CN110333195A - Water content in plant leaf detection method and device - Google Patents

Water content in plant leaf detection method and device Download PDF

Info

Publication number
CN110333195A
CN110333195A CN201910635811.1A CN201910635811A CN110333195A CN 110333195 A CN110333195 A CN 110333195A CN 201910635811 A CN201910635811 A CN 201910635811A CN 110333195 A CN110333195 A CN 110333195A
Authority
CN
China
Prior art keywords
hyperion
leaf
spectral data
sample
plant leaf
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
Application number
CN201910635811.1A
Other languages
Chinese (zh)
Inventor
王延仓
李笑芳
金永涛
顾晓鹤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
North China Institute of Aerospace Engineering
Original Assignee
North China Institute of Aerospace Engineering
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by North China Institute of Aerospace Engineering filed Critical North China Institute of Aerospace Engineering
Priority to CN201910635811.1A priority Critical patent/CN110333195A/en
Publication of CN110333195A publication Critical patent/CN110333195A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/17Systems in which incident light is modified in accordance with the properties of the material investigated
    • G01N21/25Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
    • G01N21/31Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/17Systems in which incident light is modified in accordance with the properties of the material investigated
    • G01N21/55Specular reflectivity
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84Systems specially adapted for particular applications
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N5/00Analysing materials by weighing, e.g. weighing small particles separated from a gas or liquid
    • G01N5/04Analysing 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
    • G01N5/045Analysing 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 for determining moisture content
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84Systems specially adapted for particular applications
    • G01N2021/8466Investigation of vegetal material, e.g. leaves, plants, fruits
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2201/00Features of devices classified in G01N21/00
    • G01N2201/12Circuits of general importance; Signal processing
    • G01N2201/129Using chemometrical methods

Abstract

The present invention provides a kind of water content in plant leaf detection method and device, is related to detection technique field.The detection method of the water content in plant leaf, comprising: obtain the high-spectral data of plant leaf blade to be measured;According to preset diagnostic model, the high-spectral data for treating measuring plants blade is handled, to obtain the diagnostic result of diagnostic model output, diagnostic model includes: the high-spectral data of preset plant leaf blade and the corresponding relationship of moisture content, and diagnostic result is the moisture content of the plant leaf blade to be measured.The plant leaf blade spectroscopic data of measurement is input in the diagnostic model of foundation by the application, can directly export the moisture content of the plant leaf blade, output result is accurate, and improves timeliness.

