Plant nitrogen nutrition determination method and system based on multispectral unmanned aerial vehicle platform

Abstract

The invention relates to the technical field of plant development, and discloses a plant nitrogen nutrition determination method and system based on a multispectral unmanned aerial vehicle platform, and the method comprises the steps: obtaining an image set, calculating a first nitrogen value, calculating a second nitrogen value, calculating a third nitrogen value, and outputting a plant nitrogen nutrition determination value. The system corresponds to the method. According to the invention, accurate and efficient determination of the nitrogen nutrition status of the plant is realized; a rich data basis is provided for comprehensive analysis by obtaining an image set with a time sequence; using a multispectral analysis knowledge base and a growth vigor analysis knowledge base to deeply mine information from spectrum and growth vigor dimensions, and calculating a first nitrogen value and a second nitrogen value; and a more accurate third nitrogen value is obtained based on a change capture mechanism, a capture spectrum, a growth vigor change trend and a corrected nitrogen value, so that the hysteresis and singleness problems of a traditional method are effectively overcome, the fertilization accuracy is remarkably improved, resource waste and environmental pollution risks are reduced, and high quality and high yield of crops are guaranteed.

Description

Technical Field

[0001] The present application relates to the technical field of plant development, and specifically to a method and system for measuring plant nitrogen nutrition based on a multispectral drone platform. Background Art

[0002] In modern agricultural production, accurately measuring the nitrogen nutrition status of plants is of crucial significance for rational fertilization, improving crop yield and quality, and reducing environmental pollution. Traditional methods for measuring plant nitrogen mainly rely on laboratory chemical analysis. Although this method has high precision, it has obvious limitations, specifically manifested as follows:

[0003] 1. The process is cumbersome, requiring a large number of plant samples to be collected and sent to the laboratory for complex processing and testing. This not only consumes a large amount of manpower, material resources, and time, but also the test results have obvious lag, unable to timely reflect the real-time dynamic changes of plant nitrogen, resulting in difficulty in accurately fertilizing according to the actual needs of plants, often resulting in over-fertilization or under-fertilization, which not only increases production costs but also may cause pollution to the soil and water environment;

[0004] 2. With the development of remote sensing technology, methods for monitoring plant nitrogen based on satellite remote sensing have gradually emerged. However, the spatial resolution of satellite remote sensing is relatively low, difficult to meet the needs of small-scale farmland or precision agriculture, and is greatly affected by atmospheric factors such as clouds, and the timeliness and accuracy of data acquisition are limited;

[0005] 3. The emergence of drone remote sensing technology provides a possibility to solve this problem, but most of the existing methods for measuring plant nitrogen based on drones only rely on single spectral data or simple image analysis, lacking comprehensive consideration of complex environmental factors and growth dynamic changes during the plant growth process. For example, the phenological period of plants and local meteorological conditions are not fully combined for data collection, resulting in insufficient accuracy and representativeness of spectral data; in the process of calculating nitrogen values, multi-source data and knowledge are not systematically integrated, unable to accurately capture the real situation of plant nitrogen, and unable to provide a reliable decision-making basis for precision agriculture.

[0006] Chinese Patent No. CN119147483A discloses a drip-irrigated wheat water and nitrogen intelligent monitoring system based on drone multispectral inversion, but this invention has problems such as data dependence, model limitations, incomplete consideration of environmental factors, difficult system integration, high implementation cost, and the need to balance cost-effectiveness, which limit its application and promotion in different scenarios and need further improvement and optimization.

[0007] In summary, there is an urgent need for a new, more accurate, efficient, and comprehensive technical solution for measuring plant nitrogen nutrition to overcome these problems. Summary of the Invention

[0008] The purpose of this application is to provide a method and system for measuring plant nitrogen nutrition based on a multispectral UAV platform to solve the technical problems presented in the above background art.

[0009] To achieve the above objective, this application discloses the following technical solutions:

[0010] In the first aspect, this application discloses a method for measuring plant nitrogen nutrition based on a multispectral UAV platform, and the method includes:

[0011] S1: Obtain an image set with time series for plant nitrogen nutrition measurement by using a multispectral UAV platform; wherein, the image set at least includes the multispectral image and the growth image of the plant.

[0012] S2: Analyze the multispectral image by using a preset multispectral analysis knowledge base and calculate a first nitrogen value; wherein, the multispectral analysis knowledge base is used to analyze the multispectral image based on the preset multispectral analysis knowledge.

[0013] S3: Analyze the growth image by using a preset growth analysis knowledge base, correct the first nitrogen value based on the result of this analysis, and calculate a second nitrogen value; wherein, the growth analysis knowledge base is used to analyze the growth image based on the preset growth analysis knowledge.

[0014] S4: Capture the spectral change trend of the multispectral image and the growth change trend of the growth image based on a preset change capture mechanism, correct the second nitrogen value based on the spectral change trend and the growth change trend, and calculate a third nitrogen value; wherein, the change capture mechanism is used to extract the spectral features of the multispectral image and the growth features of the growth image, and obtain the spectral change trend and the growth change trend based on the spectral features and the growth features.

[0015] S5: Output the third nitrogen value and define it as the plant nitrogen nutrition measurement value.

