Long-wave-band PolInSAR under-forest terrain inversion method and device based on non-penetration depth correction

Abstract

The invention provides a long-band PolInSAR under-forest terrain inversion method and device based on non-penetration depth correction, and the method comprises the steps: obtaining two-scene complete polarization SAR image data, carrying out the preprocessing, and generating a polarization interference complex covariance matrix; whitening the polarization interference complex covariance matrix, applying regular matrix constraint to obtain an optimal regular matrix approximate solution, and extracting two endpoint complex coherence coefficients of the phase line segment; on the basis of a polarization coherent scattering model, model parameters including earth surface phase, forest height, extinction coefficient and earth body amplitude ratio are solved by utilizing an endpoint complex coherence coefficient; in combination with sparse LiDAR elevation data, the forest height and the non-penetration depth between the RVoG inversion terrain and the LiDAR reference terrain are calculated, and a one-dimensional lookup table based on the mapping relation between the forest height and the non-penetration depth is constructed; and correcting the elevation deviation of the terrain under the forest inversed by the PolInSAR according to the one-dimensional lookup table. According to the method, the precision of forest area surface topography inversion can be remarkably improved, and large-range under-forest topography surveying and mapping can be realized only by means of single polarization interference.

Description

Technical Field

[0001] The present invention relates to the technical fields of radar interferometry and forest scene parameter inversion, and proposes a technical solution for estimating and correcting understory terrain based on long-wavelength PolInSAR data. Background Art

[0002] Understory terrain products are essential datasets for forest ecosystem monitoring. Accurate characterization of exposed surface terrain allows for more precise estimation of forest vertical structure and geomorphic features. This is crucial for monitoring geological hazards in forest areas, promoting climate change adaptation, supporting the development of carbon reduction strategies, and deepening our understanding of the function and stability of forest ecosystems. Synthetic Aperture Radar (SAR) interferometric measurement (InSAR) technology has been widely used in large-scale terrain mapping due to its all-weather and all-day imaging capabilities. However, traditional InSAR technology mostly relies on shortwave radars such as X-band or C-band. Due to limited signal penetration, it often cannot penetrate the tree canopy to the ground surface in forested areas. The interferometric phase center is often located in the tree canopy, resulting in the generated terrain products being closer to digital surface models (DSMs) than to true "bare earth" terrain models.

[0003] To accurately capture "bare earth" topography, the scientific community continues to focus on the application of long-wavelength (e.g., L-band and P-band) SAR systems, particularly in densely forested areas such as tropical rainforests. Compared to shortwave radar, long-wavelength radar has greater vegetation penetration and, in theory, can provide interferometric information that is closer to the actual terrain. Furthermore, to further separate the scattering contributions from the ground surface and vegetation, polarimetric interferometric SAR (PolInSAR) technology has been proposed and extensively studied. The Random Volume over Ground (RVoG) model, a classic polarimetric coherent scattering parameterization in PolInSAR, assumes that vegetation is composed of a penetrable random volume layer and an impenetrable ground layer. This model establishes a physical relationship between the multi-polarization complex coherence coefficient and forest structural parameters (e.g., forest height, ground phase, ground amplitude ratio, and extinction coefficient). This allows for the extraction of ground phase to a certain extent, and thus the inversion of understory topography. However, even when using P-band fully polarimetric SAR data and the RVoG model, incomplete penetration can still occur, especially in tropical rainforests with tall trees and high canopy density. In such cases, the inferred surface phase will still deviate from the actual terrain, leading to systematic deviations in the understory topography.

[0004] Therefore, how to compensate for understory terrain errors caused by insufficient penetration in the RVoG model by establishing a mapping relationship between forest height and non-penetration depth based on known PolInSAR data has become a pressing research issue. Given this, and having found no relevant methods, the present invention discloses a long-wavelength PolInSAR understory terrain inversion method based on non-penetration depth correction.

