A method and system for extrapolating multi-population crop areas by spatial sampling

Abstract

This application belongs to the field of agricultural remote sensing technology, and specifically relates to a method and system for spatially sampling and extrapolating the crop areas of multiple populations. The method includes the following steps: obtaining the spatial distribution vector map of the total cultivated land, the information table of sampling unit layers, and the statistical table of layer information; making the layer population distribution map; forming the list of survey sample sampling units; forming the sample spatial distribution vector map; obtaining remote sensing images; performing target recognition on the remote sensing images, extracting the target crop areas according to the sampling units, and forming sample observations; setting the directory table of sub-population names; generating the intermediate file of the sub-population sample observation value set; assigning a hierarchical code to each sampling unit; by circularly calling the relevant parameters in the information statistical tables of each sub-population layer and the sample observations, circularly calculating the crop areas of each sub-population, and finally forming the statistical table of the crop areas of all populations. The method and system of this application can quickly and conveniently perform spatial extrapolation of the crop areas of multiple populations at one time to obtain the estimated values of each population.

Description

Technical Field

[0001] The present application belongs to the field of agricultural remote sensing technology, and specifically relates to a method and system for spatial sampling and extrapolation of multi-population crop areas. Background Art

[0002] The agricultural application of remote sensing technology (RS) and geographic information system (GIS) technology has caused a profound revolution in the field of agricultural monitoring. As one of the important contents of agricultural monitoring, the survey of crop planting area is the first to be affected. Compared with the traditional crop planting area statistics reported and summarized by government departments step by step, the crop planting area survey using remote sensing (RS) and geographic information system (GIS) technology has great advantages such as objectivity, economy, speed and comprehensiveness.

[0003] The annual remote sensing survey of crop area for business operation usually adopts the sampling survey method. This method effectively avoids the problem that it is difficult to obtain complete remote sensing image data in a planting season or annual full coverage survey. It is cheaper and more efficient than a full coverage survey, and the results are more accurate and reliable.

[0004] However, the traditional sampling survey method cannot establish a crop area remote sensing sampling monitoring model cluster for multiple survey populations at one time, and when the sampling basic data changes, the traditional sampling survey method is not only costly but also inefficient to modify the model parameters. Therefore, the population (crop area) estimation method designed based on the sampling survey method faces the same problem, that is, it is impossible to quickly and conveniently perform spatial extrapolation of the crop area of multiple populations at one time to obtain the estimated value of each population; in addition, when it is necessary to calculate the inter-annual change rate of the crop area of multiple populations, the existing methods also have the problems of high cost and low efficiency. Summary of the invention

[0005] In order to solve at least one technical problem existing in the prior art, the present application provides a method and system for spatial sampling and extrapolation of multiple crop areas.

[0006] In a first aspect, the present application discloses a method for spatial sampling and extrapolating multi-population crop areas, comprising the following steps:

[0007] Step 101, obtaining a total cultivated land spatial distribution vector diagram of the monitoring target population, and obtaining a sampling unit layer information table and a layer information statistical table of a plurality of sub-populations contained in the population, wherein the sampling unit layer information table at least includes a sampling unit code and a layer field value, and in addition, in the step of obtaining the sampling unit layer information table and the layer information statistical table of each sub-population, a plurality of sampling units are set, and each of the sampling units has a unique identification code;

[0008] Step 102. Within the scope of the total cultivated land spatial distribution vector map, based on the sampling unit layer information table of each sub-population, create a layer population distribution map of the sub-population stratified by cultivated land area;

[0009] Step 103. According to the layer information statistical table of each sub-population, randomly select a sufficient number of sampling units for each layer on the layer population distribution map of the sub-population, record the identification codes of the selected sampling units, and form a survey sample sampling unit list of each sub-population;

[0010] Step 104. Based on the survey sample sampling unit list, obtain the geographical boundaries of each sampling unit to form a sample spatial distribution vector map of each sub-population;

[0011] Step 105. Obtain a remote sensing image, where the geographical location and quantity of the remote sensing image are adapted to the sample spatial distribution vector map;

[0012] Step 106. Perform target recognition on the remote sensing image, extract the target crop area according to the sampling unit, and form sample observation values. Among them, one sub-population corresponds to a set of sample observation values, and the set of sample observation values of all sub-populations forms a sample population file, and the file is named according to a predetermined naming method;

[0013] Step 107. Set a sub-population name directory table, where each sub-population corresponds to a record, and the field values of each record at least include the full name, short name, and code of the sub-population;

[0014] Step 108. Loop through and read the field values of each sub-population in the sub-population name directory table. Based on the read field values, extract the sub-population sample observation value set from the sample population file, and dynamically generate the sub-population sample observation value set;

[0015] Step 109. Call the sampling unit code and layer field values in the sampling unit layer information table of each sub-population, and assign a hierarchical code to each sampling unit in the sub-population sample observation value set;

[0016] Step 110. By looping through and calling the relevant parameters in the layer information statistical table of each sub-population and the sample observation values of each sub-population into the following formula (1), calculate the crop area of each sub-population in a loop, and finally form a statistical table of the crop area of all populations:

[0017]

[0018] Among them, L is the number of layers of stratified sampling; N h is the total number of each sampling layer population; h = 1, 2,..., L; n h is the sample size of each sampling layer; y hi is the observation value of the i-th sampling unit in the h-th layer; is the sub-population estimated value, i.e., the crop area.

[0019] In an alternative embodiment, step 101 includes:

[0020] Step 1011, obtain the total cultivated land spatial distribution vector map of the monitored target population and the sub-population boundary vector maps of multiple sub-populations;

[0021] Step 1012, set sampling units and obtain the sampling unit layer vector map of the sampling units;

[0022] Step 1013, obtain the sub-cultivated land spatial distribution vector maps and sub-sampling unit layer vector maps of each sub-population;

[0023] Step 1014, overlay each sub-cultivated land spatial distribution vector map with the corresponding sub-sampling unit layer vector map, extract the cultivated land areas of each sub-population in units of sampling unit individuals, and use the cultivated land area data of each sub-population as the stratified basic data of each sub-population and store it in a file directory in a predetermined format;

[0024] Step 1015, set a sub-population name directory table, where each sub-population corresponds to a record, and the field values of each record at least include the full name, short name, and code of the sub-population;