[0016] Preferably, the process of obtaining the image set includes:

[0017] During the plant growth period, in combination with the phenological period of the plant and the local meteorological conditions, plan and execute the shooting task based on the preset dynamic programming rules to obtain the image set; wherein, the dynamic programming rules at least include temperature planning rules, light planning rules, height planning rules, and terrain planning rules. The temperature planning rules are used to control the shooting frequency based on different temperature conditions, the light planning rules are used to control the shooting angle based on different light conditions, the height planning rules are used to control the shooting height based on different plant height conditions, and the terrain planning rules are used to control the shooting path based on different plant growth terrain conditions.

[0018] Preferably, the multispectral analysis knowledge base is specifically as follows:

[0019] Based on long-term multispectral monitoring and nitrogen analysis experiments on plant samples under different plant species, different soil conditions and climatic environments, the multispectral analysis knowledge is obtained;

[0020] Based on the specific growth stage of each plant, a functional relationship between the nitrogen value and the reflectance and its derivative of the spectral band is constructed. The specific functional relationship is as follows:

[0021]

[0022] where R x is the reflectance of the specific band x, and R' x (λ) is the first derivative of the specific band x at the specific wavelength λ, R' y (λ) is the first derivative of the specific band y at the specific wavelength λ, a, b and c are coefficients obtained by fitting based on long-term multispectral monitoring and nitrogen analysis experiments, and N is the corresponding nitrogen value.

[0023] Preferably, the calculation process of the first nitrogen value includes:

[0024] Obtain the multispectral image and perform preprocessing;

[0025] Based on the functional relationship between the nitrogen value and the reflectance and its derivative of the spectral band, calculate the first nitrogen value, specifically:

[0026]

[0027] where R' x (λ 1 ) is the first derivative of the specific band x at the specific wavelength λ 1 obtained from the multispectral image, R' y (λ 2 ) is the first derivative of the specific band y at the specific wavelength λ 2 obtained from the multispectral image, and N 1 is the calculated first nitrogen value.

[0028] Preferably, the growth trend analysis knowledge base is specifically as follows:

[0029] Based on the preset plant growth knowledge and long-term field observation data, the growth trend correlation relationship between the change rate of morphological characteristics and the change trend of physiological indexes of plants at different growth stages and the nitrogen value is obtained; wherein, the change rate of morphological characteristics is used to characterize the change law of the morphology of plants, and the change trend of physiological indexes is used to characterize the change law of the physiological indexes of plants.

[0030] Preferably, the calculation process of the second nitrogen value includes:

[0031] Extracting the characteristic values of the morphological and physiological changes of the plant by using image analysis and data processing algorithms, and comparing them with the growth-related relationships in the growth analysis knowledge base;

[0032] When the first nitrogen value is N 1 The calculation of the second nitrogen value is specifically:

[0033] When it is determined based on the comparison that the plant is nitrogen-deficient, a correction coefficient k is determined based on the degree of deficiency 1 , and the calculation formula of the second nitrogen value at this time is N 2 = k 1 * N 1 ;

[0034] When it is determined based on the comparison that the plant is nitrogen-excessive, a correction coefficient k is determined based on the degree of excess 2 , and the calculation formula of the second nitrogen value at this time is N 2 = k 2 * N 1 ;

[0035] When it is determined based on the comparison that the plant is nitrogen-normal, it is determined that the correction coefficient k = 1, and the calculation formula of the second nitrogen value at this time is N 2 = k * N 1 .

[0036] Preferably, the change capture mechanism is specifically:

[0037] Based on the spectral characteristics, calculate the difference in spectral reflectance between adjacent time points and the corresponding differences in the first derivative and the second derivative, and construct a spectral feature vector V s , and obtain a spectral nitrogen value correction model N based on spectral changes based on the spectral feature vector s = f(V s ), where f() is a non-linear function trained by historical spectral characteristics;

[0038] Based on the growth characteristics, calculate the difference in the change rate of the morphological and physiological indexes of the plant, and construct a growth feature vector V g , and obtain a growth nitrogen value correction model N based on growth changes based on the growth feature vector g = g(V g ), where g() is a non-linear function trained by historical growth characteristics.

[0039] Preferably, the calculation process of the third nitrogen value includes:

[0040] When the second nitrogen value is N 2 the nitrogen content correction value caused by the change in the spectral characteristics obtained based on the change capture mechanism is N s and the nitrogen content correction value caused by the change in the growth characteristics is N g then the calculation formula for the third nitrogen value at this time is:

[0041] N 3 = N 2 + N s + N g

[0042] where N 3 is the calculated third nitrogen value.

[0043] Preferably, the method further includes:

[0044] S6: Indexed by plant species, growth stage, and environmental parameters, classify and store the calculation process and related data of the first nitrogen value, the second nitrogen value, and the third nitrogen value; use an artificial intelligence algorithm to deeply mine the stored data, analyze the relationship between each parameter and the nitrogen value; configure a general algorithm framework based on the result of analyzing the relationship between each parameter and the nitrogen value, and automatically adapt and call the corresponding parameters and models based on the newly input plant and environmental data to calculate the nitrogen value.

