A fast cloud judgment method for satellite on-orbit real-time cloud detection

Abstract

The application discloses a kind of for satellite in-orbit real-time cloud detection fast cloud determination method, belong to optical remote sensing image processing field, the method utilizes the principle of cloud top reflection sunlight, by establishing look-up table in ground platform, the method is realized in-orbit fast cloud determination by storing look-up table information in cloud determination load computing platform.The application according to the reflectivity of cloud layer in visible light compared to ground object in remote sensing image is higher, the thicker cloud layer is, the stronger reflection ability is, the greater albedo is, by establishing the relationship between the luminance information represented by observation instrument response DN value and cloud optical thickness, the cloud optical thickness is inverted by response DN value, to judge the cloud coverage of observation target area according to cloud optical thickness, in the case where guaranteeing identification accuracy, the speed of satellite in-orbit cloud detection is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of optical remote sensing image processing, and particularly relates to a fast cloud judgment method for satellite on-orbit real-time cloud detection. BACKGROUND

[0002] Optical remote sensing satellite images are widely used in the fields of agriculture, environment, disaster prevention, surveying and mapping, and battlefield reconnaissance. With the development of remote sensing technology, the performance of optical remote sensing devices is continuously improved, such as spatial resolution, temporal resolution, and spectral resolution, and higher requirements are put forward for data quality and utilization. However, the global annual average cloud coverage rate is more than 60%, and due to the existence of a large number of clouds in high-resolution optical remote sensing data, the quality of remote sensing images is affected, thereby reducing the data utilization rate of the images, wasting valuable satellite resources, and reducing the service life of the satellite. Therefore, cloud detection and cloud removal have become important problems that need to be solved in the process of remote sensing image processing, and the technical route mainly includes ground processing and on-board processing. Traditional cloud detection methods mostly receive compressed image data through satellite stations, and then the decompressed images are handed over to the ground data processing center for a series of processing by data processing software, and finally the data products are generated and provided to users. With the development of technology, users have put forward higher requirements for the real-time performance and data utilization rate of optical remote sensing satellite data processing. Most existing satellite payloads generally use threshold methods for business cloud detection, but the algorithm is relatively complex, and the satellite platform needs to provide sufficient computing power, and the calculation time is relatively long. At the same time, with the development of deep learning, there are also many cloud judgment methods based on deep learning, but there are problems such as large computing power requirement and difficulty in platform implementation. SUMMARY

[0003] To solve the problems existing in the prior art, the present application provides a fast cloud judgment method for satellite on-orbit real-time cloud detection, so as to process optical remote sensing satellite images on-orbit in real time, detect the cloud coverage rate of the observed target area in advance, and solve the problems of the existing methods, such as inability to process on-orbit in real time, large computing power requirement, great difficulty in on-orbit implementation, and high cost.

[0004] The solution adopted by the present application to solve the technical problems is:

[0005] The fast cloud judgment method for satellite on-orbit real-time cloud detection of the present application is characterized by being applied to an environment composed of a ground computing platform, a cloud judgment payload computing platform, and a satellite platform, and includes the following steps:

[0006] Step 1: According to the surface albedo γ of the selected cloud judgment detection band, the solar zenith angle θ solar , the observation zenith angle θ view , and the relative azimuth angle A lookup table is established on the ground computing platform to determine the cloud optical thickness τ.

[0007] Step 1.1 First, set the array of cloud optical thickness parameters. Array of surface albedo parameters for the selected cloud detection band Array of solar zenith angle parameters Array of observed zenith angle parameters Array of relative azimuth parameters Where, τ q Represents the array τ[n CT The q-th cloud optical thickness parameter, n CT The array τ[n] represents the cloud optical thickness parameter. CT The total number of parameters in ], γ p Represents the array γ[n A The p-th surface albedo parameter in ], n A Represents the array of surface albedo parameters γ[n A The total number of parameters in the [] section; Represents array θ solar [n SZ The j-th solar zenith angle parameter in ], n SZ Represents the array of solar zenith angle parameters θ solar [n SZ The total number of parameters in the [] section; Represents array θ view [n VZ The m-th observed zenith angle parameter in ], n VZ Represents the array of observed zenith angle parameters θ view [n VZ The total number of parameters in the [] section; Represents array The i-th relative azimuth parameter in n VA Represents the array of relative azimuth parameters The total number of parameters;