[0005] When using the RVoG model to invert forest terrain, traditional methods typically rely on a three-stage algorithm or a nonlinear iterative optimization algorithm to solve the surface phase. However, due to the strong linear correlation between the complex coherence coefficients of multi-polarization interferometry under the same baseline, the inversion model matrix suffers from rank deficiency, resulting in non-unique or unstable parameter solutions. To overcome this rank deficiency, the present invention introduces a normal matrix constraint to extract the complex coherence coefficients of the two endpoints of the coherent line segment from the complex covariance matrix of the polarization interferometry. Summary of the Invention

[0006] The purpose of this invention is to propose a one-dimensional lookup table terrain correction method based on the mapping relationship between forest height and non-penetration depth, which is used to correct the surface elevation deviation problem caused by insufficient penetration in the RVoG model inversion of long-waveband PolInSAR technology.

[0007] The above technical problems of the present invention are mainly solved by the following technical solutions: A long-band PolInSAR understory terrain inversion method based on non-penetrating depth correction includes the following steps: Acquire two scenes of fully polarimetric SAR image data and perform preprocessing to generate the polarimetric interferometry complex covariance matrix; performing whitening processing on the polarization interference complex covariance matrix, applying normal matrix constraints, obtaining an optimal normal matrix approximate solution, and extracting the complex coherence coefficients of the two endpoints of the coherent line segment; Based on the polarization coherent scattering model, the endpoint complex coherence coefficient is used to solve the model parameters, wherein the model parameters include the surface phase, the forest height, the extinction coefficient and the ground amplitude ratio; Combining sparse LiDAR elevation data, we calculated forest height and the non-penetration depth between the RVoG inverted terrain and the LiDAR reference terrain, and constructed a one-dimensional lookup table based on the mapping relationship between forest height and non-penetration depth. According to the one-dimensional lookup table, the elevation deviation of the understory terrain inverted by PolInSAR is corrected, and the corrected digital terrain model is output.

[0008] Moreover, the implementation process of applying the normal matrix constraint includes performing normal matrix constraint on the whitened interference correlation matrix, generating intermediate variable parameters, solving the optimal normal matrix approximate solution, and ensuring the geometric stability of the coherent line segment within the complex unit circle.

[0009] Moreover, when the endpoint complex coherence coefficient is used to solve the model parameters, the surface phase is determined by using the sign of the interference vertical wavenumber; the endpoint complex coherence coefficient is substituted into the polarization coherent scattering model, and the forest height, extinction coefficient and ground amplitude ratio are solved through nonlinear iterative optimization.

[0010] Moreover, the construction and implementation method of the one-dimensional lookup table includes dividing the samples according to preset forest height intervals, counting the mean forest height and the mean non-penetration depth in each interval, thereby constructing a mapping relationship between forest height and non-penetration depth, and obtaining a sparse lookup table; based on the sparse lookup table, a continuous forest height and non-penetration depth mapping relationship is generated through one-dimensional interpolation.

[0011] Moreover, the implementation process of correcting the elevation deviation of the forest terrain is as follows: Based on the forest height inverted by PolInSAR, a non-penetration depth compensation value is obtained by interpolation through the one-dimensional lookup table; The compensation value is deducted from the RVoG inverted terrain to obtain the corrected understory terrain elevation.

[0012] Moreover, the interference baseline length of the two-scene fully polarized SAR image data satisfies that the blur height is greater than half of the maximum forest height.

[0013] Moreover, it is used to achieve understory terrain inversion in areas with high forest cover.

[0014] On the other hand, the present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and runnable on the processor, wherein when the processor executes the program, the long-waveband PolInSAR understory terrain inversion method based on non-penetration depth correction as described above is implemented.

[0015] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the long-waveband PolInSAR understory terrain inversion method based on non-penetration depth correction as described above.

[0016] On the other hand, the present invention also provides a computer program product, comprising a computer program, which, when executed by a processor, implements the long-waveband PolInSAR understory terrain inversion method based on non-penetration depth correction as described above.

[0017] The innovation of this invention lies in combining sparse LiDAR surface elevation information with the PolInSAR polarimetric coherent scattering model. Using a one-dimensional lookup table, this method compensates for the non-penetrating system elevation bias in the RVoG model's inversion of understory terrain. This significantly improves the accuracy of surface terrain inversion in tropical rainforests and enables large-scale understory terrain mapping using only a single polarimetric interferometer pair. This approach will facilitate current and future polarimetric spaceborne SAR (such as the BIOMASS satellite) understory terrain mapping missions. Therefore, this invention demonstrates strong engineering feasibility and practical application value, and is particularly suitable for SAR understory terrain inversion and DEM product correction in areas with high forest cover.

