CN116151454A - Method and system for predicting yield of short-forest linalool essential oil by multispectral unmanned aerial vehicle - Google Patents
Method and system for predicting yield of short-forest linalool essential oil by multispectral unmanned aerial vehicle Download PDFInfo
- Publication number
- CN116151454A CN116151454A CN202310161507.4A CN202310161507A CN116151454A CN 116151454 A CN116151454 A CN 116151454A CN 202310161507 A CN202310161507 A CN 202310161507A CN 116151454 A CN116151454 A CN 116151454A
- Authority
- CN
- China
- Prior art keywords
- model
- essential oil
- module
- yield
- forest
- 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
- 239000000341 volatile oil Substances 0.000 title claims abstract description 88
- CDOSHBSSFJOMGT-UHFFFAOYSA-N linalool Chemical compound CC(C)=CCCC(C)(O)C=C CDOSHBSSFJOMGT-UHFFFAOYSA-N 0.000 title claims abstract description 59
- 238000000034 method Methods 0.000 title claims abstract description 54
- 239000001490 (3R)-3,7-dimethylocta-1,6-dien-3-ol Substances 0.000 title claims abstract description 29
- CDOSHBSSFJOMGT-JTQLQIEISA-N (R)-linalool Natural products CC(C)=CCC[C@@](C)(O)C=C CDOSHBSSFJOMGT-JTQLQIEISA-N 0.000 title claims abstract description 29
- 229930007744 linalool Natural products 0.000 title claims abstract description 29
- 241000723346 Cinnamomum camphora Species 0.000 claims description 44
- 230000003595 spectral effect Effects 0.000 claims description 34
- 238000012545 processing Methods 0.000 claims description 32
- 238000002310 reflectometry Methods 0.000 claims description 26
- 241000196324 Embryophyta Species 0.000 claims description 23
- 238000004458 analytical method Methods 0.000 claims description 22
- 238000007637 random forest analysis Methods 0.000 claims description 14
- 238000005070 sampling Methods 0.000 claims description 14
- 238000001228 spectrum Methods 0.000 claims description 14
- DSSYKIVIOFKYAU-XCBNKYQSSA-N (R)-camphor Chemical compound C1C[C@@]2(C)C(=O)C[C@@H]1C2(C)C DSSYKIVIOFKYAU-XCBNKYQSSA-N 0.000 claims description 11
- 238000011156 evaluation Methods 0.000 claims description 11
- 238000012706 support-vector machine Methods 0.000 claims description 11
- 238000010219 correlation analysis Methods 0.000 claims description 9
- 239000003921 oil Substances 0.000 claims description 9
- 238000010276 construction Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000005259 measurement Methods 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 5
- 230000000694 effects Effects 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 3
- 238000012045 magnetic resonance elastography Methods 0.000 description 10
- 239000002689 soil Substances 0.000 description 7
- 238000012937 correction Methods 0.000 description 6
- 238000012360 testing method Methods 0.000 description 6
- 238000012795 verification Methods 0.000 description 5
- 229960000846 camphor Drugs 0.000 description 4
- 229930008380 camphor Natural products 0.000 description 4
- 238000005520 cutting process Methods 0.000 description 4
- 238000003973 irrigation Methods 0.000 description 4
- 230000002262 irrigation Effects 0.000 description 4
- 241001464837 Viridiplantae Species 0.000 description 3
- 238000010521 absorption reaction Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000005855 radiation Effects 0.000 description 3
- 239000002994 raw material Substances 0.000 description 3
- 208000027066 STING-associated vasculopathy with onset in infancy Diseases 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000004821 distillation Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000004720 fertilization Effects 0.000 description 2
- 238000003306 harvesting Methods 0.000 description 2
- 238000000691 measurement method Methods 0.000 description 2
- 239000002304 perfume Substances 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 239000002028 Biomass Substances 0.000 description 1
- 241000218195 Lauraceae Species 0.000 description 1
- 241000587213 Litsea coreana Species 0.000 description 1
- 150000001298 alcohols Chemical class 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000003153 chemical reaction reagent Substances 0.000 description 1
- 238000007635 classification algorithm Methods 0.000 description 1
- 230000001066 destructive effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 238000011065 in-situ storage Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 150000008442 polyphenolic compounds Chemical class 0.000 description 1
- 235000013824 polyphenols Nutrition 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- WPPDXAHGCGPUPK-UHFFFAOYSA-N red 2 Chemical compound C1=CC=CC=C1C(C1=CC=CC=C11)=C(C=2C=3C4=CC=C5C6=CC=C7C8=C(C=9C=CC=CC=9)C9=CC=CC=C9C(C=9C=CC=CC=9)=C8C8=CC=C(C6=C87)C(C=35)=CC=2)C4=C1C1=CC=CC=C1 WPPDXAHGCGPUPK-UHFFFAOYSA-N 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 235000013599 spices Nutrition 0.000 description 1
- 239000010902 straw Substances 0.000 description 1
- 230000001225 therapeutic effect Effects 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
- 238000009333 weeding Methods 0.000 description 1
- 239000002023 wood Substances 0.000 description 1
Images
Classifications
-
- 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
- G01N21/27—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands using photo-electric detection ; circuits for computing concentration
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/188—Vegetation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Human Resources & Organizations (AREA)
- Artificial Intelligence (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Analytical Chemistry (AREA)
- Chemical & Material Sciences (AREA)
- Marketing (AREA)
- Medical Informatics (AREA)
- Biochemistry (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Operations Research (AREA)
- Development Economics (AREA)
- Marine Sciences & Fisheries (AREA)
- Multimedia (AREA)
- Game Theory and Decision Science (AREA)
- Animal Husbandry (AREA)
- Agronomy & Crop Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Primary Health Care (AREA)
- Mining & Mineral Resources (AREA)
- Quality & Reliability (AREA)
- Spectroscopy & Molecular Physics (AREA)
Abstract
The invention discloses a method and a system for predicting the yield of a short-forest linalool essential oil by using a multispectral unmanned aerial vehicle, which relate to the technical field of agriculture.
