CN117347283A - Wheat nitrogen fertilizer recommendation method suitable for agricultural park scale - Google Patents
Wheat nitrogen fertilizer recommendation method suitable for agricultural park scale Download PDFInfo
- Publication number
- CN117347283A CN117347283A CN202310130973.6A CN202310130973A CN117347283A CN 117347283 A CN117347283 A CN 117347283A CN 202310130973 A CN202310130973 A CN 202310130973A CN 117347283 A CN117347283 A CN 117347283A
- Authority
- CN
- China
- Prior art keywords
- wheat
- nitrogen
- plant
- pna
- nitrogen fertilizer
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 241000209140 Triticum Species 0.000 title claims abstract description 125
- 235000021307 Triticum Nutrition 0.000 title claims abstract description 107
- 239000000618 nitrogen fertilizer Substances 0.000 title claims abstract description 89
- 238000000034 method Methods 0.000 title claims abstract description 72
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 claims abstract description 249
- 229910052757 nitrogen Inorganic materials 0.000 claims abstract description 125
- 241000196324 Embryophyta Species 0.000 claims abstract description 71
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 51
- 238000009825 accumulation Methods 0.000 claims abstract description 36
- 238000005070 sampling Methods 0.000 claims abstract description 25
- 230000007812 deficiency Effects 0.000 claims abstract description 24
- 238000012544 monitoring process Methods 0.000 claims abstract description 19
- 238000010801 machine learning Methods 0.000 claims abstract description 18
- 230000003595 spectral effect Effects 0.000 claims abstract description 10
- 238000001228 spectrum Methods 0.000 claims abstract description 10
- 238000012360 testing method Methods 0.000 claims description 45
- 238000011282 treatment Methods 0.000 claims description 37
- 239000003337 fertilizer Substances 0.000 claims description 21
- 238000012937 correction Methods 0.000 claims description 18
- 238000002310 reflectometry Methods 0.000 claims description 15
- 230000001186 cumulative effect Effects 0.000 claims description 14
- 230000005855 radiation Effects 0.000 claims description 10
- 238000007637 random forest analysis Methods 0.000 claims description 9
- 239000002028 Biomass Substances 0.000 claims description 8
- 230000003993 interaction Effects 0.000 claims description 8
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 6
- 102000004190 Enzymes Human genes 0.000 claims description 5
- 108090000790 Enzymes Proteins 0.000 claims description 5
- 238000001035 drying Methods 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 5
- 239000002689 soil Substances 0.000 claims description 5
- 241001116459 Sequoia Species 0.000 claims description 4
- 229930002875 chlorophyll Natural products 0.000 claims description 4
- 235000019804 chlorophyll Nutrition 0.000 claims description 4
- ATNHDLDRLWWWCB-AENOIHSZSA-M chlorophyll a Chemical compound C1([C@@H](C(=O)OC)C(=O)C2=C3C)=C2N2C3=CC(C(CC)=C3C)=[N+]4C3=CC3=C(C=C)C(C)=C5N3[Mg-2]42[N+]2=C1[C@@H](CCC(=O)OC\C=C(/C)CCC[C@H](C)CCC[C@H](C)CCCC(C)C)[C@H](C)C2=C5 ATNHDLDRLWWWCB-AENOIHSZSA-M 0.000 claims description 4
- 230000003750 conditioning effect Effects 0.000 claims description 4
- 238000010790 dilution Methods 0.000 claims description 4
- 239000012895 dilution Substances 0.000 claims description 4
- 238000009331 sowing Methods 0.000 claims description 4
- 238000005303 weighing Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000000227 grinding Methods 0.000 claims description 3
- 125000001475 halogen functional group Chemical group 0.000 claims description 3
- 238000003384 imaging method Methods 0.000 claims description 3
- 238000001556 precipitation Methods 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 238000007696 Kjeldahl method Methods 0.000 claims description 2
- 238000010521 absorption reaction Methods 0.000 claims description 2
- 238000004140 cleaning Methods 0.000 claims description 2
- 230000001066 destructive effect Effects 0.000 claims description 2
- DNGRSVWAENAWJR-UHFFFAOYSA-N dianthra[2,3-b:2',3'-f]thieno[3,2-b]thiophene Chemical compound C1=CC=C2C=C(C=C3C(C=4SC=5C(C=4S3)=CC3=CC4=CC=CC=C4C=C3C=5)=C3)C3=CC2=C1 DNGRSVWAENAWJR-UHFFFAOYSA-N 0.000 claims description 2
- QJGQUHMNIGDVPM-UHFFFAOYSA-N nitrogen group Chemical group [N] QJGQUHMNIGDVPM-UHFFFAOYSA-N 0.000 claims description 2
- 239000000843 powder Substances 0.000 claims description 2
- 238000012216 screening Methods 0.000 claims description 2
- 235000016709 nutrition Nutrition 0.000 abstract description 6
- 230000035764 nutrition Effects 0.000 abstract description 6
- 230000004927 fusion Effects 0.000 abstract 2
- 238000007726 management method Methods 0.000 description 23
- 230000004720 fertilization Effects 0.000 description 9
- 238000002790 cross-validation Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 238000003745 diagnosis Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 210000000056 organ Anatomy 0.000 description 4
- 238000010238 partial least squares regression Methods 0.000 description 4
- 235000017166 Bambusa arundinacea Nutrition 0.000 description 3
- 235000017491 Bambusa tulda Nutrition 0.000 description 3
- 241001330002 Bambuseae Species 0.000 description 3
- 235000015334 Phyllostachys viridis Nutrition 0.000 description 3
- 239000011425 bamboo Substances 0.000 description 3
- 238000010200 validation analysis Methods 0.000 description 3
- 238000010835 comparative analysis Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 239000005431 greenhouse gas Substances 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 241001024327 Oenanthe <Aves> Species 0.000 description 1
- 240000007594 Oryza sativa Species 0.000 description 1
- 235000007164 Oryza sativa Nutrition 0.000 description 1
- 240000008042 Zea mays Species 0.000 description 1
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 description 1
- 235000002017 Zea mays subsp mays Nutrition 0.000 description 1
- 230000009418 agronomic effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 235000013339 cereals Nutrition 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 235000005822 corn Nutrition 0.000 description 1
- 238000012272 crop production Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000007599 discharging Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 235000013312 flour Nutrition 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000003306 harvesting Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 235000015097 nutrients Nutrition 0.000 description 1
- 235000003715 nutritional status Nutrition 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000035790 physiological processes and functions Effects 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 235000009566 rice Nutrition 0.000 description 1
- 239000000243 solution Substances 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N1/00—Sampling; Preparing specimens for investigation
- G01N1/02—Devices for withdrawing samples
- G01N1/04—Devices for withdrawing samples in the solid state, e.g. by cutting
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N1/00—Sampling; Preparing specimens for investigation
- G01N1/28—Preparing specimens for investigation including physical details of (bio-)chemical methods covered elsewhere, e.g. G01N33/50, C12Q
- G01N1/286—Preparing specimens for investigation including physical details of (bio-)chemical methods covered elsewhere, e.g. G01N33/50, C12Q involving mechanical work, e.g. chopping, disintegrating, compacting, homogenising
