D point cloud, use cloth simulation filtering to extract the elevation normalized ground point cloud in the forest 3D point cloud, and superimpose the elevation normalized ground point cloud on the coarse digital ground model to obtain the ground point cloud in the forest 3D point cloud; The ground point cloud is projected onto a two-dimensional grid to form an understory terrain model.

2. The method for extracting understory terrain based on multi-baseline SAR data according to claim 1, characterized in that: The acquiring of the target area multi-baseline data set includes: Acquire the P-band multi-baseline dataset of the target area acquired by the aircraft.

3. The method for extracting understory terrain based on multi-baseline SAR data according to claim 1, characterized in that: The interferometric phase error in the SAR image data set is corrected, including: Remove phase error, flat ground phase and terrain phase from SAR image datasets.

4. The method for extracting understory terrain based on multi-baseline SAR data according to claim 1, characterized in that: The signal decomposition of the SAR image data set and tomography of the decomposed signal using spectrum reconstruction include: Perform SKP signal decomposition on SAR image dataset; Capon spectrum reconstruction is used to suppress SAR echo signal noise and improve the resolution of tomography in the altitude direction.

5. The method for extracting understory terrain based on multi-baseline SAR data according to claim 1, characterized in that: The step of removing the side lobe signal in the structural section and retaining the main lobe signal in the structural section comprises: The size of the sliding window is set, and the maximum power position corresponding to the maximum backscatter coefficient within the sliding window is determined in the structural section; At the maximum power position, the pixels corresponding to the backscattered signals within the upper and lower ranges of the window are extracted through a sliding window; The sidelobe signal is eliminated and the mainlobe signal is extracted pixel by pixel on the structural profile in the azimuth or distance direction.

6. The method for extracting understory terrain based on multi-baseline SAR data according to claim 1, characterized in that: The method of obtaining a reconstructed three-dimensional forest point cloud by projecting the main lobe signal in the structural section includes: According to the elevation information corresponding to the pixels of the backscattered signal in the structural section, the pixels are projected to the corresponding heights in the azimuth or distance direction to reconstruct the three-dimensional point cloud of the forest.

7. The method for extracting understory terrain based on multi-baseline SAR data according to claim 6, characterized in that: The method of generating a rough digital ground model representing terrain fluctuations comprises: The forest 3D point cloud is gridded in the horizontal direction, and the lowest point in each grid is used as the reference point to obtain the corresponding coarse digital ground model.

8. The method for extracting understory terrain based on multi-baseline SAR data according to claim 6, characterized in that: After the elevation normalization of the forest three-dimensional point cloud using the rough digital ground model, the method includes: For each target point in the forest 3D point cloud, the elevation difference between the target point elevation value and the elevation value of the corresponding point in the rough digital ground model is calculated, and the elevation difference is used as the elevation value of the target point, thereby obtaining the forest 3D point cloud after elevation normalization.

9. The method for extracting understory terrain based on multi-baseline SAR data according to claim 1, characterized in that: The projecting of the ground point cloud to a two-dimensional grid to form an understory terrain model includes: The elevation information of the ground point cloud is interpolated and then projected onto a two-dimensional grid to form an understory terrain model.

10. A device for extracting understory terrain based on multi-baseline SAR data, characterized in that: It includes an image data synthesis unit, a tomography unit, a signal extraction unit, a point cloud reconstruction unit, a ground point cloud extraction unit and an understory terrain model generation unit, wherein: The image data synthesis unit is used to obtain the collected target area multi-baseline data set, synthesize it to obtain the SAR image data set, and correct the interference phase error in the SAR image data set; The tomographic imaging unit is used to perform signal decomposition on the SAR image data set, and to perform tomographic imaging on the decomposed signal using spectrum reconstruction to obtain a forest tomographic structure profile; The signal extraction unit is used to remove the side lobe signal in the structural section and retain the main lobe signal in the structural section; The point cloud reconstruction unit is used to obtain a reconstructed three-dimensional point cloud of the forest by projecting the main lobe signal in the structural section; The ground point cloud extraction unit is used to generate a coarse digital ground model that represents the undulation of the terrain, use the coarse digital ground model to perform elevation normalization processing on the forest three-dimensional point cloud, use cloth simulation filtering to extract the elevation normalized ground point cloud in the forest three-dimensional point cloud, and superimpose the elevation normalized ground points on the coarse digital ground model to obtain the ground point cloud in the forest three-dimensional point cloud; The understory terrain model generating unit is used to project the ground point cloud onto a two-dimensional grid to form an understory terrain model.

Images