Remote sensing-based crop nitrogen fertilizer demand quantity processing method and device

Abstract

The invention provides a crop nitrogen fertilizer demand quantity processing method and device based on remote sensing, and is applied to the technical field of crop fertilization decision, and the method comprises the steps: dividing a plurality of cultivated lands at equal intervals according to the relative yield of a target crop, and obtaining the soil nutrient grade of each cultivated land in the plurality of cultivated lands; acquiring remote sensing data images of a plurality of cultivated lands in a plurality of time periods; based on the remote sensing data image, determining the maximum vegetation index of each cultivated land in each of the plurality of time periods; based on the maximum vegetation index and the meteorological data, the average yield of each cultivated land in the multiple time periods is determined; determining a product of the average yield and a preset yield annual increment rate as a target yield of each cultivated land; based on the soil nutrient grade and the target yield, determining the crop nitrogen fertilizer total demand quantity of the target crop of each cultivated land; according to the invention, the total nitrogen fertilizer demand quantity of cultivated land can be accurately determined before crop sowing.

Description

Technical Field

[0001] The present invention relates to the technical field of crop fertilization decision-making, and particularly to a method and device for processing crop nitrogen fertilizer demand based on remote sensing. Background Art

[0002] Nitrogen fertilizer is a key nutrient for crop growth and development, crucial for the synthesis of proteins and starches, and directly affects crop yield and quality. The principle of applying nitrogen fertilizer is not the more the better. Excessive application of nitrogen fertilizer will not only cause waste of fertilizer and increase production costs, but may also have a negative impact on the environment.

[0003] However, determining the total nitrogen fertilizer demand of crops through field trials has strong regional limitations. It is necessary to set up multi-point trials under different soil, climate, farming conditions and crop varieties, with poor operability and a problem of time lag.

[0004] It can be seen that the method for processing the total nitrogen fertilizer demand of crops in the related art has technical problems of relatively high limitations and poor timeliness. Summary of the Invention

[0005] The present invention provides a method and device for processing crop nitrogen fertilizer demand based on remote sensing, to solve the defects of relatively high limitations and poor timeliness in the method for processing the total nitrogen fertilizer demand of crops in the prior art, and to accurately determine the total nitrogen fertilizer demand of cultivated land before crop sowing.

[0006] The present invention provides a method for processing crop nitrogen fertilizer demand based on remote sensing, including the following steps. Divide multiple cultivated lands at equal intervals according to the relative yield of the target crop to obtain the soil nutrient grade of each cultivated land among the multiple cultivated lands; obtain remote sensing data images of the multiple cultivated lands in multiple time periods; based on the remote sensing data images, determine the maximum vegetation index of each cultivated land in each time period of the multiple time periods; based on the maximum vegetation index and meteorological data, determine the average yield of each cultivated land in the multiple time periods; multiply the average yield by the preset annual yield increase rate to determine the target yield of each cultivated land; based on the soil nutrient grade and the target yield, determine the total crop nitrogen fertilizer demand of the target crop for each cultivated land.

[0007] According to the method for processing crop nitrogen fertilizer demand based on remote sensing provided by the present invention, the step of dividing multiple cultivated lands at equal intervals according to the relative yield of the target crop to obtain the soil nutrient grade of each cultivated land among the multiple cultivated lands includes: obtaining the soil nitrogen content and relative yield of the target crop in each cultivated land among the multiple cultivated lands; determining the first linear relationship between the soil nitrogen content and the relative yield; based on the inverse function of the first linear relationship, determine the soil nutrient grade of each cultivated land.

[0008] A method for processing crop nitrogen fertilizer demand based on remote sensing according to the present invention. Determining the maximum vegetation index of each cultivated land in each of the multiple time periods based on the remote sensing data image includes: determining the vegetation index of each pixel based on the remote sensing data images of the multiple cultivated lands in the multiple time periods; constructing a time series data set in chronological order based on the vegetation index of each pixel, wherein each time series in the time series data set corresponds to the pixel respectively; preprocessing the time series corresponding to each pixel by time series harmonic analysis method to obtain the maximum vegetation index of each cultivated land in each of the multiple time periods.

[0009] A method for processing crop nitrogen fertilizer demand based on remote sensing according to the present invention. Before determining the average yield of each cultivated land in the multiple time periods based on the maximum vegetation index and meteorological data, the method further includes: obtaining the actual yield of each cultivated land in each time period of a preset time period, and the maximum vegetation index of each cultivated land in each time period of the preset time period, wherein the preset time period is the target growth period of the target crop; determining a second linear relationship between the actual yield and the maximum vegetation index, wherein the second linear relationship includes: a first coefficient and a first intercept; obtaining a first coefficient data set, a first intercept data set, and the average temperature, average precipitation, and average solar radiation of each time period of the preset time period; determining a third linear relationship between the first coefficient data set, the average temperature, the average precipitation, and the average solar radiation; determining a fourth linear relationship between the first intercept data set, the average temperature, the average precipitation, and the average solar radiation.

[0010] A method for processing crop nitrogen fertilizer demand based on remote sensing according to the present invention. The meteorological data includes temperature, precipitation, and solar radiation. Determining the average yield of each cultivated land in the multiple time periods based on the maximum vegetation index and meteorological data includes: taking each time period in the multiple time periods as a target time period respectively and performing the following operations to obtain the average yield of the target time period: based on the average temperature, average precipitation, and average solar radiation of the target time period, determining a target coefficient and a target intercept according to the third linear relationship and the fourth linear relationship; determining the average yield of the target time period based on the target coefficient, the target intercept, and the maximum vegetation index of the target time period.