[0008] Step 1.2 Establish the response value DN at pixel position (x,y) in the image data received by the cloud judgment load calculation platform according to equation (1). x,y and the radiance L it represents x,y Relationship:

[0009]

[0010] In equation (1), k represents the absolute radiation response coefficient. This represents the background response value at pixel location (x, y) in the received image data;

[0011] Step 1.3 Based on radiance Lx,y prior information, the radiance value L q at the cloud optical thickness τ q is selected from the radiance of the image data received by the cloud payload calculation platform, so as to obtain the array of radiance typical value parameters L[n CT ] = {L1, L2…L q …L nCT}, and the relationship function f1 of the array L[n CT ] and the array τ[n CT ] is constructed by using formula (2):

[0012] L q = f1(τ q ) (2)

[0013] Step 1.4, according to formula (3), the relationship function f2 of the radiance value L q , the surface albedo γ, the solar zenith angle θ solar , the observation zenith angle θ view , and the relative azimuth angle φ is established at the cloud optical thickness τ = τ q :

[0014]

[0015] Step 1.5, according to formula (4), the radiance value L p at the cloud optical thickness τ = τ q,p and the surface albedo γ = γ q is obtained:

[0016]

[0017] Step 1.6, according to formula (5), the radiance value L p at the cloud optical thickness τ = τ q,p,n , the surface albedo γ = γ q , the solar zenith angle θ , and the observation zenith angle θ is obtained:

[0018]

[0019] Step 1.7, according to formula (6), the radiance value L p at the cloud optical thickness τ = τ q,p,n,m , the surface albedo γ = γ q , the solar zenith angle θ , the observation zenith angle θ , and the relative azimuth angle φ is obtained:

[0020]

[0021] Step 1.8 According to equation (7), the optical thickness τ = τ in the cloud is obtained. q Surface albedo γ = γ p Sun zenith angle Observation zenith angle relative azimuth Radiance value L at time q,p,n,m,i :

[0022]

[0023] Step 1.9 Based on the cloud optical thickness input parameters, surface albedo input parameters of the selected cloud detection band, solar zenith angle input parameters, observation zenith angle input parameters, and relative azimuth angle input parameters preset in Step 1.1, all input parameters are traversed sequentially according to Steps 1.3 to 1.8 to obtain all data of the lookup table;

[0024] Step 2: Store the global surface albedo information for different seasons and all data from the lookup table on the cloud-based load computing platform;

[0025] Step 3: The cloud-based payload calculation platform receives the information sent by the satellite platform and calculates the pixel coordinates of the observed target area.

[0026] Step 3.1 The cloud-based payload calculation platform receives the time information, satellite position information, satellite velocity information, latitude and longitude information of the observation target area, and four-element information of satellite attitude sent by the satellite platform;

[0027] Step 3.2 The cloud-based payload calculation platform calculates the satellite attitude rotation matrix based on the quaternion information of the satellite attitude;

[0028] Step 3.3 The cloud-based load calculation platform calculates the pixel coordinates in the pixel coordinate system corresponding to the geographic coordinates of the observed target area based on the latitude and longitude information, satellite position information, satellite velocity information, satellite attitude rotation matrix, cloud-based load intrinsic parameters, and distortion coefficients of the observed target area.

[0029] Step 4: The cloud-based payload calculation platform receives remote sensing image data and performs real-time cloud-based on-orbit analysis;

[0030] Step 4.1 The cloud-based payload calculation platform receives the solar zenith angle of the observed target area in its current state from the satellite platform. Sun azimuth On-orbit correction coefficient of cloud-based instrument;

[0031] Step 4.2 The cloud judgment payload computing platform receives the pixel response value of the remote sensing image data of the observation target region, and obtains the pixel radiance value L of the observation target region after pre-processing the remote sensing image data r ;

[0032] Step 4.3 The cloud judgment payload computing platform inverses the pixel coordinate position corresponding to the geographic coordinates of the observation target region according to the lookup table to obtain the cloud optical thickness

[0033] Step 4.3.1 The cloud judgment payload computing platform calculates the observation zenith angle under the current state according to the time information, satellite position information, satellite speed, latitude and longitude information of the observation target region, and four-element information of the satellite attitude of the observation target region received by the satellite platform observation azimuth angle and reads the albedo information γ of the observation target region from the cloud judgment payload computing platform r ;

