Aod inversion method and system based on multiple vegetation indices

Abstract

The application discloses an AOD inversion method and system based on multiple vegetation indexes, and the method comprises the following steps: finding a dark pixel based on multiple vegetation indexes on the basis of satellite remote sensing image data after processing; calculating the average apparent reflectivity of red and blue bands by counting the number of dark pixels in a preset range and calculating the average apparent reflectivity of red and blue bands; obtaining the surface reflectivity of red and blue bands and calculating the ratio of the two according to the observation geometry data of satellite remote sensing and a lookup table established in advance; continuing AOD inversion when the difference between the ratio and a preset ratio is greater than a preset threshold; obtaining the surface reflectivity of red and blue bands and the ratio smaller than or equal to the preset threshold requirement through multiple iterations, and calculating AOD according to the extinction coefficient of different aerosol types. The AOD inversion method and system provided by the application use the dark pixel determined by multiple vegetation indexes, realize accurate determination of the dark pixel, and further improve the accuracy of AOD inversion.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aerosol optical depth, and particularly relates to an AOD inversion method and system based on multiple vegetation indices. BACKGROUND

[0002] Aerosol is a solid or liquid particle suspended in gas, usually with a diameter of 0.01-10 μm, a dispersed phase widely existing in air and formed by a sol with gas as a dispersion medium, having colloidal properties. Aerosol has the characteristics of light scattering, electrophoresis and Brownian motion. The dispersed phase does Brownian motion in the atmosphere and does not settle due to gravity, so that it can be suspended in the atmosphere for several months or even years. Many phenomena of the atmosphere are related to aerosol, such as atmospheric visibility, cloud formation, ozone photochemical reaction, radiation transmission, etc. Therefore, in the research of environmental and climate change, atmospheric correction of space-to-ground remote sensing and many other research processes, the influence of aerosol cannot be ignored, and the research of aerosol has important significance.

[0003] Aerosol Optical Depth (AOD) refers to the optical thickness of aerosol components due to extinction, which can also be called atmospheric turbidity, and is used to describe the degree of attenuation of light by aerosol. Since AOD can reflect the influence of aerosol on radiation balance and has a high correlation with particulate matter concentration, the measurement of AOD is also of great significance.

[0004] At present, the inversion methods of AOD used by humans mainly include two types: ground-based observation inversion and satellite remote sensing inversion. In the aspect of ground-based observation inversion, the use of a sun photometer is a common and reliable measurement method. The sun photometer can directly obtain the data of ground-based observation of solar radiation flux, and can also obtain the spectral distribution and scattering phase of aerosol particles by observing the radiation in different directions of the sky. Common inversion algorithms include: spectral extinction method (Langley method), bandwidth extinction method, Femald method, etc.

[0005] Satellite remote sensing inversion is a method of detecting aerosol optical characteristics by using satellite remote sensors carried by satellites with the development of human aerospace technology, and using the constantly improved and perfected inversion algorithm to invert AOD. Based on the AOD inversion process of satellite remote sensing, the regression model based on empirical coefficients is constructed by means of observation data based on the empirical regression method, and the aerosol optical thickness can obtain satisfactory results under the premise of rich and high-precision observation data, but the precision and range are subject to the difficulty of obtaining observation data. The method based on physical model is mainly based on the radiation transfer model, and the look-up table is constructed to determine the aerosol optical thickness at the corresponding channel through the solar zenith angle, satellite zenith angle and relative azimuth angle. Compared with ground-based observation, it has many advantages such as low cost, wide area, fast response and so on. Common inversion algorithms include: dark pixel method, structure function method, land-sea contrast method, atmospheric transmittance method, reflectance angle distribution method and polarization method.

