A high-resolution remote sensing scene three-dimensional radiance calculation method and application

Abstract

The application relates to the technical field of space remote sensing, and specifically provides a three-dimensional radiance calculation method and application of a high-resolution remote sensing scene. First, a surface cell grid of a three-dimensional space is created, spontaneous radiation of all triangular surface cells under sunlight irradiation is calculated, a traditional RGM model is improved, a plurality of threads of a GPU are used to perform parallel calculation of a visibility factor under an OpenACC parallel framework, a successive over-relaxation method is used to iteratively calculate radiation flux density of the surface cells, after the radiation flux density of each surface cell is obtained, bidirectional reflectivity BRF under different observation angles is calculated, efficient simulation of bidirectional reflectivity and radiation transmission of a complex scene is realized. The application significantly improves the calculation speed of an original RGM model, the calculation speed is improved by nearly 70 times through testing, and when a large-scale scene is simulated, ten million small surface pieces can be processed without sacrificing precision.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of space remote sensing, and particularly relates to a three-dimensional radiance calculation method for a high-resolution remote sensing scene and application. BACKGROUND

[0002] With the rapid development of remote sensing technology, three-dimensional radiative transfer (RT) models play an increasingly important role in simulating bidirectional reflectance factors (BRFs). Compared with one-dimensional RT models, three-dimensional RT models are less dependent on assumptions and can provide higher accuracy, playing an important role in forest structure and biophysical parameter extraction applications, and are widely used in the study of vegetation, forest and urban scenes. RT models can accurately calculate multiple scattering and maintain energy conservation. Currently commonly used models include the discrete ordinate method (DOM), the Monte Carlo radiative transfer method (MCRT) and the radiance model (RGM).

[0003] The RGM model has been widely used in the field of vegetation remote sensing, and is suitable for tasks such as plant index design, leaf size and shape extraction, and scale effect research. However, the RGM model has low computational efficiency, especially when simulating large heterogeneous canopy scenes at the pixel scale, the calculation of the view factor often faces the limitation of low efficiency, and even cannot be completed. This problem hinders its expansion application in high-resolution and larger scenes. For example, when the scene contains more than 100,000 complex structures, the view factor calculation may cause the simulation process to be unable to proceed.

[0004] Although methods such as the RAPID model improve the computational efficiency by simplifying the scene, excessive simplification can lead to a decrease in the accuracy of the simulation results. In heterogeneous canopy scenes, preserving the fine structure of the scene is crucial for the accuracy of BRF simulation. Therefore, there is an urgent need for a solution that can efficiently handle view factor calculation while preserving as much detail of the canopy scene as possible to meet the simulation needs of high-resolution remote sensing scenes at the pixel scale. SUMMARY

[0005] Therefore, the present application aims to provide a three-dimensional radiance calculation method for a high-resolution remote sensing scene and application, which improves the traditional radiance model (RGM) based on the OpenACC parallel framework, optimizes and rewrites the view factor calculation part of the RGM model through OpenACC technology, and optimizes the design of the polygon filling algorithm to fill the region with polygons, improves the calculation equation of the surface element radiative flux density, breaks through the efficiency limit of the view factor calculation, and realizes high-precision BRF simulation at the pixel scale.

[0006] To achieve the above purpose, the technical scheme of the present application is as follows:

[0007] This invention provides a method for calculating the three-dimensional radiometrics of a high-resolution remote sensing scene, comprising: generating a triangular mesh by defining vertices in a three-dimensional space to create a three-dimensional scene;

[0008] Calculate the flux density of spontaneous emission of a surface element under sunlight based on its optical parameters.

[0009] Visual factors are computed using the OpenACC parallel framework, including: transferring face data used to compute visual factors from the CPU to the GPU, and distributing the computation task of iteratively computing visual factors between different face data to multiple threads on the GPU to accelerate visual factor computation.

[0010] The radiative flux density of the surface element is calculated iteratively using the successive over-relaxation method. The iterative calculation formula is as follows:

[0011] ;

[0012] in, Indicates the relaxation factor. This represents the radiative flux density of the i-th surface element in the (k+1)-th iteration. This represents the radiative flux density of the i-th surface element in the k-th iteration. Let represent the radiative flux density of the j-th surface element in the k-th iteration. The reflectance or transmittance of the j-th surface element is represented by N, where N represents the total number of surface elements. Represents the visibility factor of facet j to facet i. This represents the flux density of spontaneous radiation of the i-th surface element under sunlight.

