On-orbit registration method and system for satellite polarization multi-angle imager

Abstract

The application discloses a kind of satellite polarization multi-angle imager's on-orbit registration method and system, wherein the method comprises: obtaining multi-angle polarization observation image, and pre-processing, generating image to be registered;Obtain on-orbit registration coefficient;According to the preset registration algorithm and the on-orbit registration coefficient, the image to be registered is registered on-orbit, and generates image after registration;Based on the image to be registered and the image after registration, registration result is generated.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of remote sensing technology, and in particular to an on-orbit registration method and system for a satellite polarization multi-angle imager. Background Technology

[0002] The Polarization Multi-Angle Imager (PMAI) onboard the FY-3G satellite provides polarization multi-angle observation data in the shortwave infrared band, which is of great significance for the study of aerosols, clouds, and precipitation. However, due to the satellite's inverted flight, registration errors may occur between polarization channel images, leading to a decrease in the quality of the observation data, especially in applications requiring high-precision registration, such as identifying cloud rainbows. Summary of the Invention

[0003] This invention provides an on-orbit registration method and system for a satellite polarization multi-angle imager, aiming to solve the problem of polarization channel image registration error during satellite inverted flight and improve the quality of observation data.

[0004] To achieve the above objectives, in a first aspect, the present invention provides an on-orbit registration method for a satellite polarization multi-angle imager, comprising:

[0005] Step S100: Acquire multi-angle polarization observation images and perform preprocessing to generate images to be registered;

[0006] Step S200: Obtain the on-orbit registration coefficients;

[0007] Step S300: Based on the preset registration algorithm and the on-orbit registration coefficients, perform on-orbit registration on the image to be registered to generate a registered image;

[0008] Step S400: Generate a registration result based on the image to be registered and the registered image.

[0009] In one embodiment of the present invention, step S100 includes:

[0010] Step S101: Acquire the multi-angle polarization observation image during the satellite's inverted flight process;

[0011] Step S102: Preprocess the multi-angle polarization observation image to generate an image to be registered.

[0012] In one embodiment of the present invention, the on-orbit registration coefficient is 0.3-0.4.

[0013] In one embodiment of the present invention, step S300 includes:

[0014] Step S301: Extract feature points from the image to be registered, perform matching, and generate matching results;

[0015] Step S302: Based on the matching result, the image to be registered is registered in orbit according to the preset registration algorithm and the in-orbit registration coefficient to generate the registered image.

[0016] Secondly, this invention provides an on-orbit registration system for a satellite polarization multi-angle imager, comprising: a first generation module, an acquisition module, a second generation module, and a third generation module. The first generation module acquires multi-angle polarization observation images and performs preprocessing to generate an image to be registered. The acquisition module acquires on-orbit registration coefficients. The second generation module performs on-orbit registration of the image to be registered according to a preset registration algorithm and the on-orbit registration coefficients to generate a registered image. The third generation module generates a registration result based on the image to be registered and the registered image.

[0017] In one embodiment of the present invention, the first generation module includes a first acquisition unit and a first generation unit. The first acquisition unit is used to acquire the multi-angle polarization observation image during the satellite's inverted flight process. The first generation unit is used to preprocess the multi-angle polarization observation image to generate an image to be registered.

[0018] In one embodiment of the present invention, the on-orbit registration coefficient is 0.3-0.4.

[0019] In one embodiment of the present invention, the second generation module includes a second generation unit and a third generation unit. The second generation unit is used to extract feature points in the image to be registered and perform matching to generate a matching result. The third generation unit is used to perform on-orbit registration of the image to be registered based on the matching result, according to the preset registration algorithm and the on-orbit registration coefficients, to generate a registered image.

[0020] Thirdly, the present invention provides an electronic device, comprising:

[0021] At least one processor; and

[0022] A memory that is communicatively connected to the at least one processor;

[0023] The memory stores instructions that can be executed by the at least one processor, which are then executed by the at least one processor to enable the at least one processor to perform the on-orbit registration method of the satellite polarization multi-angle imager as described above.

[0024] Fourthly, the present invention provides a computer-readable storage medium including a computer program and instructions, which, when the computer program or the instructions are executed on a computer, cause the computer to perform the on-orbit registration method of the satellite polarization multi-angle imager as described above.

