Surface reflectance data scaling method, device, equipment, medium and product

Abstract

The present application relates to the technical field of remote sensing data processing, and provides a surface reflectivity data scale conversion method, device, equipment, medium and product. The optimal matching target ground measurement spectrum is selected from a standard spectrum library by calculating a spectral correlation coefficient, the target ground measurement spectrum is taken as a standard reference spectrum of image reflectivity data, first and second scale conversion coefficients are determined, and ground measurement reflectivity data under star-ground synchronization and non-star-ground synchronization is respectively subjected to scale conversion based on the first and second scale conversion coefficients. By interpolating the first scale conversion coefficient, the deviation caused by time mismatch is reduced, the first and second scale conversion coefficients are used to reflect the spatial difference between the ground measurement position and the high-resolution pixel of the satellite, the scale conversion precision is improved, and the problem of large error caused by error accumulation of high-resolution data is solved.

Description

Technical Field

[0001] This invention relates to the field of remote sensing data processing technology, and in particular to a method, apparatus, equipment, medium, and product for scaling surface reflectance data. Background Technology

[0002] Ground reflectance, as a fundamental data source for quantitative remote sensing, is a crucial reference for retrieving ground attributes (such as vegetation cover, crop growth, and soil organic matter) and verifying the authenticity of remote sensing products. However, due to factors such as spatial heterogeneity of the ground, differences in sensor observation methods, varying observation perspectives and lighting conditions, and mismatches in data acquisition time, there are significant differences in measurement scale between ground reflectance measured at a single point and reflectance products at the satellite pixel scale. Furthermore, ground measurements are typically conducted in localized areas with limited spatial representativeness, while satellite remote sensing products mix different land cover types at the pixel scale, resulting in variations in the proportion of different land cover types measured at the same location. Ground measurements alone cannot characterize the complex spatial heterogeneity within a pixel, making it difficult to directly use ground measurement data for remote sensing product verification. To address this issue, it is necessary to scale the ground reflectance observed from ground observations to the satellite pixel scale.

[0003] Existing upscaling methods mainly include statistical methods and high-resolution imagery-based methods. Statistical methods perform scale transformation through weighted averaging or interpolation of multiple ground measurement points. High-resolution imagery-based methods use high-resolution imagery as an intermediary, establishing relationships between ground measurement data, high-resolution data, and low-resolution data through regression models, decomposition algorithms, or spatial optimization methods, thereby achieving scale transformation from ground measurement data to low-resolution data.

[0004] However, statistical methods are only applicable to areas with relatively uniform ground cover distribution and cannot accurately reflect the spatially heterogeneous ground features, resulting in limited scale transformation accuracy. While methods based on high-resolution imagery can improve scale transformation accuracy to some extent, they are limited by the error accumulation inherent in high-resolution data itself, leading to larger errors in large or complex areas. Summary of the Invention

[0005] This invention provides a method, apparatus, device, medium, and product for scaling surface reflectance data, which addresses the limitations of statistical methods in scaling accuracy and the shortcomings of high-resolution image-based methods, which are constrained by the availability of high-resolution data and the accumulation of errors in the data itself, resulting in large errors over large or complex areas.

[0006] This invention provides a method for scaling surface reflectance data, comprising the following steps.

[0007] Obtain image reflectance data for at least one satellite corresponding to the ground measurement location;

[0008] Based on the spectral correlation coefficient, the target ground measurement spectrum corresponding to the image reflectance data is obtained by screening from a pre-constructed standard spectral library, and the target ground measurement spectrum is used as the standard reference spectrum of the image reflectance data;

[0009] By comparing the standard reference spectrum with the ground measurement spectrum under the same date under satellite-to-ground synchronization, the first scale conversion coefficient under satellite-to-ground synchronization is obtained; the first scale conversion coefficient is interpolated to obtain the second scale conversion coefficient under non-satellite-to-ground synchronization.

[0010] Based on the first scale conversion coefficient, the ground measurement reflectance data under satellite-to-ground synchronization is scaled, and based on the second scale conversion coefficient, the ground measurement reflectance data under non-satellite-to-ground synchronization is scaled to obtain pixel-scale surface reflectance data.

[0011] According to the present invention, a method for scaling surface reflectance data is provided, wherein the standard spectral library includes a ground-equivalent standard spectral library and a full-spectrum standard spectral library. Correspondingly, the step of selecting the target ground measurement spectrum corresponding to the image reflectance data from the pre-constructed standard spectral library based on the spectral correlation coefficient includes:

[0012] Based on the spectral correlation coefficient between the ground equivalent standard spectral library and the image reflectance data, the ground equivalent spectrum with the best fitting effect and the corresponding measurement time are obtained.

[0013] Based on the measurement time of the ground equivalent spectrum with the best fitting effect, ground measurement spectra with the corresponding measurement time are selected from the full-spectrum standard spectrum library, and the ground measurement spectra are used as the target ground measurement spectra.

[0014] According to the present invention, a method for scaling surface reflectance data is provided, wherein the full-spectrum standard spectral library is obtained by performing spectral filtering and format normalization on time-series measured surface reflectance data, and the ground equivalent standard spectral library is obtained by spectral convolution of the full-spectrum standard spectral library and the satellite spectral response function.

[0015] According to the method for scale conversion of surface reflectance data provided by the present invention, the method further includes: spatially expanding the surface reflectance data at the pixel scale based on the image reflectance data of at least one satellite within a time scale to obtain raster reflectance data within the target area scale.

[0016] According to the present invention, a method for scaling surface reflectance data includes spatial scaling of the pixel-scale surface reflectance data based on image reflectance data from at least one satellite over a time scale to obtain raster reflectance data at the target area scale, comprising:

[0017] Based on the surface reflectance data at the pixel scale, the equivalent ground reflectance at the pixel scale is obtained;

[0018] Based on the image reflectance data of at least one satellite over a time scale, determine the central pixel and surrounding pixels within the target area scale centered on the ground measurement location;

[0019] Based on the satellite image reflectance data, a minimum cost function is constructed between the center pixel and any of the surrounding pixels. The minimum cost function is used to characterize the least squares sum of the differences between the pixel reflectance of the surrounding pixels and the ground reflectance of the center pixel.

[0020] Solving the minimum cost function yields the space expansion coefficient matrix;

[0021] Based on the spatial expansion coefficient matrix, the surface reflectance data at the pixel scale is spatially expanded to obtain the raster reflectance data at the target area scale.

