Chlorophyll content inversion method, system, equipment and medium

Abstract

The invention discloses a chlorophyll content inversion method, system, equipment and medium, and relates to the field of chlorophyll content quantitative inversion, and the chlorophyll content inversion method comprises the following steps: processing leaf spectrum data by using a fractional order derivative to obtain fractional order derivative data; performing continuous wavelet transform processing on the leaf spectral data by using a Gaussian second derivative to obtain wavelet energy coefficients of multiple scales; according to a required wave band capable of reflecting the chlorophyll content, finding out fractional order derivative data and a wavelet energy coefficient under the corresponding wave band; inputting the fractional derivative data corresponding to three wavebands in the required wavebands into a three-waveband vegetation index model to obtain a three-waveband vegetation index; inputting the selected wavelet energy coefficient into a vegetation index model corresponding to the required wave band to obtain a vegetation index; according to the evaluated three-waveband vegetation index model and vegetation index model, the chlorophyll content is obtained through inversion. The accuracy of chlorophyll content analysis can be improved.

Description

Technical Field

[0001] The invention relates to the field of quantitative inversion of chlorophyll content, and in particular to an inversion method, system, equipment and medium for chlorophyll content. Background Art

[0002] Citrus is one of the most important fruits in the world. It is suitable for growing in warm and humid climates and is one of the typical tropical and subtropical evergreen fruit trees. Chlorophyll reflects the growth status and physiological parameters of plants. It is a biological pigment involved in plant photosynthesis and plays a key role in the growth and yield of citrus. Studies have found that chlorophyll is affected by many factors on the spectral response curve, including leaf water content, carotenoids, flavonoids and their derivatives. At the same time, the baseline drift caused by the instrument itself will also affect the accuracy of chlorophyll estimation. Chlorophyll has its sensitive bands in the range of 400-500nm and 650-700nm. Due to the presence of carotenoids, anthocyanins and flavonoids, the light signals in these bands will be interfered by the absorption peaks of these pigments, especially when these absorption peaks overlap with the absorption peaks of chlorophyll, it is very difficult to extract the sensitive bands of chlorophyll. Therefore, effectively eliminating the influence of spectral overlapping peaks and baseline drift and accurately extracting the sensitive bands of chlorophyll are necessary prerequisites for improving the accuracy of chlorophyll content estimation.

[0003] In the existing technology, derivative spectroscopy is a relatively effective technology for eliminating background signals and solving overlapping spectral features in analytical chemistry, which improves the accuracy of leaf chlorophyll content inversion. It can also improve the sensitivity of spectral data, enhance spectral characteristics, and improve the spectral resolution of overlapping peaks.

[0004] However, derivative spectroscopy will cause some of the information present in the original spectrum to disappear and enhance the signal caused by high-frequency noise, resulting in poor accuracy in chlorophyll content analysis. Summary of the invention

[0005] The embodiments of the present invention provide a chlorophyll content inversion method, system, device and medium, which can solve the problem of poor accuracy of chlorophyll content analysis in the prior art.

[0006] The embodiment of the present invention provides a chlorophyll content inversion method, comprising the following steps: obtaining leaf spectral data containing different bands; processing the leaf spectral data using fractional derivatives to obtain fractional derivative data that can enhance the characteristics of the spectral data; performing continuous wavelet transform processing on the leaf spectral data using Gaussian second-order derivatives to obtain wavelet energy coefficients of multiple scales that can reflect the characteristics of the spectral data in multiple frequencies; finding the fractional derivative data and wavelet energy coefficients in the corresponding bands according to the required bands that can reflect the chlorophyll content; inputting the fractional derivative data corresponding to three bands in the required bands into the three The three-band vegetation index is obtained by using the band vegetation index model; the selected wavelet energy coefficient is input into the vegetation index model corresponding to the required band to obtain the vegetation index; the three-band vegetation index and vegetation index are respectively subjected to correlation analysis and preliminary feature screening by using the correlation coefficient method and the partial least squares method; the precision of the screened three-band vegetation index and vegetation index is respectively evaluated by the Gaussian process regression algorithm, and the three-band vegetation index model and vegetation index model that can reflect the chlorophyll content are determined according to the precision evaluation results; the chlorophyll content is inverted according to the determined three-band vegetation index model and vegetation index model.