Description
Technical Field
The invention relates to the technical field of agriculture, in particular to a method and a system for predicting the yield of short-forest linalool essential oil by a multispectral unmanned aerial vehicle.
Background
As a precious essential oil raw material tree species in China, camphor tree not only can be used for wood and landscaping, but also has unique positions in medical treatment and perfume industry. The polyphenols of camphor tree essential oil have therapeutic effect in various diseases, and the alcohols are exported as natural perfume to all countries around the world. The camphor tree is mainly distributed in Yangtze river basin and red soil area in south of China, and is a main raw material source in the fields of essence and spice, medical sanitation, gardens, food and the like, and the camphor tree essential oil industry becomes one of leading industries of forestry in south area. The planting area of industrial raw material forests mainly containing Lauraceae plants in China reaches 6.67 ten thousand hm 2 And possess international market price pricing rights for natural linalool. The growth condition of the Cinnamomum camphora affects the biomass and the oil yield, thereby affecting the high-quality and high-efficiency production of the Cinnamomum camphora industry, and the method brings challenges to the economic benefit and sustainable development of the Cinnamomum camphora industry. Therefore, by means of the emerging technological means, the essential oil yield is accurately predicted before the camphor harvest, and further the fields such as fertilization or irrigation and the like are scientifically guided, so that the method has important significance for protecting and improving the essential oil yield of the camphor.
Traditionally, camphor tree essential oil extraction and yield measurement, whether distillation or extraction, are measured by in-situ sampling investigation, which not only consumes a great deal of time and effort, but also damages plants, which is a direct measurement method for destructive sampling. In addition, due to the limitation of sampling points, the methods can only be applied to small areas, and cannot be expanded to monitor the short-forest linaloes planted in large areas; and an indirect measurement method, such as a method for acquiring economic plant spectrum information through a remote sensing technology and estimating the yield of the economic plant by utilizing an algorithm model, provides a new thought and a new way for monitoring the yield of the large-area short-forest linalool essential oil. The unmanned aerial vehicle multispectral data is utilized to screen vegetation indexes in a large quantity, more accurate and comprehensive parameter input conditions are provided for the prediction model, and the accuracy of the model prediction result is improved.
Therefore, a method and a system for predicting the yield of the short-forest linalool essential oil by using the multispectral unmanned aerial vehicle are provided, so that the difficulty in the existing large-area short-forest linalool essential oil yield prediction technology is solved, and the problem to be solved by the person skilled in the art is urgent.
Disclosure of Invention
In view of the above, the invention provides a method and a system for predicting the yield of the short-forest linalool essential oil by using a multispectral unmanned aerial vehicle, which can rapidly predict the yield of the camphor tree essential oil and manage fertigation, and provide accurate scientific and theoretical support.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
a method for predicting the yield of the short-forest linalool essential oil by a multispectral unmanned aerial vehicle comprises the following steps:
s1, data acquisition: acquiring data of the yield of the short forest linalool essential oil and multispectral image data shot by the unmanned aerial vehicle;
s2, data processing: splicing the multispectral image data in the step S1 to obtain spectral reflectivity data of different wavebands; measuring the yield of the short-forest linalool essential oil in the S1 by using a method for measuring each plant;
s3, constructing a model: the spectral reflectivity data in the step S2 are combined linearly or nonlinearly to form vegetation indexes, and multispectral vegetation indexes are selected to be used for constructing various spectral index models;
s4, selecting a model: selecting a model evaluation index to select a plurality of spectrum index models constructed in the step S3, and performing correlation analysis on the spectrum index models and the essential oil yield measured in the step S2 to obtain an optimal model;
s5, result analysis: and (3) rapidly predicting the essential oil yield of the camphor tree by using the optimal model so as to realize decision making on fertigation management of the camphor tree.
