An abundance decomposition method, system, device, and storage medium

Abstract

This invention provides an abundance decomposition method, system, device, and storage medium applicable to remote sensing images. The abundance decomposition method includes: iteratively clustering the remote sensing image using a spatial adjustment factor and a spectral feature weight adjustment factor to obtain multiple blocks; for each block: calculating the data field strength and potential energy of each pixel; selecting the pixel with the highest potential energy as the representative pixel and obtaining the spectral angles of the remaining pixels based on its spectral response; obtaining the spectral angle judgment result based on a preset angle threshold, and processing it to obtain the component proportion of the pixel. This invention, through the energy field principle, realizes a limited correlation between the data features of the spectrum and the spatial distribution features of the spectrum, making up for the shortcomings of traditional linear spectral mixing models that cannot reasonably utilize spatial feature information; at the same time, compared with traditional methods, it utilizes the energy field principle to reduce the complexity of the algorithm in the abundance decomposition process, improve the efficiency of unmixing calculation, and also improve the decomposition accuracy.

Description

Technical Field

[0001] This invention relates to the field of remote sensing image processing, and in particular to an abundance decomposition method, system, device, and storage medium. Background Technology

[0002] In the field of remote sensing image processing, hybrid pixel decomposition is an important application foundation for research on vegetation cover analysis, atmospheric and environmental monitoring, geological exploration, marine remote sensing, land use and change monitoring, water quality monitoring and crop monitoring based on remote sensing observations. The speed and accuracy of hybrid pixel abundance decomposition directly affect the accuracy of classification and recognition of remote sensing images.

[0003] With the continuous development of application demands in social production processes, a large number of mixed pixel abundance decomposition methods have emerged in the field of remote sensing image processing. Among them, the most widely used are pure pixel index algorithms, N-FINDR algorithms, fixed-point component analysis methods, and iterative error analysis methods. Although these endmember extraction algorithms cover automatic and semi-automatic processing strategies, most of them extract image endmembers from the perspective of spectral data features and do not incorporate spatial information in the pixel abundance decomposition process. However, the spectral information and spatial distribution characteristics of remote sensing images are strongly correlated. If spatial information can be reasonably integrated in the abundance decomposition process, the decomposition accuracy of mixed pixels will be further improved.

[0004] In summary, existing abundance decomposition methods suffer from problems such as low decomposition accuracy. Summary of the Invention

[0005] In view of the shortcomings of the prior art, the purpose of this invention is to provide an abundance decomposition method, system, device and storage medium to improve the technical problems of low decomposition accuracy in the existing abundance decomposition methods.

[0006] To achieve the above and other related objectives, the present invention provides an abundance decomposition method applicable to remote sensing images, wherein the abundance decomposition method includes:

[0007] The remote sensing image is iteratively clustered using spatial adjustment factors and spectral feature weight adjustment factors to obtain multiple blocks.

[0008] For each block:

[0009] The data field strength of each pixel is calculated, and the potential energy of each pixel is obtained by processing the data field strength.

[0010] The pixel with the largest potential energy is taken as the representative pixel, and the spectral angle of the remaining pixels is obtained by processing the spectral response of the representative pixel.

[0011] The spectral angle is determined based on a preset angle threshold, and the component proportion of the pixel is obtained by processing the result of the determination.

[0012] In one embodiment of the present invention, the step of iteratively clustering the remote sensing image using a spatial adjustment factor and a spectral feature weight adjustment factor to obtain multiple blocks includes:

[0013] The remote sensing image is divided using a grid of a preset size;

[0014] For each grid of the remote sensing image:

[0015] The cluster centers of the grid are obtained through processing;

[0016] The center response of each pixel within the grid is calculated using the following formula, and the spectrum at the lowest point of the center response is taken as the center spectrum:

[0017]

[0018] Where: d k I is the center response; I is the differential response of the current pixel; (i,j) is the coordinate of the current pixel; Δi is the neighborhood coordinate offset of the current pixel at coordinate i; Δj is the neighborhood coordinate offset of the current pixel at coordinate j.