[0011] A method for processing crop nitrogen fertilizer demand based on remote sensing provided by the present invention, determining the total crop nitrogen fertilizer demand of the target crop for each cultivated land based on the soil nutrient level and the target yield, includes: Wherein, represents the total crop nitrogen fertilizer demand of the target crop, represents the soil nutrient level, represents the target yield, represents the grain nitrogen content of the target crop, represents the preset nitrogen fertilizer utilization rate.

[0012] The present invention also provides a device for processing crop nitrogen fertilizer demand based on remote sensing, including the following modules: a division module for equally spacing multiple cultivated lands according to the relative yield of the target crop to obtain the soil nutrient level of each cultivated land in the multiple cultivated lands; an acquisition module for acquiring remote sensing data images of the multiple cultivated lands in multiple time periods; a vegetation index module for determining the maximum vegetation index of each cultivated land in each of the multiple time periods based on the remote sensing data images; an average yield module for determining the average yield of each cultivated land in the multiple time periods based on the maximum vegetation index and meteorological data; a target yield module for determining the product of the average yield and the preset annual yield increase rate as the target yield of each cultivated land; a total demand module for determining the total crop nitrogen fertilizer demand of the target crop for each cultivated land based on the soil nutrient level and the target yield.

[0013] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the method for processing crop nitrogen fertilizer demand based on remote sensing as described in any one of the above when executing the program.

[0014] The present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the method for processing crop nitrogen fertilizer demand based on remote sensing as described in any one of the above.

[0015] The present invention also provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the method for processing crop nitrogen fertilizer demand based on remote sensing as described in any one of the above.

[0016] The method and device for processing crop nitrogen fertilizer demand based on remote sensing provided by the present invention can accurately evaluate the soil nutrient status of each cultivated land by analyzing the relative yield of the target crop and equally dividing multiple cultivated lands at equal intervals; by using remote sensing technology to obtain the cultivated land image data in multiple time periods, the vegetation growth status can be monitored in real time, and by analyzing the maximum vegetation index, the growth potential and health status of the crop can be indirectly reflected; by combining the maximum vegetation index with meteorological data, the average yield of each cultivated land in multiple time periods can be predicted more accurately, considering the natural environmental factors of crop growth, making the yield prediction closer to the actual situation; by determining the product of the average yield and the preset annual yield growth rate as the target yield of each cultivated land, the historical yield level and the expected future yield growth are considered; based on the soil nutrient level and the target yield, the total crop nitrogen fertilizer demand required for each cultivated land can be calculated more accurately. Thus, by precise fertilization, the growth requirements of the crop can be met, and environmental pollution and cost waste caused by excessive fertilization can be avoided. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art one by one. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 It is a schematic flowchart of the method for processing crop nitrogen fertilizer demand based on remote sensing provided by the present invention.

[0019] Figure 2 It is a technical flowchart for constructing the lookup table of the total crop nitrogen fertilizer demand provided by the present invention.

[0020] Figure 3 It is a schematic structural diagram of the device for processing crop nitrogen fertilizer demand based on remote sensing provided by the present invention.

[0021] Figure 4 It is a schematic physical structure diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] To make the objectives, technical solutions, and advantages of the present invention clearer, the following will clearly and completely describe the technical solutions in the present invention with reference to the drawings in the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0023] Nitrogen fertilizer is a key nutrient for crop growth and development, crucial for the synthesis of proteins and starches, and directly affects crop yield and quality. However, the principle of applying nitrogen fertilizer is not the more the better. Excessive fertilization not only causes fertilizer waste and increased production costs, but may also have a negative impact on the environment. Therefore, for major food crops such as wheat, corn, and rice, it is necessary to determine the appropriate total nitrogen fertilizer requirement according to the soil nutrient heterogeneity of the field before sowing.

[0024] Currently, the methods for determining the total nitrogen fertilizer requirement of crops mainly include the fertilizer effect function method, the nutrient balance method, and the soil fertility difference method. The fertilizer effect function method requires a large number of experimental treatments. The nutrient balance method and the soil fertility difference method also require at least field experiments. These methods all calculate the total nitrogen fertilizer requirement of crops based on the nitrogen requirement for the target yield, and the target yield is determined by the yield results of field experiments in nitrogen-sufficient plots. However, these field experiment-based methods have strong regional limitations, require multi-point experiments to be set up under different soil, climate, tillage conditions, and crop varieties, have poor operability, and have the problem of time lag. Moreover, the fertilization decisions made based on these methods are limited to the regional scale and cannot formulate appropriate fertilization decision-making schemes for the field scale.

[0025] With the rapid development of remote sensing technology, it has been able to accurately estimate and predict crop physiological and biochemical components and yields. Nevertheless, the integration of remote sensing technology with crop nitrogen fertilizer decision-making models in actual production is still insufficient, lacking a calculation model for the total nitrogen fertilizer requirement of crops driven by remote sensing technology. Therefore, it is necessary to improve the traditional decision-making methods that highly rely on field experiments, have a time lag, and cannot adapt to the field scale.

[0026] Therefore, the present invention proposes a method for processing crop nitrogen fertilizer requirements based on remote sensing. This method constructs a lookup table for the total nitrogen fertilizer requirement of crops, aiming to formulate a nitrogen fertilizer decision-making scheme at the field scale through the soil nitrogen content range of the field and the lookup table on the basis of reasonably determining the target yield of the field by integrating remote sensing technology. This method avoids conducting field experiments in nitrogen-sufficient plots, simplifies the operation process, and is more conducive to application in agricultural practice. This method can effectively overcome the problems of incompleteness and inaccuracy of traditional nitrogen fertilizer decision-making methods and achieve total fertilizer decision-making at the field scale according to local conditions.

