A sun-avoidance angle prediction method based on upper limit of background radiance

Abstract

The present application relates to the technical field of space optical detection, atmospheric radiation modeling and observation mission planning, and particularly relates to a solar avoidance angle prediction method based on a background radiance upper limit. A background radiance calculation model is constructed to calculate the background radiance corresponding to different solar-target angles in a known scene. The complex radiation transfer result is compressed into a unified analytical formula (i.e. the background radiance calculation model), and the background radiance can be calculated by substituting the angle, atmospheric visibility and scene category into the formula when calling online. The present application can directly solve the avoidance angle under the upper limit constraint, and is safe, convenient and does not require complex deployment. The present application solves the problems of poor real-time performance, safety and easy distortion in visible light observation missions of space targets.

Description

Technical Field

[0001] This invention relates to the fields of space optical detection, atmospheric radiation modeling and observation mission planning, and in particular to a method for predicting solar avoidance angle based on the upper limit of background radiance. Background Technology

[0002] In visible light observation missions of space targets, when the target observation direction is close to the sun, the forward scattering of solar radiation by atmospheric molecules and aerosols significantly increases the background radiance. For imaging recognition, target detection, and observation mission planning, what truly determines the feasibility of the mission is not a certain average background value, but rather the upper limit of the background radiance that may occur under given observation geometry and atmospheric conditions.

[0003] Existing engineering practices mainly fall into three categories: one is to directly call radiative transfer tools such as MODTRAN to calculate on a case-by-case basis; another is to create lookup tables or interpolation tables from the simulation results for offline use; and the third is to fit the background mean or empirical values ​​for a single angle. Although the first two methods have high accuracy, online calling is costly and deployment is complex, which is not conducive to real-time task evaluation; the third method, although simple to calculate, often describes the average trend rather than the background upper limit, making it difficult to directly use for determining the safety constraints of the solar avoidance angle.

[0004] Furthermore, traditional empirical formulas often use only the sun-target angle (or simply angle) as a single independent variable, which makes it difficult to simultaneously reflect changes in visibility and differences in typical surface / aerosol scenarios. In particular, at the tail end of large angles, trend distortion or derivative divergence can easily occur, making it impossible to stably support the estimation of engineering safety boundaries. Summary of the Invention

[0005] To overcome the problems of poor real-time performance, poor security, and easy distortion in the existing technology for visible light observation of space targets, this invention proposes a solar avoidance angle prediction method based on the upper limit of background radiance. This method can directly solve for the avoidance angle under the upper limit constraint, which is safe, convenient, and does not require complex deployment.

[0006] This invention proposes a solar avoidance angle prediction method based on the upper limit of background radiance, and constructs a background radiance calculation model to calculate the background radiance corresponding to different sun-target angles in a known scene; The background radiance calculation model is as follows: ; Where θ is the angle between the sun and the target, or simply the angle; V is the atmospheric visibility; and s is the scene category. Background radiance; Main lobe amplitude; The amplitude function for the background base; Characterizing large-angle shoulder compensation under different atmospheric visibility conditions; , and All are correlation factors of atmospheric visibility V under scenario s; Let x be the correlation constant under scene s, and let x be the correlation factor of the included angle θ, x = (1 - cosθ) / 2; HG functions that associate the included angle with the scene, Let be the correlation constant under scenario s; and These are the minimum and maximum angle suppression factors for scenario s, respectively.

[0007] Preferred: ; ; ; in, , , , , and It is a fixed constant for the associated scenario s.

[0008] Preferred: ; is a fixed constant for the associated scenario s; k is a fixed constant common to the scenario.

[0009] Preferred: ; ; in, This is the intermediate value of the included angle j; and Both are constants relating the included angle j and the scene s, where b represents the minimum included angle and m represents the maximum included angle; This represents the Sigmoid function; y is the correlation factor of the included angle θ, y = (1 + cosθ) / 2; Let be the correlation constant under scenario s.

[0010] Preferably, simulation experiments are conducted for each scenario, and the simulation dataset {θ,V;} is compiled. } s Fit a background radiance calculation model on the corresponding dataset.

[0011] The preferred background radiance calculation model for desert environments is as follows: ; ; ; ; ; ; in, and This represents the constant value of the minimum included angle b in the corresponding scenario. and This represents the constant value of the maximum included angle m in the corresponding scenario.