[0019] A circle is drawn with the cluster center as the center and a radius of twice the length of the preset size to obtain a range circle;

[0020] The distance between each pixel within the range circle and the cluster center is calculated using the following formula:

[0021] z = η c z c +η s z s

[0022] Where: Z is the distance; Z c Z is a spectral distance measure between the current pixel's spectrum and the center spectrum. s η is the spatial distance metric between the current pixel and the cluster center. c η is the spectral feature weighting adjustment factor; s The spatial adjustment factor;

[0023] All pixels whose distance is less than a preset clustering threshold are taken as target pixels;

[0024] The final block is obtained by iterative clustering based on the target pixel.

[0025] In one embodiment of the present invention, the step of iteratively clustering the target pixels to obtain the final block includes:

[0026] Based on all the target pixels within the current range circle, a geometric centroid and a spectral expectation value are obtained, and the spectral expectation value is used as the central spectrum of the geometric centroid.

[0027] Using the geometric center as the center and a circle with a radius twice the length of the preset dimension as the radius, the next range circle is obtained;

[0028] Determine whether the error between the center spectrum of the next range circle and the current range circle meets a preset iteration error.

[0029] If so, then all target pixels within the specified range circle are clustered to obtain the block;

[0030] If not, then continue to obtain the subsequent range circle.

[0031] In one embodiment of the present invention, the error is calculated using the following formula:

[0032]

[0033] Where ε is the error; λ'0 is the center spectrum of the next range circle; λ0 is the center spectrum of the current range circle; and L represents a specific band.

[0034] In one embodiment of the present invention, the steps of calculating the data field strength of each pixel and processing the data field strength to obtain the potential energy of each pixel include:

[0035] The field strength of all pixels is obtained by processing the data using the following formula:

[0036]

[0037] Where: P represents the data field strength of a pixel; l represents a specific band; γ represents the external radiation factor; B(i,j) represents the data radiance at coordinate (i,j); and r represents the data radiation radius of a pixel.

[0038] The potential energy of each pixel is obtained by superimposing and accumulating the field strength of all specific wavebands.

[0039] In one embodiment of the present invention, the step of taking the pixel with the largest potential energy as the representative pixel and processing the spectral response of the representative pixel to obtain the spectral angles of the remaining pixels includes:

[0040] The pixel with the highest potential energy is selected as the representative pixel;

[0041] Based on the representative pixel, the spectral angles of other pixels are obtained using the following formula:

[0042]

[0043] Where, d SAD The spectral angle is... λ represents the spectral response of the pixel k. std These are endmember spectra from the standard ground feature spectral library.

[0044] In one embodiment of the present invention, the step of determining the spectral angle based on a preset angle threshold and processing the result to obtain the component proportion of the pixel includes:

[0045] Determine whether the spectral angle reaches the angle threshold;

[0046] Pixels with spectral angles smaller than the stated angle threshold are considered as mixed pixels;

[0047] The component proportions of all the mixed pixels are obtained using the following formula:

[0048]

[0049] Where: λ represents the spectral reflectance of the current mixed pixel; K is the component proportion of the current mixed pixel; λ STD ε is the standard spectral reflectance of the current mixed pixel; ε is the error; and M is the number of basic components contained in the current mixed pixel.

[0050] The present invention also discloses an abundance decomposition system, wherein the abundance decomposition system using the above-described abundance decomposition method comprises:

[0051] The block partitioning module is used to iteratively cluster remote sensing images using spatial adjustment factors and spectral feature weight adjustment factors to obtain multiple blocks.

[0052] The potential energy calculation module is used to calculate the data field strength of all pixels in each block, and process the data field strength to obtain the potential energy of all pixels.

[0053] The spectral angle calculation module is used to take the pixel with the largest potential energy in each block as the representative pixel, and process the spectral response of the representative pixel to obtain the spectral angle of the remaining pixels.

[0054] The abundance decomposition module is used to determine the spectral angle in each block according to a preset angle threshold, and process the result to obtain the component proportion of the pixel.