[0034] Step 4.3.2 The cloud judgment payload computing platform calculates the relative azimuth angle under the current state according to formula (8)

[0035]

[0036] Step 4.3.3 The cloud judgment payload computing platform inverses the cloud optical thickness τ of the observation target region according to the observation zenith angle under the current state sun zenith angle relative azimuth angle albedo information γ of the observation target region r and pixel radiance L r using the lookup table method and interpolation method and through formula (9) result :

[0037]

[0038] In formula (9), f3 represents a relationship function derived from f1, and f4 represents a relationship function derived from f2;

[0039] Step 4.4 According to the inversion result τ of the cloud optical thickness result and the three determination thresholds τ low , τ mid , τ high set, the cloud coverage of the pixel region corresponding to the geographic coordinates of the observation target region is calculated:

[0040] When τ result ≥ τ high , it is judged that the cloud coverage of the observation target region is determined to have cloud;

[0041] When τ high > τ result ≥ τ mid , it is determined that the cloud coverage of the observation target region is suspected to have clouds;

[0042] When τ mid > τ result ≥ τ low , the cloud coverage of the observation target region is determined to be a suspected clear sky;

[0043] When τ low > τ result , it is determined that the cloud coverage of the observation target region is determined to be a clear sky;

[0044] Step five: the cloud judgment payload computing platform transmits the cloud coverage of the pixel region corresponding to the geographic coordinates of the observation target region to the satellite platform.

[0045] The fast cloud judgment method for satellite on-orbit real-time cloud detection provided by the application is also characterized in that the size of the lookup table is n CT × n A × n SZ × n VZ × n VA × n bit , wherein n bit represents the number of stored data bits.

[0046] The pre-processing of the remote sensing image data in step 4.2 includes background subtraction, geometric correction and radiation correction of the image.

[0047] The electronic device provided by the application comprises a memory and a processor, and is characterized in that the memory is used to store a program supporting the processor to execute any of the fast cloud judgment methods, and the processor is configured to execute the program stored in the memory.

[0048] The computer readable storage medium provided by the application stores a computer program, and the computer program is characterized in that when the computer program is run by a processor, the steps of any of the fast cloud judgment methods are executed.

[0049] Compared with the prior art, the application has the following beneficial effects:

[0050] 1. The fast cloud judgment method for satellite on-orbit real-time cloud detection provided by the application, compared with the current traditional ground processing method, performs cloud recognition and cloud coverage rate discrimination on the target region in advance through on-orbit real-time cloud detection, thereby providing an observation time window and necessary pointing information for the main payload.

[0051] 2. The fast cloud judgment method for satellite on-orbit real-time cloud detection of the application, compared with the current traditional ground processing method, avoids the large amount of cloud conditions in the remote sensing data obtained by the main load through on-orbit real-time cloud detection, thereby improving the usability of the main load remote sensing data, reducing the data storage pressure and transmission pressure of the satellite platform when the main load data is sent to the ground station by the satellite platform, and reducing the bandwidth occupancy rate of invalid data.

[0052] 3. The fast cloud judgment method for satellite on-orbit real-time cloud detection of the application, which is inverted through cyclically nested data processing, and the cyclically nested process can accelerate the cloud judgment speed through data pipeline processing, thereby improving the on-orbit cloud judgment speed.

[0053] 4. The fast cloud judgment method for satellite on-orbit real-time cloud detection of the application, which simplifies the inversion process, effectively eliminates the problems of large computing power requirement, large on-orbit implementation difficulty and high hardware cost.

[0054] 5. In step 1.1 of the application, the number of parameters of the arrays τ[n CT ], γ[n A ], θ solar [n SZ ], θ view [n VZ ], can be pre-set, so that different storage size lookup tables can be realized, which is beneficial to realizing the optimal solution according to the storage space and computing capacity of the cloud judgment load calculation platform.

[0055] 6. In step two of the application, the cloud judgment load calculation platform stores albedo information of different seasons, which avoids cloud judgment errors caused by different ground albedos in different seasons, thereby improving the accuracy of cloud judgment.

[0056] 7. In step 4.2 of the application, the remote sensing image data is preprocessed, which reduces the radiation error and geometric error of the remote sensing image data, and is beneficial to improving the credibility of the data.

