A method for estimating vegetation cover based on a new remote sensing index

Abstract

The present invention discloses a vegetation cover estimation method based on a new remote sensing index. The method uses the reflectance of the green and near-infrared bands of a remote sensing image to calculate the simulated reflectance of the red band of the remote sensing image. The simulated reflectance of the red band, the red band reflectance, and the blue band reflectance are used to construct a new remote sensing index using the formula #imgabs0#: where VCI is the new remote sensing index value, #imgabs1# is the simulated reflectance of the red band, and #imgabs2# and #imgabs3# are the reflectances of the red and blue bands, respectively. A quantitative relationship between vegetation cover and the new remote sensing index is established, and the new remote sensing index is substituted into the quantitative relationship to obtain the vegetation cover of the pixel. The method has a simple principle, is easy to implement, has a high degree of automation, and has a wide range of applications. It has broad application value in the field of vegetation and ecological remote sensing, and is suitable for remote sensing estimation and mapping of vegetation cover over large regions and even on 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 in particular to a vegetation coverage remote sensing estimation method based on a novel remote sensing index. Background Art

[0002] Vegetation cover is a key parameter for characterizing surface vegetation cover and has important applications in ecosystem carbon cycle simulation, soil erosion assessment, and climate change research. Satellite remote sensing, with its ability to acquire data over large areas, in multiple temporal phases, and at low cost, has become the primary method for estimating vegetation cover. However, existing remote sensing methods for estimating vegetation cover struggle to achieve high accuracy, broad applicability, and ease of use, limiting their potential for dynamic monitoring of large-scale vegetation cover.

[0003] A series of remote sensing estimation methods for vegetation cover 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-based vegetation cover; (2) machine learning method, which uses machine learning algorithms such as random forest to construct a nonlinear mapping relationship between remote sensing reflectance data and vegetation cover; (3) mixed pixel decomposition method, which estimates vegetation cover by solving the contribution ratio of the spectral signals of vegetation and soil in mixed pixels based on the linear spectral mixture model. The empirical regression method is only applicable to vegetation cover estimation in a specific range and for a specific vegetation type and is not universal. The machine learning method is highly dependent on the quality and representativeness of the training samples, and high-quality samples usually require a large amount of ground-based measured data and high-resolution remote sensing images. The data acquisition cost is expensive and the processing process is complicated, resulting in the lack of ease of operation of this method. The main disadvantage 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 a regional scale, which greatly increases the uncertainty of the remote sensing estimation results of vegetation cover. Summary of the Invention

[0004] This paper addresses the challenges of existing remote sensing vegetation cover estimation methods, which struggle to achieve high accuracy, broad applicability, and ease of use. By proposing a vegetation cover estimation method based on a novel remote sensing index, the method aims to achieve highly accurate, efficient, and fully automated estimation of vegetation cover at regional and global scales. Its principle is simple, easy to implement, highly automated, and widely applicable.

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

[0006] A vegetation coverage estimation method based on a new remote sensing index comprises the following steps:

[0007] 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;

[0008] A new remote sensing index is constructed using the simulated reflectivity of the red band, the red band reflectivity, and the blue band reflectivity. The calculation formula is:

[0009]

[0010] in, is a new remote sensing index value, is the simulated reflectivity of the red band, and are the reflectances of the red and blue bands, respectively;

[0011] 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.

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

[0013]

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

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

[0016] Combined with the linear spectral mixture model,

[0017] in, is a 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.

[0018] Optional, vegetation ,soil , vegetation coverage New remote sensing index Direct characterization, the calculation formula is:

[0019]

[0020] .

[0021] Optionally, the method further includes the steps of obtaining the reflectivity 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 is easy to calculate.

[0024] (2) With 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;

[0025] (3) It has a wide range of applications and can be widely used in a variety of common remote sensing sensor platforms. BRIEF DESCRIPTION OF THE DRAWINGS

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

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

[0028] Figure 3 The Sentinel-2 surface reflectance image used in the vegetation cover estimation method based on the novel remote sensing index of the present invention;

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

[0030] Figure 5 It is the calculation result of the new remote sensing index based on the vegetation cover estimation method of the new remote sensing index. DETAILED DESCRIPTION

[0031] In their research on remote sensing estimation methods for vegetation cover, the inventors found that the spectral reflectance curves of green vegetation and soil in the visible to near-infrared band range have significant differences. This difference can be used to accurately separate the quantitative contributions of vegetation signals and soil signals in pixels. Specifically: (1) Due to the strong absorption of chlorophyll, the reflectance of green vegetation in the blue and red bands is low and approximately equal; (2) The reflectance of soil varies approximately linearly in the visible-near-infrared band range. Based on the above spectral characteristics, the present invention designs a vegetation cover estimation method based on a new remote sensing index. The method first uses the reflectance of the green and near-infrared bands of the remote sensing image to calculate the simulated reflectance of the red band. Then, the simulated reflectance of the red band, the reflectance of the red band, and the reflectance of the blue band are used to construct a new remote sensing index. Finally, a linear spectral mixture model is combined to establish a quantitative relationship between vegetation cover and the new remote sensing index, thereby directly obtaining the vegetation cover of each pixel in the remote sensing image.