[0013] The biaxial reflectivity (BRF) at different observation angles is calculated based on the radiant flux density of each surface element. The formula is as follows:

[0014] ;

[0015] in, Indicates the observation angle. Indicates in Diaxial reflectivity at an angle Indicates the direction of the normal to the i-th face element. Indicates from the perspective of observation The area of ​​the i-th facet seen. Let represent the radiative flux density of the i-th surface element.

[0016] Preferably, the formula for calculating the flux density of spontaneous radiation of a surface element under sunlight is:

[0017] ;

[0018] in, represents the area of the i-th facet, represents an arbitrary point on the i-th facet the flux density of spontaneous emission.

[0019] Preferably, the calculation formula of the visibility factor is:

[0020] ;

[0021] wherein, represents the derivative of with respect to, represents the derivative of with respect to, represents and the angle between the line connecting and the normal of the surface, represents and the angle between the line connecting and the normal of the surface, represents and the Euclidean distance between

[0022] Preferably, at least one of the following strategies is adopted to accelerate the calculation of the field of view factor under the OpenACC parallel framework: parallel loop instruction, optimized data layout, minimized data transmission, and dynamic memory allocation.

[0023] Preferably, the area of the facet is filled by a polygon filling algorithm, and the Cartesian coordinates of an arbitrary point in the three-dimensional space are converted into barycentric coordinates, and the barycentric coordinates are used to determine whether the point is located in a certain facet.

[0024] Preferably, in the polygon filling algorithm, the coordinates of the three vertices A, B and C of the facet are used to express an arbitrary point P as:

[0025] ;

[0026] wherein, and represent the barycentric coordinates of point P;

[0027] When and are both non-negative, point P is in the facet; otherwise, point P is outside the facet.

[0028] Preferably, the calculation formula of the barycentric coordinates and is:

[0029] ;

[0030] wherein, , and represent the components of the vectors , and in the x-direction, , and represent the components of the vectors , and in the y-direction.

[0031] Preferably, in the process of calculating the visibility factor, the storage structure in the GPU is a hash-based data structure.

[0032] Preferably, the range of the relaxation factor is 0 to 2, and in the process of solving the radiant flux density, the optimal relaxation factor is searched in the range of the relaxation factor with a step of 0.2.

[0033] The application further provides an application of the high-resolution remote sensing scene three-dimensional radiance calculation method.

[0034] Compared with the prior art, the application has the following beneficial effects:

[0035] The application improves the traditional radiance model (RGM) based on the OpenACC parallel framework, and proposes a three-dimensional radiance calculation method with higher calculation efficiency, larger calculation scale and guaranteed modeling accuracy. PS The application transfers data operation from the CPU to the GPU, and designs an improved radiance model RGM PS in the GPU by using the OpenACC parallel technology.

[0036] The application has good accuracy and reliability in simulating the bidirectional reflectance and radiation transmission of complex scenes. PSThe correlation (R2) between the original RGM model is more than 0.94, the root mean square error (RMSE) is less than 0.0038, and the cross-validation with the DART model shows high consistency (R2 up to 0.98), with RMSE of 0.0031 and 0.0340 in the red and near-infrared bands, with very high accuracy and reliability.

[0037] The parallel iterative operation of the application not only significantly improves the calculation speed, but also expands the calculable scale, improves the complexity of the simulating scene from tens of thousands of faces of the original RGM model to 14 million faces, can meet the simulation needs of larger range and high resolution remote sensing scenes, and provides the possibility for studying larger scale and more complex ecological systems, and can be currently applied to the fields of vegetation remote sensing, forest structure and biophysical parameter extraction, and urban scene research, and can realize more efficient and more accurate simulation.

[0038] The application optimizes the traditional face filling and projection area calculation method, optimizes the triangular filling algorithm (TFBCs) based on the barycentric coordinates, replaces the traditional line scanning algorithm with the triangular filling algorithm based on the barycentric coordinates, improves the efficiency of filling and projection area calculation, is more suitable for GPU acceleration, further improves the calculation performance of the model, and ensures that the model exhibits high precision in the BRF simulation of the main plane and the cross plane through the optimization of the polygon filling algorithm and the parallel strategy, and realizes the balance between precision and efficiency.