[0025] Compared with existing technologies, the on-orbit registration method and system for satellite polarization multi-angle imagers according to the present invention can effectively solve the problem of polarization channel image registration error during satellite inverted flight, and improve the quality of observation data. Especially in applications requiring high-precision registration, such as cloud rainbow identification, the present invention can significantly improve the reliability and accuracy of observation data, providing more reliable data support for the study of aerosols, clouds, and precipitation. Attached Figure Description

[0026] Figure 1 This is a flowchart illustrating an on-orbit registration method for a satellite polarization multi-angle imager according to Embodiment 1 of the present invention.

[0027] Figure 2 This is a schematic diagram of the on-orbit registration system of a satellite polarization multi-angle imager according to Embodiment 2 of the present invention;

[0028] Figure 3 This is a schematic diagram of the structure of an electronic device according to Embodiment 3 of the present invention. Detailed Implementation

[0029] The embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and not intended to limit the scope of the invention. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the embodiments of the present invention, and not all structures.

[0030] To facilitate understanding, the main implementation concepts of the various embodiments of the present invention will be briefly described first.

[0031] The Fengyun-3 precipitation satellite (FY-3G), carrying China's first polarization multi-angle imager (PMAI) with short-wave infrared channels, was launched in April 2023, providing crucial support for the aerosol-cloud-precipitation observation chain. Operating in a non-sun-synchronous inclined orbit, the instrument provides a spatial resolution of 3 km (nadir) and an image width of 700 km. The PMAI's observation channels include polarization channels at 1030 nm, 1370 nm, and 1640 nm, along with corresponding unpolarized channels, providing observational information from 14 angles. A wide-field-of-view optical head is positioned in front of a filter wheel composed of three filters and nine polarizers in the short-wave infrared band. The incident light received by the optical system is transmitted to the focal plane of an InGaAs detector. As the satellite platform moves, the same target can be observed in different fields of view and imaged on the detector at different focal plane positions. Thirteen images are captured during the rotation of a filter wheel. The first image is a dark image, corresponding to the detector's dark current reference, and the other 12 images correspond to the 12 spectral and polarization channels. Three spectral bands are equipped with polarizers: 1030 nm, 1370 nm, and 1640 nm. For each of these wavelengths, three similar spectral filters are associated with three polarizers, with the first and third polarizers oriented at +60° and -60° to the central polarizer, respectively. Using a radiative response model, the polarization characteristics of the incident light can be determined using observations in the three polarization directions.

[0032] The FY-3G platform obtains a three-polarization image of a single band from continuous detector responses. To compensate for flight lag, a wedge prism is used to register the three images together. The FY-3G platform operates in an inclined orbit with a nominal altitude of 407 km. To maintain a stable temperature and electromagnetic environment and ensure the satellite's on-orbit performance, the satellite employs an autonomous 180° yaw maneuver control scheme near the orbital plane. The polarization channel registration scheme of the polarization imager is designed under the condition of forward-facing satellite flight. Backward-facing satellite flight brings significant challenges to instrument channel registration, such as registration errors between polarization channel images, leading to a deterioration in the quality of observation data, especially in applications requiring high-precision registration, such as cloud rainbow identification.

[0033] By identifying the technical defects described in the prior art, the inventors propose an on-orbit registration method and system for a satellite polarization multi-angle imager, which can effectively solve the problem of polarization channel image registration error during satellite inverted flight and improve the quality of observation data.

[0034] Example 1

[0035] Figure 1 This is a flowchart illustrating an on-orbit registration method for a satellite polarization multi-angle imager according to Embodiment 1 of the present invention, as shown below. Figure 1As shown, Embodiment 1 provides an on-orbit registration method for a satellite polarization multi-angle imager, including:

[0036] Step S100: Acquire multi-angle polarization observation images and perform preprocessing to generate images to be registered;

[0037] Specifically, multi-angle polarization observation image data acquired during the satellite's inverted flight is obtained from a polarization multi-angle imager. This image data typically includes image data for each polarization channel (e.g., 1030nm, 1370nm, 1640nm) and corresponding metadata (such as timestamps, satellite position, solar angle, etc.). Necessary preprocessing is then performed on this image data to generate the image to be registered. Preprocessing may include noise reduction and radiometric calibration to ensure the accuracy of subsequent registration.