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N1/00—Sampling; Preparing specimens for investigation
- G01N1/28—Preparing specimens for investigation including physical details of (bio-)chemical methods covered elsewhere, e.g. G01N33/50, C12Q
- G01N1/44—Sample treatment involving radiation, e.g. heat
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/55—Specular reflectivity
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N31/00—Investigating or analysing non-biological materials by the use of the chemical methods specified in the subgroup; Apparatus specially adapted for such methods
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N5/00—Analysing materials by weighing, e.g. weighing small particles separated from a gas or liquid
- G01N5/04—Analysing materials by weighing, e.g. weighing small particles separated from a gas or liquid by removing a component, e.g. by evaporation, and weighing the remainder
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Pathology (AREA)
- Immunology (AREA)
- General Health & Medical Sciences (AREA)
- Biochemistry (AREA)
- Analytical Chemistry (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Computer Hardware Design (AREA)
- Mathematical Physics (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Geometry (AREA)
- Molecular Biology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention provides a wheat nitrogen fertilizer recommendation method suitable for agricultural park scales, which is used for generating and applying nitrogen fertilizer recommendation prescriptions in the wheat topdressing period of different area scales of agricultural parks. Through wheat field experiments, multispectral images are acquired by utilizing a high-efficiency unmanned aerial vehicle remote sensing system, a wheat plant dry matter weight PDM and plant nitrogen accumulation PNA monitoring model is built based on machine learning algorithm fusion spectrum, meteorological and management data, a wheat real-time nitrogen deficiency algorithm nitrogen fertilizer recommendation model is further built, and the method is verified in agricultural parks with different area scales. According to the method, the high-efficiency sampling performance and the multispectral band characteristics of the unmanned aerial vehicle remote sensing system are fully considered, the nitrogen nutrition condition of a large-area wheat in a key growth period can be monitored in real time in a lossless manner through a method of quickly acquiring the spectral images and the multisource data fusion, finally, a wheat nitrogen fertilizer recommendation prescription map of a park is drawn through a real-time nitrogen deficiency algorithm, and an accurate nitrogen fertilizer recommendation amount is provided.
Description
Technical Field
The invention belongs to the field of accurate regulation and control of crop nitrogen fertilizer, and particularly relates to a wheat nitrogen fertilizer recommendation method suitable for agricultural park scale.
Background
Winter wheat is a raw material for a variety of flour products and is a cereal crop of worldwide importance. The application of the nitrogen fertilizer can promote the growth of the wheat and improve the yield of the wheat, and the unreasonable application of the nitrogen fertilizer not only prevents the growth and development of the wheat and the formation of the yield, but also reduces the utilization efficiency of the nitrogen fertilizer and pollutes the environment. Therefore, scientists are devoted to developing a precise nitrogen application method to improve the nitrogen fertilizer utilization rate, increase the economic benefits of farmers and reduce the emission of greenhouse gases.
Real-time lossless nitrogen nutrition condition monitoring and diagnosis are the precondition of accurate nitrogen management, and development and application of remote sensing technology provide great potential for nondestructive monitoring of crop nitrogen conditions. The earliest research scholars mainly use the single-band reflectivity or classical vegetation index obtained by a spectrum sensor, and establish a quantitative relation model between spectrum data and nitrogen index by using a simple unitary parameter regression method, so that good diagnosis effect is obtained in the key growth period of wheat. However, in the application of remote sensing quantitative diagnosis, spectral information cannot be fully mined and utilized only based on a single wave band or vegetation index, so researchers propose to combine multi-wave band and multi-vegetation indexes to construct a spectral diagnosis model. As a result, it was found that the multiple linear regression model combined with partial vegetation index was more effective than the single vegetation index in estimating the crop nitrogen nutrient. Crop growth and nitrogen nutrient status are also affected by factors such as weather and crop management measures. Previous studies have utilized WheatGrow and AFRCWHEAT crop growth simulation models in combination with soil, weather and field management information to predict crop growth and nitrogen status, but crop growth simulation models require extensive data input in parameterization, calibration and verification, etc., which limits their practical application in large area crop on-season nitrogen management. The machine learning model can introduce various factors that affect crop growth without consideration of complex physiological processes. Therefore, the spectrum can be fused based on a machine learning algorithm, and the nitrogen nutrition condition of the wheat can be monitored by meteorological and management data, so that the accurate nitrogen regulation and control can be further realized.
In the past, the research is based on the near-end sensing technology to monitor the crop nitrogen nutrition status, and proposes a nitrogen fertilizer regulation and control method such as a field nitrogen fertilizer management method, a sufficiency index method, a nitrogen fertilizer optimization algorithm and the like, which is successfully applied to the accurate management of nitrogen fertilizer of crops such as corn, rice and the like. However, these nitrogen fertilizer recommendation methods based on the near-end sensing technology are only suitable for field scale, and along with the increasingly obvious trend of intensive operations in China, a single field management strategy cannot meet the requirements of agricultural park scale management on agricultural equipment and technology. Unmanned Aerial Vehicle (UAV) remote sensing system is simple to operate, and can acquire crop spectrum data with large area and high spatial resolution. The maximum flight time of a single battery of the multi-rotor unmanned aerial vehicle is 25 minutes, and the single flight can acquire crop growth information with the area of 10 hectares (the flight height of 120 meters). The high-efficiency fixed wing unmanned aerial vehicle adopted in the invention can cover an area (flight height 100 meters) close to 80 hectares within a single frame time duration of 55 minutes, and is more suitable for acquiring crop information of agricultural parks with different area scales. Therefore, it is necessary to develop reliable nitrogen fertilizer recommendation methods for the unmanned aerial vehicle remote sensing platform to guide the accurate management of nitrogen in farms of different scales.