[0039] The application changes the data storage structure in the simulation calculation process from the traditional array to the hash-based data structure, improves the data query efficiency from O(n) to O(1) level, and speeds up the access and processing speed of data in the view factor calculation process. BRIEF DESCRIPTION OF DRAWINGS

[0040] The drawings constituting a part of the application are used to provide further understanding of the application, the schematic embodiments of the application and the description thereof are used to explain the application, and do not constitute improper limitation on the application. In the drawings:

[0041] Figure 1 It is a flowchart of the three-dimensional radiance calculation method of the high-resolution remote sensing scene provided by the embodiment of the application;

[0042] Figure 2 It is a running strategy diagram of the view factor parallel processing under the OpenACC parallel framework provided by the embodiment of the application;

[0043] Figure 3 It is a simulation reflectivity comparison result diagram of the DART model and the RGM constructed by the method of the application PS of the embodiment of the application.

[0044] Figure 4 The RGM provided according to the embodiments of the present invention PS A comparison of simulation results from the DART and 4SAIL models with experimental results from real-world airborne measurements. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only for explaining the invention and do not constitute a limitation thereof. Similar elements in different embodiments are referred to by associated similar element reference numerals. In the following embodiments, many details are described to facilitate a better understanding of the invention. However, those skilled in the art will readily recognize that some features may be omitted in different situations, or may be replaced by other elements, materials, or methods. In some cases, some operations related to the invention are not shown or described in the specification. This is to avoid obscuring the core parts of the invention with excessive description. For those skilled in the art, detailed description of these related operations is not necessary; they can fully understand the related operations based on the description in the specification and general technical knowledge in the art.

[0046] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined to form various implementations. Furthermore, the order of the steps or actions in the method description can be changed or adjusted in a manner readily apparent to those skilled in the art. Therefore, the various orders in the specification and drawings are merely for the clear description of a particular embodiment and do not imply a mandatory order, unless otherwise stated that a particular order must be followed.

[0047] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application. In addition, the terms "first", "second" and the like are only for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined with "first", "second" and the like can be explicitly or implicitly included one or more. In the description of the present application, unless otherwise stated, the meaning of "a plurality of" is two or more.

[0048] In the description of the present application, it should be noted that unless otherwise specified and limited, the terms "mounting", "connection", "connection" should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected; it can be mechanically connected, or it can be electrically connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be connected inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood through specific circumstances.

[0049] The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0050] Please refer to Figure 1 In an embodiment of the present application, a high-resolution remote sensing scene three-dimensional radiance calculation method is provided, which optimizes and improves the traditional radiance model (RGM) by using the method, and constructs a three-dimensional radiance improved model (RGM PS ) in GPU environment by using OpenACC parallel technology, which breaks through the efficiency limitation of traditional RGM model in field factor calculation, and can realize efficient simulation of complex canopy bidirectional reflectance and radiation transmission. The three-dimensional radiance calculation method is as follows:

[0051] S1: First, create a three-dimensional scene. Specifically, in a three-dimensional space, a series of vertices are defined to generate a triangular mesh, and complex geometric shapes are decomposed into multiple triangles, each triangle being a facet, and each triangular facet being determined by three vertices.

[0052] These triangular facets can be combined to form complex geometric shapes, such as vegetation leaves, buildings, and terrain. Correspondingly, the outer vertices of the graphic formed by these facets can represent the corner points of complex geometric shapes such as vegetation leaves, buildings, and terrain.

[0053] Define spectral properties, such as reflectivity and transmittance, for the surface of each triangular element to define the surface properties of the scene. These spectral properties will be used for subsequent radiometric calculations.