[0025] Step 1016, loop to read the corresponding field values of each sub-population in the sub-population name directory table, dynamically construct the file names of the stratified basic data of each sub-population, thereby read the stratified basic data of the corresponding sub-populations, and sort the data in ascending order;

[0026] Step 1017, stratify the stratified basic data of each sub-population after ascending sorting and display the stratification result in tabular form to obtain the information table of the sampling unit layer of each sub-population, where the stratification is based on the size of the cultivated land area, the stratification method is the cumulative frequency method, and a preset grouping step size and number of levels are set. After stratification, all sampling unit individuals of each sub-population are assigned corresponding level codes;

[0027] Step 1018, loop to read the information table of the sampling unit layer of each sub-population, calculate the layer statistical values of each sub-population, thereby obtain the information statistical table of the layer of each sub-population;

[0028] Step 1019, judge whether the stratification result meets the requirements through the predetermined parameters in the information statistical table of the layer of each sub-population. If not, return to step to modify the preset grouping step size and number of levels until it meets the requirements, where the predetermined parameters at least include the total number of the layer population, the minimum sample size, and the sampling ratio.

[0029] In an alternative embodiment, in step 1013, the overall cultivated land spatial distribution vector map and the sampling unit layer vector map are respectively cut by the sub-population boundary vector maps of the multiple sub-populations on the GIS platform, so as to form the sub-cultivated land spatial distribution vector maps and sub-sampling unit layer vector maps of each sub-population.

[0030] In an alternative embodiment, in step 1014, the cultivated land areas of each sub-population with individual sampling units as the unit are two-dimensional data table sets stored in each sub-population.

[0031] In an alternative embodiment, in step 1017, the field values of the headers of each sub-population sampling unit layer information table at least include sub-population code, sampling unit code, cultivated land area, frequency, cumulative frequency, square root of cumulative frequency, and layer type.

[0032] In an alternative embodiment, in step 1019, the minimum sample size n is calculated by the following formula:

[0033]

[0034] where L is the number of layers of stratified sampling; N h is the total number of each sampling layer; h = 1, 2,..., L; is the variance of each sampling layer; N is the total population; W h = N h / N, that is, the sampling weight of the h-th layer; V is the variance of the estimator. If V is not given, but the error bound d is given, then V = (d / t) 2 , when the sample size is quite large, it can be assumed that t = u 0.025 = 1.96; d = (1 - δ)Y, where δ is the sampling precision and Y is the total population value;

[0035] The sampling ratio r is calculated by the following formula:

[0036] r = n / N;

[0037] Finally, the minimum sample size n of each layer h is calculated by the following formula:

[0038] n h = N h ×r.

[0039] In an alternative embodiment, in step 103, the number of sampling units used to extrapolate the area of each layer of each sub-population is equal to the actual sample size of each layer, and the actual sample size of each layer should not be less than the minimum sample size of each layer in the sub-population layer information statistical table.

[0040] In an alternative embodiment, in step 105, the obtained remote sensing image completely covers the geographical space of the sampling units shown in the spatial distribution vector map of each sub-population sample space.

[0041] Second, the present application also discloses a system for extrapolating the crop areas of multiple populations by spatial sampling, including:

[0042] An acquisition module, configured to acquire a total cultivated land spatial distribution vector map of a monitoring target population, and acquire a sampling unit layer information table and a layer information statistical table of multiple sub-populations included in the population, wherein the sampling unit layer information table at least includes a sampling unit code and a layer field value; and

[0043] configured to acquire a remote sensing image, the geographical location and quantity of which are adapted to the sample space distribution vector map;

[0044] A basic data processing module, configured to, within the range of the total cultivated land spatial distribution vector map, make a layer population distribution map of the sub-populations based on the stratification of cultivated land areas according to the sampling unit layer information tables of the respective sub-populations; and

[0045] configured to randomly select a sufficient number of sampling units of each layer on the layer population distribution map according to the layer information statistical tables of the respective sub-populations, record the identification codes of the selected sampling units, and form a survey sample sampling unit list; and

[0046] configured to, according to the sampling unit list, acquire the geographical boundaries of each selected sampling unit to form a sample space distribution vector map; and

[0047] configured to perform target recognition on the remote sensing image, extract the target crop areas according to the sampling units to form sample observation values, where one sub-population corresponds to a set of sample observation values, and the set of sample observation values of all sub-populations forms a sample population file, and the file is named according to a predetermined naming method; and

[0048] configured to set a sub-population name directory table, where each sub-population corresponds to a record, and the field values of each record at least include the full name, short name, and code of the sub-population;

[0049] An extrapolation module, configured to cyclically read the field values of each sub-population to dynamically generate a set of sub-population sample observation values; and

[0050] configured to call the sampling unit codes and layer field values in the sampling unit layer information tables of the respective sub-populations to assign a hierarchical code of each sampling unit to the set of sub-population sample observation values; and

[0051] By cyclically invoking relevant parameters in the information statistical tables of each sub-population layer and the observed values of each sub-population sample into the following formula (1), the crop area of each sub-population is cyclically calculated, and finally a statistical table of the total crop area of the entire population is formed:

[0052]

[0053] Among them, L is the number of layers of stratified sampling; N h is the total number of each sampling layer; h = 1, 2,..., L; n h is the sample size of each sampling layer; y hi is the observed value of the i-th sampling unit in the h-th layer; is the sub-population estimated value, that is, the crop area.

[0054] In an alternative embodiment, the acquisition module is further configured to:

[0055] Obtain the total cultivated land space distribution vector map of the monitored target population and the sub-population boundary vector maps of multiple sub-populations; and

[0056] Set the sampling unit and obtain the sampling unit layer vector map of the sampling unit; and

[0057] Obtain the sub-cultivated land space distribution vector map and sub-sampling unit layer vector map of each sub-population; and

[0058] Overlay each sub-cultivated land space distribution vector map with the corresponding sub-sampling unit layer vector map, and extract the cultivated land area of each sub-population in units of sampling unit individuals;

[0059] The basic data processing module is further configured to:

[0060] Use the cultivated land area data of each sub-population as the stratified basic data of each sub-population and store it in a file directory in a predetermined format; and

[0061] Set a sub-population name directory table, where each sub-population corresponds to a record, and the field values of each record at least include the full name, short name, and code of the sub-population; and

[0062] Cyclically read the corresponding field values of each sub-population in the sub-population name directory table, dynamically construct the file names of the stratified basic data of each sub-population, so as to read the stratified basic data of the corresponding sub-populations and sort the data in ascending order;

[0063] The system for spatially sampling and extrapolating the crop area of multiple populations further includes:

[0064] A stratification processing module, configured to stratify the stratified basic data of each sub-population after ascending sorting and display the stratification result in a table form to obtain an information table of each sub-population sampling unit layer; and

[0065] For circularly reading the information tables of sampling units at each sub - population layer, calculating the layer statistical values of each sub - population, so as to obtain the information statistical tables of each sub - population layer; and

[0066] For judging whether the stratification result meets the requirements through the predetermined parameters in the information statistical tables of each sub - population layer. If not, return to modify the preset grouping step size and the number of levels, and then execute the subsequent steps until it meets the requirements, where the predetermined parameters at least include the total number of the layer population, the minimum sample size, and the sampling ratio.