[0007] Furthermore, the inversion obtains the chlorophyll content, and the specific steps include: inverting the bands adapted to the three-band vegetation index model and the vegetation index model through the leaf spectral data to be tested, and the inverted bands can reflect the chlorophyll content; inverting the chlorophyll content through the inverted bands.

[0008] Furthermore, the use of fractional derivatives to process the leaf spectral data specifically includes the following steps: smoothing the leaf spectral data using a Savitzky-Golay filter Savitzky-Golay algorithm to remove noise; selecting the fractional derivative Grunwald-Letnikov to transform the processed leaf spectral data to obtain fractional derivative data of 0.1 to 2.0 orders.

[0009] Furthermore, the continuous wavelet transform processing of the leaf spectral data using the second-order derivative of Gaussian specifically includes: selecting the second-order derivative of Gaussian function as the basis function, performing continuous wavelet transform processing of the wavelet energy coefficients of 10 scales on the leaf spectral data, and obtaining the wavelet energy coefficients of 10 scales.

[0010] Furthermore, the calculation formula of the three-band vegetation index TBI2 is:

[0011]

[0012] Among them, R λ1 and R λ2 are two wavelengths randomly selected from 400nm to 2400nm, Rλ3 is a wavelength randomly selected from 500nm to 555nm, 630nm to 650nm, 690nm to 715nm and 735nm to 740nm, and R λ1 ≠R λ2 ≠R λ3 .

[0013] Furthermore, the correlation coefficient method uses the Pearson correlation coefficient which can measure the degree of linear correlation between two variables.

[0014] An embodiment of the present invention provides a chlorophyll content inversion system, comprising:

[0015] A data acquisition module, used to acquire leaf spectral data containing different bands;

[0016] The data processing module is used to process the leaf spectral data using fractional derivatives to obtain fractional derivative data that can enhance the characteristics of the spectral data; to perform continuous wavelet transform processing on the leaf spectral data using Gaussian second-order derivatives to obtain wavelet energy coefficients of multiple scales that can reflect the characteristics of the spectral data in multiple frequencies; and to find the fractional derivative data and wavelet energy coefficients in the corresponding band according to the required band that can reflect the chlorophyll content;

[0017] The model selection module is used to input the fractional derivative data corresponding to the three bands in the required band into the three-band vegetation index model to obtain the three-band vegetation index; input the selected wavelet energy coefficient into the vegetation index model corresponding to the required band to obtain the vegetation index; use the correlation coefficient method and the partial least squares method to perform correlation analysis and preliminary feature screening on the three-band vegetation index and vegetation index respectively; use the Gaussian process regression algorithm to perform accuracy assessment on the screened three-band vegetation index and vegetation index respectively, and determine the three-band vegetation index model and vegetation index model that can reflect the chlorophyll content according to the accuracy assessment results;

[0018] The chlorophyll content inversion module is used to invert the chlorophyll content according to the determined three-band vegetation index model and vegetation index model.

[0019] An embodiment of the present invention provides a computer device, comprising: a memory and a processor; the memory stores a computer program, and the processor implements the above-mentioned chlorophyll content inversion method when executing the computer program.

[0020] An embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned chlorophyll content inversion method is implemented.

[0021] The embodiment of the present invention provides a chlorophyll content inversion method, system, device and medium. Compared with the prior art, the beneficial effects thereof are as follows:

[0022] According to the required bands that can reflect the chlorophyll content, find the fractional derivative data and wavelet energy coefficients in the corresponding bands; input the fractional derivative data corresponding to the three bands in the required bands into the three-band vegetation index model to obtain the three-band vegetation index; input the selected wavelet energy coefficient into the vegetation index model corresponding to the required band to obtain the vegetation index; use the correlation coefficient method and partial least squares method to perform correlation analysis and preliminary feature screening on the three-band vegetation index and vegetation index respectively; use the Gaussian process regression algorithm to perform accuracy assessment on the screened three-band vegetation index and vegetation index respectively, and determine the three-band vegetation index model and vegetation index model that can reflect the chlorophyll content according to the accuracy assessment results; and invert the chlorophyll content according to the determined three-band vegetation index model and vegetation index model.