[0012] The preferred model for calculating background radiance in a marine environment is: ; ; ; ; ; ; in, and This represents the constant value of the minimum included angle b in the corresponding scenario. and This represents the constant value of the maximum included angle m in the corresponding scenario.

[0013] The preferred background radiance calculation model for rural environments is: ; ; ; ; ; ; in, and This represents the constant value of the minimum included angle b in the corresponding scenario. and This represents the constant value of the maximum included angle m in the corresponding scenario.

[0014] Preferably, the scene category and atmospheric visibility are first determined, and then the background radiance calculation model for the corresponding scene is substituted into the model. The minimum included angle θ is determined according to the following optimization objectives. min : ; ; Where θ0 is the set threshold; The set allowable background threshold; The ratio coefficient between the manually set associated scene s and atmospheric visibility V; ; For safe background radiance.

[0015] The present invention proposes a solar avoidance angle prediction system based on the upper limit of background radiance, comprising a memory and a processor. The memory stores a computer program, and the processor is connected to the memory. The processor is used to execute the computer program to implement the solar avoidance angle prediction method based on the upper limit of background radiance.

[0016] The advantages of this invention are: (1) This invention replaces a large number of case-by-case radiative transfer calculations with a unified analytical formula, so that the system can directly obtain the background radiance after inputting the sun-target angle θ, visibility V and scene category s. This invention directly models the engineering safety boundary quantity of "upper limit of background radiance" rather than the average background quantity, thus directly serving the determination of the minimum solar avoidance angle. This method is used to quickly predict the upper limit of background radiance in the target's line of sight based on the sun-target angle, visibility, and scene category, and further calculates the minimum solar avoidance angle under the acceptable background threshold of the camera system. It implements a physically inspired architecture with strong interpretability.

[0017] (2) This invention compresses complex radiative transfer results into a unified analytical formula (i.e., a background radiance calculation model). When called online, only the included angle θ, atmospheric visibility V, and scene category s need to be substituted for calculation. It is suitable for integration into task planning software, load constraint analysis programs, and embedded decision modules. The parent structure of this invention consists of the forward scattering main lobe amplitude... Background base amplitude function Large-angle shoulder compensation and endpoint convergence terms , The composition has a clear structural explanation, avoiding the trend distortion and tail divergence problems commonly found in simple polynomial fitting in high-angle intervals.

[0018] (3) The present invention further includes a system-allowed background threshold. Quickly obtain the minimum solar avoidance angle θ min This invention can also incorporate a safety factor K by combining publicly available real observation data. sf(V,s) This leads to more conservative engineering avoidance boundaries. Attached Figure Description

[0019] Figure 1Here is a flowchart of a solar avoidance angle prediction method based on the upper limit of background radiance proposed in this invention; Figure 2 This is a comparison chart of ablation experiments. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0021] like Figure 1 As shown in the figure, the solar avoidance angle prediction method based on the upper limit of background radiance proposed in this embodiment first constructs a background radiance calculation model to calculate the background radiance under a known scene; the defined parameters of the known scene include the sun-target angle (referred to as the angle) θ and the atmospheric visibility V; then the background radiance calculation model is used to predict the background radiance corresponding to the angle, and the angle is adjusted until the background radiance falls into the set background radiance safety range.

[0022] The construction of the background radiance calculation model includes the following steps S1-S3.

[0023] S1. Determine the correlation factor of background radiance and construct the mapping function: (1) Where θ is the angle between the sun and the target, or simply the angle; V is the atmospheric visibility; and s is the scene category. The background radiance.

[0024] S2. Determine a common background radiance calculation model for various scenarios: (2) in, Used to characterize the amplitude of the main lobe of forward scattering from the solar neighborhood (referred to as main lobe amplitude). The amplitude function for the background base; Used to characterize large-angle shoulder compensation under different atmospheric visibility conditions. Used to improve large-angle tail trends. Let x be the correlation constant under scenario s, and let x be the correlation factor of the sun-target angle θ, x = (1 - cosθ) / 2; , and All are correlation factors of atmospheric visibility V under scenario s; Here, HG (Henyey-Greenstein) function is the correlation factor between the angle and the scene, and θ is the correlation factor between the sun and the target angle. Let be the correlation constant under scenario s; The inhibition factor for minimizing θ under scene s, also known as the minimum angle inhibition factor; The suppression factor for θ to reach its maximum value under scene s, also known as the maximum angle suppression factor; Right now: (3) (4) (5) (6) (7) , and For the mapping function corresponding to scene s, and This is a mapping function applicable to all scenarios.

