Satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform

Abstract

The invention provides a satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform. The satellite vegetation index fusion method comprises the following steps: acquiring a meteorological vegetation index product and a high-resolution vegetation index product; taking a specified vegetation index product as a background field, and performing multi-stage interpolation processing based on optimal interpolation and plane equation interpolation on the meteorological vegetation index product to obtain a vegetation index interpolation; fast Fourier transform processing is carried out on the high-resolution vegetation index product to remove the abutted seam area contained in the high-resolution vegetation index product, and a new high-resolution vegetation index product is obtained; the vegetation index interpolation is fused with texture features corresponding to the new high-resolution vegetation index product, a vegetation index fusion result is obtained, and the texture features are used for describing the vegetation index change condition between two adjacent pixels in the new high-resolution vegetation index product. According to the method, the precision of a vegetation index fusion result can be remarkably improved in a complex terrain or an area where multiple vegetation types coexist.

Description

Technical Field

[0001] The present invention relates to the technical field of vegetation index fusion, and in particular to a satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform. Background Art

[0002] With the rapid progress of earth observation technology, remote sensing has entered the stage of multi-platform, multi-sensor and multi-angle observation development, significantly improving the ability to obtain high spatial resolution and high spectral resolution data. However, limited by the optical diffraction limit, the modulation transfer function of the imaging system and the signal-to-noise ratio, it is difficult to achieve spaceborne remote sensing images with both high spatial and high spectral resolutions. Although panchromatic and multispectral cameras can provide images with high spatial resolution, their spectral resolution is low and the number of bands is small; although hyperspectral cameras have extremely high spectral resolution and can capture spectral cube data with nanometer-level resolution, covering visible light, near-infrared, short-wave infrared, and even mid-infrared and thermal infrared bands, and can provide images with up to hundreds of narrow spectral bands, factors such as payload platform flutter, imaging blur caused by the optical system transfer function, atmospheric radiation, and cloud occlusion will also lead to a decrease in the radiation information quality of hyperspectral images, low spatial resolution, and the occurrence of mixed pixel phenomena, which are particularly prominent in hyperspectral image analysis, understanding, and pattern recognition.

[0003] Domestic Fengyun meteorological satellites (FY series) play an important role in providing vegetation index (such as NDVI, EVI, etc.) products, especially in large-scale ecological vegetation monitoring and climate change research. However, such products also have significant drawbacks, especially the limitations in spatial and temporal resolutions, making it difficult to meet the needs of refined monitoring. First, the vegetation index products of Fengyun meteorological satellites have a low spatial resolution, usually at the level of 250 meters to 1 kilometer, which has limited monitoring ability for small-scale ecosystem processes, is vulnerable to the influence of the mixed pixel effect, and it is difficult to accurately characterize local vegetation changes in areas with high surface condition heterogeneity. Second, although Fengyun meteorological satellites have a high imaging frequency, due to factors such as cloud cover and observation delay, the actual temporal resolution of the products is limited; in areas with frequent cloud cover, the spatial continuity is often affected due to sparse data. In addition, due to the low spectral resolution of the sensor, the quantitative accuracy overly depends on atmospheric correction and inversion algorithms, resulting in errors easily occurring in areas with complex terrain or coexistence of multiple vegetation types. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform, which can significantly improve the accuracy of the vegetation index fusion result in areas with complex terrain or coexistence of multiple vegetation types.

[0005] In a first aspect, the present invention provides a satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform, including:

[0006] Obtain a meteorological vegetation index product and a high-resolution vegetation index product;

[0007] Using the specified vegetation index product as the background field, perform multi-level interpolation processing on the meteorological vegetation index product based on optimal interpolation and plane equation interpolation to obtain a vegetation index interpolation;

[0008] Perform fast Fourier transform processing on the high-resolution vegetation index product to eliminate the stitching area included in the high-resolution vegetation index product, and obtain a new high-resolution vegetation index product;

[0009] Fuse the vegetation index interpolation with the texture features corresponding to the new high-resolution vegetation index product, where the texture features are used to describe the change in vegetation index between two adjacent pixels in the new high-resolution vegetation index product, to obtain a vegetation index fusion result.

[0010] In one implementation, using the specified vegetation index product as the background field, performing multi-level interpolation processing on the meteorological vegetation index product based on optimal interpolation and plane equation interpolation to obtain a vegetation index interpolation includes:

[0011] For any first grid point in the first grid to be interpolated, determine the vegetation index observation values corresponding to multiple observation points adjacent to the first grid point from the meteorological vegetation index product, and determine the vegetation index background values corresponding to each observation point from the specified vegetation index product. Based on the vegetation index observation values and the vegetation index background values, perform optimal interpolation processing on the first grid point to obtain the vegetation index analysis value corresponding to the first grid point;

[0012] For any second grid point in the second grid to be interpolated, determine the target grid in the first grid where the second grid point is located. Based on the vegetation index analysis value corresponding to the first grid point included in the target grid, perform plane equation interpolation processing on the second grid point to obtain the vegetation index interpolation corresponding to the second grid point;

[0013] Wherein, the resolution of the second grid is higher than that of the first grid.

[0014] In one implementation, based on the vegetation index observation values and the vegetation index background values, performing optimal interpolation processing on the first grid point to obtain the vegetation index analysis value corresponding to the first grid point includes:

[0015] Taking the minimum error correlation coefficient corresponding to the first grid point as the target, assign target weight coefficients to each observation point;

[0016] Determine the vegetation index difference between the vegetation index observation value and the vegetation index background value corresponding to the same observation point, and perform weighted fusion on all vegetation index differences based on the target weight coefficient to obtain the vegetation index correction value;

[0017] Use the vegetation index correction value to correct the vegetation index background value corresponding to the first grid point to obtain the vegetation index analysis value corresponding to the first grid point.

[0018] In one implementation, with the goal of minimizing the error correlation coefficient corresponding to the first grid point, assign target weight coefficients to each observation point, including:

[0019] Denote any two observation points adjacent to the first grid point as the first observation point and the second observation point respectively;

[0020] Based on the vegetation index observation value corresponding to the first observation point and the vegetation index observation value corresponding to the second observation point, determine the first error correlation coefficient; and, based on the vegetation index background value corresponding to the first observation point and the vegetation index background value corresponding to the second observation point, determine the second error correlation coefficient;

[0021] Based on the first error correlation coefficient, the second error correlation coefficient, the ratio of the error standard deviation corresponding to the first observation point, and the ratio of the error standard deviation corresponding to the second observation point, determine the target error correlation coefficient between the first observation point and the second observation point;

[0022] Use the current weight coefficient to perform weighted fusion on the target error correlation coefficient to obtain the grid point error correlation coefficient corresponding to the first grid point, adjust the current weight coefficient, and continue to use the new current weight coefficient to perform weighted fusion on the target error correlation coefficient until the obtained grid point error correlation coefficient is minimized to determine the target weight coefficient.