Description

Water content in plant leaf detection method and device
Technical field
The present invention relates to detection technique fields, in particular to a kind of water content in plant leaf detection method and dress It sets.
Background technique
Water is material element necessary to plant leaf blade growth and development, is that mineral nutrients transport and chloroplaset carry out physiology The main medium of biochemical reaction.Plant leaf blade interior tissue water content has preferable indicative function to topsoil soils soil moisture content, can Reflect influence of the drought stress to plant growth and development, and within certain limits, the reduction of water content in plant leaf facilitates Plant seed protein content is promoted, plant seed level of quality is improved.Therefore how real-time to water content in plant leaf progress, Precisely detection, is a major issue.
In the prior art, acquire multiple leaf samples, and drying and processing carried out to each leaf sample, and according to drying at Leaf sample after reason is handled, and the water content of each blade is obtained, and then obtains the leaf water content in sample region.
But the above method, when detecting to crops leaf water content, poor in timeliness is at high cost.
Summary of the invention
It is an object of the present invention in view of the deficiency of the prior art, provide a kind of water content in plant leaf inspection Survey method and device, it is poor in timeliness when solving the problems, such as leaf water content detection in the prior art, at high cost.
To achieve the above object, technical solution used in the embodiment of the present invention is as follows:
In a first aspect, the embodiment of the invention provides a kind of detection methods of water content in plant leaf, comprising: obtain to The high-spectral data of measuring plants blade;
According to preset diagnostic model, the high-spectral data for treating measuring plants blade is handled, to obtain diagnostic model The diagnostic result of output, diagnostic model include: the high-spectral data of preset plant leaf blade and the corresponding relationship of moisture content, are examined Disconnected result is the moisture content of plant leaf blade to be measured.
Further, according to preset diagnostic model, treat measuring plants blade high-spectral data handled before, should The detection method of water content in plant leaf further include:
The high-spectral data of leaf sample is obtained, and according to the high-spectral data of leaf sample, obtains EO-1 hyperion characteristic wave Section;
Correlation analysis is carried out to the moisture content of EO-1 hyperion characteristic wave bands and leaf sample;
According to correlation analysis as a result, establishing diagnostic model.
Further, according to the high-spectral data of leaf sample, EO-1 hyperion characteristic wave bands are obtained, comprising:
EO-1 hyperion sensitive band is extracted from the high-spectral data of leaf sample, the EO-1 hyperion sensitive band is to blade The influence wave band of moisture content tool;
EO-1 hyperion sensitive band is handled using Wavelet Transformation Algorithm, obtains EO-1 hyperion characteristic wave bands.
Further, leaf sample includes: trained sample sets;To the moisture content of EO-1 hyperion characteristic wave bands and leaf sample Carry out correlation analysis, comprising: to the moisture content of the training obtained EO-1 hyperion characteristic wave bands of sample sets and training sample sets Carry out correlation analysis.
Further, leaf sample further include: verification sample collection;The detection method of the water content in plant leaf is also wrapped It includes:
The EO-1 hyperion characteristic wave bands according to obtained by verification sample collection are handled using least square method, to be diagnosed The detection accuracy of model;
According to detection accuracy, processing is optimized to diagnostic model.
Further, the detection method of the water content in plant leaf, further includes:
EO-1 hyperion sensitive band is handled using spectrum smoothing method, obtains smooth EO-1 hyperion wave band;
Smooth EO-1 hyperion wave band, EO-1 hyperion wave band after being denoised are handled using derivative transformation method;
EO-1 hyperion wave band after denoising, the EO-1 hyperion wave band after being optimized are handled using continuum minimizing technology.
Further, the EO-1 hyperion sensitive band is handled using Wavelet Transformation Algorithm, obtains bloom spectrum signature Wave band, comprising: the EO-1 hyperion sensitive band after optimization is handled using Wavelet Transformation Algorithm, obtains the bloom spectrum signature Wave band.
Second aspect, the embodiment of the invention also provides a kind of detection device of water content in plant leaf, the device packets It includes:
Module is obtained, for obtaining the high-spectral data of plant leaf blade to be measured;
Processing module is used for according to preset diagnostic model, and the high-spectral data for treating measuring plants blade is handled, with The diagnostic result of diagnostic model output is obtained, diagnostic model includes: the high-spectral data and moisture content of preset plant leaf blade Corresponding relationship, diagnostic result be the plant leaf blade to be measured moisture content.
Further, the detection device of the water content in plant leaf further include: analysis module establishes module;
Module is obtained, is also used to obtain the high-spectral data of leaf sample, and according to the high-spectral data of leaf sample, obtain To EO-1 hyperion characteristic wave bands;
Analysis module carries out correlation analysis for the moisture content to EO-1 hyperion characteristic wave bands and leaf sample;
Establish module, for according to correlation analysis as a result, establishing diagnostic model.
Further, module is obtained, is also used to extract EO-1 hyperion sensitive band from the high-spectral data of leaf sample, it is high Spectrum sensitive wave band is the influence wave band to leaf water content;Using Wavelet Transformation Algorithm to EO-1 hyperion sensitive band at Reason, obtains EO-1 hyperion characteristic wave bands.
Further, the detection device of the water content in plant leaf, further includes: training module, leaf sample include: instruction Practice sample sets;Training module is used for the moisture content to the training obtained EO-1 hyperion characteristic wave bands of sample sets and training sample sets Carry out correlation analysis.
Further, the detection device of the water content in plant leaf further include: authentication module;Leaf sample further include: Verification sample collection;Authentication module is used to carry out the EO-1 hyperion characteristic wave bands according to obtained by verification sample collection using least square method Processing, to obtain the detection accuracy of diagnostic model;According to detection accuracy, processing is optimized to diagnostic model.
Further, module is obtained, is also used to handle EO-1 hyperion sensitive band using spectrum smoothing method, is obtained smooth high Spectral band;Smooth EO-1 hyperion wave band, EO-1 hyperion wave band after being denoised are handled using derivative transformation method;It is gone using continuum EO-1 hyperion wave band except EO-1 hyperion wave band after method processing denoising, after being optimized.
Further, module is obtained, is also used to carry out the EO-1 hyperion sensitive band after optimization using Wavelet Transformation Algorithm Processing, obtains EO-1 hyperion characteristic wave bands.
The beneficial effects of the present invention are: a kind of detection method of water content in plant leaf provided by the invention, comprising: obtain Take the high-spectral data of plant leaf blade to be measured;According to preset diagnostic model, the high-spectral data for treating measuring plants blade is carried out Processing, to obtain the diagnostic result of diagnostic model output, diagnostic model includes: the high-spectral data and water of preset plant leaf blade Divide the corresponding relationship of content, diagnostic result is the moisture content of the plant leaf blade to be measured.The application is by the plant leaf blade of measurement Spectroscopic data is input in the diagnostic model of foundation, can directly export the moisture content of the plant leaf blade, and output result is accurate, and Improve timeliness.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached Figure is briefly described, it should be understood that the following drawings illustrates only certain embodiments of the present invention, therefore is not construed as pair The restriction of range for those of ordinary skill in the art without creative efforts, can also be according to this A little attached drawings obtain other relevant attached drawings.
Fig. 1 is a kind of water content in plant leaf detection method flow diagram provided by the present application;
Fig. 2 is the water content in plant leaf detection method flow diagram that one embodiment of the application provides;
Fig. 3 is the water content in plant leaf detection method flow diagram that another embodiment of the application provides;
