A stable radiation calibration field extraction method and device based on multi-source optical satellite remote sensing, a terminal equipment and a storage medium

Abstract

This invention discloses a method, apparatus, terminal equipment, and storage medium for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing, belonging to the field of satellite remote sensing radiometric calibration technology. The method comprises: acquiring multi-source satellite image data covering a target area within a preset time period; fusing and processing outliers for the first surface reflectance corresponding to the same pixel in the satellite image data at the same time, and calculating the coefficient of variation and standardized statistics; determining time-stable pixels based on the coefficient of variation and standardized statistics; screening the time-stable pixels for spatial uniformity and surface type consistency to obtain stable field areas, and calculating the reflectance error of each stable field area; and using stable field areas whose reflectance errors do not exceed a preset error threshold as the final stable radiometric calibration field. By implementing this invention, the problem of distorted determination of calibration field stability due to the use of a single index for calibration field stability screening in existing technologies can be solved.

Description

Technical Field

[0001] This invention relates to the field of satellite remote sensing radiometric calibration technology, and in particular to a method, apparatus, terminal equipment, and storage medium for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing. Background Technology

[0002] Radiometric calibration is a core component of the quantitative application of satellite remote sensing data. Its purpose is to establish a precise correspondence between sensor output signals and surface physical quantities (such as reflectivity and radiance), eliminate systematic biases between data from different satellites and at different times, and ensure the accuracy and comparability of remote sensing products. A stable radiometric calibration field serves as the reference benchmark for radiometric calibration, and its quantity, spatial coverage, and stability directly determine the calibration accuracy.

[0003] Existing technologies typically use only a single dispersion index (such as standard deviation) to screen for calibration field stability, ignoring the trend of reflectivity drift. This can lead to satellites launched at different times using different field regions as benchmarks, resulting in inconsistent calibration results among multiple satellites. Therefore, existing technologies suffer from the problem of distorted calibration field stability determination. Summary of the Invention

[0004] This invention provides a method, apparatus, terminal equipment, and storage medium for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing. It can solve the problem of distortion in the determination of calibration field stability caused by using a single index for calibration field stability screening in the prior art.

[0005] One embodiment of the present invention provides a method for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing, comprising: Acquire multi-source satellite imagery data covering the target area within a preset time period; The first surface reflectance corresponding to the same pixel in the satellite image data at the same time within the preset time period is fused and outlier processing is performed to obtain the effective reflectance set of each pixel; Based on the effective reflectance set, the coefficient of variation, which reflects the dispersion of reflectance of each pixel, and the standardized statistic, which reflects the trend of drift, are calculated. Pixels whose coefficient of variation does not exceed a preset coefficient of variation threshold and whose absolute value of the standardized statistic does not exceed a preset standardized statistic threshold are determined to be time-stable pixels. Spatial homogeneity and surface type consistency were screened for all time-stable pixels to obtain several stable field areas; For each stable field area, the second surface reflectance of each pixel in each band within the current stable field area is obtained, and the reflectance error corresponding to the current stable field area is calculated based on the second surface reflectance. The stable field region where the reflectivity error does not exceed a preset error threshold is taken as the final stable radiation calibration field.

[0006] Furthermore, before fusing and outlier processing the first surface reflectance corresponding to the same pixel in satellite image data at the same time within the preset time period to obtain the effective reflectance set for each pixel, the process further includes: For each pixel, obtain the solar zenith angle, observation zenith angle, and relative azimuth angle corresponding to all first surface reflectances of the current pixel; Geometric registration and resampling are performed on all first surface reflectance values ​​to obtain preprocessed first surface reflectance values; The preprocessed first surface reflectance is corrected based on the solar zenith angle, the observed zenith angle, and the relative azimuth angle.

[0007] Furthermore, the first surface reflectance corresponding to the same pixel in the satellite image data at the same time within the preset time period is fused and outlier processing is performed to obtain the effective reflectance set for each pixel, including: Based on the first surface reflectance, the variance of surface reflectance of each type of satellite image data at each time point is calculated; Based on the surface reflectance variance and the first surface reflectance, the fused reflectance of each pixel at different times is calculated, and then the fused reflectance set of each pixel at the preset time period is generated. For each pixel, the median reflectance, mean reflectance, and standard deviation reflectance of the fused reflectance set corresponding to the current pixel are calculated based on the fused reflectance. The anomaly identification range threshold is calculated based on the mean reflectance and standard deviation reflectance. The fused reflectance exceeding the anomaly identification range threshold is taken as an anomaly value, and the median reflectance is used to replace the anomaly value to obtain the effective reflectance set of each pixel.

[0008] Furthermore, based on the effective reflectance set, a coefficient of variation reflecting the dispersion of reflectance for each pixel is calculated, along with standardized statistics reflecting trend drift, including: Obtain the number of data points in the effective reflectance set corresponding to each pixel; Based on the effective reflectance set of each pixel, the mean effective reflectance and the standard deviation of effective reflectance corresponding to the effective reflectance set are calculated; The coefficient of variation for each pixel is calculated based on the mean effective reflectance and the standard deviation of effective reflectance. Based on the effective reflectance set, the order statistics of each pixel are calculated; Based on the order statistics and the number of data points, the standardized statistics for each pixel are calculated.

[0009] Furthermore, spatial homogeneity and surface type consistency were used to screen all time-stable pixels, resulting in several stable field areas, including: Based on all time-stabilized pixels, calculate the spatial standard deviation of reflectance of each time-stabilized pixel within a preset sliding window; Based on the connectivity of all time-stable pixels, the time-stable pixels whose reflectance spatial standard deviation does not exceed the preset standard deviation are aggregated to obtain several candidate stable field areas. Obtain the area, number of pixels of dominant land surface type, and total number of pixels for each candidate stable field area; Based on the number of pixels of the dominant land surface type and the total number of pixels, the proportion of the dominant land surface type in each candidate stable field area is calculated; Candidate stable areas with a dominant land surface type proportion not less than a preset proportion threshold and an area not less than a preset area threshold are designated as stable areas.