[0027] In the related art, the traditional total fertilizer decision-making method relies on a large number of field trials with different fertilizer gradients, which is labor-intensive and cannot formulate a total fertilizer decision-making plan before crop planting. To determine the target yield traditionally, it is necessary to set up a nitrogen-sufficient plot in each region and measure its yield as the target yield, or it is necessary to calculate the target yield by the average yield-year increment rate method based on the yields of the previous three years in this region. Due to the lack of statistical data records or the lack of representativeness in the selection and layout of nitrogen-sufficient plots, the representativeness of the target yield data is insufficient. The traditional total fertilizer decision-making model can only formulate a regional total fertilizer decision-making plan for the target yield at the regional scale, and it is difficult to formulate a suitable total fertilizer decision-making plan for complex farmland conditions.

[0028] In view of the above defects, the present invention proposes a method for processing crop nitrogen fertilizer requirements based on remote sensing. This model avoids field trials with different nitrogen fertilizer gradients, especially the implementation of field trials in nitrogen-sufficient plots. Based on remote sensing data, it realizes the calculation of target yields with high spatial resolution and high precision. At the same time, according to the relative yield, soil nutrients are divided into different levels to calculate the total nitrogen fertilizer requirements for plots with different soil nutrient levels, thereby formulating a lookup table for the total nitrogen fertilizer requirements of crops considering the heterogeneity of soil fertility in the fields. The lookup table for the total nitrogen fertilizer requirements of crops constructed by the present invention is suitable for the total fertilizer decision-making of crops in different target yields and fields with different soil fertilities, simplifies the operation process in the actual production process, and is more conducive to efficient, timely, and accurate application in practice.

[0029] Optionally, the method for processing crop nitrogen fertilizer requirements based on remote sensing in the embodiments of the present application can be executed by a server, or by a terminal device, or jointly executed by a server and a terminal device. Taking the server as an example to execute the method for processing crop nitrogen fertilizer requirements based on remote sensing in this embodiment.

[0030] Figure 1 is a schematic flowchart of the method for processing crop nitrogen fertilizer requirements based on remote sensing provided by the present invention. As Figure 1 shown, the method includes the following steps.

[0031] Step 101: Divide multiple cultivated lands at equal intervals according to the relative yield of the target crop to obtain the soil nutrient level of each cultivated land among the multiple cultivated lands.

[0032] In some embodiments, the soil nitrogen content (i.e., soil nitrogen content) and relative yield of non-nitrogen-sufficient plots of the same crop (such as corn, wheat, etc.) are obtained, and the cultivated lands are divided at equal intervals according to the relative yield to obtain the soil nutrient level of each cultivated land among the multiple cultivated lands.

[0033] For example, the relative yield is divided at equal intervals of 10% using the equal division method. In the present invention, considering the actual production application of cultivated land, the relative yield is divided into 7 grades: ((95%, 100%], (85%, 95%], (75%, 85%], (65%, 75%], (55%, 65%], (45%, 55%], (0%, 45%]).

[0034] Collect data on the soil nitrogen content of corn or wheat from multiple cultivated lands, and at the same time collect data on the relative yield of these cultivated lands under the same management conditions. The relative yield is usually calculated based on the yield in the full-fertilizer area.

[0035] Correspond the soil nitrogen content and relative yield data of each cultivated land to form a data set; use the soil nitrogen content (x) as the abscissa and the relative yield (y) as the ordinate to draw a scatter plot; observe the distribution trend of the scatter plot and preliminarily judge the linear relationship between the soil nitrogen content and the relative yield.

[0036] According to the distribution of the relative yield, determine the standard for equal-interval division (in the present invention, the relative yield is divided at equal intervals of 10%); according to the determined division standard, divide the cultivated lands at equal intervals according to the relative yield. Among them, the cultivated lands within each grade have similar relative yield levels.

[0037] Step 102, obtain remote sensing data images of multiple cultivated lands over multiple time periods.

[0038] In the embodiments of the present invention, use a geographic information system or an online map tool to delimit the geographical boundaries of the target cultivated land (multiple cultivated lands); classify the cultivated land in detail according to the crop planting type, soil type, landform characteristics, etc. to ensure the accuracy of subsequent remote sensing data selection.

[0039] According to the crop growth cycle, set the time range for obtaining remote sensing data. For example, the critical periods of crop growth (such as the sowing period, growth period, and maturity period) can be selected for observation.

[0040] Step 103, based on the remote sensing data images, determine the maximum vegetation index of each cultivated land in each time period of multiple time periods.

[0041] In the embodiments of the present invention, calculate the EVI2 (Enhanced Vegetation Index 2) vegetation index of the cultivated land area based on Sentinel-2 image data, perform preprocessing on the EVI2 time series curve of the previous 3 years for each pixel using the harmonic analysis of time series (HANTS), and extract the maximum EVI2 (EVI2max) of each of the previous 3 years.

[0042] For each pixel, extract the EVI2 values for the previous 3 years to form time - series data; use the HANTS algorithm to perform harmonic analysis on the time - series data. The HANTS algorithm can identify and smooth the periodic changes in the time - series while removing noise; from the time - series data processed by HANTS, identify the EVI2 peak for each year; for each pixel, extract the maximum EVI2 value (EVI2max) for each of the previous 3 years.

[0043] Step 104: Based on the maximum vegetation index and meteorological data, determine the average yield of each cultivated land in multiple time periods.

[0044] Use remote sensing technology (such as Sentinel - 2) to calculate the maximum vegetation index (EVI2max or NDVImax) of each cultivated land in multiple time periods (such as years, seasons, or growth cycles). Collect meteorological data corresponding to the cultivated land from weather stations, meteorological satellites, or online meteorological databases, including temperature, precipitation, sunshine hours, etc.