[0022] According to the method for scaling surface reflectance data provided by the present invention, after performing spatial scale expansion based on image reflectance data of at least one satellite within a time scale and surface reflectance data at the pixel scale to obtain raster reflectance data within the target area scale, the method further includes:

[0023] Calculate the spectral correlation coefficient, and select the ground measurement spectrum of any pixel within the target area scale on any date within the time scale from the standard spectral library to obtain the spectral scale expansion result of the grid reflectance data.

[0024] The present invention also provides a surface reflectance data scaling device, comprising:

[0025] The acquisition module is used to acquire image reflectance data of at least one satellite corresponding to the ground measurement location;

[0026] The filtering module is used to filter the target ground measurement spectrum corresponding to the image reflectance data from a pre-built standard spectral library based on the spectral correlation coefficient, and use the target ground measurement spectrum as the standard reference spectrum of the image reflectance data.

[0027] The coefficient determination module is used to compare the standard reference spectrum with the ground measurement spectrum under the same date under satellite-to-ground synchronization conditions to obtain the first scale conversion coefficient under satellite-to-ground synchronization conditions; and to perform interpolation processing on the first scale conversion coefficient to obtain the second scale conversion coefficient under non-satellite-to-ground synchronization conditions.

[0028] The scale conversion module is used to perform scale conversion on ground measurement reflectance data under satellite-to-ground synchronization based on the first scale conversion coefficient, and to perform scale conversion on ground measurement reflectance data under non-satellite-to-ground synchronization based on the second scale conversion coefficient, so as to obtain surface reflectance data at the pixel scale.

[0029] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the surface reflectance data scaling method as described above.

[0030] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the surface reflectance data scaling method as described above.

[0031] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the surface reflectance data scaling method as described above.

[0032] The present invention provides a method, apparatus, device, medium, and product for scaling surface reflectance data. By calculating the spectral correlation coefficient, the optimal matching target ground measurement spectrum is selected from a standard spectral library. This target ground measurement spectrum is used as the standard reference spectrum for image reflectance data. First and second scale conversion coefficients are then determined, and scale conversion is performed on ground measurement reflectance data under both satellite-to-ground synchronization and non-satellite-to-ground synchronization conditions based on these coefficients. Interpolation processing of the first scale conversion coefficient reduces the bias introduced by time mismatch. The first and second scale conversion coefficients reflect the spatial difference between the ground measurement location and the high-resolution pixels of the satellite, improving scale conversion accuracy and solving the problem of large errors caused by error accumulation in high-resolution data. Attached Figure Description

[0033] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0034] Figure 1 This is one of the flowcharts illustrating the surface reflectance data scale conversion method provided by the present invention.

[0035] Figure 2 This is the second flowchart illustrating the surface reflectance data scale conversion method provided by the present invention.

[0036] Figure 3 This is a schematic diagram of spectral curves from a standard spectral library at a certain station.

[0037] Figure 4 This is a schematic diagram of pixel reflectance at the corresponding ground measurement location from the Sentinel2 satellite.

[0038] Figure 5 This is a schematic diagram showing the comparison between the ground equivalent reflectivity with the best fitting effect and the corresponding band of the Sentinel2 satellite.

[0039] Figure 6 This is a schematic diagram of the standard reference reflectance curve.

[0040] Figure 7 These are image maps corresponding to the B2, B3, B4, and B8 bands of the Sentinel2 satellite at the target area scale.

[0041] Figure 8 This is a schematic diagram showing the comparison of Sentinel satellite reflectance results before and after a certain day, as well as on the same day, based on ground-based spectral upscaling measurements at a certain station.

[0042] Figure 9 This is a schematic diagram of the spatial scale expansion provided by the present invention.

[0043] Figure 10 This is a schematic diagram of the spectral scale extension provided by the present invention.

[0044] Figure 11 This is a schematic diagram of the surface reflectance data scale conversion device provided by the present invention.

[0045] Figure 12 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0047] Figure 1This is one of the flowcharts illustrating the surface reflectance data scaling method provided by this invention, such as... Figure 1 As shown, the method includes steps S100-S400.

[0048] Step S100: Obtain image reflectance data of at least one satellite corresponding to the ground measurement location.

[0049] Ground measurement location refers to the location where remote sensing instruments are installed to measure ground reflectivity. Remote sensing instruments can be multispectral or hyperspectral imagers, radiometers, ground object spectrometers, etc.

[0050] The image reflectance data of at least one satellite corresponding to the ground measurement location specifically refers to the image reflectance data obtained by high-resolution satellite measurement at the ground measurement location. The image reflectance data includes reflectance value, data acquisition time, and satellite transit time.

[0051] Step S200: Based on the spectral correlation coefficient, the target ground measurement spectrum corresponding to the image reflectance data of each satellite is selected from the pre-constructed standard spectral library, and the target ground measurement spectrum is used as the standard reference spectrum of the image reflectance data.

[0052] Based on the correlation coefficient, the target ground measurement spectrum corresponding to the image reflectance data of each satellite is obtained by screening from a pre-constructed standard spectral library. Specifically, the best matching ground measurement spectrum is selected from the standard spectral library as the target ground measurement spectrum, using the correlation coefficient as the matching index.

[0053] Step S300: Compare the standard reference spectrum with the ground measurement spectrum under the same date under satellite-to-ground synchronization to obtain the first scale conversion coefficient under satellite-to-ground synchronization; perform interpolation on the first scale conversion coefficient to obtain the second scale conversion coefficient under non-satellite-to-ground synchronization.

[0054] The ground-based spectral measurements under satellite-to-ground synchronization conditions are obtained by selecting ground-based spectral curves from a standard spectral library based on the satellite transit time in the image reflectance data.

[0055] Step S400: Based on the first scale conversion coefficient, scale conversion is performed on the ground measurement reflectance data under satellite-to-ground synchronization conditions, and based on the second scale conversion coefficient, scale conversion is performed on the ground measurement reflectance data under non-satellite-to-ground synchronization conditions, to obtain pixel-scale surface reflectance data.

[0056] Pixel-scale surface reflectance data is obtained by upscaling ground-measured reflectance data to the satellite pixel scale.

[0057] Understandably, this invention uses correlation coefficients to select the optimal matching target ground measurement spectrum from a standard spectral library. This target ground measurement spectrum is then used as a standard reference spectrum for image reflectance data to determine the first and second scale conversion coefficients. Based on these coefficients, scale conversion is performed on ground measurement reflectance data under both satellite-to-ground synchronization and non-satellite-to-ground synchronization conditions. By interpolating the first scale conversion coefficient, biases introduced by time mismatch are reduced. The first and second scale conversion coefficients are used to reflect the spatial differences between the ground measurement location and the high-resolution pixels of the satellite, thus improving scale conversion accuracy and addressing the problem of large errors caused by the accumulation of errors in the high-resolution data itself.

