Method and system for estimating plant diversity based on remote sensing scale effect

Abstract

The invention discloses a method and a system for estimating plant diversity based on a remote sensing scale effect, which belong to the field of plant diversity monitoring, and are characterized in that a high-resolution remote sensing image of a to-be-monitored area is acquired, and the pixel size is approximately equal to the size of a single plant; extracting a plurality of investigation quadrats in the monitoring area, and obtaining plant diversity true values of the quadrats; preprocessing the remote sensing image to obtain surface reflectance data, and upscaling the high-resolution image by adopting a pixel averaging method to obtain a low-resolution remote sensing image; calculating a nonlinear vegetation index based on the remote sensing images, and calculating a scale deviation of the vegetation index between the remote sensing images with two resolutions; and constructing a mathematical relationship model between the scale deviation and the plant diversity truth value, and applying the mathematical relationship model to the whole remote sensing image so as to estimate the plant diversity of the whole area. According to the method, the plant diversity in a large range can be quickly and accurately estimated, heavy ground survey work is reduced, and manpower, time and economic cost are saved.

Description

Technical Field

[0001] The present invention relates to the field of plant diversity monitoring, and in particular to a method and system for estimating plant diversity based on remote sensing scale effect. Background Art

[0002] Plant diversity plays a vital role in maintaining ecosystem stability. However, we are currently facing a global biodiversity loss. China, despite its rich biodiversity resources, is also one of the countries experiencing the most severe biodiversity loss. Rapid and accurate monitoring of plant diversity is a question worth exploring.

[0003] Traditional plant diversity estimates rely on ground surveys, which involve laying out sample plots on the ground and counting the number of species within them, thereby estimating the plant diversity of a small area. However, ground surveys are labor-intensive and time-consuming, and usually require the full participation of professional personnel (such as botanists), so this method is only applicable to small areas. However, biodiversity loss is a global problem, and we need to estimate plant diversity at global, national, or at least regional scales. Therefore, there is an urgent need to develop a method for rapid estimation of plant diversity across large areas.

[0004] Remote sensing technology can provide large-scale, information-rich imagery, and estimating plant diversity based on this technology offers great potential. Although several methods have been proposed, they are often limited by ecosystem type, sensor type, and study scale. The classic spectral variability hypothesis has been tested across a wide range of ecosystem types, sensor types, and study scales. However, some studies have shown that spectral diversity is not always positively correlated with biodiversity. Therefore, given the urgency of global biodiversity decline, existing methods are insufficient to fully meet the needs of remote sensing monitoring of biodiversity, and the development of new approaches is crucial. Summary of the Invention

[0005] In response to the problems existing in the existing technology, the present invention provides a method and system for estimating plant diversity based on remote sensing scale effect, which can achieve large-scale, rapid and accurate estimation of plant diversity, and has low economic cost and strong operability.

[0006] To achieve the above technical objectives, the present invention discloses a method for estimating plant diversity based on remote sensing scale effect, comprising the following steps:

[0007] Step 1: Obtain high-resolution remote sensing images of the area to be monitored based on the ecosystem type of the monitoring area, and ensure that the pixel size of the high-resolution remote sensing image is close to the size of a single plant in the monitoring area; when the monitoring area is a grassland ecosystem, the resolution of the high-resolution remote sensing image is set to 10 cm; when the monitoring area is a shrub ecosystem, the resolution of the high-resolution remote sensing image is set to 50 cm; when the monitoring area is a forest ecosystem, the resolution of the high-resolution remote sensing image is set to 1 meter;

[0008] Step 2: Divide the entire monitoring area into multiple large blocks, and then randomly select multiple plant diversity quadrats of the same size from each large block, and obtain and mark the true value of plant diversity and coordinate information in each selected quadrat;

[0009] Step 3: Preprocess the high-resolution remote sensing image, including radiometric calibration, atmospheric correction, and geometric correction, to obtain surface reflectance data of the monitoring area; use the pixel averaging method in spatial resampling to upscale the high-resolution remote sensing image to the sample size to obtain a low-resolution remote sensing image;

[0010] Step 4: Calculate the vegetation index based on the high-resolution remote sensing image and the vegetation index based on the low-resolution remote sensing image respectively, and calculate the remote sensing scale deviation of the vegetation index between the high-resolution remote sensing image and the low-resolution remote sensing image;

[0011] Step 5: Use the remote sensing scale deviation to fit the true value of plant diversity in all selected samples to obtain a mathematical relationship model between the remote sensing scale deviation and the true value of plant diversity. Apply this mathematical relationship model to the remote sensing image of the entire monitored area to estimate the plant diversity of the entire monitored area.