[0045] Conduct a correlation analysis on the maximum vegetation index, meteorological data, and crop yield to determine the degree of correlation between them. According to the results of the correlation analysis, select a suitable yield estimation model. Commonly used models include linear regression models, non - linear regression models, machine learning models (such as random forest, support vector machine, etc.).

[0046] Use the known historical yield data, maximum vegetation index data, and meteorological data to train the model to obtain model parameters.

[0047] Input the maximum vegetation index data and meteorological data in the time period to be estimated into the trained yield estimation model; the model outputs the yield estimation value of each cultivated land in each time period, and then calculates the average yield of each cultivated land in multiple time periods.

[0048] Step 105: Multiply the average yield by the preset annual yield increase rate to determine the target yield of each cultivated land.

[0049] In the embodiment of the present invention, determine the average yield data of each cultivated land in multiple time periods (such as the past few years) and the annual yield increase rate. Among them, the annual yield increase rate is a preset value, reflecting the expected or predicted future yield growth trend.

[0050] The target yield of the target cultivated land can be determined by the following formula: Target yield = Average yield×(1 + Annual yield increase rate)^n, where n is the number of years difference between the predicted future year and the current year.

[0051] Step 106: Based on the soil nutrient level and the target yield, determine the total nitrogen fertilizer requirement of the target crop for each cultivated land.

[0052] In some embodiments, based on the soil nutrient level and the target yield, the total nitrogen fertilizer requirement for crops is calculated through the nitrogen fertilizer application rate formula.

[0053] Through the above steps of the embodiments of the present invention, by analyzing the relative yield of the target crop and equally dividing multiple cultivated lands at equal intervals, the soil nutrient status of each cultivated land can be accurately evaluated; by using remote sensing technology to obtain the cultivated land image data in multiple time periods, the vegetation growth status can be monitored in real time, and by analyzing the maximum vegetation index, the growth potential and health status of the crop can be indirectly reflected; by combining the maximum vegetation index with meteorological data, the average yield of each cultivated land in multiple time periods can be predicted more accurately, considering the natural environmental factors of crop growth, making the yield prediction closer to the actual situation; by determining the product of the average yield and the preset annual yield increase rate as the target yield of each cultivated land, the historical yield level and the expected future yield growth are considered; based on the soil nutrient level and the target yield, the total nitrogen fertilizer requirement for crops needed for each cultivated land can be calculated more accurately. Thus, through precise fertilization, the growth requirements of the crops can be met, and environmental pollution and cost waste caused by excessive fertilization can be avoided.

[0054] According to the method for processing nitrogen fertilizer requirements for crops based on remote sensing provided by the present invention, multiple cultivated lands are equally divided at equal intervals according to the relative yield of the target crop, and the soil nutrient level of each cultivated land among the multiple cultivated lands is obtained, including: Obtain the soil nitrogen content and relative yield of the target crop in each of the multiple cultivated lands. Determine the first linear relationship between the soil nitrogen content and the relative yield. Based on the inverse function of the first linear relationship, determine the soil nutrient level of each cultivated land.

[0055] In the embodiments of the present invention, consult the literature data over the years, and classify and organize the soil nitrogen content (x) and relative yield (y) of non-nitrogen sufficient plots of the same crop (such as corn, wheat, etc.).

[0056] Based on the collected relative yield (y) and soil nitrogen content (x) data, construct a regression equation between the relative yield (y) and the soil nitrogen content (x): y = f(x) (1) First, the relative yield (y) is equally divided at equal intervals of 10% by the equal division method. In the embodiments of the present invention, considering the actual production application of the cultivated land, we divide the relative yield into 7 grades ((95%, 100%], (85%, 95%], (75%, 85%], (65%, 75%], (55%, 65%], (45%, 55%], (0%, 45%]).

[0057] Solve the inverse function f according to formula (1). -1 (y), and the soil nutrient levels corresponding to the 7 relative yield levels can be calculated ((f -1 (95%), f -1 (100%)], (f -1 (85%), f -1 (95%)], (f -1 (75%), f -1 (85%)], (f -1 (65%), f -1 (75%)], (f -1 (55%), f -1 (65%)], (f -1 (45%), f -1 (55%)], (f -1 (0%), f -1 (45%)).

[0058] Through the embodiments of the present invention, by collecting data on soil nitrogen content and relative yield of the target crop in each cultivated land, a linear relationship based on actual observations can be established, and the linear regression method is used to determine the relationship between soil nitrogen content and relative yield, which can reduce the influence of random errors on the results.

[0059] According to the method for processing crop nitrogen fertilizer demand based on remote sensing provided by the present invention, based on remote sensing data images, the maximum vegetation index of each cultivated land in each time period of multiple time periods is determined, including: Based on remote sensing data images of multiple cultivated lands in multiple time periods, determine the vegetation index of each pixel; Based on the vegetation index of each pixel, construct a time series data set in chronological order, where each time series of the time series data set corresponds to the pixel respectively; Perform preprocessing on the time series corresponding to each pixel by the time series harmonic analysis method to obtain the maximum vegetation index of each cultivated land in each time period of multiple time periods.

[0060] In the embodiments of the present invention, the EVI2 vegetation index of the cultivated land area is calculated based on Sentinel-2 image data, and the time series harmonic analysis method (Harmonic Analysis of Time Series, HANTS) is used to preprocess the EVI2 time series curve of the previous 3 years for each pixel, and the maximum EVI2 (EVI2max) of each of the previous 3 years is extracted.