In the above method, optionally, in S2, the specific content of the splicing processing for the multispectral image data acquired by the unmanned aerial vehicle is:
the method comprises the steps of performing geometric correction and radiation correction pretreatment on multispectral image data by utilizing splicing software Yusense Map V2.2.2, importing the multispectral image information of the pretreated unmanned aerial vehicle into ENVI5.3, cutting out corresponding spectral images on images by taking a ground actually measured sampling area as a center, removing the shadows of soil and trees, and taking the average reflection spectrum of a dwarf linaloe sample in a region range as the spectral reflectivity of the sampling point to obtain spectral reflectivity data of different wave bands.
In the above method, optionally, the measurement formula of the yield of the short-forest linalool essential oil in the S2 is:
wherein i and j are positive integers, O i The unit is kg hm -2 ，G Bottle j + oil j The unit of G is G, G is the weight of the j tree essential oil and the bottle Leaf j Measuring the weight of the leaf sample of the oil for the j-th tree, G Total leaf j Is the total leaf weight of the j-th tree.
In the above method, optionally, the specific content of the vegetation index in S3 is: plants exhibit different spectral reflectivities due to significant differences in their own inherent plant cell biochemical parameters; in the visible light wave band, main absorption peaks of healthy green plants are formed near the red light wave band and the blue light wave band, and main reflection peaks are formed in the green light wave band; the spectral reflectivities of the wave bands are combined linearly or nonlinearly to form a vegetation index, which can be used for diagnosing the growth state of economic plants and inverting various economic plant parameters.
The method described above, optionally, the multiple models in S3 include, but are not limited to, a support vector machine model, a back propagation neural network model, and a random forest model.
In the above method, optionally, specific contents of the model evaluation index in S4 are:
the model fitting result adopts a determination coefficient R 2 The root mean square error RMSE and the average relative error MRE are evaluated to determine a coefficient R 2 The closer to 1, the higher the prediction accuracy of the model; the closer the root mean square error RMSE is to 0 with the average relative error MRE, the better the model effect is, the more the prediction result is concentrated, and the calculation formula is as follows:
in the method, in the process of the invention,
is a model predictive value; a, a i Is the actual sampling value; />
Is the average value; n is the number of samples.
A system for predicting the yield of the essential oil of the cinnamomum camphora in the short forest by using a multispectral unmanned aerial vehicle, a method for predicting the yield of the essential oil of the cinnamomum camphora in the short forest by using any one of the multispectral unmanned aerial vehicle, comprising: the system comprises a data acquisition module, a data processing module, a model construction module, a selected model module and a result analysis module;
and a data acquisition module: the system comprises a data processing module, a data processing module and a data processing module, wherein the data processing module is used for acquiring output data of the short forest linaloe essential oil and multispectral image data shot by an unmanned aerial vehicle;
and a data processing module: the device is connected with the input end of the model building module and is used for performing splicing processing on the acquired multispectral image data to obtain spectral reflectivity data of different wave bands; measuring the yield of the short-forest linalool essential oil by using a method for measuring each plant;
model construction module: the system comprises a model module, a vegetation index module, a multi-spectral vegetation index module and a data processing module, wherein the model module is connected with the input end of the model module and is used for forming a vegetation index by linear or nonlinear combination of spectral reflectivity data, and the multi-spectral vegetation index is selected to be used for constructing various spectral index models;
selecting a model module: the system comprises a result analysis module, a model evaluation index selection module, a correlation analysis module and a model analysis module, wherein the result analysis module is connected with the input end of the result analysis module and is used for selecting a plurality of constructed models and performing correlation analysis on the plurality of constructed models and the measured essential oil yield to obtain an optimal model;
and a result analysis module: the method is used for rapidly predicting the essential oil yield of the camphor tree by utilizing the optimal model so as to realize decision making on fertigation management of the camphor tree.