[0025] S3. Conduct simulation experiments for each scenario and compile the simulation dataset {θ,V; } s ; Fit formulas (2) to (7) on the corresponding dataset to obtain the background radiance calculation model for each scene.

[0026] In practice, this step can further determine the form of the above formulas (3)-(7), for example: (3.1) (4.1) (5.1) (6.1) (7.1) (7.2) in, , , , , , , , and For the associated scene s, a fixed constant; This is the intermediate value of the included angle j; and Both are constants relating the included angle j and the scene s, where b represents the minimum included angle and m represents the maximum included angle; This represents the Sigmoid function; k is a fixed constant generally used in the scenario. y is the correlation factor of the sun-target angle θ, y = (1 + cosθ) / 2; Let be the correlation constant under scenario s.

[0027] The following specific embodiments are used to verify the above-mentioned solar avoidance angle prediction method based on the upper limit of background radiance.

[0028] In this embodiment, a simulation experiment was conducted in a visible light environment, setting up 3 scenes and 13 atmospheric visibility levels; The three scenes are: desert, ocean, and countryside; The 13 atmospheric visibility levels are: 2, 4, 5, 7.5, 10, 12, 15, 18, 20, 22, 25, 28, and 30; the unit is kilometers (km).

[0029] In this embodiment, parameter fitting is first performed for each scenario. Specifically, for each scenario, a formula is fitted using the corresponding 13 atmospheric visibility simulation data. The fitting results for the parameters are shown in Table 1 below: Table 1: Simulation Experiment Fitting Results

[0030] When the scene category s = desert, the background radiance calculation model is as follows: ; ; ; ; W desert(V) =exp(3.069990 + 0.033470 lnV); D desert(V) = exp(3.227031 + 0.011004 lnV); S desert(V) = exp(4.372469 - 0.322714 lnV); = σ(-2.954418 - 0.464557 lnV); = σ(-1.102110 - 0.050193 lnV); g desert = 0.530664; = 0.339337; p desert = 0.718130; in, The visible light background radiance in a desert scene. W desert(V) D desert(V) and S desert(V) These are the HG function, main lobe amplitude, background base amplitude function, and large-angle shoulder compensation for desert scenes; T b,desert(V) and These are the minimum angle suppression factor and intermediate value in the desert scenario, respectively; T m,desert(V) and These are the maximum angle suppression factor and intermediate value in the desert scene, respectively; g desert , and p desert This is a constant for desert scenarios.

[0031] When the scene category s = ocean, the background radiance calculation model is as follows: ; ; ; ; W ocean(V) = exp(2.926328 + 0.104183 lnV); D ocean(V) = exp(4.463416 - 0.773614 lnV); S ocean(V) = exp(3.232939 - 0.085632 lnV); = σ(-0.014173 - 1.264808 lnV); = σ(-3.945933 - 0.383476 lnV); g ocean = 0.551212; = 0.605744; p ocean = 0.973635; in, The visible light background radiance in an ocean scene. W ocean(V) D ocean(V) and S ocean(V)These are the HG function, main lobe amplitude, background base amplitude function, and large-angle shoulder compensation for marine scenes; T b,ocean(V) and These are the minimum angle suppression factor and intermediate value in the marine scenario, respectively; T m,ocean(V) and These are the maximum angle suppression factor and intermediate value in the marine scenario, respectively; g ocean , and p ocean This is a constant for the marine scenario.

[0032] When the scene category s = rural, the background radiance calculation model is as follows: ; ; ; ; W rural(V) = exp(3.021878 + 0.127315 lnV); D rural(V) = exp(4.282872 - 0.673943 lnV); S rural(V) = exp(3.374096 - 0.145799 lnV); = σ(0.082892 - 1.667010 lnV); = σ(-0.413678 - 0.753088 lnV); g rural = 0.458785; = 1.996516; p rural = 0.621052; in, For visible light background radiance in rural scenes, W rural(V) D rural(V) and S rural(V) These are the HG function, main lobe amplitude, background base amplitude function, and large-angle shoulder compensation for rural scenes; T b,rural(V) and These are the minimum angle inhibition factor and intermediate value in the rural scenario, respectively; T m,rural(V) and These are the maximum angle inhibition factor and intermediate value in the rural scenario, respectively; g rural , arural and p rural This is a constant for rural settings.