[0018] The core technology of this invention lies in using statistical regression methods to construct a one-dimensional lookup table mapping model between forest height and non-penetration depth, which is beneficial for achieving rapid and accurate correction of understory terrain over a wide area. SAR imagery observations offer the advantages of large-scale, continuous area observations, while LiDAR observations offer the advantages of point-based observations and high terrain accuracy. This method fully combines the advantages of both SAR and LiDAR technologies. Based on the parameter inversion of the polarization coherent scattering model (RVoG model), combined with localized or sparsely acquired LiDAR terrain information, it is extended to large areas covered only by full polarimetric interferometric SAR imagery but lacking high-density LiDAR data. This method has excellent practicality and potential for expansion.

[0019] The solution of the present invention is simple and convenient to implement and has strong practicality. It solves the problems of low practicality and inconvenience in actual application existing in related technologies, can improve user experience, and has important market value. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a flow chart of a long-band PolInSAR understory terrain inversion method based on non-penetration depth correction according to an embodiment of the present invention.

[0021] Figure 2 This is a one-dimensional lookup table and interpolation version between the non-penetration depth and the forest height according to an embodiment of the present invention.

[0022] Figure 3 Comparison chart of DTM products inverted by the embodiment method of the present invention and different methods: (a) TanDEM-X DEM; (b) DTM obtained by RVoG; (c) DTM corrected using the present invention.

[0023] Figure 4 Comparison of DTM accuracy between the method according to the embodiment of the present invention and different methods: (a) TanDEM-X DEM; (b) DTM obtained by RVoG; (c) DTM corrected using the present invention. DETAILED DESCRIPTION

[0024] The following will further illustrate the concept, specific structure and technical effects of the present invention in conjunction with the accompanying drawings and embodiments, so as to fully understand the purpose, characteristics and effects of the present invention.

[0025] In densely forested areas, even RVoG model inversion based on longwave Polarimetric Interferometry (PolInSAR) data still exhibits systematic errors in the resulting terrain. This invention defines the difference between the RVoG model-estimated terrain and the sparse Light Detection and Ranging (LiDAR) reference terrain as the "non-penetration depth." Furthermore, a one-dimensional lookup table mapping relationship between forest height and non-penetration depth is established. This RVoG model-estimated forest height is then generalized to the entire area covered by Polarimetric Interferometry (PolISAR) imagery, enabling high-precision mapping of large-scale understory terrain. This invention is particularly suitable for improving the accuracy of surface terrain mapping and inversion in tropical rainforest areas.

[0026] See also Figure 1 The embodiment of the present invention provides a long-band PolInSAR forest terrain inversion method based on non-penetration depth correction, comprising the following steps: Step 1: Data preprocessing: Preferably, the interference baseline length is such that the blur height is greater than half of the maximum forest height.

[0027] Example: Two fully polarimetric SAR images covering the study area were obtained with an interferometric baseline length of 40 m. One of the appropriate SAR images was selected as the primary image. The image was then interferometrically processed with the other SAR image, and multiple data processing steps including removing the flat ground phase, removing the reference terrain phase, and performing multi-look filtering were performed to estimate the polarimetric interferometric complex covariance matrix. , which can be expressed as

[0028] in, and are the polarization complex covariance matrices of the master and slave images, is the interference cross-correlation matrix, the superscript Represents the conjugate transpose operation.

[0029] In this example, the study area is a region on the west coast of a country in west-central Africa, with geographic coordinates ranging from approximately 1.88° to 1.92°S and 10.17° to 10.22°E. This region is one of four typical tropical rainforest study areas within the European Space Agency's AfriSAR airborne SAR experimental project. This project conducted an airborne, fully polarimetric P-band SAR mission in 2016, acquiring data covering approximately 60 km². The data, featuring excellent polarimetric configuration and interferometric baseline design, provide favorable conditions for high-quality polarimetric interferometric imaging analysis. The region is primarily hilly, with relatively gentle surface elevation variations ranging from approximately 10 to 200 meters. The vegetation is a typical tropical rainforest ecosystem, dominated by dense, high-canopy closure natural forests, with an average forest height of approximately 30 meters. The TanDEM-X DEM elevation product was used for data processing to remove terrain phases and geocode the data. While the project was running, in February and March 2016, agencies including ESA, DLR, ACN and NASA acquired 20 m resolution LiDAR data over the experimental area using the Land, Vegetation and Ice Sensor (LVIS) on a NASA Lengley B200 aircraft.