Compared with the prior art, the invention provides a method and a system for predicting the yield of the short-forest linalool essential oil by using the multispectral unmanned aerial vehicle, which are known from the technical scheme, wherein the method comprises the following steps of:
1) The method comprises the steps of obtaining multispectral images by using an unmanned aerial vehicle, synchronously collecting data of the yield of the camphor tree essential oil, analyzing the correlation between the yield of the camphor tree essential oil and the vegetation index, further screening out a vegetation index group with high correlation degree with the yield of the camphor tree essential oil, respectively constructing a support vector machine model, a back propagation neural network model and a random forest model of the yield of the camphor tree essential oil by utilizing the obtained vegetation index group, comparing the prediction precision of 3 models by selecting model evaluation indexes, finally obtaining the most suitable vegetation index and prediction model, and providing decision support for the field irrigation and fertilization management of the camphor tree essential oil.
2) The precision of the short forest linalool essential oil yield prediction model constructed based on the random forest model is obviously higher than that of a support vector machine model and a back propagation neural network model. Therefore, the random forest model can be used as a first-choice method for modeling the yield of the camphor essential oil in the large short forest.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present invention, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for predicting the yield of the short-forest linalool essential oil by a multispectral unmanned aerial vehicle;
fig. 2 is a prediction result of a different modeling method for a camphor tree essential oil yield in a short forest according to an embodiment of the present invention, wherein 2a is a camphor tree essential oil yield estimation model of a support vector machine, 2b is a camphor tree essential oil yield estimation model of a random forest, and 2c is a camphor tree essential oil yield estimation model of a back propagation neural network;
fig. 3 is a block diagram of a system for predicting the yield of the short-forest linalool essential oil by the multispectral unmanned aerial vehicle.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, the invention discloses a method for predicting the yield of short-forest linalool essential oil by a multispectral unmanned aerial vehicle, which comprises the following steps:
s1, data acquisition: acquiring data of the yield of the short forest linalool essential oil and multispectral image data shot by the unmanned aerial vehicle;
s2, data processing: splicing the multispectral image data in the step S1 to obtain spectral reflectivity data of different wavebands; measuring the yield of the short-forest linalool essential oil in the S1 by using a method for measuring each plant;
s3, constructing a model: the spectral reflectivity data in the step S2 are combined linearly or nonlinearly to form vegetation indexes, and multispectral vegetation indexes are selected to be used for constructing various spectral index models;
s4, selecting a model: selecting a model evaluation index to select a plurality of spectrum index models constructed in the step S3, and performing correlation analysis on the spectrum index models and the essential oil yield measured in the step S2 to obtain an optimal model;
s5, result analysis: and (3) rapidly predicting the essential oil yield of the camphor tree by using the optimal model so as to realize decision making on fertigation management of the camphor tree.
Further, in S2, the specific content of the splicing processing for the multispectral image data acquired by the unmanned aerial vehicle is:
the method comprises the steps of performing geometric correction and radiation correction pretreatment on multispectral image data by utilizing splicing software Yusense Map V2.2.2, importing the multispectral image information of the pretreated unmanned aerial vehicle into ENVI5.3, cutting out corresponding spectral images on images by taking a ground actually measured sampling area as a center, removing soil and shadow, and taking the average reflection spectrum of a dwarf linaloe sample in a region range as the spectral reflectivity of the sampling point to obtain spectral reflectivity data of different wave bands.
Further, the determination formula of the essential oil yield of the short-forest linalool in the S2 is as follows:
wherein i and j are positive integers, O i The unit is kg hm -2 ，G Bottle j + oil j The unit of G is G, G is the weight of the j tree essential oil and the bottle Leaf j Measuring the weight of the leaf sample of the oil for the j-th tree, G Total leaf j Is the total leaf weight of the j-th tree.
Further, the vegetation index in S3 specifically includes: plants exhibit different spectral reflectivities due to significant differences in their own inherent plant cell biochemical parameters; in the visible light wave band, main absorption peaks of healthy green plants are formed near the red light wave band and the blue light wave band, and main reflection peaks are formed in the green light wave band; the spectral reflectivities of the wave bands are combined linearly or nonlinearly to form a vegetation index, which can be used for diagnosing the growth state of economic plants and inverting various economic plant parameters.
Further, the multiple models in S3 include, but are not limited to, a support vector machine model, a back propagation neural network model, and a random forest model.
Further, the specific content of the model evaluation index in S4 is:
the model fitting result adopts a determination coefficient R 2 The root mean square error RMSE and the average relative error MRE are evaluated to determine a coefficient R 2 The closer to 1, the higher the prediction accuracy of the model; the closer the root mean square error RMSE is to 0 with the average relative error MRE, the better the model effect is, the more the prediction result is concentrated, and the calculation formula is as follows:
in the method, in the process of the invention,
is a model predictive value; a, a i Is the actual sampling value; />
Is the average value; n is the number of samples.