Fig. 4 is the water content in plant leaf detection method flow diagram that the another embodiment of the application provides;
Fig. 5 is the experimental result picture that one embodiment of the application provides;
Fig. 6 is the experimental result picture that another embodiment of the application provides;
Fig. 7 is the water content in plant leaf detection method flow diagram that the another embodiment of the application provides;
Fig. 8 is water content in plant leaf detection device schematic diagram provided by the present application;
Fig. 9 is the water content in plant leaf detection device schematic diagram that one embodiment of the application provides;
Figure 10 is the water content in plant leaf detection device schematic diagram that the another embodiment of the application provides;
Figure 11 is the network equipment infrastructure schematic diagram that one embodiment of the application provides.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is A part of the embodiment of the present invention, instead of all the embodiments.
Fig. 1 is a kind of water content in plant leaf detection method flow diagram provided by the present application, and this method can be by There is the equipment of processing function to execute for computer, server etc., as shown in Figure 1, this method comprises:
S110, the high-spectral data for obtaining plant leaf blade to be measured.
Since the moisture content of plant leaf blade to the growth and development of plant and has instruction well to topsoil soils soil moisture content Effect, therefore the moisture content for studying plant leaf blade is of great significance.
Wherein, above-mentioned plant leaf blade to be measured can be the various plants such as trees, crops, flower, grass, to indicated by measuring plants Specific plant variety herein with no restrictions.
For being the winter wheat in crops to measuring plants.It obtains the process for doing the high-spectral data of wheat leaf blade Are as follows: Wheat Leavess when choosing the field winter wheat plot that area is larger, landform is flat, and choosing in blooming stage are as sample Product.Wherein, a certain number of Wheat Leavess are won at random in the biggish plot of area, for example, winning at random in plot 30 Wheat Leavess use the EO-1 hyperion number of field spectroradiometer measurement measurement 30 Wheat Leavess samples as sample According to real-time transmission to server.
Further, the interference due to uncontrollable factors such as steam, aerosols by external environment and spectrum measurement instruments Itself existing error causes the high-spectral data obtained to be inevitably present noise;Utmostly to weaken high-frequency noise Negative effect promotes spectroscopic data signal-to-noise ratio, extracts Screening Treatment to the wave band in high-spectral data.
It should be noted that the spectral region of optical spectrum instrumentation measurement is 350nm~2500nm, in high-spectral data The process that wave band extracts Screening Treatment can adopt wave band of the low-pass filter to spectroscopic data in the spectrum stage of telling somebody what one's real intentions are carry out it is flat Sliding processing, and positioned at the wave band in high light spectrum stage because there are a large amount of high-frequency noises, signal-to-noise ratio is too low, it can be by higher spectrum number It is handled according to using deletion mode.
Explanation is needed further exist for, if spectroscopic data in the range of 350nm~1074nm frequency range, can determine the light Modal data is located at the lower spectrum stage, is smoothed to the spectroscopic data for being located at the stage using low-pass filter.If light Modal data then can determine that the spectroscopic data is located at the high light spectrum stage, to being located in the range of 1075nm~2500nm frequency range The spectroscopic data in the stage can carry out delete processing.It should be pointed out that above-mentioned lower spectrum stage and high light compose the stage It is limitation, above-mentioned only possible example with above range that data, which divide range not,.
Further, the selection of above-mentioned plant leaf blade is not limitation, and the sample of Wheat Leavess with Wheat Leavess Choosing quantity is also limitation not with 30 leaf samples.
S120, according to preset diagnostic model, the high-spectral data for treating measuring plants blade is handled, with obtain diagnosis Model output diagnostic result, diagnostic model be preset plant leaf blade high-spectral data and moisture content corresponding relationship, Diagnostic result is the moisture content of plant leaf blade to be measured.
Above-mentioned diagnostic model be based on least-squares algorithm obtain about plant leaf blade with the diagnostic model for dividing content.It should In method, after the high-spectral data of the plant leaf blade to be measured obtained through step S110 being handled, it is input to diagnostic model In, it can get the moisture content of the Wheat Leavess of the random acquisition.
Wherein, in this method, the high-spectral data of the plant leaf blade to be measured can be screened, and according to preset diagnosis Model handles obtained high-spectral data is screened, can effectively reduce the difficulty and complexity of data processing.
Further, above-mentioned diagnostic model includes: the high-spectral data of preset plant leaf blade and the correspondence of moisture content Relationship, the acquisition process of the corresponding relationship can EO-1 hyperion numbers on the moisture content to obtained plant leaf blade and its plant leaf blade According to correlation analysis is carried out, further to obtain the relationship of high-spectral data Yu leaf water content.
The detection method of water content in plant leaf provided by the present application passes through the moisture content and spectrum to plant leaf blade Data establish diagnostic model using least-squares algorithm, only the plant spectral data of real-time measurement need to be input to diagnostic model In, the moisture content of the plant leaf blade will be exported, testing result is accurate, and saves the time.
Fig. 2 is the water content in plant leaf detection method flow diagram that one embodiment of the invention provides, such as Fig. 2 institute Show, before step S120, this method comprises:
S111, the high-spectral data for obtaining leaf sample, and according to the high-spectral data of leaf sample, obtain EO-1 hyperion spy Levy wave band.
Since the high-spectral data measurement detection test period span of leaf sample is long, the leaf sample of each period is raw Educating the phase can all be slightly changed, therefore carry out time standard to test data using linear interpolation.Wherein, by optical spectrum instrumentation Collected EO-1 hyperion digital signal mainly includes response signal and system noise of the measuring instrument to blade Wheat Leavess sample. Measuring instrument is valuable signal to the response signal of winter wheat, needs to retain, and system noise needs to remove, in addition to this, leaf Ambient noise and spectrum noise in piece sample spectra curve are also required to be eliminated.
Further, it for high-spectral data, need to be calculated by spectrum smoothing algorithm, derivative transformation algorithm and continuum removal The transformation of three processes of method removes system noise, and the high-spectral data after transformation is further used wavelet transformation EO-1 hyperion characteristic wave bands can be obtained in technology.
S112, correlation analysis is carried out to the moisture content of EO-1 hyperion characteristic wave bands and leaf sample.
Wherein, the moisture content of sample blade can be for using the obtained water content of physical detection methods.Then to leaf It, can be to each leaf of acquisition before the moisture content of piece sample and its high-spectral data on leaf sample carry out correlation analysis Piece sample is weighed as weighed using electronic scale, obtains the weight in wet base of each leaf sample, and to each leaf sample into Leaf sample is such as placed in baking box by row baking preset time carries out baking certain time, so that not depositing in each leaf sample In moisture, then weigh each leaf sample after baking to obtain the dry weight of each leaf sample, and according to each The weight in wet base and dry weight of leaf sample, obtain the moisture content of each leaf sample, such as according to weight in wet base subtract dry weight can obtain it is each The practical moisture content of leaf sample.
Further, it may be determined that pair between the spectroscopic data of each leaf sample and the moisture content of each leaf sample It should be related to.Correlation analysis is such as carried out by the moisture content of spectroscopic data and each leaf sample to each leaf sample, It is determined according to correlation results corresponding between the spectroscopic data of each leaf sample and the moisture content of each leaf sample Relationship.So far, the corresponding relationship between the spectroscopic data of each leaf sample and the moisture content of each leaf sample is obtained, Establish the diagnostic model.After the diagnostic model is established, can the obtained high-spectral data of plant leaf blade to be measured of input measurement, It can determine that moisture content corresponding to the high-spectral data of the plant leaf blade to be measured is the moisture content of the plant leaf blade to be measured.