[0010] Furthermore, the reflectance error corresponding to the current stable field area is calculated based on the second surface reflectance, including: Obtain the field-measured reflectance for each band; Based on the second surface reflectance, the signal-to-noise ratio of each pixel in the current stable field area in each band is calculated; Bands with a signal-to-noise ratio less than a preset signal-to-noise ratio threshold are removed to obtain high-quality bands after screening. The weighted average reflectance of each pixel in each high-quality band is calculated based on the second surface reflectance and signal-to-noise ratio of each high-quality band. The reflectance error is calculated based on the weighted average reflectance of each high-quality band and the actual reflectance measured in the field.

[0011] Furthermore, it also includes: Acquire field area information, reflectivity data, and satellite data for each stable radiometric calibration field; Based on the field information, reflectivity data, and satellite data, a stable radiometric calibration field database is constructed.

[0012] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments; This invention provides a stable radiometric calibration field extraction device based on multi-source optical satellite remote sensing, comprising: The system includes a satellite imagery data acquisition module, a data processing module, a coefficient of variation and standardized statistics calculation module, a time-stable pixel determination module, a stable field area screening module, a reflectivity error calculation module, and a stable radiometric calibration field determination module. The satellite image data acquisition module is used to acquire multi-source satellite image data covering the target area within a preset time period; The data processing module is used to fuse and process outliers of the first surface reflectance corresponding to the same pixel in the satellite image data at the same time within the preset time period, so as to obtain the effective reflectance set of each pixel. The coefficient of variation and standardized statistics calculation module is used to calculate, based on the effective reflectance set, the coefficient of variation reflecting the dispersion of reflectance of each pixel and the standardized statistics reflecting the trend of drift. The time-stable pixel determination module is used to determine pixels whose coefficient of variation does not exceed a preset coefficient of variation threshold and whose absolute value of the standardized statistic does not exceed a preset standardized statistic threshold as time-stable pixels. The stable field area screening module is used to screen all time-stable pixels for spatial uniformity and surface type consistency to obtain several stable field areas. The reflectivity error calculation module is used to obtain the second surface reflectivity of each pixel in each band in the current stable field area for each stable field area, and calculate the reflectivity error corresponding to the current stable field area based on the second surface reflectivity. The stable radiation calibration field determination module is used to take the stable field region where the reflectivity error does not exceed a preset error threshold as the final stable radiation calibration field.

[0013] Based on the above method embodiments, the present invention provides a corresponding terminal device embodiment; The present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the stable radiometric calibration field extraction method based on multi-source optical satellite remote sensing described in any embodiment of the present invention.

[0014] Based on the above method embodiments, the present invention provides a corresponding storage medium embodiment; The present invention provides a storage medium including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the stable radiometric calibration field extraction method based on multi-source optical satellite remote sensing described in any embodiment of the present invention.

[0015] The embodiments of the present invention have the following beneficial effects: This invention provides a method, apparatus, terminal device, and storage medium for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing. The method includes: acquiring multi-source satellite image data covering a target area within a preset time period; fusing and outlier processing the first surface reflectance corresponding to the same pixel in the satellite image data at the same time within the preset time period to obtain an effective reflectance set for each pixel; calculating a coefficient of variation reflecting the reflectance dispersion of each pixel and a standardized statistic reflecting trend drift based on the effective reflectance set; identifying pixels whose coefficient of variation does not exceed a preset threshold and whose absolute value of the standardized statistic does not exceed a preset threshold as temporally stable pixels; performing spatial homogeneity and surface type consistency screening on all temporally stable pixels to obtain several stable field areas; for each stable field area, acquiring the second surface reflectance of each pixel in each band within the current stable field area, and calculating the reflectance error corresponding to the current stable field area based on the second surface reflectance; finally, using the stable field area whose reflectance error does not exceed a preset error threshold as the final stable radiometric calibration field. Therefore, in this invention, pixels are screened by judging the reflectivity dispersion, trend drift, spatial uniformity, and surface type consistency of the pixels. As a result, the final stable radiometric calibration field is more accurate than the result obtained by using a single dispersion index for screening. Attached Figure Description

[0016] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating a method for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing, provided by an embodiment of the present invention.

[0018] Figure 2 This is a schematic diagram of a stable radiometric calibration field extraction device based on multi-source optical satellite remote sensing, provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0021] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0022] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0023] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0024] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0025] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0026] See Figure 1 To address the problem of distorted stability assessment in existing technologies that use a single index for calibration field stability screening, an embodiment of this invention provides a method for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing, comprising: Step S101: Acquire multi-source satellite imagery data covering the target area within a preset time period; Specifically, the aforementioned target area refers to a candidate geographical area that may have a stable radiometric calibration field, pre-delineated according to calibration requirements. The aforementioned multi-source satellite imagery data needs to provide foundational data for subsequent stability assessments, and therefore must meet the following requirements: First, this imagery data must be continuous for more than 12 months, with at least one cloud-free image per month. For cloud cover determination, the existing Fmask cloud detection algorithm can be used, considering images with less than 10% cloud cover as cloud-free images. Second, the cloud-free coverage area of ​​a single image must be no less than 80% to avoid statistical errors due to insufficient effective data. Finally, multi-sensor, multi-resolution optical satellite data should be selected as the aforementioned multi-source satellite imagery data, such as Sentinel-2 satellite Level-2A products, Landsat 8 / 9 satellite OLI / TIRS sensor data, and domestic Gaofen-1 (GF-1) and Gaofen-6 (GF-6) wide-swath satellite data. PlanetLabs small satellite constellation data can also be used to improve temporal resolution.

[0027] Step S102: In the satellite image data of the same time within the preset time period, the first surface reflectance corresponding to the same pixel is fused and outlier is processed to obtain the effective reflectance set of each pixel; Specifically, since the aforementioned steps acquire satellite imagery data over a period of time, each pixel in this satellite imagery data needs to be processed hour by hour to obtain the set of effective reflectance of each pixel over this period of time.