[0012] Furthermore, when the monitoring area is a grassland ecosystem, the sample size is at least 1x1 m; when the monitoring area is a shrub ecosystem, the sample size is at least 5x5 m; when the monitoring area is a forest ecosystem, the sample size is at least 10x10 m.

[0013] Furthermore, the specific process of upscaling high-resolution remote sensing images using the pixel averaging method in spatial resampling is as follows:

[0014] The following formula is used to average all pixel values ​​corresponding to the high-resolution image to obtain the pixel value of the low-resolution image;

[0015]

[0016] Among them, x i is the pixel value of the high-resolution remote sensing image, is the pixel value of the upscaled low-resolution remote sensing image. A pixel of a low-resolution remote sensing image is composed of n pixels of a high-resolution remote sensing image.

[0017] Furthermore, when calculating the vegetation index in step 4, the vegetation index should be a nonlinear vegetation index, and the calculation formula of the vegetation index should include the near-infrared band;

[0018] Vegetation indices include:

[0019] Green ratio vegetation index (GRVI):

[0020]

[0021] Green atmospherically resistant vegetation index (GARI):

[0022]

[0023] Green Normalized Difference Vegetation Index (GNDVI):

[0024]

[0025] Normalized Difference Vegetation Index (NDVI):

[0026]

[0027] Among them, NIR is the reflectivity of the near-infrared band, Green is the reflectivity of the green light band, Red is the reflectivity of the red light band, and Blue is the reflectivity of the blue light band.

[0028] Furthermore, the remote sensing scale deviation method for calculating vegetation index between high-resolution remote sensing images and low-resolution remote sensing images is as follows:

[0029]

[0030] Among them, F(x) is the scale deviation of vegetation index between remote sensing images with different spatial resolutions, and f is used to represent the calculation formula of nonlinear vegetation index. is the vegetation index of a single pixel corresponding to the plant diversity survey quadrat on the low-resolution remote sensing image. is the average value of vegetation index of multiple pixels corresponding to the plant diversity survey quadrat on the high-resolution remote sensing image, f(x i ) is the vegetation index of the i-th pixel corresponding to the plant diversity survey sample on the high-resolution remote sensing image. A sample includes n high-resolution pixels.

[0031] Furthermore, the mathematical model between scale deviation and the true value of plant diversity is:

[0032] The Taylor formula can be used to expand the calculation formula f(x) of the nonlinear vegetation index at a certain point x0, where x represents the pixel value and R n is the remainder:

[0033]

[0034] Therefore, any nonlinear vegetation index can be expressed as: f(x) = ax 2 +bx+c;

[0035] The average value of vegetation index of multiple pixels corresponding to the plant diversity survey sample on high-resolution remote sensing images Expressed as:

[0036]

[0037] On low-resolution remote sensing images, the vegetation index f(x) of a single pixel corresponding to a plant diversity survey quadrat is expressed as:

[0038] The scale deviation of vegetation index between remote sensing images with different spatial resolutions is obtained:

[0039]

[0040] Where D(X) represents the degree of surface heterogeneity, and a is a constant. There is a significant positive correlation between surface heterogeneity D(X) and the true value of plant diversity. Therefore, D(X) is used to represent plant diversity at the same time.

[0041] The scale deviation and plant diversity satisfy the following mathematical relationship model:

[0042] y=Am+B

[0043] Where y is plant diversity, m is scale deviation, and A and B are constants obtained by fitting the true value of plant diversity in the study area with the scale deviation.

[0044] A system for estimating plant diversity based on remote sensing scale effect, comprising a plant diversity true value calculation module, an image processing module, a scale deviation calculation module, and a model construction and inversion module connected in sequence;

[0045] Plant diversity true value calculation module, used to input plant diversity ground survey results and sample coordinates, and calculate the plant diversity index;

[0046] The image processing module is used to pre-process the input remote sensing image to obtain surface reflectance data, and to upscale the high-resolution remote sensing image to the ground sample size to obtain a low-resolution remote sensing image;

[0047] The scale deviation calculation module is used to calculate the nonlinear vegetation index on remote sensing images of two spatial resolutions and calculate the scale deviation of the vegetation index between the remote sensing images of the two resolutions;

[0048] The model building and inversion module constructs a mathematical model between scale deviation and the true value of plant diversity, and applies the model to the entire remote sensing image to estimate the plant diversity of the entire monitoring area.