S113, according to correlation analysis as a result, establishing diagnostic model.
The analysis result of the correlation is for example can include: at least one set of high-spectral data and every group of high-spectral data institute are right The leaf water content answered.For example, it is a that the high-spectral data of leaf sample, which is leaf water content corresponding to A,;Leaf sample High-spectral data be leaf water content corresponding to B be b.
It should be noted that above-mentioned high-spectral data A, B can be specific spectra values, or EO-1 hyperion numerical value Interval value, specific value or specific section indicated by A, B herein with no restrictions, the data that actual tests of being subject to obtain.
Correspondingly, above-mentioned leaf water content a, b can be specific moisture content value, or moisture content section Value, specific value or specific section indicated by a, b herein with no restrictions, the data that actual tests of being subject to obtain.
The EO-1 hyperion characteristic wave bands that will be obtained through step S111, using least-squares algorithm according to EO-1 hyperion characteristic wave bands The structure for carrying out correlation analysis with the moisture content of leaf sample constructs moisture content diagnostic model.
Fig. 3 is the water content in plant leaf detection method flow diagram that one embodiment of the application provides, such as Fig. 3 institute Show, step S111 includes:
S1111, EO-1 hyperion sensitive band is extracted from the high-spectral data of leaf sample, EO-1 hyperion sensitive band is to leaf Piece moisture content has the wave band of influence.
As follows by taking the blade of winter wheat as an example, to the sensitive frequency range of extraction EO-1 hyperion from the high-spectral data of leaf sample It reason and is handled based on the EO-1 hyperion sensitivity frequency range extracted to obtain being said to the effect for establishing diagnostic model It is bright.
By taking Wheat Leavess sample as an example, divide different phase in growth and development according to winter wheat, external morphology therein and Some variations occur for external structure, so that winter wheat causes reflection bent in different growing or when coercing by varying environment The difference of line.With the normal development of winter wheat, leaves of winter wheat chip architecture can be caused to change due to factors such as climatic environments, Internal eucaryotic cell structure changes, and the chlorophyll content in Wheat Leavess reduces, and photosynthesis weakens, anthocyanidin, lutein Equal pigment contents increase, and red, blue light reflection is caused to increase, and emit near infrared light and reduce, and it is visible for spectrally just cashing The reflectivity of light region wave band obviously increases, and near infrared band range internal reflection rate reduces, and red side is mobile to shortwave direction, occurs Blue shifts of red edge phenomenon.Therefore, it is necessary to choose sensitive band of the winter wheat in development to be analyzed and processed, so that the diagnosis established Model is more accurate, to obtain more accurately water content in plant leaf testing result.
It is the reflectivity on each wave band to the hyperspectral information of plant leaf blade as acquired in optical spectrum instrumentation, as Basis carries out data mart modeling to hyper spectral reflectance, obtains various types of EO-1 hyperion parameters, tends to preferably reflect and plant The physicochemical property of object.The face that commonly used EO-1 hyperion parameter has characteristic parameter, bloom spectral curve based on bloom spectral position to constitute The hyperspectral index etc. that product characteristic parameter, the combination of different-waveband EO-1 hyperion are constituted.
Bloom spectral curve, which contains information abundant, different-waveband information and atural object difference element or significant condition, to be had not Same correlation, in order to effectively extract Target scalar feature, it usually needs various combinatorial operations are carried out to hyper spectral reflectance, Required information can be enhanced, eliminating interference information bring influences, and improves information extraction precision.
The building of traditional vegetation index is based on visible light region inner chlorophyll more and absorbs and blade knot near infrared band Relationship between structure Multiple Scattering, but physical and chemical index specific for crops, need to find that more optimized, susceptibility is higher Hyperspectral index.This research is joined from hyperspectral index construction mechanism, by different-waveband combination with winter wheat Physiology and biochemistry Number establishes correlativity, finds EO-1 hyperion sensitive band, constructs novel spectral index.
Therefore it after the high-spectral data obtained delete smoothly equal processing, also needs further to extract EO-1 hyperion sensitivity wave Section.Due to water content in plant leaf detection method provided by the present application, when extracting EO-1 hyperion sensitive stage, extraction Sensitive band is the wave band for having influence to the height of leaf water content.
S1112, EO-1 hyperion sensitive band is handled using Wavelet Transformation Algorithm, obtains EO-1 hyperion characteristic wave bands.
The hyper spectral reflectance of all Wheat Leavess samples of acquisition is subjected to difference by using continuous wavelet technology The processing of scale obtains the wavelet coefficient under multiple scales, and the continuous wavelet frequency spectrum of the signal after continuous wavelet transform is flexible is still Can be by the flexible acquisition of the wavelet spectrum of original signal, i.e. continuous wavelet transform can keep stretching constant, from conversion process Multiplication rule and inverse transformation reconstruction formula it is found that continuous wavelet is similar to Fourier transformation, can be realized signal transformation and letter Number reconstruct, and be able to maintain the complete of original information and have invertibity.
Wavelet transformation can extract the characteristic information most sensitive to plant physical and chemical parameter from spectral reflectance data. Continuous wavelet transform can be obtained on different-waveband and different scale by decomposing on multiple scales to high-spectral data Obtained small echo capacity factor is decomposed, is used by the spectral signature information that correlation analysis can preferably extract some sensitivities In the water and basis weight inverting of winter wheat.
Wherein, above-mentioned small echo capacity factor indicates to carry out wavelet transformation to a Setting signal, is exactly by the signal by certain Signal, i.e., is expressed as a series of linear combination of the wavelet function of different scales and different location by the expansion of one wavelet function cluster, Wherein the coefficient of each single item is known as small echo capacity factor, and under same scale the wavelet function of all different locations linear combination Referred to as Wavelet Component of the signal under the scale.
Further, Continuous Wavelet Transform by using a mother wavelet function ψ (λ) by spectral reflectivity f (λ), (λ=1,2 ..., n are wave band number) is decomposed into the wavelet energy coefficient of different scale, and the wavelet basis used is as follows:
In formula, the wave band for the curve of spectrum that λ is made of high-spectral data, a and b are positive real number.Wherein, a is default Contraction-expansion factor, indicate small echo width;B is preset shift factor, for determining the position of small echo.
Wavelet coefficient as a > 1, after decompositionWave-length coverage be greater than the wave-length coverage of ψ (λ), with a value by It is cumulative big,Wave-length coverage ratio ψ (λ) wave-length coverage increase amplitude become larger, wavelet transformation reflects wavelength at this time It is comparatively rough, and to the opposite fine of frequency reflection, this is corresponded to each other just with low frequency situation.As a < 1,'s Wave-length coverage is less than the wave-length coverage of ψ (λ), with being gradually reduced for a value,The wave-length coverage of wave-length coverage ratio ψ (λ) subtract Small amplitude becomes smaller, and wavelet transformation reflects frequency comparatively rough at this time, and then reflects to wavelength comparatively smart Carefully.