Disclosure of Invention
The technical problem to be solved by the invention is to provide a wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale, which is based on a high-efficiency unmanned aerial vehicle remote sensing system to acquire multispectral images of a wheat planting park, synchronously acquire meteorological and management data, utilize a machine learning method to fuse the spectra, the meteorological and the management data to construct a wheat plant biomass and plant nitrogen accumulation PNA monitoring model, simultaneously construct a real-time accumulated nitrogen deficiency algorithm nitrogen fertilizer recommendation model, draw a nitrogen fertilizer recommendation amount prescription diagram of the wheat topdressing period of the agricultural park, and guide farmers to conduct accurate nitrogen fertilizer management.
The technical solution for realizing the purpose of the invention is as follows:
a wheat nitrogen fertilizer recommendation method suitable for agricultural park scale comprises the following steps:
step 1: setting a wheat field nitrogen fertilizer and variety interaction test and a farm scattered point sampling test, sampling destructive plants in the early stage of wheat jointing, and measuring yield in the mature stage to obtain yield data;
step 2: removing roots of the wheat plant sample, cleaning, deactivating enzyme in an oven, drying, and weighing to obtain the biomass of the overground part of the wheat plant sample. Further grinding into powder, and determining plant nitrogen accumulation PNA data by a Kjeldahl method;
step 3: acquiring multispectral image data of wheat canopy:
acquiring multispectral image data of a wheat canopy in the early stage of wheat jointing by adopting a high-efficiency unmanned aerial vehicle remote sensing system;
step 4: preprocessing the wheat canopy multispectral image data, and calculating to obtain wheat canopy multispectral image reflectivity data;
step 4-1: extracting the reflectance of multispectral wave bands of each cell or sampling point according to the region of interest (ROI);
step 4-2: calculating a vegetation index related to monitoring of plant dry matter weight PDM and plant nitrogen accumulation PNA according to the reflectivity in the step 4-1;
step 5: acquiring weather and management data of wheat in growing seasons;
step 6: utilizing 4 machine learning algorithms, fusing spectral vegetation indexes, weather and management parameters, and constructing a wheat plant dry matter weight PDM and plant nitrogen accumulation PNA monitoring model;
step 7: according to the field test 1-3 data, a ten-fold cross validation method is adopted to perform precision validation on plant dry matter weight PDM and plant nitrogen accumulation PNA monitoring models constructed based on 4 machine learning algorithms, and the decision coefficient R of each model validation result is calculated and compared 2 Screening out a machine learning algorithm with optimal performance and a plant dry matter weight PDM and plant nitrogen accumulation PNA monitoring model corresponding to the machine learning algorithm;
step 8: and constructing a nitrogen-chasing regulation model based on a real-time accumulated nitrogen deficiency algorithm. Based on the estimated plant dry matter weight PDM (PDM), calculating critical nitrogen concentration (N) by means of wheat critical nitrogen concentration dilution curve model (formula 1) c ) The method comprises the steps of carrying out a first treatment on the surface of the Based on PDM and N c Can calculate the critical plant nitrogen accumulation PNA (PNA) c The method comprises the steps of carrying out a first treatment on the surface of the Equation 2); PNA-based c The difference from PNA can calculate the cumulative nitrogen deficiency (AND; equation 3); the nitrogen pick-up in the season can be calculated by the formula 4.
N c ＝4.17*PDM -0.39 (1)
PNA c ＝PDM*N c (2)
AND＝PNA c -PNA (3)
N r ＝N local -AND/NUE (4)
Wherein N is c Indicating critical nitrogen concentration of plant, PNA c Represents the critical nitrogen accumulation of plants, AND represents the accumulated nitrogen deficiency of plants, N local Indicating the nitrogen trace of the plant in a proper nitrogen state, referring to the recommended dosage of the local agricultural department, 105kg ha -1 NUE is the utilization efficiency of nitrogen fertilizer of wheat jointing fertilizer, 0.6, n is taken here r The dosage is recommended for the wheat jointing fertilizer.
Further, the wheat field interaction test in the step 1 comprises wheat of different varieties and nitrogen application levels.
Further, the wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale provided by the invention comprises the steps of (2) obtaining wheat plant biomass and plant nitrogen accumulation PNA:
in the early stage of wheat jointing, 20 wheat strains are randomly fetched in a test cell or a 10 x 10m area of a sampling point, and the wheat tillering number of double rows of 1 meter is counted. Separating a wheat plant sample into leaves and stems, deactivating enzymes in an oven at 108 ℃ for 30 minutes, and drying at 80 ℃ to constant weight to obtain plant dry matter weight PDM data; and (3) measuring the nitrogen content of the ground plant sample by an AA3 flow analyzer to obtain PNA data of the nitrogen accumulation amount of the plant.
Furthermore, according to the wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale, in the step 3, the high-efficiency unmanned aerial vehicle remote sensing system is a high-efficiency fixed wing unmanned aerial vehicle carrying a Sequoia multispectral camera, the single battery can last for 55 minutes, and the spectrum sampling area can reach 80 hectares when the single battery is used at the flight height of 100 meters. The Sequoia multispectral camera can simultaneously acquire the spectral reflectance images of the crop canopy at the green light (570 nm), the red light (675 nm), the red edge (730 nm) and the near infrared (850 nm) wave bands, and can acquire the crop growth information of different area-scale agricultural parks in real time.
Furthermore, the wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale provided by the invention has the advantages that step 3, the wheat canopy multispectral image data come from a test field of nitrogen fertilizer and variety interaction, and wheat planting parks with different area scales.
Further, according to the wheat nitrogenous fertilizer recommendation method suitable for the agricultural park scale, in the step 4, the pretreatment of the wheat canopy multispectral image data is specifically as follows:
1) Noise cancellation: placing an imaging sensor in a dark room, setting different exposure time to shoot, taking the extracted DN value as a noise image, and subtracting the noise image from the original image to perform denoising treatment;
2) Halo correction: taking the average value of brightness values of pixels adjacent to the spots as spot brightness, or adopting a cubic convolution method for correcting the spot brightness;
3) Lens distortion correction: adopting enough black-and-white chessboard patterns, calculating internal parameters and external parameters of the lens by using a least square method, and solving distortion related parameters according to the rest point coordinates to correct;
4) Band registration: extracting features of a reference image and an image to be registered respectively by adopting a SIFT feature registration method, carrying out feature description, then carrying out feature matching, calculating to obtain transformation model parameters, and carrying out image transformation registration;
5) Radiation calibration: and carrying out radiation correction through a correction white board with specific reflectivity, wherein the correction white board is placed in a shooting area before each unmanned aerial vehicle flies.
