Vegetation coverage estimation method based on novel remote sensing index

Abstract

The invention discloses a vegetation coverage estimation method based on a novel remote sensing index, and the method comprises the steps: calculating the simulated reflectivity of a remote sensing image red wave band through employing the reflectivity of a remote sensing image green wave band and a near-infrared wave band; a novel remote sensing index is constructed by using the simulated reflectivity of the red wave band, the reflectivity of the red wave band and the reflectivity of the blue wave band, and the calculation formula is # imgabs0 #: VCI is a novel remote sensing index value, # imgabs1 # is the simulated reflectivity of the red wave band, # imgabs2 # and # imgabs3 # are the reflectivity of the red wave band and the reflectivity of the blue wave band respectively; and establishing a quantitative relationship between the vegetation coverage and the novel remote sensing index, and substituting the novel remote sensing index into the quantitative relationship to obtain the vegetation coverage of the pixels. The method is simple in principle, easy to implement, high in automation degree and wide in application range, has wide application value in the field of vegetation and ecological remote sensing, and is suitable for remote sensing estimation and mapping of the vegetation coverage of a large area and even a global scale.

Description

Technical Field

[0001] The present invention relates to the fields of remote sensing digital image processing and vegetation remote sensing, and particularly to a method for remotely estimating vegetation coverage based on a novel remote sensing index. Background Art

[0002] Vegetation coverage is an important parameter for characterizing surface vegetation cover and has important application values in fields such as ecosystem carbon cycle simulation, soil erosion assessment, and climate change research. Satellite remote sensing has become the main technical means for estimating vegetation coverage due to its capabilities of large-scale, multi-temporal, and low-cost data acquisition. However, existing methods for remotely estimating vegetation coverage are difficult to simultaneously possess high accuracy, broad universality, and easy operability, which restricts their application potential in large-scale dynamic monitoring of vegetation coverage.

[0003] Currently, a series of methods for remotely estimating vegetation coverage have been developed, which can be roughly divided into three categories: (1) empirical regression method, which estimates by establishing an empirical regression model between remote sensing features and ground-measured vegetation coverage; (2) machine learning method, which uses machine learning algorithms such as random forest to construct a non-linear mapping relationship between remote sensing reflectance data and vegetation coverage; (3) mixed pixel decomposition method, which estimates vegetation coverage based on a linear spectral mixture model by calculating the spectral signal contribution ratios of vegetation and soil in mixed pixels. The empirical regression method is only applicable to estimating vegetation coverage within a specific range and for specific vegetation types and does not have universality. The machine learning method highly depends on the quality and representativeness of training samples, and high-quality samples usually require a large amount of ground-measured data and high-resolution remote sensing images, with expensive data acquisition costs and complex processing procedures, resulting in the lack of simplicity in operation for this method. The main drawback of the mixed pixel decomposition method is that it is difficult to accurately determine the spectral characteristics of vegetation endmembers and soil endmembers pixel by pixel at the regional scale, which will greatly increase the uncertainty of the remotely estimated vegetation coverage results. Summary of the Invention

[0004] Aiming at the problem that existing methods for remotely estimating vegetation coverage are difficult to simultaneously possess high accuracy, broad universality, and easy operability, the present invention proposes a method for estimating vegetation coverage based on a novel remote sensing index, aiming to achieve high-precision, high-efficiency, and fully automated estimation of vegetation coverage at regional and global scales. Its principle is simple, easy to implement, highly automated, and has a wide application range.

[0005] To achieve this purpose, the present invention adopts the following technical solutions:

[0006] A method for estimating vegetation coverage based on a novel remote sensing index, which includes the steps of:

[0007] Using the reflectances of the green band and the near-infrared band of a remote sensing image to calculate the simulated reflectance of the red band of the remote sensing image;

[0008] Construct a new remote sensing index using the simulated reflectance in the red band, the reflectance in the red band, and the reflectance in the blue band. The calculation formula is:

[0009]

[0010] Among them, is the value of the new remote sensing index, is the simulated reflectance in the red band, and are the reflectances in the red band and the blue band respectively;

[0011] Establish a quantitative relationship between the vegetation coverage and the new remote sensing index, and substitute the new remote sensing index into the quantitative relationship to obtain the vegetation coverage of the pixel.

[0012] Optionally, use the reflectances in the green band and the near-infrared band of the remote sensing image to calculate the simulated reflectance in the red band of the remote sensing image. The calculation formula is:

[0013]

[0014] Among them, is the simulated reflectance in the red band, and are the reflectances in the green band and the near-infrared band respectively, 、 and are the central wavelengths of the green band, the red band, and the near-infrared band of the remote sensing sensor respectively.

[0015] Optionally, the method for establishing the quantitative relationship between the vegetation coverage and the new remote sensing index is:

[0016] Combined with the linear spectral mixture model,

[0017] Among them, is the value of the new remote sensing index, is the vegetation coverage of the pixel, is the value of the new remote sensing index of the vegetation in the pixel, is the value of the new remote sensing index of the soil in the pixel.