[0067] The present application has at least the following beneficial technical effects:

[0068] 1) In the method and system for spatially sampling and extrapolating the crop areas of multiple populations of the present application, one sub - population corresponds to a set of sample observation values, and a directory table of sub - population names is set. Finally, by circularly calling the relevant parameters in the information statistical tables of each sub - population layer and the sample observation values of each sub - population, the crop areas of each sub - population are circularly calculated. Therefore, the crop areas of multiple populations can be spatially extrapolated quickly and conveniently at one time to obtain the estimated values of each population. Moreover, the present application can estimate the areas of multiple populations for two consecutive years, and then calculate the inter - annual change rate of the crop areas of multiple populations, effectively meeting the requirements of operational operation;

[0069] 2) The sampling method adopted in the method and system for spatially sampling and extrapolating the crop areas of multiple populations of the present application is established based on scientific statistical theories. The operation results of the model have been tested by production practice and are considered reliable and accurate. Compared with traditional crop area statistical methods and other remote sensing methods for crop areas, it has the characteristics of low cost, high efficiency, and good stability. Description of the Drawings

[0070] Figure 1 is the flowchart of the method for creating a crop area sampling statistical model cluster of the present application;

[0071] Figure 2 is the schematic diagram of the sampling unit design process in the method for spatially sampling and extrapolating the crop areas of multiple populations of the present application;

[0072] Figure 3 is the flowchart of data stratification for modeling and generating basic model parameters in the method for spatially sampling and extrapolating the crop areas of multiple populations of the present application;

[0073] Figure 4 is the flowchart of calculating the minimum sample size and the sampling ratio in the method for spatially sampling and extrapolating the crop areas of multiple populations of the present application;

[0074] Figure 5In Example 1 of the method for extrapolating multi-population crop areas by spatial sampling in this application, a stratified sampling layer population distribution map of single-season rice in Heilongjiang Province was modeled;

[0075] Figure 6 In Example 1 of this application, it is a spatial distribution map of the stratified sampling samples of single-season rice in Heilongjiang Province;

[0076] Figure 7 It is a schematic diagram of the call process of various data and parameters during the operation of the model in this application;

[0077] Figure 8 It is a schematic diagram of the basic calculation process during the operation of the model in this application;

[0078] Figure 9 It is a schematic diagram of the interval estimation calculation process during the operation of the model in this application;

[0079] Figure 10 It is a composition diagram of the system for extrapolating multi-population crop areas by spatial sampling in this application. Detailed implementation manners

[0080] To make the purpose, technical solutions, and advantages of the implementation of this application clearer, the technical solutions in the embodiments of this application will be described in more detail below with reference to the accompanying drawings in the embodiments of this application. The described embodiments are some, but not all, of the embodiments of this application. The embodiments described below by referring to the accompanying drawings are exemplary and are intended to explain this application and should not be construed as limiting this application.

[0081] Example 1

[0082] The focus of this example is to introduce the methods for obtaining each parameter in step 101 (i.e., steps 1011 - 1019) in the method for extrapolating multi-population crop areas by spatial sampling in this application. Here, it can also be called the method for creating a crop area sampling statistical model cluster.

[0083] Taking the sampling survey of the planting areas of single-season rice and double-season late rice in 25 provincial administrative units out of 34 provincial administrative units in China as an example, among them, the survey target is rice, and the 25 provincial administrative units out of 34 provincial administrative units in China surveyed are designed as sub-populations, with a quantity of 25.

[0084] Such as Figures 1 - 4 As shown, the method for creating a crop area sampling statistical model cluster based on the cultivated land area (the cultivated land can be dry land or paddy field, and in this example, it is paddy field) in this application specifically includes the following steps:

[0085] Step 101: Obtain the paddy field areas of each provincial administrative unit in 25 provinces.

[0086] Furthermore, this step 101 is specifically further divided into the following steps:

[0087] Step 1011: Obtain the total paddy field spatial distribution vector map (i.e., the total cultivated land spatial distribution vector map, specifically from the second national land survey data in this embodiment and obtained through data exchange) of 25 provincial administrative units out of 34 provincial administrative units in China where rice is mainly planted through data exchange, purchase, or self-production.

[0088] Furthermore, determine the quantity and geographical locations of 25 provinces (sub-populations) within the scope of the total paddy field spatial distribution vector map, and obtain the sub-population boundary vector map (or administrative boundary vector map) of 25 provincial administrative units (sub-populations) out of 34 provincial administrative units in China through data exchange, purchase, or self-production.

[0089] It should be noted that in this step 1011, the following settings are also included:

[0090] 1) Set the spatial geographic coordinate system as CGCS2000 (China Geodetic Coordinate System 2000), and perform Albers projection within this spatial geographic coordinate system, that is, the projection is "orthoaxial equal-area double-standard parallel conic projection"; among them, the projection parameters are: the first standard parallel is 25, the second standard parallel is 47, and the central meridian is 105.

[0091] 2) Let the original universe U, U1, U2... be subsets of U, and x ∈ U; then:

[0092] U = {U1, U2,., Un}; U = {x1, x2,., xn};

[0093] Among them, U is the survey population (i.e., China), Un is the sampled survey sub-population (i.e., provincial administrative units), and x is the sampling unit.