[0023] Among them, the use of fractional derivatives to process leaf spectral data can ensure that the signal-to-noise ratio in the spectral data changes slowly, obtain more effective information in the leaf spectral data, and thus improve the accuracy; the continuous wavelet transform has the advantages of excellent frequency filters and intensity filters, and is effective in the analysis and denoising of leaf spectral data, which can further improve the accuracy of the analysis results; ultimately, accurate analysis of chlorophyll content is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 A flowchart of fractional derivative processing of leaf spectral data provided by an embodiment of the present invention;

[0025] Figure 2 A flowchart of processing leaf spectral data using Gaussian second-order derivatives provided in an embodiment of the present invention;

[0026] Figure 3 A flow chart for obtaining a three-band vegetation index provided by an embodiment of the present invention;

[0027] Figure 4 A flow chart for obtaining vegetation index provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific embodiments disclosed below.

[0029] See also Figure 1 to Figure 4 The embodiment of the present invention provides a method for inverting chlorophyll content, comprising the following steps:

[0030] Step 1: Obtain leaf spectral data containing different bands.

[0031] Step 2: Use fractional derivatives to process the leaf spectral data to obtain fractional derivative data that can enhance the characteristics of the spectral data; use Gaussian second-order derivatives to perform continuous wavelet transform on the leaf spectral data to obtain wavelet energy coefficients of multiple scales that can reflect the characteristics of the spectral data in multiple frequencies.

[0032] Step 3: According to the required bands that can reflect the chlorophyll content, find the fractional derivative data and wavelet energy coefficients under the corresponding bands, input the fractional derivative data corresponding to the three bands in the required bands into the three-band vegetation index model to obtain the three-band vegetation index; input the selected wavelet energy coefficient into the vegetation index model corresponding to the required band to obtain the vegetation index. The three-band vegetation index and vegetation index are analyzed for correlation and preliminary feature screening using the correlation coefficient method and partial least squares method respectively; the three-band vegetation index and vegetation index are evaluated for accuracy by the Gaussian process regression algorithm, and the three-band vegetation index model and vegetation index model that can reflect the chlorophyll content are determined based on the accuracy evaluation results.

[0033] Step 4: Based on the evaluated three-band vegetation index model and vegetation index model, invert the chlorophyll content.

[0034] The detailed explanations of some of the above steps are as follows:

[0035] The S1 data were transformed into fractional derivatives with a step size of 0.1 through the Grunwald-Letnikov derivative to obtain fractional derivatives of 0.1-2.0 order. The second-order derivative of the Gaussian function was selected as the basis function, and the original spectrum was transformed and decomposed into 10 scales to obtain wavelet energy coefficients of 10 scales.

[0036] Three-band vegetation index TBI:

[0037]

[0038] Among them, R λ1 and R λ2 are any two wavelengths between 400 and 2400 nm, R λ3 is a wavelength randomly selected from 500nm to 555nm, 630nm to 650nm, 690nm to 715nm and 735nm to 740nm, and R λ1 ≠R λ2 ≠R λ3 .

[0039] The best determination coefficient R2 of the fractional derivative validation set predicted based on the linear regression model is 0.853, the root mean square error (RMSE) is 3.447, and the relative predictive deviation (RPD) is 2.662. The best determination coefficient R2 of the continuous wavelet transform validation set predicted based on the linear regression model is 0.79. The test set of 5-fold cross validation based on Gaussian process regression of the combined optimal features of fractional derivatives and continuous wavelet transform: R2 is 0.887, RMSE is 3.283, and RPD is 2.702. The above can show that the double difference ratio index proposed in the present invention can effectively estimate the chlorophyll content of citrus leaves.