[0006] The dark pixel method is usually used for AOD inversion. However, the dark pixel method usually has better AOD inversion results only in the area with high vegetation coverage. The inversion accuracy of the existing dark pixel method depends on the accurate determination or measurement of the reflectivity of the dark pixel. The dark pixel method proposed by Kaufman is based on the normalized vegetation index to determine the dark pixel. However, the normalized vegetation index is greatly affected by the background of the plant canopy (such as soil, rotten earth, snow, dry leaves, etc.). When used as a basis for determining the dark pixel, the background radiation often causes inaccurate determination of the dark pixel, thereby affecting the accuracy and results of AOD inversion. Therefore, the normalized vegetation index for determining the dark pixel is only applicable to low reflectivity pixels, and requires high-precision prior knowledge of known surface reflectivity. It is sensitive to the difference between pixels, especially when there are many error introduction factors in determining the dark pixel, which ultimately leads to inaccurate dark pixels, and essentially affects the inversion accuracy of AOD, especially when the vegetation coverage is low or there are many high-surface reflectivity pixels. SUMMARY

[0007] Therefore, the present application provides an AOD inversion method and system based on multiple vegetation indices, which adds multiple vegetation indices as a basis for determining the dark pixel, can accurately determine the dark pixel, and improves the accuracy of AOD inversion to solve the technical problems in the above background.

[0008] To achieve the above purpose, the present application provides the following technical solutions:

[0009] In a first aspect, the present application provides an AOD inversion method based on multiple vegetation indices, comprising:

[0010] Finding the dark pixel based on multiple vegetation indices after processing the satellite remote sensing image data;

[0011] counting the number of dark pixels in the preset range and calculating the average apparent reflectance of the red and blue bands to obtain the average apparent reflectance of the red and blue bands;

[0012] According to the observation geometry data of satellite remote sensing and the pre-established lookup table, the surface reflectance of the red and blue bands is obtained and the ratio of the two is calculated; when the difference between the ratio and the preset ratio is greater than the preset threshold, the AOD inversion is continued; through multiple iterations until the surface reflectance of the red and blue bands and the ratio that are less than or equal to the preset threshold requirement are obtained, and the AOD is calculated according to the extinction coefficients of different aerosol types.

[0013] Preferably, the plurality of vegetation indices include: normalized difference vegetation index, ratio vegetation index, greenness vegetation index and vertical vegetation index; the process of finding dark pixels based on the plurality of vegetation indices includes:

[0014] The dark pixels are preliminarily screened based on the normalized difference vegetation index of the processed whole scene satellite remote sensing image to obtain a high vegetation coverage area with a normalized difference vegetation index greater than a first preset value, which is recorded as a first area;

[0015] The first area is filtered using the ratio vegetation index to obtain a green and healthy vegetation coverage area with a ratio vegetation index greater than a second preset value, which is recorded as a second area;

[0016] The greenness vegetation index is used to obtain a two-bit map composed of a first component and a second component, and the vegetation coverage area with characteristic differences between vegetation and soil spectra is extracted from the two-bit map, which is recorded as a third area; and the vertical vegetation index is used to obtain a vegetation coverage area with soil background eliminated, which is recorded as a fourth area; the third area and the fourth area are subjected to difference mode calculation to obtain the final dark pixels.

[0017] Preferably, the satellite remote sensing image data is high-resolution satellite image data.

[0018] Preferably, the proportion of the final dark pixels in the whole scene satellite remote sensing image meets a preset proportion range.

[0019] Preferably, the number of dark pixels in the preset range is obtained by mathematical statistics, and the average apparent reflectance of the red and blue bands is calculated to obtain the average apparent reflectance of the red and blue bands.

[0020] Preferably, the observation geometry data includes: solar zenith angle, satellite zenith angle and relative azimuth angle.

[0021] Preferably, the process of inverting to obtain the AOD includes: increasing in a fixed preset step size based on a preset initial value to obtain the corresponding AOD.

[0022] In a second aspect, the embodiment of the present application provides an AOD inversion system based on multiple vegetation indices, comprising:

[0023] The dark pixel searching module is configured to search for dark pixels based on multiple vegetation indices after processing of the satellite remote sensing image data.

[0024] The average apparent reflectance obtaining module is configured to count the number of dark pixels in a preset range and calculate the average apparent reflectance of red and blue bands to obtain the average apparent reflectance of red and blue bands.