[0049] Beneficial effects: The present invention provides a new method for estimating plant diversity using remote sensing technology. Compared with traditional ground survey methods, this method takes advantage of the advantages of remote sensing technology and can quickly and accurately estimate plant diversity in a large area in a short period of time. This method saves a lot of manpower and time costs, and can provide basic data for biodiversity conservation work at the regional scale and even the national scale. Compared with existing plant diversity remote sensing monitoring methods, this method can be applied to a variety of ecosystem types, sensor types and research scales, and has a wide range of applicability. It does not rely on expensive hyperspectral sensors, and multispectral sensors that are inexpensive and easy to operate can also meet the requirements of this method. This greatly reduces the economic cost of remote sensing monitoring of plant diversity and improves operability. Therefore, the present invention can achieve a comprehensive, objective, efficient and dynamic assessment of plant diversity, and provide technical support for my country's biodiversity conservation work. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 Schematic diagram of the process of estimating plant diversity based on remote sensing scale effect in an embodiment of the present invention.

[0051] Figure 2 Schematic diagram of a system for estimating plant diversity based on remote sensing scale effect in an embodiment of the present invention.

[0052] Figure 3 Schematic diagram of scatter plots between species number and remote sensing scale deviation in an embodiment of the present invention.

[0053] Figure 4 Schematic diagram of species number estimation using remote sensing scale effect in an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The embodiments of the present invention are further described below with reference to the accompanying drawings:

[0055] Example 1:

[0056] The present invention provides an embodiment of a method for estimating plant diversity based on remote sensing scale effects. It should be noted that although the present invention shows a logical order in the flow chart, in some cases, the steps shown or described may be performed in an order different from that shown here.

[0057] Figure 1 is a flow chart of a method for estimating plant diversity based on remote sensing scale effect according to an embodiment of the present invention. Figure 1 As shown, the method includes the following steps:

[0058] Step S10: Acquire a high-resolution remote sensing image of the area to be monitored.

[0059] Remote sensing images can be airborne remote sensing images or satellite-borne remote sensing images; they can be multispectral remote sensing images or hyperspectral remote sensing images.

[0060] Remote sensing images must include near-infrared bands, and the pixel size of high-resolution images should be close to the size of a single plant in the monitoring area.

[0061] For example, if you want to estimate plant diversity in a forest area, you can choose airborne or satellite-borne remote sensing images, whose image spatial resolution can be around 1 meter; if you want to estimate plant diversity in a grassland area, you can only choose airborne remote sensing images, whose image spatial resolution is around 0.1 meter.

[0062] This embodiment of the present invention uses the Parrot Sequoia multispectral sensor, produced by Parrot Drones SAS of France, to estimate plant diversity in the Hulunbuir grasslands of China. This sensor features high precision, compact size, lightweight, and ease of operation, making it compatible with a variety of drone types. The sensor includes four multispectral bands: green, red, red-edge, and near-infrared. Because the goal is to estimate plant diversity in grassland areas, the spatial resolution of the imagery was set to 0.1 m. The images were acquired in calm, cloudless weather, close to the time of the plant diversity survey described below.

[0063] Step S20: Arrange a certain number of plant diversity quadrats in the monitoring area, obtain the true value of plant diversity and record the coordinates of the quadrats.

[0064] If the area to be estimated is primarily herbaceous, the sample size should be at least 1x1 m; if the area to be estimated is primarily shrubby, the sample size should be at least 5x5 m; if the area to be estimated is primarily tree-based, the sample size should be at least 10x10 m. In this example, the sample size is 1x1 m, as the grassland is the grassland.

[0065] The number of samples should be determined based on the area to be monitored and should ideally be evenly distributed. The more samples, the better, provided time and labor costs are controlled.

[0066] The sample plot is generally square, and high-precision instruments (such as GPS) should be used to record the coordinates of the four corner points of the sample plot, and the accuracy error should be controlled at the centimeter level.

[0067] Plant diversity includes multiple indicators, and the number of species can be used to indicate plant diversity. The number of species is the number of different plant species in a sample plot.

[0068] Step S30 , pre-processing the remote sensing image to obtain surface reflectance data, and upscaling the high-resolution remote sensing image to the sample size to obtain a low-resolution remote sensing image.