[0058] Based on the above embodiments, as an optional embodiment, the standard spectral library includes a ground-equivalent standard spectral library and a full-spectrum standard spectral library. The full-spectrum standard spectral library covers a set of standard spectral data in the 350-1000 nm spectral range, representing the optical reflectance or radiation characteristics of various surface materials (such as soil, vegetation, water bodies, rocks, etc.) in different spectral bands. The ground-equivalent standard spectral library is a dataset that reflects the spectral characteristics of real ground objects in actual environments, measured and verified under specific standards and conditions. It is typically used to simulate the sensor observation effects on space remote sensing platforms.

[0059] Optionally, the full-spectrum standard spectral library is obtained by filtering and formatting time-series measured surface reflectance data, and the ground equivalent standard spectral library is obtained by spectral convolution of the full-spectrum standard spectral library with the spectral response function of the satellite.

[0060] The steps for determining the full-spectrum standard spectral library include data acquisition, preliminary filtering, secondary filtering, and format conversion.

[0061] Data acquisition: Acquire time series (in this embodiment, it refers to the time period at the daily granularity) ground-measured surface reflectance data. The surface reflectance data contains necessary temporal and spatial information, such as the latitude and longitude of the ground station, the ground measurement time, and the spectral data of the ground measurement.

[0062] Preliminary filtering: Read the raw surface reflectance data, filter out the effective ground spectral data, and remove measurement records with no data or obvious outliers to complete the preliminary filtering process of the surface reflectance data.

[0063] Secondary filtering: A polynomial fitting method is used to perform cloud filtering on all daily measurement records after the initial filtering process, removing measurement records affected by clouds, thus completing the secondary filtering process. Specifically, a quadratic curve is used to fit the reflectance data to filter out data points with large residuals. The specific steps are as follows:

[0064] Establish a time series of surface reflectance data, assuming the reflectance dataset is { }, the corresponding time series is { }, thus obtaining the dataset { , };

[0065] Perform quadratic curve fitting on the entire dataset to obtain the quadratic curve. coefficient , , and fitting residuals ;when At that time, according to the coefficient , , Find Remove all The data. Among them, To fit the residual threshold, N is the weighting coefficient. Based on the information provided by the automatic measuring instrument used, Set it to 0.25, and N to 1.64;

[0066] Repeat the quadratic curve fitting and data removal process until no more data is removed.

[0067] Format conversion: The measurement records after secondary filtering are saved in a unified format, including the measurement date, measurement time, and ground reflectance value, to obtain a full-spectrum standard spectral library.

[0068] The ground-equivalent standard spectral library is obtained by spectrally convolving the full-spectrum standard spectral library with the spectral response function of the satellite. Specifically, the spectral response function of the high-resolution satellite is obtained, and spectral convolution is performed on all measurement records of the full-spectrum standard spectral library to obtain the ground-equivalent standard spectral library. The formula for spectral convolution is as follows:

[0069]

[0070] in, For satellite data The equivalent ground reflectance corresponding to the band, For satellite data The spectral response function corresponding to the band, This is surface reflectance data.

[0071] Optionally, after constructing the standard spectral library, intermediate data required for scale transformation can be pre-determined before step S100 for direct use later.

[0072] The intermediate data required for scale conversion include the reflectance of the corresponding pixels at the ground measurement location, the matching sample dataset of the ground and the satellite, and the equivalent ground spectrum when not in satellite-ground synchronization.

[0073] The reflectance of the corresponding pixel at the ground measurement location is obtained by extracting multi-temporal reflectance images from the image reflectance data, which is multi-band reflectance data.

[0074] The ground-satellite matching sample dataset is selected from the full-spectrum standard spectral library based on the acquisition date and time of the high-resolution satellite image reflectance data. Spectral data from the same date and 30 minutes before and after the satellite's transit are taken, and their mean is used as the ground spectral curve under satellite-ground synchronization. Spectral convolution is then performed on this curve to obtain the equivalent ground reflectance under satellite-ground synchronization. This reflectance is then compared with the reference spectrum in the standard reference spectrum to obtain the first-scale conversion coefficient.

[0075] For non-satellite-to-ground synchronization, the ground equivalent spectrum is obtained by acquiring the average reflectance data 30 minutes before and after the local time of satellite passage from the full-spectrum standard spectral library for dates where no satellite passage is recorded. The corresponding ground equivalent spectrum is then calculated. The local time of satellite passage refers to the time when the satellite passes a specific location. Due to the Earth's rotation causing differences in the sun's position at different longitudes, the passage time of the same satellite will vary in different longitudes.

[0076] It is understood that this invention provides a construction scheme for a full-spectrum standard spectral library and a ground-based standard equivalent spectral library. Through filtering and format processing, the error of the full-spectrum standard spectral library is reduced. The ground-based standard equivalent spectral library is obtained by spectral convolution of the full-spectrum standard spectral library, which can ensure the consistency of reflectance data after scale transformation.

[0077] Based on the above embodiments, as an optional embodiment, the step of selecting the target ground measurement spectrum corresponding to the image reflectance data of each satellite from a pre-constructed standard spectral library based on the spectral correlation coefficient includes the following steps.

[0078] Based on the spectral correlation coefficient between the ground equivalent standard spectral library and the image reflectance data, the ground equivalent spectrum with the best fitting effect and the corresponding measurement time are obtained.

[0079] Based on the measurement time of the ground equivalent spectrum with the best fitting effect, ground measurement spectra with the corresponding measurement time are selected from the full-spectrum standard spectrum library, and the ground measurement spectra are used as the target ground measurement spectra.

[0080] Optionally, the correlation coefficient is the Pearson correlation coefficient, which can be represented as r. The Pearson correlation coefficient ranges from -1 to 1, with positive values ​​indicating a positive correlation and negative values ​​indicating a negative correlation. The closer the correlation is to 1, the stronger the correlation. The closer the value is to 1, the higher the degree of matching between the ground equivalent reflectivity and the reflectivity of the satellite pixels. The highest measured ground reflectance data represents the ground measurement record with the best fitting effect.

[0081] Correspondingly, the ground equivalent standard spectral data in the ground equivalent standard spectral library are subjected to spectral fitting on a daily basis. The Pearson correlation coefficient is used as the matching index to select the ground measurement record with the best fitting effect and obtain the measurement time of the ground measurement record.