[0057] 8. In step five of the application, the cloud coverage of the observation target area is transmitted to the satellite platform, which is beneficial to the satellite platform distributing to other loads of the satellite, and is beneficial to the other loads switching on and off according to the cloud coverage, thereby improving the on-orbit service life of the satellite load. BRIEF DESCRIPTION OF DRAWINGS

[0058] Figure 1 is an embodiment schematic diagram of the fast cloud judgment method for satellite on-orbit real-time cloud detection of the application;

[0059] Figure 2 is a data pipeline processing mode diagram of the inversion process in the embodiment of the application. Detailed Implementation

[0060] In this embodiment, a rapid cloud judgment method for real-time cloud detection in satellite orbit is applied to an environment consisting of a ground computing platform, a cloud judgment payload computing platform, and a satellite platform, such as... Figure 1 As shown, it includes the following steps:

[0061] Step 1: Based on the selected cloud detection band, determine the surface albedo γ and the solar zenith angle θ. solar Observing the zenith angle θ view Relative azimuth A lookup table is established on the ground computing platform to determine the cloud optical thickness τ.

[0062] Step 1.1 First, set the array of cloud optical thickness parameters. Array of surface albedo parameters for the selected cloud detection band Array of solar zenith angle parameters Array of observed zenith angle parameters Array of relative azimuth parameters Where, τ q Represents the array τ[n CT The q-th cloud optical thickness parameter, n CT The array τ[n] represents the cloud optical thickness parameter. CT The total number of parameters in ], γ p Represents the array γ[n A The p-th surface albedo parameter in ], n A Represents the array of surface albedo parameters γ[n A The total number of parameters in the [] section; Represents array θ solar [n SZ The j-th solar zenith angle parameter in ], n SZ Represents the array of solar zenith angle parameters θ solar [n SZ The total number of parameters in the [] section; Represents array θ view [n VZ The m-th observed zenith angle parameter in ], n VZ Represents the array of observed zenith angle parameters θ view [n VZ The total number of parameters in the [] section; Represents array The i-th relative azimuth parameter in n VA Represents the array of relative azimuth parameters The total number of parameters.

[0063] In specific implementation, n is setCT = 11, n A = 7, n SZ = 6, n VZ = 8, n VA = 5, τ

[11] = {0.0125, 0.0125*2, 0.0125*3, 0.0125*4, 0.0125*5, 0.0125*7, 0.0125*9, 0.0125*12, 0.0125*15, 0.0125*20, 0.0125*50}, γ[7] = {0, 0.1, 0.2, 0.3, 0.5, 0.75, 1}, θ solar [6] = {5, 15, 30, 45, 60, 80}, θ view [8] = {1, 20, 30, 35, 40, 45, 60, 80},

[0064] Step 1.2. Establishing the relationship between the response value DN at the pixel position (x, y) in the image data received by the cloud discrimination payload computing platform according to formula (1) and the representative radiance L x,y : x,y

[0065]

[0066] In formula (1), k represents the absolute radiation response coefficient, represents the background response value at the pixel position (x, y) in the received image data.

[0067] Step 1.3. According to the prior information of the radiance L x,y , selecting the radiance value L q when the cloud optical thickness is τ q from the radiance of the image data received by the cloud discrimination payload computing platform, thereby obtaining the array of radiance typical value parameters L[n CT ] = {L1, L2…L q …L nCT}, and thereby constructing the relationship function f1 between the array L[n CT ] and the array τ[n CT ]:

[0068] L q = f1(τ q ) (2)

[0069] In this embodiment, the relationship function f1 is a linear function.

[0070] Step 1.4. According to formula (3), establishing the radiance value L q when the cloud optical thickness τ = τ q :with the surface albedo γ, the solar zenith angle θ solar , the observed zenith angle θ view , the relative azimuth angle φ the relationship function f2:

[0071]

[0072] Step 1.5 The radiance value L q at the cloud optical thickness τ = τ p , the surface albedo γ = γ q,p is obtained according to formula (4):

[0073]

[0074] Step 1.6 The radiance value L q,p,n at the cloud optical thickness τ = τ q , the surface albedo γ = γ p , the solar zenith angle θ is obtained according to formula (5):

[0075]

[0076] Step 1.7 The radiance value L q,p,n,m at the cloud optical thickness τ = τ q , the surface albedo γ = γ p , the solar zenith angle θ , the observed zenith angle θ , the relative azimuth angle φ is obtained according to formula (6):