[0018] Optionally, the vegetation 、soil , the vegetation coverage can be directly characterized by the new remote sensing index . The calculation formula is:

[0019]

[0020] .

[0021] Optionally, it further includes the steps of obtaining the reflectance of the green band, near-infrared band, red band, and blue band of the remote sensing image.

[0022] The present invention has the following characteristics:

[0023] (1) The principle is simple and easy to implement. The new remote sensing index is designed based on the spectral curve characteristics of green vegetation and soil, and the calculation is simple;

[0024] (2) Good accuracy and high automation. The new remote sensing index can directly and accurately obtain the vegetation coverage of each pixel in the remote sensing image;

[0025] (3) Wide application range, and can be widely used in a variety of common remote sensing sensor platforms. Description of the Drawings

[0026] Figure 1 It is a schematic diagram of the spectral curve characteristics of green vegetation and the simulation principle of the red band reflectance;

[0027] Figure 2 It is a schematic diagram of the spectral curve characteristics of soil and the simulation principle of the red band reflectance;

[0028] Figure 3 It is the Sentinel-2 surface reflectance image used in the vegetation coverage estimation method based on the new remote sensing index of the present invention;

[0029] Figure 4 It is the calculation result of the simulated reflectance of the red band of the remote sensing image by the vegetation coverage estimation method based on the new remote sensing index;

[0030] Figure 5 It is the calculation result of the new remote sensing index by the vegetation coverage estimation method based on the new remote sensing index. Detailed Embodiments

[0031] In the research on the remote sensing estimation method of vegetation coverage, the inventor found that there are obvious differences in the spectral reflectance curve shapes of green vegetation and soil in the visible to near-infrared band range. Using this difference, the quantitative contributions of vegetation signals and soil signals in pixels can be accurately separated. Specifically: (1) Due to the strong absorption of chlorophyll, the reflectance of green vegetation in the blue band and the red band is low and approximately equal; (2) The reflectance of soil shows an approximately linear change in the visible to near-infrared band range. Based on the above spectral characteristics, the present invention designs a vegetation coverage estimation method based on a new remote sensing index. This method first calculates the simulated reflectance of the red band using the reflectance of the green band and the near-infrared band of the remote sensing image, then constructs a new remote sensing index using the simulated reflectance of the red band, the reflectance of the red band, and the reflectance of the blue band, and finally combines the linear spectral mixture model to establish a quantitative relationship between the vegetation coverage and this new remote sensing index, so as to directly obtain the vegetation coverage of each pixel in the remote sensing image.

[0032] In order to further illustrate the technical implementation scheme of the present invention, the following takes a Sentinel-2 surface reflectance image as an example and introduces in detail the technical implementation process of a vegetation coverage estimation method based on a new remote sensing index in combination with the accompanying drawings, including the following steps:

[0033] Calculate the simulated reflectance of the red band of the remote sensing image using the reflectance of the green band and the near-infrared band of the remote sensing image;

[0034] Construct a new remote sensing index using the simulated reflectance of the red band, the reflectance of the red band, and the reflectance of the blue band,

[0035] In the research, the inventor considered that mainly based on the spectral differences between vegetation and soil, that is, (1) the reflectance of green vegetation in the blue band and the red band is low and approximately equal, then ; (2) the reflectance of soil shows an approximately linear change in the visible to near-infrared band range, then 0, and further obtain . In order to strictly ensure , VCI needs to square the numerator and denominator respectively first, and then do the division. Therefore, the vegetation signal and soil signal in the pixel are quantified as 1 and 0 respectively using the following formula. Combining the linear spectral mixture model, the quantitative relationship between the vegetation coverage and the new remote sensing index can be deduced.

[0036] The calculation formula is:

[0037]

[0038] Among them, is the value of the new remote sensing index, is the simulated reflectance of the red band, and The reflectance of the red band and the blue band respectively;

[0039] Establish a quantitative relationship between the vegetation coverage and the new remote sensing index, and substitute the new remote sensing index into the quantitative relationship to obtain the vegetation coverage of the pixel.

[0040] In one embodiment, it further includes the steps of: obtaining the reflectance of the green band, the near-infrared band, the red band, and the blue band of the remote sensing image.

[0041] In one embodiment, step one: calculate the simulated reflectance of the red band of the remote sensing image. Use the reflectance of the green band and the near-infrared band of the remote sensing image to perform linear interpolation at the position of the red band to obtain the simulated reflectance of the red band. The calculation formula is:

[0042]

[0043] In the formula, is the simulated reflectance of the red band, and are the reflectances of the green band and the near-infrared band respectively, , and are the central wavelengths of the green band, the red band, and the near-infrared band of the remote sensing sensor respectively. The simulation principle of the red band reflectance refers to Figure 1 and Figure 2 schematic diagrams. Figure 1 is a schematic diagram of the spectral curve characteristics of green vegetation and the simulation principle of red band reflectance; Figure 2 is a schematic diagram of the spectral curve characteristics of soil and the simulation principle of red band reflectance.