[0038] Step S200: Obtain the on-orbit registration coefficients;

[0039] Specifically, the satellite's trajectory and velocity changes during its inverted flight are acquired (e.g., through analysis of its orbital parameters such as altitude, period, and yaw angle). The design parameters of the polarization multi-angle imager are also obtained, including pixel size, field of view, and spectral band characteristics, to determine the relative displacement between images from different polarization channels. Based on the orbital parameters and the imager design parameters, an on-orbit registration coefficient R is generated. This coefficient is used in subsequent registration algorithms to represent the proportion of energy redistribution between adjacent pixels.

[0040] Step S300: Based on the preset registration algorithm and the on-orbit registration coefficients, perform on-orbit registration on the image to be registered to generate a registered image;

[0041] Specifically, for the three polarization channel images of each spectral band (images corresponding to the central polarizer and the +60° and -60° polarizers), an on-orbit registration algorithm is applied, and the radiance values ​​of adjacent pixels are redistributed according to the registration coefficient R. For the central polarizer image, the pixel values ​​on both sides are affected by the corresponding pixel values ​​of the adjacent polarization channel images. After energy redistribution, a registered image (the registered polarization channel image) is generated.

[0042] Step S400: Generate a registration result based on the image to be registered and the registered image;

[0043] Specifically, to verify the effectiveness of the registration algorithm during the inverted flight process, a typical observation scenario was used for pre-calibration (image to be registered) and post-calibration (image after registration) tests. This involved evaluating the polarization channel registration accuracy based on a specific observation scenario. Using the Stokes vectors obtained from the three polarization channels, and according to the reflectance threshold, a water cloud scenario was selected to analyze the distribution of calculated polarization reflectance with the scattering angle. The focus was on analyzing the improvement in clarity and continuity of features such as the maximum polarization reflectance at a scattering angle of 140 degrees (i.e., cloud rainbow). Specific evaluation metrics (such as structural similarity index and peak signal-to-noise ratio) were used to quantitatively analyze the registration effect, quantifying the performance of the registration algorithm. Furthermore, the registered data was applied to actual aerosol and cloud studies to determine whether it improved the accuracy of data analysis and inversion.

[0044] In this embodiment, step S100 includes:

[0045] Step S101: Acquire the multi-angle polarization observation image during the satellite's inverted flight process;

[0046] Specifically, as the satellite flies in orbit, PMAI collects multi-angle polarization observation data in real time through its internal sensors and detectors. The data is usually transmitted in digital form, including image data for each polarization channel (e.g., 1030nm, 1370nm, 1640nm, etc. in the shortwave infrared band) and associated metadata (such as timestamps, satellite position, solar angle, etc.).

[0047] In one specific embodiment, the PMAI aboard the Fengyun-3 precipitation satellite (FY-3G) continuously collects polarization image data in different spectral bands during its inverted flight. For example, within a specific time window, it collects image data of the central polarizer, +60° polarizer, and -60° polarizer in the 1030nm band, as well as related metadata such as satellite orbit information, geographical location, and time information.

[0048] Typically, a time-division multi-angle polarization observation channel consists of three similar spectral filters and three polarizers. The first and third polarizers are oriented at +60° and -60° to the central polarizer, respectively. Through the continuous response of the detector, a three-polarization image of one band is obtained. To compensate for image lag during satellite flight, a wedge prism is used to register the three images together. To solve the polarization channel registration problem during satellite inverted flight, this invention discloses an on-orbit registration algorithm. Based on the instrument design, the energy received by adjacent pixels of the three polarization channels in the same band is redistributed at a fixed ratio, achieving effective polarization channel registration during satellite inverted flight.

[0049] Based on the above analysis, receiving raw data directly from PMAI ensures data integrity and authenticity, providing a reliable foundation for subsequent data processing and analysis. Real-time data reception means the ability to quickly obtain the latest observational information, which is crucial for applications requiring real-time monitoring, such as meteorology and environmental studies.

[0050] Step S102: Preprocess the multi-angle polarization observation image to generate an image to be registered;

[0051] Specifically, preprocessing may include, for example, denoising and radiometric calibration. The denoising process eliminates noise in the original data (multi-angle polarization observation images), thereby improving image quality. Specifically, filtering algorithms such as median filtering and Gaussian filtering can be applied to smooth the image and reduce noise. Frequency domain analysis methods can be used to identify and remove high-frequency noise, resulting in a clearer denoised image, reducing the impact of noise on the registration algorithm, and improving registration accuracy. The radiometric calibration establishes a quantitative relationship between the original observation data and known radiance, achieving data quantification and standardization. Specifically, absolute calibration and relative calibration can be used. Absolute calibration uses absolute values ​​such as radiance and irradiance to quantitatively describe the target, while relative calibration determines the relative radiance relationship between images from different polarization channels. This can be achieved by comparing observation data of the same target from different channels.