[0028] In a preferred embodiment, before fusing and outlier processing the first surface reflectance corresponding to the same pixel in the satellite image data at the same time within the preset time period to obtain the effective reflectance set for each pixel, the method further includes: For each pixel, obtain the solar zenith angle, observation zenith angle, and relative azimuth angle corresponding to all first surface reflectances of the current pixel; Geometric registration and resampling are performed on all first surface reflectance values ​​to obtain preprocessed first surface reflectance values; Specifically, the first surface reflectance can be geometrically registered using a quadratic polynomial registration algorithm based on ground control points (GCPs), with the registration error controlled within one pixel. Subsequently, bilinear interpolation is used to resample the registered first surface reflectance to a spatial resolution of 10m, with the unified projection coordinate system being WGS84 / UTM. This process is existing technology and will not be elaborated further here.

[0029] The preprocessed first surface reflectance is corrected based on the solar zenith angle, the observed zenith angle, and the relative azimuth angle.

[0030] Specifically, in this invention, the Ross-Thick / Maignan (RTM) model is used to eliminate the influence of observation angle differences, thereby correcting the preprocessed first surface reflectance. In the formula, The solar zenith angle is represented as The observed zenith angle is The relative azimuth is The first surface reflectance after preprocessing, This represents the intrinsic surface reflectance retrieved by the nuclear-driven model, i.e., the corrected first surface reflectance. and These represent the kernel weight coefficients, obtained through Sentinel-2 data inversion, with an inversion error not exceeding 3%. The solar zenith angle is represented as The observed zenith angle is The relative azimuth is Geometric optics kernel function, The solar zenith angle is represented as The observed zenith angle is The relative azimuth is The volume scattering kernel function.

[0031] In this preferred embodiment, the first surface reflectance was corrected using the Ross-Thick / Maignan (RTM) model before fusion and outlier processing.

[0032] In another preferred embodiment, the first surface reflectance corresponding to the same pixel in the satellite image data at the same time within the preset time period is fused and outlier processing is performed to obtain the effective reflectance set for each pixel, including: Based on the first surface reflectance, the variance of surface reflectance of each type of satellite image data at each time point is calculated; Specifically, the aforementioned surface reflectance variance can be calculated based on the corresponding sensor accuracy parameters, which is existing technology and will not be elaborated further here.

[0033] Based on the surface reflectance variance and the first surface reflectance, the fused reflectance of each pixel at different times is calculated, and then the fused reflectance set of each pixel at the preset time period is generated. Specifically, each pixel has a corresponding fused reflectance at each moment, thus forming a set of fused reflectance over the entire preset time period.

[0034] Specifically, this invention employs a multi-source data fusion algorithm based on Bayesian estimation to integrate the advantages of different satellite data and obtain the fused reflectance: In the formula, The value represents the fused reflectance, and n represents the number of types of satellite image data involved in the fusion. This represents the first surface reflectance corresponding to the i-th type of satellite image data. Let represent the variance of the surface reflectance of the i-th type of satellite imagery data.

[0035] For each pixel, the median reflectance, mean reflectance, and standard deviation reflectance of the fused reflectance set corresponding to the current pixel are calculated based on the fused reflectance. Specifically, each pixel has a corresponding set of fused reflectances. Based on the fused reflectances in this set, the median reflectance, mean reflectance, and standard deviation reflectance can be determined.

[0036] The anomaly identification range threshold is calculated based on the mean reflectance and standard deviation reflectance. The fused reflectance exceeding the anomaly identification range threshold is taken as an anomaly value, and the median reflectance is used to replace the anomaly value to obtain the effective reflectance set of each pixel.

[0037] Specifically, in this invention, outlier detection is performed using the 3σ principle, and the median replacement method is used for processing to obtain the aforementioned effective reflectance set. That is, the difference between the mean reflectance and three times the standard deviation of the reflectance is used as the minimum threshold in the outlier detection range, and the sum of the mean reflectance and three times the standard layer reflectance is used as the maximum threshold in the outlier detection range, forming the outlier detection range threshold. Subsequently, outlier detection and replacement are performed based on this threshold. This replacement process can be expressed by the following formula: In the formula, This represents the effective reflectance after outlier identification and processing. Represents the fused reflectivity at time t. Represents the set of fused reflectances The median reflectance.

[0038] Preferably, when identifying outliers, extreme outliers may exist, and these extreme outliers are directly discarded. It should be noted that in this invention, fused reflectance greater than 3 times the standard deviation of reflectance is considered an extreme outlier.

[0039] Preferably, the present invention employs BRDF correction and Bayesian multi-source data fusion algorithm to eliminate systematic bias caused by differences in observation angle and sensors; it adopts the temporal stability evaluation index of "CV+Mann-Kendall trend test" and combines spatial screening based on spatial homogeneity and surface type consistency to form a four-dimensional screening standard, which significantly improves the accuracy of stability judgment compared with traditional methods.

[0040] In this preferred embodiment, the effective reflectance set of each pixel is obtained by fusing the first surface reflectance and processing outliers.

[0041] Step S103: Based on the effective reflectance set, calculate the coefficient of variation to reflect the dispersion of reflectance of each pixel, and the standardized statistic to reflect the trend of drift. Specifically, this invention employs both the coefficient of variation and standardized statistics as indicators to quantitatively evaluate the effective reflectance set, while also reflecting the dispersion and trend drift of reflectance.

[0042] In a preferred embodiment, based on the effective reflectance set, a coefficient of variation reflecting the dispersion of reflectance for each pixel and a standardized statistic reflecting trend drift are calculated, including: Obtain the number of data points in the effective reflectance set corresponding to each pixel; Based on the effective reflectance set of each pixel, the mean effective reflectance and the standard deviation of effective reflectance corresponding to the effective reflectance set are calculated; The coefficient of variation for each pixel is calculated based on the mean effective reflectance and the standard deviation of effective reflectance. Specifically, the coefficient of variation mentioned above reflects the dispersion of the data; the smaller the value, the lower the dispersion. The coefficient of variation is calculated using the following formula: In the formula, Represents the coefficient of variation. Indicates the standard deviation of effective reflectance. This represents the average effective reflectance.