[0025] The AOD inversion module is configured to obtain the surface reflectance of red and blue bands and calculate the ratio of the two according to the observation geometry data of satellite remote sensing and a lookup table established in advance, continue the AOD inversion when the difference between the ratio and a preset ratio is greater than a preset threshold, and obtain the surface reflectance of red and blue bands and the ratio that meet the requirement of being less than or equal to the preset threshold through multiple iterations, and calculate the AOD according to the extinction coefficients of different aerosol types.

[0026] In a third aspect, the embodiment of the present application provides a computer device, comprising at least one processor and a memory connected with the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the AOD inversion method based on multiple vegetation indices according to the first aspect of the present application.

[0027] In a fourth aspect, the embodiment of the present application provides a computer readable storage medium, which stores computer instructions for enabling a computer to execute the AOD inversion method based on multiple vegetation indices according to the first aspect of the present application.

[0028] The technical scheme of the present application has the following advantages:

[0029] The application provides an AOD inversion method and system based on multiple vegetation indexes, which comprises the following steps: finding dark pixels based on multiple vegetation indexes after processing satellite remote sensing image data; counting the number of dark pixels in a preset range and calculating the average apparent reflectivity of red and blue bands to obtain the average apparent reflectivity of red and blue bands; obtaining the surface reflectivity of red and blue bands and calculating the ratio of the two according to the observation geometry data of satellite remote sensing and a lookup table established in advance; continuing AOD inversion when the difference between the ratio and a preset ratio is greater than a preset threshold; obtaining the surface reflectivity of red and blue bands and the ratio by multiple iterations until the surface reflectivity of red and blue bands and the ratio meet the requirement of being less than or equal to the preset threshold, and calculating AOD according to the extinction coefficient of different aerosol types. The AOD inversion method and system provided by the application use dark pixels determined by multiple vegetation indexes to realize accurate determination of dark pixels, can more accurately exclude the interference of background radiation on dark pixel recognition, improve the quality of dark pixels, and further improve the accuracy of AOD inversion. BRIEF DESCRIPTION OF DRAWINGS

[0030] In order to more clearly illustrate the specific embodiments of the application or the technical solutions in the prior art, the drawings needed in the description of the specific embodiments or the prior art will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0031] Figure 1 The flowchart of the AOD inversion method based on multiple vegetation indexes provided in the embodiments of the application;

[0032] Figure 2 The flowchart of finding dark pixels based on multiple vegetation indexes provided in the embodiments of the application;

[0033] Figure 3 The flowchart of the AOD inversion method based on multiple vegetation indexes provided in the embodiments of the application;

[0034] Figure 4 The module composition diagram of the AOD inversion system based on multiple vegetation indexes provided in the embodiments of the application;

[0035] Figure 5 The composition diagram of one specific example of the computer device provided in the embodiments of the application. DETAILED DESCRIPTION

[0036] In the following, the technical solutions in the embodiments of the present application will be described clearly and completely with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, and not all embodiments. They are not intended to limit the scope of the present application. In addition, in the following description, the description of well-known structures and techniques is omitted to avoid unnecessary confusion of the concepts disclosed in the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work should fall within the scope of the present application.

[0037] In addition, the technical features involved in the different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0038] Embodiment 1

[0039] The embodiments of the present application provide an AOD inversion method based on multiple vegetation indices, as shown in the method, which comprises: Figure 1

[0040] Step S1: Finding dark pixels based on multiple vegetation indices after processing satellite remote sensing image data. It should be noted that theoretically, satellite remote sensing image data containing red, blue and near-red three bands can use the multiple vegetation indices of the embodiments of the present application to find dark pixels and perform AOD inversion. For example, image data of foreign MODIS satellites and Landsat satellites, but it is difficult for end users to directly modify the inversion process parameters based on mature aerosol products synthesized from such satellite remote sensing image data. The embodiments of the present application prefer to use high-resolution satellite image data as the data basis, which has more abundant prior knowledge of related parameters, greatly ensuring the national information security; using domestic high-resolution series satellites launched by China as data source for AOD inversion, compared with using MODIS satellites, it can be completely self-controlled in data, satellite, orbit and technology. In this embodiment, the data processing process of high-resolution satellite image data is a conventional processing flow in the art, which is not specifically limited here.