[0052] Based on the above analysis, it is evident that the data after radiometric calibration possesses a unified unit of radiometric measurement, facilitating comparison and analysis between different data sets. Simultaneously, the calibration process eliminates inconsistencies in instrument response and other error factors, improving data accuracy and reliability, and providing high-quality data input for subsequent registration algorithms.

[0053] In this embodiment, the optimal value range of the on-orbit registration coefficient is 0.3-0.4.

[0054] Specifically, the on-orbit registration coefficient can be automatically obtained, and a value within the range of 0.3-0.4 is sufficient. Alternatively, it can be generated using the following methods.

[0055] Obtain the satellite orbit parameters and the polarization multi-angle imager design parameters;

[0056] Specifically, the orbital parameters of the FY-3G satellite were obtained, such as an orbital altitude of 780 km, an orbital period of approximately 100 minutes, and changes in yaw angle during inverted flight. The design parameters of PMAI were also obtained, such as a pixel size of 25 μm, a field of view of ±50°, and an operating band of shortwave infrared.

[0057] Based on the satellite orbit parameters and the polarization multi-angle imager design parameters, an image displacement model is generated;

[0058] Specifically, based on the satellite's orbital parameters (such as orbital altitude, orbital period, yaw angle, etc.), the satellite's trajectory and velocity changes during inverted flight are calculated. The impact of the satellite's trajectory on PMAI imaging is analyzed, particularly how changes in the satellite's position at different time points affect the relative displacement between images. The PMAI's design parameters (such as pixel size, field of view, spectral band characteristics, etc.) are analyzed to understand the imager's imaging characteristics and image quality. Combined with the imager's design parameters, the potential differences in geometric position between images from different polarization channels are evaluated. Based on the satellite orbital parameters and the polarization multi-angle imager's design parameters, a relative displacement model between images from different polarization channels is established. This model can consider various factors, such as satellite velocity, Earth's rotation, and atmospheric disturbances, to more accurately describe the displacement between images. A small portion of actually acquired PMAI image data is used for testing to verify the accuracy of the displacement model. By comparing corresponding points or features in images from different polarization channels, the relative displacement between images is quantitatively evaluated, generating an image displacement model.

[0059] In one specific embodiment, a displacement model is established based on the acquired satellite orbit parameters and the design parameters of the polarization multi-angle imager to describe the relative displacement between images from different polarization channels. Considering the inverted flight characteristics of the FY-3G satellite and the imaging parameters of PMAI, it can be predicted that during the inverted flight, the ±60° polarization angle image will have a certain displacement relative to the 0° polarization angle image. To verify the accuracy of the displacement model, several sets of actually acquired image data can be selected for testing. In these images, some corresponding points or features (such as cloud rainbows, terrain boundaries, etc.) can be identified, and their positional differences in different polarization channel images can be measured. By comparing these positional differences with the prediction results of the displacement model, the accuracy of the displacement model can be evaluated.

[0060] Based on the image displacement model, the on-orbit registration coefficients are generated;

[0061] Specifically, in the initial calculation of the registration coefficient, it is usually necessary to combine the previously established displacement model with the specific characteristics and requirements of the imager. In the previous steps, a displacement model was established based on the satellite's orbital parameters and the imager's design parameters. This model describes the relative displacement between images from different polarization channels. This model is, for example, a complex function that depends on multiple variables such as time, satellite position, and imager parameters. It is necessary to clarify the specific relationship between the registration coefficient R and the relative displacement between images. The registration coefficient R is typically used to guide how image registration algorithms handle displacement between images; it represents the proportion of energy redistribution between adjacent pixels, or it is related to parameters of the geometric transformation model (such as translation, rotation angle, etc.). The relevant parameters from the displacement model are substituted into the calculation formula for the registration coefficient R. This calculation formula can be designed, for example, based on the imager's working principle, image characteristics, and the requirements of the registration algorithm. For example, if the registration coefficient R represents the proportion of energy redistribution between adjacent pixels, then it is related to factors such as pixel displacement and pixel size. For example, there is a simple displacement model that describes the relative displacement (Δx, Δy) between images in the x and y directions caused by satellite motion. This displacement model is derived from the satellite's orbital parameters and the imager's design parameters. In this simplified example, the registration factor R does not directly represent the proportion of energy redistribution between adjacent pixels, but rather is related to the parameters of the geometric transformation model (such as translation). In this case, the registration factor R directly corresponds to the displacement (Δx, Δy), or is a factor used to adjust or scale the displacement. Assuming the displacement model gives the relative displacements Δx and Δy in the x and y directions, the registration factor R is used to adjust these displacements to suit a specific registration algorithm or imager characteristics. A scaling factor α (alpha) can be defined to adjust the displacement; this scaling factor can be considered a form of the registration factor R.