[0040] The advantages of the present invention and the method used compared to other methods are:

[0041] High-resolution spectral data also means that the noise and redundancy in the data are more serious. When obtaining sample spectral information, hyperspectral data will be affected by the special state of the sample (such as humidity and particle size) and experimental conditions, resulting in different degrees of influence on the spectral curve. Generally speaking, the spectral preprocessing methods such as multivariate scattering correction, normal variation and continuum removal transformation improve the accuracy of leaf chlorophyll content (LCC) inversion, but still cannot solve the problem of the influence of overlapping peaks. Derivative spectroscopy is a more effective technology for eliminating background signals and solving overlapping spectral features in analytical chemistry. It can improve the sensitivity of spectral data, enhance spectral characteristics, improve the spectral resolution of overlapping peaks, and help eliminate the reflectivity difference caused by material particle size differences. The first-order and second-order derivatives are the most commonly used methods in spectral preprocessing, which can improve the signal-to-noise ratio of spectral intensity. However, the traditional integer-order derivative cannot effectively detect useful information for the target variable in the case of slow slope or curvature changes, because the shapes of the nth and (n+1)th differential curves are different, and the integer-order derivative causes the information in the original spectrum to disappear, and the signal caused by high-frequency noise is easily enhanced. In contrast, the fractional order derivative (FOD) moves at smaller intervals to ensure that the signal-to-noise ratio changes slowly and is able to obtain more effective information about the target variable in the spectral curve by narrowing the spectral absorption band and sharpening the peak shape, thereby effectively reducing the influence of various radiation scattering to achieve the purpose of eliminating overlapping peaks and baseline drift.

[0042] Compared with derivative analysis, continuous wavelet transforms (CWT) are equally sensitive and have the additional advantage of excellent frequency filters and intensity filters. CWT has attracted more and more attention in image and signal analysis, decomposition, compression and denoising due to its rich wavelet basis functions, multi-resolution analysis, time-frequency positioning and other advantages. It can decompose the leaf reflectance spectrum into many scale components, and establish a model between the wavelet coefficients of each component and LCC. Studies have demonstrated the potential of CWT in estimating vegetation parameters from hyperspectral data measured in the laboratory. In LCC estimation, the original spectrum is transformed by different frequencies to separate richer spectral information, thereby obtaining more useful information. Due to the large number of hyperspectral bands, how to extract the bands suitable for LCC estimation from a large amount of hyperspectral data is the key to effectively exert the advantages of hyperspectral data. After a lot of research, it is found that the establishment of vegetation index can effectively estimate LCC, reduce the influence of interference factors, and refine the target information. A large number of studies have achieved an accuracy of 0.88-0.99 in LCC inversion, but due to the limitation of the number of sample data and the extraction of multiple characteristic variables, the robustness of the model needs to be enhanced.

[0043] An embodiment of the present invention provides a chlorophyll content inversion system, comprising:

[0044] The data acquisition module is used to acquire leaf spectral data containing different bands.

[0045] The data processing module is used to process the leaf spectral data using fractional derivatives to obtain fractional derivative data that can enhance the characteristics of the spectral data; to perform continuous wavelet transform on the leaf spectral data using Gaussian second-order derivatives to obtain wavelet energy coefficients of multiple scales that can reflect the characteristics of the spectral data within multiple frequencies; and to find the fractional derivative data and wavelet energy coefficients in the corresponding bands based on the required bands that can reflect the chlorophyll content.

[0046] The model selection module is used to input the fractional derivative data corresponding to the three bands in the required band into the three-band vegetation index model to obtain the three-band vegetation index; input the selected wavelet energy coefficient into the vegetation index model corresponding to the required band to obtain the vegetation index; use the correlation coefficient method and the partial least squares method to perform correlation analysis and preliminary feature screening on the three-band vegetation index and vegetation index respectively; use the Gaussian process regression algorithm to perform accuracy assessment on the screened three-band vegetation index and vegetation index respectively, and determine the three-band vegetation index model and vegetation index model that can reflect the chlorophyll content according to the accuracy assessment results;

[0047] The chlorophyll content inversion module is used to invert the chlorophyll content according to the determined three-band vegetation index model and vegetation index model.

[0048] An embodiment of the present invention provides a computer device, including: a memory and a processor; the memory stores a computer program, and the processor implements the steps of a chlorophyll content inversion method when executing the computer program.

[0049] An embodiment of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of a chlorophyll content inversion method are implemented.

[0050] A specific embodiment is as follows:

[0051] S1. The original leaf spectrum is smoothed using the Savitzky-Golay algorithm to remove noise.