[0041] ​In the embodiment, the multiple vegetation indices include a normalized difference vegetation index (NDVI), a relative vegetation index (RVI), a green vegetation index (GVI), and a perpendicular vegetation index (PVI). Specifically, the four vegetation indices differ in sensitivity. The NDVI has low sensitivity in a high vegetation coverage area and is sensitive to soil, humus, snow, and other vegetation coverage areas, and has a large background radiation difference. The RVI is a sensitive parameter of green plants and has high correlation with a leaf area index, leaf dry biomass, and chlorophyll content, and can be used to detect and estimate plant biomass. The GVI is obtained by separating the spectral characteristics of vegetation and soil through K-T transformation, and the two components constitute a two-dimensional diagram that can well reflect the difference between the spectral characteristics of vegetation and soil. The GVI is a weighted sum of the radiation brightness values of each wave band, and the radiation brightness is a comprehensive result of atmospheric radiation, solar radiation, and environmental radiation, so the GVI is most sensitive to changes in external conditions. The PVI better eliminates the influence of the soil background and has smaller sensitivity to the atmosphere than other vegetation indices.

[0042] In the embodiment, the process of finding dark pixels based on the multiple vegetation indices includes the following steps, as shown in FIG. 2. Figure 2

[0043] The processed whole-scene high-resolution satellite image data is subjected to preliminary screening of dark pixels based on the normalized difference vegetation index (NDVI), to obtain a high-vegetation-coverage area with NDVI greater than a first preset value, denoted as a first area; and areas that do not meet the preliminary screening condition are determined as non-dark pixels. Specifically, the first preset value is 0.6.

[0044] The first area is subjected to plant biomass sensitivity filtering using the relative vegetation index (RVI), to obtain a green and healthy vegetation coverage area with RVI greater than a second preset value, denoted as a second area; and areas that do not meet the filtering condition (low RVI sensitivity) are determined as non-dark pixels. Specifically, the second preset value is 2. The RVI sensitivity analysis excludes severely diseased and dead vegetation, and further filters dark pixels that pass the preliminary screening and have poor filtering conditions, to obtain fine-screened dark pixels.

[0045] ​The green vegetation index GVI is used on the second region to obtain a two-bit image composed of a first component and a second component, and a vegetation coverage region with characteristic differences between vegetation and soil spectra is extracted from the two-bit image and is denoted as a third region; and the perpendicular vegetation index PVI is used to obtain a vegetation coverage region with soil background eliminated and is denoted as a fourth region; and a difference mode calculation is performed on the third region and the fourth region to obtain the final dark pixel. Specifically, the GVI is highly sensitive to changes in the external environment, and thus can exclude pixels polluted by the external environment.

[0046] It should be noted that the process of using the NDVI, RVI, PVI and GVI vegetation indexes to find the dark pixel in the embodiments of the present application has a sequence requirement. First, the NDVI is used to preliminarily screen the region of the dark pixel, because the NDVI has the advantages of fast calculation speed and good generalization, and compared with other vegetation indexes, the NDVI is more widely used, is more suitable for the existing computer framework and software, and can quickly screen out non-sensitive regions. Then, the RVI is used for filtering and screening, which greatly reduces the calculation process of the RVI and thus reduces the overall calculation amount. If the order of the two kinds of vegetation indexes is interchanged, too many non-vegetation pixels will participate in the operation when the RVI is run, which has little effect on the processing of the whole scene image with high coverage. However, for an image containing high-vegetation-coverage regions, low-vegetation-coverage regions or even no-vegetation-coverage regions, errors are easily introduced and the calculation amount is increased. The order of the PVI and the GVI is not important when they are used to participate in the calculation and screening of the dark pixel, but the two kinds of vegetation indexes should be used after the NDVI and the RVI.