[0045] In a second aspect, the present application discloses a plant nitrogen nutrition determination system based on a multispectral drone platform. This system is applicable to the plant nitrogen nutrition determination method based on a multispectral drone platform as described above. The system includes an image set acquisition module, a first nitrogen value calculation module, a second nitrogen value calculation module, a third nitrogen value calculation module, and a plant nitrogen nutrition determination value output module connected in sequence;

[0046] The image set acquisition module is configured to: use a multispectral drone platform to acquire an image set with time series for plant nitrogen nutrition determination; wherein, the image set at least includes the multispectral image and the growth image of the plant;

[0047] The first nitrogen value calculation module is configured to: analyze the multispectral image using a preset multispectral analysis knowledge base to calculate the first nitrogen value; wherein, the multispectral analysis knowledge base is used to analyze the multispectral image based on preset multispectral analysis knowledge;

[0048] The second nitrogen value calculation module is configured to: analyze the growth image using a preset growth analysis knowledge base, correct the first nitrogen value based on the result of this analysis, and calculate the second nitrogen value; wherein, the growth analysis knowledge base is used to analyze the growth image based on preset growth analysis knowledge;

[0049] The third nitrogen value calculation module is configured to: capture the spectral change trend of the multispectral image and the growth change trend of the growth image based on a preset change capture mechanism, correct the second nitrogen value based on the spectral change trend and the growth change trend, and calculate the third nitrogen value; wherein, the change capture mechanism is used to extract the spectral features of the multispectral image and the growth features of the growth image, and obtain the spectral change trend and the growth change trend based on the spectral features and the growth features;

[0050] The plant nitrogen nutrition measurement value output module is configured to: output the third nitrogen value and define it as the plant nitrogen nutrition measurement value.

[0051] Beneficial effects: The plant nitrogen nutrition measurement method and system based on the multispectral UAV platform of the present application utilize the multispectral UAV platform combined with the measurement method to achieve accurate and efficient measurement of the plant nitrogen nutrition status; by obtaining an image set with time series, including multispectral images and growth images, it provides a rich data basis for comprehensive analysis; using the multispectral analysis knowledge base and the growth analysis knowledge base, it deeply excavates information from the spectral and growth dimensions respectively, calculates the first nitrogen value and the second nitrogen value, and initially determines the nitrogen content range; then based on the change capture mechanism, it captures the spectral and growth change trends, corrects the nitrogen value to obtain a more accurate third nitrogen value, effectively overcoming the lag and single problem of traditional methods, providing a reliable basis for precision agriculture decision-making, significantly improving the fertilization accuracy, reducing the risk of resource waste and environmental pollution, and ensuring high-quality and high-yield crops. Description of the Drawings

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present application. For those skilled in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0053] Figure 1 It is a flowchart of the plant nitrogen nutrition measurement method based on the multispectral UAV platform provided by the embodiment of the present application;

[0054] Figure 2 It is a structural block diagram of the plant nitrogen nutrition measurement system based on the multispectral UAV platform provided by the embodiment of the present application. Detailed Embodiments

[0055] The technical solutions in the embodiments of the present application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0056] In this document, the term "comprising" is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, the elements defined by the statement "comprising..." do not exclude the presence of additional identical elements in the process, method, article or device comprising the said elements.

[0057] The first aspect of this embodiment discloses a method for measuring plant nitrogen nutrition based on a multispectral drone platform as Figure 1 shown, and the method includes:

[0058] S1: Using the multispectral drone platform to obtain an image set with time series for plant nitrogen nutrition measurement; wherein, the image set at least includes the multispectral image and the growth image of the plant;

[0059] S2: Using a preset multispectral analysis knowledge base to analyze the multispectral image and calculate a first nitrogen value; wherein, the multispectral analysis knowledge base is used to analyze the multispectral image based on the preset multispectral analysis knowledge;

[0060] S3: Using a preset growth analysis knowledge base to analyze the growth image, and correcting the first nitrogen value based on the result of the analysis to calculate a second nitrogen value; wherein, the growth analysis knowledge base is used to analyze the growth image based on the preset growth analysis knowledge;

[0061] S4: Based on a preset change capture mechanism, capturing the spectral change trend of the multispectral image and the growth change trend of the growth image, and correcting the second nitrogen value based on the spectral change trend and the growth change trend to calculate a third nitrogen value; wherein, the change capture mechanism is used to extract the spectral features of the multispectral image and the growth features of the growth image, and obtain the spectral change trend and the growth change trend based on the spectral features and the growth features;

[0062] S5: Output the third nitrogen value and define it as the plant nitrogen nutrition measurement value.

[0063] Through the above, this embodiment utilizes a multispectral drone platform combined with a measurement method to achieve precise and efficient measurement of the nitrogen nutritional status of plants; by obtaining an image set with time series, including multispectral images and growth images, it provides a rich data basis for comprehensive analysis; using a multispectral analysis knowledge base and a growth analysis knowledge base, it deeply mines information from the spectral and growth dimensions respectively, calculates the first nitrogen value and the second nitrogen value, and preliminarily determines the nitrogen content range; then based on the change capture mechanism, it captures the spectral and growth change trends, corrects the nitrogen value to obtain a more accurate third nitrogen value, effectively overcoming the lag and single problems of traditional methods, providing a reliable basis for precision agriculture decision-making, significantly improving the fertilization accuracy, reducing the risks of resource waste and environmental pollution, and ensuring high-quality and high-yield crops.