[0069] Image preprocessing includes radiation calibration, atmospheric correction and geometric correction. After preprocessing, the surface reflectance data of the orthophoto can be obtained.

[0070] When upscaling images, the pixel averaging method in spatial resampling should be used, and the pixel size after resampling should be consistent with the size of the plant diversity survey sample.

[0071] The pixel averaging method means that the pixel value of the output low-resolution image is the average value of all pixel values ​​corresponding to the input high-resolution image.

[0072]

[0073] Among them, x i is the pixel value of the high-resolution remote sensing image, x is the pixel value of the low-resolution remote sensing image after upscaling, and a pixel of a low-resolution remote sensing image is composed of n pixels of a high-resolution remote sensing image.

[0074] In this embodiment, the image with a spatial resolution of 0.1 m needs to be resampled to 1 m. The resampling is performed in the professional remote sensing image processing software ENVI, and the resampling method is the pixel averaging method.

[0075] Step S40 , respectively calculating vegetation indices on the remote sensing images at two scales, and calculating the scale deviation of the vegetation indices between the remote sensing images at the two scales.

[0076] When calculating the vegetation index, the vegetation index should be a nonlinear vegetation index and should include the near-infrared band.

[0077] In this embodiment, the vegetation index formula includes:

[0078] Green ratio vegetation index (GRVI):

[0079]

[0080] Green atmospherically resistant vegetation index (GARI):

[0081]

[0082] Green Normalized Difference Vegetation Index (GNDVI):

[0083]

[0084] Normalized Difference Vegetation Index (NDVI):

[0085]

[0086] Among them, NIR is the reflectivity of the near-infrared band, Green is the reflectivity of the green light band, Red is the reflectivity of the red light band, and Blue is the reflectivity of the blue light band.

[0087] When calculating scale deviation, the calculation formula is:

[0088]

[0089] Among them, F(x) is the scale deviation of vegetation index between remote sensing images with different spatial resolutions, f is the calculation formula of nonlinear vegetation index, is the vegetation index of a single pixel corresponding to the plant diversity survey quadrat on the low-resolution remote sensing image. is the average value of vegetation index of multiple pixels corresponding to the plant diversity survey quadrat on the high-resolution remote sensing image, f(x i ) is the vegetation index of the i-th pixel corresponding to the plant diversity survey sample on the high-resolution remote sensing image. One ground sample corresponds to n high-resolution pixels.

[0090] In this embodiment, The GRVI index value of the low-resolution pixel corresponding to the plant diversity survey sample. One ground sample corresponds to one low-resolution pixel. The GRVI is the average value of the high-resolution pixels corresponding to the plant diversity survey quadrat. A ground quadrat corresponds to 10x10 high-resolution pixels.

[0091] Step S50 , constructing a mathematical model between scale deviation and true plant diversity value, and applying the model to the entire remote sensing image, thereby estimating the plant diversity of the entire monitoring area.

[0092] Based on the recorded sample coordinates, the ground survey results of each plant diversity sample are matched one by one with the scale deviation F(x) calculated from the image corresponding to the sample, and a mathematical model is constructed between the two. Theoretical studies have shown that there is a strong correlation between plant diversity and remote sensing scale deviation. The constructed model should pass the significance test and can be operated in professional statistical software such as SPSS. Figure 3 As shown, the mathematical model constructed in this embodiment is:

[0093] N=19.481×scale-effect_GRVI+8.760

[0094] like Figure 4 As shown, where N is the number of species and scale-effect_GRVI is the scale deviation of the GRVI index.

[0095] After calculating the scale deviation of the remote sensing image of the entire monitored area, the constructed model is applied to the entire image to estimate the plant diversity of the entire area.

[0096] In an embodiment of the present invention, remote sensing technology is used to estimate plant diversity, which avoids a large amount of ground survey work and saves manpower and time costs. Since ground surveys require the full participation of people, human subjectivity has a greater impact on the results. The present invention makes full use of remote sensing images to estimate plant diversity more objectively. Compared with ground surveys, this method can estimate a larger range at one time, so it can better provide basic data on plant diversity at a regional scale. The present invention creatively proposes to use the scale effect of remote sensing to estimate plant diversity, and estimates the plant diversity of the entire remote sensing image by constructing a mathematical relationship between scale deviation and plant diversity. The present invention can use vegetation index to calculate scale deviation, and the vegetation index requires fewer bands, thereby reducing dependence on hyperspectral sensors, which greatly reduces the cost of remote sensing monitoring of plant diversity and improves operability. The invention is less affected by ecosystem type, sensor type and research scale, and greatly enriches the remote sensing monitoring method of plant diversity.