[0023] In one implementation, based on the vegetation index analysis value corresponding to the first grid point included in the target grid, perform plane equation interpolation on the second grid point to obtain the vegetation index interpolation corresponding to the second grid point, including:

[0024] Divide the target grid into an upper triangle and a lower triangle;

[0025] Use the vegetation index analysis value corresponding to the first grid point included in the upper triangle to fit the triangular plane coefficients and intercepts corresponding to the upper triangle; and, use the vegetation index analysis value corresponding to the first grid point included in the lower triangle to fit the triangular plane coefficients and intercepts corresponding to the lower triangle;

[0026] Judge whether the second grid point is inside the upper triangle or inside the lower triangle, and based on the triangular plane coefficients and intercepts corresponding to the target triangle where the second grid point is located, determine the vegetation index interpolation corresponding to the second grid point.

[0027] In one embodiment, a fast Fourier transform is performed on the high-resolution vegetation index product to remove the seam regions included in the high-resolution vegetation index product, and a new high-resolution vegetation index product is obtained, including:

[0028] Using the fast Fourier transform, the high-resolution vegetation index product is converted from the spatial domain to the frequency domain to obtain a vegetation index spectrogram;

[0029] The low-frequency components in the vegetation index spectrogram are moved from the edge position of the vegetation index spectrogram to the center position of the vegetation index spectrogram to obtain a new vegetation index spectrogram;

[0030] The low-frequency components are filtered out from the new vegetation index spectrogram through a filter to remove the seam regions included in the high-resolution vegetation index product;

[0031] Using the inverse frequency domain conversion, the vegetation index spectrogram with the low-frequency components filtered out is converted from the frequency domain to the spatial domain to obtain a new high-resolution vegetation index product.

[0032] In one embodiment, the vegetation index interpolation is fused with the texture features corresponding to the new high-resolution vegetation index product to obtain a vegetation index fusion result, including:

[0033] The texture features corresponding to the new high-resolution vegetation index product are normalized;

[0034] The vegetation index interpolation is adjusted using the normalized texture features so that the change between the adjusted vegetation index interpolations of two adjacent pixels meets the vegetation index change described by the texture features, and a vegetation index fusion result is obtained.

[0035] In a second aspect, the present invention also provides a satellite vegetation index fusion device based on optimal interpolation and fast Fourier transform, including:

[0036] A product acquisition module for acquiring a meteorological vegetation index product and a high-resolution vegetation index product;

[0037] An interpolation module for performing multi-level interpolation processing on the meteorological vegetation index product based on optimal interpolation and plane equation interpolation with a specified vegetation index product as the background field to obtain a vegetation index interpolation;

[0038] A fast Fourier transform module for performing a fast Fourier transform on the high-resolution vegetation index product to remove the seam regions included in the high-resolution vegetation index product and obtain a new high-resolution vegetation index product;

[0039] A fusion module is used to fuse the vegetation index interpolation with the texture features corresponding to the new high - resolution vegetation index product to obtain a vegetation index fusion result. The texture features are used to describe the change of the vegetation index between two adjacent pixels in the new high - resolution vegetation index product.

[0040] In a third aspect, the present invention further provides an electronic device, including a processor and a memory. The memory stores computer - executable instructions that can be executed by the processor. The processor executes the computer - executable instructions to implement the method according to any one of the first aspect.

[0041] In a fourth aspect, the present invention further provides a computer - readable storage medium. The computer - readable storage medium stores computer - executable instructions. When the computer - executable instructions are called and executed by the processor, the computer - executable instructions cause the processor to implement the method according to any one of the first aspect.

[0042] The satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform provided by the present invention first obtains a meteorological vegetation index product and a high - resolution vegetation index product; then, using the specified vegetation index product as the background field, performs multi - level interpolation processing on the meteorological vegetation index product based on optimal interpolation and plane - equation interpolation to obtain a vegetation index interpolation; then performs fast Fourier transform processing on the high - resolution vegetation index product to eliminate the mosaic area included in the high - resolution vegetation index product and obtain a new high - resolution vegetation index product; finally, fuses the vegetation index interpolation with the texture features corresponding to the new high - resolution vegetation index product to obtain a vegetation index fusion result. The texture features are used to describe the change of the vegetation index between two adjacent pixels in the new high - resolution vegetation index product. The above - mentioned method can not only keep the meteorological vegetation index product from being distorted during the fusion process to the greatest extent by using plane - equation interpolation, but also solve the problem of unevenness of the vegetation index fusion result. And using optimal interpolation can effectively make up for the information loss problem caused by the observation angle and cloud occlusion during the fusion process, improving the spatial consistency and temporal continuity of the vegetation index fusion result. By combining optimal interpolation and plane - equation interpolation, the present invention not only maintains the details and accuracy of the meteorological vegetation index product, but also can improve the quality and application reliability of the vegetation index fusion result in space and time; in addition, the application of fast Fourier transform is to extract texture from high - resolution satellite climate (monthly, weekly, pentad, etc.) mosaic images. Due to the obvious systematic errors, daily variations, etc. in the climate synthesis data of the high - resolution vegetation index product, the texture features of the high - resolution vegetation index product are extracted through fast Fourier frequency - domain conversion, further improving the accuracy of the vegetation index fusion result in areas with complex terrain or co - existing multiple vegetation types.

[0043] Other features and advantages of the present invention will be set forth in the following description, and in part will be obvious from the description, or may be learned by practice of the present invention. The objectives and other advantages of the present invention are realized and attained by the structure particularly pointed out in the specification, claims as well as the drawings.

[0044] In order to make the above objectives, features and advantages of the present invention more obvious and understandable, the following specifically gives preferred embodiments and, in conjunction with the accompanying drawings, the detailed description is as follows. Brief Description of the Drawings

[0045] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0046] Figure 1 It is a schematic flowchart of a satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform provided by an embodiment of the present invention;

[0047] Figure 2 It is a technical framework diagram of a satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform provided by an embodiment of the present invention;

[0048] Figure 3 It is a schematic diagram of the preprocessing of a meteorological vegetation index product provided by an embodiment of the present invention;

[0049] Figure 4 It is a schematic structural diagram of a satellite vegetation index fusion device based on optimal interpolation and fast Fourier transform provided by an embodiment of the present invention;

[0050] Figure 5 It is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. Detailed Description of the Embodiments

[0051] In order to make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions of the present invention in conjunction with the embodiments. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0052] At present, the vegetation index provided by the domestic Fengyun meteorological satellite (FY series) is prone to errors in areas with complex terrain or coexistence of multiple vegetation types. In view of these shortcomings, data fusion combined with high-resolution satellites has become an effective solution.

[0053] High-resolution satellites (such as the Gaofen series, Sentinel-2, etc.) have higher spatial resolution and spectral sensitivity, which can significantly improve the accuracy and scope of application of vegetation index products. By integrating the fine spatial information of high-resolution satellites, the spatial resolution of meteorological satellite vegetation index products can be effectively improved to expand their application to small-scale vegetation dynamic monitoring; using the data interpolation and completion capabilities of high-resolution satellites, the spatiotemporal breakpoint problems caused by cloud cover or observation intervals of domestic Fengyun meteorological satellites can also be improved. In addition, the hyperspectral data of high-resolution satellites can correct the mixed pixel effects and inversion errors of Fengyun meteorological satellites, thereby improving the quantitative accuracy of vegetation indices and realizing the organic unity of large-scale background monitoring and local refined analysis. The coordinated application of high-resolution satellites and meteorological satellites not only makes up for the shortcomings of single satellite products, but also shows significant advancement and application potential in the fields of ecological monitoring, precision agriculture and disaster response.

[0054] Based on this, the present invention provides a satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform, which can significantly improve the accuracy of vegetation index fusion results in areas with complex terrain or coexistence of multiple vegetation types.

[0055] To facilitate understanding of this embodiment, firstly, a satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform disclosed in an embodiment of the present invention is described in detail, see Figure 1 The flowchart of a satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform is shown in FIG. 1 , and the method mainly includes the following steps S102 to S108:

[0056] Step S102, obtaining meteorological vegetation index products and high-resolution vegetation index products.

[0057] Among them, the meteorological vegetation index product includes the vegetation index observation values corresponding to multiple observation points in the study area, and the high-resolution vegetation index product also includes the vegetation index observation values corresponding to multiple observation points in the study area. Compared with the meteorological vegetation index product, the high-resolution vegetation index product has higher spatial resolution and spectral sensitivity.

[0058] Step S104, using the designated vegetation index product as the background field, performing multi-level interpolation processing based on optimal interpolation and plane equation interpolation on the meteorological vegetation index product to obtain vegetation index interpolation.

[0059] Exemplarily, the vegetation products of FY4 or FY3D can be used as the background field. In one example, first, using the vegetation products of FY4 or FY3D as the background field, the meteorological vegetation index product is processed based on optimal interpolation to obtain the vegetation index analysis value corresponding to each grid point in the first grid; then, plane equation interpolation is performed on the vegetation index analysis value corresponding to each grid point in the first grid to obtain the vegetation index interpolation corresponding to each grid point in the second grid, and the resolution of the second grid is higher than that of the first grid.

[0060] Step S106: Perform fast Fourier transform processing on the high-resolution vegetation index product to remove the stitching areas included in the high-resolution vegetation index product, obtaining a new high-resolution vegetation index product.

[0061] In one example, fast Fourier transform processing can be performed on the high-resolution vegetation index product to obtain a vegetation index spectrogram. After centering and low-frequency component filtering of the vegetation index spectrogram, the stitching areas included in the high-resolution vegetation index product can be removed. Then, inverse frequency domain conversion is performed on the vegetation index spectrogram at this time to obtain a new high-resolution vegetation index product without stitching areas.

[0062] Step S108: Fuse the vegetation index interpolation with the texture features corresponding to the new high-resolution vegetation index product to obtain a vegetation index fusion result.

[0063] Among them, the texture features are used to describe the change of vegetation index between two adjacent pixels in the new high-resolution vegetation index product. In one example, the texture features of the new high-resolution vegetation product index can be normalized and then fused with the vegetation index interpolation to make the change of vegetation index in the obtained vegetation index fusion result conform to the change of vegetation index described by the texture features.

[0064] The satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform provided by the embodiments of the present invention can not only maintain the meteorological vegetation index product without distortion to the greatest extent during the fusion process by using plane equation interpolation, but also solve the problem of unevenness of the vegetation index fusion result. The use of optimal interpolation can effectively make up for the information loss problems caused by the observation angle and cloud occlusion during the fusion process, improve the spatial consistency and temporal continuity of the vegetation index fusion result. It predicts and fills the unobserved area from adjacent data through weight calculation to ensure the smoothness and continuity of the fusion result. By combining the optimal interpolation with the plane equation interpolation, the present invention not only maintains the details and accuracy of the meteorological vegetation index product, but also improves the quality and application reliability of the vegetation index fusion result in space and time. In addition, the application of the fast Fourier transform is to extract the texture from the high-resolution satellite climate (monthly, weekly, dekadal, etc.) mosaicked images. Due to the obvious systematic errors, daily variations, etc. in the climate synthesis data of the high-resolution vegetation index product, the texture features of the high-resolution vegetation index product are extracted through the fast Fourier frequency domain conversion, further improving the accuracy of the vegetation index fusion result in areas with complex terrain or coexistence of multiple vegetation types.

[0065] For ease of understanding, the embodiments of the present invention provide an implementation manner of a satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform. The embodiments of the present invention propose to combine the meteorological vegetation index product with the high-resolution vegetation index product in the same area, and mainly use methods such as plane equation difference, optimal interpolation and fast Fourier transform to realize the fusion of multi-source meteorological vegetation index products and high-resolution vegetation index products. Refer to Figure 2 The technical framework diagram of a satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform shown in the figure includes: reading the data required for vegetation index fusion, that is, the meteorological vegetation index product and the high-resolution vegetation index product; performing optimal interpolation fusion on the meteorological vegetation index product; performing plane equation interpolation on the analyzed value of the fused vegetation index to prepare for subsequent fusion with the high-resolution vegetation index product; performing Fourier transform on the high-resolution vegetation index product to obtain texture features; realizing the fusion of the meteorological vegetation index product and the high-resolution vegetation index product through data normalization, texture superposition, etc.; outputting vegetation index fusion-related products and information and other processes.

[0066] The specific implementation process is as follows:

[0067] For the foregoing step S104, the embodiments of the present invention provide an implementation manner of performing multi-level interpolation processing on the meteorological vegetation index product based on optimal interpolation and plane equation interpolation with a specified vegetation index product as the background field to obtain the vegetation index interpolation. Refer to the following steps 1 to 2:

[0068] Step 1. For any first grid point in the first grid to be interpolated, determine the vegetation index observation values corresponding to multiple observation points adjacent to the first grid point from the meteorological vegetation index product, and determine the vegetation index background value corresponding to each observation point from the specified vegetation index product. Based on the vegetation index observation values and the vegetation index background values, perform optimal interpolation processing on the first grid point to obtain the vegetation index analysis value corresponding to the first grid point. Specifically, refer to the following steps 1.1 to 1.3:

[0069] Step 1.1. With the goal of minimizing the error correlation coefficient corresponding to the first grid point, assign target weight coefficients to each observation point. Among them, the target weight coefficient W i is the value that minimizes the error of the vegetation index analysis value corresponding to the grid point. The determination process of the target weight coefficient W i is as follows:

[0070] (I) Denote any two observation points adjacent to the first grid point as the first observation point i and the second observation point j respectively.

[0071] (II) Based on the vegetation index observation value corresponding to the first observation point i and the vegetation index observation value corresponding to the second observation point j, determine the first error correlation coefficient; and, based on the vegetation index background value corresponding to the first observation point i and the vegetation index background value corresponding to the second observation point j, determine the second error correlation coefficient. Among them, the first error correlation coefficient is also the error correlation coefficient between the vegetation index observation values corresponding to the first observation point i and the second observation point j, and the second error correlation coefficient is also the correlation coefficient error between the vegetation index background values corresponding to the first observation point i and the second observation point j.

[0072] The embodiment of the present invention provides an implementation manner for determining the second error correlation coefficient and determines the second error correlation coefficient according to the following formula

[0073]

[0074] where r Z and r m respectively represent the longitudinal and latitudinal distances between the first observation point i and the second observation point j, and L Z and L m respectively represent the influence radii in the two directions. Similarly, the first error correlation coefficient can be determined

[0075] (III) Determine the target error correlation coefficient between the first observation point i and the second observation point j based on the first error correlation coefficient, the second error correlation coefficient, the ratio of the error standard deviation corresponding to the first observation point i, and the ratio of the error standard deviation corresponding to the second observation point j.

[0076] The ratio of the error standard deviation can be expressed as: where σ O and σ B represent the error standard deviation of the observation point and the error standard deviation of the background field respectively, and the value of λ is taken as 1, that is, it is assumed that the error standard deviation of the observed value is consistent with the error standard deviation of the background field.

[0077] Based on this, the expression of the target error correlation coefficient between the first observation point i and the second observation point j is as follows:

[0078] (IV) Use the current weight coefficient to perform weighted fusion on the target error correlation coefficient to obtain the grid error correlation coefficient corresponding to the first grid point. Adjust the current weight coefficient, and continue to use the new current weight coefficient to perform weighted fusion on the target error correlation coefficient until the grid error correlation coefficient obtained is the smallest to determine the target weight coefficient.

[0079] where the grid error correlation coefficient corresponding to the first grid point g is calculated as follows:

[0080]

[0081] where N is the number of observation points (specifically valid observation points) adjacent to the first grid point g, is the correlation coefficient error between the vegetation index background values corresponding to the first observation point i and the second observation point j, is the error correlation coefficient between the vegetation index observed values corresponding to the first observation point i and the second observation point j, λ i is the ratio of the error standard deviation corresponding to the first observation point i, λ j is the ratio of the error standard deviation corresponding to the second observation point j, W i is the target weight coefficient, is the grid error correlation coefficient corresponding to the first grid point g.

[0082] Step 1.2, determine the vegetation index difference between the vegetation index observed value and the vegetation index background value corresponding to the same observation point, and perform weighted fusion on all vegetation index differences based on the target weight coefficient to obtain the vegetation index correction value. The expression of the vegetation index correction value is as follows: where O i is the vegetation index observed value of the first observation point i, B iis the vegetation index background value of the first observation point i.

[0083] Step 1.3: Use the vegetation index correction value to correct the vegetation index background value corresponding to the first grid point, and obtain the vegetation index analysis value corresponding to the first grid point.

[0084] Considering the characteristics such as the non - continuity of vegetation coverage data, the embodiment of the present invention considers using the optimal interpolation algorithm for fusion. Here, the vegetation products of FY4 or FY3D are used as the background field. The vegetation index analysis value participating in the analysis is finally calculated by adding the vegetation index background value at the first grid point and the vegetation index correction value at the first grid point. The correction value is obtained by weighting the deviation between the vegetation index observation value and the vegetation index background value of each effective observation point around the first grid point. The vegetation index analysis value A g corresponding to the first grid point g is calculated as follows:

[0085]

[0086] where, A g is the vegetation index analysis value corresponding to the first grid point g, B g is the vegetation index background value corresponding to the first grid point g, N is the number of observation points (specifically, effective observation points) adjacent to the first grid point g, O i is the vegetation index observation value of the first observation point i, B i is the vegetation index background value of the first observation point i, W i is the target weight coefficient. When there is no data in the observation area, the target weight coefficient W i is 0. In the embodiment of the present invention, the background field is an initial estimate of the vegetation index observation value, and the background field is corrected by the vegetation index observation value of the sensor and the weight function.

[0087] Step 2: For any second grid point in the second grid to be interpolated, determine the target grid where the second grid point is located in the first grid. Based on the vegetation index analysis value corresponding to the first grid point included in the target grid, perform plane equation interpolation processing on the second grid point to obtain the vegetation index interpolation corresponding to the second grid point.

[0088] The meteorological vegetation index product has higher accuracy compared with the high - resolution vegetation index product, but the spatial resolution is relatively low. Therefore, before fusing the data of the two, it is necessary to interpolate or downscale the meteorological vegetation index product. See Figure 3Schematic diagram of the preprocessing of a meteorological vegetation index product. The intersection points of the black solid lines represent the satellite observation points, and the white solid lines represent the data points (such as point (px, py)) that we need to obtain after preprocessing. In order to make the preprocessed data have better continuity, after testing and comparison of various methods (such as bilinear interpolation, inverse distance weighting, etc.), the conclusion is that the interpolation method using the plane equation can not only maintain the original value of the data to the greatest extent, but also obtain the smoothest interpolation result. For details, please refer to the following steps 2.1 to 2.4:

[0089] Step 2.1, determine the target grid in the first grid where the second grid point is located. Exemplarily, please continue to refer to Figure 3 , for the second grid point p(px, py), the four vertices (i.e., the first grid points) of the target grid where it is located are p1(x1, y1, z1), p2(x2, y2, z2), p3(x3, y3, z3), p4(x4, y4, z4) respectively, where x and y are the plane coordinates of the grid point, and z is the vegetation index analysis value of the grid point.

[0090] Step 2.2, divide the target grid into an upper triangle and a lower triangle. Exemplarily, take the triangle composed of p2, p3, and P4 as the lower triangle, and the triangle composed of p1, p2, and p3 as the upper triangle.

[0091] Step 2.3, use the vegetation index analysis values corresponding to the first grid points included in the upper triangle to fit the triangular plane coefficients and intercepts corresponding to the upper triangle; and, use the vegetation index analysis values corresponding to the first grid points included in the lower triangle to fit the triangular plane coefficients and intercepts corresponding to the lower triangle.

[0092] For example, given three points p1(x1, y1, z1), p2(x2, y2, z2), p3(x3, y3, z3), to determine the plane equation, the key is to find a normal vector of the plane. For this purpose, make vectors p1p2(x2 - x1, y2 - y1, z2 - z1), p1p3(x3 - x1, y3 - y1, z3 - z1). The plane normal is perpendicular to these two vectors. Therefore, the normal vector n is:

[0093]

[0094] Among them, i, j, k represent p1, p2, p3, and a, b, c represent the triangular plane coefficients.

[0095] According to the above plane equation, the triangular plane coefficients and intercepts corresponding to the upper and lower triangles are respectively:

[0096] Upper triangle:

[0097] a1 = y1 * z2 - y1 * z3 - y2 * z1 + y2 * z3 + y3 * z1 - y3 * z2;

[0098] b1 = -x1 * z2 + x1 * z3 + x2 * z1 - x2 * z3 - x3 * z1 + x3 * z2;

[0099] c1 = x1 * y2 - x1 * y3 - x2 * y1 + x2 * y3 + x3 * y1 - x3 * y2;

[0100] d1 = -x1 * y2 * z3 + x1 * y3 * z2 + x2 * y1 * z3 - x2 * y3 * z1 - x3 * y1 * z2 + x3 * y2 * z1;

[0101] Lower triangle:

[0102] a2 = y4 * z2 - y4 * z3 - y2 * z4 + y2 * z3 + y3 * z4 - y3 * z2;

[0103] b2 = -x4 * z2 + x4 * z3 + x2 * z4 - x2 * z3 - x3 * z4 + x3 * z2;

[0104] c2 = x4 * y2 - x4 * y3 - x2 * y4 + x2 * y3 + x3 * y4 - x3 * y2;

[0105] d2 = -x4 * y2 * z3 + x4 * y3 * z2 + x2 * y4 * z3 - x2 * y3 * z4 - x3 * y4 * z2 + x3 * y2 * z4.

[0106] Step 2.4, determine whether the second lattice point is inside the upper triangle or the lower triangle, so as to determine the vegetation index interpolation corresponding to the second lattice point based on the triangular plane coefficients and intercepts corresponding to the target triangle where the second lattice point is located.

[0107] In one example, the intersection point of the y vector where (px, py) is located and the line connecting the two points (x3, y3) and (x2, y2) can be calculated:

[0108] y32 = y3 - (x3 - px) / (x3 - x2) * (y3 - y2);

[0109] If py is less than y32, it is determined that the second lattice point is inside the upper triangle. At this time, the vegetation index interpolation zo can be calculated using the triangular plane coefficients and intercepts corresponding to the upper triangle:

[0110] zo = (-d1 - a1 * px - b1 * py) / c1

[0111] Otherwise, it is determined that the second lattice point is within the lower triangle. At this time, the vegetation index interpolation zo can be calculated using the triangular plane coefficients and intercepts corresponding to the lower triangle:

[0112] zo = (-d2 - a2 * px - b2 * py) / c2.

[0113] For the aforementioned step S106, the embodiments of the present invention provide a specific implementation manner of performing fast Fourier transform processing on the high - resolution vegetation index product to remove the stitched areas included in the high - resolution vegetation index product and obtain a new high - resolution vegetation index product.

[0114] This fusion algorithm extracts textures based on high - resolution satellite climate (monthly, weekly, dekadal, etc.) mosaic images. Since there are obvious stitches in the climate synthesis data of the high - resolution vegetation index product, these stitches contain systematic errors of the instrument, daily changes, etc., and these changes belong to low - frequency changes. Therefore, it is proposed to extract texture features in a specific frequency range from the high - resolution vegetation index product through steps of frequency domain conversion, spectrogram centering, filter design and application, and inverse frequency domain conversion.

[0115] Specifically, it includes the following steps a to d:

[0116] Step a, using the fast Fourier transform, convert the high - resolution vegetation index product from the spatial domain to the frequency domain to obtain the vegetation index spectrogram. The fast Fourier transform (FFT) converts the image from the spatial domain to the frequency domain, and the image is decomposed into a superposition of sine waves and cosine waves of different frequencies. The low - frequency components in the vegetation index spectrogram represent the overall structure of the image, such as smooth regions, uniform backgrounds, and large - scale brightness changes, while the high - frequency components represent the details, edges, and textures of the image.

[0117] Step b, move the low - frequency components in the vegetation index spectrogram from the edge positions of the vegetation index spectrogram to the center position of the vegetation index spectrogram to obtain a new vegetation index spectrogram.

[0118] This process is also spectrogram centering. In the vegetation index spectrogram calculated by FFT, the low - frequency components are default located at the four corners of the image, which makes the intuitive analysis of the vegetation index spectrogram less convenient. For the convenience of observation and processing, first move the low - frequency part to the center position of the vegetation index spectrogram. The central part of the vegetation index spectrogram contains low - frequency information, and the edge part contains high - frequency information. By centering the vegetation index spectrogram, it is easier to perform frequency filtering. Especially in image texture extraction, attention is paid to the low - frequency (overall structure) and high - frequency (detail texture) parts. Therefore, spectrogram centering of the vegetation index spectrogram can help to perform frequency domain processing more intuitively.

[0119] Step c: Filter out the low-frequency components from the new vegetation index spectrogram to eliminate the stitching areas included in the high-resolution vegetation index products.

[0120] First, design the filter, including: Frequency domain filtering enhances or suppresses specific frequency components by operating on the spectrum, thereby achieving effects such as texture extraction, denoising, and smoothing. Common frequency domain filters include low-pass filters, high-pass filters, and band-pass filters. Low-pass filters are used to remove high-frequency noise and retain the overall structure of the image, and are commonly used for image smoothing or detail removal; high-pass filters remove low-frequency components, retain the details and textures of the image, remove the smooth background, and only retain edge and texture information; band-pass filters are used to extract textures within a specific frequency range, especially suitable for extracting texture features existing within a certain specific spatial frequency range.

[0121] Then, apply the filter, including: By designing the filter and applying it to the vegetation index spectrogram, selectively remove specific low-frequency components in the frequency domain. The filter is usually implemented through a mask, which covers a specific area in the vegetation index spectrogram corresponding to the frequency range of interest. Using high-pass filtering masks the low-frequency area (at the center) and retains the high-frequency area (spectrum edge), highlighting the details and edges of the image.

[0122] Step d: Use inverse frequency domain transformation to transform the vegetation index spectrogram with low-frequency components filtered out from the frequency domain to the spatial domain, obtaining a new high-resolution vegetation index product.

[0123] After processing the image in the frequency domain, it is usually necessary to convert the frequency domain result back to the spatial domain through inverse fast Fourier transform. Through the inverse transformation, the processed vegetation index spectrogram is converted into a spatial domain image, which will reflect the effect of the frequency domain processing. The process of inverse transformation is to reverse-transform the frequency domain data back to the spatial domain and restore the specific content of the image. At this time, the high-frequency components in the frequency domain will be converted into the detail parts in the image.

[0124] For the aforementioned step S108, the embodiment of the present invention provides a specific implementation manner of fusing the vegetation index interpolation with the texture features corresponding to the new high-resolution vegetation index product to obtain a vegetation index fusion result. Normalize the texture features corresponding to the new high-resolution vegetation index product, and use the normalized texture features to adjust the vegetation index interpolation, so that the change situation between the adjusted vegetation index interpolations of two adjacent pixels meets the vegetation index change situation described by the texture features, obtaining the vegetation index fusion result.

[0125] In practical applications, to ensure the comparability and consistency of texture information during the fusion process, the extracted texture features are normalized to eliminate the dimensional differences and scale effects between data. After the normalization process, the processed texture features are superimposed and fused with the vegetation index interpolation to generate a vegetation index fusion result that complements each other between the texture and the product, thereby enhancing the expression ability and analysis value of remote sensing data.

[0126] In summary, the multi-source fusion algorithm for satellite vegetation index proposed in the embodiments of the present invention innovatively bases on domestic Fengyun series meteorological satellite products and effectively makes up for the deficiencies in the spatial distribution of data and systematic error problems by using the optimal interpolation method. The optimal interpolation algorithm fills the unobserved areas caused by factors such as observation angles and cloud occlusion through weight calculation, improving the spatial consistency and temporal continuity of the fused data. The embodiments of the present invention also innovatively apply the fast Fourier transform (FFT) technology to extract the texture features of high-resolution satellite climate data (such as monthly, weekly, and pentad scales), and successfully extract and correct the systematic errors and daily changes in the high-resolution data through frequency domain conversion, filter design, and inverse frequency domain conversion. By combining plane equation interpolation, normalization processing, etc. to fuse meteorological satellite products and high-resolution texture features, the spatial resolution and temporal continuity of the vegetation index products are improved, and the application effect and reliability of domestic remote sensing data in high resolution and multi-temporal are expanded.

[0127] Based on the foregoing embodiments, the embodiments of the present invention provide a satellite vegetation index fusion device based on optimal interpolation and fast Fourier transform, see Figure 4 the structural schematic diagram of a satellite vegetation index fusion device based on optimal interpolation and fast Fourier transform shown. The device mainly includes the following parts:

[0128] A product acquisition module 402, configured to acquire meteorological vegetation index products and high-resolution vegetation index products;

[0129] An interpolation module 404, configured to perform multi-level interpolation processing on the meteorological vegetation index product based on optimal interpolation and plane equation interpolation with a specified vegetation index product as the background field to obtain a vegetation index interpolation;

[0130] A fast Fourier transform module 406, configured to perform fast Fourier transform processing on the high-resolution vegetation index product to eliminate the mosaic area included in the high-resolution vegetation index product and obtain a new high-resolution vegetation index product;

[0131] A fusion module 408, configured to fuse the vegetation index interpolation with the texture features corresponding to the new high-resolution vegetation index product to obtain a vegetation index fusion result, where the texture features are used to describe the vegetation index change between two adjacent pixels in the new high-resolution vegetation index product.

[0132] The satellite vegetation index fusion device based on optimal interpolation and fast Fourier transform provided by the embodiments of the present invention can not only maximize the preservation of the meteorological vegetation index product without distortion during the fusion process by using plane equation interpolation, but also solve the problem of unevenness of the vegetation index fusion result. By using optimal interpolation, it can effectively make up for the information loss problem caused by the observation angle and cloud occlusion during the fusion process, improve the spatial consistency and temporal continuity of the vegetation index fusion result. It predicts and fills the unobserved area from adjacent data through weight calculation to ensure the smoothness and continuity of the fusion result. By combining the optimal interpolation with the plane equation interpolation, the present invention not only maintains the details and accuracy of the meteorological vegetation index product, but also can improve the quality and application reliability of the vegetation index fusion result in space and time. In addition, the application of the fast Fourier transform is to extract the texture from the high-resolution satellite climate (monthly, weekly, pentad, etc.) mosaic images. Due to the obvious systematic errors and daily variations in the climate synthesis data of the high-resolution vegetation index products, the texture features of the high-resolution vegetation index products are extracted through fast Fourier frequency domain conversion, further improving the accuracy of the vegetation index fusion result in areas with complex terrain or coexistence of multiple vegetation types.

[0133] In one implementation manner, the interpolation module 404 is specifically configured to:

[0134] For any first grid point in the first grid to be interpolated, determine the vegetation index observation values corresponding to multiple observation points adjacent to the first grid point from the meteorological vegetation index product, and determine the vegetation index background values corresponding to each observation point from the specified vegetation index product. Based on the vegetation index observation values and the vegetation index background values, perform optimal interpolation processing on the first grid point to obtain the vegetation index analysis value corresponding to the first grid point;

[0135] For any second grid point in the second grid to be interpolated, determine the target grid where the second grid point is located in the first grid. Based on the vegetation index analysis value corresponding to the first grid point included in the target grid, perform plane equation interpolation processing on the second grid point to obtain the vegetation index interpolation corresponding to the second grid point;

[0136] Wherein, the resolution of the second grid is higher than that of the first grid.

[0137] In one implementation manner, the interpolation module 404 is specifically configured to:

[0138] Taking the minimum error correlation coefficient corresponding to the first grid point as the target, assign target weight coefficients to each observation point;

[0139] Determine the vegetation index difference between the vegetation index observation value and the vegetation index background value corresponding to the same observation point, and perform weighted fusion on all vegetation index differences based on the target weight coefficients to obtain the vegetation index correction value;

[0140] Using the vegetation index correction value, correct the vegetation index background value corresponding to the first grid point to obtain the vegetation index analysis value corresponding to the first grid point.

[0141] In one implementation, the interpolation module 404 is specifically configured to:

[0142] Denote any two observation points adjacent to the first grid point as the first observation point and the second observation point respectively;

[0143] Based on the vegetation index observation value corresponding to the first observation point and the vegetation index observation value corresponding to the second observation point, determine the first error correlation coefficient; and, based on the vegetation index background value corresponding to the first observation point and the vegetation index background value corresponding to the second observation point, determine the second error correlation coefficient;

[0144] Based on the first error correlation coefficient, the second error correlation coefficient, the ratio of the error standard deviation corresponding to the first observation point, and the ratio of the error standard deviation corresponding to the second observation point, determine the target error correlation coefficient between the first observation point and the second observation point;

[0145] Use the current weight coefficient to perform weighted fusion on the target error correlation coefficient to obtain the grid error correlation coefficient corresponding to the first grid point, adjust the current weight coefficient, and continue to use the new current weight coefficient to perform weighted fusion on the target error correlation coefficient until the obtained grid error correlation coefficient is the smallest to determine the target weight coefficient.

[0146] In one implementation, the interpolation module 404 is specifically configured to:

[0147] Divide the target grid into an upper triangle and a lower triangle;

[0148] Use the vegetation index analysis value corresponding to the first grid point included in the upper triangle to fit the triangular plane coefficient and intercept corresponding to the upper triangle; and, use the vegetation index analysis value corresponding to the first grid point included in the lower triangle to fit the triangular plane coefficient and intercept corresponding to the lower triangle;

[0149] Determine whether the second grid point is inside the upper triangle or inside the lower triangle, and based on the triangular plane coefficient and intercept corresponding to the target triangle where the second grid point is located, determine the vegetation index interpolation corresponding to the second grid point.

[0150] In one implementation, the fast Fourier transform module 406 is specifically configured to:

[0151] Use the fast Fourier transform to convert the high-resolution vegetation index product from the spatial domain to the frequency domain to obtain the vegetation index frequency spectrum diagram;

[0152] Move the low-frequency components in the vegetation index spectrogram from the edge position of the vegetation index spectrogram to the center position of the vegetation index spectrogram to obtain a new vegetation index spectrogram;

[0153] Filter out the low-frequency components from the new vegetation index spectrogram through a filter to remove the stitching areas included in the high-resolution vegetation index products;

[0154] Use inverse frequency-domain conversion to convert the vegetation index spectrogram with low-frequency components filtered out from the frequency domain to the spatial domain to obtain a new high-resolution vegetation index product.

[0155] In one implementation manner, the fusion module 408 is specifically configured to:

[0156] Perform normalization processing on the texture features corresponding to the new high-resolution vegetation index product;

[0157] Use the normalized texture features to adjust the vegetation index interpolation so that the change situation between the adjusted vegetation index interpolations of two adjacent pixels meets the vegetation index change situation described by the texture features to obtain a vegetation index fusion result.

[0158] The device provided by the embodiments of the present invention has the same implementation principle and the same technical effects as those of the foregoing method embodiments. For the sake of brief description, for the parts not mentioned in the device embodiments, reference may be made to the corresponding contents in the foregoing method embodiments.

[0159] The embodiments of the present invention provide an electronic device. Specifically, the electronic device includes a processor and a storage device; a computer program is stored on the storage device, and the computer program, when run by the processor, executes the method according to any one of the foregoing implementation manners.

[0160] Figure 5 FIG. is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. The electronic device 100 includes: a processor 50, a memory 51, a bus 52, and a communication interface 53. The processor 50, the communication interface 53, and the memory 51 are connected through the bus 52; the processor 50 is used to execute an executable module stored in the memory 51, such as a computer program.

[0161] Among them, the memory 51 may include a high-speed random access memory (RAM, Random Access Memory), and may also include a non-volatile memory, such as at least one disk memory. Through at least one communication interface 53 (which may be wired or wireless), a communication connection is realized between the system network element and at least one other network element, and the Internet, a wide area network, a local area network, a metropolitan area network, etc. can be used.

[0162] The bus 52 can be an ISA bus, a PCI bus, an EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of convenience of representation, Figure 5 only a bidirectional arrow is used in Figure 5 , but it does not mean that there is only one bus or one type of bus.

[0163] Among them, the memory 51 is used to store a program. After receiving an execution instruction, the processor 50 executes the program. The method executed by the device defined by the flow process disclosed in any embodiment of the foregoing embodiments of the present invention can be applied to the processor 50 or implemented by the processor 50.

[0164] The processor 50 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the integrated logic circuit in the hardware of the processor 50 or by the instructions in software form. The above-mentioned processor 50 may be a general-purpose processor, including a central processing unit (CPU for short), a network processor (NP for short), etc.; it may also be a digital signal processor (DSP for short), an application specific integrated circuit (ASIC for short), a field-programmable gate array (FPGA for short), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present invention can be directly embodied as being executed and completed by a hardware decoding processor, or can be executed and completed by a combination of hardware and software modules in the decoding processor. The software module may be located in a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, and other well-known storage media in the art. This storage media is located in the memory 51, and the processor 50 reads the information in the memory 51 and combines its hardware to complete the steps of the above method.

[0165] The computer program product of the readable storage medium provided by the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the method described in the foregoing method embodiments. For the specific implementation, reference can be made to the foregoing method embodiments, which will not be elaborated here.

[0166] If the above-described functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this 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 for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs.

[0167] Finally, it should be noted that the above-described embodiments are only specific implementation manners of the present invention, used to illustrate the technical solutions of the present invention, rather than limiting it. The protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed by the present invention can still modify the technical solutions described in the foregoing embodiments, or can easily conceive of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes, or replacements 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, and should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform, characterized in that Including: Obtaining a meteorological vegetation index product and a high-resolution vegetation index product; Using the specified vegetation index product as a background field, performing multi-level interpolation processing on the meteorological vegetation index product based on optimal interpolation and plane equation interpolation to obtain a vegetation index interpolation; Performing fast Fourier transform processing on the high-resolution vegetation index product to remove the seam regions included in the high-resolution vegetation index product, obtaining a new high-resolution vegetation index product; Fusing the vegetation index interpolation with the texture features corresponding to the new high-resolution vegetation index product, where the texture features are used to describe the change in vegetation index between two adjacent pixels in the new high-resolution vegetation index product, to obtain a vegetation index fusion result.

2. The satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform according to claim 1, wherein Using the specified vegetation index product as a background field, performing multi-level interpolation processing on the meteorological vegetation index product based on optimal interpolation and plane equation interpolation to obtain a vegetation index interpolation, including: For any first grid point in the first grid to be interpolated, determining the vegetation index observation values corresponding to multiple observation points adjacent to the first grid point from the meteorological vegetation index product, and determining the vegetation index background values corresponding to each of the observation points from the specified vegetation index product, and performing optimal interpolation processing on the first grid point based on the vegetation index observation values and the vegetation index background values to obtain the vegetation index analysis value corresponding to the first grid point; For any second grid point in the second grid to be interpolated, determining the target grid where the second grid point is located in the first grid, and performing plane equation interpolation processing on the second grid point based on the vegetation index analysis value corresponding to the first grid point included in the target grid to obtain the vegetation index interpolation corresponding to the second grid point; Wherein, the resolution of the second grid is higher than the resolution of the first grid.

3. The satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform according to claim 2, characterized in that Performing optimal interpolation processing on the first grid point based on the vegetation index observation values and the vegetation index background values to obtain the vegetation index analysis value corresponding to the first grid point, including: Taking the minimum error correlation coefficient corresponding to the first grid point as the target, assigning target weight coefficients to each of the observation points; Determining the vegetation index difference between the vegetation index observation value and the vegetation index background value corresponding to the same observation point, and performing weighted fusion on all the vegetation index differences based on the target weight coefficients to obtain a vegetation index correction value; Using the vegetation index correction value to correct the vegetation index background value corresponding to the first grid point to obtain the vegetation index analysis value corresponding to the first grid point.

4. The satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform according to claim 3, characterized in that Taking the minimum error correlation coefficient corresponding to the first grid point as the target, assigning target weight coefficients to each of the observation points, including: Denoting any two observation points adjacent to the first grid point as a first observation point and a second observation point respectively; Determining a first error correlation coefficient based on the vegetation index observation value corresponding to the first observation point and the vegetation index observation value corresponding to the second observation point; and determining a second error correlation coefficient based on the vegetation index background value corresponding to the first observation point and the vegetation index background value corresponding to the second observation point; Determine a target error correlation coefficient between the first observation point and the second observation point based on the first error correlation coefficient, the second error correlation coefficient, a ratio of error standard deviations corresponding to the first observation point, and a ratio of error standard deviations corresponding to the second observation point; Perform weighted fusion on the target error correlation coefficient using a current weight coefficient to obtain a grid error correlation coefficient corresponding to the first grid point. Adjust the current weight coefficient, and continue to perform weighted fusion on the target error correlation coefficient using the new current weight coefficient until the obtained grid error correlation coefficient is minimized to determine a target weight coefficient.

5. The satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform according to claim 2, characterized in that Based on the vegetation index analysis value corresponding to the first grid point included in the target grid, perform plane equation interpolation processing on the second grid point to obtain a vegetation index interpolation corresponding to the second grid point, including: Divide the target grid into an upper triangle and a lower triangle; Use the vegetation index analysis value corresponding to the first grid point included in the upper triangle to fit the triangular plane coefficients and intercept corresponding to the upper triangle; and use the vegetation index analysis value corresponding to the first grid point included in the lower triangle to fit the triangular plane coefficients and intercept corresponding to the lower triangle; Determine whether the second grid point is inside the upper triangle or the lower triangle, and based on the triangular plane coefficients and the intercept corresponding to the target triangle where the second grid point is located, determine the vegetation index interpolation corresponding to the second grid point.

6. The satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform according to claim 1, characterized in that Perform fast Fourier transform processing on the high-resolution vegetation index product to remove the stitching area included in the high-resolution vegetation index product, obtaining a new high-resolution vegetation index product, including: Use fast Fourier transform to transform the high-resolution vegetation index product from the spatial domain to the frequency domain to obtain a vegetation index frequency spectrum diagram; Move the low-frequency components in the vegetation index frequency spectrum diagram from the edge position of the vegetation index frequency spectrum diagram to the center position of the vegetation index frequency spectrum diagram to obtain a new vegetation index frequency spectrum diagram; Filter out the low-frequency components from the new vegetation index frequency spectrum diagram through a filter to remove the stitching area included in the high-resolution vegetation index product; Use inverse frequency domain conversion to transform the vegetation index frequency spectrum diagram with the low-frequency components filtered out from the frequency domain to the spatial domain to obtain a new high-resolution vegetation index product.

7. The satellite vegetation index fusion method based on optimal interpolation and fast Fourier transform according to claim 1, wherein Fuse the vegetation index interpolation with the texture features corresponding to the new high-resolution vegetation index product to obtain a vegetation index fusion result, including: Perform normalization processing on the texture features corresponding to the new high-resolution vegetation index product; Adjust the vegetation index interpolation using the normalized texture features so that the change situation between the adjusted vegetation index interpolations of two adjacent pixels meets the vegetation index change situation described by the texture features, obtaining a vegetation index fusion result.

8. A satellite vegetation index fusion device based on optimal interpolation and fast Fourier transform, characterized in that Including: A product acquisition module for acquiring a meteorological vegetation index product and a high-resolution vegetation index product; An interpolation module, configured to use a specified vegetation index product as a background field, and perform multi-level interpolation processing on the meteorological vegetation index product based on optimal interpolation and plane equation interpolation to obtain a vegetation index interpolation; A fast Fourier transform module, configured to perform fast Fourier transform processing on the high-resolution vegetation index product to eliminate the seam regions included in the high-resolution vegetation index product and obtain a new high-resolution vegetation index product; A fusion module, configured to fuse the vegetation index interpolation with the texture features corresponding to the new high-resolution vegetation index product, where the texture features are used to describe the change in the vegetation index between two adjacent pixels in the new high-resolution vegetation index product, to obtain a vegetation index fusion result.

9. An electronic device, characterized in that, It includes a processor and a memory. The memory stores computer-executable instructions that can be executed by the processor, and the processor executes the computer-executable instructions to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions cause the processor to implement the method according to any one of claims 1 to 7.

Images