In a specific embodiment, taking the test type of the Litsea coreana as Ganfang No. 1 as an example, the specific contents are as follows:
the variety of the test short forest linaloe is Ganfang No. 1. 66 cells are distributed in the research area, 9 short-forest linaloes are planted in each cell, the row spacing of the short-forest linaloes is 1 multiplied by 1m, and 594 short-forest linaloes are planted in the whole cell. The test sets up the irrigation mode and adds the supplementary sprinkling irrigation for the rain to support. In month 4 of 2021, annual cutting seedlings of Ganfang No. 1 are transplanted manually, no medicine is applied, and other field management is consistent with local management by manual weeding. The camphor tree essential oil is extracted by breaking and sampling at the period of 2 days of 10 months of 2022, and the ground stems are left. The acquisition time of the unmanned aerial vehicle multispectral image is 2022, 9 late days, and the Ganfang No. 1 is in the harvest period. The flight time is 11:00-14:00, the sky is clear and cloudless in flight, and the wind power is small. And before taking off, manually controlling the unmanned aerial vehicle to fly to a position about 1m above the calibration white board, and shooting the standard white board by adopting a camera single shooting mode. The flight is set to be S-shaped, the two flying heights are 50m, the course and the side overlapping degree are respectively set to be 75%, the camera lens and the ground are 90 degrees, and the photographing mode is equal time interval.
The measurement of the yield of the Ganfang No. 1 essential oil adopts a method of measuring each plant. 594 10ml brown glass reagent bottles, 6 essential oil stills, several large mesh bags were prepared for the test. The Ganfang No. 1 is harvested, immediately transported back to a laboratory by a net bag, the fresh weight of the tree trunk of each sample is weighed after the leaves and stems of the plants are separated after the total weight is weighed, and 300g of each leaf sample is collected. When the laboratory is used for processing, the content of essential oil in the Ganfang No. 1 leaf is measured by adopting a distillation method, 300g of a weighed sample is placed into a distiller, distilled for 1 hour in a timing mode, then a disposable test straw is used for sucking the essential oil in the upper water layer in the shunt, and the essential oil is weighed by an electronic scale with the precision of 0.01 g. The leaf essential oil yield of each cell of Ganfang No. 1 is measured according to the formula (1).
Where i=1, 2,3, … … 66, j=1, 2,3, … … 9,O i The unit is kg hm -2 ，G Bottle j + oil j The unit of G is G, G is the weight of the j tree essential oil and the bottle Leaf j Measuring the weight of the leaf sample of the oil for the j-th tree, G Total leaf j Is the total leaf weight of the j-th tree.
And splicing the multispectral images acquired by the unmanned aerial vehicle by using splicing software YuusenseMapV2.2.2. The main pre-processing of the multispectral image in stitching includes geometric correction and radiation correction. And extracting spectral reflectivity software to be ENVI5.3, and importing the preprocessed unmanned aerial vehicle multispectral image information into ENVI 5.3. And cutting out a corresponding spectral image on an image by taking a ground actually measured sampling area as a center, removing the shadow of soil and trees, and taking the average reflection spectrum of a sample in the region of interest (ROI) as the spectral reflectivity of the sampling point to obtain spectral reflectivity data of different wavebands.
Plants exhibit different spectral reflectivities due to the apparent differences in their own inherent plant cell biochemical parameters. In the visible light band, the main absorption peak of healthy green plants is formed in the vicinity of the red light band and the blue light band, and the main reflection peak is formed in the green light band. The spectral reflectivities of the wave bands are combined linearly or nonlinearly to form a vegetation index, which can be used for diagnosing the growth state of economic plants and inverting various economic plant parameters. The obtained multispectral image band characteristics select 27 multispectral vegetation indexes from the existing multispectral vegetation indexes for model construction, and a specific calculation formula is shown in table 1.
TABLE 1 empirical vegetation index and formula
B、G、R、RE 1，2 NIR represents the reflectance in the blue, green, red 1, red 2 and near infrared bands, respectively.
A Support Vector Machine (SVM) is a binary linear classifier based on an optimal interval classification technique, and has been generalized to the use of nonlinear data and multi-class data; the Back Propagation Neural Network (BPNN) is a multi-layer feedforward neural network and is widely applied to the aspects of pattern recognition and classification, system simulation, intelligent fault diagnosis, image processing, function fitting, optimal prediction and the like; the Random Forest (RF) is a supervised learning classification algorithm which randomly generates a plurality of classification trees proposed by Breiman and predicts the category of a sample to be detected according to a model obtained by training, and is characterized by strong generalization capability, good robustness, high speed, high precision and the like.
The model fitting result uses the decision coefficient (R 2 ) Evaluation of Root Mean Square Error (RMSE), relative error to average (MRE), R 2 The closer to 1, the higher the prediction accuracy of the model; the closer the RMSE and the MRE are to 0, the better the model effect is, the more the prediction result is concentrated, and the calculation formula is as follows:
in the middle of
Is a model predictive value; a, a i Is the actual sampling value; />
Is the average value; n is the number of samples.