[0082] Based on the measurement time, the corresponding ground measurement spectrum is selected from the full-spectrum standard spectral library and used as the standard reference spectrum for the day the high-resolution satellite reflectance data is acquired during the upscaling process from ground measurement to high-resolution satellite pixel scale.

[0083] Preferably, a high-resolution satellite pixel-scale standard reference spectrum dataset is constructed by integrating the standard reference spectra obtained on the day of all high-resolution satellite transits.

[0084] In step S300, the standard reference spectrum is compared with the ground measurement spectrum under the same date for satellite-ground synchronization to obtain the first scale conversion coefficient under satellite-ground synchronization. Specifically, the upscaling coefficient under satellite-ground synchronization is obtained by comparing the above-mentioned high-resolution satellite pixel-scale standard reference spectrum under the same date with the ground measurement spectrum under satellite-ground synchronization.

[0085] Interpolation is performed on the first scale conversion coefficients to obtain the second scale conversion coefficients under non-satellite-ground synchronization conditions. Specifically, a linear interpolation method is used to perform time-series interpolation on the first scale conversion coefficients to obtain the second scale conversion coefficients. The first scale conversion coefficients and the second scale conversion coefficient can be used to characterize the spatial differences between a single ground point and a high-resolution pixel, such as the proportion of vegetation and bare land, avoiding errors caused by single-point or multi-point statistical methods.

[0086] In step S400, the ground measurement reflectance data under satellite-to-ground synchronization is scaled based on the first scale conversion coefficient. Specifically, the first scale conversion coefficient is multiplied by the ground measurement reflectance data under satellite-to-ground synchronization to complete the upscaling process and obtain high-resolution pixel-scale surface reflectance data under satellite-to-ground synchronization.

[0087] The scale transformation of ground-measured reflectance data under non-satellite-ground synchronization conditions is based on the second scale transformation coefficient. Specifically, the second scale transformation coefficient is multiplied by the ground-measured reflectance data under non-satellite-ground synchronization conditions to obtain high-resolution pixel-scale surface reflectance data under non-satellite-ground synchronization conditions.

[0088] Combining the upscaling results under both satellite-to-ground and non-satellite-to-ground synchronization conditions, the true ground reflectance at the high-resolution pixel scale is finally obtained. This data spatially matches the pixel scale while maintaining spectral characteristics consistent with the ground-measured spectrum.

[0089] Understandably, this invention uses the Pearson correlation coefficient as a screening criterion. The index value directly reflects spectral similarity, and the process is transparent and traceable, overcoming the shortcomings of machine learning-based methods that rely on black-box models and struggle to trace the physical mechanisms of scale transformation. By introducing a similarity matching index when calculating the scale transformation coefficient, and using a spectral matching method based on a historical ground-based measured standard spectral library to obtain reference spectra, the authenticity and reliability of the data are ensured.

[0090] Based on the above embodiments, as an optional embodiment, the present invention provides a method for scaling surface reflectance data, which further includes the following steps.

[0091] Spatial scale expansion of the surface reflectance data at the pixel scale is performed based on the image reflectance data of at least one satellite within a time scale to obtain the raster reflectance data at the target area scale.

[0092] It is understood that this invention is based on high-resolution satellite reflectance products of long time series, and spatially extends the high-resolution pixel-scale surface reflectance data to a spatial range of 500m×500m, so as to obtain raster reflectance data with spectral range and spatial resolution consistent with high-resolution satellites, and spatial range of the target area scale.

[0093] Based on the above embodiments, as an optional embodiment, the step of spatially expanding the surface reflectance data at the pixel scale based on the image reflectance data of at least one satellite within a time scale to obtain the raster reflectance data at the target area scale includes the following steps.

[0094] Based on the surface reflectance data at the pixel scale, the equivalent ground reflectance at the pixel scale is obtained.

[0095] Specifically, spectral convolution is performed on the surface reflectance data at the pixel scale to obtain high-resolution equivalent ground reflectance at the pixel scale.

[0096] Based on the image reflectance data of at least one satellite over a time scale, determine the central pixel and surrounding pixels within the target area scale centered on the ground measurement location.

[0097] Optionally, the spatial range of the target area is 500m, and the central pixel centered on the pixel where the ground measurement location is located and the surrounding pixels within this spatial range are extracted from the image reflectance data.

[0098] Based on the satellite image reflectance data within the specified regional scale, a minimum cost function is constructed for the central pixel and any of the surrounding pixels, wherein the minimum cost function is used to characterize the least squares sum of the differences between the pixel reflectance of the surrounding pixels and the ground reflectance of the central pixel.

[0099] The expression for the minimum cost function J is shown below:

[0100]

[0101] in, The time scale refers to the length of time, i.e., the length of the time series of high-resolution satellite reflectivity. wavelength In position The date is High-resolution pixel-scale reflectivity, wavelength Transforming ground-measured data on date d into ground-equivalent reflectance at a high-resolution pixel scale The space expansion coefficient matrix A needs to be derived.

[0102] Solving the minimum cost function yields the spatial expansion coefficient matrix.

[0103] The minimum cost function is solved using the least squares regression method, and the spatial expansion coefficient vector is obtained as shown below:

[0104]

[0105] in, This involves converting measured ground data into equivalent ground reflectance at a high-resolution pixel scale. wavelength In position High-resolution pixel-scale reflectivity.

[0106] Based on the spatial expansion coefficient matrix, the surface reflectance data at the pixel scale is spatially expanded to obtain the raster reflectance data at the target area scale.

[0107] Spatial extension of the pixel-scale surface reflectance data based on the spatial extension coefficient matrix specifically refers to... Surface reflectance data applied at the pixel scale is transformed to the target area scale using matrix operations to generate raster reflectance data.

[0108] The spatial resolution and spectral range of the raster reflectance data are consistent with those of high-resolution satellite reflectance products, and the spatial range is the target area scale.

[0109] Understandably, this invention establishes the spatial location mapping relationship between corresponding pixels of ground stations in high-resolution satellites and surrounding pixels through least squares regression, generating a spatial scale expansion matrix to extend the reflectance at the high-resolution pixel scale to the regional scale. Furthermore, this invention comprehensively considers temporal and spatial information, using high-resolution satellite data as a bridge to obtain spatial expansion coefficients to assist in the spatial reconstruction of low-resolution spectral information. In this process, traditional remote sensing auxiliary parameters such as DEM, albedo, land cover, and NDVI are not introduced, thus avoiding dependence on multi-source samples and external data. Simultaneously, high-resolution data is only used in the derivation of expansion coefficients; once these coefficients are obtained, subsequent processing no longer relies on this data, reducing the continuous dependence on specific remote sensing samples. Users can verify the rationality of the scale transformation by checking the distribution of intermediate parameters such as the Pearson correlation coefficient value, upscaling coefficient, and spatial expansion matrix.