[0030] Step 2: After the whitening complex covariance matrix is processed, the normal matrix constraint method of the dual phase center is applied to the polarization interference complex covariance matrix to obtain the optimal normal matrix approximate solution and extract the two endpoints of the visible coherence line segment corresponding to the maximum separation of the complex coherence coefficient.

[0031] When using the RVoG model to invert forest terrain, traditional methods usually rely on a three-stage algorithm or a nonlinear iterative optimization algorithm to solve the surface phase. However, due to the strong linear correlation between the complex coherence coefficients of multi-polarization interference under the same baseline, the inversion model matrix has a rank deficiency problem, which makes the parameter solution non-unique or unstable. In order to overcome the rank deficiency problem of RVoG model parameter solution, the present invention imposes a normal matrix constraint on the whitened interference cross-correlation matrix to obtain an approximate solution that best approximates the normal matrix, ensuring that the complex coherence coefficients of all polarization channels appear as geometrically stable coherent line segments in the complex unit circle plane, thereby more reliably extracting the complex coherence coefficients of the two endpoints of the coherence segment.

[0032] First, the polarization interference complex covariance matrix Perform whitening processing to obtain the whitened interference correlation matrix :

[0033] in, is the mean of the polarimetric covariance matrices of the master and slave images.

[0034] Compute the intermediate variable parameters under the constraints of the normal matrix of the biphase center:

[0035]

[0036]

[0037]

[0038]

[0039]

[0040]

[0041] in, 、 、 、 、 、 、 、 、 、 、 、 、 is the intermediate variable parameter; is the imaginary unit, ( ) is the trace of the matrix, () is the inverse tangent function.

[0042] Next, the optimal normal matrix approximates the solution It can be expressed as:

[0043] in, is the identity matrix, for The conjugation of .

[0044] The above optimal normal matrix approximation solution ensures the geometric stability of the visible line segments of the complex coherent line in the complex unit circle plane, thereby more reliably extracting the complex coherence coefficients corresponding to the two endpoints of the coherent line segment.

[0045] Step 3: Based on the polarization coherent scattering model RVoG, the complex coherence coefficients corresponding to the two endpoints of the coherent line segment are used as observations, and the model parameters are solved through nonlinear iterative optimization, including surface phase, forest height, extinction coefficient, and the amplitude ratio of the two ground bodies.

[0046] First, based on the two complex coherence observations obtained in step 2 (the two endpoints of the corresponding coherence line segment), the two intersection points of the corresponding coherence line and the unit complex coherence circle are calculated (i.e. and ), and adaptively determine the surface phase by interfering the sign of the vertical wavenumber.

[0047] For a single-station system, the interferometric vertical wave number The calculation formula is as follows:

[0048] in, represents the radar wavelength, Indicates the vertical baseline length, Indicates the slant range between the sensor and the ground target.

[0049] According to the interferometric geometric relationship, the criterion for determining the surface phase using the sign of the interferometric vertical wave number is: against :

[0050]

[0051] against :

[0052]

[0053] in, express The conjugate of express The conjugate of represents the exponential function, ( ) indicates the scalar operation.

[0054] After obtaining the surface phase through the above method, the surface phase can be As known values, they are substituted into the following polarization coherent scattering model formula (i.e., RVoG model).

[0055]

[0056] in, represents a mathematical constant, Indicates polarization mode The complex coherence coefficient observation of represents the pure body decoherence coefficient, represents the extinction coefficient, Indicates polarization mode The ground amplitude ratio, Indicates the forest height.

[0057] The complex coherence coefficients corresponding to the two endpoints of the coherent line segment obtained by the optimal normal matrix approximation are used as the observation quantity of the RVoG model and substituted into the nonlinear iterative optimizer to solve the forest height, that is:

[0058] in, represents the RVoG model, represents the norm of the Euclidean vector, Indicates finding the minimum value of a function.

[0059] Step 4: Based on the external TanDEM-X DEM elevation product, the surface phase is converted using the phase height conversion factor to obtain the surface elevation obtained by RVoG model inversion; combined with the accurate elevation information of sparse airborne LiDAR, the non-penetration depth is obtained and a one-dimensional mapping lookup table between forest height and non-penetration depth is constructed.