[0047] In the embodiments, the numerical values of the related parameters involved in the process of finding the dark pixel based on multiple vegetation indexes are set according to prior knowledge and are empirical values, and the present application is not limited thereto, and the numerical values can be adaptively adjusted according to actual application requirements.

[0048] In the embodiments, the proportion of the final dark pixel in the whole scene satellite remote sensing image meets a preset proportion range, and the preset proportion range is set according to the requirements of AOD inversion, for example, the preset proportion range is 5%-10%. It should be noted that if the proportion range is set too low, it is not conducive to the inversion calculation of the AOD, and if the proportion range is set too high, the obtained dark pixel is not pure enough. Only as an example, the proportion range is determined according to actual requirements.

[0049] Step S2: The number of dark pixels in the preset range is counted, and the average apparent reflectivity of the red and blue bands is calculated to obtain the average apparent reflectivity of the red and blue bands.

[0050] In the embodiments, the number of dark pixels in the preset range is counted by mathematical statistics, and the average apparent reflectivity of the red and blue bands is calculated to obtain the average apparent reflectivity of the red and blue bands. Specifically, the preset range is 20x20, which is only as an example and is not limited thereto.

[0051] It should be noted that the dark pixel determined by the plurality of vegetation indices in the embodiment of the present application can not only accurately exclude the interference caused by background radiation, but also find the dark pixel least affected by the surrounding adjacent pixels according to the average apparent reflectance of the red and blue bands (i.e. the average value of the apparent reflectance of the dark pixel band in the preset range), thereby reducing the error between pixels, maximizing the accuracy of the dark pixel, and avoiding the system error introduced due to the inaccuracy of the apparent reflectance.

[0052] In addition, in the AOD inversion process using high-resolution series satellites, such as the GF-4 satellite, since it only provides observation geometry data of the image center, using only the observation geometry data of the center pixel will cause a large error in the radiation transfer calculation. By statistically averaging the apparent reflectance of the red and blue bands of the dark pixels in the preset range, the radiation transfer calculation error can be reduced, and the AOD inversion accuracy can be improved. In the embodiment, the observation geometry data includes the solar zenith angle, the satellite zenith angle, and the relative azimuth angle. Specifically, the GF-4 satellite has a high orbit and a large imaging range (about 400 kilometers in width), and the angle difference between different pixels of the observation geometry data of the image center includes an angle difference of 8 ° in the zenith angle and an angle difference of 10 ° in the azimuth angle, which are only used as examples for illustration and can be adjusted adaptively according to actual application scenarios.

[0053] Step S3: Obtain the surface reflectance of the red and blue bands and calculate the ratio of the two according to the observation geometry data of the satellite remote sensing and the pre-established lookup table; continue the AOD inversion when the difference between the ratio and the preset ratio is greater than the preset threshold; obtain the red and blue band surface reflectance and the ratio by multiple iterations until the requirement that the red and blue band surface reflectance and the ratio are less than or equal to the preset threshold is met, and calculate the AOD according to the extinction coefficient of different aerosol types.

[0054] In the embodiment, the process of inverting the AOD includes increasing the preset initial value by a fixed preset step to obtain the corresponding AOD.

[0055] It should be noted that the preset initial value is the initial value of the AOD to be solved, and the setting basis is the weather condition at that time. For example, the initial value of the AOD to be solved is set to 0 in sunny weather, and the initial value of the AOD to be solved is set to any non-zero value in other weather conditions, which is not limited specifically herein and can be determined according to actual application requirements.

[0056] In the embodiment, it is assumed that the initial value of the AOD to be solved is 0, the surface reflectance of the red and blue bands is obtained according to the observation geometry data of the high-resolution satellite and a lookup table established in advance, and the ratio of the two is calculated; when the difference is greater than a preset threshold, the initial value of the AOD to be solved is updated in the form of fixed preset step 0.01 increase; the surface reflectance of the red and blue bands corresponding to the updated AOD to be solved is obtained by searching the lookup table again, and the ratio of the two is calculated; the difference between the ratio and the preset ratio 0.5 is calculated, and then the preset threshold 0.001 is judged, and the AOD to be solved is updated again; through multiple iterations of the above process, when the difference is less than or equal to the preset threshold 0.001, the iteration is stopped, and the AOD to be solved that meets the preset threshold requirement is obtained; and the AOD at 0.55 μm is calculated according to the extinction coefficient of different aerosol types. It should be noted that the AOD at 0.55 μm is a wavelength value recognized by those skilled in the art, and other values involved in the above process are all for illustration only, not limited thereto.