A total of 66 sets of yield samples were obtained based on field trials, as well as a sample of light spectrum data for Ganfang No. 1 at maturity (66 sets total). And (3) sequencing 66 groups of yield samples from small to large, randomly selecting 2/3 samples as a modeling set, and using the rest 1/3 samples as a verification set, wherein the statistical characteristics of the number of samples of the modeling set and the verification set and the essential oil yield data are shown in table 2.
TABLE 2 descriptive statistics of camphor tree essential oil yield
The variation coefficients of the modeling set and the verification set are both larger than 15%, which indicates that the yield data is influenced by test processing, the degree of dispersion is large, and modeling analysis has smaller influence on algorithms with large noise tolerance.
Based on the 27 spectrum indexes obtained through calculation, the correlation between the vegetation indexes and the essential oil yield is analyzed (table 3), five indexes with the highest correlation coefficient (r) are screened out, and the modified soil adjustment vegetation indexes MSAVI, the optimized soil adjustment vegetation indexes OSAVI, the renormalized vegetation indexes RDVI, the soil adjustment vegetation indexes SAVI and the nonlinear vegetation indexes NLI are used as input variables of a model, wherein the correlation coefficients between the vegetation indexes and the essential oil yield are 0.7651, 0.8131, 0.7711, 0.7794 and 0.8183.
TABLE 3 dependence of essential oil yield of camphor tree on vegetation index
The index is 5 vegetation indexes screened.
The screened vegetation index combination is taken as an independent variable, camphor tree essential oil yield is taken as a response variable, an SVM established based on a Libsvm toolbox, a BPNN established based on RF of parameter optimization and training and a Neal-Net-work toolbox based on MATLAB are respectively adopted to establish an estimation model, and R is taken as a model 2 And comprehensively evaluating model accuracy in the aspects of RMSE and MRE 3.
The results show that SVM, RF, BPNN three model results are compared with the measured values, as shown in fig. 2a, 2b, 2c, where the decision coefficients of the modeled set (R 2 ) 0.723, 0.853, 0.770, respectively; root Mean Square Error (RMSE) of 11.649k respectivelyg measurement pair -2 Measured value of 9.179kgg -2 10.484kg measurement pair -2 The method comprises the steps of carrying out a first treatment on the surface of the Average relative error (MRE) is 7.204, 10.808, 7.181, respectively; and R of the verification set 2 0.688, 0.869, 0.732, respectively; RMSE is 7.951kgg measured value -2 Measured value of 5.809kgg -2 Measured value of 8.483kgg -2 The method comprises the steps of carrying out a first treatment on the surface of the MREs were 6.914, 5.545, 7.999, respectively. From the inversion results, a comparison of the accuracy between the models was made, as shown in table 4. RF inversion accuracy is highest in comprehensive view, R of modeling set and verification set 2 0.853, 0.869, respectively, and the RMSE and MRE of the validation set are smaller than the other two sets of algorithms.
Table 4 model accuracy comparison
Conclusion: (1) The vegetation index MSAVI, OSAVI, RDVI, SAVI and NLI of the short-forest linalool are good in correlation with the yield, and the yield and growth vigor of the short-forest linalool can be monitored in real time through the change of the 5 vegetation indexes.
(2) The yield of the short-forest camphor tree essential oil is closely related to the reflectivity R of a red light wave band (the central wavelength is 660 nm) and the reflectivity NIR of a near infrared wave band (the central wavelength is 840 nm), and the vegetation index obtained based on the two wave bands can well invert the yield of the short-forest camphor tree essential oil.
(3) The accuracy of the short forest linalool essential oil yield prediction model constructed based on the RF model is obviously higher than that of SVM and BPNN models. Thus, the RF model can be used as a first-choice method for modeling the yield of the camphor essential oil in the large short forest.