[0054] S2: After the 3D scene is created, it is necessary to first determine the radiance of all facets in the 3D scene, that is, the spontaneous radiation flux density generated by each facet under sunlight. Specifically, assuming the total number of facets in the 3D scene is N, the radiative flux density of each facet is affected by both solar radiation and the radiation from other facets. Specifically, the radiance of any discrete triangular facet i can be expressed as:

[0055] ;

[0056] in, Represents the area of ​​the i-th face element. Let the area of ​​the j-th face be denoted as . Indicates to Differential, that is, the differential area of ​​surface element i. Let represent the radiative flux density of the i-th surface element. This represents the flux density of spontaneous radiation of the i-th surface element under sunlight. This represents the reflectivity or transmittance of the i-th surface element, i.e., the optical parameter of the surface element. This represents the visibility factor of the differential area from the j-th facet to the i-th facet. It's important to note that i can take the value 2N, because we need to consider the front and back faces of N facets.

[0057] Therefore, in this step, we can first calculate the flux density of spontaneous radiation formed by the surface element under sunlight. The flux density of spontaneous radiation of any i-th surface element under sunlight can be solved by the following formula:

[0058] ;

[0059] in, Represents any point on the i-th face element The flux density of spontaneous radiation.

[0060] The visibility factor in the radiance calculation formula is a fundamental parameter in radiative transfer analysis, quantifying the radiative transfer from an infinitesimal surface. The emitted radiation flux density reaches an infinitesimal surface The proportion of visibility factor is given by the formula:

[0061] ;

[0062] wherein, denotes the angle between the line connecting and the normal of the surface of , denotes the angle between the line connecting and the normal of the surface of , denotes the Euclidean distance between .

[0063] Through the above, the visibility factor of the differential surface element can be determined, and then the visibility factor between the surface elements can be obtained by integration. At present, various models and methods mainly use the ray projection method, use the line scanning algorithm for area filling and subsequent projection area method for calculation, and the calculation efficiency of the visibility factor between a large number of surface elements in a three-dimensional scene is extremely low. The calculation of the visibility factor often faces the problem of low efficiency or even cannot be completed. The problem of low efficiency leads to the inability to realize the three-dimensional radiosity calculation of high resolution or larger scenes, which limits the existing models and methods to only process scenes with no more than 100,000 surface elements. However, the number of surface elements of high resolution and large scenes usually needs to be greater than 100,000. If the number of surface elements is compressed and reduced, the simulation accuracy will be affected, thereby greatly limiting the application of the RGM model.

[0064] S3: Since the calculation of the visibility factor accounts for about 85% of the total calculation time of remote sensing simulation, the iterative calculation process of the visibility factor is optimized in the present application. OpenACC parallel acceleration technology is used to speed up the running time without affecting the calculation accuracy. Specifically, the OpenACC parallel framework is used to calculate the visibility factor. The OpenACC parallel framework can transfer the calculation data and process from the CPU to the GPU, and use the powerful computing power of the GPU to distribute the visibility factor calculation tasks between different surface elements to multiple threads of the GPU, thereby accelerating the visibility factor calculation. The OpenACC framework aims to facilitate code development and simplify the porting of existing codes on heterogeneous computing systems to achieve better parallel performance. In addition, the present application also adopts an incremental method and a combination of various strategies to improve the correctness and processing speed of the RGM model under the OpenACC framework. PS .

[0065] For specific acceleration strategies, please refer to Figure 2 , including:

[0066] Parallel loop instructions of OpenACC are used to distribute iterations of the loop to multiple threads, thereby speeding up the execution of the loop. The number of threads started by each GPU core function and its clustering can be optimized in a fine-grained manner through vector, worker and gang instructions.

[0067] Compared with the structure array (AOS) scheme in the CPU working mode, the GPU working mode is more suitable for the array structure (SoA) scheme to meet the merging memory access requirement in the RGMPS implementation. The implementation of the distribution function index formula needs to convert the array into to adapt to the single instruction multiple thread (SIMTs) execution model. Therefore, the data layout of the bin data and the calculation data needs to be optimized, and the data layout can be optimized through data compression, data alignment and data packaging technologies to arrange data in a way of maximizing performance and minimizing memory usage.

[0068] For data transmission between the CPU and the GPU, the data transmission optimization is also minimized, and the data amount to be processed is reduced or a more efficient data transmission method is used through optimized code. Specifically, the copyin clause is used to copy the array in the CPU cache or memory to the GPU before calculation, and the copyout clause is used to allocate an empty buffer and copy it back to the host memory after leaving. In this way, the amount of data to be transmitted between different parts of the system can be reduced.