[0043] Based on the effective reflectance set, the order statistics of each pixel are calculated; Specifically, this invention uses the Mann-Kendall trend test to determine whether there is a significant trend shift. This test is a non-parametric test method that does not require the data to satisfy a normal distribution.

[0044] Specifically, the order statistic is calculated using the following formula: In the formula, Represents order statistics. This represents the number of data points in the effective reflectance set. This represents the j-th data point in the set of effective reflectance. This represents the i'-th data point in the effective reflectance set. Represents a sign function, if If the value is greater than 0, then the function value is 1; if it is equal to 0... =0, then the function value is 0, if If the value is less than 0, then the function value is -1.

[0045] Based on the order statistics and the number of data points, the standardized statistics for each pixel are calculated.

[0046] Specifically, the standardized statistic is calculated using the following formula: In the formula, Represents standardized statistics. It represents an intermediate quantity.

[0047] In this preferred embodiment, the coefficient of variation and standardized statistics for each pixel are calculated based on the effective reflectance set.

[0048] Step S104: Pixels whose coefficient of variation does not exceed a preset coefficient of variation threshold and whose absolute value of the standardized statistic does not exceed a preset standardized statistic threshold are determined as time-stable pixels. Specifically, in this invention, a preset threshold for the coefficient of variation is set to 3%, and a preset threshold for the standardized statistic is set to 1.96 (i.e., the Mann-Kendall trend test significance level p≥0.05). Pixels with a coefficient of variation ≤3% and an absolute value of the standardized statistic ≤1.96 are considered time-stable pixels.

[0049] Step S105: Perform spatial homogeneity and surface type consistency screening on all time-stable pixels to obtain several stable field areas; Specifically, for the time-stable pixels selected in the aforementioned steps, further spatial homogeneity verification and surface type consistency screening are required, along with pixel aggregation and field area screening, to finally obtain a stable radiation calibration field that meets the calibration requirements.

[0050] In a preferred embodiment, all time-stable pixels are screened for spatial homogeneity and surface type consistency to obtain several stable field areas, including: Based on all time-stabilized pixels, calculate the spatial standard deviation of reflectance of each time-stabilized pixel within a preset sliding window; Specifically, the spatial standard deviation of reflectance for each time-stable pixel within a 3×3 sliding window is calculated. This spatial standard deviation reflects the degree of spatial heterogeneity of the pixel; a smaller spatial standard deviation indicates better spatial homogeneity. The spatial standard deviation of reflectance is calculated using the following formula: In the formula, Represents the spatial standard deviation of reflectivity. This represents the effective reflectance corresponding to the i''th time-stable pixel. This represents the mean effective reflectance of time-stable pixels within a 3×3 window.

[0051] Based on the connectivity of all time-stable pixels, the time-stable pixels whose reflectance spatial standard deviation does not exceed the preset standard deviation are aggregated to obtain several candidate stable field areas. Specifically, firstly, pre-defined time-stable pixels whose reflectance spatial standard deviation exceeds the preset standard deviation are removed from all time-stable pixels. Then, the connectivity between the remaining time-stable pixels is analyzed, and spatial clustering is performed according to the connectivity of four-neighbor or eight-neighbor domains to obtain several candidate stable field areas.

[0052] Obtain the area, number of pixels of dominant land surface type, and total number of pixels for each candidate stable field area; Specifically, the aforementioned land surface types can be bare land, uniform vegetation, etc.

[0053] Based on the number of pixels of the dominant land surface type and the total number of pixels, the proportion of the dominant land surface type in each candidate stable field area is calculated; Specifically, this invention introduces high-precision land use classification data (such as ESACCIL and Cover) to calculate the proportion of dominant land surface types in each candidate stable area, avoiding additional calibration errors introduced by mixed land surfaces. The proportion of dominant land surface types is calculated using the following formula: In the formula, Indicates the proportion of dominant land surface types. Indicates the number of pixels representing the dominant land surface type. This represents the total number of pixels.

[0054] Candidate stable areas with a dominant land surface type proportion not less than a preset proportion threshold and an area not less than a preset area threshold are designated as stable areas.

[0055] Specifically, the aforementioned preset percentage threshold is set to 90%, and the preset area threshold is 1 square kilometer. That is, candidate stable areas with a dominant surface type percentage of not less than 90% and an area of ​​not less than 1 square kilometer are selected as stable areas.

[0056] Preferably, small, fragmented areas can be eliminated by filtering by area.

[0057] In this preferred embodiment, several stable field areas were obtained by screening all time-stable pixels for spatial homogeneity and surface type consistency.

[0058] Step S106: For each stable field area, obtain the second surface reflectance of each pixel in each band within the current stable field area, and calculate the reflectance error corresponding to the current stable field area based on the second surface reflectance. Specifically, for each stable field area, the corresponding hyperspectral image data is extracted, and then the second surface reflectance of each pixel in each band is obtained.

[0059] Preferably, hyperspectral satellite data such as Hyperion, domestic Gaofen-5 (GF-5), Gaofen-11 (GF-11), and ZY-3 (ZY-3) can be used, covering the spectral range of 400-2500nm. At the same time, based on the geographical location of the stable field area, multi-temporal hyperspectral images of the corresponding area should be extracted, requiring at least 5 images with a time span covering 3 seasons to ensure the seasonal representativeness of the second surface reflectance.

[0060] Preferably, when acquiring these hyperspectral image data, what is actually extracted is the raw product distributed by the satellite ground station, namely the dimensionless DN value output by the sensor in response to ground radiance. Subsequently, the DN value is converted into top atmospheric radiance using the sensor's radiative transfer equation. Finally, the MODTRAN model is used to eliminate the effects of atmospheric absorption and scattering, resulting in the aforementioned second surface reflectance. This process is existing technology and will not be described in detail here.

[0061] In a preferred embodiment, the reflectance error corresponding to the current stable field area is calculated based on the second surface reflectance, including: Obtain the field-measured reflectance for each band; Based on the second surface reflectance, the signal-to-noise ratio of each pixel in the current stable field area in each band is calculated; Preferably, before calculating the signal-to-noise ratio, these second surface reflectances can be geometrically registered to maintain spatial consistency with the first surface reflectances, with a registration error of no more than 1 pixel.