[0057] In a specific embodiment, the GF-4 satellite observation data in the high-resolution satellite image data is used as a data source for AOD inversion, as shown in FIG. 1, the process includes: Figure 3

[0058] Step 1: A variety of vegetation indices are used for dark pixel determination on the processed GF-4 image data.

[0059] Specifically, according to the GF-4 cloud mask data, the image data of the blue, red and near-infrared three bands of GF-4 is masked, which can clearly identify the cloud-polluted data without damaging the original image data, and after removing the contaminated data, the processed GF-4 image data is obtained, which improves the data quality.

[0060] Step 2: The number of dark pixels in a 20x20 preset range is counted, and the average apparent reflectance of the red and blue bands is calculated.

[0061] Step 3: It is assumed that the initial value of the AOD to be solved is 0, the surface reflectance of the red and blue bands is obtained by searching the lookup table according to the observation geometry data (solar zenith angle, satellite zenith angle and relative azimuth angle) of GF-4, and the ratio of the red band surface reflectance and the blue band surface reflectance is calculated; the ratio is compared to determine whether it meets the condition, when it does not meet the condition (i.e. the difference between the ratio and the preset ratio is greater than the preset threshold), the initial value of the AOD to be solved is updated by increasing the step size to 0.01; multiple iterations of the process are performed, when the condition is met (i.e. the difference between the ratio and the preset ratio is less than or equal to the preset threshold), the corresponding surface reflectance of the red and blue bands and the ratio (i.e. the AOD to be solved that meets the preset threshold requirement) are obtained.

[0062] ​Step 4: Calculate the AOD at 0.55 μm according to the extinction coefficient of different aerosol types.

[0063] In summary, compared with the prior dark pixel determination process using only a single vegetation index, the AOD inversion method based on multiple vegetation indices provided by the embodiments of the present application increases the number of vegetation indices to improve the identification accuracy of dark pixels. In addition, by calculating the average value of the apparent reflectance of the dark pixels in the preset range, not only the inter-pixel error is reduced, but also the radiation transfer calculation error caused by the observation geometry of the center pixel of the high-resolution satellite is reduced, further improving the accuracy of AOD inversion.

[0064] Embodiment 2

[0065] The embodiments of the present application provide an AOD inversion system based on multiple vegetation indices, as shown in Figure 4 The embodiments of the present application provide an AOD inversion system based on multiple vegetation indices, as shown in

[0066] The dark pixel finding module is used to find dark pixels based on multiple vegetation indices after processing satellite remote sensing image data. This module performs the method described in step S1 of embodiment 1, and will not be described here.

[0067] The average apparent reflectance obtaining module is used to calculate the average value of the red and blue band apparent reflectance of the dark pixels in the preset range, and obtain the average apparent reflectance of the red and blue bands. This module performs the method described in step S2 of embodiment 1, and will not be described here.

[0068] The AOD inversion module is used to set the initial value of AOD, obtain the surface reflectance of the red and blue bands according to the observation geometry data of satellite remote sensing and the pre-established lookup table, and calculate the ratio of the two. When the difference between the ratio and the preset ratio is greater than the preset threshold, the AOD is updated. Through multiple iterations, the AOD that meets the requirement of being less than or equal to the preset threshold is obtained. The AOD is calculated according to the extinction coefficient of different aerosol types. This module performs the method described in step S3 of embodiment 1, and will not be described here.