[0094] Step 1012: Determine the sampling unit as the 1:50,000 topographic map frame of the standard map sheet (as shown in Figure 2 ), among which, the number of individual sampling units is multiple, and each sampling unit has a unique identification code (i.e., the map sheet number);

[0095] Furthermore, obtain the sampling unit layer vector map; among them, the sampling unit layer vector map is a vector map with the geographical coordinates described in step 1011. According to the sampling unit type, the sampling unit layer vector map can be purchased, exchanged from relevant units, or produced by oneself in GIS. In this embodiment, it is produced by oneself on the GIS platform; in addition, geospatially, the sampling unit layer vector map is consistent with the monitoring target population range.

[0096] Step 1013: On the GIS platform, use the sub - total boundary vector maps of 25 provincial administrative units (sub - populations) among the 34 provincial administrative units in China to cut the total paddy - field spatial distribution vector map, forming the paddy - field spatial distribution vector maps (i.e., sub - cultivated - land spatial distribution vector maps) of each provincial administrative unit (sub - population); similarly, on the GIS platform, use the sub - total boundary vector maps of 25 provincial administrative units (sub - populations) among the 34 provincial administrative units in China to cut the sampling - unit layer vector map, forming the sub - sampling - unit layer vector maps of each provincial administrative unit (sub - population).

[0097] Among them, the total paddy - field spatial distribution map and the sub - total boundary vector maps of 25 provincial administrative units (sub - populations) among the 34 provincial administrative units in China are vector data files, and their coordinate information conforms to the geographic coordinates described in Step 1011.

[0098] Step 1014: On the GIS platform, overlay the paddy - field spatial distribution vector maps of each provincial administrative unit (sub - population) with the corresponding sub - sampling - unit layer vector maps (as shown in Figure 2 ), and extract the paddy - field areas of each provincial administrative unit (sub - population) with sampling - unit individuals as units.

[0099] Moreover, the paddy - field areas of each provincial administrative unit (sub - population) extracted in this step with sampling - unit individuals as units are two - dimensional data table sets stored by provincial administrative units (see the first 3 columns of Table 2 below). There is one table for each provincial administrative unit (sub - population), and there are a total of 25 tables for 25 provincial administrative units (sub - populations) among the 34 provincial administrative units in China.

[0100] Furthermore, take the paddy - field area data of each provincial administrative unit (sub - population) statistically by sampling unit as the stratified basic data of each provincial administrative unit (sub - population). The data files are named according to the agreement and stored in a file directory.

[0101] Among them, the specific naming method can be set appropriately according to needs, which is convenient for subsequent programs to automatically read files. For example, the data file of Heilongjiang Province can be named: sthlj23callayer.

[0102] Step 1015: Set up a directory table of provincial administrative unit names (see Table 1 below). There is one record for each provincial administrative unit (sub - population), and the record fields include the full name of the provincial administrative unit (such as Heilongjiang, Hunan, etc.), abbreviation (such as HLJ, HN, etc.) and code (such as 23, 43, etc.). The abbreviation is agreed in the program, and the code conforms to the requirements of the GB2260 standard.

[0103] Table 1: Directory Table of Provincial Administrative Unit Names

[0104]

[0105]

[0106] Step 1016: Design program code to open the directory table of the names of provincial administrative units (sub-populations), loop to read the field values of each provincial administrative unit (sub-population), and dynamically construct the hierarchical basic data file names of each provincial administrative unit (sub-population) according to the agreement (strings, e.g., for Heilongjiang Province: sthlj23callayer). Read this string to open the hierarchical basic data files of each provincial administrative unit named according to the agreement stored in the agreed directory, and sort the data in the files in ascending order.

[0107] Step 1017: Stratify the hierarchical basic data of each provincial administrative unit (sub-population) sorted in ascending order in Step 1016, and display the stratification results in tabular form (using the preset table shown in Table 6 below) to obtain the sampling unit layer information table of each provincial administrative unit (sub-population).

[0108] Furthermore, the stratification method designed in this application is the method of cumulative frequency. The mathematical principle and application process of cumulative frequency will not be elaborated in detail here. The stratification process is briefly described as follows:

[0109] As shown in Table 2 below, for the need of automatic program execution, the header field names of the sampling unit layer information table of each provincial administrative unit (sub-population) are preset as sub-population code (DM), sampling unit code (BM), paddy field area (AREA_HA), grouping identifier (GROUP_FLAG), frequency (FREQUENCY), cumulative frequency (SUM_FRE), square root of cumulative frequency (SQRTSUM_F), and layer number (LAYER).

[0110] Specifically, initially set the grouping step size to 4 hectares and the number of layers to 6. Design a program to read the "paddy field area" in Table 2 below one by one, group by a step size of 4 hectares, and calculate the frequency, cumulative frequency, and square root of cumulative frequency of each group.

[0111] Design the number of layers to be 6, divide all sampling units into 6 layers according to the "square root of cumulative frequency", assign layer codes to each sampling unit individual, and thus dynamically generate the sampling unit layer information table of each provincial administrative unit (sub-population) (example: see Table 2 below for the sampling unit layer information table of the first-season rice in the sub-population of Heilongjiang Province). The names of the layer information tables of each provincial administrative unit (sub-population) are executed according to the agreement. There is one layer information table for each provincial administrative unit (sub-population), and there are a total of 25 layer information tables for 25 provincial administrative units (sub-populations) among the 34 provincial administrative units in China.

[0112] Table 2: Sampling Unit Layer Information Table of the First-Season Rice in Heilongjiang Province (Excerpt)

[0113]

[0114] Step 1018: Design the program code. Using the preset table shown in Table 7 below, loop through and read the information table files of each provincial administrative unit (sub-population) layer with hierarchical codes generated in Step 105, and calculate the information statistical values of each provincial administrative unit (sub-population) layer, including: the total number of cells in each layer total_cell; the minimum sample size min_s_cell, which is calculated according to the aforementioned formula, and the sampling precision is designed to be 95%; the layer mean ave_a_ha; the sampling ratio. Calculate the above values for each provincial administrative unit (sub-population) once, so as to dynamically generate the information statistical table of each provincial administrative unit (sub-population) layer (example: see Table 3 below for the information statistical table of the single-season rice layer in Heilongjiang Province), and the table name is automatically executed according to the agreed procedure. There is one information statistical table for each provincial administrative unit (sub-population), and there are a total of 25 information statistical tables for 25 provincial administrative units (sub-populations) among the 34 provincial administrative units in China.