[0064] Specifically, the process of obtaining the image set includes:

[0065] During the plant growth period, in combination with the phenological period of the plant and the local meteorological conditions, the shooting task is planned and executed based on preset dynamic programming rules to obtain the image set; among them, the dynamic programming rules at least include a temperature planning rule, a light planning rule, a height planning rule, and a terrain planning rule. The temperature planning rule is used to control the shooting frequency based on different temperature conditions, the light planning rule is used to control the shooting angle based on different light conditions, the height planning rule is used to control the shooting height based on different plant height conditions, and the terrain planning rule is used to control the shooting path based on different plant growth terrain conditions.

[0066] In a specific example, during the rapid growth stage of the plant (such as the vigorous vegetative growth period), when the average temperature for 3 consecutive days is between 20 - 30 °C and the sunshine duration exceeds 8 hours, the shooting frequency is increased to 2 times a day, and the shooting is carried out at 9 - 11 am and 3 - 5 pm respectively; during the relatively slow growth stage (such as the dormant period or low-temperature period), it is shot once every 7 days, and the shooting is selected at the noon period with higher temperature (12 - 2 pm). When shooting, the drone platform adjusts the flight height according to the terrain and plant height changes, keeping the relative height with the top of the plant canopy between 50 - 100 meters. At the same time, using the inertial navigation and satellite positioning system, ensure that the horizontal accuracy of the shooting position is within 0.5 meters and the vertical accuracy is within 0.3 meters, and obtain an image set with time series to ensure the accuracy and representativeness of the image data.

[0067] With the above, this embodiment utilizes the preset dynamic programming rules to optimize the process of acquiring the image set according to the phenological period and meteorological conditions during the plant growth period. Based on the temperature planning rule, light planning rule, height planning rule, and terrain planning rule, it is possible to accurately control the shooting frequency, angle, height, and path in different environments, ensuring that the acquired image data has high quality and is representative. This not only improves the accuracy of spectral data but also provides a more reliable data source for subsequent nitrogen value calculation, enhances the adaptability of the entire measurement method to complex environments, makes the nitrogen nutrition measurement more accurate, and strongly supports precision agriculture practices.

[0068] Specifically, the multispectral analysis knowledge base is specifically as follows:

[0069] Based on long-term multispectral monitoring and nitrogen analysis experiments on plant samples under different plant species, different soil conditions, and climatic environments, multispectral analysis knowledge is obtained;

[0070] Based on the specific growth stage of each plant, a functional relationship of nitrogen value is constructed based on the spectral band reflectance and its derivative, and this functional relationship is specifically:

[0071]

[0072] where, R x is the reflectance of the specific band x, R' x (λ) is the first derivative of the specific band x at the specific wavelength λ, R' y (λ) is the first derivative of the specific band y at the specific wavelength λ, a, b, and c are coefficients obtained by fitting based on long-term multispectral monitoring and nitrogen analysis experiments, and N is the corresponding nitrogen value. In a specific example, the specific bands in this embodiment can be but are not limited to the blue light band, near-infrared band, etc.

[0073] With the above, this embodiment utilizes the multispectral analysis knowledge base constructed based on long-term multispectral monitoring and nitrogen analysis experiments to achieve accurate nitrogen value calculation for different plant species and growth stages. By establishing a functional relationship based on the spectral band reflectance and its derivative, the key information in the multispectral image can be accurately extracted, and the change in nitrogen content can be accurately reflected. The coefficients obtained by fitting based on a large amount of experimental data make the calculation of the first nitrogen value more scientific and more in line with the actual situation, lay a solid foundation for subsequent nitrogen nutrition measurement, improve the accuracy and reliability of the measurement method, and contribute to accurately guiding the fertilization management in agricultural production.

[0074] Specifically, the calculation process of the first nitrogen value includes:

[0075] Obtain the multispectral image and perform preprocessing;

[0076] Based on the functional relationship between the reflectance and its derivative of the spectral band and the nitrogen value, calculate the first nitrogen value, specifically:

[0077]

[0078] Wherein, R' x (λ 1 ) is the first derivative at a specific wavelength λ of a specific band x obtained from the multispectral image, and R' 1 (λ y ) is the first derivative at a specific wavelength λ of a specific band y obtained from the multispectral image, and N 2 2 1 1 2 1 2

[0079] is the calculated first nitrogen value.

[0079] Through the above, this embodiment realizes the accurate calculation of the first nitrogen value by using the preprocessing of the multispectral image and the functional relationship based on the reflectance and its derivative of the spectral band. By acquiring the multispectral image and performing preprocessing to remove noise and interference, the data quality is guaranteed, and then substituting it into the calculation of the first nitrogen value, fully considering the influence of spectral characteristics and their changes on the nitrogen value. This enables the first nitrogen value to more accurately reflect the initial nitrogen status of plants, provides a reliable starting point for subsequent correction and optimization, improves the accuracy and stability of the entire measurement method, and provides strong data support for precision agriculture.