Furthermore, according to the wheat nitrogenous fertilizer recommendation method suitable for the agricultural park scale, in the step 4-1, the ROI area is the area of the corresponding cell in the cell test, and the scattered sampling area is a range of 10 x 10 meters around the sampling point.
Further, the wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale provided by the invention comprises the following vegetation indexes in the step 4-2:
normalized differential red edge index ndre= (NIR-RE)/(nir+re)
Normalized differential vegetation index ndvi= (NIR-R)/(nir+r)
Red edge soil conditioning vegetation index resavi=1.5 (NIR-RE)/(nir+re+0.5)
Green soil conditioning vegetation index gsavi=1.5 (NIR-G)/(nir+g+0.5)
Red edge chlorophyll index cire= (NIR/RE) -1
DATT＝(NIR-RE)/(NIR-R)
Corrected chlorophyll absorption and reflectance index mcari1= [ (NIR-RE) -0.2 (NIR-G) ] + (NIR/RE)
Wherein G, R, RE and NIR represent the reflectivities at the 570nm, 675nm, 730nm and 850nm bands, respectively.
Further, according to the wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale, the meteorological parameters in the step 5 comprise: average over 30 days prior to testingDay temperature (T) ave ) Average daily minimum temperature (T min ) Average daily maximum temperature (T max ) Cumulative average temperature (T sum ) Cumulative precipitation (Prep sum ) Cumulative radiation quantity (Rad sum ) And an accumulated growth day (AGDD) from sowing to testing. The management data are: the nitrogen fertilizer application amount before the test.
Furthermore, the wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale provided by the invention comprises the following 4 machine learning methods in the step 6: random forest algorithm, lasso regression algorithm, artificial neural network algorithm, partial least squares regression algorithm.
Furthermore, the invention relates to a wheat nitrogen fertilizer recommendation method suitable for agricultural park scale, wherein the coefficient R is determined in the step 7 2 The calculation formulas of the relative root mean square error RMSE and the relative error RE are as follows:
wherein P is i And O i The predicted value and the measured value of the ith sample, respectively, and n is the total number of samples.
Furthermore, the wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale provided by the invention has the advantages that in the step 8, the critical nitrogen concentration dilution curve model is a model constructed by researchers, namely N c ＝4.17*DM -0.39 ；N local Indicating the nitrogen trace of the plant in a proper nitrogen state, referring to the recommended dosage of the local agricultural department, 105kg ha -1 NUE is the wheat jointing fertilizer nitrogen fertilizer utilization efficiency, here 0.60.
Compared with the prior art, the technical scheme provided by the invention has the following technical effects:
1. according to the wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale, the spectrum, the weather and the management data are fused by using a random forest algorithm, and the constructed plant dry matter weight PDM and plant nitrogen accumulation PNA monitoring model has higher precision;
2. the wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale can monitor the growth and nitrogen nutrition conditions of wheat in real time and in a lossless manner before topdressing of the wheat, and simultaneously fully considers the spatial variability of the growth vigor of the wheat to perform variable fertilization recommendation;
3. the wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale adopts the variable nitrogen fertilizer regulation and control strategy constructed by combining the high-efficiency unmanned aerial vehicle remote sensing system with the real-time accumulated nitrogen deficiency algorithm, and can realize the accurate management of crop nitrogen in the agricultural park with different area scales compared with the field-level nitrogen management strategy based on the handheld sensor.
Drawings
Fig. 1 is a flow chart of a wheat nitrogen fertilizer recommendation method suitable for agricultural park scale according to the present invention.
Fig. 2 is a ten-fold cross-validation result of a model constructed based on 4 machine learning algorithms in an embodiment.
Fig. 3 is a drawing of recommended prescriptions for the jointing fertilizer for fishing (a), bamboo body (b) and Zhouzhuang (c) agricultural parks.
Detailed Description
Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary only for explaining the present invention and are not to be construed as limiting the present invention.
The invention constructs a wheat nitrogen fertilizer recommendation method suitable for agricultural park scale by implementing wheat field interaction tests of different varieties and nitrogen application level treatment in Jiangsu provinces. Specific test design information is shown in table 1, tests 1-3 are used for building a nitrogen-tracing regulation model of a real-time accumulated nitrogen deficiency algorithm, and tests 4 and 5 are used for verifying a nitrogen fertilizer regulation algorithm.
Table 1 basic information of test design
The test comprises interaction treatment of different varieties and nitrogen application amount and farm scattered sampling treatment, so that modeling data comprise more possibilities, and the universality of the model is improved. Due to differences in breed and nitrogen fertilizer levels, the data presents a range of variation, which represents a coverage of most possible scenarios. Thus, the data set can provide support for establishing a reliable wheat nitrogen fertilizer recommendation model.
As shown in FIG. 1, the specific steps of the wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale are as follows:
step 1, acquiring agronomic data
Plant dry weight PDM and plant nitrogen accumulation PNA data: in the early stage of wheat jointing, the tillering number of the wheat in 1 meter double rows is counted and is used for calculating the average tillering number of the wheat in unit area (per hectare). Meanwhile, in each test cell, 20 representative wheat plants are randomly selected, and in a field scattered point sampling test, trimble GeoXH6000 (Trimble, CA, USA) is used for recording position information of sampling points, and 20 wheat plants are randomly sampled within a range of 10 x 10m with the sampling points as the center. Separating plant samples into leaves and stems, deactivating enzyme at 105 ℃ for 30 minutes, drying at 80 ℃ for 24 hours to constant weight, and weighing to obtain the dry weight of each organ of wheat. Plant dry weight PDM is the sum of leaf and stem weight. Crushing the dried sample, and measuring the total nitrogen content (%) of leaves and stems of different tissues and organs of the wheat plant by using a semi-trace Kjeldahl nitrogen determination method. The nitrogen accumulation amount of each organ is the product of the nitrogen content of each organ and the dry matter weight. The nitrogen accumulation PNA of the overground plant is the sum of the leaf and stem nitrogen accumulation.
Yield data: the maturity period is that 3 1m are taken from the non-sampling position in the range of 10 x 10m taking the sampling point as the center of each cell 2 The actual yield is measured by threshing and weighing the wheat ears.