[0033] This embodiment proposes a solar avoidance angle prediction method based on the upper limit of background radiance. In further implementation, it first obtains the background radiance calculation model for the corresponding scene, and then combines the following optimization target search to find the optimal angle θ. min : ; ; Where θ0 is a set threshold, representing the lower limit of the included angle θ, which can be set to θ0=15°. This is the set allowable background threshold, which is an inherent parameter of the camera; As an intermediate variable; The ratio coefficient between the manually set associated scene s and atmospheric visibility V; This is used to reserve conservative avoidance boundaries.

[0034] This can be determined based on publicly available, real-world observations. Current consistency results from publicly available real-world data indicate that, under representative medium-to-high visibility scenarios, Typically, a value of 1.0-1.5 is sufficient to create a conservative avoidance boundary with approximately 99% accuracy; under low visibility conditions, It needs to be increased accordingly, and generally increases as visibility decreases.

[0035] In this embodiment, the engineering consistency of the fitting results shown in Table 1 is verified using publicly available AERONET PPL00 principal-plane real observation data. It should be noted that AERONET provides the nominal center wavelength spectral background radiance for two narrowband channels at 440nm and 675nm, rather than the readily available 400-750nm broadband visible light background radiance. Therefore, this embodiment uses a dual-channel trapezoidal approximation to construct the visible light engineering surrogate quantity, i.e.: L vis_proxy = 0.01 x (1.575 x L 440 + 1.925 x L 675 ); Among them, L 440 and L 675 The background spectral radiance at 440 nm and 675 nm, respectively, provided by the AERONET dataset, are L vis_proxy denoted as visible light background radiance. The coefficients in this formula are experimental constants.

[0036] In this embodiment, the original formula achieved 87.91% point-level envelope coverage of the selected representative real data on the entire AERONET dataset, meaning that 87.91% of the dataset samples could satisfy the background radiance calculation model for the corresponding scene, and only 12.09% of the sampling points exceeded the curve; thus proving the usability of the present invention.

[0037] In the AERONET dataset, a representative site was selected for validation for ocean, desert, and rural areas respectively. The results are shown in Table 2 below.

[0038] Table 2: Validation results of the AERONET dataset

[0039] In this embodiment, an optimization objective is adopted. Search θ1 min , adopt optimization objective Search θ2 min Set θ0 = 15°; In Table 2, the envelope ratio of the original curve refers to the "angle θ - visible light background radiance L" of the sample curve in the corresponding scene. vis_proxy The angle θ1 between the curve and the predicted curve calculated according to the formula in the corresponding scenario shown in Table 1 min -Visible background radiance "The probability of a match; 99% envelope rate corresponds to the proportionality coefficient This refers to the predicted curve "angle θ2" obtained using the corresponding scaling factor. min -Visible background radiance "Angle θ with the sample curve in the corresponding scene - visible light background radiance L" vis_proxy "The matching rate can reach 99%."

[0040] As can be seen from Table 2, the predicted angle is in high agreement with the actual angle, which supports the usability of the method of the present invention as a method for predicting engineering upper limits and evaluating avoidance angles.

[0041] In this embodiment, an ablation experiment was also conducted.

[0042] Compared to the method of this invention, the ablation model simplifies the background radiance calculation model by removing the endpoint convergence constraint. The ablation model is as follows: ; In this embodiment, parameter fitting and model validation were performed based on 39 full-scale visible light simulation scenarios. The method of this invention was compared with the ablation model. The average R² of the back-substitution fitting of the method of this invention was 0.9958, while that of the ablation model was 0.9816. The average NRMSE of the method of this invention in the high-angle region (120°-180°) was 0.0198, significantly lower than the 0.1471 of the ablation model. In the tail section of 150°-180°, the average NRMSE of the method of this invention was 0.0305, significantly lower than the 0.2358 of the ablation model. See details... Figure 2 .

[0043] In the method of this invention, by adjusting the atmospheric visibility V, it can be seen that the atmospheric visibility in the worst simulation scenario under marine environment is 4km, and its R² is still 0.9797, indicating that the method of this invention is more reliable in describing the large-angle tail trend and endpoint convergence.