[0110] Based on the above embodiments, as an optional embodiment, after spatially expanding the surface reflectance data at the pixel scale based on the image reflectance data of at least one satellite at the time scale to obtain the raster reflectance data at the target area scale, the following steps are also included.

[0111] Based on the correlation coefficient, the ground measurement spectrum of any pixel within the target area scale on any date within the time scale is obtained by screening from the standard spectral library, thus obtaining the spectral scale extension result of the grid reflectance data.

[0112] Specifically, raster reflectance data for any given date is selected, and all bands and the location information of any given pixel are extracted. Based on a ground-equivalent standard spectral library, Pearson correlation coefficient is used as a similarity fitting index for spectral matching. The equivalent spectrum with the highest Pearson correlation coefficient value is extracted, and the date and time of the spectral measurement are recorded. According to the date and time of the spectral measurement, the corresponding ground-measured spectrum is selected from the full-spectrum standard spectral library as the spectral extension result for that pixel. This yields high-resolution pixel-scale raster reflectance data for any given date. The raster reflectance data for all dates are combined to obtain the spectral scale extension result of the time-series high-resolution pixel-scale raster reflectance data. The spectral range of the spectral scale extension result of the raster reflectance data is consistent with the ground measurement, and the spatial range is consistent with the target area scale.

[0113] Understandably, due to the spectral differences between high-resolution satellite and ground-based measurement data, traditional methods relying on fixed response functions or simplified band matching can lead to the loss of full-spectral spectral features of the ground. Furthermore, obtaining hyperspectral data from low-spectral data often involves mathematical or machine learning simulations. This invention, however, performs spectral matching based on historical measured spectral data, achieving spectral expansion from low to high, ensuring the authenticity and reliability of the spectrum. This invention achieves progressive expansion of the ground, pixels, space, and spectrum in stages, avoiding the error amplification problem of direct cross-scale conversion and improving the accuracy of the final data. The resulting multi-scale reflectance data is highly consistent across time, space, and spectral dimensions, and can be directly used for multi-source remote sensing data fusion and long-term trend analysis.

[0114] Figure 2 This is the second flowchart of the surface reflectance data scale conversion method provided by the present invention. The preferred embodiments of the present invention will be illustrated below with reference to the accompanying drawings.

[0115] Taking the surface reflectance data measured by a fully automatic ground reflectance meter at a certain station in 20XX and the image reflectance data corresponding to the 10m resolution bands (B2, B3, B4, and B8 bands) of the Sentinel2 satellite for all dates in 20XX as an example, the preferred embodiment of the surface reflectance data scale conversion method provided by the present invention can be divided into four stages: data preparation and preprocessing stage, ground to satellite pixel scale conversion stage, reflectance spatial scale expansion stage, and reflectance spectral scale expansion stage.

[0116] 1. The data preparation and preprocessing stage specifically refers to acquiring and preprocessing surface reflectance data, establishing a standard spectral library, and other intermediate result data. The standard spectral library includes a full-spectrum standard spectral library and a ground equivalent spectral library.

[0117] (101) Read the raw ground reflectance measurement files for any given day at a certain station, including ground-measured surface reflectance data and high-resolution satellite image reflectance data. Perform batch processing on all data, including preprocessing steps such as effective data filtering, cloud removal, and format standardization, and construct a standard spectral library for the station. Figure 3 The solid line represents the surface reflectance data curve in the standard spectrum of the spectral band.

[0118] (102) Input the spectral response function of the Sentinel2 satellite ( Figure 3 (Middle dashed line) Perform spectral convolution on all measurement records in the full-spectrum standard spectral library to obtain the ground-equivalent standard spectral library ( Figure 3 The "○" in the figure represents the reflectance in the ground-equivalent standard spectral library.

[0119] (103) Obtain 10m resolution reflectance products from Sentinel-2 satellite (B2, B3, B4, and B8 bands respectively), including its multi-temporal reflectance data, and extract the reflectance of the corresponding pixels at the ground measurement location, such as... Figure 4 As shown, Figure 4 The dashed lines represent the spectral response functions corresponding to the B2, B3, B4, and B8 bands of the Sentinel2 satellite, and the "○" represents the reflectance of that pixel in the Sentinel2 image.

[0120] (104) Based on the date and time obtained from the Sentinel2 satellite reflectance, the ground spectral curve for satellite-to-ground synchronization is obtained from the aforementioned full-spectrum standard spectral library, and further spectral convolution is performed to obtain the equivalent reflectance for satellite-to-ground synchronization, such as... Figure 3 The "--" indicates the spectral response of Sentinel2 in the B2, B3, B4, and B8 bands.

[0121] (105) For dates in the full-spectrum standard spectral library where no satellite passes over, obtain the average reflectance data for 30 minutes before and after the local time of 11:20 (local time of satellite passage) as the ground spectral curve when not synchronized with satellite, and calculate the corresponding ground equivalent reflectance.

[0122] 2. The ground-to-satellite pixel-scale conversion stage specifically refers to upscaling the ground-measured spectrum to the Sentinel2 satellite pixel scale (10m) to obtain high-resolution pixel-scale reflectance products. Specific steps include:

[0123] (201) Daily readings of the 10m resolution (B2, B3, B4, and B8 bands) pixel-scale reflectance data of the Sentinel2 satellite corresponding to the ground measurement points were performed. Spectral matching was conducted with the ground equivalent standard spectral library in Phase 1. The Pearson correlation coefficient was used as the similarity matching index. The record with the highest correlation coefficient was selected as the ground measurement spectrum with the best fitting effect, and its measurement date and time were recorded. The scatter plot corresponding to its fitting effect is shown below. Figure 5 As shown.

[0124] (202) Based on the ground measurement date and time in the screening records, match the corresponding ground measurement spectrum from the full-spectrum standard spectral library, and use it as the standard reference spectrum for the day the Sentinel2 satellite reflectance data was acquired (e.g., September 3, 20XX) during the upscaling process from ground measurement to high-resolution satellite pixel scale. Figure 6 The figure shows the corresponding standard reference reflectance curve; the solid line represents the standard reference reflectance curve, "○" represents the reflectance value of the corresponding pixel of the Sentinel2 satellite on that day, and "□" represents the equivalent reflectance obtained by spectral convolution of the standard reference reflectance.