[0115] Table 3: Information Statistical Table of the Single-Season Rice Layer in Heilongjiang Province

[0116]

[0117]

[0118] Step 1019: According to the automatic operation result of the program, check the information statistical table of the layer, and check whether the total number of cells in the layer, the minimum sample size, the sampling ratio, etc. meet the requirements. If they do not meet the requirements, reset the grouping step size and the number of levels, and execute Step 105 and Step 106 until the result meets the requirements. In the example of this application, the grouping step size is finally determined to be 4.5 hectares and the number of levels is 6. Finally, the information statistical table of the sub-population layer of Heilongjiang Province automatically generated by the program is shown in Table 3 above.

[0119] There is one information table and one information statistical table for each provincial administrative unit (sub-population). Among the 34 provincial administrative units in China, there are a total of 25 information tables and 25 information statistical tables for 25 provincial administrative units (sub-populations). The sampling survey model cluster for the planting areas of single-season rice and double-season late rice in 25 provincial administrative units among the 34 provincial administrative units in China has been created.

[0120] Finally, it should be noted that in order to establish the model and calculate the model parameters, a series of preset tables need to be constructed. The structures of the preset tables are shown in Tables 4 - 8 below:

[0121] Table 4: Structure of the Sub-Population Name Directory Table (Provinces_list.dbf)

[0122] field_name field_type field_len field_dec DM N 2 0 PNAME C 10 0 NAME_JC C 3 0

[0123] Table 5: Structure of the Sub-Population Sampling Basic Data Table (XXcallayer.dbf)

[0124] field_name field_type field_len field_dec BM N 13 0 AREA_HA N 14 2

[0125] Table 6: Stratified Information Process File Table and Structure of Layer Information Table (Frequency_cal_layer.dbf)

[0126]

[0127]

[0128] Table 7: Structure of Layer Information Statistical Table (Parameters_layer.dbf)

[0129] field_name field_type field_len field_dec LAYER N 1 0 TOTAL_CELL N 11 0 MIN_S_CELL N 11 0 AVE_A_HA N 18 1 RATE_SAMPL N 10 4

[0130] Table 8: Structure of Process File Table for Data Exchange of Layer Information Statistics (Data_shifting.dbf)

[0131] field_name field_type field_len field_dec LAYER N 1 0 TOTAL_CELL N 8 0 DATAVAR N 10 1 NVAR N 14 1

[0132] Among them, Table 4 corresponds to Table 1 in Step 103; Table 5 corresponds to the two-dimensional data table set obtained in Step 1014, that is, the first 3 columns of Table 2; Table 6 corresponds to Table 2 in Step 105; Table 7 corresponds to Table 3 in Step 106; Table 8 is used when the program exchanges data in Step 105. After the program runs to completion, the table is cleared.

[0133] Embodiment 2

[0134] This embodiment focuses on introducing the complete steps of the method for extrapolating the area of multiple overall crops by spatial sampling of the present application, including the above-mentioned Step 101 (i.e., Steps 1011 - 1019). In this embodiment, the sampled survey is the planting areas of single-season rice and double-season late rice in 25 provincial administrative units (sub-populations) among 34 provincial administrative units in China in 2021. For the models and parameters used, refer to Steps 1011 - 1019.

[0135] First aspect, as Figures 5 - 9 shown, the method for extrapolating the area of multiple overall crops by spatial sampling of the present application specifically includes the following steps:

[0136] Step 101 (i.e., including all the steps in the above-mentioned Embodiment 1), obtain the total cultivated land spatial distribution vector map of the monitored target population, and obtain the sampling unit layer information table and layer information statistical table of each provincial administrative unit. Among them, the sampling unit layer information table includes at least the sampling unit code and the layer field value.

[0137] Step 102. Within the scope of the total cultivated land spatial distribution vector map, based on the sampling unit layer information table of each provincial administrative unit, create a layer population distribution map of each provincial administrative unit stratified by cultivated land area (see Figure 5 as shown).

[0138] Step 103. According to the layer information statistical table of each provincial administrative unit, randomly select a sufficient number of sampling units from each layer on the layer population distribution map of each provincial administrative unit, record the identification codes of the selected sampling units, and form a list of survey sample sampling units for each provincial administrative unit (an example of a sub-population is shown in Table 9 below).

[0139] It should be emphasized that the number of sampling units used to extrapolate the area for each layer of each sub-population is equal to the actual sample size of each layer, and the actual sample size of each layer should not be less than the minimum sample size of each layer in the layer information statistical table of each sub-population.

[0140] Table 9: List of Stratified Sampling Units for Single-season Rice in Heilongjiang Province (Partial)

[0141] Provincial administrative unit code Sampling unit code 23 11520111 23 12510101 23 12510102 23 12510111 23 12510112 23 12510484 23 12520373 23 12520484 23 12520492 23 12520494 23 12520501 23 12520502 23 12520504 23 ……

[0142] Step 104. Based on the list of sampling units, obtain the geographical boundaries of each sampling unit to form a sample spatial distribution vector map (see Figure 6 as shown).

[0143] Step 105. Based on the sample spatial distribution vector map, according to the survey requirements, purchase appropriate remote sensing images, that is, the geographical location and quantity of the remote sensing images are determined by the sample spatial distribution vector map. For the remote sensing survey of single-season rice and double-season late rice in 25 out of 34 provincial administrative units in China in 2021, the selected remote sensing images are Landsat8 and Sentinel-2A. The total number of remote sensing images purchased for 25 out of 34 provincial administrative units in China is 1060 sampling units, and 297 for Heilongjiang Province.

[0144] Step 106. Conduct target recognition on the remote sensing images, extract the target crop area according to the sampling units to form sample observations. Among them, one provincial administrative unit corresponds to a set of sample observations, and the set of sample observations of all provincial administrative units forms a sample population file, and the file is named according to a predetermined naming method.