[0097] Example 2:

[0098] The embodiment of the present invention provides a system for estimating plant diversity based on remote sensing scale effect, such as Figure 2 As shown, the system includes:

[0099] The plant diversity true value calculation module is used to input the plant diversity ground survey results and the coordinates of the sample plots, and calculate the plant diversity index.

[0100] The image processing module is used to pre-process the input remote sensing image to obtain the surface reflectance data of the orthophoto image, and to upscale the high-resolution image to obtain a low-resolution remote sensing image.

[0101] The pixel size of the input high-resolution image should be close to the size of a single plant. The upscaling method should use the pixel averaging method in spatial resampling. The pixel size of the resampled low-resolution image should be consistent with the size of the plant diversity survey plot.

[0102] The scale deviation calculation module is used to calculate the vegetation index on remote sensing images of two scales respectively, and calculate the scale deviation of the vegetation index between the remote sensing images of the two scales.

[0103] The vegetation index should be a nonlinear vegetation index and should include the near-infrared band.

[0104] The formula for calculating scale deviation is:

[0105]

[0106] Among them, F(x) is the scale deviation of vegetation index between remote sensing images with different spatial resolutions, f is the calculation formula of nonlinear vegetation index, is the vegetation index of a single pixel corresponding to the plant diversity survey quadrat on the low-resolution remote sensing image. is the average value of vegetation index of multiple pixels corresponding to the plant diversity survey quadrat on the high-resolution remote sensing image, f(x i ) is the vegetation index of the i-th pixel corresponding to the plant diversity survey sample on the high-resolution remote sensing image. One ground sample corresponds to n high-resolution pixels.

[0107] The model building and inversion module constructs a mathematical model between scale deviation and the true value of plant diversity, and applies the model to the entire remote sensing image to estimate the plant diversity of the entire monitoring area.

[0108] The system provided in the embodiment of the present invention has the same implementation principles and technical effects as the aforementioned method embodiment. For the sake of brevity, any matters not mentioned in the system embodiment can be referred to the corresponding contents in the aforementioned method embodiment 1. Those skilled in the art will clearly understand that for the convenience and brevity of description, the specific working processes of the aforementioned systems and units can all refer to the corresponding processes in the aforementioned method embodiment and will not be repeated here.

[0109] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be included therein. Any reference sign in a claim should not be construed as limiting the claim to which it relates.

[0110] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any minor modifications, equivalent replacements, and improvements made to the above embodiments based on the technical essence of the present invention shall be included in the scope of protection of the technical solution of the present invention.

Claims

1. A method for estimating plant diversity based on remote sensing scale effect, characterized in that: The following steps are involved: Step 1: Obtain high-resolution remote sensing images of the area to be monitored according to the ecosystem type of the monitored area, and ensure that the pixel size of the high-resolution remote sensing image is close to the size of a single plant in the monitored area; when the monitored area is a grassland ecosystem, the resolution of the high-resolution remote sensing image is set to 10 cm; when the monitored area is a shrub ecosystem, the resolution of the high-resolution remote sensing image is set to 50 cm; when the monitored area is a forest ecosystem, the resolution of the high-resolution remote sensing image is set to 1 m; Step 2: Divide the entire monitoring area into multiple large blocks, and then randomly select multiple plant diversity plots of the same size from each large block, and obtain and mark the true value and coordinate information of the plant diversity in each selected plot; Step 3: Preprocess the high-resolution remote sensing images, including radiation calibration, atmospheric correction and geometric correction, to obtain the surface reflectance data of the monitoring area; The pixel averaging method in spatial resampling is used to upscale the high-resolution remote sensing image to the sample size to obtain a low-resolution remote sensing image. Step 4, respectively calculating the vegetation index based on the high-resolution remote sensing image and the vegetation index based on the low-resolution remote sensing image, and calculating the remote sensing scale deviation of the vegetation index between the high-resolution remote sensing image and the low-resolution remote sensing image; Step 5: Use the remote sensing scale deviation to fit the true value of plant diversity in all selected sample plots to obtain a mathematical relationship model between the remote sensing scale deviation and the true value of plant diversity. Apply this mathematical relationship model to the remote sensing image of the entire monitored area to estimate the plant diversity of the entire monitored area.