[0125] (203) Integrate the standard reference spectra obtained on the day of all high-resolution satellite transits to construct a high-resolution satellite pixel-scale standard reference spectrum dataset.

[0126] (204) Calculate the ratio of the high-resolution satellite pixel-scale standard reference spectrum and the satellite-ground synchronous ground measurement spectrum on the same date, and use it as the upscaling factor value of the ground single station to the pixel scale on that day; multiply it with the satellite-ground synchronous ground measured reflectance data to complete the upscaling process and obtain the high-resolution pixel-scale surface reflectance data under the satellite-ground synchronous condition.

[0127] (205) Linear interpolation is used to perform time series interpolation on the upscaling coefficient values ​​obtained in the previous step, and the interpolated upscaling coefficients are multiplied by the measured ground reflectance data under non-satellite-ground synchronization to obtain high-resolution pixel-scale surface reflectance data under non-satellite-ground synchronization.

[0128] (206) Combining the upscaling results under both synchronous and asynchronous conditions, the true ground reflectance at the high-resolution pixel scale is finally obtained. This data spatially matches the 10m pixel scale of Sentinel 2 while maintaining consistency with the spectral characteristics of the ground-measured spectrum. The results are as follows: Figure 7 As shown.

[0129] 3. The reflectance spatial scale expansion stage specifically refers to the spatial expansion of the obtained high-resolution pixel-scale ground reflectance true values ​​based on the 10m resolution reflectance products from the Sentinel 2 satellite over a long time series. This involves expanding the spatial range from a single 10m resolution pixel to a 500m × 500m spatial range, resulting in raster reflectance data with the same spectral range and spatial resolution as the high-resolution satellite, but at the target area scale. The specific steps are as follows:

[0130] (301) Perform spectral convolution on the obtained high-resolution pixel-scale surface reflectance data to obtain the high-resolution pixel-scale equivalent ground reflectance.

[0131] (302) In the long-term Sentinel2 satellite reflectivity data, extract the 50×50 pixels surrounding the pixel where the ground measurement point is located. Using the least squares regression method, combined with temporal and spatial information, calculate the mapping relationship and reflectivity difference between the corresponding pixel of the ground measurement point and the surrounding pixels, and generate a spatial expansion coefficient matrix A from the satellite pixel scale to a spatial range of 500m. Figure 7 This is a comparison of the reflectance of the Sentinel satellite on April 24, 20XX, before and after the ground-based spectral upscaling (from a single ground station to a 10m pixel scale) at a certain station. The blue curve represents the ground spectrum before upscaling, the red curve represents the ground spectrum after upscaling, and "○" represents the reflectance value of the Sentinel2 satellite on that day.

[0132] (303) Apply matrix A to the reflectance data at the high-resolution pixel scale, perform matrix operations to achieve the conversion to the target area scale (500m), complete the spatial scale expansion, and obtain the raster reflectance data with a resolution of 10m, a spatial range of 500m×500m, and 4 bands (B2, B3, B4, B8), as shown in the schematic diagram. Figure 8 and Figure 9 As shown.

[0133] 4. In the reflectance spectral scale extension stage, the grid reflectance data of the four bands are subjected to spectral scale extension to expand them to the range of the ground-measured spectrum. Specific steps include:

[0134] (401) Based on the ground equivalent standard spectral library, the Pearson correlation coefficient is used as the similarity matching index. The generated grid reflectance data is spectrally matched pixel by pixel, and the date and time of the best-fit measurement record are recorded.

[0135] (402) Based on the full-spectrum standard spectral library, according to the date and time obtained in the previous step, the corresponding ground measurement spectrum is selected as the spectral extension result of the corresponding pixel.

[0136] (403) Repeat this operation for all pixels to obtain high-resolution pixel-scale grid reflectance data, whose spectral range is consistent with the ground measurement, spatial range is 500m×500m, and spatial resolution is 10m, as shown in the schematic diagram. Figure 10 As shown.

[0137] In summary, this invention fully considers the representativeness issue of single-site measurements at the high-resolution pixel scale, relying on high-resolution satellite data as a bridge for spatial information to obtain spatial spread coefficients to assist in the spatial reconstruction of low-resolution spectral information. In this process, traditional remote sensing auxiliary parameters such as DEM, albedo, land cover, and NDVI are not introduced, thus avoiding dependence on multi-source samples and external data. Simultaneously, high-resolution image reflectance data is only used in the derivation of spatial spread coefficients; once the spatial spread coefficients are obtained, subsequent processing no longer relies on this image reflectance data, reducing the continuous dependence on specific remote sensing samples. The final spectral information is obtained from ground-measured spectra inversion, rather than reflectance data from high-resolution satellite imagery, therefore the results are more physically meaningful and practically representative, effectively mitigating the error accumulation problem caused by sample inconsistency. A similarity matching index is introduced when calculating the upscaling coefficient, and a method based on spectral matching from a historical ground-measured standard spectral library is used to obtain reference spectra, ensuring the authenticity and reliability of the data. By comprehensively utilizing temporal and spatial information, and using long-term high-resolution satellite reflectance products as a bridge, and fully considering the spatial relationship and reflectance variations between ground measurement points and surrounding pixels, the upscaled high-resolution pixel-scale reflectance is further expanded in both spatial and spectral scales. This extends the spatial range to the pixel-scale range of the target resolution (500m), resulting in raster reflectance data. The raster reflectance data not only expands the spatial scale but also retains the spectral information from ground measurements, achieving an expansion of the spectral range of high-resolution satellite reflectance.

[0138] This invention addresses the spatial range difference between ground-based single-point measurements and high-resolution satellite pixel scales by upscaling ground-based single-point measurements to the high-resolution satellite pixel scale, avoiding the direct use of single-point measurements as the true pixel-scale value. Simultaneously, by combining time-series information and spatial distribution characteristics, the spatial range of pixel-scale reflectance data is expanded to obtain raster reflectance data. This invention first upscales ground-based single-point measurements to the high-resolution pixel scale, and then further expands to a 500m spatial range, achieving multi-level scale conversion and providing data support for obtaining pixel-scale reflectance products from coarser-resolution satellites. This invention constructs a standard spectral library and selects standard reference spectra based on similarity matching evaluation indicators, calculating upscaling coefficients. In other words, it extracts the true spectra of ground objects based on actual measurement data, improving the accuracy and reliability of spectral information during the scale conversion process. This provides multi-scale, highly consistent surface reflectance data support for remote sensing inversion, surface process simulation, and environmental monitoring.