[0145] Furthermore, this step specifically includes:

[0146] Step 1061. On the spatial calculation standard platform, crop and splice the remote sensing images using the sampling unit vector files of each provincial administrative unit;

[0147] Step 1062: Conduct target recognition on the remote sensing image by sampling unit, adopting a combination of computer automatic classification and manual visual interpretation;

[0148] Step 1063: Extract the area of the target crop by sampling unit to form sample observations. To reduce data redundancy and simplify program design, the sample observation value sets of 25 out of 34 provincial administrative units in China (see Table 10 below) form a sample population file, and the file name is named according to the convention;

[0149] Table 10: Observation Values of Stratified Sampling of Single-season Rice in Heilongjiang Province

[0150] Provincial administrative unit code Sampling unit code Rice area in 2020 Rice area in 2021 23 11520111 1260.72 1550.23 23 12510101 3370.81 3230.18 23 12510102 1245.25 1052.38 23 12510111 2567.29 2411.06 23 12510112 5244.9 4830.62 23 12510484 46.41 43.41 23 12520373 372.6 714.6 23 12520484 12953.27 12695.23 23 12520492 5107.35 5107.35 23 12520494 4204.95 4094.49 23 12520501 16794.59 16368.06 23 12520502 5478.68 5242.02 23 12520504 3254.28 3254.28 … … … …

[0151] Step 1064: In this embodiment, according to the survey requirements, the government is more concerned about the annual change rate of crop area; for this reason, the remote sensing interpretation of crops is for two consecutive years.

[0152] Step 107: Set up a directory table of provincial administrative unit names. Among them, each provincial administrative unit corresponds to a record, and the field values of each record include at least the full name, abbreviation, and code of the provincial administrative unit. This step can be carried out again, or directly adopt the directory table of provincial administrative unit names set in the above Step 1015.

[0153] Step 108: Design program code to open the sample population file, open the directory table of provincial administrative unit names, loop to read the field values of each provincial administrative unit, and dynamically generate the sample observation value set of each provincial administrative unit (this sample observation value set belongs to an intermediate file, see Table 10 above for details).

[0154] Step 109: Design program code to call the sampling unit code and layer field values in the sampling unit layer information table of each provincial administrative unit (see Table 2 above), and assign each sampling unit level code in the sample observation value set of each provincial administrative unit.

[0155] Step 110: By looping to call the relevant parameters in the layer information statistical table of each provincial administrative unit and the sample observations of each provincial administrative unit into the following formula (1), thus loop to calculate the crop area of each provincial administrative unit, and finally form a statistical table of the total crop area of the whole population:

[0156]

[0157] Among them, L is the number of layers of stratified sampling; N h is the total number of each sampling layer; h = 1, 2,..., L; n h is the sample size of each sampling layer; y hi is the observation value of the i-th sampling unit in the h-th layer; is the estimated value of the sub-population, i.e., the crop area.

[0158] Furthermore, this step 110 may specifically include:

[0159] Step 1101: Design program code to loop through and open the set of sample observation values (i.e., the intermediate file) for each provincial administrative unit, and based on the sampling unit stratification field that has been assigned values in step 109, use formula (1) to loop through and calculate the mean of the sample observation values for each layer of each provincial administrative unit; store the calculation results in the data exchange process file table Result_stat.dbf.

[0160] Step 1102: Design program code to loop through and call the values of LAYER (stratification) and TOTAL_CELL (total population of the layer) for each layer in the information statistics table of each provincial administrative unit (see Table 3 above), and based on the "stratification" value, fill in the "total population of the layer" value into the data exchange process file table Result_stat.dbf.

[0161] Step 1103: Design program code to use the above formula (1) to loop through and calculate the extrapolation value for each layer of each provincial administrative unit, the estimated value of each provincial administrative unit, the inter-annual change rate, and the interval estimate value, and fill the results into the area result table (Result.dbf).

[0162] Finally, the results of extrapolating the rice planting areas of 25 out of 34 provincial administrative units in China in 2021 based on the cluster sampling of the spatial statistical model are shown in Table 11 below:

[0163] Table 11: Results of the remote sensing sampling survey of rice planting areas of 25 out of 34 provincial administrative units in China in 2021 Unit (hectare)

[0164]

[0165]

[0166] In the second aspect, as Figure 10 shown, the present application also discloses a system for spatially sampling and extrapolating multi-population crop areas, including an acquisition module 201, a basic data processing module 202, an extrapolation module 203, and a stratification module 204.

[0167] Among them, the acquisition module 201 is used for:

[0168] acquiring the total cultivated land spatial distribution vector map of the monitored target population, and acquiring the sampling unit layer information table and the layer information statistics table of multiple sub-populations included in the population, where the sampling unit layer information table includes at least the sampling unit code and the stratification field value; and

[0169] Obtain a remote sensing image, where the geographical location and quantity of the remote sensing image are adapted to the sample spatial distribution vector map.

[0170] The basic data processing module 202 is used for:

[0171] Within the scope of the total cultivated land spatial distribution vector map, based on the sampling unit layer information table of each sub-population, produce a layer population distribution map of the sub-population stratified by cultivated land area; and

[0172] According to the layer information statistical table of each sub-population, randomly select a sufficient number of extrapolation sampling units for each layer on the layer population distribution map, record the identification codes of the selected sampling units, and form a survey sample sampling unit list; and

[0173] Based on the sampling unit list, obtain the geographical boundaries of each extrapolation sampling unit to form a sample spatial distribution vector map; and

[0174] Perform target recognition on the remote sensing image, extract the target crop area according to the extrapolation sampling unit to form a sample observation value. Among them, one sub-population corresponds to a set of sample observation values, and the set of sample observation values of all sub-populations forms a sample population file, and the file is named according to a predetermined naming method; and

[0175] Set a sub-population name directory table, where each sub-population corresponds to a record, and the field values of each record at least include the full name, short name, and code of the sub-population.

[0176] The extrapolation module 203 is used for:

[0177] Loop to read the field values of each sub-population and dynamically generate a set of sample observation values for each sub-population; and

[0178] Call the sampling unit code and layer field value in the sampling unit layer information table of each sub-population, and assign a sampling unit level code to each sampling unit in the set of sample observation values of each sub-population; and

[0179] By looping to call the relevant parameters in the layer information statistical table of each sub-population and the sample observation values of each sub-population into the following formula (1), thereby loop calculating the crop area of each sub-population, and finally forming a statistical table of the crop area of all populations:

[0180]

[0181] Among them, L is the number of layers of stratified sampling; N h is the total number of each sampling layer population; h = 1, 2,..., L; n h is the sample size of each sampling layer; y hi is the observation value of the i-th sampling unit in the h-th layer; is the sub-population estimated value, that is, the crop area.