[0060] Based on the surface phase obtained in step 3, the phase height conversion operator and the TanDEM-X DEM elevation product used in step 1 are used to convert the surface phase into the elevation product. , and obtain the surface elevation based on the RVoG model inversion :

[0061] in, This is the RVoG understory terrain product.

[0062] LiDAR ground elevation data acquired with the help of the airborne LVIS , which can obtain the non-penetration depth of SAR signal ,Right now:

[0063] With the help of the forest height obtained in step 3, a series of forest height sampling intervals are pre-set, such as 0-5 m, 5-10 m, 10-15 m, 15-20 m, 20-25 m, 25-30 m, 30-35 m, 35-40 m, 40-45 m, 45-50 m, 50-55 m, 55-60 m, etc. The mean forest height and the mean non-penetration depth of all sample points in each forest height interval are counted, so as to construct a one-dimensional lookup table mapping relationship between forest height and non-penetration depth. ,Right now:

[0064] Figure 2A one-dimensional lookup table and interpolated version showing the relationship between non-penetration depth and forest height. P-band SAR signals fully penetrate low-lying trees in tropical rainforests, but lack penetration in taller trees. As forest height increases, the non-penetration depth also increases.

[0065] Step 5: Based on the forest height results obtained from PolInSAR image inversion, the elevation deviation contained in the RVoG understory terrain product is corrected through a one-dimensional lookup table, thereby achieving the goal of large-scale high-precision understory terrain mapping.

[0066] The sparse lookup table obtained in step 4 is used to obtain the continuous mapping relationship between forest height and non-penetration depth through one-dimensional interpolation. This process can also be completed in advance in step 4; then, the forest height obtained by PolInSAR surface observation and continuous cover inversion is used to correct the RVoG understory terrain product. The elevation deviation caused by insufficient SAR penetration is eliminated, thus obtaining a large-scale, high-precision forest terrain product. ,Right now

[0067] in, is the non-penetrating depth obtained from the forest height map according to the interpolated version of the one-dimensional lookup table.

[0068] Figure 3 and Figure 4 The DTM products inverted by different methods and the elevation accuracy map with LiDAR DTM as reference are given respectively: (a) TanDEM-X DEM; (b) DTM obtained by RVoG; (c) DTM corrected by the present invention. Figure 4 (a) The TanDEM-X DEM product exhibits significant systematic terrain overestimation, with a root mean square error (RMSE) of up to 17.14 meters and a bias of 15.95 meters. This result highlights the limited penetration capability of X-band SAR signals. The interferometric phase center is often located above the canopy rather than near the actual ground, resulting in severe overestimation in densely forested areas. Figure 4 (b) shows the terrain results obtained by the linear fitting method based on RVoG. Compared with the LiDAR DTM, the RMSE is reduced to 9.57 meters and the deviation is reduced to 6.46 meters. This result shows that even if the full polarization P-band PolInSAR data and the parameterized RVoG model inversion method are used, the incomplete penetration of the radar signal will still lead to overestimation of the acquired terrain; because even if the polarization coherent scattering model is used for surface phase compensation, the P-band SAR signal still cannot fully penetrate to the surface. Figure 4As shown in (c), the terrain correction results of the present invention further reduce the RMSE to 5.90 meters and the elevation deviation to -1.12 meters; the deviation of the DTM product corrected by the present invention is basically close to zero, which verifies that the proposed method can solve the problem of terrain elevation overestimation caused by insufficient SAR signal penetration, and demonstrates its great potential for high-precision mapping of understory terrain in dense forest areas.

[0069] Thus, the present invention has completed and implemented a long-wavelength PolInSAR understory terrain inversion method based on non-penetrating depth correction. Through the above process, based on the polarization coherent scattering (RVoG) model, leveraging localized or sparse LiDAR surface elevation information, and using a one-dimensional lookup table from tree height to non-penetrating depth, PolInSAR technology can achieve large-scale understory terrain mapping. This method is particularly effective in correcting surface terrain in tropical rainforests.

[0070] In specific implementation, the method proposed in the technical solution of the present invention can be automatically run by those skilled in the art using computer software technology. System devices that implement the method, such as computer-readable storage media that store the corresponding computer program of the technical solution of the present invention and computer equipment that runs the corresponding computer program, should also be within the scope of protection of the present invention.