[0069] The AOD inversion system based on multiple vegetation indices provided by the embodiments of the present application realizes accurate determination of dark pixels, can more accurately exclude the interference of background radiation on dark pixel identification, and improves the quality of dark pixels. At the same time, the average apparent reflectance is used to reduce the radiation transfer calculation error caused by the observation geometry of the center pixel of the high-resolution satellite, and improve the accuracy of AOD inversion.

[0070] Embodiment 3

[0071] The embodiments of the present application provide a computer device, as shown in Figure 5As shown, it comprises at least one processor 501, at least one communication interface 503, memory 504 and at least one communication bus 502. Among them, the communication bus 502 is used to realize the connection communication between the components, the communication interface 503 can include a display screen and a keyboard, and the optional communication interface 503 can also include a standard wired interface, a wireless interface. The memory 504 can be a high-speed volatile random access memory, or a non-volatile memory, or at least one storage device located away from the aforementioned processor 501. Among them, the processor 501 can execute the AOD inversion method based on multiple vegetation indices in embodiment 1. The memory 504 stores a set of program codes, and the processor 501 calls the program codes stored in the memory 504 for executing the AOD inversion method based on multiple vegetation indices in embodiment 1.

[0072] Among them, the communication bus 502 can be a peripheral component interconnect (Peripheral Component Interconnect, abbreviated as PCI) bus or an extended industry standard architecture (Extended Industry Standard Architecture, abbreviated as EISA) bus, etc. The communication bus 502 can be divided into an address bus, a data bus, a control bus, etc. For convenience of representation, Figure 5 In the figure, only one line is used to represent, but it does not mean that there is only one bus or one type of bus.

[0073] Among them, the memory 504 can include a volatile memory (Volatile Memory), such as a random access memory (Random Access Memory, abbreviated as RAM); the memory can also include a non-volatile memory (Non-volatile Memory), such as a flash memory (Flash Memory), a hard disk (Hard Disk Drive, abbreviated as HDD) or a solid-state disk (Solid-state Drive, abbreviated as SSD); the memory 504 can also include a combination of the above types of memories.

[0074] Among them, the processor 501 can be a central processing unit (Central Processing Unit, abbreviated as CPU), a network processor (Network Processor, abbreviated as NP) or a combination of CPU and NP.

[0075] The processor 501 can further include a hardware chip. The hardware chip can be an application-specific integrated circuit (ASIC), a programmable logic device (PLD) or a combination thereof. The PLD can be a complex programmable logic device (CPLD), a field programmable gate array (FPGA), a generic array logic (GAL) or any combination thereof.

[0076] Optionally, the memory 504 is further configured to store program instructions. The processor 501 can invoke the program instructions to implement the AOD retrieval method based on multiple vegetation indices in the embodiment of the present application.

[0077] The embodiment of the present application further provides a computer readable storage medium, and the computer readable storage medium stores computer executable instructions. The computer executable instructions can execute the AOD retrieval method based on multiple vegetation indices in the embodiment 1. The storage medium can be a disk, an optical disk, a read only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD) or a solid state drive (SSD), etc. The storage medium can also include a combination of the above-mentioned storage devices.

[0078] Obviously, the above embodiment is only an example for clear illustration, and does not limit the embodiments. Based on the above description, other different forms of changes or variations can be made by those skilled in the art. Here, all the embodiments do not need to be exhausted, and the obvious changes or variations still fall within the protection scope of the present application.