[0055] The present invention also discloses an abundance decomposition device, including a processor coupled to a memory, the memory storing program instructions, which, when executed by the processor, implement the above-described abundance decomposition method.

[0056] The present invention also discloses a computer-readable storage medium, characterized in that it includes a program that, when run on a computer, causes the computer to perform the above-described abundance decomposition method.

[0057] In summary, the abundance decomposition method, system, device, and storage medium provided by this invention have the following beneficial effects:

[0058] The abundance decomposition method, system, device, and storage medium provided by this invention realizes a limited correlation between the data features and spatial distribution features of the spectrum through the principle of energy field, making up for the shortcomings of traditional linear spectral mixing models that cannot make reasonable use of spatial feature information. At the same time, compared with traditional methods, it reduces the complexity of the algorithm in the abundance decomposition process by utilizing the principle of energy field, improves the efficiency of demixing calculation, and also improves the accuracy of decomposition. Attached Figure Description

[0059] To more clearly illustrate the technical solutions in the embodiments of the present 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 only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0060] Figure 1 The diagram shown is a system flowchart of the abundance decomposition method of the present invention.

[0061] Figure 2 The flowchart shown is for step S10 of the present invention;

[0062] Figure 3 The flowchart shown is for step S40 of the present invention;

[0063] Figure 4 The diagram shown is a schematic diagram of the principle structure of the abundance decomposition system of the present invention.

[0064] Figure 5 The diagram shown is a schematic representation of the principle structure of the abundance decomposition device of the present invention.

[0065] Component designation explanation

[0066] Abundance decomposition system 100:

[0067] Block partitioning module 110;

[0068] Potential energy calculation module 120;

[0069] Spectral angle calculation module 130;

[0070] Abundance decomposition module 140;

[0071] Abundance decomposition equipment 200;

[0072] Processor 210;

[0073] Memory 220. Detailed Implementation

[0074] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features can be combined with each other. It should also be understood that the terminology used in the embodiments of the present invention is for describing specific implementation schemes and not for limiting the scope of protection of the present invention. Test methods in the following embodiments that do not specify specific conditions are generally performed under conventional conditions or according to the conditions recommended by the respective manufacturers.

[0075] Please see Figures 1 to 4 It should be understood that the structures, proportions, sizes, etc., depicted in the accompanying drawings are merely for illustrative purposes to aid those skilled in the art and to facilitate understanding. They are not intended to limit the scope of the invention and therefore have no substantial technical significance. Any modifications to the structure, changes in proportions, or adjustments to size, without affecting the effectiveness and purpose of the invention, should still fall within the scope of the technical content disclosed herein. Furthermore, the terms such as "upper," "lower," "left," "right," "middle," and "one" used in this specification are merely for clarity and not intended to limit the scope of the invention. Changes or adjustments to their relative relationships, without substantially altering the technical content, should also be considered within the scope of the invention.

[0076] When numerical ranges are given in the embodiments, it should be understood that, unless otherwise stated in the present invention, both endpoints of each numerical range and any value between the two endpoints may be selected. Unless otherwise defined, all technical and scientific terms used in this invention, as well as the prior art known to those skilled in the art and the description of this invention, may be implemented using any prior art methods, devices, and materials similar to or equivalent to those described, used, or made of materials in the embodiments of this invention.

[0077] Abundance, or the abundance of an element in a natural body, refers to the relative proportion of a chemical element in the total amount of that natural body. Abundance decomposition refers to the method of extracting various land cover components (endmembers) and the proportion of each component (i.e., abundance) from the actual spectral data of mixed pixels. Among them, endmember extraction and abundance estimation are two important processes in the abundance decomposition of mixed pixels.

[0078] Please see Figure 1 This embodiment provides an abundance decomposition method, which includes:

[0079] Step S10: Iteratively cluster the remote sensing image using spatial adjustment factor and spectral feature weight adjustment factor to obtain multiple blocks;

[0080] Remote sensing images refer to films or photographs that record the magnitude of electromagnetic waves emitted by various ground features. They are mainly divided into aerial photographs and satellite photographs. Due to their macroscopic, objective, periodic, and convenient characteristics, remote sensing images have played a significant role in monitoring land cover, forest cover, wetland cover, and grassland cover.