[0139] The surface reflectance data scale conversion device provided by the present invention is described below. The surface reflectance data scale conversion device described below and the surface reflectance data scale conversion method described above can be referred to in correspondence.

[0140] Figure 11 This is a schematic diagram of the surface reflectance data scaling device provided by the present invention, as shown below. Figure 11 As shown, the present invention also provides a surface reflectance data scaling device, comprising:

[0141] The acquisition module 1110 is used to acquire image reflectance data of at least one satellite corresponding to the ground measurement location;

[0142] The filtering module 1120 is used to filter the target ground measurement spectrum corresponding to the image reflectance data from a pre-constructed standard spectral library based on the spectral correlation coefficient, and use the target ground measurement spectrum as the standard reference spectrum of the image reflectance data.

[0143] The coefficient determination module 1130 is used to compare the standard reference spectrum with the ground measurement spectrum under the same date under satellite-to-ground synchronization to obtain the first scale conversion coefficient under satellite-to-ground synchronization; and to perform interpolation processing on the first scale conversion coefficient to obtain the second scale conversion coefficient under non-satellite-to-ground synchronization.

[0144] The scale conversion module 1140 is used to perform scale conversion on ground measurement reflectance data under satellite-to-ground synchronization based on the first scale conversion coefficient, and to perform scale conversion on ground measurement reflectance data under non-satellite-to-ground synchronization based on the second scale conversion coefficient, so as to obtain surface reflectance data at the pixel scale.

[0145] As one embodiment, the standard spectral library includes a ground-equivalent standard spectral library and a full-spectrum standard spectral library. Correspondingly, the screening module 1120 is further used for:

[0146] Based on the spectral correlation coefficient between the ground equivalent standard spectral library and the image reflectance data, the ground equivalent spectrum with the best fitting effect and the corresponding measurement time are obtained.

[0147] Based on the measurement time of the ground equivalent spectrum with the best fitting effect, ground measurement spectra with the corresponding measurement time are selected from the full-spectrum standard spectrum library, and the ground measurement spectra are used as the target ground measurement spectra.

[0148] As an example, the full-spectrum standard spectral library is obtained by performing spectral filtering and format normalization on measured surface reflectance data of the time series, and the ground equivalent standard spectral library is obtained by spectral convolution of the full-spectrum standard spectral library and the satellite spectral response function.

[0149] As one embodiment, it also includes:

[0150] The spatial scale extension module is used to spatially extend the surface reflectance data at the pixel scale based on the image reflectance data of at least one satellite within a time scale, so as to obtain the raster reflectance data within the target area scale.

[0151] As one embodiment, the spatial scale expansion module is also used for:

[0152] Based on the surface reflectance data at the pixel scale, the equivalent ground reflectance at the pixel scale is obtained;

[0153] Based on the image reflectance data of at least one satellite over a time scale, determine the central pixel and surrounding pixels within the target area scale centered on the ground measurement location;

[0154] Based on the image reflectance data, a minimum cost function is constructed between the center pixel and any of the surrounding pixels. The minimum cost function is used to characterize the least squares sum of the differences between the pixel reflectance of the surrounding pixels and the ground reflectance of the center pixel.

[0155] Solving the minimum cost function yields the space expansion coefficient matrix;

[0156] Based on the spatial expansion coefficient matrix, the surface reflectance data at the pixel scale is spatially expanded to obtain the raster reflectance data at the target area scale.

[0157] As one embodiment, it also includes:

[0158] The spectral scale extension module is used to calculate the spectral correlation coefficient, and to select the ground measurement spectrum of any pixel within the target area scale on any date within the time scale from the standard spectral library, so as to obtain the spectral scale extension result of the grid reflectance data.

[0159] Figure 12 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 12 As shown, the electronic device may include: a processor 1210, a communications interface 1220, a memory 1230, and a communication bus 1240, wherein the processor 1210, the communications interface 1220, and the memory 1230 communicate with each other via the communication bus 1240. The processor 1210 can call logical instructions in the memory 1230 to execute a surface reflectance data scaling method, which includes:

[0160] Obtain image reflectance data for at least one satellite corresponding to the ground measurement location;

[0161] Based on the spectral correlation coefficient, the target ground measurement spectrum corresponding to the image reflectance data is obtained by screening from a pre-constructed standard spectral library, and the target ground measurement spectrum is used as the standard reference spectrum of the image reflectance data;

[0162] By comparing the standard reference spectrum with the ground measurement spectrum under the same date under satellite-to-ground synchronization, the first scale conversion coefficient under satellite-to-ground synchronization is obtained; the first scale conversion coefficient is interpolated to obtain the second scale conversion coefficient under non-satellite-to-ground synchronization.

[0163] Based on the first scale conversion coefficient, the ground measurement reflectance data under satellite-to-ground synchronization is scaled, and based on the second scale conversion coefficient, the ground measurement reflectance data under non-satellite-to-ground synchronization is scaled to obtain pixel-scale surface reflectance data.

[0164] Furthermore, the logical instructions in the aforementioned memory 1230 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0165] On the other hand, the present invention also provides a computer program product, the computer program product comprising a computer program that can be stored on a non-transitory computer-readable storage medium, wherein when the computer program is executed by a processor, the computer is able to execute the surface reflectance data scaling method provided by the above methods, the method comprising:

[0166] Obtain image reflectance data for at least one satellite corresponding to the ground measurement location;

[0167] Based on the spectral correlation coefficient, the target ground measurement spectrum corresponding to the image reflectance data is obtained by screening from a pre-constructed standard spectral library, and the target ground measurement spectrum is used as the standard reference spectrum of the image reflectance data;

[0168] By comparing the standard reference spectrum with the ground measurement spectrum under the same date under satellite-to-ground synchronization, the first scale conversion coefficient under satellite-to-ground synchronization is obtained; the first scale conversion coefficient is interpolated to obtain the second scale conversion coefficient under non-satellite-to-ground synchronization.

[0169] Based on the first scale conversion coefficient, the ground measurement reflectance data under satellite-to-ground synchronization is scaled, and based on the second scale conversion coefficient, the ground measurement reflectance data under non-satellite-to-ground synchronization is scaled to obtain pixel-scale surface reflectance data.