2. A method for estimating plant diversity based on remote sensing scale effect according to claim 1, characterized in that; When the monitoring area is a grassland ecosystem, the sample size shall be at least 1x1m; when the monitoring area is a shrub ecosystem, the sample size shall be at least 5x5m; when the monitoring area is a forest ecosystem, the sample size shall be at least 10x10m.

3. The method for estimating plant diversity based on remote sensing scale effect according to claim 1, characterized in that: The specific process of upscaling high-resolution remote sensing images using the pixel averaging method in spatial resampling is as follows: The following formula is used to average all pixel values ​​corresponding to the high-resolution image to obtain the pixel value of the low-resolution image; Among them, x i is the pixel value of the high-resolution remote sensing image, is the pixel value of the upscaled low-resolution remote sensing image. A pixel of a low-resolution remote sensing image is composed of n pixels of high-resolution remote sensing images.

4. The method for estimating plant diversity based on remote sensing scale effect according to claim 1, characterized in that: When calculating the vegetation index in step 4, the vegetation index should be a nonlinear vegetation index, and the calculation formula of the vegetation index should include the near infrared band; Vegetation indices include: Green ratio vegetation index (GRVI): Green atmospherically resistant vegetation index (GARI): Green Normalized Difference Vegetation Index (GNDVI): Normalized Difference Vegetation Index (NDVI): Among them, NIR is the reflectivity of the near-infrared band, Green is the reflectivity of the green light band, Red is the reflectivity of the red light band, and Blue is the reflectivity of the blue light band.

5. The method for estimating plant diversity based on remote sensing scale effect according to claim 4, characterized in that: The method for calculating the remote sensing scale deviation of vegetation index between high-resolution remote sensing images and low-resolution remote sensing images is as follows: Among them, F(x) is the scale deviation of vegetation index between remote sensing images with different spatial resolutions, and f is used to represent the calculation formula of nonlinear vegetation index. is the vegetation index of a single pixel corresponding to the plant diversity survey plot on a low-resolution remote sensing image. is the average vegetation index of multiple pixels corresponding to the plant diversity survey plot on the high-resolution remote sensing image, f(x i ) is the vegetation index of the ith pixel corresponding to the plant diversity survey plot on the high-resolution remote sensing image. A plot includes n high-resolution pixels.

6. The method for estimating plant diversity based on remote sensing scale effect according to claim 5, characterized in that: The mathematical model between scale deviation and true value of plant diversity is: The Taylor formula can be used to expand the calculation formula f(x) of the nonlinear vegetation index at a certain point x0, where x represents the pixel value and R n is the remainder: Therefore, any nonlinear vegetation index is expressed as: f(x) = ax 2 +bx+c; The average value of vegetation index of multiple pixels corresponding to the plant diversity survey plot on high-resolution remote sensing images It is expressed as: Vegetation index of a single pixel corresponding to a plant diversity survey plot on a low-resolution remote sensing image It is expressed as: The scale deviation of vegetation index between remote sensing images with different spatial resolutions is obtained: Among them, D(X) represents the degree of surface heterogeneity, and a is a constant. There is a significant positive correlation between surface heterogeneity D(X) and the true value of plant diversity. Therefore, D(X) is used to represent plant diversity at the same time. The scale deviation and plant diversity satisfy the following mathematical relationship model: y=Am+B Among them, y is plant diversity, m is scale deviation; A and B are constants, which are obtained by fitting the true value of plant diversity in the study area and the scale deviation.

7. A system for implementing the method for estimating plant diversity based on remote sensing scale effect as claimed in claim 1, characterized in that: It includes a plant diversity truth value calculation module, an image processing module, a scale deviation calculation module and a model building and inversion module which are sequentially connected; Plant diversity true value calculation module, used to input plant diversity ground survey results and sample plot coordinates, and calculate plant diversity index; The image processing module is used to pre-process the input remote sensing image to obtain the surface reflectance data, and to upscale the high-resolution remote sensing image to the ground sample size to obtain the low-resolution remote sensing image; The scale deviation calculation module is used to calculate the nonlinear vegetation index on the remote sensing images of two spatial resolutions respectively, and calculate the scale deviation of the vegetation index between the remote sensing images of two resolutions; The model building and inversion module constructs a mathematical model between scale deviation and the true value of plant diversity, and applies the model to the entire remote sensing image to estimate the plant diversity of the entire monitoring area.

Images