[0061] In some embodiments, determining the vegetation index of each pixel includes: collecting remote sensing data images within multiple time periods, where the remote sensing data images cover the target cultivated area; performing preprocessing such as radiometric correction and geometric correction on the remote sensing data images to obtain preprocessed remote sensing data to ensure the accuracy and consistency of the data; determining the vegetation index of each pixel based on the preprocessed remote sensing data. For example, common vegetation indices include the Normalized Difference Vegetation Index (NDVI), the Enhanced Vegetation Index (EVI), etc. The vegetation index can reflect the growth status and health of vegetation.

[0062] Construct a time series dataset of the vegetation index of each pixel in chronological order, where each time series in the time series dataset corresponds to a specific pixel and records the change in the vegetation index of that pixel over different time periods.

[0063] In the embodiments of the present invention, selecting the time series harmonic analysis method as a preprocessing means can extract the periodic components in the time series, which helps to analyze the change trend and seasonal fluctuations of the vegetation index.

[0064] Perform harmonic analysis on the time series corresponding to each pixel to extract the harmonic components of each time period; based on the harmonic components of each time period, determine the maximum vegetation index of each cultivated land in each time period. The maximum vegetation index usually corresponds to the period when the vegetation grows most vigorously.

[0065] Through the embodiments of the present invention, the construction of the time series dataset can intuitively observe and analyze the change trend of the vegetation index over time, revealing the growth law and seasonal fluctuations of the vegetation; the time series harmonic analysis method can extract the periodic components in the time series, which helps to improve the depth and breadth of data analysis.

[0066] According to the method for processing crop nitrogen fertilizer demand based on remote sensing provided by the present invention, before determining the average yield of each cultivated land over multiple time periods based on the maximum vegetation index and meteorological data, the above method further includes: Obtain the actual yield of each cultivated land in each time period within a preset time period, and the maximum vegetation index of each cultivated land in each time period within the preset time period, where the preset time period is the target growth period of the target crop; Determine the second linear relationship between the actual yield and the maximum vegetation index, where the second linear relationship includes: a first coefficient and a first intercept; Obtain a first coefficient dataset, a first intercept dataset, and the average temperature, average precipitation, and average solar radiation in each time period of the preset time period; Determine the third linear relationship between the first coefficient dataset, the average temperature, the average precipitation, and the average solar radiation; Determine the fourth linear relationship among the first intercept dataset, average temperature, average precipitation, and average solar radiation.

[0067] In an embodiment of the present invention, a hierarchical linear model for crop yield prediction is constructed.

[0068] In an embodiment of the present invention, based on experimental data or literature data, the actual yield of each cultivated land in each time period within a preset time period, and the maximum vegetation index of each cultivated land in each time period within the preset time period are obtained, and the first-level linear model (i.e., the second linear relationship) of the hierarchical linear model for crop yield prediction is constructed: Level1: yield = a0 × EVI2 max + a1 where Level1 represents the first-level linear model, yield represents the actual yield, a0 represents the first coefficient, EVI2 max represents the maximum vegetation index, and a1 represents the first intercept.

[0069] In an embodiment of the present invention, meteorological data is an important input variable for yield prediction, including the monthly average temperature (T mean )), monthly average precipitation (P mean ), and monthly average solar radiation (R mean ) in the critical growth periods of different crops (such as March - May for winter wheat, July - August for summer maize, June - August for single-season rice, etc.), all of which are generated by interpolating meteorological data provided by meteorological bureaus or meteorological stations. Based on the cumulative meteorological data T mean , P mean , and R mean in the critical growth periods of different crops and their corresponding slopes (k), the second-level linear model of the hierarchical linear model for crop yield prediction is constructed: Level 2: a i = b0 + b1 × T mean + b2 × P mean + b3 × R mean In the formula, Level 2 represents the second-level linear model, a i represents a0 and a1 in Level 1, i is 0 and 1 respectively, b0 is the intercept, and b1, b2, and b3 are the slopes of the cumulative T mean , P mean , and R mean in the critical growth period.

[0070] In an embodiment of the present invention, in the first-level linear model, for each crop, in different years of different regions, a linear model can be constructed between EVI2 max and yield: yield = a0 × EVI2 max+a1, which will generate a dataset of a0 and a1 (each region will correspond to an a0 and an a1 every year).

[0071] In the second - layer linear model, for each crop, each region has a definite a0 and a1 every year, and the monthly average temperature (T mean ), monthly average precipitation (P mean ), and monthly average solar radiation (R mean ) during the critical growth period of the crop (such as March - May for winter wheat, July - August for summer maize, June - August for single - season rice, etc.) can be calculated.

[0072] The a0 dataset (i.e., the first - coefficient dataset) and its corresponding T mean , P mean , R mean can be used to construct a linear model (i.e., the third - linear relationship): a0 = b0 + b1×T mean +b2×P mean +b3×R mean .

[0073] The a1 dataset (i.e., the first - intercept dataset) and its corresponding T mean , P mean , R mean can be used to construct a linear model (i.e., the fourth - linear relationship): a1 = b0 + b1×T mean +b2×P mean +b3×R mean .

[0074] Among them, b0, b1, b2, and b3 are the parameters of the second - layer linear model respectively.

[0075] Through the embodiments of the present invention, by determining the linear relationships between climate factors (such as average temperature, average precipitation, average solar radiation) and the first coefficient and the first intercept, the impact of climate factors on the actual yield can be quantified. By determining the linear relationship between the actual yield and the maximum vegetation index, a prediction model can be established for subsequent prediction of the target yield.