[0032] To further illustrate the technical implementation of the present invention, the following, in conjunction with the accompanying drawings, takes a Sentinel-2 surface reflectance image as an example to describe in detail the technical implementation process of a vegetation cover estimation method based on a new remote sensing index, including the following steps:

[0033] 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;

[0034] A new remote sensing index is constructed using the simulated reflectivity of the red band, the reflectivity of the red band, and the reflectivity of the blue band.

[0035] The inventors considered in their research: mainly based on the difference between vegetation and soil in the spectrum, that is, (1) the reflectivity of green vegetation in the blue band and the red band is low and approximately equal, then ; (2) The reflectivity of the soil changes approximately linearly in the visible-near infrared band, then 0, further get In order to strictly ensure , VCI needs to square the numerator and denominator respectively and then perform division. Therefore, the following formula is used to quantify the vegetation signal and soil signal in the pixel to 1 and 0 respectively. Combined with the linear spectral mixture model, the quantitative relationship between vegetation cover and the new remote sensing index can be derived.

[0036] The calculation formula is:

[0037]

[0038] in, is a new remote sensing index value, is the simulated reflectivity of the red band, and are the reflectances of the red and blue bands, respectively;

[0039] A quantitative relationship between vegetation coverage and the new remote sensing index is established, and the new remote sensing index is introduced into the quantitative relationship to obtain the vegetation coverage of the pixel.

[0040] In one embodiment, the step is further included: obtaining the reflectivity 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 1 is to calculate the simulated reflectance of the red band of the remote sensing image. The reflectance of the green band and the near-infrared band of the remote sensing image is linearly interpolated at the red band position to obtain the simulated reflectance of the red band. The calculation formula is:

[0042]

[0043] Where, is the simulated reflectivity of the red band, and are the reflectances of the green band and the near-infrared band, 、 and These are the central wavelengths of the green band, red band, and near-infrared band of the remote sensing sensor. Figure 1 and Figure 2 Schematic diagram. Figure 1 It is a schematic diagram of the green vegetation spectral curve characteristics and the red band reflectivity simulation principle; Figure 2 It is a schematic diagram of soil spectral curve characteristics and red band reflectivity simulation principle.

[0044] In this step, a Sentinel-2 surface reflectance image is input, and its true color composite image is as follows: Figure 3 As shown in Figure 2. The Sentinel-2 image contains blue, green, red, and near-infrared bands, with the central wavelengths of each band being 492 nm, 559 nm, 665 nm, and 833 nm, respectively. The image has a spatial resolution of 10 m and a total of 1604 columns × 1145 rows of pixels. The simulated reflectance of the red band is then calculated using the Sentinel-2 surface reflectance image, which is: Figure 4 .

[0045] In one embodiment, step 2 is to construct a new remote sensing index. The new remote sensing index is constructed using the simulated reflectivity of the red band, the reflectivity of the red band, and the reflectivity of the blue band. The calculation formula is:

[0046]

[0047] Where, is a new remote sensing index value, is the simulated reflectivity of the red band, and are the reflectance of the red and blue bands respectively; the spectral curve characteristics of green vegetation and soil are as follows Figure 1 As shown in , for vegetation, since the reflectance of the blue band is approximately equal to the reflectance of the red band, that is, , so vegetation ; For soil, since the simulated red band reflectivity and the red band reflectivity are approximately equal, that is, , so the soil .

[0048] This case uses the simulated reflectance of the red band of Sentinel-2 remote sensing images, the red band reflectance, and the blue band reflectance to calculate the new remote sensing index, namely Figure 5 .

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

[0050]

[0051] Where, is a new remote sensing index value, is the vegetation coverage of the pixel, is the new remote sensing index value of vegetation in the pixel, is a new remote sensing index value of soil in the pixel. ,soil , so the vegetation coverage New remote sensing index Direct characterization, the calculation formula is:

[0052]

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

[0054]

[0055] The reflectivity of the green band, near infrared band, red band and blue band of remote sensing images is used to achieve Calculation of vegetation coverage of the pixel.

[0056] The present invention has the following characteristics:

[0057] (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 is easy to calculate.

[0058] (2) With 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) It has a wide range of applications and can be widely used in a variety of common remote sensing sensor platforms.

[0060] The above is merely one specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by anyone familiar with the art within the technical scope disclosed by the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A vegetation coverage estimation method based on a new remote sensing index, characterized in that: Including 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; A new remote sensing index is constructed using the simulated reflectivity of the red band, the red band reflectivity, and the blue band reflectivity. The calculation formula is: in, is a new remote sensing index value, is the simulated reflectivity of the red band, and are the reflectances of the red and blue bands, respectively; Establishing a quantitative relationship between vegetation coverage and a new remote sensing index, and substituting the new remote sensing index into the quantitative relationship to obtain the vegetation coverage of the pixel; Combined with the linear spectral mixing model, the quantitative relationship between vegetation cover and the new remote sensing index is expressed as the following calculation formula: in, is a new remote sensing index value, is the vegetation coverage of the pixel, is the new remote sensing index value of vegetation in the pixel, is the new remote sensing index value of soil in the pixel; 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 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 are the central wavelengths of the green band, red band and near-infrared band of the remote sensing sensor respectively; vegetation ,soil , vegetation coverage New remote sensing index Direct characterization, the calculation formula is:

2. The vegetation coverage estimation method based on the novel remote sensing index according to claim 1, 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