Further, Wavelet Transformation Algorithm is analyzed and processed using localization, is mainly based upon temporal frequency and sky Between refinement analysis is carried out to function in frequency, to effectively extract relevant information.Using continuous wavelet transform (Continuous Wavelet Transform, CWT) transformation of high-spectral data signal is carried out, and select a threshold value that will convert obtained small echo Coefficient carries out threshold value selection, and the coefficient after then being chosen according to threshold value carries out signal reconstruction, the signal after being denoised.Small echo becomes Function or signal can be decomposed by operations such as flexible and translations on multiple scales by changing, and preferably solution Fourier becomes Change cannot simultaneously analysis time domain and frequency domain defect.
Further, the application is write using matrix labotstory (Matrix Laboratory, abbreviation MATLAB) language Data processor based on Ha Er (Harr-like) wavelet basis completes the processing to canopy of winter wheat spectroscopic data, and with reality It surveys Wheat Leavess moisture data and carries out correlation analysis, screen and extract characteristic wave bands.Wherein Harr small echo is by a component The collection of functions of section constant function composition.This collection of functions be defined on half open interval [0,1) on, the number of each piecewise constant function Value is 1 in a small range, is elsewhere 0, now illustrates by taking image as an example and using the vector space in linear algebra Ha Er basic function.The Haar small echo wherein used is defined as follows:
Wherein, ψ (t) indicates Haar wavelet transform function related with time t.
Optionally, leaf water detection method of content provided by the present application, the leaf sample of acquisition include: trained sample Collection.Training sample sets can be all of the leaf sample of acquisition, alternatively, the part of the leaf sample of acquisition, such as therein three / bis-.The parameter in diagnostic model is established using the sample data in training sample sets.
Further, correlation analysis is carried out to the moisture content of EO-1 hyperion characteristic wave bands and leaf sample, comprising: to instruction The moisture content for practicing the obtained EO-1 hyperion characteristic wave bands of sample sets and training sample sets carries out correlation analysis.
For example, the Wheat Leavess sample of acquisition is 30, using 20 positions therein training sample sets, the part is used Leaf sample participate in diagnostic model training process.Wherein, it is wrapped in the sample number of Wheat Leavess sample and training sample sets The quantity contained is not limitation with the example above numerical value.
Referring to figure 4., Fig. 4 is the water content in plant leaf detection method process signal that the another embodiment of the application provides Figure.
After establishing diagnostic model using the data in training sample sets, model need to be further verified, therefore, the application mentions The object leaf water detection method of content of confession, leaf sample further include: verification sample collection.The verification sample collection can be to collect Leaf sample in part, as acquisition leaf sample in one third.
The detection method of the water content in plant leaf, further includes:
S210, the EO-1 hyperion characteristic wave bands according to obtained by verification sample collection are handled using least square method, with To the detection accuracy of diagnostic model.
The application determines the detection accuracy of diagnostic model using cross-validation method, determines formula are as follows:
Wherein,Indicate that detection accuracy, PRESS are Prediction sum squares, SS is error sum of squares, h is sample survey Number.
S220, according to detection accuracy, processing is optimized to diagnostic model.
Referring to figure 5., Fig. 5 is the experimental result picture that one embodiment of the application provides.Plant leaf blade spectroscopic data is through continuous After wavelet algorithm decomposes, by taking the spectroscopic data under the 1st scale as an example, the application uses least square method verification method are as follows: is building Before each step calculating of mould terminates, the cross validation of validity is carried out, if had in h stepThen model Precision reaches requirement, stops extracting principal component, otherwise needs to continue to extract.
Further, it is verified with the model accuracy that least square method constructs.It is extracted using correlation analysis algorithm sensitive Characteristic wave bands, using the diagnostic model of least square method building Wheat Leavess water content, the diagnostic model precision of building is used Return evaluation index: the coefficient of determination (R2) evaluated jointly with root-mean-square error (RMSE), the wherein coefficient of determination (R2) method be by Predicted value with only using mean value in the case where compare, observation performance quality.Its section is usually between (0,1).0 expression might as well The case where what is not predicted, directly takes mean value, and 1 indicates the case where all predictions are with legitimate reading perfect matching.It is calculated Method is as follows:
Wherein root-mean-square error (RMSE) is the mean value of the square-error root of predicted value and true value, is observation and true value The square root of the quadratic sum detection frequency n ratio of deviation, in actual measurement, observation frequency n is always limited, and true value can only be used Most believable (best) value replaces.Square error is very sensitive to the especially big or special small error reflection in one group of measurement, so, Root-mean-square error can be well reflected out the precision measured.
Above formula, n are detection number, SOMiFor actual value, SOMP is predicted value,For the average value of actual value.
Fig. 6 is please referred to, Fig. 6 is the experimental result picture that the another embodiment of the application provides.
With plant leaf blade spectroscopic data after the decomposition of continuous wavelet algorithm, the data of obtained 1-20 scale and diagnostic model If the Data Detection that as a result verification sample is concentrated goes out precision less than preset value, need to optimize diagnostic model processing, needs weight New screening EO-1 hyperion sensitive stage, and spectrum smoothing algorithm, derivative transformation algorithm and continuum are used to EO-1 hyperion sensitive stage Removal algorithm does optimization processing, and further uses the model after least-squares algorithm is optimized.
Fig. 7 is please referred to, Fig. 7 is the water content in plant leaf detection method process signal that the another embodiment of the application provides Figure.This method further include:
S310, EO-1 hyperion sensitive band is handled using spectrum smoothing method, obtains smooth EO-1 hyperion wave band.
By the collected spectral digital signal of optical spectrum instrumentation can include: two parts, i.e. optical spectrum instrumentation respond atural object Signal and system noise.It is generated when wherein system noise is mainly worked by each component part of detection system, in addition to this atural object The curve of spectrum also includes ambient noise and spectrum noise.The presence of noise brings very big to the analysis, detection, differentiation of object spectrum Interference.In order to eliminate these interference, required useful information is extracted from object spectrum, needs to be permitted existing for spectrum More " burr " noises are smoothly pre-processed.The principle of evaluation smoothing method superiority-inferiority is: in the spy of utmostly spectral preservation The curve of spectrum is as smooth as possible under the principle of value indicative, and it is smoothed after the curve of spectrum to the Physiology and biochemistry parameter of winter wheat Premeasure is more preferable.
According to the application selection by building least square method Mathematics Optimization Method to EO-1 hyperion feature-sensitive wave band into The smooth pretreatment of row.
It should be noted that spectrum smoothing algorithm provided by the present application is not to use the mathematics of building least square method excellent Change method be limitation, can also for the method for moving average, convolution (Savitaky-Golay, abbreviation SG) exponential smoothing, median filtering method, Gauss (Gaussian, abbreviation GS) filter method, low pass filtering method and Wavelet-denoising Method etc., herein with no restrictions.
S320, smooth EO-1 hyperion wave band, EO-1 hyperion wave band after being denoised are handled using derivative transformation method.
In leaf spectra analysis, derivative transformation method is the Derivative Operation that a kind of pair of curve of spectrum carries out different orders Method.Its purpose includes: can eliminate different degrees of ambient noise by the Derivative Operation of different end, can be improved not With the contrast of Absorption Characteristics, the spectrum characteristic parameters such as spectrum bending point, maximum value and minimum value can be determined.
Wherein, first derivative can eliminate linear and quadratic form ambient noise in EO-1 hyperion wave band region of variation.Second order is led Number can eliminate the influence of quadratic term ambient noise.First derivative calculation formula therein is as follows:
In formula, λ is wavelength, λiFor the wavelength of wave band i;R(λi) be i wave band wave band reflectivity;Δ λ is wavelength Xi-1It arrives Wavelength XiDistance;R′(λi) it is wavelength XiFirst derivative spectrum value.Wherein the too small then noise of Δ λ is big, influences subsequent modeling matter Amount;Useful information is lost it will cause smooth transition on the contrary, therefore the value of Δ λ is needed to carry out repeatedly to attempt could really It is fixed.
S330, EO-1 hyperion wave band after denoising, the EO-1 hyperion wave band after being optimized are handled using continuum minimizing technology.
Continuum is carried out to bloom spectral curve and goes division operation, the overall shape feature of available winter wheat spectrum.Due to The curve of spectrum is made of discrete sampling point, and the envelope of the curve of spectrum can be indicated with continuous spectrum curve approximation.Continuum The definition of removal is the peak point for connecting absorption or reflection protrusion with wavelength change point by point with line segment, and broken line is on peak point Diplomacy be greater than 180 °, accounting equation is as follows:
Rc(λ)=a λ+b
Wherein RcFor continuum line reflection rate;λ is wavelength location;A is continuum line intercept;B is continuum line slope.
Continuum removes reflectivity R ' (λi) reflectance value (R (λ)) that is equivalent at the spectral absorption at each wave band removes With continuum line reflection rate value (R at corresponding band po sitionc(λ)):
Curve of spectrum Absorption Characteristics after continuum removal normalization are highlighted, and the Absorption Characteristics of the curve of spectrum are brighter Aobvious, this method can correct position excursion caused by wave band reflectivity is relied on due to wave band, and by continuum removal, treated The curve of spectrum can be used for Spectral Characteristics Analysis and waveband selection.
EO-1 hyperion sensitive band after above-mentioned steps are handled, after being optimized.
Further, after the EO-1 hyperion sensitive band after step S330 is optimized, using Wavelet Transformation Algorithm to excellent EO-1 hyperion sensitive band after change is handled, and EO-1 hyperion characteristic wave bands are obtained, obtain the processes of EO-1 hyperion characteristic wave bands with it is upper It is identical to state embodiment process, does not add to repeat herein.
As shown in figure 8, the device includes: to obtain the present invention also provides a kind of detection device of water content in plant leaf Modulus block 501, processing module 502, in which:
Module 501 is obtained, for obtaining the high-spectral data of plant leaf blade to be measured.
Processing module 502 is used for according to preset diagnostic model, at the high-spectral data for treating measuring plants blade Reason, to obtain the diagnostic result of diagnostic model output, diagnostic model is that the high-spectral data of preset plant leaf blade and moisture contain The corresponding relationship of amount, diagnostic result are the moisture content of the plant leaf blade to be measured.
Fig. 9 is please referred to, Fig. 9 is the water content in plant leaf detection device schematic diagram that one embodiment of the application provides, should Device further include: analysis module 503 establishes module 504, in which:
Module 501 is obtained, is also used to obtain the high-spectral data of leaf sample, and according to the EO-1 hyperion number of leaf sample According to obtaining EO-1 hyperion characteristic wave bands.
Analysis module 503 carries out correlation analysis for the moisture content to EO-1 hyperion characteristic wave bands and leaf sample.
Establish module 504, for according to correlation analysis as a result, establishing diagnostic model.
Optionally, module 501 is obtained, is also used to extract EO-1 hyperion sensitive band from the high-spectral data of leaf sample, EO-1 hyperion sensitive band is the wave band for having influence to leaf water content.It is sensitive to EO-1 hyperion using Wavelet Transformation Algorithm Wave band is handled, and EO-1 hyperion characteristic wave bands are obtained.
Figure 10 is please referred to, Figure 10 is the water content in plant leaf detection device signal that the another embodiment of the application provides Figure.The device further include: training module 505.Leaf sample includes: trained sample sets, and training module 505 is used for training sample The moisture content for collecting obtained EO-1 hyperion characteristic wave bands and training sample sets carries out correlation analysis.
Further, please continue to refer to Figure 10, the detection device of the water content in plant leaf further include: authentication module 506, leaf sample further include: verification sample collection, authentication module 506 are used for using least square method to according to verification sample collection institute It obtains EO-1 hyperion characteristic wave bands to be handled, to obtain the detection accuracy of diagnostic model.According to detection accuracy, to diagnostic model into Row optimization processing.
Further, module 501 is obtained, is also used to handle EO-1 hyperion feature-sensitive wave band using spectrum smoothing method, obtain Obtain smooth EO-1 hyperion wave band.Smooth EO-1 hyperion wave band, EO-1 hyperion wave band after being denoised are handled using derivative transformation method.Using EO-1 hyperion wave band after the processing denoising of continuum minimizing technology, the EO-1 hyperion wave band after being optimized.
Further, obtain module 501, be also used to using Wavelet Transformation Algorithm to the EO-1 hyperion sensitive band after optimization into Row processing, obtains EO-1 hyperion characteristic wave bands.
The method that above-mentioned apparatus is used to execute previous embodiment offer, it is similar that the realization principle and technical effect are similar, herein not It repeats again.
The above module can be arranged to implement one or more integrated circuits of above method, such as: one Or multiple specific integrated circuits (Application Specific Integrated Circuit, abbreviation ASIC), or, one Or multi-microprocessor (digital singnal processor, abbreviation DSP), or, one or more field programmable gate Array (Field Programmable Gate Array, abbreviation FPGA) etc..For another example, when some above module passes through processing elements When the form of part scheduler program code is realized, which can be general processor, such as central processing unit (Central Processing Unit, abbreviation CPU) or it is other can be with the processor of caller code.For another example, these modules can integrate Together, it is realized in the form of system on chip (system-on-a-chip, abbreviation SOC).
Figure 11 is the network equipment infrastructure schematic diagram that one embodiment of the application provides.The device can integrate in terminal device Or the chip of terminal device, the terminal can be the calculating equipment for having processing function.
The device includes: memory 601, processor 602.
Memory 601 is for storing program, the program that processor 602 calls memory 601 to store, to execute the above method Embodiment.Specific implementation is similar with technical effect, and which is not described herein again.
Optionally, the present invention also provides a kind of program product, such as computer readable storage medium, including program, the journeys Sequence is when being executed by processor for executing above method embodiment.
In several embodiments provided by the present invention, it should be understood that disclosed device and method can pass through it Its mode is realized.For example, the apparatus embodiments described above are merely exemplary, for example, the division of the unit, only Only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components can be tied Another system is closed or is desirably integrated into, or some features can be ignored or not executed.Another point, it is shown or discussed Mutual coupling, direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING or logical of device or unit Letter connection can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme 's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list Member both can take the form of hardware realization, can also realize in the form of hardware adds SFU software functional unit.
The above-mentioned integrated unit being realized in the form of SFU software functional unit can store and computer-readable deposit at one In storage media.Above-mentioned SFU software functional unit is stored in a storage medium, including some instructions are used so that a computer Equipment (can be personal computer, server or the network equipment etc.) or processor (English: processor) execute this hair The part steps of bright each embodiment the method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (English: Read-Only Memory, abbreviation: ROM), random access memory (English: Random Access Memory, letter Claim: RAM), the various media that can store program code such as magnetic or disk.