Claims

1. AOD inversion method based on multiple vegetation indices, characterized in that, The method comprises the following steps: finding dark pixels based on multiple vegetation indices after processing satellite remote sensing image data; counting the number of dark pixels in a preset range and calculating the average apparent reflectance of red and blue bands to obtain the average apparent reflectance of red and blue bands; obtaining the surface reflectance of red and blue bands and calculating the ratio of the two according to the observation geometry data of satellite remote sensing and a pre-established lookup table; when the difference between the ratio and a preset ratio is greater than a preset threshold, AOD inversion is continued; through multiple iterations, the surface reflectance of red and blue bands and the ratio that meet the requirement of being less than or equal to the preset threshold are obtained, and AOD is calculated according to the extinction coefficients of different aerosol types; wherein the multiple vegetation indices include normalized difference vegetation index, ratio vegetation index, greenness vegetation index and vertical vegetation index; the process of finding dark pixels based on multiple vegetation indices comprises the following steps: performing initial screening of dark pixels based on normalized difference vegetation index on the processed whole-scene satellite remote sensing image to obtain a high-vegetation-covering area with normalized difference vegetation index greater than a first preset value, which is recorded as a first area; performing plant biomass sensitivity filtering on the first area using ratio vegetation index to obtain a green and healthy vegetation covering area with ratio vegetation index greater than a second preset value, which is recorded as a second area; obtaining a two-bit map composed of a first component and a second component using greenness vegetation index, extracting the two-bit map to obtain a vegetation covering area with characteristic difference between vegetation and soil spectrum, which is recorded as a third area; and obtaining a vegetation covering area with soil background eliminated using vertical vegetation index, which is recorded as a fourth area; performing difference mode calculation on the third area and the fourth area to obtain the final dark pixels.

2. The method according to claim 1, wherein, The satellite remote sensing image data is high-resolution satellite image data.

3. The method according to claim 1, wherein, The proportion of the final dark pixels in the whole-scene satellite remote sensing image meets a preset proportion range.

4. The method according to claim 1, wherein, The number of dark pixels in a preset range is obtained through mathematical statistics, and the average apparent reflectance of red and blue bands is calculated by calculating the average apparent reflectance of red and blue bands.

5. The method according to claim 1, wherein, The observation geometry data includes solar zenith angle, satellite zenith angle and relative azimuth angle.

6. The method according to claim 1, wherein, The process of obtaining AOD by inversion comprises the following steps: increasing in a fixed preset step length based on a preset initial value to obtain the corresponding AOD.

7. An AOD inversion system based on multiple vegetation indices, characterized in that, The method comprises the following steps: a dark pixel finding module for finding dark pixels based on multiple vegetation indices after processing satellite remote sensing image data; an average apparent reflectance obtaining module for counting the number of dark pixels in a preset range and calculating the average apparent reflectance of red and blue bands to obtain the average apparent reflectance of red and blue bands; an AOD inversion module for obtaining the surface reflectance of red and blue bands and calculating the ratio of the two according to the observation geometry data of satellite remote sensing and a pre-established lookup table; when the difference between the ratio and a preset ratio is greater than a preset threshold, AOD inversion is continued; through multiple iterations, the surface reflectance of red and blue bands and the ratio that meet the requirement of being less than or equal to the preset threshold are obtained, and AOD is calculated according to the extinction coefficients of different aerosol types; The multiple vegetation indices include a normalized difference vegetation index, a ratio vegetation index, a greenness vegetation index, and a vertical vegetation index. The process of finding the dark pixel based on the multiple vegetation indices comprises: Performing dark pixel preliminary screening on the processed whole scene satellite remote sensing image based on a normalized difference vegetation index to obtain a high vegetation coverage area with a normalized difference vegetation index greater than a first preset value, denoted as a first area; Performing plant biomass sensitivity filtering on the first area using a ratio vegetation index to obtain a green and healthy vegetation coverage area with a ratio vegetation index greater than a second preset value, denoted as a second area; 8. A computer device, comprising: Using a greenness vegetation index to obtain a two-bit map composed of a first component and a second component, extracting the two-bit map to obtain a vegetation coverage area with characteristic differences between vegetation and soil spectrum, denoted as a third area, and using a vertical vegetation index to obtain a vegetation coverage area with soil background eliminated, denoted as a fourth area; and performing difference mode calculation on the third area and the fourth area to obtain the final dark pixel. Comprise:

9. A computer-readable storage medium, characterized in that, At least one processor and a memory connected in communication with the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to cause the at least one processor to perform the AOD inversion method based on multiple vegetation indices according to any one of claims 1-6. The computer readable storage medium stores computer instructions for causing the computer to perform the AOD inversion method based on multiple vegetation indices according to any one of claims 1-6.

Images