Step 2, acquiring and preprocessing multispectral image data of unmanned aerial vehicle
According to the invention, the remote sensing system of the ebee fixed wing unmanned aerial vehicle is used for acquiring the multispectral image of the wheat canopy, the maximum single-frame endurance time of the system is about 55 minutes, and the coverage area is about 80 hectares. The Parrot Sequoia multispectral camera in the ebee fixed wing unmanned aerial vehicle remote sensing system is composed of five channels, comprises an RGB camera with 1600 ten thousand pixels, a single-band camera with resolution of 4608 multiplied by 3456 pixels and four 150 ten thousand pixels, a single-band camera with resolution of 1280 multiplied by 960 pixels, and four wave bands respectively comprise green light (550 nm and bandwidth of 40 nm), red light (660 nm and bandwidth of 40 nm), red edge (735 nm and bandwidth of 10 nm) and near infrared (790 nm and bandwidth of 40 nm) wave bands. And (3) performing flight test in the early stage of wheat jointing, wherein the time is selected from 10 am to 2 pm under the condition of stable illumination condition and no wind and no cloud. The corresponding fly height is set to 100 meters and the sideways and heading image overlap is set to 70%. The camera is set to an auto mode and the image is saved in TIFF format. After the unmanned aerial vehicle image is acquired, the unmanned aerial vehicle image is exported from the multispectral camera to a personal computer. And further introducing all the spectral images and the radiation correction images in the test area into the Pix4 Dapper Ag software, and splicing to obtain the orthographic images of the test area after the spectral correction. According to a plurality of ground control points uniformly distributed in the test area, geographic correction is carried out on the orthographic images of each test area in ArcGIS software, and the accuracy of geographic positions of the orthographic images is ensured.
The pretreatment is specifically as follows:
1) Noise cancellation: placing an imaging sensor in a dark room, setting different exposure time to shoot, taking the extracted DN value as a noise image, and subtracting the noise image from the original image to perform denoising treatment;
2) Halo correction: taking the average value of brightness values of pixels adjacent to the spots as spot brightness, or adopting a cubic convolution method for correcting the spot brightness;
3) Lens distortion correction: adopting enough black-and-white chessboard patterns, calculating internal parameters and external parameters of the lens by using a least square method, and solving distortion related parameters according to the rest point coordinates to correct;
4) Band registration: extracting features of a reference image and an image to be registered respectively by adopting a SIFT feature registration method, carrying out feature description, then carrying out feature matching, calculating to obtain transformation model parameters, and carrying out image transformation registration;
5) Radiation calibration: and carrying out radiation correction through a correction white board with specific reflectivity, wherein the correction white board is placed in a shooting area before each unmanned aerial vehicle flies.
Step 3, constructing vegetation index
After preprocessing the 4-band multispectral image acquired by the unmanned aerial vehicle remote sensing system, extracting 4 spectral band reflectivity information of each test cell and sampling point in the multispectral image by adopting a region of interest (ROI) method in ENVI 5.2 software. Averaging the total reflectivity of each cell in a cell test to obtain the average reflectivity of each wave band of the cell; in a field sampling test, the 4-band reflectivity of a sampling point is extracted in a range of 10 x 10m of the sampling point. The vegetation index was calculated using reflectance values for each band according to the formula definition of the vegetation index, as shown in table 2.
TABLE 2 Vegetation index for monitoring wheat plant biomass and plant Nitrogen accumulation PNA
Step 4, acquiring meteorological and management data
Meteorological data from the wheat sowing to harvesting period (11 months to 6 months of the next year) was obtained by a local climate observation station. Calculating relevant temperature parameters including average daily temperature (T) over 30 days prior to testing ave ) Average day minimum temperatureDegree (T) min ) Average daily maximum temperature (T max ) Cumulative average temperature (T sum ) Cumulative precipitation (Prep sum ) Cumulative radiation quantity (Rad sum ). And cumulative growth days (AGDD) from sowing to testing. The management data is the nitrogen fertilizer application amount before the test.
Step 5, construction and inspection of plant biomass and plant nitrogen accumulation PNA estimation model
Using 4 machine learning methods: the method comprises the steps of a random forest algorithm, a Lasso regression algorithm, an artificial neural network algorithm, a partial least squares regression algorithm, merging spectral vegetation indexes, weather and management parameters, constructing a wheat plant dry weight PDM and plant nitrogen accumulation PNA estimation model, and verifying the model, wherein the specific description is as follows:
the wheat canopy multispectral data acquired by the high-efficiency unmanned aerial vehicle remote sensing system in the test 1-3 are combined with meteorological and management parameters, a random forest algorithm, a Lasso regression algorithm, a support vector machine regression algorithm and an artificial neural network algorithm are utilized to construct a plant dry matter weight PDM and plant nitrogen accumulation PNA prediction model.
And (3) precision inspection:
the ten-fold cross validation method is utilized to carry out precision validation on plant biomass and plant nitrogen accumulation PNA monitoring models constructed based on 4 machine learning methods, and R is used for 2 The RMSE and RE values determine the best model. R is R 2 The higher the RMSE and RE values, the smaller the model accuracy. Wherein R is 2 The specific calculation formulas of RMSE and RE are as follows:
wherein P is i And O i The predicted value and the measured value of the ith sample, respectively, and n is the total number of samples.
As shown in fig. 2, ten-fold cross-validation results of a model constructed based on 4 machine learning algorithms are shown, wherein the right subscripts are respectively referred to as: the early stage of jointing is based on random forests (a), lasso (b), an artificial neural network (c) and partial least squares regression (d) algorithm to construct a ten-fold cross validation result of a plant dry matter weight PDM monitoring model; based on random forests (e) and Lasso (f), an artificial neural network (g) and partial least squares regression (h) algorithm is used for constructing a ten-fold cross validation result of the plant nitrogen accumulation PNA monitoring model.
As can be seen from the results of fig. 2, among the 4 machine learning algorithms, the random forest algorithm was found to be effective in the plant dry weight PDM (R 2 ＝0.78,RMSE＝0.43t ha -1 Re=24.49%) and plant nitrogen accumulation PNA (R 2 ＝0.83,RMSE＝13.00kg ha -1 Re=25.44%) the monitoring effect is relatively best, and the model verified R 2 The values are higher and the RMSE and RE values are lower. Therefore, the plant dry matter weight PDM and plant nitrogen accumulation PNA monitoring model based on the random forest algorithm is used for constructing the wheat nitrogen tracking regulation model.