[0081] Specifically:

[0082] Step S11: Divide the remote sensing image into a grid of preset size N×N;

[0083] An S×S grid is constructed over the entire remote sensing image. In a preferred embodiment, S can be equal to 7.

[0084] Furthermore, steps S12-S17 are performed for each grid of the remote sensing image:

[0085] Step S12: Process to obtain the cluster centers of the grid;

[0086] In this step, common techniques are used to obtain the cluster centers of the grid, such as the k-means clustering algorithm.

[0087] Step S13: Calculate the center response of each pixel within the grid using the following formula, and take the spectrum at the lowest center response as the center spectrum:

[0088]

[0089] Where: d k I is the center response; I is the differential response of the current cell; (i,j) is the coordinate of the current cell; Δi is the neighborhood coordinate offset of the current cell at coordinate i; Δj is the neighborhood coordinate offset of the current cell at coordinate j.

[0090] The center pixel extracted within the grid is not necessarily located at the geometric center of the grid. To ensure that the spectral information of the center pixel is more representative, it is preferable to extract the center response d of the center pixel relative to the neighboring pixels in an 8×8 block. k .

[0091] Step S14: Draw a circle with the cluster center as the center and a radius of twice the length of the preset size to obtain a range circle;

[0092] To ensure smoother segmentation after clustering among different grids, the same type of clustering is performed with the pixels at coordinates i and j as the center and 2S as the search range.

[0093] Step S15: Calculate the distance between each pixel within the range circle and the cluster center using the following formula:

[0094] z = η c z c +η s z s

[0095] Where: z is the distance; z c z is a spectral distance measure between the current pixel's spectrum and the center spectrum. s η is the spatial distance metric between the current pixel and the cluster center. c η is the spectral feature weighting adjustment factor; s Spatial adjustment factor;

[0096] Specifically:

[0097] z c It can be represented by the Euclidean distance between the spectra of pixels under a specific band L, and the formula is as follows:

[0098]

[0099] Where, λ l (i,j) represents the spectral response of the l-th band at coordinate position (i,j). The center spectrum of the l-th band of the cluster center.

[0100] z s This can be represented by the actual coordinates of each point and the Euclidean distance between the cluster centers, using the following formula:

[0101]

[0102] Where x is the x-coordinate of the current cell; y is the y-coordinate of the current cell; x0 is the x-coordinate of the cluster center; and y0 is the y-coordinate of the cluster center.

[0103] To adjust the relative importance of spectral feature similarity and pixel spatial similarity in the overall similarity measure, two normalized weights are introduced here: η s As a spatial adjustment factor, to eliminate its relationship with specific image dimensions, η can be set... s =1 / S 2 In a preferred embodiment, S = 7. Where S is the number of grid cells; η c This is the spectral feature weighting adjustment factor, which is specifically affected by the spectral and data dimensions. It can be set as a constant based on practical experience; here, we set it to η. c =1 / L 2 In a preferred embodiment, Where L is a specific band in the dimension.

[0104] Step S16: Select all pixels whose distance is less than a preset clustering threshold as target pixels;

[0105] When the distance is less than the clustering threshold, the cell corresponding to that distance is taken as the target cell, that is, it is considered to be a cell that forms a block together with the cluster center.

[0106] In a preferred embodiment, the clustering threshold can be 0.3.

[0107] Step S17: Perform iterative clustering based on the target pixels to obtain the final block;

[0108] Based on all target pixels within the current range circle, a geometric centroid and spectral expectation value are obtained through processing. The spectral expectation value is used as the central spectrum of the geometric centroid.

[0109] Using the geometric center as the center and a radius twice the length of the preset size, draw a circle to obtain the next range circle;

[0110] Determine whether the error between the center spectrum of the next range circle and the current range circle meets a preset iteration error.

[0111] If so, then cluster all target pixels within the range circle to obtain a block;

[0112] If not, then continue to obtain the subsequent range circle.