[0069] In order to improve the running efficiency of parallel computing and avoid memory leakage or excessive memory usage, the data allocation rules are also optimized, and dynamic memory allocation is used to save memory space and reduce the amount of memory required for temporary variables between calculation regions.

[0070] In the calculation process of the visibility factor, the traditional RGM model uses a line scanning algorithm for region filling and subsequent projection area calculation, but this method is low in calculation efficiency and difficult to adapt to GPU acceleration. Therefore, in the embodiment of the present application, the traditional line scanning algorithm is replaced by a triangle filling algorithm (TFBCs) based on the barycentric coordinates for region filling and projection area calculation. Specifically, the Cartesian coordinates of any point in a three-dimensional space are converted into barycentric coordinates, and the barycentric coordinates are used to determine whether the point is inside a certain triangular bin. In the polygon filling algorithm, the coordinates of the three vertices A, B and C of the bin can be used to express any point P as:

[0071]

[0072] wherein, and represent the barycentric coordinates of point P;

[0073] When ​And If both are non-negative, then point P is inside the facet; otherwise, point P is outside the facet.

[0074] As for the barycentric coordinates And It can be calculated by the following formula:

[0075] ;

[0076] Wherein, , And Respectively represent the components of the vector , And In the x direction, , And Respectively represent the components of the vector , And In the y direction.

[0077] By area filling and projection area calculation, the visibility factor of each facet can be determined, for the calculated data, the embodiment of the application adopts a hash-based data structure for storage, through this way instead of traditional array storage structure, the data query efficiency is improved from O(n) to O(1) level, greatly improving the data access speed and processing efficiency.

[0078] S4: After obtaining the visibility factor, the successive over-relaxation method (SOR) is used to iteratively calculate the radiation flux density of the facet. The successive over-relaxation method (SOR) is a weighted average method introduced in the Gauss-Seidel method used in the traditional radiation flux density calculation in the process of traditional radiation flux density iterative calculation. The iterative calculation formula of the radiation flux density of the facet is:

[0079] ;

[0080] Wherein, Indicates the relaxation factor, Indicates the (k+1)th iteration radiation flux density of the i-th facet, Indicates the kth iteration radiation flux density of the i-th facet, Indicates the kth iteration radiation flux density of the j-th facet, Indicates the reflectivity or transmissivity of the j-th facet, and N indicates the total number of facets, Indicates the visibility factor from the j-th facet to the i-th facet, Indicates the flux density of the i-th facet self-emitted under the illumination of sunlight. The value range of the relaxation factor Is 0 to 2, and the relaxation factor The value of the radiation flux density can be optimized and adjusted, and in the process of solving the radiation flux density, the optimal relaxation factor is searched in the range of the relaxation factor with a step of 0.2, so that the fastest iteration process is ensured, and the calculation accuracy is ensured.

[0081] S5: After the radiation flux density of each face element is calculated, the bidirectional reflectance BRF of the face element under different observation angles can be further calculated, and the calculation formula of the BRF is as follows:

[0082] ;

[0083] Wherein, represents the observation angle, represents the bidirectional reflectance at the observation angle, represents the normal direction of the i-th face element, represents the area of the i-th face element seen from the observation angle .

[0084] The three-dimensional radiance calculation method of the high-resolution remote sensing scene can optimize the traditional RGM model, and an improved GPU-based three-dimensional radiance graphic model (RGM PS ) can be constructed, and the model can be applied to complex heterogeneous canopy simulation.Through simulation using the model, the calculation speed of the visibility factor can be improved, the processing accuracy of the model can be ensured, and in addition, due to the great improvement in the calculation speed of the visibility factor, the RGM PS model can significantly improve the complexity of the simulating scene from tens of thousands of faces of the original model to 14 million faces, so that the simulation demand of a large range and high-resolution remote sensing scene can be met.

[0085] The three-dimensional radiance calculation method of the high-resolution remote sensing scene provided by the embodiment of the application has been experimentally verified, and the RGM PS model is constructed by the above method, and through real single tree experiments and synthetic heterogeneous canopy experiments, the correlation (R²) between the RGM PS model and the original RGM model is more than 0.94, the root mean square error (RMSE) is less than 0.0038, and the cross-validation with the DART model shows high consistency (R² is as high as 0.98). The airborne multi-angle measurement evaluation result shows that the RGM PS model has good accuracy in BRF simulation, and the RMSE of the red light and near-infrared waveband is 0.0031 and 0.0340 respectively.