[0062] Specifically, the signal-to-noise ratio is calculated using the following formula: In the formula, Indicates the signal-to-noise ratio. This represents the average second surface reflectance of a homogeneous surface area on the band. This represents the noise standard deviation, which is obtained through statistics of dark pixel regions.

[0063] Bands with a signal-to-noise ratio less than a preset signal-to-noise ratio threshold are removed to obtain high-quality bands after screening. Specifically, a preset signal-to-noise ratio (SNR) threshold of 50 is set, which means low-quality bands with an SNR less than 50 are eliminated to avoid them affecting the reflectivity accuracy, thus obtaining the remaining high-quality bands. For illustration, the selected high-quality bands can be 400nm-2500nm, with a spectral resolution not exceeding 10nm.

[0064] The weighted average reflectance of each pixel in each high-quality band is calculated based on the second surface reflectance and signal-to-noise ratio of each high-quality band. Specifically, a signal-to-noise ratio-based weighted averaging algorithm is used to calculate the weighted average of reflectance, replacing the traditional arithmetic average, to improve reflectance accuracy. The weighted average of reflectance is calculated using the following formula: In the formula, Indicates that a pixel is in The weighted average reflectance over the band, where m represents the number of hyperspectral images included in the calculation. This indicates that a pixel in the k-th hyperspectral image is in band [band missing]. The second surface reflectance, This indicates that a pixel in the k-th hyperspectral image is in band [band missing]. Weight on, This indicates that a pixel in the k-th hyperspectral image is in band [band missing]. The signal-to-noise ratio.

[0065] The reflectance error is calculated based on the weighted average reflectance of each high-quality band and the actual reflectance measured in the field.

[0066] Specifically, the weighted average reflectance is cross-validated using field measurement data from a predetermined calibration field. The accuracy of this measurement data must not exceed 1%, and the reflectance error must not exceed 5%. Therefore, the validation formula is: In the formula, Indicates that a pixel is in Reflectivity error in the band express Field measured reflectance of the band.

[0067] Preferably, the present invention uses SNR to filter out low-quality bands and combines a weighted average algorithm to improve the accuracy of hyperspectral reflectance data and expand the application range of spectral bands.

[0068] In this preferred embodiment, the reflectance error corresponding to each stable field area is calculated using the second surface reflectance.

[0069] Step S107: The stable field region where the reflectivity error does not exceed the preset error threshold is taken as the final stable radiation calibration field.

[0070] Specifically, a preset error threshold of 5% is set, meaning that a stable field region with a reflectivity error of no more than 5% is used as the final stable radiation calibration field.

[0071] Preferably, the stable radiometric calibration field extraction method of the present invention integrates multiple types of optical satellite data, and the entire process can be automated. It can quickly screen out stable calibration fields in a wide area, making up for the shortcomings of the scarcity of existing calibration fields and limited spatial coverage, and adapting to the needs of multi-satellite constellation collaborative observation.

[0072] Preferably, the present invention effectively resists interference factors such as cloud pollution, atmospheric anomalies, and changes in observation angle through multiple mechanisms such as outlier processing, BRDF correction, and land surface type constraints.

[0073] In a preferred embodiment, it further includes: Acquire field area information, reflectivity data, and satellite data for each stable radiometric calibration field; Specifically, the aforementioned field information represents the core geographic and stability attributes of the stable radiometric calibration field, including: the field ID, latitude and longitude range, land surface type, land surface type percentage, stabilization duration, CV value, p-value, spatial standard deviation (i.e., the aforementioned spatial standard deviation of reflectance), and validation error (i.e., the aforementioned cross-validation reflectance error). The aforementioned reflectance data represents the hyperspectral reflectance attribute of the stable radiometric calibration field, including: band wavelength, weighted average of hyperspectral reflectance, standard deviation of hyperspectral reflectance, and mean SNR. The aforementioned satellite data represents the satellites and data parameters involved in the processing, including: satellite type, sensor parameters, data time range, and preprocessing parameters. These preprocessing parameters include: radiometric calibration parameters, atmospheric correction parameters, and topographic parameters.

[0074] It should be noted that the above-mentioned standard deviation of hyperspectral reflectance refers to the standard deviation of hyperspectral reflectance of the stable radiative calibration field over multiple time periods; the weighted average of hyperspectral reflectance refers to the weighted average of hyperspectral reflectance of the stable radiative calibration field over multiple time periods.

[0075] Based on the field information, reflectivity data, and satellite data, a stable radiometric calibration field database is constructed.

[0076] Specifically, to achieve standardized storage, dynamic updates, spatial indexing, and visual querying of calibration field data, and to provide a standardized and reusable reference benchmark for multi-satellite collaborative observation, a stable radiometric calibration field database based on PostgreSQL and PostGIS can be built based on the aforementioned field area information, reflectivity data, and satellite data. The database's infrastructure uses a PostgreSQL relational database and PostGIS spatial extensions to support spatial data storage, indexing, and querying, adapting to the geospatial attributes of the calibration field. Then, based on the aforementioned field area information, reflectivity data, and satellite data, and combined with a dynamic update log table (i.e., database update and audit records), four core data tables are designed to achieve standardized and categorized data storage. The tables are linked through field area IDs to facilitate multi-table joint queries. The core data tables are shown in the table below: Preferably, for the aforementioned stable radiometric calibration field database, the GiST (Generalized Search Tree) spatial index is used to optimize regional query efficiency. Simultaneously, newly added multi-source optical satellite data and hyperspectral data are automatically added monthly, recalculating stability indices and average reflectance values ​​to achieve automatic updates of calibration field attributes. A web-based visual query interface can also be developed, supporting multi-condition queries, including by region (rectangular or circular area), band, stability level, and surface type. Query results are output in a combination of tables and spatial distribution maps, intuitively displaying the spatial distribution and attribute information of the calibration field.

[0077] In this preferred embodiment, a stable radiometric calibration field database is constructed based on the field area information, reflectivity data, and satellite data of the stable radiometric calibration field.