[0062] The calculation formula is as follows:

[0063] delta y =α*Δy

[0064] delta y It is the displacement adjusted by the registration coefficient R (represented here as the scaling factor α), which will be used in the subsequent image registration process.

[0065] The specific value of the scaling factor α needs to be determined based on the actual situation. It can be an empirical value, or the optimal value can be searched using optimization algorithms (such as gradient descent, genetic algorithms, etc.). In the initial calculation, an initial value can be set, such as α = 1 (indicating no adjustment of displacement), and then adjusted according to the registration effect.

[0066] Based on the above analysis, it is evident that registration coefficients can significantly reduce the relative displacement and deformation between images from different polarization channels, thereby improving registration accuracy. This is particularly important for applications requiring high-precision registration, such as research on aerosols, clouds, and precipitation. The calculation of registration coefficients is based on satellite orbital parameters and imager design parameters, factors that are known or predictable during satellite operation. Therefore, pre-calculating registration coefficients can greatly save computation time in subsequent registration processes, improving data processing efficiency. Accurate registration coefficients ensure that the registered image data maintains a high degree of geometric consistency, which helps eliminate data interpretation errors caused by image misalignment or deformation, thus enhancing data reliability and credibility. High-quality registration results are the foundation for subsequent data analysis and applications. Accurate registration coefficients ensure that the registered image data meets the needs of subsequent applications, such as feature extraction, target recognition, and change detection, providing strong support for scientific research and technological applications. Therefore, this step not only improves registration accuracy and data processing efficiency but also enhances data reliability and supports subsequent analysis, making it an indispensable and crucial step in the remote sensing data processing workflow.

[0067] In this embodiment, step S300 includes:

[0068] Step S301: Extract feature points from the image to be registered, perform matching, and generate matching results;

[0069] Specifically, the registration coefficient R is loaded, and image pairs to be registered are selected based on the imager design (e.g., images corresponding to the central polarizer and ±60° polarizers). For each spectral band, ensure that image pairs acquired at the same or similar times are selected to reduce registration difficulties caused by time differences. Feature points or feature regions for registration are extracted from the image pairs. Feature extraction methods can be selected based on the image characteristics and registration requirements, such as SIFT, SURF, and ORB algorithms. For polarized images, polarization information can also be used to assist feature extraction. The extracted feature points or feature regions are matched to find corresponding points or regions between image pairs. Algorithms such as nearest neighbor matching and RANSAC can be used to eliminate mismatched points and improve matching accuracy.

[0070] Step S302: Based on the matching result, the image to be registered is registered in orbit according to the preset registration algorithm and the in-orbit registration coefficient to generate the registered image.

[0071] Specifically, based on the matched feature points or feature regions, a geometric transformation model between image pairs is estimated. For polarization multi-angle imagers, a two-dimensional translation transformation needs to be considered because images from different polarization channels are mainly displaced in the horizontal or vertical directions. The parameters of the transformation model are solved using the least squares method or other optimization algorithms. Based on the estimated geometric transformation model, the reference image is resampled and interpolated to geometrically align it with the image to be registered. The resampling and interpolation methods can be selected according to the characteristics of the image and the registration requirements, such as bilinear interpolation, bicubic interpolation, etc. After image resampling, the radiometric values ​​of adjacent pixels are redistributed according to the registration coefficients (R) to generate the registered image.

[0072] In one specific embodiment, based on the instrument design and satellite orbital altitude, the satellite displacement is equal to one-third of a pixel during the time interval between two polarization images. Since the mission objective requires good registration between three polarization images of the same wavelength, the three polarization images are directly registered onto the detector matrix using a prism equipped with filters. During orbital operation, there is a one-pixel gap between the two sets of consecutive superimposed three images; that is, images of different polarization bands are positioned one row apart in the flight direction.