[0080] Specifically, the growth trend analysis knowledge base is specifically:

[0081] Based on the preset plant growth knowledge and long-term field observation data, the growth trend correlation relationship between the change rate of morphological characteristics and the change trend of physiological indexes of plants at different growth stages and the nitrogen value is obtained; among them, the change rate of morphological characteristics is used to characterize the change law of the morphology of plants, and the change trend of physiological indexes is used to characterize the change law of the physiological indexes of plants.

[0082] In a specific example, the change rate of morphological characteristics in this embodiment can be indirectly estimated through spectral characteristics such as the leaf growth rate, stem thickening rate, etc., and the change trend of physiological indexes can be such as the change rate of chlorophyll content, the change trend of photosynthesis efficiency, etc. Specifically, for example, for wheat at the jointing stage, under normal circumstances, the daily growth rate of the leaf area should be between A 1 -A 2 , and the relative change rate of chlorophyll content is within the range of C 1 -C 2 . If it exceeds this range, it indicates that there may be an abnormality in nitrogen nutrition.

[0083] With the above, the growth trend analysis knowledge base constructed by the present embodiment using preset plant growth knowledge and long-term field observation data realizes the effective correction of the first nitrogen value. By analyzing the correlation between the change rate of morphological characteristics and the change trend of physiological indexes of plants at different growth stages and the nitrogen value, the nitrogen nutrition status can be accurately judged according to the actual growth situation. When compared with the first nitrogen value, the correction coefficient is determined based on the degree of deficiency or excess, so as to obtain a second nitrogen value that is more in line with the actual situation, improve the accuracy of nitrogen determination, avoid errors caused by single spectral analysis, and provide a more accurate basis for precise fertilization.

[0084] Specifically, the calculation process of the second nitrogen value includes:

[0085] Using image analysis and data processing algorithms to extract the characteristic values of the morphological and physiological changes of plants, and comparing them with the growth trend correlation relationship in the growth trend analysis knowledge base;

[0086] When the first nitrogen value is N 1 The calculation of the second nitrogen value is specifically as follows:

[0087] When it is judged based on the comparison that the plant is nitrogen-deficient, the correction coefficient k 1 is determined based on the degree of deficiency. At this time, the calculation formula of the second nitrogen value is N 2 =k 1 *N 1 ;

[0088] When it is judged based on the comparison that the plant is nitrogen-excessive, the correction coefficient k 2 is determined based on the degree of excess. At this time, the calculation formula of the second nitrogen value is N 2 =k 2 *N 1 ;

[0089] When it is judged based on the comparison that the plant is nitrogen-normal, the correction coefficient k = 1 is determined. At this time, the calculation formula of the second nitrogen value is N 2 =k*N 1 .

[0090] In a specific example, the correction coefficient k 1 of the present embodiment can be, for example, k 1 =0.8 for mild deficiency, k 1 =0.6 for moderate deficiency, k 1 =0.4 for severe deficiency; the correction coefficient k 2 can be, for example, k 2 =1.2 for mild excess, k 2 =1.4 for moderate excess, k 2 =1.6 for severe excess.

[0091] With the above, this embodiment uses image analysis and data processing algorithms in combination with a growth analysis knowledge base to achieve the accurate calculation of the second nitrogen value. By extracting the morphological and physiological change characteristic values of plants and comparing them with the growth correlation relationships in the knowledge base, and according to the judgment results of nitrogen deficiency, excess or normal, using the corresponding correction coefficient calculation formula, the nitrogen value can be adjusted specifically. This makes the second nitrogen value more accurately reflect the true nitrogen demand of plants, further optimizing the nitrogen nutrition measurement results, providing more accurate guidance for reasonable fertilization in agricultural production, improving fertilizer utilization rate, and reducing resource waste.

[0092] Specifically, the change capture mechanism is specifically as follows:

[0093] Based on spectral characteristics, calculate the difference in spectral reflectance between adjacent time points and the corresponding differences in the first derivative and second derivative, and construct a spectral feature vector V s , and obtain a spectral nitrogen value correction model N based on spectral changes based on the spectral feature vector s = f(V s ), where f() is a non-linear function trained through historical spectral characteristics;

[0094] Based on growth characteristics, calculate the difference in the change rate of the morphological and physiological indicators of plants, and construct a growth feature vector V g , and obtain a growth nitrogen value correction model N based on growth changes based on the growth feature vector g = g(V g ), where g() is a non-linear function trained through historical growth characteristics.

[0095] It should be noted that the spectral feature vector V of this embodiment s can be an array formed by calculating the difference in spectral reflectance between adjacent time points and the corresponding differences in the first derivative and second derivative using the numerical values of the spectral reflectance in each band. This array is used to characterize spectral changes, so as to obtain the non-linear function f() trained through historical spectral characteristics; the growth feature vector V of this embodiment g can be an array formed by using the difference in the change rate of the morphological and physiological indicators of plants. This array is used to characterize growth changes, so as to obtain the non-linear function g() trained through historical growth characteristics.

[0096] Through the above, this embodiment utilizes the correction model constructed based on the changes in spectral characteristics and growth characteristics to achieve dynamic and accurate correction of nitrogen values. By calculating the spectral reflectance and derivative difference between adjacent time points to construct a spectral feature vector, a nitrogen value correction model based on spectral changes is obtained; by calculating the difference in the change rates of plant morphological and physiological indicators to construct a growth characteristic vector, a correction model based on growth changes is obtained. These two models can sensitively capture the change trend during the plant growth process, timely correct the nitrogen value, make the measurement result more in line with the actual dynamic change of plant nitrogen nutrition, provide real-time and accurate data support for precision agriculture decision-making, and ensure the healthy growth of crops.

