Method for calculating the value of cultivated land ecosystem services based on multi-source remote sensing technology

Through multi-source remote sensing technology, the advantageous feature information of cultivated land plots is extracted and the crop yield correction model coupled to topographic factors is constructed, which solves the problem of low accuracy in accounting of the value of cultivated land resources ecosystem services in the existing technology, and realizes the value of cultivated land resources accurate assessment and scientific guidance for ecological protection.

CN118731935BActive Publication Date: 2024-11-15自然资源部第六地形测量队
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411214107.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-31
Publication Date
2024-11-15
Estimated Expiration
2044-08-31

AI Technical Summary

Technical Problem

The existing arable land resource ecosystem service value accounting methods cannot accurately reflect the spatial location differences between crop types and topography of a single arable land plot, resulting in inaccurate correction of equivalent factors, which in turn reduces the accuracy of arable land resource ecosystem service value accounting.

Method used

Using a method based on multi-source remote sensing technology, advantageous feature information is extracted through timing optical imaging, radar imaging and field sample data, a crop yield correction model coupled to topographic factors is constructed, and arable land resource ecosystem services are provided through the equivalent factor regional adaptive calculation model. Accounting of value.

Benefits of technology

It improves the accuracy of the value accounting of arable land resource ecosystem services, and can more accurately correct the impact of crop categories and topographic factors on yield, thereby achieving scientific and practical guidance on the "three-in-one" protection of arable land quantity, quality and ecology.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118731935B_ABST
    Figure CN118731935B_ABST
Patent Text Reader

Abstract

The present invention provides a method for calculating the service value of an ecosystem of cultivated land resources based on multi-source remote sensing technology, and belongs to the technical field of ecological value accounting of cultivated land resource assets. In order to solve the problem that the existing accounting methods cannot correct the influence of factors such as the types of cultivated land crops and topography, the present invention introduces time-series multi-source remote sensing data, extracts the types of crops and the planting area based on the differences in crop phenological characteristics, constructs a crop yield correction model that couples factors such as slope and cultivated land type, further calculates the equivalent factors of various ecological service functions through the corrected yield, and finally realizes the accurate calculation of the service value of the ecosystem of cultivated land resources, which has important scientific and practical guiding significance for realizing the "three-in-one" protection of cultivated land quantity, quality and ecology and the outgoing audit of natural resource assets.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of ecological value accounting of cultivated land resource assets, and in particular to a method for calculating the service value of cultivated land resource ecosystem based on multi-source remote sensing technology. Background Art

[0002] The accounting and assessment of the ecological value of cultivated land resource assets is the key to implementing the "three-in-one" protection of cultivated land quantity, quality and ecology, and is also an important part of the outgoing audit of natural resource assets. It comprehensively considers the comprehensive value of the supply, regulation, culture and support service functions of cultivated land resources as an ecosystem, and assesses the ecological value of cultivated land resource assets by calculating the cultivated land ecosystem service value, which can improve the scientificity and rationality of the assessment results.

[0003] The existing method for calculating the value of cultivated land resources ecosystem services is mainly based on statistical data such as cultivated land categories, cultivated land grades, and standard grain output to revise the equivalent factors for calculation. However, it is unable to reflect the spatial location differences such as the types of crops planted in individual cultivated land plots and topography, resulting in inaccurate revisions to the equivalent factors and low accuracy in the calculation of the value of cultivated land resources ecosystem services. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the present invention provides a method for calculating the value of cultivated land resources ecosystem services based on multi-source remote sensing technology, which solves the problem of low accuracy in calculating the value of cultivated land resources ecosystem services.

[0005] In order to achieve the above-mentioned purpose, in the first aspect, the present invention proposes a method for calculating the service value of cultivated land resources ecosystem based on multi-source remote sensing technology, comprising:

[0006] Step S1, extracting dominant feature information based on time-series optical images, radar images and field sample data, and extracting crop categories and planting areas based on the dominant feature information;

[0007] Step S2, constructing a crop yield correction model for coupled terrain factors based on slope and cultivated land category;

[0008] Step S3, comprehensively considering the crop categories, planting areas, and yields, and constructing a regional adaptive calculation model for equivalent factors;

[0009] Step S4: Calculate the ecosystem service value of cultivated land resources based on the regional adaptive calculation model of equivalent factors.

[0010] Furthermore, step S1 further comprises:

[0011] Step S11, obtaining band information in the optical image, polarization data of the radar image, and yield data of different crops in the field sample;

[0012] Step S12, calculating the Normalized Difference Vegetation Index (NDVI) based on the band data in the optical image;

[0013] Step S13, extracting texture feature information based on radar polarization data;

[0014] Step S14, calculating the importance scores of time series spectral band information, radar polarization data, normalized difference vegetation index and texture features in crop classification based on Gini impurity contribution;

[0015] Step S15, filtering out dominant feature information by weighted calculation of the importance score of each feature;

[0016] Step S16: extracting the crop category and planting area based on the dominant feature information.

[0017] Further, extracting texture feature information based on radar polarization data in step S13 includes: generating a gray level co-occurrence matrix based on radar polarization data, and extracting texture feature information based on the gray level co-occurrence matrix; generating the gray level co-occurrence matrix, by calculating the frequency of occurrence of selected pixel pairs in two windows, and obtaining the probability value of each cell of the entire image by sliding the window, and the calculation formula is: , where i is the row number, j is the column number, and V i,j is the value in cell (i, j) of the image window, P i,j is the probability value recorded for cell (i, j). The selected pixel pairs are determined by the analysis direction and distance selected by the center pixel of the window.

[0018] Based on the gray level co-occurrence matrix, 8 types of texture feature information including mean, variance, homogeneity, contrast, difference, entropy, second-order moment and correlation are extracted. The calculation formula is:

[0019]

[0020]

[0021]

[0022]

[0023]

[0024]

[0025]

[0026]

[0027] in, For homogeneity, is the contrast, For the difference, is entropy, is the second-order moment, For relevance, is the mean of the i-th row, is the mean of the jth column, is the variance of the i-th row, is the variance of the jth column, is the probability value recorded for cell (i, j), and N is the number of rows and columns in the image.

[0028] Furthermore, step S14 further includes: for the M surface categories in the sample set S, input all feature information extracted in steps S11-S13, calculate the Gini impurity at each node of the decision tree, find a value by searching to distinguish different categories and minimize the Gini impurity, and obtain an importance score according to the contribution of each feature to reducing the Gini impurity. The Gini impurity calculation formula is: , where Gini(S) represents the Gini impurity calculation formula, S is the data sample set, and p i is the proportion of the i-th surface category, and M is the number of surface categories.

[0029] Furthermore, step S15 further includes:

[0030] The phase and type proportion of the features are calculated based on the weighted statistical features based on the importance score of each feature. The calculation formula is:

[0031]

[0032]

[0033] Among them, Score i,j represents the importance score of the j-type feature in phase i, phase i It represents the percentage of the sum of the scores of all M features in phase i to the total score, type j It represents the percentage of the sum of the scores of the j-th feature in all N phases to the total score, where N is the number of phases and M is the number of features;

[0034] The proportions of time phases and types are added up in descending order to construct five time series feature combination schemes with a total proportion of 100%, more than 80%, more than 60%, more than 40%, and more than 20%.

[0035] Furthermore, step S16 further includes:

[0036] Based on five time series feature combination schemes, crop categories and planting areas were extracted through machine learning algorithms, and the overall accuracy OA and Kappa coefficient were used for accuracy analysis. The calculation formula is:

[0037]

[0038]

[0039] Among them, x ii is the number of pixels correctly classified into class i, n is the number of classes, N is the total number of validation samples, x +i is the number of pixels of the true i class in the validation sample, x i+ is the number of pixels in class i in the entire classification result; the solution with the highest OA and Kappa coefficients is selected, and the crop category and planting area are extracted.

[0040] Furthermore, in step S2, constructing a crop yield correction model of coupled terrain factors based on slope and cultivated land category also includes:

[0041] Step S21, grade the slope and assign an initial coefficient;

[0042] Step S22, classifying the cultivated land and assigning initial coefficients;

[0043] Step S23: construct a crop yield correction model.

[0044] Furthermore, in step S21, the slope classification and initial coefficient assignment include: dividing the cultivated land slope i into 5 levels: ≤2°, 2°-6°, 6°-15°, 15°-25°, and >25°, and assigning the initial coefficients according to the gradient as: 1, 0.8, 0.6, 0.4, and 0.2.

[0045] Furthermore, step S22 further includes: dividing the cultivated land category j into three categories: dry land, paddy field, and irrigated land, and assigning initial coefficients to 1, 1, and 1 according to the category.

[0046] Furthermore, step S23 further includes:

[0047] Comprehensively considering the influencing factors such as standard grain yield of different crops, slope and type of cultivated land, a functional relationship is established and regression analysis is performed. The calculation formula is:

[0048]

[0049] Among them, y m,i.j is the measured average yield per mu of the mth crop in plot (i, j), α i0 is the initial slope coefficient of plot (i, j), β j0 is the initial category coefficient of plot (i, j), is the average yield per mu of the mth crop in the statistical data, ε is the random error, and regression analysis is carried out based on the actual yield data of the mth crop in N groups of field samples to obtain the final slope coefficient α of the mth crop in plot (i, j). i and the category coefficient β j , forming a crop yield correction model coupled with terrain factors, the calculation formula is:

[0050]

[0051] Among them, Y m,i.j is the average yield per mu of the mth crop in plot (i, j) after correction, α i is the corrected slope coefficient of plot (i, j), β j is the corrected category coefficient of plot (i, j), is the average yield per mu of the mth crop in the statistical data.

[0052] Furthermore, the formula for calculating the equivalent factor of the m-th crop area adaptation in step S3 is as follows:

[0053]

[0054] Among them, f m,n is the calculated equivalent factor of the value of the nth ecological service function per unit area of ​​cultivated land for the mth crop, n=1, 2, …, 9, representing the 9 ecosystem service functions of gas regulation, climate regulation, water conservation, soil formation and protection, waste treatment, biodiversity protection, food production, raw materials and entertainment culture, respectively. Y m,i.j is the actual per-acre yield of the mth crop in plot (i, j), A m,i.j is the area of ​​plot (i,j), is the average yield per mu of the mth crop in the statistical data, F n It is the equivalent factor of the ecosystem service value per unit area of ​​cultivated land in the country.

[0055] Furthermore, step S4 further includes:

[0056] The 9 ecosystem service function value equivalent factors of each crop within the scope of cultivated land are summarized and counted, and the ecosystem service value of cultivated land resources is calculated by combining the average price of different crops. The formula for calculating the ecosystem service value of cultivated land resources is:

[0057]

[0058] Among them, ESA is the ecosystem service value of cultivated land resources, 1 / 7 refers to the conversion coefficient of the economic value provided by the natural environment without human input, and f m,nis the calculated value equivalent factor of the nth ecological service function per unit area of ​​cultivated land for the mth crop, A m is the planting area of ​​the mth crop, p m is the average price of the mth crop, m=1, 2, …, M, representing M different types of crops; n=1, 2, …, 9 represents 9 ecosystem service functions, namely gas regulation, climate regulation, water conservation, soil formation and protection, waste treatment, biodiversity protection, food production, raw materials and entertainment culture.

[0059] In order to solve the problem that the existing accounting methods cannot correct the influence of factors such as crop types and topography of cultivated land, the present invention introduces time-series multi-source remote sensing data, extracts crop types and planting areas based on differences in crop phenological characteristics, and constructs a crop yield correction model that couples factors such as slope and cultivated land types. The equivalent factors of various ecological service functions are further calculated through the corrected yield, and ultimately the accurate accounting of the ecosystem service value of cultivated land resources is achieved, which has important scientific and practical guiding significance for the realization of the "trinity" protection of cultivated land quantity, quality, and ecology and the outgoing audit of natural resource assets. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0061] Figure 1 It is a flow chart of a method for calculating the service value of cultivated land resources ecosystem based on multi-source remote sensing technology in the present invention.

[0062] Figure 2 This is the national cultivated land per unit area ecosystem service value equivalent factor diagram in the present invention. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the solutions in the embodiments of the present invention are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0064] It should be noted that the embodiments of the present invention and the features therein may be combined with each other as long as there is no conflict between the combinations.

[0065] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.

[0066] like Figure 1As shown in, in an embodiment of the present invention, the present invention discloses a method for calculating the service value of cultivated land resources ecosystem based on multi-source remote sensing technology, comprising the following steps:

[0067] Step S1, extracting dominant feature information based on time-series optical images, radar images and field sample data, and extracting crop categories and planting areas based on the dominant feature information;

[0068] Step S2, constructing a crop yield correction model for coupled terrain factors based on slope and cultivated land category;

[0069] Step S3, comprehensively considering the crop categories, planting areas, and yields, and constructing a regional adaptive calculation model for equivalent factors;

[0070] Step S4: Calculate the ecosystem service value of cultivated land resources based on the regional adaptive calculation model of equivalent factors.

[0071] Furthermore, step S1 also includes:

[0072] Step S11, obtaining band information in the optical image, polarization data of the radar image, and yield data of different crops in the field sample;

[0073] Step S12, calculating the Normalized Difference Vegetation Index NDVI based on the band data in the optical image;

[0074] Step S13, extracting texture feature information based on radar polarization data;

[0075] Step S14, calculating the importance scores of time series spectral band information, radar polarization data, normalized difference vegetation index and texture features in crop classification based on Gini impurity contribution;

[0076] Step S15, filtering out dominant feature information by weighted calculation of the importance score of each feature;

[0077] Step S16: extracting the crop category and planting area based on the dominant feature information.

[0078] In this embodiment, the commonly used band information of the optical image includes a red band, a green band, a blue band and a near-infrared band. The above band information can be used to calculate the normalized vegetation index and other required feature information.

[0079] Furthermore, the NDVI calculation formula in step S12 is:

[0080]

[0081] Among them, Band NIR is the pixel value of the near infrared band, Band R is the pixel value of the red band.

[0082] Further, extracting texture feature information based on radar polarization data in step S13 includes: generating a gray level co-occurrence matrix based on the radar polarization data, and extracting texture feature information based on the gray level co-occurrence matrix;

[0083] The gray-level co-occurrence matrix is ​​generated by calculating the frequency of occurrence of the selected pixel pairs in the two windows and sliding the windows to obtain the probability value of each cell in the entire image. The calculation formula is:

[0084]

[0085] Where i is the row number, j is the column number, and V i,j is the value in cell (i, j) of the image window, P i,j is the probability value recorded for the cell (i, j). In the embodiment of the present invention, the sliding window size constructed is 3×3, and the calculation is performed at a scanning angle of 45°, that is, in this embodiment, the selected pixel pairs are composed of pixels selected by the central window along the 45° direction with a window distance of 3.

[0086] Based on the generated gray-level co-occurrence matrix, eight texture features, including mean, variance, homogeneity, contrast, difference, entropy, second-order moment and correlation, are extracted. The texture feature calculation formula of cell (i, j) is:

[0087]

[0088]

[0089]

[0090]

[0091]

[0092]

[0093]

[0094]

[0095] in, For homogeneity, is the contrast, For the difference, is entropy, is the second-order moment, For relevance, is the mean of the i-th row, is the mean of the jth column, is the variance of the i-th row, is the variance of the jth column, is the probability value recorded for cell (i, j), and N is the number of row and column numbers in the image obtained by the gray-level co-occurrence matrix.

[0096] Furthermore, in step S4, the importance scores of time series spectral band information, radar polarization data, normalized vegetation index and texture features in crop classification are calculated based on the contribution of Gini impurity. For the M surface categories in the sample set S, all feature information extracted in steps S11-S13 is input. At each node of the decision tree, the Gini impurity is calculated, and a value is found by searching to distinguish different categories and minimize the Gini impurity. The importance score is obtained according to the contribution of each feature to reducing the Gini impurity. The Gini impurity calculation formula is: , where Gini(S) represents the Gini impurity calculation formula, S is the data sample set, and p i is the ratio of the i-th surface category, and M is the number of surface categories. In the embodiment of the present invention, in order to control the sample capacity, the number of input field sample points for each surface category in the sample set S is 100.

[0097] Furthermore, in step S15, the dominant feature information is screened by weighted calculation of the importance score of each feature, and the phase and type proportion of the feature are weighted and statistically calculated based on the importance score of each feature. The calculation formula is:

[0098]

[0099]

[0100] Score i,j represents the importance score of the j-type feature in phase i, phase i It represents the percentage of the sum of the scores of all M features in phase i to the total score, type j It represents the percentage of the sum of the scores of the j-th feature in all N phases;

[0101] The proportions of time phases and types are added up in descending order to construct five time series feature combination schemes with a total proportion of 100%, more than 80%, more than 60%, more than 40%, and more than 20%.

[0102] Furthermore, in step S16, crop categories and planting areas are extracted based on the dominant feature information. Based on the five time series feature combination schemes constructed in step S15, crop categories and planting areas are extracted through machine learning algorithms, and precision analysis is performed using overall accuracy OA and Kappa coefficient. The calculation formula is:

[0103]

[0104]

[0105] Among them, x ii is the number of pixels correctly classified into class i, N is the total number of validation samples, x +i is the number of pixels of the true i class in the validation sample, x i+ is the number of pixels of class i in the entire classification result.

[0106] In the embodiment of the present invention, 30 field sample points are used for verification of each type of land surface category, the solution with the highest OA and Kappa coefficient is selected, and the crop category and planting area are extracted.

[0107] When extracting crop categories, this embodiment takes into account the phenological characteristics of crops and selects time-series multi-source remote sensing images. On the other hand, it takes into account the efficiency of extraction and the accuracy of the results. Feature information that contributes greatly to crop identification is selected through feature optimization. This can not only achieve accurate extraction of crop types and planting areas, but also reduce the computational complexity of the extraction process and improve efficiency.

[0108] Furthermore, in step S2, constructing a crop yield correction model of coupled terrain factors based on slope and cultivated land category also includes:

[0109] Step S21, grade the slope and assign an initial coefficient;

[0110] Step S22, classifying the cultivated land and assigning initial coefficients;

[0111] Step S23: construct a crop yield correction model.

[0112] Furthermore, step S21 further includes: dividing the cultivated land slope i into 5 levels: ≤2°, 2°-6°, 6°-15°, 15°-25°, and >25°, and assigning the initial coefficients according to the gradient as: 1, 0.8, 0.6, 0.4, and 0.2.

[0113] Furthermore, in step S22, the cultivated land is classified and assigned initial coefficients, and the cultivated land category j is divided into three categories: dry land, paddy field, and irrigated land, and the initial coefficients are assigned as 1, 1, and 1 according to the category.

[0114] Furthermore, in step S23, the crop yield correction model is constructed by integrating the standard grain yield of different crops, the slope and type of cultivated land and other influencing factors, establishing a functional relationship and performing regression analysis. The calculation formula is:

[0115]

[0116] Among them, y m,i.j is the measured average yield per mu of the mth crop in plot (i, j), α i0 is the initial slope coefficient of plot (i, j), βj0 is the initial category coefficient of plot (i, j), is the average yield per mu of the mth crop in the statistical data, ε is the random error, and regression analysis is carried out based on the actual yield data of the mth crop in N groups of field samples to obtain the final slope coefficient α of the mth crop in plot (i, j). i and the category coefficient β j , forming a crop yield correction model coupled with terrain factors, the calculation formula is:

[0117]

[0118] Among them, Y m,i.j is the average yield per mu of the mth crop in plot (i, j) after correction, α i is the corrected slope coefficient of plot (i, j), β j is the corrected category coefficient of plot (i, j), is the average yield per mu of the mth crop in the statistical data.

[0119] When estimating crop yields, the present invention takes into account the differences in growth potential of crops in different cultivated land slopes and categories, incorporates the slope coefficient and category coefficient into regression analysis, and realizes reasonable correction of crop yields within a regional range.

[0120] Furthermore, in step S3, the crop category, planting area, and yield are integrated to construct an equivalent factor regional adaptive calculation model. The equivalent factor adaptive calculation formula for the mth crop in the region is as follows:

[0121]

[0122] Among them, f m,n is the calculated equivalent factor of the value of the nth ecological service function per unit area of ​​cultivated land for the mth crop, where n=1, 2, …, 9 represent the 9 ecosystem service functions of gas regulation, climate regulation, water conservation, soil formation and protection, waste treatment, biodiversity protection, food production, raw materials, and entertainment culture, respectively. Y m,i.j is the actual per-acre yield of the mth crop in plot (i, j), A m,i.j is the area of ​​plot (i,j), is the average yield per mu of the mth crop in the statistical data, F n The ecosystem service value equivalent factor per unit area of ​​cultivated land in China (e.g. Figure 2As shown in Figure 1). In this field, ecological service functions include material supply services, regulation services, support services, cultural services, biological control services, habitat services, genetic resource services, information services and aesthetic services. In this example, nine ecosystem service functions, including gas regulation, climate regulation, water conservation, soil formation and protection, waste treatment, biodiversity protection, food production, raw materials and entertainment culture, were selected to participate in the calculation of ecological service value.

[0123] When correcting the equivalent factor of the value of cultivated land resource ecosystem services, the present invention takes into account the types of crops planted on the cultivated land, the planting area and the yield differences caused by terrain differences, corrects the errors caused by calculating the equivalent factor based on statistical data, and realizes regional adaptive calculation and correction of the equivalent factor of the value of cultivated land resource ecosystem services.

[0124] Furthermore, in step S4, based on the regional adaptive calculation model of equivalent factors, the value of cultivated land resources ecosystem services is calculated, and the equivalent factors of the nine ecosystem service function values ​​of various crops within the cultivated land range are summarized and counted. Combined with the average price of different crops, the value of cultivated land resources ecosystem services is calculated. The formula for calculating the value of cultivated land resources ecosystem services is:

[0125]

[0126] Among them, ESA is the ecosystem service value of cultivated land resources, 1 / 7 refers to the conversion coefficient of the economic value provided by the natural environment without human input, and f m,n is the calculated value equivalent factor of the nth ecological service function per unit area of ​​cultivated land for the mth crop, A m is the planting area of ​​the mth crop, p m is the average price of the mth crop, m=1, 2, …M, representing M different crop categories, and n=1, 2, …, 9, representing 9 ecosystem service functions: gas regulation, climate regulation, water conservation, soil formation and protection, waste treatment, biodiversity protection, food production, raw materials, and entertainment culture.

[0127] In order to solve the problem that the existing accounting methods for the value of cultivated land resource ecosystem services cannot correct the impact of differences in spatial information such as crop planting categories and cultivated land topography on the accounting results, the present invention introduces multi-source remote sensing images such as time-series optics and radar, considers changes in crop phenological characteristics, and realizes the extraction of crop categories and planting areas based on time-series optimization features; considers spatial and attribute information such as the slope and category of the cultivated land, constructs a crop yield correction model coupled with terrain factors, and further corrects the equivalent factor based on the corrected yield to realize the accounting of the value of cultivated land resource ecosystem services, thereby improving the precision and accuracy of the accounting results.

[0128] The main features and advantages of the present invention are shown and described above. It is obvious to those skilled in the art that the present invention is limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes that fall within the meaning and scope of the equivalent elements of the claims are included in the present invention.

[0129] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in the embodiments may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A method for calculating the value of cultivated land ecosystem services based on multi-source remote sensing technology. It is characterized in that include: Step S1, extracting dominant feature information based on time-series optical images, radar images and field sample data, and extracting crop categories and planting areas based on the dominant feature information; The step S1 further comprises: Step S11, obtaining band information in the optical image, polarization data of the radar image, and yield data of different crops in the field sample; Step S12, calculating the Normalized Difference Vegetation Index NDVI based on the band data in the optical image; Step S13, extracting texture feature information based on radar polarization data; Step S14, calculating the importance scores of time series spectral band information, radar polarization data, normalized difference vegetation index and texture features in crop classification based on Gini impurity contribution; Step S15, filtering out dominant feature information by weighted calculation of the importance score of each feature; Step S16, extracting the crop category and planting area based on the dominant feature information; Step S2, constructing a crop yield correction model of coupled terrain factors based on slope and cultivated land category; said step S2 further includes: Step S21, grading the slope and assigning an initial coefficient; the step S21 further includes: grading the cultivated land slope i into 5 levels: ≤2°, 2°-6°, 6°-15°, 15°-25°, >25°, and assigning the initial coefficients as 1, 0.8, 0.6, 0.4, and 0.2 according to the gradient; Step S22, classifying the cultivated land and assigning initial coefficients; the step S22 further includes: classifying the cultivated land category j into three categories: dry land, paddy field, and irrigated land, and assigning initial coefficients according to the categories as: 1, 1, 1; Step S23, constructing a crop yield correction model; the step S23 further includes: Comprehensively considering the influencing factors such as standard grain yield of different crops, slope and type of cultivated land, a functional relationship is established and regression analysis is performed. The calculation formula is: Among them, y m,i.j is the measured average yield per mu of the mth crop in plot (i, j), α i0 is the initial slope coefficient of plot (i, j), β j0 is the initial category coefficient of plot (i, j), is the average yield per mu of the mth crop in the statistical data, is a random error. Based on the actual yield data of the mth crop in N groups of field samples, a regression analysis is carried out to obtain the final slope coefficient α of the mth crop in plot (i, j). i and the category coefficient β j , forming a crop yield correction model coupled with terrain factors, the calculation formula is: Among them, Y m,i.j is the average yield per mu of the mth crop in plot (i, j) after correction, α i is the corrected slope coefficient of plot (i, j), β j is the corrected category coefficient of plot (i, j), is the average yield per mu of the mth crop in the statistical data; Step S3, comprehensively considering the crop categories, planting areas, and yields, and constructing a regional adaptive calculation model for equivalent factors; Step S4: Calculate the ecosystem service value of cultivated land resources based on the regional adaptive calculation model of equivalent factors.

2. According to claim 1, a method for calculating the service value of cultivated land resources ecosystem based on multi-source remote sensing technology, It is characterized in that The step S13 further comprises: A gray-level co-occurrence matrix is ​​generated based on radar polarization data, and texture feature information is extracted based on the gray-level co-occurrence matrix; the gray-level co-occurrence matrix is ​​generated by calculating the frequency of occurrence of pixel pairs selected in two windows, where the selected pixel pairs are determined by the analysis direction and distance selected by the central pixel of the window, and then the probability value of each cell of the entire image is obtained by sliding the window. The calculation formula is: , where i is the row number, j is the column number, and V i,j is the value in cell (i, j) of the image window, P i,j is the probability value recorded for cell (i, j), where N is the number of rows and columns in the image; Based on the gray level co-occurrence matrix, 8 types of texture feature information including mean, variance, homogeneity, contrast, difference, entropy, second-order moment and correlation are extracted. The specific calculation formula is as follows: in, For homogeneity, is the contrast, For difference, is entropy, is the second-order moment, For relevance, is the mean of the i-th row, is the mean of the jth column, is the variance of the i-th row, is the variance of the jth column, is the probability value recorded for cell (i, j), and N is the number of rows and columns in the image.

3. According to claim 1, a method for calculating the service value of cultivated land resources ecosystem based on multi-source remote sensing technology, It is characterized in that The step S14 further comprises: For the M surface categories in the sample set S, all feature information extracted in steps S11-S13 is input. At each node of the decision tree, the Gini impurity is calculated. A value is searched to distinguish different categories and minimize the Gini impurity. The importance score is obtained according to the contribution of each feature to reducing the Gini impurity. The Gini impurity calculation formula is: , where Gini(S) represents the Gini impurity calculation formula, S is the data sample set, and p i is the proportion of the i-th surface category, and M is the number of surface categories.

4. According to claim 1, a method for calculating the service value of cultivated land resources ecosystem based on multi-source remote sensing technology, It is characterized in that The step S15 further comprises: The phase and type proportion of the features are calculated based on the weighted statistical features based on the importance score of each feature. The calculation formula is: Among them, Score i,j represents the importance score of the j-type feature in phase i, phase i It represents the percentage of the sum of the scores of all M features in phase i to the total score, type j It represents the percentage of the sum of the scores of the j-th feature in all N phases to the total score, where N is the number of phases and M is the number of features; The proportions of time phases and types are added up in descending order to construct five time series feature combination schemes with a total proportion of 100%, more than 80%, more than 60%, more than 40%, and more than 20%.

5. According to claim 1, a method for calculating the service value of cultivated land resources ecosystem based on multi-source remote sensing technology, It is characterized in that The step S16 further comprises: Based on five time series feature combination schemes, crop categories and planting areas were extracted through machine learning algorithms, and the overall accuracy OA and Kappa coefficient were used for accuracy analysis. The calculation formula is: Among them, x ii is the number of pixels correctly classified into class i, n is the number of classes, N is the total number of validation samples, x +i is the number of pixels of the true i class in the validation sample, x i+ is the number of pixels in class i in the entire classification result; the solution with the highest OA and Kappa coefficients is selected, and the crop category and planting area are extracted.

6. According to claim 1, a method for calculating the service value of cultivated land resources ecosystem based on multi-source remote sensing technology, It is characterized in that The formula for calculating the regional adaptation equivalent factor of the mth crop in step S3 is as follows: Among them, f m,n is the calculated value equivalent factor of the nth ecological service function per unit area of ​​cultivated land of the mth crop, n=1, 2, …, 9, representing the 9 categories of ecosystem service functions, respectively, Y m,i.j is the actual per-acre yield of the mth crop in plot (i, j), A m,i.j is the area of ​​plot (i,j), is the average yield per mu of the mth crop in the statistical data, F n It is the equivalent factor of the ecosystem service value per unit area of ​​cultivated land in the country.

7. According to claim 1, a method for calculating the service value of cultivated land resources ecosystem based on multi-source remote sensing technology, It is characterized in that The step S4 further comprises: The 9 ecosystem service function value equivalent factors of each crop within the scope of cultivated land are summarized and counted, and the ecosystem service value of cultivated land resources is calculated by combining the average price of different crops. The formula for calculating the ecosystem service value of cultivated land resources is: Among them, ESA is the ecosystem service value of cultivated land resources, 1 / 7 refers to the conversion coefficient of the economic value provided by the natural environment without human input, and f m,n is the calculated value equivalent factor of the nth ecological service function per unit area of ​​cultivated land for the mth crop, A m is the planting area of ​​the mth crop, p m is the average price of the mth crop, m=1, 2, …, M, representing M different types of crops; n=1, 2, …, 9, representing 9 ecosystem service functions: gas regulation, climate regulation, water conservation, soil formation and protection, waste treatment, biodiversity protection, food production, raw materials and entertainment culture.

Citation Information

Cited By

  • Land ecological condition monitoring system and method based on remote sensing data

    CN120894702A

  • Land ecological condition monitoring system and method based on remote sensing data

    CN120894702B