[0113] Specifically:

[0114] The expected value refers to the sum of the probabilities of each possible outcome in a discrete random variable experiment multiplied by the result. In step S17, by calculating the expected value of the spectra of all target pixels, the central spectrum of the geometric centroid can be obtained more accurately.

[0115] The error is calculated using the following formula:

[0116]

[0117] Where ε is the error; λ'0 is the center spectrum of the next range circle; λ0 is the center spectrum of the current range circle; and L represents a specific band.

[0118] Iteration error T ε This is the classification iteration precision set according to user needs during actual calculations. The larger the value, the faster the pixel clustering speed. When the error ε is less than the iteration error T... ε When the iteration ends, all target pixels within the range circle are clustered to obtain a block.

[0119] Furthermore, for each block, steps S20-S40 are performed:

[0120] Step S20: Calculate the data field strength of each pixel, and process the data field strength to obtain the potential energy of each pixel;

[0121] The field strength of all pixels is obtained by processing the data using the following formula:

[0122]

[0123] Where: p represents the data field strength of a pixel; l represents a specific band; γ represents the external radiation factor; B(ij) represents the data radiance at coordinates (i,j); and r represents the data radiation radius of a pixel.

[0124] The potential energy of each pixel is obtained by superimposing and accumulating the field strength of all specific wavebands.

[0125] Based on the principles of independence and superposition of pixels in a data field, the field strength of a single pixel in the data field across multiple specific bands can be accumulated to obtain the potential energy of the spatial data field for each pixel:

[0126]

[0127] Step S30: Select the pixel with the largest potential energy as the representative pixel, and process the spectral response of the representative pixel to obtain the spectral angle of the remaining pixels.

[0128] The pixel with the highest potential energy is selected as the representative pixel;

[0129] Based on the representative pixel, the spectral angles of other pixels are obtained using the following formula:

[0130]

[0131] Where, d SAD The spectral angle is... λ represents the spectral response of the pixel k. std These are endmember spectra from the standard ground feature spectral library.

[0132] Step S40: Determine the spectral angle according to a preset angle threshold, and process the result to obtain the component ratio of the pixel.

[0133] Step S41: Determine whether the spectral angle has reached the angle threshold;

[0134] Set the angle threshold to T SAD In a preferred embodiment, the angle threshold is T. SAD It is 0.1.

[0135] Step S42: Pixels with spectral angles smaller than the angle threshold are taken as mixed pixels;

[0136] Mixed pixels serve as end-member points for the actual ground cover spectral correspondences of paired ground cover species.

[0137] Step S43: The component proportions of the mixed pixels are obtained using the following formula:

[0138]

[0139] Where: λ represents the spectral reflectance of the current mixed pixel; K is the component proportion of the current mixed pixel; λ STD ε is the standard spectral reflectance of the current mixed pixel; ε is the error of the center spectrum; and M is the number of basic components contained in the current mixed pixel.

[0140] Furthermore, the minimum error ε can be calculated using the least squares method, thereby calculating the component proportions of all mixed pixels.

[0141] Please see Figure 4 This embodiment also provides an abundance decomposition system 100. Using the above-described abundance decomposition method, the abundance decomposition system includes:

[0142] The block partitioning module 110 is used to iteratively cluster the remote sensing image using spatial adjustment factors and spectral feature weight adjustment factors to obtain multiple blocks.

[0143] The potential energy calculation module 120 is used to calculate the data field strength of all pixels in each block and process the data field strength to obtain the potential energy of all pixels.

[0144] The spectral angle calculation module 130 is used to take the pixel with the largest potential energy in each block as the representative pixel, and process the spectral response of the representative pixel to obtain the spectral angle of the remaining pixels.

[0145] The abundance decomposition module 140 is used to determine the spectral angle in each block according to a preset angle threshold, and process the result of the determination to obtain the component proportion of the pixel.