[0044] This embodiment also employs the leave-one-out method for verification. The leave-one-out method involves combining the scene and atmospheric visibility to obtain 39 samples. Simulation fitting is performed on 38 of these samples, with the remaining sample used as a test sample. In this verification, the average indicators were calculated across all 39 test samples, yielding the following results: average R² of 0.9945, average NRMSE of 0.0136, average NRMSE of 0.0232 in the high-angle region, and average NRMSE of 0.0330 in the 150°-180° tail region. Even the worst verification scene (rural scene, atmospheric visibility 2km) still showed an R² of 0.9755, indicating that the model does not rely solely on a single scene for backfitting.

[0045] Of course, those skilled in the art will recognize that the present invention is not limited to the details of the exemplary embodiments described above, but also includes the same or similar structures that can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0046] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

[0047] The technologies, shapes, and structures not described in detail in this invention are all known technologies.

Claims

1. A method for predicting solar avoidance angle based on the upper limit of background radiance, characterized in that, A background radiance calculation model is constructed to calculate the background radiance corresponding to different sun-target angles in a known scene. The background radiance calculation model is as follows: Where θ is the angle between the sun and the target, or simply the angle; V is the atmospheric visibility; and s is the scene category. Background radiance; Main lobe amplitude; The amplitude function for the background base; Characterizing large-angle shoulder compensation under different atmospheric visibility conditions; , and All are correlation factors of atmospheric visibility V under scenario s; Let x be the correlation constant under scene s, and let x be the correlation factor of the included angle θ, x = (1 - cosθ) / 2; HG functions that associate the included angle with the scene, Let be the correlation constant under scenario s; and These are the minimum and maximum angle suppression factors for scenario s, respectively.

2. The solar avoidance angle prediction method based on the upper limit of background radiance as described in claim 1, characterized in that: in, , , , , and It is a fixed constant for the associated scenario s.

3. The solar avoidance angle prediction method based on the upper limit of background radiance as described in claim 1, characterized in that: is a fixed constant for the associated scenario s; k is a fixed constant common to the scenario.

4. The solar avoidance angle prediction method based on the upper limit of background radiance as described in claim 3, characterized in that: in, This is the intermediate value of the included angle j; and Both are constants relating the included angle j and the scene s, where b represents the minimum included angle and m represents the maximum included angle; This represents the Sigmoid function; y is the correlation factor of the included angle θ, y = (1 + cosθ) / 2; Let be the correlation constant under scenario s.

5. The solar avoidance angle prediction method based on the upper limit of background radiance as described in any one of claims 1-4, characterized in that, Simulation experiments were conducted for various scenarios, and the simulation dataset {θ,V;} was compiled. } s Fit a background radiance calculation model on the corresponding dataset.

6. The solar avoidance angle prediction method based on the upper limit of background radiance as described in claim 5, characterized in that, The background radiance calculation model in a desert environment is as follows: in, and This represents the constant value of the minimum included angle b in the corresponding scenario. and This represents the constant value of the maximum included angle m in the corresponding scenario.

7. The solar avoidance angle prediction method based on the upper limit of background radiance as described in claim 5, characterized in that: The background radiance calculation model in a marine environment is as follows: in, and This represents the constant value of the minimum included angle b in the corresponding scenario. and This represents the constant value of the maximum included angle m in the corresponding scenario.

8. The solar avoidance angle prediction method based on the upper limit of background radiance as described in claim 5, characterized in that: The background radiance calculation model for rural environments is as follows: in, and This represents the constant value of the minimum included angle b in the corresponding scenario. and This represents the constant value of the maximum included angle m in the corresponding scenario.

9. The solar avoidance angle prediction method based on the upper limit of background radiance as described in claim 5, characterized in that, First, determine the scene category and atmospheric visibility, then substitute them into the background radiance calculation model for the corresponding scene, and determine the minimum included angle θ based on the following optimization objectives. min : Where θ0 is the set threshold; The set allowable background threshold; The ratio coefficient between the manually set associated scene s and atmospheric visibility V; ; For safe background radiance.

10. A solar avoidance angle prediction system based on the upper limit of background radiance, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, the processor is connected to the memory, and the processor is used to execute the computer program to implement the solar avoidance angle prediction method based on the upper limit of background radiance as described in any one of claims 6-9.

Images