[0076] According to the method for processing crop nitrogen fertilizer demand based on remote sensing provided by the present invention, the meteorological data includes temperature, precipitation, and solar radiation. Based on the maximum vegetation index and the meteorological data, the average yield of each cultivated land in multiple time periods is determined, including: Taking each time period in the multiple time periods as the target time period respectively to perform the following operations to obtain the average yield of the target time period: Based on the average temperature, average precipitation, and average solar radiation of the target time period, according to the third - linear relationship and the fourth - linear relationship, determine the target coefficient and the target intercept; Determine the average yield for the target time period based on the target coefficient, target intercept, and the maximum vegetation index for the target time period.

[0077] In an embodiment of the present invention, during actual application, after obtaining the average temperature, average precipitation, and average solar radiation for the target time period, the first slope (a0) and the first intercept (a1) can be determined through the second - layer linear model of the hierarchical linear model for crop yield prediction. Subsequently, substituting the first slope (a0) and the first intercept (a1) into the first - layer linear model of the hierarchical linear model for crop yield prediction and combining with the maximum vegetation index EVI2max, the average yield for the target time period can be determined.

[0078] In an embodiment of the present invention, the crop yields for the previous three years in the cultivated area are predicted through the hierarchical linear model for crop yield prediction based on remote sensing data. The target yield for each pixel is calculated based on the average yield - annual increase rate method. The formula is as follows: GY = (Mean(yield) (1 + 13%)) In the formula, GY is the target yield, Mean(yield) is the average of the crop yields in the previous 3 years, and 13% is the annual increase rate of the yield.

[0079] Through the target yield calculation method of the embodiment of the present invention, the hierarchical linear model for crop yield prediction based on remote sensing data improves the average yield - annual increase rate method, avoiding the problems of insufficient representation of nitrogen - sufficient plots and imperfect statistical data.

[0080] According to the method for processing crop nitrogen fertilizer requirements based on remote sensing provided by the present invention, based on the soil nutrient level and the target yield, determine the total crop nitrogen fertilizer requirement for the target crop of each cultivated land, including: Among them, represents the total crop nitrogen fertilizer requirement for the target crop, represents the soil nutrient level, represents the target yield, represents the grain nitrogen content of the target crop, represents the preset nitrogen fertilizer utilization rate.

[0081] In an embodiment of the present invention, based on the above - mentioned soil nutrient level and target yield, calculate the total crop nitrogen fertilizer requirement through the nitrogen fertilizer application rate formula to construct a lookup table. The formula is as follows: In the formula, Let \(N\) be the total fertilizer demand for the crop, \(y\) be the target yield corresponding to different soil nutrient levels. In the present invention, the relative yields corresponding to the 7 soil nutrient levels are 100%, 90%, 80%, 70%, 60%, 50%, and 25% in sequence, \(GY\) is the target yield. Let \(N_{g}\) be the nitrogen content per kilogram of grain (a fixed biological parameter representing the proportion of nitrogen in the crop grain), and \(NUE\) be the nitrogen fertilizer utilization rate (i.e., what proportion of the applied nitrogen fertilizer can be effectively utilized by the crop).

[0082] In some embodiments, the nitrogen content of the grain is determined according to the variety of the crop, and the fertilizer utilization rate is basically fixed at 40% according to experience or literature research.

[0083] Reference Figure 2 , Figure 2 FIG. Figure 2 is a technical flow chart for constructing the lookup table of the total nitrogen fertilizer demand of the crop provided by the present invention. Among them, each parameter, process or formula in the figure can refer to the above embodiments, and the present invention will not elaborate herein.

[0084] In summary, the lookup table of the total nitrogen fertilizer demand of the crop constructed by integrating remote sensing data and agronomic knowledge in the present invention is shown in Table 1: Table 1 Lookup Table of Total Nitrogen Fertilizer Demand of Crops

[0085] In the embodiments of the present invention, the soil nutrient level (\(y\)) reflects the existing nitrogen nutrient level in the soil. The higher the soil nutrient level, the more existing nitrogen nutrients in the soil, and thus the lower the demand for additional nitrogen fertilizer by the crop.

[0086] The target yield (\(GY\)) represents the expected crop yield. The higher the target yield, the corresponding increase in the demand for nitrogen fertilizer by the crop.

[0087] The nitrogen content of the grain ( ) is a biological parameter representing the proportion of nitrogen in the crop grain, and the nitrogen content of the grain is different for different crop types.

[0088] The nitrogen fertilizer utilization rate (\(NUE\)) is an efficiency parameter reflecting the effect of nitrogen fertilizer application. The higher the nitrogen fertilizer utilization rate, the more crop yield can be produced by the same amount of nitrogen fertilizer, and thus the relatively reduced amount of nitrogen fertilizer actually required.

[0089] Through the embodiments of the present invention, by comprehensively considering the soil nutrient level, target yield, nitrogen content of the grain and nitrogen fertilizer utilization rate, the nitrogen fertilizer demand of each cultivated land can be accurately calculated, which helps to improve agricultural production efficiency and reduce production costs.

[0090] Through the above embodiments of the present invention, the present invention realizes the calculation of the high spatial resolution of the target yield (GY) based on remote sensing data and hierarchical linear models, which is more conducive to the calculation of the total nitrogen demand of crops at the field block scale. The present invention does not rely on a large number of field experiments, especially the experimental results of nitrogen-sufficient plots, and can calculate the accurate total nitrogen demand before crop planting. The present invention constructs a lookup table, simplifies the operation process in the actual production process, and is more conducive to popularization and application in agricultural practice.