[0073] The 180° yaw maneuver allows for autonomous switching between forward and backward flight of the FY-3G satellite. The satellite's 180° yaw turn is autonomously controlled, completing one orbit every 28 days, 13 times per year. This application presents an on-orbit registration algorithm to address the polarization channel registration problem of the polarization imager during the satellite's inverted flight. The registration algorithm redistributes the energy received by adjacent pixels at a fixed ratio. Each polarization channel of the polarization imager (PMAI) forms a raw observation image. PMAI is an area array imager, and each observation image has an m-row, n-column layout. After radiometric calibration, the raw DN values ​​are converted to radiance values. Due to the satellite's reverse flight, the observations of the three polarization channels corresponding to the same ground target are not imaged on the same pixel. Therefore, the radiance of adjacent pixels needs to be redistributed to achieve the purpose of registration of the three polarization channels. The calculated radiance of each polarization channel in the whole-image plane is shown below (i.e., redistribution).

[0074]

[0075] Where L ′ 1(m,n),L ′ 2(m,n),L ′3(m,n) represents the calibrated radiance value for each polarization channel in a given band. m is the row number of the image plane, and n is the column number. L1(m,n), L2(m,n), and L3(m,n) are the initial radiance values ​​for the polarization channels, calculated from the relationship between the DN value and the radiative response. R represents the registration coefficient, with an optimal range of 0.3–0.4. (The determination of the R value stems from the instrument design. Because the three polarization channels are observed in a time-division manner, their images on the detector differ by 1 / 3 of a pixel. To ensure consistent imaging across the three polarization channels, a wedge design is used to offset the optical paths of the first and third polarization channels, aligning them with the central polarization channel. Therefore, the R value is an empirical value; the optimal value requires calculating the degree of linear polarization using the Stokes vectors calculated from the three polarization channels and evaluating it using the cloud rainbow characteristics of water clouds.)

[0076] Based on the above analysis, it can be seen that step S300 can effectively apply the on-orbit registration algorithm to the data processing of the polarization multi-angle imager, achieving accurate registration between images of different polarization channels; it significantly improves registration accuracy, optimizes data processing efficiency, enhances data reliability, and supports subsequent data analysis and applications, which is of great significance for remote sensing data processing and scientific research.

[0077] In one specific embodiment, step S400, to verify the effectiveness of the registration algorithm during the inverted flight process, can use a typical observation scenario for pre- and post-calibration tests. Multi-angle polarization observation data utilizes the main cloud rainbow feature generated by cloud droplets to identify water clouds from ice clouds. The main cloud rainbow is the peak of polarization reflectance observed at a scattering angle of approximately 140° with the incident direction.

[0078] These specific wavebands are equipped with three polarizers to fully determine the radiation field characteristics (reflectivity and linear polarization state) of the Stokes vector components I, Q, and U. Total reflectivity I and polarization reflectivity I0 p The definition is as follows:

[0079]

[0080] Where L is radiance, E s Here, θ represents solar irradiance, and θ represents the solar zenith. The angle between the incident and scattered light is called the scattering angle, which is crucial in cloud radiation. The phase function and polarization phase function within the effective scattering angle range provide important information for the inversion of different cloud parameters. The formula for calculating the scattering angle Θ is as follows:

[0081]

[0082] Where θ s and θ v This indicates the solar zenith angle and the observed zenith angle, while and This indicates the solar azimuth and the observation azimuth. In areas of ocean flare, polarization reflection is high due to specular reflection from the ocean surface, so it is necessary to mark the areas of ocean flare to distinguish them from the polarization information of clouds.

[0083] In one specific embodiment, during the satellite's forward flight, PMAI acquired cloud rainbows from the 1030 nm and 1640 nm polarization channels. The distribution of polarization reflectance with scattering angle exhibits a distinct characteristic: clear cloud rainbows are visible near a scattering angle of 140°. The presence of cloud rainbows indicates water clouds containing spherical particles.

[0084] However, during backward flight, the rainbow was obscured by noise due to channel registration errors between subpixels. The region with a 140° scattering angle should have shown a rainbow. However, because the three polarization channels were not effectively registered, the observed image was too noisy to recognize the rainbow. Furthermore, it can be seen that the registration error of the polarization channels has a significant impact on polarization reflectivity, but a small impact on the total reflectivity.