Claims (10)

1. a kind of detection method of water content in plant leaf, which is characterized in that the described method includes:
Obtain the high-spectral data of plant leaf blade to be measured;
According to preset diagnostic model, the high-spectral data for treating measuring plants blade is handled, to obtain the diagnostic model The diagnostic result of output, the diagnostic model include: that the high-spectral data of preset plant leaf blade and the corresponding of moisture content are closed System, the diagnostic result are the moisture content of the plant leaf blade to be measured.
2. the method according to claim 1, wherein described according to preset diagnostic model, to the plant to be measured Before the high-spectral data of object blade is handled, the method also includes:
The high-spectral data of leaf sample is obtained, and according to the high-spectral data of the leaf sample, obtains EO-1 hyperion characteristic wave Section;
Correlation analysis is carried out to the moisture content of the EO-1 hyperion characteristic wave bands and the leaf sample;
According to the correlation analysis as a result, establishing the diagnostic model.
3. according to the method described in claim 2, it is characterized in that, the high-spectral data according to the leaf sample, obtains To EO-1 hyperion characteristic wave bands, comprising:
EO-1 hyperion sensitive band is extracted from the high-spectral data of the leaf sample, the EO-1 hyperion sensitive band is to blade The influence wave band of moisture content;
The EO-1 hyperion sensitive band is handled using Wavelet Transformation Algorithm, obtains the EO-1 hyperion characteristic wave bands.
4. according to the method described in claim 2, it is characterized in that, the leaf sample includes: trained sample sets;
The moisture content to the EO-1 hyperion characteristic wave bands and the leaf sample carries out correlation analysis, comprising:
It is related to the progress of the moisture content of the trained sample sets to the trained obtained EO-1 hyperion characteristic wave bands of sample sets Property analysis.
5. according to the method described in claim 4, it is characterized in that, the leaf sample further include: verification sample collection;
The method also includes:
The EO-1 hyperion characteristic wave bands according to obtained by the verification sample collection are handled using least square method, it is described to obtain The detection accuracy of diagnostic model;
According to the detection accuracy, processing is optimized to the diagnostic model.
6. according to the method described in claim 3, it is characterized in that, described sensitive to the EO-1 hyperion using Wavelet Transformation Algorithm Wave band is handled, before obtaining the EO-1 hyperion characteristic wave bands, the method also includes:
The EO-1 hyperion sensitive band is handled using spectrum smoothing method, obtains smooth EO-1 hyperion wave band;
The smooth EO-1 hyperion wave band, EO-1 hyperion wave band after being denoised are handled using derivative transformation method;
The EO-1 hyperion wave band using EO-1 hyperion wave band after the continuum minimizing technology processing denoising, after being optimized.
7. according to the method described in claim 6, it is characterized in that, described sensitive to the EO-1 hyperion using Wavelet Transformation Algorithm Wave band is handled, and EO-1 hyperion characteristic wave bands are obtained, comprising:
The EO-1 hyperion sensitive band after optimization is handled using Wavelet Transformation Algorithm, obtains the EO-1 hyperion characteristic wave Section.
8. a kind of detection device of water content in plant leaf, which is characterized in that described device includes:
Module is obtained, for obtaining the high-spectral data of plant leaf blade to be measured;
Processing module, for handling the high-spectral data of the plant leaf blade to be measured according to preset diagnostic model, with Obtain the diagnostic result of diagnostic model output, the diagnostic model include: preset plant leaf blade high-spectral data and The corresponding relationship of moisture content, the diagnostic result are the moisture content of the plant leaf blade to be measured.
9. device according to claim 8, which is characterized in that described device further include: analysis module establishes module, In:
The acquisition module is also used to obtain the high-spectral data of leaf sample, and according to the EO-1 hyperion number of the leaf sample According to obtaining EO-1 hyperion characteristic wave bands;
The analysis module carries out correlation point for the moisture content to the EO-1 hyperion characteristic wave bands and the leaf sample Analysis;
It is described to establish module, for according to the correlation analysis as a result, establishing the diagnostic model.
10. device according to claim 9, which is characterized in that the acquisition module is also used to, from the leaf sample High-spectral data in extract EO-1 hyperion sensitive band, the EO-1 hyperion sensitive band is influence wave to leaf water content Section;The EO-1 hyperion sensitive band is handled using Wavelet Transformation Algorithm, obtains the EO-1 hyperion characteristic wave bands.
CN201910635811.1A 2019-07-15 2019-07-15 Water content in plant leaf detection method and device Pending CN110333195A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910635811.1A CN110333195A (en) 2019-07-15 2019-07-15 Water content in plant leaf detection method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910635811.1A CN110333195A (en) 2019-07-15 2019-07-15 Water content in plant leaf detection method and device