[0170] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the surface reflectance data scaling method provided by the methods described above, the method comprising:

[0171] Obtain image reflectance data for at least one satellite corresponding to the ground measurement location;

[0172] Based on the spectral correlation coefficient, the target ground measurement spectrum corresponding to the image reflectance data is obtained by screening from a pre-constructed standard spectral library, and the target ground measurement spectrum is used as the standard reference spectrum of the image reflectance data;

[0173] By comparing the standard reference spectrum with the ground measurement spectrum under the same date under satellite-to-ground synchronization, the first scale conversion coefficient under satellite-to-ground synchronization is obtained; the first scale conversion coefficient is interpolated to obtain the second scale conversion coefficient under non-satellite-to-ground synchronization.

[0174] Based on the first scale conversion coefficient, the ground measurement reflectance data under satellite-to-ground synchronization is scaled, and based on the second scale conversion coefficient, the ground measurement reflectance data under non-satellite-to-ground synchronization is scaled to obtain pixel-scale surface reflectance data.

[0175] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; 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. Those skilled in the art can understand and implement this without any creative effort.

[0176] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0177] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for scaling surface reflectance data, the method comprising: The method comprises the following steps: acquiring image reflectivity data corresponding to at least one satellite at a ground measurement position; based on the spectral correlation coefficient, a target ground measurement spectrum corresponding to the image reflectivity data is selected from a pre-constructed standard spectrum library, and the target ground measurement spectrum is taken as a standard reference spectrum of the image reflectivity data; comparing the standard reference spectrum with a ground measurement spectrum under a satellite-ground synchronization condition on the same date to obtain a first scale conversion coefficient under the satellite-ground synchronization condition; and performing interpolation processing on the first scale conversion coefficient to obtain a second scale conversion coefficient under a non-satellite-ground synchronization condition; based on the first scale conversion coefficient, performing scale conversion on ground measurement reflectivity data under the satellite-ground synchronization condition, and based on the second scale conversion coefficient, performing scale conversion on ground measurement reflectivity data under the non-satellite-ground synchronization condition to obtain pixel-scale ground surface reflectivity data; the standard spectrum library comprises a ground equivalent standard spectrum library and a full-spectrum standard spectrum library, and correspondingly, the target ground measurement spectrum corresponding to the image reflectivity data is selected from the pre-constructed standard spectrum library based on the spectral correlation coefficient, which comprises the following steps: based on the spectral correlation coefficient between the ground equivalent standard spectrum library and the image reflectivity data, a ground equivalent spectrum with the optimal fitting effect and a corresponding measurement time are obtained; based on the measurement time of the ground equivalent spectrum with the optimal fitting effect, a ground measurement spectrum corresponding to the measurement time is selected from the full-spectrum standard spectrum library, and the ground measurement spectrum is taken as the target ground measurement spectrum.

2. The method of claim 1, wherein The full-spectrum standard spectrum library is obtained by performing spectral filtering and format standardization processing on time-series actual measurement ground surface reflectivity data, and the ground equivalent standard spectrum library is obtained by spectral convolution of the full-spectrum standard spectrum library and a satellite spectral response function.

3. The method of claim 1, wherein Further comprising: based on the image reflectivity data of at least one satellite within a time scale, the pixel-scale ground surface reflectivity data is spatially scaled to obtain grid reflectivity data within a target area scale.

4. The method of claim 3, wherein The method for spatially scaling the pixel-scale ground surface reflectivity data based on the image reflectivity data of at least one satellite within a time scale to obtain grid reflectivity data within a target area scale comprises the following steps: based on the pixel-scale ground surface reflectivity data, pixel-scale ground equivalent reflectivity is obtained; based on the image reflectivity data of at least one satellite within a time scale, a center pixel within a target area scale centered on the ground measurement position and a surrounding pixel are determined; based on the image reflectivity data, a minimum cost function between the center pixel and any surrounding pixel is constructed, and the minimum cost function is used to represent the least square sum of the difference between the pixel reflectivity of the surrounding pixel and the ground reflectivity of the center pixel; the minimum cost function is solved to obtain a spatial expansion coefficient matrix; based on the spatial expansion coefficient matrix, the pixel-scale ground surface reflectivity data is spatially expanded to obtain the grid reflectivity data within the target area scale.

5. The method of claim 3, wherein After the satellite image reflectivity data of at least one satellite is used to expand the ground surface reflectivity data of the pixel scale to the target area scale, the method further comprises: calculating a spectral correlation coefficient, and screening a ground measurement spectrum of any pixel in the target area scale on any date in the time scale from the standard spectrum library to obtain a spectral scale expansion result of the grid reflectivity data.

6. An apparatus for scaling surface reflectance data, the apparatus comprising: The method comprises: an acquisition module configured to acquire satellite image reflectivity data of at least one satellite corresponding to a ground measurement position; a screening module configured to screen a target ground measurement spectrum corresponding to the satellite image reflectivity data from a pre-constructed standard spectrum library based on a spectral correlation coefficient, and use the target ground measurement spectrum as a standard reference spectrum of the satellite image reflectivity data; a coefficient determination module configured to compare the standard reference spectrum with a ground measurement spectrum in a satellite-ground synchronous condition on the same date to obtain a first scale conversion coefficient in the satellite-ground synchronous condition, and perform interpolation processing on the first scale conversion coefficient to obtain a second scale conversion coefficient in a non-satellite-ground synchronous condition; a scale conversion module configured to perform scale conversion on ground measurement reflectivity data in a satellite-ground synchronous condition based on the first scale conversion coefficient, and perform scale conversion on ground measurement reflectivity data in a non-satellite-ground synchronous condition based on the second scale conversion coefficient to obtain ground surface reflectivity data of a pixel scale; the standard spectrum library comprises a ground equivalent standard spectrum library and a full-spectrum standard spectrum library, and correspondingly, the target ground measurement spectrum corresponding to the satellite image reflectivity data is screened from the pre-constructed standard spectrum library based on a spectral correlation coefficient, which comprises: based on the spectral correlation coefficient between the ground equivalent standard spectrum library and the satellite image reflectivity data, obtaining a ground equivalent spectrum with the optimal fitting effect and a corresponding measurement time; based on the measurement time of the ground equivalent spectrum with the optimal fitting effect, screening a ground measurement spectrum corresponding to the measurement time from the full-spectrum standard spectrum library, and using the ground measurement spectrum as the target ground measurement spectrum.

7. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor executes the computer program to realize the ground surface reflectivity data scale conversion method according to any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the ground surface reflectivity data scale conversion method according to any one of claims 1 to 5.

9. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to realize the ground surface reflectivity data scale conversion method according to any one of claims 1 to 5.

Images