Step 6: and constructing a nitrogen-chasing regulation model based on a real-time accumulated nitrogen deficiency algorithm. Based on the estimated plant dry matter weight PDM (PDM), calculating critical nitrogen concentration (N) by means of wheat critical nitrogen concentration dilution curve model (formula 1) c ) The method comprises the steps of carrying out a first treatment on the surface of the Based on PDM and N c Can calculate the critical plant nitrogen accumulation PNA (PNA) c The method comprises the steps of carrying out a first treatment on the surface of the Equation 2); PNA-based c The difference from PNA can calculate the cumulative nitrogen deficiency (AND; equation 3); the nitrogen pick-up in the season can be calculated by the formula 4.
N c ＝4.17*PDM -0.39 (1)
PNA c ＝PDM*N c (2)
AND＝PNA c -PNA (3)
N r ＝N local -AND/NUE (4)
Wherein N is c Indicating critical nitrogen concentration of plant, PNA c Represents the critical nitrogen accumulation of plants, AND represents the accumulated nitrogen deficiency of plants, N local Indicating the nitrogen trace of the plant in a proper nitrogen state, referring to the recommended dosage of the local agricultural department, 105kg ha -1 NUE is the utilization efficiency of nitrogen fertilizer of wheat jointing fertilizer, 0.6, n is taken here r The dosage is recommended for the wheat jointing fertilizer.
Step 7: tests 4 and 5 were used to verify a wheat nitrogen fertilizer recommendation method for agricultural park scale constructed in the present invention, with fishing, bamboo and Zhouzhuang agricultural parks areas of 70, 25 and 47 hectares, respectively.
Drawing all field boundary shp files of fishing, bamboo and Zhouzhuang agricultural parks based on ArcGIS software, extracting average spectrum data of each field in the parks, calculating recommended nitrogen fertilizer consumption of each field in 3 agricultural parks, and drawing a recommended prescription space distribution map of each farm jointing fertilizer (figure 3). Wherein the right subscripts are respectively referred to as: and (3) recommending a prescription chart for the jointing fertilizer of the agricultural park by fishing (a), zhu hong (b) and Zhouzhuang (c).
When the nitrogen tracing amount is verified in the field, variable nitrogen fertilizer application is carried out on 3 base fertilizer treatment fields arranged in a fishing park according to the nitrogen fertilizer recommendation method constructed in the invention at the early stage of jointing, and the effect of the variable nitrogen fertilizer application is evaluated by comparing with the conventional nitrogen fertilizer application. In the areas of the Hongzhuang and Zhouzhuang, the farmer base fertilizer is randomly selected to treat the field, the variable fertilizer is applied by utilizing the nitrogen fertilizer recommendation method constructed by the invention in the topdressing period, and the effect of the variable nitrogen fertilizer application is evaluated by comparing the conventional nitrogen fertilizer application treatment. The fertilization amounts of the treatments are shown in Table 3, N1, N2 and N3, and 120, 225,330kg ha respectively -1 The fixed amount of base fertilizer and additional fertilizer treatment, N1_M, N2_M and N3_M respectively represent variable additional fertilizer treatment for constructing the recommended method of the wheat nitrogen fertilizer in the jointing period under the fixed amount of base fertilizer treatment. FT represents the fertilization treatment of farmers, and FT_M represents the variable topdressing treatment of the recommendation method for constructing the wheat nitrogen fertilizer in the invention.
Table 3 wheat nitrogen fertilizer recommendation method for different campuses base fertilizer (N basal ) And topdressing (N) top ) Quantitative, yield, nitrogen bias productivity (PFP), and net yield (NP) comparative analysis
As can be seen from the table, the nitrogen deficiency treatment field in the fishing park was regulated based on the real-time cumulative nitrogen deficiency algorithm (N1_M) to increase the nitrogen pick-up amount relative to the control group (N1). The variable fertilization treatment (n2_m) fine-tuned the nitrogen pick-up under the appropriate nitrogen treatment (N2). Under the condition of excessive nitrogen application, the real-time accumulated nitrogen deficiency algorithm regulation model (N3M) reduces the nitrogen tracking amount compared with the control treatment (N3). In the Zhu hong and Zhouzhuang parks in test 5, the nitrogen fertilizer regulation and control treatment (FT_M) applied nitrogen amount based on the real-time accumulated nitrogen deficiency algorithm is obviously lower than that of Farmers (FT). Compared with conventional fertilization treatment, the nitrogen fertilizer recommendation method constructed by the invention reduces nitrogen fertilizer investment, but the wheat yield is still higher, so that the nitrogen bias productivity (PFP) and net income (NP) indexes of variable fertilization treatment based on a real-time accumulated nitrogen deficiency algorithm are obviously better than those of conventional fertilization treatment, the nitrogen fertilizer recommendation method constructed by the invention is fully proved to be capable of accurately recommending the nitrogen fertilizer consumption according to the remote sensing monitoring result of the real-time nitrogen nutrition condition of crops, the fertilizer utilization efficiency and economic benefit can be improved at the same time, and the nitrogen fertilizer recommendation method has higher practical value for on-season nitrogen fertilizer management in field crop production.
Table 4 wheat nitrogen fertilizer recommendation methods for different campuses Net Energy (NE), energy Use Efficiency (EUE), energy Productivity (EP) and CO validated in the field 2 Emission comparative analysis
Table 4 calculates and analyzes energy input and output for 3 agricultural parks under different nitrogen fertilizer treatments. In the fishing park of trial 4, the Net Energy (NE) of the suitable nitrogen treatment (N2) and the excess nitrogen treatment (N3) was higher than that of the nitrogen deficiency treatment. The Energy Utilization Efficiency (EUE) and the Energy Productivity (EP) of the nitrogen fertilizer excessive treatment are the lowest, and the EUE and the EP are obviously improved by the nitrogen application treatment based on the real-time accumulated nitrogen deficiency algorithm variable. In the Zhu hong and Zhou village (Table 4) of test 5, there were higher EUE and EP for the variable nitrogen application based on the real-time cumulative nitrogen deficiency algorithm of the present invention relative to the peasant household fertilizer application treatment.
In addition, CO for treatment with different nitrogenous fertilizers 2 The emissions were evaluated (table 4). In run 4, CO 2 The emission increases with the increase of the nitrogen application amount, and the nitrogen fertilizer recommendation method constructed by the invention obviously reduces CO under the treatment of excessive fertilization (N3) 2 And (5) discharging. In the Zhenghong and Zhouzhuang park of test 5, CO is applied to the nitrogen treatment based on the real-time accumulated nitrogen deficiency algorithm variable 2 The emission is reduced by 7.60-10.11% compared with the fertilization treatment of farmers. The analysis results fully demonstrate that the nitrogen fertilizer recommendation method constructed by the invention can reduce energy input, reduce greenhouse gas emission and is beneficial to energy conservation and emission reduction.
