A green crop recognition method, system, device and medium

Abstract

The application discloses a green crop identification method, system, device and medium, and relates to the field of crop identification.The method comprises the following steps: acquiring a multispectral remote sensing image; calculating a vegetation coverage index in the multispectral remote sensing image according to a red light band and a near-infrared band in the multispectral remote sensing image; identifying each pixel in the multispectral remote sensing image one by one to obtain a normalized vegetation index of a target pixel; calculating a slope of a reflection spectrum curve of the target pixel from the red light band to the near-infrared band according to spectral reflectivity of the target pixel in the red light band and the near-infrared band; constructing a green crop identification index according to the slope, the vegetation coverage index and the normalized vegetation index of the target pixel; and determining a green crop identification result according to the green crop identification index.The application can solve the problems of low green crop identification precision and quantification difficulty.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of crop identification, and in particular to a green crop identification method, system, device and medium. BACKGROUND

[0002] The remote sensing method based on spectral analysis technology has the advantages of non-destructive, high throughput, large area, etc., and can accurately monitor the target object in real time. When the spectrum information is extracted from the target area for crop growth monitoring, the neglect of the soil background will inevitably introduce impurities and noise, which will eventually seriously interfere with the crop growth monitoring results. Therefore, accurate identification of crops in remote sensing images will be a key step to determine whether a series of subsequent research results are accurate.

[0003] The main crop identification methods at present are: color model-based identification method, machine learning-based identification method and index threshold-based identification method, but the above three methods also have obvious shortcomings. The color model-based method is not accurate enough; the machine learning-based method has high training cost and cannot guarantee universality; the index threshold-based method is easily affected by light conditions and vegetation coverage changes. So far, there is still no stable and quantitative crop identification method suitable for different growth periods and complex light environments. SUMMARY

[0004] The purpose of the present application is to provide a green crop identification method, system, device and medium to solve the problems of low green crop identification accuracy and quantification difficulty.

[0005] To achieve the above purpose, the present application provides the following solutions:

[0006] A green crop identification method, comprising:

[0007] acquiring a multispectral remote sensing image;

[0008] calculating a vegetation coverage index in the multispectral remote sensing image according to a red light band and a near-infrared band in the multispectral remote sensing image;

[0009] identifying each pixel in the multispectral remote sensing image one by one to obtain a normalized vegetation index of a target pixel;

[0010] calculating a slope of a reflection spectrum curve of the target pixel from the red light band to the near-infrared band according to the spectral reflectance of the target pixel in the red light band and the near-infrared band;

[0011] constructing a green crop identification index according to the slope, the vegetation coverage index and the normalized vegetation index of the target pixel;

[0012] Determine a green crop identification result according to the green crop identification index; the green crop identification result is that the target pixel to be identified in the image is a green crop or the target pixel to be identified in the image is not a green crop.

[0013] Optionally, a vegetation coverage index in the multispectral remote sensing image is calculated according to a red light band and a near-infrared band in the multispectral remote sensing image, and specifically includes:

[0014] A normalized vegetation index of all pixels is calculated according to the red light band and the near-infrared band in the multispectral remote sensing image.

[0015] A normalized vegetation index maximum value, a normalized vegetation index minimum value and a normalized vegetation index average value are determined according to the normalized vegetation index of all pixels.

[0016] The vegetation coverage index in the multispectral remote sensing image is calculated according to the normalized vegetation index maximum value, the normalized vegetation index minimum value and the normalized vegetation index average value.

[0017] Optionally, an expression of the vegetation coverage index is:

[0018] C v ((Nave-Nmin) / (Nmax-Nave)) ^0.5

[0019] wherein, C v is the vegetation coverage index, Nave is the normalized vegetation index average value, Nmin is the normalized vegetation index minimum value, and Nmax is the normalized vegetation index maximum value.

[0020] Optionally, an expression of the slope is:

[0021] K p (I Nir -I R ) / 0.19

[0022] wherein, K p is the slope, I Nir is the spectral reflectance of the near-infrared band, and I R is the spectral reflectance of the red light band.

[0023] Optionally, a green crop identification index is constructed according to the slope, the vegetation coverage index and the normalized vegetation index of the target pixel, and specifically includes:

[0024] Determine whether the slope is greater than the vegetation coverage index to obtain a first determination result;

[0025] if the first determination result is yes, then the sum of the normalized vegetation index of the target pixel and the average value of the normalized vegetation index is taken as a green crop identification index of the target pixel;

[0026] if the first determination result is no, then the difference between the normalized vegetation index of the target pixel and the average value of the normalized vegetation index is taken as a green crop identification index of the target pixel.