[0077]

[0078] Step 1.8 The radiance value L q,p,n,m,i at the cloud optical thickness τ = τ q , the surface albedo γ = γ p , the solar zenith angle θ , the observed zenith angle θ , the relative azimuth angle φ is obtained according to formula (7):

[0079]

[0080] Step 1.9. According to the cloud optical thickness input parameters set in advance in step 1.1, the surface albedo input parameters of the selected cloud detection wave band, the solar zenith angle input parameters, the observation zenith angle input parameters, and the relative azimuth angle input parameters, all the data of the lookup table are obtained by traversing all the input parameters in sequence according to steps 1.3 to step 1.8. In this embodiment, the size of the lookup table is 11x7x6x8x5x16bit≈36.1KB.

[0081] Step two: store the global surface albedo information of different seasons and all the data of the lookup table on the cloud judgment payload computing platform.

[0082] Step three: the cloud judgment payload computing platform receives the information sent by the satellite platform and calculates the pixel coordinate position corresponding to the observation target area.

[0083] Step 3.1. The cloud judgment payload computing platform receives the time information, satellite position information, satellite speed information, latitude and longitude information of the observation target area, and four-element information of the satellite attitude sent by the satellite platform;

[0084] Step 3.2. The cloud judgment payload computing platform calculates the satellite attitude rotation matrix according to the four-element information of the satellite attitude;

[0085] Step 3.3. The cloud judgment payload computing platform calculates the pixel coordinate position corresponding to the geographical coordinates of the observation target area according to the latitude and longitude information of the observation target area, the satellite position information, the satellite speed information, the satellite attitude rotation matrix, the cloud judgment payload internal parameters and the distortion coefficient.

[0086] Step four: the cloud judgment payload computing platform receives the remote sensing image data and performs real-time cloud judgment on orbit;

[0087] Step 4.1. The cloud judgment payload computing platform receives the solar zenith angle of the observation target area in the current state sent by the satellite platform the solar azimuth angle the on-orbit correction coefficient of the cloud detector.

[0088] Step 4.2. The cloud judgment payload computing platform receives the pixel response value of the remote sensing image data of the observation target area, and obtains the pixel radiance value L r of the observation target area after pre-processing the remote sensing image data; in this embodiment, the pre-processing includes background subtraction, geometric correction and radiation correction.

[0089] Step 4.3. The cloud judgment payload computing platform inverses the pixel coordinate position corresponding to the geographical coordinates of the observation target area according to the lookup table to obtain the cloud optical thickness.

[0090] The step 4.3.1 cloud judgment load calculation platform calculates the observation zenith angle in the current state according to the time information of the observation target area, the satellite position information, the satellite speed, the latitude and longitude information of the observation target area and the four element information of the satellite attitude sent by the satellite platform received The observation azimuth angle And read the albedo information γ of the observation target area from the cloud judgment load calculation platform r ;

[0091] The step 4.3.2 cloud judgment load calculation platform calculates the relative azimuth angle in the current state according to formula (8)

[0092]

[0093] The step 4.3.3 cloud judgment load calculation platform calculates the cloud optical thickness inversion result τ of the observation target area according to the observation zenith angle in the current state The sun zenith angle The relative azimuth angle The albedo information γ of the observation target area r And the pixel radiance L r Using the look-up table method and the interpolation method and through formula (9), the cloud optical thickness inversion result τ of the observation target area is obtained result :

[0094]

[0095] In formula (9), f3 represents the relationship function according to f1, and f4 represents the relationship function according to f2.

[0096] In this embodiment, as shown in Figure 2 Clk represents the clock period, op1, op2 and op3 represent operation steps 1, 2 and 3 in one cycle respectively, RD0, CMP and WRO represent the read operation, the calculation operation and the write operation in one operation step respectively, and the data flow pipeline processing mode shown in Figure 2 The sequential execution is changed into the parallel mechanism by the data flow pipeline processing mode, the hardware resources of the cloud judgment load calculation platform can be fully utilized, the data throughput is effectively improved, and the cloud judgment speed is improved.