[0052] S2. The S1 data was transformed into fractional derivatives with a step size of 0.1 by Grunwald-Letnikov derivative to obtain fractional derivatives of 0.1-2.0 order. The second-order derivative of the Gaussian function was selected as the basis function, and the original spectrum was transformed and decomposed into 10 scales to obtain wavelet energy coefficients of 10 scales.

[0053] S3. This paper constructs a three-band vegetation index Where R λ1 and R λ2 are any two wavelengths between 400 and 2400 nm, R λ3 is a wavelength randomly selected from 500nm to 555nm, 630nm to 650nm, 690nm to 715nm and 735nm to 740nm, and R λ1 ≠R λ2 ≠R λ3 .

[0054] S4. Calculate each scale wavelet coefficient through different vegetation index calculation formulas, and obtain a total of 120 vegetation indices. The calculation formula is shown in Table 1:

[0055] Table 1 Calculation formulas for different vegetation indices

[0056]

[0057]

[0058] S5. Divide the data sets in S3 and S4 into training set and test set according to the ratio of 7:3, search for characteristic band combinations on the original spectrum and the training set of different fractional derivatives, and use the correlation coefficient evaluation method to evaluate the searched characteristic band combinations.

[0059] The Pearson correlation coefficient is a method to measure the linear correlation between two variables and is often used to evaluate the relationship between data variables. Its value range is between -1 and 1. The closer its value is to 1 or -1, the stronger the positive correlation between the two variables; the lower the correlation between the two, the closer it is to zero.

[0060] TBI λ1 =571nm; R λ2 =1697nm; R λ3 =740nm) has a maximum correlation coefficient of 0.927 at the 0.9th order derivative; the maximum correlation coefficient of MTCI at scale 6 is 0.889.

[0061] S6. Use partial least squares to perform regression analysis on the feature sets in S3 and S4 to obtain the determination coefficient R of different vegetation indices in different fractional derivatives. 2 , extract the feature band combination with the highest accuracy, and verify the effectiveness of obtaining the optimal features through S3 and S4.

[0062] Partial least squares is a multivariate statistical analysis method that integrates linear regression, principal component analysis, and canonical correlation analysis. The PLS method overcomes the problem of high collinearity, and identifying effective wavelength variables can help eliminate redundant spectral information, reduce model complexity and facilitate interpretation, and improve model accuracy.

[0063] Y=Xb+E

[0064] Where Y is the mean center vector of LCC, X is the mean center matrix of spectral reflectance, b is the matrix of coefficients, and E is the residual matrix.

[0065] In this paper, the coefficient of determination (R 2 ), Root Mean Square Error (RMSE) and Relative Predictive Deviation (RPD) are used to evaluate the performance of the prediction model. The RPD value is divided into three levels: RPD ≤ 1.4 (model without prediction ability); 1.4 ≤ RPD ≤ 2 (model with good prediction ability); 2 ≤ RPD (model with excellent prediction ability). The calculation formula is as follows:

[0066]

[0067] In the formula, y i represents the true value of chlorophyll, represents the predicted value of chlorophyll, represents the average value of chlorophyll in all band combinations, N represents the number of band combinations, and SD represents the standard deviation of the true value of chlorophyll in all band combinations.

[0068] TBI has the largest R at the 0.9-order derivative 2 is 0.853; MTCI has the largest R in scale 6 after continuous wavelet transform 2 It is 0.79.

[0069] S7. The prediction accuracy of different fractional derivatives is obtained by jointly inverting the multidimensional features of different fractional derivatives and the continuous wavelet transform feature set through the partial Gaussian process regression algorithm. The test set of the 5-fold cross-validation based on the Gaussian process regression of the optimal features of the combined fractional derivatives and continuous wavelet transform: R 2 The average value of the proposed method is 0.887, the RMSE is 3.283, and the RPD is 2.702.

[0070] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for inverting chlorophyll content, characterized in that: The following steps are involved: Obtain leaf spectrum data containing different bands; The leaf spectral data are processed by using fractional derivatives to obtain fractional derivative data that can enhance the characteristics of the spectral data; the leaf spectral data are processed by continuous wavelet transform using Gaussian second-order derivatives to obtain wavelet energy coefficients of multiple scales that can reflect the characteristics of the spectral data in multiple frequencies; According to the required band that can reflect the chlorophyll content, find the fractional derivative data and wavelet energy coefficient under the corresponding band; input the fractional derivative data corresponding to the three bands in the required band into the three-band vegetation index model to obtain the three-band vegetation index; input the selected wavelet energy coefficient into the vegetation index model corresponding to the required band to obtain the vegetation index; The correlation coefficient method and partial least squares method were used to conduct correlation analysis and preliminary feature screening of the three-band vegetation index and vegetation index respectively; the Gaussian process regression algorithm was used to conduct accuracy assessment of the screened three-band vegetation index and vegetation index respectively, and the three-band vegetation index model and vegetation index model that can reflect the chlorophyll content were determined according to the accuracy assessment results; According to the determined three-band vegetation index model and vegetation index model, the chlorophyll content is inverted and obtained.

2. A chlorophyll content inversion method as claimed in claim 1, characterized in that: The inversion to obtain the chlorophyll content specifically comprises the following steps: Using the leaf spectrum data to be tested, inverting the bands adapted to the three-band vegetation index model and the vegetation index model, wherein the inverted bands can reflect the chlorophyll content; The chlorophyll content is obtained by inverting the inferred bands.

3. The chlorophyll content inversion method according to claim 1, characterized in that: The method of processing the leaf spectral data by using fractional derivatives comprises the following specific steps: The leaf spectral data were smoothed by Savitzky-Golay filter algorithm to remove noise; The fractional derivative Grunwald-Letnikov transform was selected to transform the processed leaf spectral data to obtain fractional derivative data of 0.1 to 2.0 orders.

4. The chlorophyll content inversion method according to claim 1, characterized in that: The continuous wavelet transform processing of the leaf spectral data using the Gaussian second-order derivative specifically includes: The second-order derivative of the Gaussian function was selected as the basis function, and the leaf spectral data was processed by continuous wavelet transform of wavelet energy coefficients at 10 scales to obtain wavelet energy coefficients at 10 scales.

5. The chlorophyll content inversion method according to claim 1, characterized in that: The calculation formula of the three-band vegetation index TBI2 is: Among them, R λ1 and R λ2 are two wavelengths randomly selected from 400nm to 2400nm, R λ3 is a wavelength randomly selected from 500nm to 555nm, 630nm to 650nm, 690nm to 715nm and 735nm to 740nm, and R λ1 ≠R λ2 ≠R λ3 .

6. The chlorophyll content inversion method according to claim 1, characterized in that: The correlation coefficient method uses the Pearson correlation coefficient which can measure the degree of linear correlation between two variables.

7. A chlorophyll content inversion system, characterized in that: include: A data acquisition module, used to acquire leaf spectral data containing different bands; The data processing module is used to process the leaf spectral data using fractional derivatives to obtain fractional derivative data that can enhance the characteristics of the spectral data; to perform continuous wavelet transform processing on the leaf spectral data using Gaussian second-order derivatives to obtain wavelet energy coefficients of multiple scales that can reflect the characteristics of the spectral data in multiple frequencies; and to find the fractional derivative data and wavelet energy coefficients in the corresponding band according to the required band that can reflect the chlorophyll content; The model selection module is used to input the fractional derivative data corresponding to the three bands in the required band into the three-band vegetation index model to obtain the three-band vegetation index; input the selected wavelet energy coefficient into the vegetation index model corresponding to the required band to obtain the vegetation index; use the correlation coefficient method and the partial least squares method to perform correlation analysis and preliminary feature screening on the three-band vegetation index and vegetation index respectively; use the Gaussian process regression algorithm to perform accuracy assessment on the screened three-band vegetation index and vegetation index respectively, and determine the three-band vegetation index model and vegetation index model that can reflect the chlorophyll content according to the accuracy assessment results; The chlorophyll content inversion module is used to invert the chlorophyll content according to the determined three-band vegetation index model and vegetation index model.

8. A computer device comprising: Memory and processor; The memory stores a computer program, wherein the processor implements a chlorophyll content inversion method according to any one of claims 1 to 6 when executing the computer program.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, a chlorophyll content inversion method according to any one of claims 1 to 6 is implemented.

Images