[0086] Specific verification experiments and experimental results are as follows:

[0087] Experiment goal one: verify the progress of the RGM PS model in calculation speed and accuracy.

[0088] The experimental simulation target is a real single-tree scene, and the performance efficiency of the original RGM model and the RGM PS The experiment was performed on a DELL notebook computer equipped with an Intel1 Core2 i7-12700H CPU @2.30 GHz, 32G RAM, and a NVIDIA GeForce RTX3060 GPU. The BRF simulation time of the original RGM model was 2 hours 53 minutes 18 seconds (10398 seconds), while the RGM PS model only took 2 minutes 32 seconds (152 seconds), achieving about 68.41 times acceleration.

[0089] Experimental target two: verify the effectiveness of the acceleration strategy under the OpenACC parallel framework.

[0090] To verify the effectiveness of different improvement strategies, an ablation experiment was conducted to determine the performance characteristics:

[0091] The speedup ratio of the parallel loop instruction was 4.32; the speedup ratio of the data layout optimization was 4.30; the speedup ratio of the data transmission minimization was 3.57; the speedup ratio of the data distribution was 1.33; the speedup ratio of the improved polygon filling algorithm was 53.68; and the speedup ratio of the new surface element radiance flux density solving equation was 1.21. The polygon filling algorithm and the data structure based on hash mapping had a significant impact on performance efficiency, showing significant acceleration effect. In contrast, the parallel loop instruction and the optimized data layout strategy, which are common transplantation operations of the OpenACC framework, showed suboptimal acceleration effect, achieving about 4.3 times improvement. The remaining three strategies showed relatively moderate parallel computing performance, but collectively contributed about 6 times overall computing speed improvement to the RGM PS model. The experiment studied different surface element scales from 1000 to 100000 to investigate the efficiency of the acceleration strategy. The results confirmed that the overall speedup ratio remained stable throughout the experiment, ranging from 60.2 to 71.1 times.

[0092] Experimental target three: cross-validation with the existing three-dimensional radiation transfer model DART

[0093] The RGM PS model and the DART model were used to simulate the reflectance results as shown in Figure 3 Figure 3 where a, b, c, and d are the principal plane reflectance results, and the principal plane is the plane containing the sensor optical axis and the nadir point, which can be understood as the plane containing the sun, the sensor, and the ground. Figure 3 ​In the figure, e, f, g, and h represent the reflectance results of the intersecting principal plane, which is a plane perpendicular to the principal plane and containing the nadir. As can be seen from the figure, there is a significant similarity between the two models; the overall shape and trend of the reflectance BRF curve as a function of the observation azimuth are comparable. Therefore, the RGM can be obtained. PS The conclusion that the simulated reflectance values ​​obtained from the DART model are highly consistent with those obtained from the present invention verifies that the method of the present invention significantly improves the calculation speed without reducing the accuracy of reflectance simulation calculation.

[0094] Experimental Objective 4: To conduct a comparative experiment with ground-based airborne measurements.

[0095] like Figure 4 As shown, RGM PS The simulation results of the DART and 4SAIL models are compared and analyzed with the airborne measurement results in real-world scenarios. For the red band and near-infrared band, RGM... PS The model's R-squared for the canopy BRF 2 The values ​​were 0.90 and 0.83, respectively, and the RMSE values ​​were 0.0031 and 0.0340, respectively. The results showed a high degree of consistency with the airborne measurement results in the real scene, proving that the RGMPS model has high spatial simulation accuracy and can be used for simulation of large-scale, high-resolution remote sensing scenes such as vegetation, forests and urban scenes.

[0096] In summary, the above description is merely a preferred embodiment of this specification and is not intended to limit the scope of protection of this specification. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of protection of this specification.

[0097] The systems, apparatuses, modules, or units described in one or more of the above embodiments may be implemented by a computer chip or entity, or by a product having a certain function. A typical implementation device is a computer. Specifically, a computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0098] It is also to be noted that the terms "comprising", "including", and any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises a... " does not, without more constraints, exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the recited element.