[0097] Specifically, the calculation process of the third nitrogen value includes:

[0098] When there is a second nitrogen value of N 2 At this time, the nitrogen content correction value caused by the spectral feature change obtained based on the change capture mechanism is N s And the nitrogen content correction value caused by the growth characteristic change is N g Then, the calculation formula for the third nitrogen value at this time is:

[0099] N 3 = N 2 + N s + N g

[0100] Wherein, N 3 Is the calculated third nitrogen value.

[0101] Through the above, this embodiment utilizes the correction value obtained by the change capture mechanism to achieve high-precision calculation of the third nitrogen value. When there is a second nitrogen value, based on the nitrogen content correction value caused by the spectral feature change and the nitrogen content correction value caused by the growth characteristic change, the third nitrogen value is obtained through a specific calculation formula. This process fully considers various dynamic factors during the plant growth process, comprehensively and accurately adjusts the nitrogen value, makes the final measured value of plant nitrogen nutrition closer to the real situation, greatly improves the accuracy and reliability of the measurement method, and provides a solid technical guarantee for precision agriculture.

[0102] Specifically, this method further includes:

[0103] S6: Indexed based on plant species, growth stage, and environmental parameters, classify and store the calculation processes and related data of the first nitrogen value, second nitrogen value, and third nitrogen value; use artificial intelligence algorithms to deeply mine the stored data, analyze the relationship between each parameter and the nitrogen value; configure a general algorithm framework based on the results of analyzing the relationship between each parameter and the nitrogen value, and automatically adapt and call the corresponding parameters and models based on the newly input plant and environmental data to calculate the nitrogen value.

[0104] It should be noted that in this embodiment, the existing artificial intelligence algorithm is used to deeply mine and analyze the stored data and the relationship between various parameters and the nitrogen value, and the existing general algorithm framework is continued to automatically adapt and call the corresponding parameters and models based on the newly input plant and environmental data to calculate the nitrogen value.

[0105] Through the above, this embodiment realizes the improvement of the universality and adaptability of the measurement method by storing, mining and algorithm framework configuration of the nitrogen value data. By classifying and storing the nitrogen value calculation process and related data based on plant species, growth stages and environmental parameters, using artificial intelligence algorithms to deeply mine the relationship between parameters and nitrogen values, and configuring the general algorithm framework accordingly, it is possible to automatically adapt models and parameters according to new plant and environmental data for nitrogen value calculation. This makes this measurement method can be widely applied to different agricultural scenarios, improves the popularization and practicality of the technology, and provides a more convenient and efficient means for measuring nitrogen nutrition in agricultural production.

[0106] The second aspect of this embodiment discloses a Figure 2 plant nitrogen nutrition measurement system based on a multispectral unmanned aerial vehicle platform as shown. This system is applicable to the above-mentioned plant nitrogen nutrition measurement method based on a multispectral unmanned aerial vehicle platform. The system includes an image set acquisition module, a first nitrogen value calculation module, a second nitrogen value calculation module, a third nitrogen value calculation module, and a plant nitrogen nutrition measurement value output module connected in sequence;

[0107] The image set acquisition module is configured to: use a multispectral unmanned aerial vehicle platform to acquire an image set with time series for plant nitrogen nutrition measurement; wherein, the image set at least includes the multispectral image and the growth image of the plant;

[0108] The first nitrogen value calculation module is configured to: analyze the multispectral image using a preset multispectral analysis knowledge base and calculate the first nitrogen value; wherein, the multispectral analysis knowledge base is used to analyze the multispectral image based on the preset multispectral analysis knowledge;

[0109] The second nitrogen value calculation module is configured to: analyze the growth image using a preset growth analysis knowledge base, correct the first nitrogen value based on the result of this analysis, and calculate the second nitrogen value; wherein, the growth analysis knowledge base is used to analyze the growth image based on the preset growth analysis knowledge;

[0110] The third nitrogen value calculation module is configured to: capture the spectral change trend of the multispectral image and the growth change trend of the growth image based on a preset change capture mechanism, correct the second nitrogen value based on this spectral change trend and this growth change trend, and calculate the third nitrogen value; wherein, the change capture mechanism is used to extract the spectral features of the multispectral image and the growth features of the growth image, and obtain the spectral change trend and the growth change trend based on the spectral features and the growth features;

[0111] The plant nitrogen nutrition measurement value output module is configured to: output a third nitrogen value and define it as the plant nitrogen nutrition measurement value.

[0112] It should be noted that the plant nitrogen nutrition measurement system based on the multispectral UAV platform in this embodiment corresponds to the aforementioned plant nitrogen nutrition measurement method based on the multispectral UAV platform. Therefore, for the content not specifically described in the plant nitrogen nutrition measurement system based on the multispectral UAV platform in this embodiment, such as but not limited to function definition, working principle, and background technology, etc., reference can be made to the records of the aforementioned plant nitrogen nutrition measurement method based on the multispectral UAV platform, and this text will not elaborate here.