[0097] The step 4.4 calculates the cloud coverage of the pixel area corresponding to the geographical coordinates of the observation target area according to the cloud optical thickness inversion result τ result And the three judgment thresholds τ low , τ mid , τ high Set:

[0098] When τ result ≥ τ highAt that time, determining the cloud cover in the observed target area is considered to confirm the presence of clouds;

[0099] When τ high >τ result ≥τ mid At that time, the cloud cover in the observed target area was judged as suspected to be cloudy;

[0100] When τ mid >τ result ≥τ low At that time, the cloud cover in the observed target area was determined to be suspected to be clear sky;

[0101] When τ low >τ result At that time, the cloud cover of the observed target area is determined to be clear sky;

[0102] In this embodiment, a judgment threshold τ is set. low =0.000625, τ mid =0.001875, τ high =0.00375;

[0103] Step 5: The cloud-based payload calculation platform transmits the cloud coverage information of the pixel area corresponding to the geographic coordinates of the observed target area to the satellite platform.

[0104] In this embodiment, an electronic device includes a memory and a processor. The memory is used to store a program that supports the processor in executing the above-described fast cloud judgment method. The processor is configured to execute the program stored in the memory.

[0105] In this embodiment, a computer-readable storage medium stores a computer program, which is executed by a processor to perform the steps of the above-described fast cloud judgment method.