[0078] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments.

[0079] like Figure 2 As shown, one embodiment of the present invention provides a stable radiometric calibration field extraction device based on multi-source optical satellite remote sensing, comprising: The system includes a satellite imagery data acquisition module, a data processing module, a coefficient of variation and standardized statistics calculation module, a time-stable pixel determination module, a stable field area screening module, a reflectivity error calculation module, and a stable radiometric calibration field determination module. The satellite image data acquisition module is used to acquire multi-source satellite image data covering the target area within a preset time period; Specifically, the aforementioned target area refers to a candidate geographical area that may have a stable radiometric calibration field, pre-delineated according to calibration requirements. The aforementioned multi-source satellite imagery data needs to provide foundational data for subsequent stability assessments, and therefore must meet the following requirements: First, this imagery data must be continuous for more than 12 months, with at least one cloud-free image per month. For cloud cover determination, the existing Fmask cloud detection algorithm can be used, considering images with less than 10% cloud cover as cloud-free images. Second, the cloud-free coverage area of ​​a single image must be no less than 80% to avoid statistical errors due to insufficient effective data. Finally, multi-sensor, multi-resolution optical satellite data should be selected as the aforementioned multi-source satellite imagery data, such as Sentinel-2 satellite Level-2A products, Landsat 8 / 9 satellite OLI / TIRS sensor data, and domestic Gaofen-1 (GF-1) and Gaofen-6 (GF-6) wide-swath satellite data. PlanetLabs small satellite constellation data can also be used to improve temporal resolution.

[0080] The data processing module is used to fuse and process outliers of the first surface reflectance corresponding to the same pixel in the satellite image data at the same time within the preset time period, so as to obtain the effective reflectance set of each pixel. Specifically, since the aforementioned module acquires satellite imagery data over a period of time, the data processing module needs to process the data hourly for each pixel in the satellite imagery data to obtain the set of effective reflectance of each pixel over this period of time.

[0081] The coefficient of variation and standardized statistics calculation module is used to calculate, based on the effective reflectance set, the coefficient of variation reflecting the dispersion of reflectance of each pixel and the standardized statistics reflecting the trend of drift. Specifically, in the coefficient of variation and standardized statistics calculation module, the coefficient of variation and standardized statistics are used as dual indicators to quantitatively evaluate the effective reflectance set, while reflecting the dispersion and trend drift of reflectance.

[0082] The time-stable pixel determination module is used to determine pixels whose coefficient of variation does not exceed a preset coefficient of variation threshold and whose absolute value of the standardized statistic does not exceed a preset standardized statistic threshold as time-stable pixels. Specifically, in the time-stable pixel determination module, a preset threshold for the coefficient of variation is set to 3%, and a preset threshold for the standardized statistic is set to 1.96 (i.e., the Mann-Kendall trend test significance level p≥0.05). Pixels with a coefficient of variation ≤3% and an absolute value of the standardized statistic ≤1.96 are considered time-stable pixels.

[0083] The stable field area screening module is used to screen all time-stable pixels for spatial uniformity and surface type consistency to obtain several stable field areas. Specifically, for the selected time-stable pixels, the stable field area screening module will further conduct spatial homogeneity verification and surface type consistency screening, and perform pixel aggregation and field area screening to finally obtain a stable radiation calibration field that meets the calibration requirements.

[0084] The reflectivity error calculation module is used to obtain the second surface reflectivity of each pixel in each band in the current stable field area for each stable field area, and calculate the reflectivity error corresponding to the current stable field area based on the second surface reflectivity. Specifically, for each stable field area, the reflectance error calculation module extracts the corresponding hyperspectral image data, and then obtains the second surface reflectance of each pixel in each band.

[0085] Preferably, hyperspectral satellite data such as Hyperion, domestic Gaofen-5 (GF-5), Gaofen-11 (GF-11), and ZY-3 (ZY-3) can be used, covering the spectral range of 400-2500nm. At the same time, based on the geographical location of the stable field area, multi-temporal hyperspectral images of the corresponding area should be extracted, requiring at least 5 images with a time span covering 3 seasons to ensure the seasonal representativeness of the second surface reflectance.

[0086] Preferably, when the reflectance error calculation module acquires these hyperspectral image data, it essentially extracts the raw product distributed by the satellite ground station, namely the dimensionless DN value output by the sensor in response to ground radiance. Subsequently, the DN value is converted into top atmospheric radiance using the sensor's radiative transfer equation. Finally, the MODTRAN model is used to eliminate the effects of atmospheric absorption and scattering, resulting in the aforementioned second surface reflectance. This process is existing technology and will not be described in detail here.

[0087] The stable radiation calibration field determination module is used to take the stable field region where the reflectivity error does not exceed a preset error threshold as the final stable radiation calibration field.

[0088] Specifically, the stable radiation calibration field determination module is set with a preset error threshold of 5%, meaning that a stable field region with a reflectivity error of no more than 5% is used as the final stable radiation calibration field.

[0089] In a preferred embodiment, it further includes: a data correction module; The data correction module is used to, before fusing and outlier processing the first surface reflectance corresponding to the same pixel in the satellite image data at the same time within the preset time period to obtain the effective reflectance set of each pixel, obtain the solar zenith angle, observation zenith angle and relative azimuth angle corresponding to all the first surface reflectances of the current pixel for each pixel. Geometric registration and resampling are performed on all first surface reflectance values ​​to obtain preprocessed first surface reflectance values; The preprocessed first surface reflectance is corrected based on the solar zenith angle, the observed zenith angle, and the relative azimuth angle.

[0090] In another preferred embodiment, the data processing module includes: Variance calculation unit, set generation unit, set parameter calculation unit, threshold calculation unit, and outlier handling unit; The variance calculation unit is used to calculate the variance of the surface reflectance of each type of satellite image data at each time point based on the first surface reflectance. The set generation unit is used to calculate the fused reflectance of each pixel at different times based on the surface reflectance variance and the first surface reflectance, and then generate a set of fused reflectance of each pixel in the preset time period. The set parameter calculation unit is used to calculate, for each pixel, the median reflectance, mean reflectance, and standard deviation reflectance of the set of fused reflectances corresponding to the current pixel based on the fused reflectance. The threshold calculation unit is used to calculate the anomaly identification range threshold based on the mean reflectance and the standard deviation reflectance. The outlier processing unit is used to identify fused reflectance values ​​exceeding the outlier identification range threshold as outliers, and replace the outliers with the median reflectance to obtain the effective reflectance set for each pixel.