[0044] In this step, input a Sentinel-2 surface reflectance image, and its true color composite image is as Figure 3 shown. The Sentinel-2 image contains blue, green, red, and near-infrared bands. The central wavelengths of each band are 492 nm, 559 nm, 665 nm, and 833 nm in sequence. The image spatial resolution is 10 m, and there are a total of 1604 columns × 1145 rows of pixels. Then use the Sentinel-2 surface reflectance image to calculate the simulated reflectance of the red band, that is, Figure 4 .

[0045] In one embodiment, step two: construct a new remote sensing index. Use the simulated reflectance of the red band, the red band reflectance, and the blue band reflectance to construct a new remote sensing index. The calculation formula is:

[0046]

[0047] In the formula, is the value of the new remote sensing index, is the simulated reflectance in the red band, and are the reflectances in the red and blue bands respectively; The spectral curve characteristics of green vegetation and soil are as Figure 1 shown. For vegetation, since the reflectance in the blue band and the reflectance in the red band are approximately equal, that is , so vegetation ; For soil, since the simulated reflectance in the red band and the reflectance in the red band are approximately equal, that is , so soil .

[0048] In this case, the new remote sensing index is calculated using the simulated reflectance in the red band, the reflectance in the red band, and the reflectance in the blue band of the Sentinel-2 remote sensing image, that is Figure 5 .

[0049] In one embodiment, step three is to establish a quantitative relationship between the vegetation coverage and the new remote sensing index. Combining the linear spectral mixture model, the quantitative relationship between the vegetation coverage and the new remote sensing index can be expressed by the following calculation formula:

[0050]

[0051] In the formula, is the value of the new remote sensing index, is the vegetation coverage of the pixel, is the value of the new remote sensing index of the vegetation in the pixel, is the value of the new remote sensing index of the soil in the pixel. Since vegetation , soil , so the vegetation coverage can be directly characterized by the new remote sensing index , and its calculation formula is:

[0052]

[0053] In the formula, is the vegetation coverage of the pixel, is the value of the new remote sensing index, that is

[0054]

[0055] Thus, the calculation of as the vegetation coverage of the pixel is realized by using the reflectances in the green band, near-infrared band, red band, and blue band of the remote sensing image.

[0056] The present invention has the following characteristics:

[0057] (1) Simple principle and easy to implement. The new remote sensing index is designed based on the spectral curve characteristics of green vegetation and soil, and is easy to calculate;

[0058] (2) Good accuracy and high degree of automation. The new remote sensing index can directly and accurately obtain the vegetation coverage of each pixel in the remote sensing image;

[0059] (3) Wide range of applications, and can be widely used in a variety of common remote sensing sensor platforms.

[0060] As mentioned above, it is only a specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those familiar with the technology within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A vegetation coverage estimation method based on a new remote sensing index, characterized in that: Includes steps: The simulated reflectance of the red band of the remote sensing image is calculated using the reflectance of the green band and near-infrared band of the remote sensing image. The simulated reflectivity of the red band, the reflectivity of the red band and the reflectivity of the blue band are used to construct a new remote sensing index. The calculation formula is: in, is the new remote sensing index value, is the simulated reflectivity of the red band, and are the reflectances of the red and blue bands, respectively; A quantitative relationship between vegetation coverage and a new remote sensing index is established, and the new remote sensing index is substituted into the quantitative relationship to obtain the vegetation coverage of the pixel.

2. The vegetation coverage estimation method based on the novel remote sensing index according to claim 1, characterized in that: The simulated reflectivity of the red band of the remote sensing image is calculated using the reflectivity of the green band and near-infrared band of the remote sensing image. The calculation formula is: in, is the simulated reflectivity of the red band, and are the reflectances of the green band and the near-infrared band, , and They are the center wavelengths of the green band, red band and near-infrared band of the remote sensing sensor respectively.

3. The vegetation coverage estimation method based on the novel remote sensing index according to claim 1, characterized in that: The method to establish the quantitative relationship between vegetation coverage and the new remote sensing index is: Combined with the linear spectral mixture model, in, is the new remote sensing index value, is the vegetation coverage of the pixel, is the new remote sensing index value of vegetation in the pixel, It is a new remote sensing index value of soil in the pixel.

4. The vegetation coverage estimation method based on the novel remote sensing index according to claim 3 is characterized in that: vegetation ,soil , vegetation coverage New remote sensing index Direct characterization, the calculation formula is: 。 5. The vegetation coverage estimation method based on the novel remote sensing index according to claim 3 is characterized in that: The method also includes the steps of obtaining the reflectance of the green band, the near infrared band, the red band and the blue band of the remote sensing image.

Images