Corresponding to the method shown in fig. 1, the embodiment of the invention also provides a system for predicting the yield of the short-forest linaloe essential oil by the multispectral unmanned aerial vehicle, which is used for realizing the method shown in fig. 1, and the structural schematic diagram is shown in fig. 3, and specifically comprises the following steps: the system comprises a data acquisition module, a data processing module, a model construction module, a selected model module and a result analysis module;
and a data acquisition module: the system comprises a data processing module, a data processing module and a data processing module, wherein the data processing module is used for acquiring output data of the short forest linaloe essential oil and multispectral image data shot by an unmanned aerial vehicle;
and a data processing module: the device is connected with the input end of the model building module and is used for performing splicing processing on the acquired multispectral image data to obtain spectral reflectivity data of different wave bands; measuring the yield of the short-forest linalool essential oil by using a method for measuring each plant;
model construction module: the system comprises a model module, a vegetation index module, a multi-spectral vegetation index module and a data processing module, wherein the model module is connected with the input end of the model module and is used for forming a vegetation index by linear or nonlinear combination of spectral reflectivity data, and the multi-spectral vegetation index is selected to be used for constructing various spectral index models;
selecting a model module: the system comprises a result analysis module, a model evaluation index selection module, a correlation analysis module and a model analysis module, wherein the result analysis module is connected with the input end of the result analysis module and is used for selecting a plurality of constructed models and performing correlation analysis on the plurality of constructed models and the measured essential oil yield to obtain an optimal model;
and a result analysis module: the method is used for rapidly predicting the essential oil yield of the camphor tree by utilizing the optimal model so as to realize decision making on fertigation management of the camphor tree.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (5)
1. A method for predicting the yield of the essential oil of the short-forest linaloe by a multispectral unmanned aerial vehicle, which is characterized by comprising the following steps:
s1, data acquisition: acquiring data of the yield of the short forest linalool essential oil and multispectral image data shot by the unmanned aerial vehicle;
s2, data processing: splicing the multispectral image data in the step S1 to obtain spectral reflectivity data of different wavebands; measuring the yield of the short-forest linalool essential oil in the S1 by using a method for measuring each plant;
s3, constructing a model: the spectral reflectivity data in the step S2 are combined linearly or nonlinearly to form vegetation indexes, and multispectral vegetation indexes are selected to be used for constructing various spectral index models;
s4, selecting a model: selecting a model evaluation index to select a plurality of spectrum index models constructed in the step S3, and performing correlation analysis on the spectrum index models and the essential oil yield measured in the step S2 to obtain an optimal model;
s5, result analysis: and (3) rapidly predicting the essential oil yield of the camphor tree by using the optimal model so as to realize decision making on fertigation management of the camphor tree.
2. The method for predicting the yield of the essential oil of the short-forest linaloe by the multispectral unmanned aerial vehicle according to claim 1, wherein,
the measurement formula of the content of the short-forest linalool essential oil in the S2 is as follows:
wherein i and j are positive integers, O i The unit is kg hm -2 ，G Bottle j + oil j The unit of G is G, G is the weight of the j tree essential oil and the bottle Leaf j Measuring the weight of the leaf sample of the oil for the j-th tree, G Total leaf j Is the total leaf weight of the j-th tree.
3. The method for predicting the yield of the essential oil of the short-forest linaloe by the multispectral unmanned aerial vehicle according to claim 1, wherein,
the index model in S3 includes, but is not limited to, a support vector machine model, a back propagation neural network model, and a random forest model.
4. The method for predicting the yield of the essential oil of the short-forest linaloe by the multispectral unmanned aerial vehicle according to claim 1, wherein,
s4, specific contents of model evaluation indexes are as follows:
model fitting results are adoptedDetermining the coefficient R 2 The root mean square error RMSE and the average relative error MRE are evaluated to determine a coefficient R 2 The closer to 1, the higher the prediction accuracy of the model; the closer the root mean square error RMSE is to 0 with the average relative error MRE, the better the model effect is, the more the prediction result is concentrated, and the calculation formula is as follows:
in the method, in the process of the invention,
is a model predictive value; a, a i Is the actual sampling value; />
Is the average value; n is the number of samples.