[0085] After calibration using the on-orbit registration algorithm during backward flight, the cloud rainbow appeared at a scattering angle of 140°. Some noise points may exist around the cloud rainbow. That is, while calibration using the registration algorithm significantly improved the registration effect, it did not fully achieve the performance of forward flight. This is because the registration algorithm redistributes the energy received by adjacent pixels at a fixed ratio. However, in reality, the targets observed by adjacent detectors are usually not completely identical. Therefore, when the observed target changes, the calibrated polarization reflectance still contains some noise.

[0086] Example 2

[0087] Figure 2 This is a schematic diagram of the on-orbit registration system of a satellite polarization multi-angle imager according to Embodiment 2 of the present invention, as shown below. Figure 2 As shown in Embodiment 2, an on-orbit registration system for a satellite polarization multi-angle imager is provided, including: a first generation module 201, an acquisition module 202, a second generation module 203, and a third generation module 204. The first generation module 201 is used to acquire multi-angle polarization observation images and perform preprocessing to generate images to be registered. The acquisition module 202 is used to acquire on-orbit registration coefficients. The second generation module 203 is used to perform on-orbit registration of the images to be registered according to a preset registration algorithm and the on-orbit registration coefficients to generate a registered image. The third generation module 204 is used to generate a registration result based on the images to be registered and the registered image.

[0088] In this embodiment, the first generation module 201 includes a first acquisition unit and a first generation unit. The first acquisition unit is used to acquire the multi-angle polarization observation image during the satellite's inverted flight process. The first generation unit is used to preprocess the multi-angle polarization observation image to generate an image to be registered.

[0089] In this embodiment, the on-orbit registration coefficient is 0.3-0.4.

[0090] In this embodiment, the second generation module 203 includes a second generation unit and a third generation unit. The second generation unit is used to extract feature points from the image to be registered and perform matching to generate a matching result. The third generation unit is used to perform on-orbit registration of the image to be registered based on the matching result, according to the preset registration algorithm and the on-orbit registration coefficients, to generate a registered image.

[0091] The various variations and specific examples of the on-orbit registration method for the satellite polarization multi-angle imager provided in Embodiment 1 are also applicable to the on-orbit registration system for the satellite polarization multi-angle imager provided in this embodiment. Through the foregoing detailed description of an on-orbit registration method for a satellite polarization multi-angle imager, those skilled in the art can clearly understand the implementation method of the on-orbit registration system for a satellite polarization multi-angle imager in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0092] Example 3

[0093] Figure 3 This is a schematic diagram of the structure of an electronic device according to Embodiment 3 of the present invention, as shown below. Figure 3 As shown, Embodiment 3 also provides an electronic device 300, which may include a processor 301 and a memory 302.

[0094] Memory 302 is used to store programs. Memory 302 may include volatile memory, such as random-access memory (RAM), such as static random-access memory (SRAM), double data rate synchronous dynamic random-access memory (DDR SDRAM), etc.; memory may also include non-volatile memory, such as flash memory. Memory 302 is used to store computer programs (such as application programs, functional modules, etc. that implement the above methods), computer instructions, etc. The computer programs, computer instructions, etc., can be partitioned and stored in one or more memories 302. Furthermore, the computer programs, computer instructions, data, etc., can be accessed by processor 301.

[0095] The aforementioned computer programs and instructions can be stored in one or more partitions of memory 302. Furthermore, the aforementioned computer programs and instructions can be invoked by processor 301.

[0096] The processor 301 is configured to execute the computer program stored in the memory 302 to implement the various steps in the methods described in the above embodiments.

[0097] For details, please refer to the relevant descriptions in the preceding method embodiments.

[0098] The processor 301 and the memory 302 can be independent structures or integrated structures. When the processor 301 and the memory 302 are independent structures, the memory 302 and the processor 301 can be coupled together via bus 303.

[0099] The electronic device in this embodiment can execute the technical solution in the above method. Its specific implementation process and technical principle are the same, and will not be repeated here.

[0100] Example 4

[0101] Embodiment 4 also provides a computer-readable storage medium including a computer program and instructions, which, when executed on a computer, cause the computer to perform the on-orbit registration method of the satellite polarization multi-angle imager according to any embodiment of the present invention.