[0146] Please see Figure 5 This embodiment also proposes an abundance decomposition device 200, which includes a processor 210 and a memory 220. The processor 210 and the memory 220 are coupled, and the memory 220 stores program instructions. When the program instructions stored in the memory 220 are executed by the processor 210, the above-mentioned abundance decomposition method is implemented. The processor 210 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The memory 220 may include random access memory (RAM) and may also include non-volatile memory, such as at least one disk storage device. The memory 220 can also be an internal memory of the Random Access Memory (RAM) type. The processor 210 and the memory 220 can be integrated into one or more independent circuits or hardware, such as Application Specific Integrated Circuits (ASICs). It should be noted that the computer program in the memory 220 can be implemented as a software functional unit and, when sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, 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, electronic device, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention.

[0147] This embodiment also proposes a computer-readable storage medium storing computer instructions for instructing a computer to execute the task management method described above. The storage medium can be an electronic medium, magnetic medium, optical medium, electromagnetic medium, infrared medium, or a semiconductor system or propagation medium. The storage medium may also include semiconductor or solid-state memory, magnetic tape, removable computer disk, random access memory (RAM), read-only memory (ROM), hard disk, and optical disc. Optical discs may include optical disc-read-only memory (CD-ROM), optical disc-read / write (CD-RW), and DVD.

[0148] The abundance decomposition method, system, device, and storage medium provided by this invention realizes a limited correlation between the data features and spatial distribution features of the spectrum through the principle of energy field, making up for the shortcomings of traditional linear spectral mixing models that cannot make reasonable use of spatial feature information. At the same time, compared with traditional methods, it reduces the complexity of the algorithm in the abundance decomposition process by utilizing the principle of energy field, improves the efficiency of demixing calculation, and also improves the accuracy of decomposition.

[0149] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by those skilled in the art without departing from the spirit and technical concept disclosed in the present invention should still be covered by the claims of the present invention.

Claims

1. An abundance decomposition method, characterized in that, For remote sensing images, the abundance decomposition method includes: The remote sensing image is iteratively clustered using a spatial adjustment factor and a spectral feature weight adjustment factor to obtain multiple blocks, including: dividing the remote sensing image using a grid of a preset size; For each grid of the remote sensing image: The cluster centers of the grid are obtained through processing; The center response of each pixel within the grid is calculated using the following formula, and the spectrum at the lowest point of the center response is taken as the center spectrum: in: I represents the center response; I represents the differential response of the current pixel. The coordinates of the current cell; This represents the neighborhood coordinate offset of the current pixel at coordinate i. This represents the neighborhood coordinate offset of the current pixel at coordinate j. A circle is drawn with the cluster center as the center and a radius of twice the length of the preset size to obtain a range circle; The distance between each pixel within the range circle and the cluster center is calculated using the following formula: in: The distance is mentioned; The spectral distance between the current pixel's spectrum and the center spectrum is measured. This is a measure of the spatial distance between the current pixel and the cluster center. The spectral feature weighting adjustment factor; The spatial adjustment factor; All pixels whose distance is less than a preset clustering threshold are taken as target pixels; Iterative clustering is performed based on the target pixels to obtain the final blocks, including: Based on all the target pixels within the current range circle, a geometric centroid and a spectral expectation value are obtained, and the spectral expectation value is used as the central spectrum of the geometric centroid. Using the geometric center as the center and a circle with a radius twice the length of the preset dimension as the radius, the next range circle is obtained; Determine whether the error between the center spectrum of the next range circle and the current range circle satisfies a preset iteration error. If so, then all target pixels within the specified range circle are clustered to obtain the block; If not, then continue to obtain the subsequent range circles; For each block: The data field strength of each pixel is calculated, and the potential energy of each pixel is obtained by processing the data field strength. The pixel with the largest potential energy is taken as the representative pixel, and the spectral angle of the remaining pixels is obtained by processing the spectral response of the representative pixel. The spectral angle is determined based on a preset angle threshold, and the component proportion of the pixel is obtained by processing the result of the determination.

2. The abundance decomposition method according to claim 1, characterized in that, The error is calculated using the following formula: in, The error is described above; The central spectrum of the next range circle; The center spectrum of the current range circle; L represents a specific band.