[0091] The present invention proposes a method for constructing a lookup table for the total nitrogen demand of crops by integrating remote sensing data and agronomic knowledge, which can accurately determine the total nitrogen demand of cultivated land before crop sowing, and has the following advantages: Field heterogeneity: It can use the characteristics of wide coverage, periodicity, and multi-scale of remote sensing data to calculate the high spatial resolution of the target yield, effectively avoiding the problem of coarse resolution in traditional fertilization decision-making. Timeliness: Using the hierarchical linear yield estimation model and the algorithm of average yield - annual increment rate can avoid the dependence of target yield calculation on nitrogen-sufficient plots. High efficiency: When constructing the hierarchical linear yield estimation model, the method based on literature research can efficiently establish the hierarchical linear yield estimation model, efficiently calculate the target yield at the field block scale, and avoid the disadvantages of consuming a large amount of manpower, material resources, and strong subjectivity in a large number of nitrogen gradient field experiments and manual statistics of field historical yields.

[0092] The following describes the device for processing crop nitrogen fertilizer demand based on remote sensing provided by the present invention. The device for processing crop nitrogen fertilizer demand based on remote sensing described below can be mutually referred to the method for processing crop nitrogen fertilizer demand based on remote sensing described above.

[0093] Reference Figure 3 , Figure 3 is a schematic structural diagram of the device for processing crop nitrogen fertilizer demand based on remote sensing provided by the present invention.

[0094] The division module 301 is used to equally divide multiple cultivated lands according to the relative yield of the target crop to obtain the soil nutrient grade of each cultivated land among the multiple cultivated lands; The acquisition module 302 is used to acquire remote sensing data images of multiple cultivated lands within multiple time periods; The vegetation index module 303 is used to determine the maximum vegetation index of each cultivated land in each time period among multiple time periods based on the remote sensing data image; The average yield module 304 is used to determine the average yield of each cultivated land within multiple time periods based on the maximum vegetation index and meteorological data; The target yield module 305 is used to determine the product of the average yield and the preset annual yield increment rate as the target yield of each cultivated land; The total demand module 306 is configured to determine the total demand for crop nitrogen fertilizer of the target crop for each cultivated land based on the soil nutrient level and the target yield.

[0095] The device for processing crop nitrogen fertilizer demand based on remote sensing provided by the present invention can accurately evaluate the soil nutrient status of each cultivated land by analyzing the relative yield of the target crop and equally dividing multiple cultivated lands at equal intervals; by using remote sensing technology to obtain the cultivated land image data in multiple time periods, it can monitor the vegetation growth status in real time, and by analyzing the maximum vegetation index, it can indirectly reflect the growth potential and health status of the crop; by combining the maximum vegetation index with meteorological data, it can more accurately predict the average yield of each cultivated land in multiple time periods, taking into account the natural environmental factors of crop growth, making the yield prediction closer to the actual situation; by determining the product of the average yield and the preset annual yield growth rate as the target yield for each cultivated land, it takes into account the historical yield level and the expected future yield growth; based on the soil nutrient level and the target yield, it can more accurately calculate the total demand for crop nitrogen fertilizer required for each cultivated land. Thus, through precise fertilization, it can not only meet the growth needs of the crop, but also avoid environmental pollution and cost waste caused by excessive fertilization.

[0096] Specifically, the device for processing crop nitrogen fertilizer demand based on remote sensing provided by the present invention can implement all the method steps realized by the above-mentioned method embodiment for processing crop nitrogen fertilizer demand based on remote sensing, and can achieve the same technical effects. The same parts and beneficial effects as those in the method embodiment will not be specifically described in this embodiment.

[0097] Figure 4 is a schematic structural diagram of the entity of the electronic device provided by the present invention, as Figure 4As shown in the figure, the electronic device may include: a processor 410, a communications interface 420, a memory 430, and a communication bus 440. Among them, the processor 410, the communications interface 420, and the memory 430 complete communication with each other through the communication bus 440. The processor 410 may call the logical instructions in the memory 430 to execute a method for processing the nitrogen fertilizer demand of crops based on remote sensing. The method includes: equally spacing and dividing multiple cultivated lands according to the relative yield of the target crop to obtain the soil nutrient grade of each cultivated land among the multiple cultivated lands; acquiring remote sensing data images of the multiple cultivated lands in multiple time periods; based on the remote sensing data images, determining the maximum vegetation index of each cultivated land in each time period of the multiple time periods; based on the maximum vegetation index and meteorological data, determining the average yield of each cultivated land in the multiple time periods; multiplying the average yield by the preset annual yield increase rate to determine the target yield of each cultivated land; based on the soil nutrient grade and the target yield, determining the total nitrogen fertilizer demand of the target crop of each cultivated land.

[0098] In addition, when the logical instructions in the above-mentioned memory 430 can be implemented in the form of software functional units and sold or used as an independent product, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0099] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the remote sensing-based crop nitrogen fertilizer demand processing method provided by the above-mentioned various methods. The method includes: equally spacing and dividing multiple cultivated lands according to the relative yield of the target crop to obtain the soil nutrient grade of each cultivated land among the multiple cultivated lands; acquiring remote sensing data images of the multiple cultivated lands in multiple time periods; based on the remote sensing data images, determining the maximum vegetation index of each cultivated land in each time period of the multiple time periods; based on the maximum vegetation index and meteorological data, determining the average yield of each cultivated land in the multiple time periods; multiplying the average yield by the preset annual yield growth rate to determine the target yield of each cultivated land; based on the soil nutrient grade and the target yield, determining the total crop nitrogen fertilizer demand of the target crop of each cultivated land.