Claims

1. A rapid cloud judgment method for real-time on-orbit cloud detection of satellites, characterized by: It is applied in an environment consisting of a ground computing platform, a cloud-based payload computing platform, and a satellite platform, and includes the following steps: Step 1: Based on the selected cloud detection band, determine the surface albedo γ and the solar zenith angle θ. solar Observing the zenith angle θ view Relative azimuth A lookup table is established on the ground computing platform to determine the cloud optical thickness τ. Step 1.1 First, set the array of cloud optical thickness parameters. Array of surface albedo parameters for the selected cloud detection band Array of solar zenith angle parameters Array of observed zenith angle parameters Array of relative azimuth parameters Where, τ q Represents the array τ[n CT The q-th cloud optical thickness parameter, n CT The array τ[n] represents the cloud optical thickness parameter. CT The total number of parameters in ], γ p Represents the array γ[n A The p-th surface albedo parameter in ], n A Represents the array of surface albedo parameters γ[n A The total number of parameters in the [] section; Represents array θ solar [n SZ The j-th solar zenith angle parameter in ], n SZ Represents the array of solar zenith angle parameters θ solar [n SZ The total number of parameters in the [] section; Represents array θ view [n VZ The m-th observed zenith angle parameter in ], n VZ Represents the array of observed zenith angle parameters θ view [n VZ The total number of parameters in the [] section; Represents array The i-th relative azimuth parameter in n VA Represents the array of relative azimuth parameters The total number of parameters; Step 1.2 Establish the response value DN at pixel position (x,y) in the image data received by the cloud judgment load calculation platform according to equation (1). x,y and the radiance L it represents x,y Relationship: In equation (1), k represents the absolute radiation response coefficient. This represents the background response value at pixel location (x, y) in the received image data; Step 1.3 Based on radiance L x,y Based on prior information, the cloud optical thickness τ is selected from the radiance of the image data received by the cloud-based load computing platform. q Radiance value L at time q This yields an array of typical radiance parameters. Therefore, the array L[n] is constructed using equation (2). CT ] and array τ[n CT The relational function f1: L q =f1(τ q ) (2) Step 1.4 In the cloud optical thickness τ=τ q At that time, the radiance value L is established according to equation (3). q Related to surface albedo γ and solar zenith angle θ solar Observing the zenith angle θ view Relative azimuth Relational function f2: Step 1.5 According to equation (4), the cloud optical thickness τ = τ q Surface albedo γ = γ p Radiance value L at time q,p : Step 1.6 According to equation (5), the optical thickness τ = τ in the cloud is obtained. q Surface albedo γ = γ p Sun zenith angle Radiance value L at time q,p,n : Step 1.7 According to equation (6), the optical thickness τ = τ in the cloud is obtained. q Surface albedo γ = γ p Sun zenith angle Observation zenith angle Radiance value L at time q,p,n,m : Step 1.8 According to equation (7), the optical thickness τ = τ in the cloud is obtained. q Surface albedo γ = γ p Sun zenith angle Observation zenith angle relative azimuth Radiance value L at time q,p,n,m,i : Step 1.9 Based on the cloud optical thickness input parameters, surface albedo input parameters of the selected cloud detection band, solar zenith angle input parameters, observation zenith angle input parameters, and relative azimuth angle input parameters preset in Step 1.1, all input parameters are traversed sequentially according to Steps 1.3 to 1.8 to obtain all data of the lookup table; Step 2: Store the global surface albedo information for different seasons and all data from the lookup table on the cloud-based load computing platform; Step 3: The cloud-based payload calculation platform receives the information sent by the satellite platform and calculates the pixel coordinates of the observed target area. Step 3.1 The cloud-based payload calculation platform receives the time information, satellite position information, satellite velocity information, latitude and longitude information of the observation target area, and four-element information of satellite attitude sent by the satellite platform; Step 3.2 The cloud-based payload calculation platform calculates the satellite attitude rotation matrix based on the quaternion information of the satellite attitude; Step 3.3 The cloud-based load calculation platform calculates the pixel coordinates in the pixel coordinate system corresponding to the geographic coordinates of the observed target area based on the latitude and longitude information, satellite position information, satellite velocity information, satellite attitude rotation matrix, cloud-based load intrinsic parameters, and distortion coefficients of the observed target area. Step 4: The cloud-based payload calculation platform receives remote sensing image data and performs real-time cloud-based on-orbit analysis; Step 4.1 The cloud-based payload calculation platform receives the solar zenith angle of the observed target area in its current state from the satellite platform. Sun azimuth On-orbit correction coefficient of cloud-based instrument; Step 4.2 The cloud-based load calculation platform receives the pixel response values ​​of the remote sensing image data of the observation target area, and obtains the pixel radiance value L of the observation target area after preprocessing the remote sensing image data. r ; Step 4.3 The cloud load calculation platform inverts the pixel coordinates corresponding to the geographic coordinates of the observed target area according to the lookup table to obtain the cloud optical thickness; Step 4.3.1 The cloud-based payload calculation platform calculates the observation zenith angle under the current state based on the four elements of the observation target area, including time information, satellite position information, satellite velocity, latitude and longitude information of the observation target area, and satellite attitude information received from the satellite platform. Observation azimuth And read the albedo information γ of the observed target area from the cloud-based payload computing platform. r ; Step 4.3.2 The cloud-based load calculation platform calculates the relative azimuth angle under the current state according to equation (8). Step 4.3.3 The cloud-based load calculation platform calculates the zenith angle based on the current state of the observed zenith angle. Solar zenith angle relative azimuth Albedo information γ of the observed target area r and pixel radiance L r The inversion result τ of the cloud optical thickness in the observed target area is obtained by using the lookup table method and interpolation method and by inversion through equation (9). result : In equation (9), f3 represents the relational function derived from f1, and f4 represents the relational function derived from f2; Step 4.4 Based on the inversion result τ of cloud optical thickness result and the three judgment thresholds τ set low τ mid τ high Calculate the cloud cover of the pixel area corresponding to the geographic coordinates of the observed target area: When τ result ≥τ high At that time, determining the cloud cover in the observed target area is considered to confirm the presence of clouds; When τ high >τ result ≥τ mid At that time, the cloud cover in the observed target area was judged as suspected to be cloudy; When τ mid >τ result ≥τ low At that time, the cloud cover in the observed target area was determined to be suspected to be clear sky; When τ low >τ result At that time, the cloud cover of the observed target area is determined to be clear sky; Step 5: The cloud-based payload calculation platform transmits the cloud coverage information of the pixel area corresponding to the geographic coordinates of the observed target area to the satellite platform.

2. The rapid cloud judgment method for real-time on-orbit cloud detection of satellites according to claim 1, characterized in that: The size of the lookup table is n. CT ×n A ×n SZ ×n VZ ×n VA ×n bit , where n bit Indicates the number of bits used to store data.

3. The rapid cloud judgment method for real-time on-orbit cloud detection of satellites according to claim 1, characterized in that: The preprocessing of remote sensing image data in step 4.2 includes background subtraction, geometric correction, and radiometric correction.

4. An electronic device, comprising a memory and a processor, characterized in that, The memory is used to store a program that supports the processor in executing any of the fast cloud determination methods of claims 1-3, and the processor is configured to execute the program stored in the memory.

5. A computer-readable storage medium storing a computer program thereon, characterized in that, The computer program is executed by the processor to perform the steps of any of the fast cloud judgment methods described in claims 1-3.

Images