Publications (1)

Publication Number Publication Date
CN110333195A true CN110333195A (en) 2019-10-15

Family

ID=68144977

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910635811.1A Pending CN110333195A (en) 2019-07-15 2019-07-15 Water content in plant leaf detection method and device

Country Status (1)

Country Link
CN (1) CN110333195A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110907367A (en) * 2019-11-09 2020-03-24 华东师范大学 Soil moisture content detection method based on hyperspectral remote sensing
CN112504977A (en) * 2020-11-17 2021-03-16 华南农业大学 Tea water content detection method and model construction method, medium and equipment thereof

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102495005A (en) * 2011-11-17 2012-06-13 江苏大学 Method for diagnosing crop water deficit through hyperspectral image technology
CN103472009A (en) * 2013-09-16 2013-12-25 南京农业大学 Method for monitoring wheat plant water content under different plant nitrogen content levels
CN103528965A (en) * 2013-08-28 2014-01-22 南京农业大学 Wheat leaf equivalent water thickness hyperspectral monitoring method
WO2019023712A1 (en) * 2017-07-28 2019-01-31 Cornell University In situ sensing of water potential

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102495005A (en) * 2011-11-17 2012-06-13 江苏大学 Method for diagnosing crop water deficit through hyperspectral image technology
CN103528965A (en) * 2013-08-28 2014-01-22 南京农业大学 Wheat leaf equivalent water thickness hyperspectral monitoring method
CN103472009A (en) * 2013-09-16 2013-12-25 南京农业大学 Method for monitoring wheat plant water content under different plant nitrogen content levels
WO2019023712A1 (en) * 2017-07-28 2019-01-31 Cornell University In situ sensing of water potential

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
ZHOU, XIN ET AL.,: ""Visualization research of moisture content in leaf lettuce leaves based on WT-PLSR and hyperspectral imaging technology"", 《JOURNAL OF FOOD PROCESS ENGINEERING 》 *
董霁红 等: "《矿区复垦土壤重金属光谱解析与迁移特征研究》", 31 May 2018 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110907367A (en) * 2019-11-09 2020-03-24 华东师范大学 Soil moisture content detection method based on hyperspectral remote sensing
CN112504977A (en) * 2020-11-17 2021-03-16 华南农业大学 Tea water content detection method and model construction method, medium and equipment thereof

Similar Documents

Publication Publication Date Title
Padarian et al. Using deep learning to predict soil properties from regional spectral data
Feilhauer et al. Multi-method ensemble selection of spectral bands related to leaf biochemistry
Castro‐Esau et al. Variability in leaf optical properties of Mesoamerican trees and the potential for species classification
Herrmann et al. LAI assessment of wheat and potato crops by VENμS and Sentinel-2 bands
CN108875913B (en) Tricholoma matsutake rapid nondestructive testing system and method based on convolutional neural network
CN110376167A (en) Rice leaf nitrogen content monitoring method based on unmanned plane EO-1 hyperion
CN110333195A (en) Water content in plant leaf detection method and device
Li et al. Assessment of unified models for estimating leaf chlorophyll content across directional-hemispherical reflectance and bidirectional reflectance spectra
Tan et al. Analysis of different hyperspectral variables for diagnosing leaf nitrogen accumulation in wheat
CN109754127A (en) Rice grain amylose content estimating and measuring method based on unmanned plane imaging EO-1 hyperion
CN110455722A (en) Rubber tree blade phosphorus content EO-1 hyperion inversion method and system
Lin et al. Predicting copper content in chicory leaves using hyperspectral data with continuous wavelet transforms and partial least squares
Wen et al. Estimation of the vertically integrated leaf nitrogen content in maize using canopy hyperspectral red edge parameters
CN105608296B (en) A kind of blade potassium concn inversion method based on lichee canopy spectra
CN110082310A (en) A kind of near infrared band EO-1 hyperion diagnostic method of rubber tree LTN content
CN110363119A (en) Based on wavelet transformation-random forests algorithm tobacco mildew method for quickly identifying
Wang et al. Derivative hyperspectral vegetation indices in characterizing forest biophysical and biochemical quantities
Wang et al. Assessing canopy nitrogen and carbon content in maize by canopy spectral reflectance and uninformative variable elimination
CN110779875B (en) Method for detecting moisture content of winter wheat ear based on hyperspectral technology
CN111650130A (en) Prediction method and prediction system for magnesium content of litchi leaves
Ren et al. Quantitative identification of yellow rust in winter wheat with a new spectral index: Development and validation using simulated and experimental data
US11054368B2 (en) System for detection of disease in plants
Kothari et al. Plant spectra as integrative measures of plant phenotypes
Wang et al. Prediction of N, P, and K Contents in Sugarcane Leaves by VIS-NIR Spectroscopy and Modeling of NPK Interaction Effects
CN109520964A (en) Seed Vigor Index, system and storage medium based on NIRS

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