CN109916845A - Rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength - Google Patents
Rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength Download PDFInfo
- Publication number
- CN109916845A CN109916845A CN201910046755.8A CN201910046755A CN109916845A CN 109916845 A CN109916845 A CN 109916845A CN 201910046755 A CN201910046755 A CN 201910046755A CN 109916845 A CN109916845 A CN 109916845A
- Authority
- CN
- China
- Prior art keywords
- diagnostic
- rice
- module
- stress intensity
- stress
- 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
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The present invention provides a kind of rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength, it is characterized by comprising blade folder, spectrum sensor, microcomputers, GPS, blade folder, spectrum sensor, GPS are separately connected microcomputer, blade folder is made of two panels spectrum sensor probe, and acquisition module (10), spectroscopic data preprocessing module (20), the characteristic wave bands index value computing module (30), Cd stress intensity diagnostic module (40) of rice canopy near-infrared specific band spectroscopic data are arranged in microcomputer；Pressed from both sides using blade, can the lossless specific band spectroscopic data for easily acquiring rice canopy, using the SVM diagnostic model quick diagnosis of pre-training and differentiation, the final Cd stress intensity of output diagnosis gained.The device is light portable, harmless to crops (rice), only one need to gently press from both sides on rice leaf, can be detected out the intensity that rice is coerced by heavy metal cadmium.
Description
Technical field
The present invention relates to remote sensing fields, examined more particularly, to a kind of rice Cd stress intensity of near-infrared specific wavelength
Disconnected blade folder.
Background technique
In China, by the cultivated area of heavy metal pollution surpassed 10,000,000 hectares (Luo Yongming Teng and answers 2006, Zhang,
Huang et al.2010), more than 1,000 ten thousand tons of grain drop in production (Sun Tiehang, Li Peijun et al.2005) are caused every year.And cadmium is dirty
Dye is the soil heavy metals in farmland pollution category (Chen, Teng et al.2015) of current most serious.How soil is quickly detected
Whether earth is particularly important to timely discovery and subsequent improvement by Cd stress and the intensity of stress.
Currently, the diagnostic method of heavy metals in farmland pollution is mainly chemical analysis detection, for having resulted in the region of pollution
The sampling for carrying out soil sample is collected, and grinds chemically examine in laboratory later, is related to a large amount of soil sample acquisition and chemical examination,
Low efficiency, timeliness is slow, is difficult to use in practical applications.
The present invention can not damage it is considered that remote sensing technology can obtain the canopy information of rice leaf harmlessly
In the case where rice normal growth, the information such as texture, the spectrum of rice leaf are detected, it is possible to for whether detecting rice
By heavy metal cadmium stress and coercive intensity.
Therefore, the present invention utilizes the near-infrared specific wavelength data of rice leaf, the diagnosis leaf provided with Cd stress degree
Piece clamp device quickly can harmlessly identify the intensity that rice is coerced by heavy metal cadmium.
Pertinent literature:
Chen,H.,Y.Teng,S.Lu,Y.Wang and J.Wang(2015)."Contamination features
and health risk of soil heavy metals in China."Sci Total Environ512-513:143-
153.
Li,G.,E.M.Sproviero,R.C.S.Iii,N.Iguchi,J.D.Blakemore,R.H.Crabtree,
G.W.Brudvig and V.S.Batista(2009)."Deposition of an oxomanganese water
oxidation catalyst on TiO2 nanoparticles:computational modeling,assembly and
characterization."Energy&Environmental Science2(2):230-238.
Zhang,X.,C.Huang,B.Liu and Q.Tong(2010).Inversion of soil Cu concentration based on band selection of hyperspetral data.Geoscience and
Remote Sensing Symposium.
Luo Yongming Teng and answers (2006) " China's Status of Soil Pollution Degradation and Preventing Countermeasures "Soil38(5):505-
508.
Sun Tiehang, Li Peijun and Zhou Qixing (2005)Soil pollution formation mechenism and recovery technique, Science Press
Summary of the invention
It is an object of the present invention to realize that the near-infrared specific wavelength based on rice leaf carries out the quick of Cd stress degree
Diagnosis, by light portable, the blade clamp device harmless to rice finds the intensity that rice is coerced by heavy metal cadmium in time,
Reliable technical support is provided for subsequent timely improvement.
The present invention provides a kind of rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength, including leaf
Piece folder, spectrum sensor, microcomputer and GPS, blade folder, spectrum sensor, GPS are separately connected microcomputer, blade
Folder is made of two panels spectrum sensor probe, and rice canopy near-infrared specific band spectroscopic data is arranged in microcomputer
Acquisition module 10, spectroscopic data preprocessing module 20, characteristic wave bands index value computing module 30, Cd stress intensity diagnostic module
40；
Acquisition module 10, for being pressed from both sides using blade, the lossless specific band spectroscopic data for easily acquiring rice canopy is defeated
Enter Pretreated spectra module 20；
Pretreated spectra module 20 carries out water for receiving and processing the multiband light modal data of the transmission of acquisition module 10
The elimination of rice leaf spectra noise and redundancy；
Characteristic wave bands index value computing module 30, the pretreated spectrum for being inputted based on Pretreated spectra module 20
Data calculate diagnostic value of the characteristic wave bands calculated value as index, input Cd stress intensity diagnostic module 40；
Cd stress intensity diagnostic module 40 is diagnosed for the characteristic wave bands calculated value for input using the SVM of pre-training
Model quick diagnosis and differentiation, the final Cd stress intensity of output diagnosis gained.
Moreover, pre-selecting is suitable for the diagnostic factro for diagnosing Cd stress intensity, corresponding rice is acquired using blade folder and is preced with
The specific band spectroscopic data of layer, selection implementation is as follows,
1) determine potential sensitive factor range, including from all more than 2000 near-infrareds in the difference of wave band, by
Pre-stage test Analysis of variance ANOVA2 screens to obtain several potential sensitive factors；
2) potential sensitive factor resulting to step 1), according to the size sequencing selection of the feature confidence level of random forest RF
Several preceding characterization factors；
3) based on the diagnostic accuracy of SVM, a characterization factor for selecting comprehensive diagnos precision optimal is as determining diagnosis
The factor determines the diagnostic factro obtained by wave band A and B difference；
4) the wider wave-length coverage near wave band A and B is investigated, can reach satisfied in the difference of two wave band reflectivity and examine
Under conditions of disconnected precision, it is determined as the specific band of final diagnostic factro.
Moreover, selecting two wave-length coverages is respectively the specific band of 754nm-771nm, 772nm-789nm as final
Diagnostic factro.
Moreover, the processing that Pretreated spectra module 20 carries out is to ask first derivative, Savitzky-Golay smoothing processing, mark
The two or above combination in standardization.
Moreover, the diagnostic value of 30 parameter of characteristic wave bands index value computing module, the spectrum including calculating two wave bands
The difference of reflectivity obtains characteristic wave bands calculated value divided by 2, as the spectral value for final diagnosis.
Moreover, the diagnostic result of SVM diagnostic model includes no Cd stress, low concentration Cd in Cd stress intensity diagnostic module 40
Stress, middle concentration C d stress, high concentrations of Cd coerce four classifications, and SVM diagnostic model is calculated based on the characteristic wave bands of rice canopy
Value, obtain rice whether the stress by any of them, diagnostic categories highest for confidence level, as final diagnosis
Stress classification out.
The invention discloses a kind of diagnosis blade of soil heavy metal cadmium coercive intensity folders, and device is light portable, to farming
Object (rice) is harmless, only one need to gently press from both sides on rice leaf, can be detected out the intensity that rice is coerced by heavy metal cadmium.This
Blade folder hardware device can be realized effectively: 1, according to the spectroscopic data of the near-infrared specific band of crops (rice) blade,
Differentiate to field real-time on-site crops (rice) whether by Cd stress；2, to crops (rice) whether by Cd stress
Carry out quickly identification early warning；3, the substantially strength range for judging Cd stress provides preliminary foundation to subsequent improvement.The present apparatus is
It is a kind of small and exquisite easy-to-use, simple portable Cd stress diagnostic device, can in the case where lossless to crops (rice) blade,
It is efficiently diagnosed to be the coercive intensity of Cd stress, is a kind of crops (rice) soil huge sum of money based on near-infrared specific band
Belong to the diagnosis blade clamp device of Cd stress.
Detailed description of the invention
Fig. 1 is the function structure chart of the embodiment of the present invention；
Fig. 2 is that the blade of the embodiment of the present invention presss from both sides hardware device figure；
Fig. 3 is that the rice canopy spectra collection of the embodiment of the present invention is intended to；
Fig. 4 is the rice canopy near-infrared specific band Pretreated spectra schematic diagram of the embodiment of the present invention；
Fig. 5 is the rice canopy spectrum intelligence computation schematic diagram of the embodiment of the present invention
Fig. 6 is the diagnosis schematic diagram of the Cd stress intensity of the embodiment of the present invention.
Specific embodiment
The inventive technique scheme is described in detail below in conjunction with drawings and examples.
The present invention proposes the diagnosis leaf that heavy metal cadmium coercive intensity is carried out by rice canopy near-infrared specific wavelength data
Piece folder, realizes the fast slowdown monitoring of growth period rice Cd stress intensity.The device first connects sensor and backstage microcomputer
It is connected together, obtains rice canopy blade spectroscopic data using specific band reflectivity sensor, deposited using microcomputer
Store up and record data acquisition time, place etc.；Then automation pretreatment is carried out to spectrum, comprising: pick including exceptional spectrum
It removes, spectrum first differential, Savitzky-Golay are smooth, standardization.Combination calculating is automated to pretreated spectral value to examine
Disconnected value simultaneously inputs last Cd stress intensity diagnostic module, realizes the quick identification of Cd stress intensity.
Such as Fig. 1, embodiment provides the internal mode of the blade clamp device of rice canopy spectroscopic data quick diagnosis Cd stress
Block schematic diagram, acquisition module 10, spectroscopic data preprocessing module including rice canopy near-infrared specific band spectroscopic data
20, characteristic wave bands index value computing module 30, Cd stress intensity diagnostic module 40.
In embodiment, the construction of entire instrument as shown in Fig. 2, include blade folder, spectrum sensor, microcomputer and
GPS, blade folder, spectrum sensor, GPS are separately connected microcomputer, and the display of display additionally can be set and remain
Remaining electric-quantity display device, power supply power switch and charging plug.When it is implemented, can be by spectrum sensor, miniature calculating
Machine, GPS etc. are integrally disposed inside housings, and blade sandwiched, which is set, to be connected in hull outside with spectrum sensor.It can also be in shell
The handle conveniently taken is set on body.
Blade folder is located at upper-right position in embodiment, is made of the spectrum sensor probe of upper and lower two panels thumbnail size, just
In normal situation in opened condition, data only need to slightly exert oneself when acquisition, can be closed and gently clamp blade acquisition data, number
It is stored according to microcomputer after spectrum sensor obtains, is passed to；The right hand that handle is resorted to ordinary person's habit is had an effect mode,
Instrument comfortably is caught, and index finger and thumb is facilitated to control having an effect for blade folder.
In embodiment, microcomputer can be used miniature personal PC and realize, the CPU including central processing unit, general interior
It, can be with the rice canopy high-spectral data of real-time storage sensor measurement equipped with storage unit.GPS, display, spectrum sensing
Device is all connect with microcomputer, specifically, be can be considered and is connect with CPU.The GPS position information and spectral information of acquisition can be real
When show over the display, display can be always on state in spectra collection information after storage, and data enter through microcomputer
Backstage is pre-processed, characteristic wave bands calculate, diagnosis, and after the completion of diagnostic result, through Microcomputer control, result is recorded
And it shows over the display.Diagnostic result is shown always over the display, until data acquisition next time starts.
Microcomputer can realize adopting for rice canopy near-infrared specific band spectroscopic data using software modularity mode
Collect module 10, spectroscopic data preprocessing module 20, characteristic wave bands index value computing module 30, Cd stress intensity diagnostic module 40.
10 the data obtained of acquisition module is transferred to spectroscopic data preprocessing module 20, is transmitted to characteristic wave bands index value meter after processing again
Module 30 is calculated, is most transmitted to CPU through the diagnosis of Cd stress intensity diagnostic module 40 afterwards, and be shown on display.In order to facilitate number
It transmits and carries out according to the automation of acquisition and processing, above-mentioned modules are all connect with CPU.The use of internal module can be below
It is described in detail.
Illustrate that each module is realized, individually below for implementation reference.
The acquisition module 10 of rice canopy near-infrared specific band spectroscopic data is used for the acquisition of rice canopy spectroscopic data,
It is stored data into the storage unit set in microcomputer by forms such as data transmission, it is pre- as subsequent spectroscopic data
The input data of processing module 20.
It, can the lossless specific band spectroscopic data for easily acquiring rice canopy using blade folder and spectrum sensor.
Before starting measurement, the environmental parameter (solar radiation, atmospheric vapour content, local time, weather conditions etc.) to the same day have
One preliminary record, such as table 1, warms up machinery later, reaches spectroscopic probe head and instrument operating status most preferably, guarantees spectrum number
According to acquisition phase electricity, sufficient, instrument operating temperature is suitable for, and while preheating, can set the acquisition ginseng of blade folder hardware
Number, the pre- Warm status of instrument and engineering parameter setting table are as shown in table 2, and the record form for acquiring data is as shown in table 3.
1 environment parameter recording table of table
| Date | Time | ||
| Solar radiation | Temperature | ||
| Atmospheric vapour content | Weather conditions |
Record sheet is arranged in 2 engineering parameter of table
| Time started | Preheating time | ||
| Ambient temperature | Instrument temperature | ||
| Time for exposure | Single acquisition number | ||
| Storage location | Storage format | ||
| Number form | Number opening code | ||
| Remaining capacity |
3 canopy spectra data record sheet of table
| Collection point number | Spectral value | ||
| Acquisition time | Longitude | ||
| Latitude |
In embodiment, local time 10:00-14:00 of ceiling unlimited is selected, is preheated instrument 3-5 minutes, in field rice
Tanaka gently clamps blade 3s-5s, and spectrum sensor the data obtained passes to acquisition module 10, and 10 data of acquisition module have acquired
At by collected spectroscopic data input spectrum data preprocessing module.
What acquisition module 10 acquired is the spectral value of 2 special wave bands respectively obtained by the probe of 2 blades.Spectrum
Wave band has the difference (the latter subtracts the former) of more than 2000 neighbouring wave bands, to improve efficiency, can be embodied by as diagnostic factro
When can therefrom find the potential sensitive factor of suitable Cd diagnosis in advance, then selection is suitable for examining inside potential sensitive factor again
The disconnected factor.In embodiment, determine the specific implementation of the last diagnostic factor the following steps are included:
1) potential sensitive factor range is determined, including the difference from all more than 2000 visual and near-infrareds adjacent to wave band
(the so-called difference for closing on wave band has the value of the latter's reflectivity to subtract obtained by the former that is, for 2 wave bands of arbitrary neighborhood)
In, it is obtained by pre-stage test Analysis of variance (Two-way analysis of variance, ANOVA2) screening several potential
Sensitive factor；The stress group of zero low middle high four concentration based on the setting of Cd stress tests, ANOVA2 is by difference in reflectivity
The summation of the sum of square of deviations of mean value and overall average between different groups indicates that result (significance of difference) is to sentence
Not waveband difference value foundation whether sensitive to Cd stress.According to the conspicuousness (setting of difference between different Cd coercive intensities
0.05), i.e., effective to distinguish difference in reflectivity caused by Cd coercive intensity changes, it obtains coercing sensitivity to single heavy metal Cd
Potential characterization factor.
2) potential sensitive factor resulting to step 1), according to the feature confidence of random forest (Random Forest, RF)
10 characterization factors before the size sequencing selection of degree (Z-Value).
3) based on the diagnostic accuracy of SVM, a characterization factor for selecting comprehensive diagnos precision optimal is as determining diagnosis
The factor determines the diagnostic factro obtained by wave band A and B difference.In embodiment, the factor is by wave band 772nm and 771nm difference
Gained.
Therefore two specific bands selected in embodiment are 771nm and 772nm.
4) consider from manufacturing cost, due to the relatively narrow sensor higher cost of spectral bandwidth, investigate in 771nm and
Wider wave-length coverage near 772nm, when range 754-771nm, 772-789nm, the difference using two wave band reflectivity still may be used
Reach satisfied diagnostic accuracy, therefore the last diagnostic factor of Cd stress diagnosis, i.e., two wavelength selected in embodiment have been determined
Range is respectively the specific band of 754nm-771nm, 772nm-789nm.
Spectroscopic data preprocessing module 20, for carrying out rice leaf spectrum and making an uproar to 10 gained spectroscopic data of acquisition module
The elimination of sound and redundancy improves the reliability of data acquisition, to obtain the diagnostic accuracy of higher Cd stress intensity.
It may include first derivative (1st-Derivative) when specific implementation, standardize (Std), Savitzky-Golay smoothly (Sg)
In two kinds or more the combinations handled.
Characteristic wave bands index value computing module 30, the pretreated spectrum for being inputted based on Pretreated spectra module 20
Data, distribution calculate memory, calculate diagnostic value of the characteristic wave bands calculated value as index, input Cd stress intensity diagnostic module
40。
The calculating of characteristic wave bands derives from specific band spectrum proposed by the present invention, and specific wave band and calculation method are such as
Under, based on the pretreated spectroscopic data that the transmission of spectroscopic data preprocessing module 20 is come in, the diagnosis of index is calculated
Value, inputs next diagnostic module；
Diagnose spectral value calculation formula used are as follows: (R772-789nm-R754-771nm)/2
That is: the difference of the spectral reflectivity of two wave bands (wave-length coverage 754nm-771nm, 772nm-789nm) is calculated, most
The spectral value for final diagnosis is obtained divided by 2 afterwards.
When it is implemented, the computing function of microcomputer can be relied on, according to data calculation amount, (data volume assessment divides
With storage, storage cell etc.), such as table 4, transmit after the data of Pretreated spectra module 20 are calculated into diagnostic module.
4 calculating parameter record form of table
| Data volume | Storage cell | ||
| The total memory of computer | Committed memory | ||
| Running memory distribution |
Cd stress intensity diagnostic module 40 is diagnosed for the characteristic wave bands calculated value for input using the SVM of pre-training
Model quick diagnosis and differentiation, the final Cd stress intensity of output diagnosis gained.
Based on the spectral value used of diagnosis obtained by characteristic wave bands index computing module 30, it is passed to internal support vector machines
(SVM) whether model classifiers are quickly determined by heavy metal cadmium coercive intensity that rice, diagnostic accuracy is highest will
Intensity as final rice by Cd stress.
Diagnostic module includes built-in model call unit and diagnostic result display unit.
Built-in model call unit: it is adjusted to suitable by parameter-embedded according to the SVM diagnostic model parameter of pre-training and examines
The design parameter setting of disconnected parameter, SVM diagnostic model shares four groups 12, altogether includes that 12 groups of complete SVM are examined in built-in unit
The parameter, including kernel function classification, deviation etc. of disconnected model, svmModel1001, svmModel2001, svmModel3001
It is the model whether Diagnosis Rice is coerced by any concentration Cd, svmModel1002, svmModel2002, svmModel3002
It is to diagnose the model whether coerced by low concentration Cd, svmModel1003, svmModel2003, svmModel3003 are to examine
The disconnected model whether coerced by middle concentration C d, svmModel1004, svmModel2004, svmModel3004 are that diagnosis is
The no model coerced by high concentrations of Cd.
In embodiment, entire model training and classification are operated on Matlab 2015a, the SVM for selecting platform included
Training, it is contemplated that number of samples is limited, and the SVM of platform stays to the form assessment models precision of a verifying instead, i.e., N number of sample,
Each sample is used as validation data set one by one, and as training set, training, final precision on platform take N number of discrimination model to others
Average value.In SVM training pattern, diagnosis for zero/low/medium/high concentration, kernel function KernelFunction is kept not
Become, select linear kernel function linear kernel, other can have difference with the variation of training label: group
GroupsName is training label, and changing unit includes supporting vector (SupportVectors), Lagrangian vector
(Alpha), hyperplane intercept (Bias), vector index (SupportVectorIndices), data scale (ScaleData),
And ScaleData includes row vector mean value (shift) and the inverse of shift --- scale factor (scaleFactor)
Two factors.Specific implementation can be found in Matlab 2015a and accordingly encapsulate, and it will not go into details by the present invention.
For the model of four intensity stress (each including 3 svmModel), the 1st is diagnosed when " -001 " is training pattern
The diagnostic model of a sample, diagnoses the diagnostic model of the last one sample when " -003 " is training pattern, " -002 " is trained mould
The diagnostic model of middle trained sample (removing first and last 2) is diagnosed when type.In specific implementation, the training that training pattern is obtained is joined
Number is built into blade clamp device, and a similar SVM multi-categorizer collects the sample in actual acquisition as verifying
Coercive intensity verifying is carried out, obtains its " classification " --- coercive intensity.
Diagnostic result unit, including four classifications: no Cd stress, low concentration Cd stress, middle concentration C d stress, high concentration
Cd stress.Diagnostic model can based on the peculiar spectrum of rice canopy, obtain rice whether the stress by any of them, it is right
In the highest diagnostic categories of confidence level, the as stress classification that is diagnosed to be of last blade clamp device as a result will be aobvious with confidence level
Show screen display.
For the sake of ease of implementation, user is provided using rice Cd stress intensity real-time diagnosis provided by the present invention
It is as follows that blade presss from both sides hardware device process:
Step 1, rice canopy spectrum data gathering step, including according to parameter set by user, it is close to obtain rice canopy
The spectroscopic data of infrared specific band；
Step 2, spectroscopic data pre-treatment step, including first derivative, standardization, Savitzky-Golay it is smooth and they
Combination, obtain not spectroscopic data after by the pretreatment of such environmental effects；
Step 3, characteristic wave bands index value calculates step, and microcomputer will be according to the parameter set before starting to acquire
(table 2), intelligence distribution calculate space, automate parameter value and are passed to next module；
Step 4, Cd stress intensity diagnostic step, which is also that the automation of microcomputer differentiates, using embedded
SVM diagnostic model assesses the confidence level of zero-dose, low concentration, middle concentration, high concentration Cd stress respectively, by miniature
The comparison of computer embedded program, the highest coercive intensity confidence level that final blade folder differentiates is rice reality by Cd stress
Intensity, show the intensity and corresponding confidence level over the display.
Software technology can be used when specific implementation and realize automatic running process.
Specific embodiment described herein is only an example for the spirit of the invention.The neck of technology belonging to the present invention
The technical staff in domain can make various modifications or additions to the described embodiments or replace by a similar method
In generation, however, it does not deviate from the spirit of the invention or beyond the scope of the appended claims.
Claims (6)
1. a kind of rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength, it is characterised in that: including leaf
Piece folder, spectrum sensor, microcomputer and GPS, blade folder, spectrum sensor, GPS are separately connected microcomputer, blade
Folder is made of two panels spectrum sensor probe, and adopting for rice canopy near-infrared specific band spectroscopic data is arranged in microcomputer
Collect module (10), spectroscopic data preprocessing module (20), characteristic wave bands index value computing module (30), Cd stress intensity diagnostic mould
Block (40)；
Acquisition module (10), for being pressed from both sides using blade, the lossless specific band spectroscopic data for easily acquiring rice canopy, input
Pretreated spectra module (20)；
Pretreated spectra module (20) carries out water for receiving and processing the multiband light modal data of acquisition module (10) transmission
The elimination of rice leaf spectra noise and redundancy；
Characteristic wave bands index value computing module (30), for the pretreated spectrum based on Pretreated spectra module (20) input
Data calculate diagnostic value of the characteristic wave bands calculated value as index, input Cd stress intensity diagnostic module (40)；
Cd stress intensity diagnostic module (40) diagnoses mould using the SVM of pre-training for the characteristic wave bands calculated value for input
Type quick diagnosis and differentiation, the final Cd stress intensity of output diagnosis gained.
2. the rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength according to claim 1, special
Sign is: pre-selecting is suitable for the diagnostic factro for diagnosing Cd stress intensity, and the spy of corresponding rice canopy is acquired using blade folder
Standing wave section spectroscopic data, selection implementation is as follows,
1) determine potential sensitive factor range, including from all more than 2000 near-infrareds in the difference of wave band, by early period
Test Analysis of variance ANOVA2 screens to obtain several potential sensitive factors；
2) potential sensitive factor resulting to step 1), if before according to the size sequencing selection of the feature confidence level of random forest RF
Dry characterization factor；
3) based on the diagnostic accuracy of SVM, diagnostic factro of the characterization factor for selecting comprehensive diagnos precision optimal as determination,
Determine the diagnostic factro obtained by wave band A and B difference；
4) the wider wave-length coverage near wave band A and B is investigated, satisfied diagnosis essence can be reached in the difference of two wave band reflectivity
Under conditions of degree, it is determined as the specific band of final diagnostic factro.
3. the rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength according to claim 2, special
Sign is: selecting two wave-length coverages is respectively the specific band of 754nm-771nm, 772nm-789nm as final diagnosis
The factor.
4. the rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength according to claim 1, special
Sign is: the processing that Pretreated spectra module (20) carries out is to ask first derivative, Savitzky-Golay smoothing processing, standardization
Both middle or above combination.
5. the rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength according to claim 1, special
Sign is: the diagnostic value of characteristic wave bands index value computing module (30) parameter, the spectral reflectance including calculating two wave bands
The difference of rate obtains characteristic wave bands calculated value divided by 2, as the spectral value for final diagnosis.
6. the according to claim 1 or 2 or 3 or 4 or 5 diagnosis blades based on near-infrared specific wavelength rice Cd stress intensity
Clamp device, it is characterised in that: in Cd stress intensity diagnostic module (40) diagnostic result of SVM diagnostic model include no Cd stress,
Low concentration Cd stress, middle concentration C d stress, high concentrations of Cd coerce four classifications, feature of the SVM diagnostic model based on rice canopy
Wave band calculated value, obtain rice whether the stress by any of them, diagnostic categories highest for confidence level, as most
The stress classification being diagnosed to be afterwards.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910046755.8A CN109916845A (en) | 2019-01-18 | 2019-01-18 | Rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910046755.8A CN109916845A (en) | 2019-01-18 | 2019-01-18 | Rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109916845A true CN109916845A (en) | 2019-06-21 |
Family
ID=66960382
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910046755.8A Pending CN109916845A (en) | 2019-01-18 | 2019-01-18 | Rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109916845A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101750281A (en) * | 2008-12-05 | 2010-06-23 | 中国科学院沈阳应用生态研究所 | Rice Fe2+ stress spectrum quantification identification method |
| CN102830072A (en) * | 2012-08-13 | 2012-12-19 | 中国计量学院 | Identification method for rice leaves contaminated by heavy metals based on near infrared spectroscopy |
| CN104089926A (en) * | 2014-07-18 | 2014-10-08 | 湖南省食品测试分析中心 | NIR (Near Infrared Ray) spectral analysis model and method for identifying excessive content of cadmium in rice |
| CN104111264A (en) * | 2014-08-01 | 2014-10-22 | 广东省粮食科学研究所 | Methods for quickly detecting content of heavy metal elements in rice and evaluating comprehensive pollution index of heavy metal elements |
| CN108519339A (en) * | 2018-03-26 | 2018-09-11 | 江苏大学 | A kind of blade cadmium content Vis-NIR spectral signature modeling methods based on WT-LSSVR |
-
2019
- 2019-01-18 CN CN201910046755.8A patent/CN109916845A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101750281A (en) * | 2008-12-05 | 2010-06-23 | 中国科学院沈阳应用生态研究所 | Rice Fe2+ stress spectrum quantification identification method |
| CN102830072A (en) * | 2012-08-13 | 2012-12-19 | 中国计量学院 | Identification method for rice leaves contaminated by heavy metals based on near infrared spectroscopy |
| CN104089926A (en) * | 2014-07-18 | 2014-10-08 | 湖南省食品测试分析中心 | NIR (Near Infrared Ray) spectral analysis model and method for identifying excessive content of cadmium in rice |
| CN104111264A (en) * | 2014-08-01 | 2014-10-22 | 广东省粮食科学研究所 | Methods for quickly detecting content of heavy metal elements in rice and evaluating comprehensive pollution index of heavy metal elements |
| CN108519339A (en) * | 2018-03-26 | 2018-09-11 | 江苏大学 | A kind of blade cadmium content Vis-NIR spectral signature modeling methods based on WT-LSSVR |
Non-Patent Citations (17)
| Title |
|---|
| 修丽娜 等: "镉污染水稻高光谱诊断分析与建模", 《光谱学与光谱分析》 * |
| 修丽娜 等: "镉污染水稻高光谱诊断分析与建模", 《光谱学与光谱分析》, vol. 31, no. 1, 31 January 2011 (2011-01-31), pages 192 - 196 * |
| 吕杰 等: "利用支持向量机构建水稻镉含量高光谱预测模型", 《利用支持向量机构建水稻镉含量高光谱预测模型》 * |
| 吕杰 等: "利用支持向量机构建水稻镉含量高光谱预测模型", 《利用支持向量机构建水稻镉含量高光谱预测模型》, vol. 30, no. 1, 31 January 2012 (2012-01-31), pages 105 - 110 * |
| 安欣 等: "复合支持向量机方法及其在光谱分析中的应用", 《光谱学与光谱分析》 * |
| 安欣 等: "复合支持向量机方法及其在光谱分析中的应用", 《光谱学与光谱分析》, vol. 27, no. 8, 31 August 2007 (2007-08-31), pages 1619 - 1621 * |
| 张双印 等: "基于成像高光谱数据的温室水稻重金属胁迫诊断研究", 《安徽农业科学》 * |
| 张双印 等: "基于成像高光谱数据的温室水稻重金属胁迫诊断研究", 《安徽农业科学》, vol. 46, no. 1, 31 January 2018 (2018-01-31), pages 6 - 9 * |
| 张龙 等: "基于近红外光谱的重金属汞、镉和铅污染水稻叶片鉴别", 《浙江大学学报(农业与生命科学版)》 * |
| 张龙 等: "基于近红外光谱的重金属汞、镉和铅污染水稻叶片鉴别", 《浙江大学学报(农业与生命科学版)》, vol. 39, no. 1, 31 January 2013 (2013-01-31), pages 50 - 55 * |
| 赵思颖 等: "土壤重金属污染对水稻高光谱的影响", 《江苏农业科学》 * |
| 赵思颖 等: "土壤重金属污染对水稻高光谱的影响", 《江苏农业科学》, vol. 44, no. 11, 30 November 2016 (2016-11-30), pages 423 - 426 * |
| 金铭 等: "基于冠层多维光谱的水稻镉污染胁迫诊断模型研究", 《中国环境科学》 * |
| 金铭 等: "基于冠层多维光谱的水稻镉污染胁迫诊断模型研究", 《中国环境科学》, vol. 31, no. 1, 31 January 2011 (2011-01-31), pages 137 - 143 * |
| 马淏: "光谱及高光谱成像技术在作物特征信息提取中的应用研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
| 马淏: "光谱及高光谱成像技术在作物特征信息提取中的应用研究", 《中国博士学位论文全文数据库 信息科技辑》, no. 7, 15 July 2015 (2015-07-15), pages 9 - 38 * |
| 黄继红: "《抗性淀粉生产技术及其应用》", 31 January 2017, 河南科学技术出版社, pages: 234 - 236 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Bock et al. | From visual estimates to fully automated sensor-based measurements of plant disease severity: status and challenges for improving accuracy | |
| Tan et al. | Estimating the distribution trend of soil heavy metals in mining area from HyMap airborne hyperspectral imagery based on ensemble learning | |
| CN102788752B (en) | Non-destructive detection device and method of internal information of crops based on spectrum technology | |
| Li et al. | Assessment of unified models for estimating leaf chlorophyll content across directional-hemispherical reflectance and bidirectional reflectance spectra | |
| CN102621118B (en) | Early warning method of greenhouse vegetable diseases and insect pests | |
| CN110717368A (en) | Qualitative classification method for textiles | |
| US11221320B2 (en) | Method and apparatus for measuring leaf nitrogen content | |
| CN105158186B (en) | A kind of method detected based on high spectrum image to ternip evil mind | |
| CN101971006A (en) | Method and apparatus of evaluating fitness-for-plucking of tea leaf, system of evaluating fitness-for-plucking of tea leaf, and computer-usable medium | |
| CN107392920B (en) | Plant health distinguishing method and device based on visible light-terahertz light | |
| CN109669023A (en) | A kind of soil attribute prediction technique based on Multi-sensor Fusion | |
| CN107045637A (en) | A kind of blood species identification instrument and recognition methods based on spectrum | |
| CN107679569A (en) | Raman spectrum substance automatic identifying method based on adaptive hypergraph algorithm | |
| CN102841063B (en) | Method for tracing and identifying charcoal based on spectrum technology | |
| CN107748144A (en) | The middle infrared spectrum detecting system of quick measure SOIL CARBON AND NITROGEN and its stable isotope | |
| Huang et al. | Detection of scab in wheat ears using in situ hyperspectral data and support vector machine optimized by genetic algorithm | |
| CN105866050A (en) | Low-cost lossless and fast detecting equipment for apple moldy core | |
| CN103278467A (en) | Rapid nondestructive high-accuracy method with for identifying abundance degree of nitrogen element in plant leaf | |
| CN115905881A (en) | Method and device for classifying yellow pearls, electronic equipment and storage medium | |
| CN205719951U (en) | Detect black box and utilize the lossless quick detection equipment of Apple Mould Core of this detection black box | |
| CN111965167A (en) | Method and device for predicting element composition and calorific value of solid waste | |
| CN109916845A (en) | Rice Cd stress intensity diagnostic blade clamp device based on near-infrared specific wavelength | |
| CN117491293A (en) | High-steep bank slope carbonate rock corrosion rapid evaluation method based on hyperspectrum | |
| CN109668842A (en) | A kind of shellfish heavy metal pollution rapid detection method | |
| CN116310788A (en) | Plant leaf nitrogen content hyperspectral modeling method based on map fusion |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190621 |
|
| RJ01 | Rejection of invention patent application after publication |