[0091] In another preferred embodiment, the coefficient of variation and standardized statistic calculation module includes: The system includes a data count unit, an effective reflectance parameter calculation unit, a coefficient of variation calculation unit, an order statistic calculation unit, and a standardized statistic calculation unit. The data counting unit is used to obtain the number of data points in the effective reflectance set corresponding to each pixel; The effective reflectance parameter calculation unit is used to calculate the mean effective reflectance and the standard deviation of effective reflectance corresponding to the effective reflectance set based on the effective reflectance set of each pixel. The coefficient of variation calculation unit is used to calculate the coefficient of variation of each pixel based on the mean effective reflectance and the standard deviation of effective reflectance. The order statistics calculation unit is used to calculate the order statistics of each pixel based on the effective reflectance set. The standardized statistics calculation unit is used to calculate the standardized statistics of each pixel based on the order statistics and the number of data.

[0092] In another preferred embodiment, the stable field screening module includes: The unit includes a spatial standard deviation calculation unit for reflectance, a candidate stable field area determination unit, a candidate stable field area data acquisition unit, a dominant land surface type proportion calculation unit, and a stable field area determination unit. The reflectance spatial standard deviation calculation unit is used to calculate the reflectance spatial standard deviation of each time-stable pixel within a preset sliding window based on all time-stable pixels. The candidate stable field area determination unit is used to aggregate the pre-prepared time-stable pixels whose reflectance spatial standard deviation does not exceed a preset standard deviation based on the connectivity of all time-stable pixels, and obtain a number of candidate stable field areas. The candidate stable field area data acquisition unit is used to acquire the area, number of dominant land surface type pixels, and total number of pixels of each candidate stable field area. The dominant surface type proportion calculation unit is used to calculate the dominant surface type proportion of each candidate stable field area based on the number of pixels of the dominant surface type and the total number of pixels. The stable field area determination unit is used to identify candidate stable field areas whose dominant surface type accounts for a proportion not less than a preset proportion threshold and whose area is not less than a preset area threshold as stable field areas.

[0093] In another preferred embodiment, the reflectivity error calculation module includes: The system includes a field-measured reflectance acquisition unit, a signal-to-noise ratio calculation unit, a high-quality band selection unit, a weighted average reflectance calculation unit, and a reflectance error determination unit. The field-measured reflectance acquisition unit is used to acquire the field-measured reflectance of each band. The signal-to-noise ratio calculation unit is used to calculate the signal-to-noise ratio of each pixel in the current stable field area in each band based on the second surface reflectivity. The high-quality band filtering unit is used to remove bands with a signal-to-noise ratio less than a preset signal-to-noise ratio threshold to obtain filtered high-quality bands. The reflectance weighted average calculation unit is used to calculate the reflectance weighted average of each pixel in each high-quality band based on the second surface reflectance and signal-to-noise ratio of each high-quality band. The reflectivity error determination unit is used to calculate the reflectivity error based on the weighted average reflectivity corresponding to each high-quality band and the field-measured reflectivity.

[0094] In another preferred embodiment, it further includes: a database construction module; The database construction module is used to acquire field area information, reflectivity data, and satellite data for each stable radiometric calibration field; Based on the field information, reflectivity data, and satellite data, a stable radiometric calibration field database is constructed.

[0095] It should be noted that the device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationship between modules indicates that they have a communication connection, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without creative effort. The above schematic diagram is merely an example of a stable radiometric calibration field extraction device based on multi-source optical satellite remote sensing and does not constitute a limitation on a stable radiometric calibration field extraction device based on multi-source optical satellite remote sensing. It may include more or fewer components than illustrated, or combine certain components, or use different components.

[0096] Based on the above method embodiments, the present invention provides corresponding terminal device embodiments.

[0097] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the stable radiometric calibration field extraction method based on multi-source optical satellite remote sensing described in any embodiment of the present invention.

[0098] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the device. The aforementioned terminal devices may be computing devices such as desktop computers, laptops, handheld computers, and cloud servers. These devices may include, but are not limited to, processors and memory. The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the device, connecting various parts of the device via various interfaces and lines. The aforementioned memory can be used to store the aforementioned computer programs and / or modules. The aforementioned processor implements various functions of the aforementioned device by running or executing the computer programs and / or modules stored in the aforementioned memory, and by calling data stored in the memory. The aforementioned memory may mainly include a program storage area and a data storage area, wherein the program storage area may store the operating system, at least one application program required for a function, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0099] Based on the above method embodiments, the present invention provides corresponding storage medium embodiments.

[0100] Another embodiment of the present invention provides a storage medium including a stored computer program, wherein, when the computer program is running, it controls the device where the storage medium is located to execute the stable radiometric calibration field extraction method based on multi-source optical satellite remote sensing described in any embodiment of the present invention.

[0101] In this embodiment, the storage medium is a computer-readable storage medium, and the computer program includes computer program code, which may be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium may include any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0102] The above are preferred embodiments of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing, characterized in that, include: Acquire multi-source satellite imagery data covering the target area within a preset time period; The first surface reflectance corresponding to the same pixel in the satellite image data at the same time within the preset time period is fused and outlier processing is performed to obtain the effective reflectance set of each pixel; Based on the effective reflectance set, the coefficient of variation, which reflects the dispersion of reflectance of each pixel, and the standardized statistic, which reflects the trend of drift, are calculated. Pixels whose coefficient of variation does not exceed a preset coefficient of variation threshold and whose absolute value of the standardized statistic does not exceed a preset standardized statistic threshold are determined to be time-stable pixels. Spatial homogeneity and surface type consistency were screened for all time-stable pixels to obtain several stable field areas; For each stable field area, the second surface reflectance of each pixel in each band within the current stable field area is obtained, and the reflectance error corresponding to the current stable field area is calculated based on the second surface reflectance. The stable field region where the reflectivity error does not exceed a preset error threshold is taken as the final stable radiation calibration field.