[0099] Each of the embodiments in the present specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other. Each of the embodiments focuses on the difference from other embodiments. In particular, for the system embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the part of the method embodiments.

[0100] The above describes specific embodiments of the present specification. Other embodiments are within the scope of the appended claims. In some cases, the acts or steps recited in the claims can be performed in a different order than the order in which the acts or steps are recited in the embodiments and still achieve desirable results. In addition, the processes depicted in the figures do not necessarily require the particular order shown, or sequential order, to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous.

Claims

1. A method for high-resolution 3D radiance computation of a remotely sensed scene, characterized in that, The method comprises the following steps: generating a triangular facet mesh by defining vertices in a three-dimensional space to create a three-dimensional scene; calculating the flux density of the facet self-emission under the sunlight based on the optical parameters of the facet; calculating the visibility factor through an OpenACC parallel framework, including: transferring the facet data for calculating the visibility factor from the CPU to the GPU, distributing the calculation tasks of the visibility factor cycle iteration between different facets to multiple threads of the GPU, and accelerating the calculation of the visibility factor; calculating the radiation flux density of the facet by using the successive over-relaxation method, and the iterative calculation formula is: ； wherein, represents a relaxation factor, represents the k+1 iteration radiative flux density of the i-th surface element, represents the k iteration radiative flux density of the i-th surface element, represents the k iteration radiative flux density of the j-th surface element, represents the reflectivity or transmissivity of the j-th surface element, and N represents the total number of surface elements, represents the visibility factor from the j-th surface element to the i-th surface element, represents the flux density of the i-th surface element spontaneously radiating under solar irradiation; calculating the bidirectional reflectance distribution function (BRF) under different observation angles based on the radiation flux density of each facet, and the calculation formula is: ； in, Indicates the observation angle. Indicates in Diaxial reflectivity at an angle Indicates the direction of the normal to the i-th face element. Indicates from the perspective of observation The area of ​​the i-th facet seen. Let represent the radiative flux density of the i-th surface element.

2. The method of claim 1, wherein, the calculation formula of the flux density of the facet self-emission under the sunlight is: ； wherein, represents the area of the i-th facet, represents the flux density of spontaneous emission at an arbitrary point on the i-th facet spontaneous emission.

3. The method of claim 1, wherein, the calculation formula of the visibility factor is: ； wherein, denotes the derivative with respect to time, denotes the derivative with respect to time, denotes and the angle between the line connecting and the normal to the surface of denotes and the angle between the line connecting and the normal to the surface of denotes and the Euclidean distance between 4. The method of claim 3, wherein, Under the OpenACC parallel framework, at least one of the following strategies is used to accelerate the calculation of the field of view factor: parallel loop instruction, optimized data layout, minimized data transmission and dynamic memory allocation.

5. The method of claim 4, wherein, The region of the facet is filled by using a polygon filling algorithm, and the Cartesian coordinates of any point in the three-dimensional space are converted into barycentric coordinates, and the barycentric coordinates are used to determine whether the point is located in a certain facet.

6. The method of claim 5, wherein, In the polygon filling algorithm, the coordinates of the three vertices A, B and C of the facet are used to represent an arbitrary point P as: ； wherein and represent the barycentric coordinates of point P; When and are both non-negative, then point P is inside the facet; otherwise, point P is outside the facet.

7. The method of claim 6, wherein, barycentric coordinates and The calculation formula is: ； wherein , and denote the components of the vectors , and in the x-direction, , and denote the components of the vectors , and in the y-direction.

8. The method of claim 1, wherein, In the process of calculating the visibility factor, the storage structure in the GPU is a hash-based data structure.

9. The method of claim 1, wherein, The range of the relaxation factor is 0 to 2, and the optimal relaxation factor is searched in the range of the relaxation factor with a step of 0.2 during the solution of the radiation flux density.

10. Use of a method for the calculation of the three-dimensional radiance of a high-resolution remotely sensed scene according to any one of claims 1 to 9, characterized in that, The three-dimensional radiance calculation method of the high-resolution remote sensing scene according to any one of claims 1 to 9 is used to construct a GPU-based three-dimensional radiance graphic model, and the three-dimensional radiance graphic model is used for heterogeneous canopy simulation.

Images