3. The abundance decomposition method according to claim 1, characterized in that, The steps of calculating the data field strength of each pixel and processing the data field strength to obtain the potential energy of each pixel include: The field strength of all pixels is obtained by processing the data using the following formula: Where: P represents the data field strength of a pixel; l represents a specific wavelength band; γ represents the external radiation factor; The value represents the data radiance at coordinates (i,j); r represents the data radiance radius of the pixel. The potential energy of each pixel is obtained by superimposing and accumulating the field strength of all specific wavebands.

4. The abundance decomposition method according to claim 1, characterized in that, The step of taking the pixel with the largest potential energy as the representative pixel and processing the spectral response of the representative pixel to obtain the spectral angles of the remaining pixels includes: The pixel with the highest potential energy is selected as the representative pixel; Based on the representative pixel, the spectral angles of other pixels are obtained using the following formula: in, The spectral angle is... This represents the spectral response of the pixel k. These are endmember spectra from the standard ground feature spectral library.

5. The abundance decomposition method according to claim 2, characterized in that, The step of determining the spectral angle based on a preset angle threshold and processing the result to obtain the component proportion of the pixel includes: Determine whether the spectral angle reaches the angle threshold; Pixels with spectral angles smaller than the stated angle threshold are considered as mixed pixels; The component proportions of all the mixed pixels are obtained using the following formula: in: This represents the spectral reflectance of the current mixed pixel; This represents the component percentage of the current mixed pixels; The standard spectral reflectance of the current mixed pixel; The error is M, where M is the number of basic components contained in the current mixed pixel.

6. An abundance decomposition system, characterized in that, If the abundance decomposition method according to any one of claims 1-5 is used, then the abundance decomposition system includes: The block partitioning module is used to iteratively cluster the remote sensing image using a spatial adjustment factor and a spectral feature weight adjustment factor to obtain multiple blocks, including: dividing the remote sensing image using a grid of a preset size; For each grid of the remote sensing image: The cluster centers of the grid are obtained through processing; The center response of each pixel within the grid is calculated using the following formula, and the spectrum at the lowest point of the center response is taken as the center spectrum: in: I represents the center response; I represents the differential response of the current pixel. The coordinates of the current cell; This represents the neighborhood coordinate offset of the current pixel at coordinate i. This represents the neighborhood coordinate offset of the current pixel at coordinate j. A circle is drawn with the cluster center as the center and a radius of twice the length of the preset size to obtain a range circle; The distance between each pixel within the range circle and the cluster center is calculated using the following formula: in: The distance is mentioned; The spectral distance between the current pixel's spectrum and the center spectrum is measured. This is a measure of the spatial distance between the current pixel and the cluster center. The spectral feature weighting adjustment factor; The spatial adjustment factor; All pixels whose distance is less than a preset clustering threshold are taken as target pixels; Iterative clustering is performed based on the target pixels to obtain the final blocks, including: Based on all the target pixels within the current range circle, a geometric centroid and a spectral expectation value are obtained, and the spectral expectation value is used as the central spectrum of the geometric centroid. Using the geometric center as the center and a circle with a radius twice the length of the preset dimension as the radius, the next range circle is obtained; Determine whether the error between the center spectrum of the next range circle and the current range circle satisfies a preset iteration error. If so, then all target pixels within the specified range circle are clustered to obtain the block; If not, then continue to obtain the subsequent range circles; The potential energy calculation module is used to calculate the data field strength of all pixels in each block, and process the data field strength to obtain the potential energy of all pixels. The spectral angle calculation module is used to take the pixel with the largest potential energy in each block as the representative pixel, and process the spectral response of the representative pixel to obtain the spectral angle of the remaining pixels. The abundance decomposition module is used to determine the spectral angle in each block according to a preset angle threshold, and process the result to obtain the component proportion of the pixel.

7. An abundance decomposition device, characterized in that, The method includes a processor coupled to a memory, the memory storing program instructions that, when executed by the processor, implement the abundance decomposition method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, Includes a program that, when run on a computer, causes the computer to perform the abundance decomposition method as described in any one of claims 1 to 5.

Images