[0071] The following describes the long-band PolInSAR understory terrain inversion electronic device based on non-penetration depth correction provided by the present invention. The long-band PolInSAR understory terrain inversion electronic device based on non-penetration depth correction described below and the long-band PolInSAR understory terrain inversion method based on non-penetration depth correction described above can be referenced to each other.

[0072] The electronic device may include: a processor, a communications interface, a memory, and a communications bus, wherein the processor, the communications interface, and the memory communicate with each other via the communications bus. The processor may invoke logic instructions in the memory to execute a long-wavelength PolInSAR understory terrain inversion method based on non-penetration depth correction, primarily including the software processing portion of the aforementioned steps.

[0073] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage media include various media capable of storing program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0074] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the software processing part of the long-wave band PolInSAR understory terrain inversion method based on non-penetration depth correction provided by the above methods.

[0075] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the software processing part of the long-waveband PolInSAR understory terrain inversion method based on non-penetration depth correction provided by the above-mentioned methods.

[0076] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0077] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0078] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A long-band PolInSAR understory terrain inversion method based on non-penetration depth correction, characterized in that: The following processes are included: Acquire two scenes of fully polarimetric SAR image data and perform preprocessing to generate the polarimetric interferometry complex covariance matrix; performing whitening processing on the polarization interference complex covariance matrix, applying normal matrix constraints, obtaining an optimal normal matrix approximate solution, and extracting the complex coherence coefficients of the two endpoints of the coherent line segment; Based on the polarization coherent scattering model, the endpoint complex coherence coefficient is used to solve the model parameters, wherein the model parameters include the surface phase, the forest height, the extinction coefficient and the ground amplitude ratio; Combining sparse LiDAR elevation data, we calculated forest height and the non-penetration depth between the RVoG inverted terrain and the LiDAR reference terrain, and constructed a one-dimensional lookup table based on the mapping relationship between forest height and non-penetration depth. According to the one-dimensional lookup table, the elevation deviation of the understory terrain inverted by PolInSAR is corrected, and the corrected digital terrain model is output.

2. The method according to claim 1, wherein: The implementation process of applying the normal matrix constraint includes performing the normal matrix constraint on the whitened interference correlation matrix, generating intermediate variable parameters, solving the optimal normal matrix approximate solution, and ensuring the geometric stability of the coherent line segment within the complex unit circle.

3. The method according to claim 1, wherein: When the endpoint complex coherence coefficient is used to solve the model parameters, the surface phase is determined by using the sign of the interferometric vertical wavenumber; the endpoint complex coherence coefficient is substituted into the polarization coherent scattering model, and the forest height, extinction coefficient and ground amplitude ratio are solved through nonlinear iterative optimization.

4. The method according to claim 1, wherein: The one-dimensional lookup table is constructed by dividing samples into preset forest height intervals, and calculating the mean forest height and the mean non-penetration depth in each interval, thereby constructing a mapping relationship between forest height and non-penetration depth to obtain a sparse lookup table; Based on a sparse lookup table, a continuous mapping relationship between forest height and non-penetration depth is generated through one-dimensional interpolation.

5. The method according to claim 1, wherein: The process of correcting the elevation deviation of the understory terrain is as follows: based on the forest height inverted by PolInSAR, a non-penetration depth compensation value is obtained by interpolation of the one-dimensional lookup table; the compensation value is deducted from the RVoG inverted terrain to obtain the corrected understory terrain elevation.

6. The method according to claim 1, wherein: The interference baseline length of the two-scene fully polarized SAR image data satisfies that the blur height is greater than half of the maximum forest height.

7. The method according to claim 1, wherein: Used to achieve understory terrain inversion in high forest cover areas.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the long-band PolInSAR understory terrain inversion method based on non-penetration depth correction as described in any one of claims 1 to 7 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the long-band PolInSAR understory terrain inversion method based on non-penetration depth correction as claimed in any one of claims 1 to 7 is implemented.

10. A computer program product comprising a computer program, characterized in that: When the computer program is executed by a processor, the long-band PolInSAR understory terrain inversion method based on non-penetration depth correction as claimed in any one of claims 1 to 7 is implemented.

Images