[0102] Computer-readable storage media include various media that can store program code, such as USB flash drives, external hard drives, ROM, RAM, magnetic disks, or optical disks.

[0103] This embodiment also provides a computer program product, which includes: a computer program stored in a readable storage medium, at least one processor of an electronic device can read the computer program from the readable storage medium, and the at least one processor executes the computer program to cause the electronic device to perform the solution provided in any of the above embodiments.

[0104] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this invention disclosure can be executed in parallel, sequentially, or in different orders.

[0105] This document does not impose any restrictions as long as the desired results of the technical solution disclosed in this invention can be achieved.

[0106] In summary, the on-orbit registration method and system for a satellite polarization multi-angle imager of the present invention can effectively solve the problem of polarization channel image registration error during satellite inverted flight, and improve the quality of observation data. Especially in applications requiring high-precision registration, such as cloud rainbow identification, the present invention can significantly improve the reliability and accuracy of observation data, providing more reliable data support for the study of aerosols, clouds, and precipitation.

[0107] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

Claims

1. An on-orbit registration method for a satellite polarization multi-angle imager, characterized in that, include: Step S100: Acquire multi-angle polarization observation images and perform preprocessing to generate images to be registered; Step S200: Obtain the on-orbit registration coefficients; Step S300: Based on the preset registration algorithm and the on-orbit registration coefficients, perform on-orbit registration on the image to be registered to generate a registered image; Step S400: Based on the image to be registered and the registered image, generate a registration result; Step S100 includes: Step S101: Acquire the multi-angle polarization observation image during the satellite's inverted flight process; Step S102: Preprocess the multi-angle polarization observation image to generate an image to be registered; The on-orbit registration coefficient is 0.3-0.

4. Specifically, the preset registration algorithm is as follows: For the three polarization channels of each spectral band, the radiance values ​​of adjacent pixels are redistributed using the following formula: ； in, , , These are the calibrated radiation values ​​for the three polarization channels within a single wavelength band. These are the row numbers of the image plane. It's the column number. , , These are the initial radiation values ​​for the three polarization channels. This represents the on-orbit registration coefficient.

2. The on-orbit registration method for a satellite polarization multi-angle imager as described in claim 1, characterized in that, Step S300 includes: Step S301: Extract feature points from the image to be registered, perform matching, and generate matching results; Step S302: Based on the matching result, perform on-orbit registration on the matched image to be registered according to the preset registration algorithm and the on-orbit registration coefficient to generate the registered image.

3. An on-orbit registration system for a satellite polarization multi-angle imager, characterized in that, include: The first generation module is used to acquire multi-angle polarization observation images, perform preprocessing, and generate images to be registered; The acquisition module is used to obtain the on-orbit registration coefficients; The second generation module is used to perform on-orbit registration of the image to be registered according to the preset registration algorithm and the on-orbit registration coefficients, and generate the registered image. The third generation module is used to generate a registration result based on the image to be registered and the registered image; The first generation module includes: The first acquisition unit is used to acquire the multi-angle polarization observation images during the satellite's inverted flight process; The first generation unit is used to preprocess the multi-angle polarization observation image to generate an image to be registered; The on-orbit registration coefficient is 0.3-0.

4. Specifically, the preset registration algorithm is as follows: For the three polarization channels of the image in each spectral band, the radiance values ​​of adjacent pixels are redistributed using the following formula: ； in, , , These are the calibrated radiation values ​​for the three polarization channels within a single wavelength band. These are the row numbers of the image plane. It's the column number. , , These are the initial radiation values ​​for the three polarization channels. This represents the on-orbit registration coefficient.

4. The on-orbit registration system for a satellite polarization multi-angle imager as described in claim 3, characterized in that, The second generation module includes: The second generation unit is used to extract feature points from the image to be registered, perform matching, and generate matching results. The third generation unit is used to perform on-orbit registration on the matched image to be registered based on the matching result, according to the preset registration algorithm and the on-orbit registration coefficient, and generate the registered image.

5. An electronic device, characterized in that, include: At least one processor; as well as A memory that is communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the on-orbit registration method of the satellite polarization multi-angle imager according to any one of claims 1-2.

6. A computer-readable storage medium, characterized in that, It includes computer programs and instructions that, when the computer program or the instructions are run on a computer, cause the computer to perform the on-orbit registration method of the satellite polarization multi-angle imager as described in any one of claims 1-2.

Images