[0113] In summary, the plant nitrogen nutrition measurement method and system based on the multispectral UAV platform in this embodiment utilize the multispectral UAV platform combined with the measurement method to achieve accurate and efficient measurement of the plant nitrogen nutrition status; by obtaining an image set with time series, including multispectral images and growth images, it provides a rich data basis for comprehensive analysis; using the multispectral analysis knowledge base and the growth analysis knowledge base, it deeply excavates information from the spectral and growth dimensions respectively, calculates the first nitrogen value and the second nitrogen value, and preliminarily determines the nitrogen content range; then based on the change capture mechanism, it captures the spectral and growth change trends, corrects the nitrogen value to obtain a more accurate third nitrogen value, effectively overcomes the lag and single - ness problems of traditional methods, provides a reliable basis for precision agriculture decision - making, significantly improves the fertilization accuracy, reduces the risks of resource waste and environmental pollution, and ensures high - quality and high - yield crops.

[0114] In the embodiments provided in the present application, it should be understood that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code, or any appropriate combination thereof. For hardware implementation, the processor can be implemented in one or more of the following units: application specific integrated circuit (ASIC), digital signal processor (DSP), digital signal processing device (DSPD), programmable logic device (PLD), field programmable gate array (FPGA), processor, controller, microcontroller, microprocessor, other electronic units designed to implement the functions described herein, or a combination thereof. For software implementation, part or all of the processes of the embodiments can be completed by a computer program instructing the relevant hardware. When implemented, the above program can be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. The computer-readable storage medium includes computer storage media and communication media, where the communication media includes any medium that facilitates the transfer of a computer program from one place to another. The storage media can be any available medium that can be accessed by a computer. The computer-readable storage medium can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM, or other optical disk storage, magnetic disk storage media, or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer.

[0115] Finally, it should be noted that the above are only the preferred embodiments of the present application and are not used to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for determining plant nitrogen nutrition based on a multispectral UAV platform, characterized in that: The method includes: S1: using a multispectral UAV platform to obtain a time-series image set for plant nitrogen nutrition determination; wherein the image set includes at least a multispectral image and a growth image of the plant; S2: Analyze the multispectral image using a preset multispectral analysis knowledge base to calculate a first nitrogen value; wherein the multispectral analysis knowledge base is used to analyze the multispectral image based on preset multispectral analysis knowledge; S3: Analyze the growth image using a preset growth analysis knowledge base, correct the first nitrogen value based on the analysis result, and calculate a second nitrogen value; wherein the growth analysis knowledge base is used to analyze the growth image based on preset growth analysis knowledge; S4: Based on a preset change capture mechanism, the spectral change trend of the multispectral image and the growth change trend of the growth image are captured, and the second nitrogen value is corrected based on the spectral change trend and the growth change trend, and a third nitrogen value is calculated; wherein the change capture mechanism is used to extract the spectral features of the multispectral image and the growth features of the growth image, and obtain the spectral change trend and the growth change trend based on the spectral features and the growth features; S5: Output the third nitrogen value and define it as a plant nitrogen nutrition determination value.

2. The method for determining plant nitrogen nutrition based on a multispectral UAV platform according to claim 1, characterized in that: The process of acquiring the image set includes: During the plant growth period, in combination with the plant phenological period and local meteorological conditions, the shooting task is planned and executed based on preset dynamic programming rules to obtain an image set; wherein the dynamic programming rules include at least temperature planning rules, lighting planning rules, height planning rules and terrain planning rules, the temperature planning rules are used to control the shooting frequency based on different temperature conditions, the lighting planning rules are used to control the shooting angle based on different lighting conditions, the height planning rules are used to control the shooting height based on different plant height conditions, and the terrain planning rules are used to control the shooting path based on different plant growth terrain conditions.

3. The method for determining plant nitrogen nutrition based on a multispectral UAV platform according to claim 1, characterized in that: The multi-spectral analysis knowledge base is specifically: The multispectral analysis knowledge is obtained based on long-term multispectral monitoring and nitrogen analysis experiments on plant samples of different plant species, different soil conditions and climate environments; Based on the specific growth stage of each plant, a functional relationship based on the spectral band reflectance and its derivative with nitrogen value is constructed. The specific functional relationship is: Among them, R x is the reflectivity of a specific wavelength band x, R' x (λ) is the first-order derivative of a specific band x at a specific wavelength λ, R' y (λ) is the first derivative of a specific band y at a specific wavelength λ, a, b and c are coefficients obtained based on long-term multi-spectral monitoring and nitrogen analysis experimental fitting, and N is the corresponding nitrogen value.

4. The method for determining plant nitrogen nutrition based on a multispectral UAV platform according to claim 3, characterized in that: The calculation process of the first nitrogen value includes: Acquire the multispectral image and perform preprocessing; Based on the functional relationship of the spectral band reflectance and its derivative with the nitrogen value, the first nitrogen value is calculated, specifically: Among them, R' x (λ1) is the first-order derivative of a specific band x at a specific wavelength λ1 obtained based on the multispectral image, R' y (λ2) is the first derivative of the specific band y at the specific wavelength λ2 obtained based on the multispectral image, and N1 is the first nitrogen value obtained by calculation.