2. The method for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing according to claim 1, characterized in that, Before fusing and outlier processing the first surface reflectance corresponding to the same pixel in the satellite image data at the same time within the preset time period to obtain the effective reflectance set for each pixel, the method further includes: For each pixel, obtain the solar zenith angle, observation zenith angle, and relative azimuth angle corresponding to all first surface reflectances of the current pixel; Geometric registration and resampling are performed on all first surface reflectance values ​​to obtain preprocessed first surface reflectance values; The preprocessed first surface reflectance is corrected based on the solar zenith angle, the observed zenith angle, and the relative azimuth angle.

3. The method for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing according to claim 2, characterized in that, For satellite imagery data at the same time within the preset time period, the first surface reflectance corresponding to the same pixel is fused and outlier processed to obtain the effective reflectance set for each pixel, including: Based on the first surface reflectance, the variance of surface reflectance of each type of satellite image data at each time point is calculated; Based on the surface reflectance variance and the first surface reflectance, the fused reflectance of each pixel at different times is calculated, and then the fused reflectance set of each pixel at the preset time period is generated. For each pixel, the median reflectance, mean reflectance, and standard deviation reflectance of the fused reflectance set corresponding to the current pixel are calculated based on the fused reflectance. The anomaly identification range threshold is calculated based on the mean reflectance and standard deviation reflectance. The fused reflectance exceeding the anomaly identification range threshold is taken as an anomaly value, and the median reflectance is used to replace the anomaly value to obtain the effective reflectance set of each pixel.

4. The method for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing according to claim 3, characterized in that, Based on the set of effective reflectance, a coefficient of variation reflecting the dispersion of reflectance for each pixel is calculated, along with standardized statistics reflecting trend drift, including: Obtain the number of data points in the effective reflectance set corresponding to each pixel; Based on the effective reflectance set of each pixel, the mean effective reflectance and the standard deviation of effective reflectance corresponding to the effective reflectance set are calculated; The coefficient of variation for each pixel is calculated based on the mean effective reflectance and the standard deviation of effective reflectance. Based on the effective reflectance set, the order statistics of each pixel are calculated; Based on the order statistics and the number of data points, the standardized statistics for each pixel are calculated.

5. The method for extracting a stable radiometric calibration field based on multi-source optical satellite remote sensing according to claim 4, characterized in that, Spatial homogeneity and surface type consistency were screened for all time-stable pixels, resulting in several stable field areas, including: Based on all time-stabilized pixels, calculate the spatial standard deviation of reflectance of each time-stabilized pixel within a preset sliding window; Based on the connectivity of all time-stable pixels, the time-stable pixels whose reflectance spatial standard deviation does not exceed the preset standard deviation are aggregated to obtain several candidate stable field areas. Obtain the area, number of pixels of dominant land surface type, and total number of pixels for each candidate stable field area; Based on the number of pixels of the dominant land surface type and the total number of pixels, the proportion of the dominant land surface type in each candidate stable field area is calculated; Candidate stable areas with a dominant land surface type proportion not less than a preset proportion threshold and an area not less than a preset area threshold are designated as stable areas.

6. The method for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing according to claim 5, characterized in that, The reflectance error corresponding to the current stable field area is calculated based on the second surface reflectance, including: Obtain the field-measured reflectance for each band; Based on the second surface reflectance, the signal-to-noise ratio of each pixel in the current stable field area in each band is calculated; Bands with a signal-to-noise ratio less than a preset signal-to-noise ratio threshold are removed to obtain high-quality bands after screening. The weighted average reflectance of each pixel in each high-quality band is calculated based on the second surface reflectance and signal-to-noise ratio of each high-quality band. The reflectance error is calculated based on the weighted average reflectance of each high-quality band and the actual reflectance measured in the field.

7. The method for extracting stable radiometric calibration fields based on multi-source optical satellite remote sensing according to claim 6, characterized in that, Also includes: Acquire field area information, reflectivity data, and satellite data for each stable radiometric calibration field; Based on the field information, reflectivity data, and satellite data, a stable radiometric calibration field database is constructed.

8. A stable radiometric calibration field extraction device based on multi-source optical satellite remote sensing, characterized in that, include: The system includes a satellite imagery data acquisition module, a data processing module, a coefficient of variation and standardized statistics calculation module, a time-stable pixel determination module, a stable field area screening module, a reflectivity error calculation module, and a stable radiometric calibration field determination module. The satellite image data acquisition module is used to acquire multi-source satellite image data covering the target area within a preset time period; The data processing module is used to fuse and process outliers of the first surface reflectance corresponding to the same pixel in the satellite image data at the same time within the preset time period, so as to obtain the effective reflectance set of each pixel. The coefficient of variation and standardized statistics calculation module is used to calculate, based on the effective reflectance set, the coefficient of variation reflecting the dispersion of reflectance of each pixel and the standardized statistics reflecting the trend of drift. The time-stable pixel determination module is used to determine pixels whose coefficient of variation does not exceed a preset coefficient of variation threshold and whose absolute value of the standardized statistic does not exceed a preset standardized statistic threshold as time-stable pixels. The stable field area screening module is used to screen all time-stable pixels for spatial uniformity and surface type consistency to obtain several stable field areas. The reflectivity error calculation module is used to obtain the second surface reflectivity of each pixel in each band in the current stable field area for each stable field area, and calculate the reflectivity error corresponding to the current stable field area based on the second surface reflectivity. The stable radiation calibration field determination module is used to take the stable field region where the reflectivity error does not exceed a preset error threshold as the final stable radiation calibration field.

9. A terminal device, characterized in that, The method includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement a stable radiometric calibration field extraction method based on multi-source optical satellite remote sensing as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device where the storage medium is located to perform a stable radiometric calibration field extraction method based on multi-source optical satellite remote sensing as described in any one of claims 1 to 7.

Images