[0182] In an alternative embodiment, the obtaining module 201 is further configured to:

[0183] Obtain the total cultivated land space distribution vector map of the monitored target population and the sub-population boundary vector maps of multiple sub-populations; and

[0184] Set sampling units and obtain the sampling unit layer vector map of the sampling units; and

[0185] Obtain the sub-cultivated land space distribution vector maps and sub-sampling unit layer vector maps of each sub-population; and

[0186] Overlay each sub-cultivated land space distribution vector map with the corresponding sub-sampling unit layer vector map, and extract the cultivated land area of each sub-population with the sampling unit individual as the unit.

[0187] Correspondingly, the above basic data processing module 202 is further configured to:

[0188] Use the cultivated land area data of each sub-population as the stratified basic data of each sub-population, and store it in a file directory in a predetermined format; and

[0189] Set a sub-population name directory table, where each sub-population corresponds to a record, and the field values of each record at least include the full name, short name, and code of the sub-population; and

[0190] Loop to read the corresponding field values of each sub-population in the sub-population name directory table, dynamically construct the file names of the stratified basic data of each sub-population, so as to read the corresponding stratified basic data of each sub-population, and sort the data in ascending order.

[0191] Furthermore, the system for spatially sampling and extrapolating the crop areas of multiple populations in this application further includes a stratification processing module 204, and this stratification processing module 204 is configured to:

[0192] Stratify the stratified basic data of each sub-population after ascending sorting, and display the stratification results in tabular form to obtain the sub-population sampling unit layer information table; and

[0193] Loop to read the sub-population sampling unit layer information tables, calculate the layer statistical values of each sub-population, so as to obtain the sub-population layer information statistical tables; and

[0194] Judge whether the stratification result meets the requirements through the predetermined parameters in the sub-population layer information statistical tables. If not, return to modify the preset grouping step size and number of levels, and then execute the subsequent steps until it meets the requirements, where the predetermined parameters at least include the total number of layer populations, the minimum sample size, and the sampling ratio.

[0195] The above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present application should be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims described above.

Claims

1. A method for extrapolating the crop areas of multiple populations by spatial sampling, characterized in that, It includes the following steps: Step 101: Obtain the total cultivated land spatial distribution vector map of the overall monitoring target, and obtain the sampling unit layer information table and layer information statistical table of multiple sub-populations included in the overall population. Among them, the sampling unit layer information table at least includes the sampling unit code and the layer field value. In addition, in the step of obtaining the sampling unit layer information table and layer information statistical table of each sub-population, multiple sampling units are set, and each sampling unit has a unique identification code; Step 102: Within the range of the total cultivated land spatial distribution vector map, based on the sampling unit layer information table of each sub-population, make a layer population distribution map of the sub-population stratified by cultivated land area; Step 103: According to the layer information statistical table of each sub-population, randomly select a sufficient number of sampling units for each layer on the layer population distribution map of the sub-population, record the identification codes of the selected sampling units, and form a survey sample sampling unit list for each sub-population; Step 104: Based on the survey sample sampling unit list, obtain the geographical boundary of each sampling unit to form a sample spatial distribution vector map for each sub-population; Step 105: Obtain a remote sensing image, and the geographical location and quantity of the remote sensing image are adapted to the sample spatial distribution vector map; Step 106: Perform target recognition on the remote sensing image, extract the target crop area according to the sampling unit to form a sample observation value. Among them, one sub-population corresponds to a set of sample observation values, and the set of sample observation values of all sub-populations forms a sample population file, and the file is named according to a predetermined naming method; Step 107: Set a sub-population name directory table, where each sub-population corresponds to a record, and the field values of each record at least include the full name, short name, and code of the sub-population; Step 108: Loop to read the field values of each sub-population in the sub-population name directory table, and based on the read field values, extract the sub-population sample observation value set from the sample population file to dynamically generate the sub-population sample observation value set; Step 109: Call the sampling unit code and layer field value in the sampling unit layer information table of each sub-population, and assign a hierarchical code to each sampling unit in the sub-population sample observation value set; Step 110: By looping to call the relevant parameters in the layer information statistical table of each sub-population and the sample observation values of each sub-population into the following formula (1), calculate the crop area of each sub-population in a loop, and finally form a total crop area statistical table for all populations: where L is the number of layers for stratified sampling; N h is the total number of the population in each sampling layer; h = 1, 2, …, L; j = 1, 2, …, N h ; n h is the sample size of each sampling layer; y hi is the observed value of the i-th sampling unit in the h-th layer; is the sub-population estimate value, that is, the crop area.

2. The method for extrapolating the area of multiple overall crops by spatial sampling according to claim 1, wherein The said Step 101 includes: Step 1011: Obtain the total cultivated land spatial distribution vector map of the overall monitoring target and the sub-population boundary vector maps of multiple sub-populations; Step 1012: Set sampling units and obtain the sampling unit layer vector map of the sampling units; Step 1013: Obtain the sub-cultivated land spatial distribution vector map and sub-sampling unit layer vector map of each sub-population; Step 1014: Overlay the sub-cultivated land spatial distribution vector maps with the corresponding sub-sampling unit layer vector maps, extract the cultivated land area of each sub-population in units of sampling unit individuals, and use the cultivated land area data of each sub-population as the stratified basic data of each sub-population and store it in a file directory in a predetermined format; Step 1015: Set up the sub-population name directory table. Each sub-population corresponds to one record, and the field values of each record at least include the full name, short name, and code of the sub-population. Step 1016: Read the corresponding field values of each sub-population in the sub-population name directory table in a loop, dynamically construct the hierarchical basic data file names of each sub-population, so as to read the hierarchical basic data of the corresponding sub-populations, and sort the data in ascending order. Step 1017: Stratify the hierarchical basic data of each sub-population after ascending sorting, and display the stratification results in tabular form to obtain the sampling unit layer information table of each sub-population. Among them, the stratification is based on the size of the cultivated land area, the stratification method is the cumulative frequency method, and the preset grouping step size and number of levels are set. After stratification, all sampling unit individuals of each sub-population are assigned corresponding level codes. Step 1018: Read the sampling unit layer information table of each sub-population in a loop, calculate the layer statistical values of each sub-population, so as to obtain the layer information statistical table of each sub-population. Step 1019: Judge whether the stratification result meets the requirements through the predetermined parameters in the layer information statistical table of each sub-population. If not, return to the previous step to modify the preset grouping step size and number of levels until it meets the requirements. Among them, the predetermined parameters at least include the total number of the layer population, the minimum sample size, and the sampling ratio.