5. The method for determining plant nitrogen nutrition based on a multispectral UAV platform according to claim 1, characterized in that: The growth analysis knowledge base is specifically: Based on preset plant growth knowledge and long-term field observation data, the correlation between the change rate of morphological characteristics and the change trend of physiological indicators of plants at different growth stages and the growth potential of nitrogen values ​​is obtained; wherein the change rate of morphological characteristics is used to characterize the changing law of plant morphology, and the changing trend of physiological indicators is used to characterize the changing law of plant physiological indicators.

6. The method for determining plant nitrogen nutrition based on a multispectral UAV platform according to claim 5, characterized in that: The calculation process of the second nitrogen value includes: Extracting characteristic values ​​of morphological and physiological changes of plants by using image analysis and data processing algorithms, and comparing them with the growth potential correlation relationships in the growth potential analysis knowledge base; When the first nitrogen value is N1, the calculation of the second nitrogen value is specifically as follows: When the plant is judged to be nitrogen deficient based on the comparison, the correction coefficient k1 is determined based on the degree of deficiency, and the calculation formula of the second nitrogen value at this time is N2=k1*N1; When the plant is judged to have excessive nitrogen based on the comparison, the correction coefficient k2 is determined based on the degree of excess, and the calculation formula of the second nitrogen value at this time is N2=k2*N1; When it is determined based on the comparison that the nitrogen content of the plant is normal, the correction coefficient k=1 is determined, and the calculation formula for the second nitrogen value is N2=k*N1.

7. The method for determining plant nitrogen nutrition based on a multispectral UAV platform according to claim 1, characterized in that: The change capture mechanism is specifically: Based on the spectral features, the spectral reflectance difference between adjacent time points and the corresponding first-order derivative and second-order derivative difference are calculated to construct a spectral feature vector V s Based on the spectral feature vector, a spectral nitrogen value correction model N based on spectral changes is obtained. s =f(V s ), where f() is a nonlinear function obtained by training historical spectral features; Based on the growth characteristics, the difference in the change rate of the plant's morphological and physiological indicators is calculated to construct a growth characteristic vector V g Based on the growth potential characteristic vector, a growth potential nitrogen value correction model N based on growth potential changes is obtained. g =g(V g ), where g() is a nonlinear function trained through historical growth characteristics.

8. The method for determining plant nitrogen nutrition based on a multispectral UAV platform according to claim 7, characterized in that: The calculation process of the third nitrogen value includes: When the second nitrogen value is N2, the nitrogen content correction value caused by the change in the spectral feature obtained based on the change capture mechanism is N s The nitrogen content correction value caused by the change in growth characteristics is N g , then the calculation formula of the third nitrogen value at this time is: <h2 style=";text-align:left;direction:ltr">N3 = N2 + N<h2 style=";text-align:left;direction:ltr"> s <h2 style=";text-align:left;direction:ltr"> +N<h2 style=";text-align:left;direction:ltr"> g Wherein, N3 is the third nitrogen value obtained by calculation.

9. The method for determining plant nitrogen nutrition based on a multispectral UAV platform according to claim 1, characterized in that: The method further includes: S6: Based on the plant species, growth stage and environmental parameters as indexes, the calculation process and related data of the first nitrogen value, the second nitrogen value and the third nitrogen value are classified and stored; artificial intelligence algorithms are used to deeply mine the stored data and analyze the relationship between each parameter and the nitrogen value; a general algorithm framework is configured based on the results of the analysis of the relationship between each parameter and the nitrogen value, and based on the newly input plant and environmental data, the corresponding parameters and models are automatically adapted and called to calculate the nitrogen value.

10. A plant nitrogen nutrition determination system based on a multispectral UAV platform, the system being applicable to the plant nitrogen nutrition determination method based on a multispectral UAV platform as described in any one of claims 1 to 9, characterized in that: The system comprises an image set acquisition module, a first nitrogen value calculation module, a second nitrogen value calculation module, a third nitrogen value calculation module and a plant nitrogen nutrition determination value output module which are connected in sequence; The image set acquisition module is configured to: use a multispectral UAV platform to acquire an image set with a time series for plant nitrogen nutrition determination; wherein the image set includes at least a multispectral image and a growth image of the plant; The first nitrogen value calculation module is configured to: analyze the multispectral image using a preset multispectral analysis knowledge base to calculate a first nitrogen value; wherein the multispectral analysis knowledge base is used to analyze the multispectral image based on preset multispectral analysis knowledge; The second nitrogen value calculation module is configured to: analyze the growth image using a preset growth analysis knowledge base, correct the first nitrogen value based on the analysis result, and calculate the second nitrogen value; wherein the growth analysis knowledge base is used to analyze the growth image based on preset growth analysis knowledge; The third nitrogen value calculation module is configured to: capture the spectral change trend of the multispectral image and the growth change trend of the growth image based on a preset change capture mechanism, correct the second nitrogen value based on the spectral change trend and the growth change trend, and calculate the third nitrogen value; wherein the change capture mechanism is used to extract the spectral features of the multispectral image and the growth features of the growth image, and obtain the spectral change trend and the growth change trend based on the spectral features and the growth features; The plant nitrogen nutrition measurement value output module is configured to output the third nitrogen value and define it as the plant nitrogen nutrition measurement value.

Images