[0100] In another aspect, the present invention also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it is implemented to execute the remote sensing-based crop nitrogen fertilizer demand processing method provided by the above-mentioned various methods. The method includes: equally spacing and dividing multiple cultivated lands according to the relative yield of the target crop to obtain the soil nutrient grade of each cultivated land among the multiple cultivated lands; acquiring remote sensing data images of the multiple cultivated lands in multiple time periods; based on the remote sensing data images, determining the maximum vegetation index of each cultivated land in each time period of the multiple time periods; based on the maximum vegetation index and meteorological data, determining the average yield of each cultivated land in the multiple time periods; multiplying the average yield by the preset annual yield growth rate to determine the target yield of each cultivated land; based on the soil nutrient grade and the target yield, determining the total crop nitrogen fertilizer demand of the target crop of each cultivated land.

[0101] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative labor.

[0102] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0103] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for processing crop nitrogen fertilizer demand based on remote sensing, characterized in that: include: Dividing a plurality of cultivated lands into equal intervals according to the relative yield of target crops, and obtaining the soil nutrient grade of each of the plurality of cultivated lands; Acquire remote sensing data images of the plurality of cultivated lands within a plurality of time periods; Determining the maximum vegetation index of each of the cultivated lands in each of the multiple time periods based on the remote sensing data image; Determining the average yield of each cultivated land in the multiple time periods based on the maximum vegetation index and the meteorological data; The product of the average yield and the preset annual yield increase rate is determined as the target yield of each cultivated land; Based on the soil nutrient level and the target yield, the total crop nitrogen fertilizer requirement of the target crop in each arable land is determined.

2. The method for processing crop nitrogen fertilizer demand based on remote sensing according to claim 1, characterized in that: The step of dividing the plurality of cultivated lands into equal intervals according to the relative yield of the target crop to obtain the soil nutrient level of each of the plurality of cultivated lands comprises: Obtaining the soil nitrogen content and relative yield of target crops in each of a plurality of cultivated lands; determining a first linear relationship between the soil nitrogen content and the relative yield; Based on the inverse function of the first linear relationship, the soil nutrient level of each cultivated land is determined.

3. The method for processing crop nitrogen fertilizer demand based on remote sensing according to claim 1, characterized in that: The step of determining the maximum vegetation index of each cultivated land in each time period of the multiple time periods based on the remote sensing data image includes: Determine the vegetation index of each pixel based on the remote sensing data images of the plurality of cultivated lands in a plurality of time periods; Based on the vegetation index of each pixel, construct a time series data set in time order, wherein each time series of the time series data set corresponds to the pixel respectively; The time series corresponding to each pixel is preprocessed by time series harmonic analysis method to obtain the maximum vegetation index of each cultivated land in each time period of the multiple time periods.

4. The method for processing crop nitrogen fertilizer demand based on remote sensing according to claim 1, characterized in that: Before determining the average yield of each cultivated land in the multiple time periods based on the maximum vegetation index and the meteorological data, the method further includes: Obtaining the actual yield of each of the cultivated lands in each time period in a preset time period, and the maximum vegetation index of each of the cultivated lands in each time period in the preset time period, wherein the preset time period is the target growth period of the target crop; Determining a second linear relationship between the actual yield and the maximum vegetation index, wherein the second linear relationship includes: a first coefficient and a first intercept; Obtaining a first coefficient data set, a first intercept data set, and an average temperature, an average precipitation, and an average solar radiation in each of the preset time periods; determining a third linear relationship between the first coefficient data set, the average temperature, the average precipitation, and the average solar radiation; A fourth linear relationship is determined between the first intercept data set, the average air temperature, the average precipitation, and the average solar radiation.

5. The method for processing crop nitrogen fertilizer demand based on remote sensing according to claim 4, characterized in that: The meteorological data includes temperature, precipitation and solar radiation. The average yield of each cultivated land in the multiple time periods is determined based on the maximum vegetation index and the meteorological data, including: The following operations are performed on each of the multiple time periods as a target time period to obtain the average output of the target time period: Based on the average temperature, average precipitation and average solar radiation in the target time period, and according to the third linear relationship and the fourth linear relationship, determine a target coefficient and a target intercept; An average yield of the target time period is determined based on the target coefficient, the target intercept, and the maximum vegetation index of the target time period.

6. The method for processing crop nitrogen fertilizer demand based on remote sensing according to claim 1, characterized in that: The determining of the total nitrogen fertilizer requirement of the target crop of each cultivated land based on the soil nutrient grade and the target yield includes: in, represents the total crop nitrogen fertilizer requirement of the target crop, represents the soil nutrient level, represents the target yield, represents the nitrogen content of the grain of the target crop, Indicates the preset nitrogen utilization rate.

7. A crop nitrogen fertilizer demand processing device based on remote sensing, characterized in that: include: A division module, used for dividing a plurality of cultivated lands into equal intervals according to the relative yield of target crops, and obtaining the soil nutrient level of each of the plurality of cultivated lands; An acquisition module, used for acquiring remote sensing data images of the plurality of cultivated lands within a plurality of time periods; A vegetation index module, used for determining the maximum vegetation index of each cultivated land in each time period of the multiple time periods based on the remote sensing data image; An average yield module, used for determining the average yield of each cultivated land in the multiple time periods based on the maximum vegetation index and meteorological data; A target yield module, used to determine the target yield of each cultivated land by multiplying the average yield by a preset annual yield increase rate; The total demand module is used to determine the total demand for crop nitrogen fertilizer of the target crop in each cultivated land based on the soil nutrient level and the target yield.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the remote sensing-based crop nitrogen fertilizer demand processing method as described in any one of claims 1 to 6 is implemented.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the remote sensing-based crop nitrogen fertilizer demand processing method as described in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the remote sensing-based crop nitrogen fertilizer demand processing method as described in any one of claims 1 to 6 is implemented.

Images