While only a few embodiments of the present invention have been described, it should be noted that modifications could be made by those skilled in the art without departing from the principles of the present invention, which modifications are to be regarded as being within the scope of the invention.
Claims (10)
1. The wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale is characterized by comprising the following steps of:
step 1: setting a wheat field nitrogen fertilizer and variety interaction test and a farm scattered point sampling test, sampling destructive plants in the early stage of wheat jointing to obtain a wheat plant sample, and measuring yield in the mature stage;
step 2: removing roots of the wheat plant sample, cleaning, deactivating enzyme in an oven, drying, and weighing to obtain the biomass of the overground part of the wheat plant sample; further grinding into powder, and measuring the nitrogen content of the plant by a Kjeldahl method to obtain PNA data of the nitrogen accumulation amount of the plant;
step 3: acquiring multispectral image data of a wheat canopy in the early stage of wheat jointing by adopting a high-efficiency unmanned aerial vehicle remote sensing system;
step 4: preprocessing the wheat canopy multispectral image data, and calculating to obtain wheat canopy multispectral image reflectivity data;
step 4-1: based on the reflectance data of the wheat canopy multispectral image, extracting the multispectral band reflectivities of the test cells and sampling points of the farm scatter sampling test according to the region of interest ROI;
step 4-2: calculating a vegetation index related to plant dry matter weight PDM and plant nitrogen accumulation PNA estimation according to the reflectivity in the step 4-1;
step 5: acquiring weather and management data of wheat in growing seasons;
step 6: using 4 machine learning algorithms, including: random forest algorithm, lasso regression algorithm, artificial neural network algorithm, partial least square regression algorithm; and the spectrum vegetation index, the weather and the management parameters are fused, and a wheat plant dry weight PDM and a plant nitrogen accumulation PNA monitoring model are respectively constructed;
step 7: verifying plant dry matter weight PDM and plant nitrogen accumulation PNA monitoring models constructed based on 4 machine learning algorithms by adopting the data obtained in the steps 1-3 and a ten-fold cross verification method, and calculating and comparing the decision coefficients R of the verification results of the models 2 Screening out a machine learning algorithm with optimal performance and a plant dry matter weight PDM and plant nitrogen accumulation PNA monitoring model corresponding to the machine learning algorithm;
step 8: constructing a real-time accumulated nitrogen deficiency algorithm nitrogen-chasing regulation model;
based on the estimated plant dry matter weight PDM, calculating the critical nitrogen concentration N by means of a wheat critical nitrogen concentration dilution curve model c ；
N c ＝4.17*PDM -0.39 (1)
Based on PDM and N c Calculating the product of critical plant nitrogen accumulation PNA c ；
PNA c ＝ PDM*N c (2)
PNA-based c Calculating the real-time accumulated nitrogen deficiency AND of the plant according to the difference value between the accumulated nitrogen deficiency AND PNA;
AND＝ PNA c -PNA (3)
when the quaternary nitrogen pick-up is calculated by the following formula:
N r ＝N local -AND/NUE (4)
wherein N is c Representing the critical nitrogen concentration of plants; PNA (PNA) c Representing the critical nitrogen accumulation amount of plants; PNA represents plant nitrogen accumulation PNA; AND represents the real-time accumulated nitrogen deficiency of the plants; n (N) local Indicating the nitrogen tracing amount of the plant in a proper nitrogen state, and referring to the recommended dosage of the local agricultural department; NUE is the utilization efficiency of the nitrogen fertilizer of the wheat jointing fertilizer; n (N) r The dosage is recommended for the wheat jointing fertilizer.
2. A wheat nitrogen fertilizer recommendation method suitable for use on an agricultural campus scale as claimed in claim 1, wherein the wheat field interaction test in step 1 includes wheat of different varieties and nitrogen application levels.
3. The wheat nitrogen fertilizer recommendation method applicable to agricultural park scale according to claim 1, wherein the step of obtaining the aboveground biomass and plant nitrogen accumulation PNA of the wheat plant sample in step 2 comprises the steps of:
randomly taking 20 wheat strains in a test cell or a 10 x 10m area of a sampling point in the early stage of wheat jointing, and simultaneously counting the tillering number of the wheat with 1 meter double rows; separating a wheat plant sample into leaves and stems, deactivating enzymes in an oven at 108 ℃ for 30 minutes, and drying at 80 ℃ to constant weight to obtain plant dry matter weight PDM data; and after grinding the plant samples, determining the nitrogen content by an AA3 flow analyzer to obtain plant nitrogen accumulation PNA data.
4. The wheat nitrogen fertilizer recommendation method suitable for the agricultural park scale according to claim 1, wherein the high-efficiency unmanned aerial vehicle remote sensing system used in the step 3 is an efficient fixed wing unmanned aerial vehicle carrying a Sequoia multispectral camera, the Sequoia multispectral camera simultaneously acquires the spectral reflectance images of crops at the wave bands of green light 570nm, red light 675nm, red light 730nm and near infrared 850nm, and crop growth information of the agricultural park with different area scales is acquired in real time.
5. The wheat nitrogen fertilizer recommendation method suitable for agricultural park scale according to claim 1, wherein the step 3 wheat canopy multispectral image data is from test fields of nitrogen fertilizer and variety interaction, and wheat parks of different area scales.
6. The wheat nitrogen fertilizer recommendation method suitable for agricultural park scale according to claim 1, wherein the preprocessing of the wheat canopy multispectral image data in step 4 is specifically as follows:
1) Noise cancellation: placing an imaging sensor in a dark room, setting different exposure time to shoot, taking the extracted DN value as a noise image, and subtracting the noise image from the original image to perform denoising treatment;
2) Halo correction: taking the average value of brightness values of pixels adjacent to the spots as spot brightness, or adopting a cubic convolution method for correcting the spot brightness;
3) Lens distortion correction: adopting enough black-and-white chessboard patterns, calculating internal parameters and external parameters of the lens by using a least square method, and solving distortion related parameters according to the rest point coordinates to correct;
4) Band registration: extracting features of a reference image and an image to be registered respectively by adopting a SIFT feature registration method, carrying out feature description, then carrying out feature matching, calculating to obtain transformation model parameters, and carrying out image transformation registration;
5) Radiation calibration: and carrying out radiation correction through a correction white board with specific reflectivity, wherein the correction white board is placed in a shooting area before each unmanned aerial vehicle flies.