[0027] Optionally, the green crop identification result is determined according to the green crop identification index, and specifically includes:

[0028] the green crop identification index is normalized to obtain a normalized green crop identification index;

[0029] it is determined whether the normalized green crop identification index is greater than a set threshold value, to obtain a second determination result;

[0030] if the second determination result is yes, then it is determined that the target pixel to be identified in the image is a green crop;

[0031] if the second determination result is no, then it is determined that the target pixel to be identified in the image is not a green crop.

[0032] The present application also provides a green crop identification system, comprising:

[0033] an acquisition module configured to acquire a multispectral remote sensing image;

[0034] a vegetation coverage index calculation module configured to calculate a vegetation coverage index in the multispectral remote sensing image according to a red light band and a near-infrared band in the multispectral remote sensing image;

[0035] a normalized vegetation index calculation module configured to identify each pixel in the multispectral remote sensing image one by one to obtain a normalized vegetation index of a target pixel;

[0036] a slope calculation module configured to calculate a slope of a reflection spectrum curve of the target pixel from the red light band to the near-infrared band according to spectral reflectance of the target pixel in the red light band and the near-infrared band;

[0037] a construction module configured to construct a green crop identification index according to the slope, the vegetation coverage index and the normalized vegetation index of the target pixel;

[0038] an identification module configured to determine a green crop identification result according to the green crop identification index; the green crop identification result is that the target pixel to be identified in the image is a green crop or that the target pixel to be identified in the image is not a green crop.

[0039] Optionally, the vegetation coverage index calculation module specifically comprises:

[0040] a normalized vegetation index calculation unit for calculating normalized vegetation indexes of all pixels according to red light bands and near-infrared bands in the multispectral remote sensing image;

[0041] a statistics unit for determining a normalized vegetation index maximum value, a normalized vegetation index minimum value and a normalized vegetation index average value according to the normalized vegetation indexes of all pixels;

[0042] a vegetation coverage index calculation unit for calculating a vegetation coverage index in the multispectral remote sensing image according to the normalized vegetation index maximum value, the normalized vegetation index minimum value and the normalized vegetation index average value.

[0043] The application further provides an electronic device comprising:

[0044] one or more processors;

[0045] a storage device having one or more programs stored thereon;

[0046] which, when executed by the one or more processors, cause the one or more processors to implement the method as described.

[0047] The application further provides a computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method as described.

[0048] According to the embodiments of the application, the following technical effects are achieved:

[0049] The application obtains a multispectral remote sensing image; calculates a vegetation coverage index in the multispectral remote sensing image according to a red light band and a near-infrared band in the multispectral remote sensing image; identifies each pixel in the multispectral remote sensing image one by one to obtain a normalized vegetation index of a target pixel; calculates a slope of a reflection spectrum curve of the target pixel from the red light band to the near-infrared band according to spectral reflectivity of the target pixel in the red light band and the near-infrared band; constructs a green crop identification index according to the slope, the vegetation coverage index and the normalized vegetation index of the target pixel; and determines a green crop identification result according to the green crop identification index; the green crop identification result is that the target pixel to be identified in the image is a green crop or the target pixel to be identified in the image is not a green crop. The vegetation coverage index is used to quantitatively describe the ground crop vegetation coverage, the slope characteristics of the reflection spectrum of the green crop from the red light band to the near-infrared band are used to construct the green crop identification index which can stably identify the green crop pixel, and the problems of low precision and difficulty in quantification in the green crop identification research are solved. BRIEF DESCRIPTION OF DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0051] Figure 1 The green crop identification method flowchart provided by the present application;

[0052] Figure 2 The crop identification result under the condition of strong light and low vegetation coverage;

[0053] Figure 3 The crop identification result under the condition of dark light and high vegetation coverage. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present application will be described clearly and completely in the following with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0055] The purpose of the present application is to provide a green crop identification method, system, device and medium to solve the problems of low identification precision and difficulty in quantification of green crops.

[0056] In order to make the above objectives, characteristics and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0057] As shown in Figure 1 The present application provides a green crop identification method, comprising:

[0058] Obtaining a multispectral remote sensing image.

[0059] Calculating a vegetation coverage index in the multispectral remote sensing image according to a red light band and a near-infrared band in the multispectral remote sensing image.

[0060] The vegetation coverage index is calculated according to the red light band and the near-infrared band in the multispectral remote sensing image, specifically comprising: calculating the normalized difference vegetation index of all pixels according to the red light band and the near-infrared band in the multispectral remote sensing image; determining the maximum value, the minimum value and the average value of the normalized difference vegetation index according to the normalized difference vegetation index of all pixels; and calculating the vegetation coverage index in the multispectral remote sensing image according to the maximum value, the minimum value and the average value of the normalized difference vegetation index.

[0061] In actual application, data of two bands of red light R (650 nm) and near-infrared Nir (840 nm) are selected from the multispectral remote sensing image. The NDVI value of all pixels in the multispectral remote sensing image is calculated according to the normalized difference vegetation index (NDVI) formula NDVI=(Nir-R) / (Nir+R), and the maximum value Nmax, the average value Nave and the minimum value Nmin are counted.

[0062] The expression of the vegetation coverage index is:

[0063] C v = ((Nave-Nmin) / (Nmax-Nave)) ^0.5

[0064] wherein, C v is the vegetation coverage index, Nave is the average value of the normalized difference vegetation index, Nmin is the minimum value of the normalized difference vegetation index, and Nmax is the maximum value of the normalized difference vegetation index.

[0065] Each pixel in the multispectral remote sensing image is identified one by one to obtain the normalized difference vegetation index of the target pixel, denoted as NDVI p .

[0066] The slope from red to near-infrared is calculated according to the spectral reflectance of the target pixel in the red light band and the near-infrared band. The spectral reflectance of the pixel in the red light R (650 nm) and the near-infrared Nir (840 nm) is extracted, respectively denoted as I R and I Nir , and the slope of the pixel from red to near-infrared is calculated K p .

[0067] The expression of the slope is:

[0068] K p =(I Nir -I R ) / 0.19

[0069] wherein, K p is the slope, I Nir is the spectral reflectance in the near-infrared band, and I R is the spectral reflectance in the red light band.

[0070] A green crop recognition index is constructed according to the slope, the vegetation coverage index, and the normalized difference vegetation index of the target pixel.

[0071] A green crop recognition index is constructed according to the slope, the vegetation coverage index, and the normalized difference vegetation index of the target pixel, specifically including: judging whether the slope is greater than the vegetation coverage index to obtain a first judgment result; if the first judgment result is yes, the sum of the normalized difference vegetation index of the target pixel and the average value of the normalized difference vegetation index is taken as the green crop recognition index of the target pixel; if the first judgment result is no, the difference between the normalized difference vegetation index of the target pixel and the average value of the normalized difference vegetation index is taken as the green crop recognition index of the target pixel.

[0072] In actual application, the slope from red to near-infrared of the pixel to be recognized K p is compared with the size of the vegetation coverage of the multispectral remote sensing image Cv , and a green crop recognition index Q is constructed. If K p > Cv , let Q =NDVI p +Nave; if K p < Cv , let Q =NDVI p-Nave. The amount of chlorophyll in a crop directly affects the stretching and slope of its reflectance spectrum from the red to the near-infrared band. K p This can reflect the degree of stretching. When K p > Cv When the pixel to be identified has a high chlorophyll content in the entire multispectral image, it can be considered that the pixel to be identified is closer to a green crop pixel; when K p < Cv In this case, the pixel to be identified can be considered to have a low chlorophyll content in the entire multispectral image, which also means that the pixel to be identified is less likely to be a green crop pixel. ±Nave is used to increase the green crop identification index for different chlorophyll contents. Q The differences in the data allow for the subsequent differentiation of green crop pixels from other pixels using an intermediate threshold.

[0073] The green crop identification result is determined based on the green crop identification index; the green crop identification result is either that the target pixel to be identified in the image is a green crop or that the target pixel to be identified in the image is not a green crop.

[0074] Determining the green crop identification result based on the green crop identification index specifically includes: normalizing the green crop identification index to obtain a normalized green crop identification index; determining whether the normalized green crop identification index is greater than a set threshold to obtain a second judgment result; if the second judgment result is yes, then the target pixel to be identified in the image is determined to be a green crop; if the second judgment result is no, then the target pixel to be identified in the image is determined not to be a green crop.

[0075] In practical applications, the green crop identification index Q Normalization was performed. The difference between green crop pixels and non-green crop pixels was further widened. Q Value differences make green crop pixels Q Values ​​approaching 1 indicate non-green crop pixels. Q The value approaches 0. If the normalized green crop identification index of the target pixel to be identified... Q If the value is greater than 0.5, the pixel is identified as a green crop pixel; otherwise, it is identified as another object.

[0076] The main crop recognition methods at present include a color model-based recognition method, a machine learning-based recognition method and an index threshold-based recognition method, but the above three methods have obvious shortcomings, the color model-based method is not accurate enough, the machine learning-based method has high training cost and cannot guarantee universality, and the recognition accuracy of the index threshold-based method is easily affected by light conditions and vegetation coverage changes.

[0077] The method provided by the application fully utilizes the reflection spectrum characteristics of green crops in the red light band to the near-infrared band, can quantitatively recognize green crop pixels under different light conditions and different vegetation coverage conditions, has high speed and high accuracy in the whole recognition process and does not need feedback training, is especially suitable for solving the green crop recognition problem based on unmanned aerial vehicle and satellite remote sensing platforms, and has high popularization and application value.

[0078] The method provided by the application is applied to green crop recognition of a multispectral remote sensing image, and the recognition result is as shown in Figure 2 and Figure 3 .

[0079] The application further provides a green crop recognition system, comprising:

[0080] An acquisition module is configured to acquire a multispectral remote sensing image.

[0081] A vegetation coverage index calculation module is configured to calculate a vegetation coverage index in the multispectral remote sensing image according to a red light band and a near-infrared band in the multispectral remote sensing image.

[0082] A target pixel normalized vegetation index calculation module is configured to identify each pixel in the multispectral remote sensing image one by one to obtain a normalized vegetation index of a target pixel.

[0083] A slope calculation module is configured to calculate a slope of a reflection spectrum curve of the target pixel from the red light band to the near-infrared band according to spectral reflectance of the target pixel in the red light band and the near-infrared band.

[0084] A construction module is configured to construct a green crop recognition index according to the slope, the vegetation coverage index and the normalized vegetation index of the target pixel.

[0085] An identification module is configured to determine a green crop recognition result according to the green crop recognition index, wherein the green crop recognition result is that a target pixel to be recognized in the image is a green crop or that the target pixel to be recognized in the image is not a green crop.

[0086] As an optional implementation, the vegetation coverage index calculation module specifically comprises:

[0087] A normalized difference vegetation index calculation unit is configured to calculate a normalized difference vegetation index of all pixels according to a red band and a near-infrared band in the multispectral remote sensing image.

[0088] A statistics unit is configured to determine a normalized difference vegetation index maximum value, a normalized difference vegetation index minimum value and a normalized difference vegetation index average value according to the normalized difference vegetation index of all pixels.

[0089] A vegetation fraction index calculation unit is configured to calculate a vegetation fraction index in the multispectral remote sensing image according to the normalized difference vegetation index maximum value, the normalized difference vegetation index minimum value and the normalized difference vegetation index average value.

[0090] The present application also provides an electronic device, comprising: one or more processors; a storage device having one or more programs stored thereon; and when the one or more programs are executed by the one or more processors, the one or more processors implement the method as described.

[0091] The present application also provides a computer storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement the method as described.

[0092] The present application proposes a vegetation fraction index C v , which realizes quantitative description of crop vegetation fraction on the ground. A green crop recognition method using a slope feature of a reflection spectrum of the crop from a red band to a near-infrared band is proposed. The method overcomes the problem of quantitative difficulty in green crop recognition using a normalized difference vegetation index (NDVI) and realizes quantitative recognition of green crops with 0.5 as a threshold under complex vegetation cover and light conditions.

[0093] The embodiments in the specification are described in a progressive manner, and each embodiment focuses on the difference from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiments, the description is relatively simple because it corresponds to the method disclosed in the embodiments. The relevant parts can be referred to the description of the method.

[0094] The principles and implementation manners of the present application are described by using specific examples. The above description of the embodiments is only used to help understand the method of the present application and its core idea. For those skilled in the art, the specific implementation manners and application scope can be changed according to the idea of the present application. In conclusion, the content of the specification should not be understood as a limitation of the present application.

Claims

1. A method for identifying green crops, characterized in that, include: Acquire multispectral remote sensing images; The vegetation coverage index in the multispectral remote sensing image is calculated based on the red and near-infrared bands in the image. Each pixel in the multispectral remote sensing image is identified individually to obtain the normalized vegetation index of the target pixel; The slope of the reflectance spectrum curve of the target pixel from the red band to the near-infrared band is calculated based on the spectral reflectance of the target pixel in the red band and the near-infrared band. A green crop identification index is constructed based on the slope, the vegetation coverage index, and the normalized vegetation index of the target pixel. The green crop identification result is determined based on the green crop identification index; the green crop identification result is either that the target pixel to be identified in the image is a green crop or that the target pixel to be identified in the image is not a green crop. The vegetation cover index in the multispectral remote sensing image is calculated based on the red and near-infrared bands, specifically including: The normalized vegetation index of all pixels is calculated based on the red and near-infrared bands in the multispectral remote sensing image. The maximum, minimum, and average normalized vegetation index (NVI) values ​​are determined based on the normalized vegetation index of all pixels. The vegetation coverage index in the multispectral remote sensing image is calculated based on the maximum value of the normalized vegetation index, the minimum value of the normalized vegetation index, and the average value of the normalized vegetation index. The expression for the vegetation coverage index is: C v =((Nave-Nmin) / (Nmax-Nave)) ^0.5 in, C v Nave is the vegetation coverage index, Nmin is the normalized average vegetation index, and Nmax is the normalized average vegetation index. The expression for the slope is: K p =(I Nir -I R ) / 0.19 in, K p Let I be the slope. Nir I represents the spectral reflectance in the near-infrared band. R The spectral reflectance is in the red light band. A green crop identification index is constructed based on the slope, the vegetation coverage index, and the normalized vegetation index of the target pixel, specifically including: Determine whether the slope is greater than the vegetation coverage index to obtain a first determination result; If the first judgment result is yes, then the sum of the normalized vegetation index of the target pixel and the average value of the normalized vegetation index is taken as the green crop identification index of the target pixel. If the first determination result is negative, then the difference between the normalized vegetation index of the target pixel and the average value of the normalized vegetation index is taken as the green crop identification index of the target pixel.

2. The green crop identification method according to claim 1, characterized in that, The green crop identification result is determined based on the green crop identification index, specifically including: The green crop identification index is normalized to obtain the normalized green crop identification index; Determine whether the normalized green crop identification index is greater than a set threshold to obtain a second determination result; If the second judgment result is yes, then the target pixel to be identified in the image is determined to be a green crop; If the second judgment result is negative, then it is determined that the target pixel to be identified in the image is not a green crop.

3. A green crop identification system, characterized in that, include: The acquisition module is used to acquire multispectral remote sensing images; The vegetation coverage index calculation module is used to calculate the vegetation coverage index in the multispectral remote sensing image based on the red light band and near-infrared band in the multispectral remote sensing image. The normalized vegetation index (NVI) calculation module for target pixels is used to identify each pixel in the multispectral remote sensing image one by one to obtain the normalized vegetation index of the target pixels. The slope calculation module is used to calculate the slope of the reflectance spectrum curve of the target pixel from the red light band to the near infrared band based on the spectral reflectance of the target pixel in the red light band and the near infrared band. A construction module is used to construct a green crop identification index based on the slope, the vegetation coverage index, and the normalized vegetation index of the target pixel. The identification module is used to determine the green crop identification result based on the green crop identification index; the green crop identification result is either that the target pixel to be identified in the image is a green crop or that the target pixel to be identified in the image is not a green crop. The vegetation coverage index calculation module specifically includes: The normalized vegetation index calculation unit for all pixels is used to calculate the normalized vegetation index for all pixels based on the red band and near-infrared band in the multispectral remote sensing image. The statistical unit is used to determine the maximum, minimum, and average normalized vegetation index (NVI) values ​​based on the NVI values ​​of all pixels. The vegetation coverage index calculation unit is used to calculate the vegetation coverage index in the multispectral remote sensing image based on the maximum value of the normalized vegetation index, the minimum value of the normalized vegetation index, and the average value of the normalized vegetation index. The expression for the vegetation coverage index is: C v =((Nave-Nmin) / (Nmax-Nave)) ^0.5 in, C v Nave is the vegetation coverage index, Nmin is the normalized average vegetation index, and Nmax is the normalized average vegetation index. The expression for the slope is: K p =(I Nir -I R ) / 0.19 in, K p Let I be the slope. Nir I represents the spectral reflectance in the near-infrared band. R The spectral reflectance is in the red light band. A green crop identification index is constructed based on the slope, the vegetation coverage index, and the normalized vegetation index of the target pixel, specifically including: Determine whether the slope is greater than the vegetation coverage index to obtain a first determination result; If the first judgment result is yes, then the sum of the normalized vegetation index of the target pixel and the average value of the normalized vegetation index is taken as the green crop identification index of the target pixel. If the first determination result is negative, then the difference between the normalized vegetation index of the target pixel and the average value of the normalized vegetation index is taken as the green crop identification index of the target pixel.

4. An electronic device, characterized in that, include: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors cause the one or more processors to implement the method as described in any one of claims 1 to 2.

5. A computer storage medium, characterized in that, It stores a computer program thereon, wherein the computer program, when executed by a processor, implements the method as described in any one of claims 1 to 2.

Images