5. A system for predicting the yield of the essential oil of the cinnamomum camphora in a short forest by using the multi-spectrum unmanned aerial vehicle, which is characterized in that the method for predicting the yield of the essential oil of the cinnamomum camphora in a short forest by using the multi-spectrum unmanned aerial vehicle according to any one of claims 1 to 4 comprises the following steps: the system comprises a data acquisition module, a data processing module, a model construction module, a selected model module and a result analysis module;
and a data acquisition module: the system comprises a data processing module, a data processing module and a data processing module, wherein the data processing module is used for acquiring output data of the short forest linaloe essential oil and multispectral image data shot by an unmanned aerial vehicle;
and a data processing module: the device is connected with the input end of the model building module and is used for performing splicing processing on the acquired multispectral image data to obtain spectral reflectivity data of different wave bands; measuring the yield of the short-forest linalool essential oil by using a method for measuring each plant;
model construction module: the system comprises a model module, a vegetation index module, a multi-spectral vegetation index module and a data processing module, wherein the model module is connected with the input end of the model module and is used for forming a vegetation index by linear or nonlinear combination of spectral reflectivity data, and the multi-spectral vegetation index is selected to be used for constructing various spectral index models;
selecting a model module: the system comprises a result analysis module, a model evaluation index selection module, a correlation analysis module and a model analysis module, wherein the result analysis module is connected with the input end of the result analysis module and is used for selecting a plurality of constructed models and performing correlation analysis on the plurality of constructed models and the measured essential oil yield to obtain an optimal model;
and a result analysis module: the method is used for rapidly predicting the essential oil yield of the camphor tree by utilizing the optimal model so as to realize decision making on fertigation management of the camphor tree.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310161507.4A CN116151454A (en) | 2023-02-24 | 2023-02-24 | Method and system for predicting yield of short-forest linalool essential oil by multispectral unmanned aerial vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310161507.4A CN116151454A (en) | 2023-02-24 | 2023-02-24 | Method and system for predicting yield of short-forest linalool essential oil by multispectral unmanned aerial vehicle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116151454A true CN116151454A (en) | 2023-05-23 |
Family
ID=86350455
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310161507.4A Pending CN116151454A (en) | 2023-02-24 | 2023-02-24 | Method and system for predicting yield of short-forest linalool essential oil by multispectral unmanned aerial vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116151454A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117746168A (en) * | 2024-02-21 | 2024-03-22 | 深圳块织类脑智能科技有限公司 | Intelligent pine forest nematode disease detection method based on unmanned aerial vehicle multispectral |
-
2023
- 2023-02-24 CN CN202310161507.4A patent/CN116151454A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117746168A (en) * | 2024-02-21 | 2024-03-22 | 深圳块织类脑智能科技有限公司 | Intelligent pine forest nematode disease detection method based on unmanned aerial vehicle multispectral |
| CN117746168B (en) * | 2024-02-21 | 2024-05-07 | 深圳块织类脑智能科技有限公司 | Intelligent pine forest nematode disease detection method based on unmanned aerial vehicle multispectral |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110287944B (en) | Crop pest monitoring method based on multispectral remote sensing image of deep learning | |
| Pôças et al. | Predicting grapevine water status based on hyperspectral reflectance vegetation indices | |
| Kızıl et al. | Lettuce (Lactuca sativa L.) yield prediction under water stress using artificial neural network (ANN) model and vegetation indices. | |
| Tatsumi et al. | Prediction of plant-level tomato biomass and yield using machine learning with unmanned aerial vehicle imagery | |
| CN107271372A (en) | A kind of Apple Leaves chlorophyll remote sensing estimation method | |
| CN112287886B (en) | Wheat plant nitrogen content estimation method based on hyperspectral image fusion map features | |
| CN117036088A (en) | Data acquisition and analysis method for identifying growth situation of greening plants by AI | |
| CN114821321A (en) | Blade hyperspectral image classification and regression method based on multi-scale cascade convolution neural network | |
| Tosin et al. | Canopy VIS-NIR spectroscopy and self-learning artificial intelligence for a generalised model of predawn leaf water potential in Vitis vinifera | |
| CN116151454A (en) | Method and system for predicting yield of short-forest linalool essential oil by multispectral unmanned aerial vehicle | |
| Zhao et al. | Detection of cotton waterlogging stress based on hyperspectral images and convolutional neural network | |
| Xiao et al. | Comparison of leaf chlorophyll content retrieval performance of citrus using FOD and CWT methods with field-based full-spectrum hyperspectral reflectance data | |
| Quan et al. | Multimodal remote sensing application for weed competition time series analysis in maize farmland ecosystems | |
| Ranjan et al. | Thermal-RGB imagery and in-field weather sensing derived sweet cherry wetness prediction model | |
| Guo et al. | Predicting Fv/Fm and evaluating cotton drought tolerance using hyperspectral and 1D-CNN | |
| CN116883874A (en) | Evaluation method and system for comprehensive growth vigor of cinnamomum camphora dwarf forest | |
| Prasad et al. | Hyperspectral vegetation indices offer insights for determining economically optimal time of harvest in Mentha arvensis | |
| CN115049902B (en) | Visual prediction method, system, equipment and storage medium for water content of citrus blade | |
| Galvez-Valencia et al. | Predictive model of water stress in Tenera oil palm by means of spectral signature methods | |
| Wang et al. | Optimal waveband identification for estimation of leaf area index of paddy rice | |
| Li et al. | Establishment of deep learning model for the growth of tea cutting seedlings based on hyperspectral imaging technique | |
| CN114199793B (en) | Hyperspectral-based apple leaf nitrogen content detection method | |
| Jo et al. | Determining canopy growth conditions of paddy rice via ground-based remote sensing | |
| Yu et al. | Collaborative inverse modeling of nitrogen and phosphorus content in rice based on WOA-ELM | |
| CN116187478A (en) | Corn overground biomass estimation method and system |
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 |