3. The method for spatially sampling and extrapolating multi-population crop areas according to claim 2, characterized in that, In the step 1013, on the GIS platform, the total cultivated land spatial distribution vector map and the sampling unit layer vector map are respectively cut by the sub-population boundary vector maps of the multiple sub-populations, so as to form the sub-cultivated land spatial distribution vector map and the sub-sampling unit layer vector map of each sub-population.

4. The method for spatially sampling and extrapolating multi-population crop areas according to claim 2, characterized in that In the step 1014, the cultivated land areas of each sub-population with sampling unit individuals as units are two-dimensional data table sets stored in each sub-population.

5. The method for extrapolating multi-population crop areas by spatial sampling according to claim 2, characterized in that, In the step 1017, the field values of the table head of each sub-population sampling unit layer information table at least include sub-population code, sampling unit code, cultivated land area, frequency, cumulative frequency, square root of cumulative frequency, and level.

6. The method for extrapolating the area of multiple populations of crops by spatial sampling according to claim 2, characterized in that In the step 1019, the minimum sample size n is calculated by the following formula: Among them, L is the number of layers of stratified sampling; N h is the total number of each sampling layer; h = 1, 2, …, L; is the variance of each sampling layer; N is the total population; W h = N h / N, that is, the sampling weight of the h-th layer; V is the variance of the estimator. If V is not given, but the error bound d is given, then V = (d / t) 2 , when the sample size is quite large, it can be assumed that t = 1.96; d = (1 - δ)Y, where δ is the sampling precision and Y is the total population value; The sampling ratio r is calculated by the following formula: r = n / N; Finally, the minimum sample size \(n\) of each layer h is calculated by the following formula: n h = N h × r.

7. The method for extrapolating the area of multiple overall crops by spatial sampling according to claim 2, wherein In the step 103, the number of sampling units used to extrapolate the area in each layer of each sub-population is equal to the actual sample size of each layer, and the actual sample size of each layer should not be less than the minimum sample size of each layer in the layer information statistical table of each sub-population.

8. The method for extrapolating the area of multiple overall crops by spatial sampling according to claim 2, wherein In the step 105, the obtained remote sensing image completely covers the geographical space of the sampling units shown in the sample space distribution vector map of each sub-population.

9. A system for extrapolating the crop areas of multiple populations by spatial sampling, characterized in that, Including: An acquisition module (201) for acquiring the total cultivated land spatial distribution vector map of the monitoring target population, and acquiring the sampling unit layer information table and the layer information statistical table of multiple sub-populations included in the population. Among them, at least the sampling unit code and the layer field value are included in the sampling unit layer information table; and For acquiring a remote sensing image, the geographical location and quantity of the remote sensing image are adapted to the sample space distribution vector map. The basic data processing module (202) is used to, within the scope of the total cultivated land spatial distribution vector map, make a layer overall distribution map of each sub-population based on the cultivated land area stratification according to the sampling unit layer information table of each sub-population; and is used to randomly select a sufficient number of sampling units for each layer on the layer overall distribution map according to the layer information statistical table of each sub-population, record the identification codes of the selected sampling units, and form a survey sample sampling unit list; and is used to obtain the geographical boundaries of each selected sampling unit according to the sampling unit list, and form a sample spatial distribution vector map; and is used to perform target recognition on the remote sensing image, extract the target crop area according to the sampling unit, and form sample observations. Among them, one sub-population corresponds to a set of sample observations, and the set of sample observations of all sub-populations forms a sample overall file, and the file is named according to a predetermined naming method; and is used to set a sub-population name directory table, where each sub-population corresponds to a record, and the field values of each record at least include the full name, short name, and code of the sub-population; The extrapolation module (203) is used to loop through and read the field values of each sub-population to dynamically generate a set of sub-population sample observations; and is used to call the sampling unit codes and layer field values in the sampling unit layer information table of each sub-population, and assign a hierarchical code to each sampling unit in the set of sub-population sample observations; and is used to loop through and calculate the crop areas of each sub-population by looping through and calling the relevant parameters in the layer information statistical table of each sub-population and the sample observations of each sub-population into the following formula (1), and finally form a total overall crop area statistical table: Among them, L is the number of layers of stratified sampling; N h is the total number of each sampling layer; h = 1, 2, …, L; j = 1, 2, …, N h ; n h is the sample size of each sampling layer; y hi is the observed value of the i-th sampling unit in the h-th layer; is the sub-population estimated value, that is, the crop area.

10. The system for spatially sampling and extrapolating multi-population crop areas according to claim 9, wherein The obtaining module (201) is further used for: obtaining the total cultivated land spatial distribution vector map of the monitored target population and the sub-population boundary vector maps of multiple sub-populations; and setting sampling units, and obtaining the sampling unit layer vector map of the sampling units; and obtaining the sub-cultivated land spatial distribution vector maps and sub-sampling unit layer vector maps of each sub-population; and overlaying each sub-cultivated land spatial distribution vector map with the corresponding sub-sampling unit layer vector map, and extracting the cultivated land areas of each sub-population in units of sampling unit individuals; The basic data processing module (202) is further used for: taking the cultivated land area data of each sub-population as the stratified basic data of each sub-population, and storing it in a file directory in a predetermined format; and setting a sub-population name directory table, where each sub-population corresponds to a record, and the field values of each record at least include the full name, short name, and code of the sub-population; and looping through and reading the corresponding field values of each sub-population in the sub-population name directory table, dynamically constructing the file names of the stratified basic data of each sub-population, thereby reading the stratified basic data of the corresponding sub-populations, and sorting the data in ascending order; The system for spatially sampling and extrapolating the crop areas of multiple populations further includes: a stratification processing module (204) for stratifying the stratified basic data of each sub-population after sorting in ascending order, and displaying the stratification results in tabular form to obtain the sampling unit layer information table of each sub-population; and For cyclically reading the information tables of sampling unit layers of each sub-population, calculating the layer statistical values of each sub-population, so as to obtain the information statistical tables of each sub-population layer; and For judging whether the stratification result meets the requirements through the predetermined parameters in the information statistical tables of each sub-population layer. If not, return to modify the preset grouping step size and the number of levels, and then execute the subsequent steps until the requirements are met, where the predetermined parameters at least include the total number of the layer population, the minimum sample size, and the sampling ratio.

Images