7. The wheat nitrogen fertilizer recommendation method suitable for agricultural park scale according to claim 1, wherein in step 4-1, the ROI area is the corresponding cell area in the cell test, and the scattered point sampling area is the range of 10 x 10 meters around the sampling point.
8. A wheat nitrogen fertilizer recommendation method suitable for use on an agricultural campus scale as claimed in claim 1, wherein the vegetation index in step 4-2 comprises:
normalized differential red edge index ndre= (NIR-RE)/(nir+re)
Normalized differential vegetation index ndvi= (NIR-R)/(nir+r)
Red edge soil conditioning vegetation index resavi=1.5 (NIR-RE)/(nir+re+0.5)
Green soil conditioning vegetation index gsavi=1.5 (NIR-G)/(nir+g+0.5)
Red edge chlorophyll index cire= (NIR/RE) -1
DATT＝(NIR-RE)/(NIR-R)
Corrected chlorophyll absorption and reflection index mcari1= [ (NIR-RE) -0.2 x (NIR-G) ]x (NIR/RE)
Wherein G, R, RE and NIR represent the reflectivities at the 570nm, 675nm, 730nm and 850nm bands, respectively.
9. The wheat nitrogen fertilizer recommendation method applicable to agricultural park scales as claimed in claim 1, wherein the weather parameters in the step 5 are obtained through a weather station of a test point, comprising: average daily temperature T over 30 days prior to testing ave Average day minimum temperature T min Average day maximum temperature T max Cumulative average temperature T sum Cumulative precipitation Prep sum Cumulative radiation amount Rad sum And an accumulated growth day AGDD from sowing to testing; the management data are: the nitrogen fertilizer application amount before the test.
10. The wheat nitrogen fertilizer recommendation method suitable for agricultural park scale according to claim 1, wherein the coefficient R is determined in step 7 2 The calculation formulas of the relative root mean square error RMSE and the relative error RE are as follows:
wherein P is i And O i The predicted value and the measured value of the ith sample, respectively, and n is the total number of samples.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310130973.6A CN117347283A (en) | 2023-02-17 | 2023-02-17 | Wheat nitrogen fertilizer recommendation method suitable for agricultural park scale |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310130973.6A CN117347283A (en) | 2023-02-17 | 2023-02-17 | Wheat nitrogen fertilizer recommendation method suitable for agricultural park scale |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117347283A true CN117347283A (en) | 2024-01-05 |
Family
ID=89354569
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310130973.6A Pending CN117347283A (en) | 2023-02-17 | 2023-02-17 | Wheat nitrogen fertilizer recommendation method suitable for agricultural park scale |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117347283A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118010648A (en) * | 2024-03-11 | 2024-05-10 | 扬州大学 | Rice and wheat nitrogen nutrition multispectral diagnosis method for unmanned aerial vehicle precise fertilization |
| CN118067709A (en) * | 2024-04-17 | 2024-05-24 | 北京市农林科学院信息技术研究中心 | Synchronous monitoring method and device for crop water fertilizer surplus and shortage |
-
2023
- 2023-02-17 CN CN202310130973.6A patent/CN117347283A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118010648A (en) * | 2024-03-11 | 2024-05-10 | 扬州大学 | Rice and wheat nitrogen nutrition multispectral diagnosis method for unmanned aerial vehicle precise fertilization |
| CN118067709A (en) * | 2024-04-17 | 2024-05-24 | 北京市农林科学院信息技术研究中心 | Synchronous monitoring method and device for crop water fertilizer surplus and shortage |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112903600A (en) | Rice nitrogen fertilizer recommendation method based on multispectral image of fixed-wing unmanned aerial vehicle | |
| CN110222475B (en) | Method for inverting moisture content of winter wheat plants based on multispectral remote sensing of unmanned aerial vehicle | |
| Zhang et al. | High throughput analysis of leaf chlorophyll content in sorghum using RGB, hyperspectral, and fluorescence imaging and sensor fusion | |
| CN109187441B (en) | Method for constructing summer corn nitrogen content monitoring model based on canopy spectral information | |
| Qi et al. | Monitoring of peanut leaves chlorophyll content based on drone-based multispectral image feature extraction | |
| CN113268923B (en) | Summer corn yield estimation method based on simulated multispectral | |
| CN117347283A (en) | Wheat nitrogen fertilizer recommendation method suitable for agricultural park scale | |
| CN112345458A (en) | Wheat yield estimation method based on multispectral image of unmanned aerial vehicle | |
| CN111241912A (en) | Multi-vegetation index rice yield estimation method based on machine learning algorithm | |
| CN110376167A (en) | Rice leaf nitrogen content monitoring method based on unmanned plane EO-1 hyperion | |
| CN112557393B (en) | Wheat leaf layer nitrogen content estimation method based on hyperspectral image fusion map features | |
| CN111044516B (en) | Remote sensing estimation method for chlorophyll content of rice | |
| CN108985588A (en) | A kind of crop yield prediction remote sensing estimation method, device and system | |
| CN106372592A (en) | Winter wheat plantation area calculation method based on winter wheat area index | |
| Zhang et al. | Estimating wheat yield by integrating the WheatGrow and PROSAIL models | |
| CN112287886B (en) | Wheat plant nitrogen content estimation method based on hyperspectral image fusion map features | |
| CN116602106A (en) | Unmanned aerial vehicle-based variable fertilization method in paddy field | |
| CN116108318B (en) | Rape nitrogen fertilizer recommended dressing amount calculation method based on unmanned aerial vehicle multispectral image | |
| CN112816618A (en) | Method for screening nitrogen-efficient wheat varieties | |
| Liu et al. | UAV multispectral images for accurate estimation of the maize LAI considering the effect of soil background | |
| CN114549881B (en) | Wheat early tiller number estimation method based on regional gradient vegetation index | |
| Lyu et al. | UAV time-series imagery with novel machine learning to estimate heading dates of rice accessions for breeding | |
| Liao et al. | A double-layer model for improving the estimation of wheat canopy nitrogen content from unmanned aerial vehicle multispectral imagery | |
| Zhangzhong et al. | Development of an evapotranspiration estimation method for lettuce via mobile phones using machine vision: Proof of concept | |
| CN116883874A (en) | Evaluation method and system for comprehensive growth vigor of cinnamomum camphora dwarf forest |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |