CN105675534A - Method for quickly and nondestructively identifying polished grains - Google Patents
Method for quickly and nondestructively identifying polished grains Download PDFInfo
- Publication number
- CN105675534A CN105675534A CN201610180371.1A CN201610180371A CN105675534A CN 105675534 A CN105675534 A CN 105675534A CN 201610180371 A CN201610180371 A CN 201610180371A CN 105675534 A CN105675534 A CN 105675534A
- Authority
- CN
- China
- Prior art keywords
- sample
- polishing
- training set
- characteristic vector
- infrared absorption
- 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
- 238000000034 method Methods 0.000 title claims abstract description 53
- 238000012549 training Methods 0.000 claims abstract description 45
- 239000013598 vector Substances 0.000 claims abstract description 43
- 238000000862 absorption spectrum Methods 0.000 claims abstract description 34
- 239000002480 mineral oil Substances 0.000 claims abstract description 31
- 235000010446 mineral oil Nutrition 0.000 claims abstract description 31
- 238000001514 detection method Methods 0.000 claims abstract description 18
- 238000005498 polishing Methods 0.000 claims description 69
- 238000012360 testing method Methods 0.000 claims description 37
- 235000013339 cereals Nutrition 0.000 claims description 29
- 241000209094 Oryza Species 0.000 claims description 26
- 235000007164 Oryza sativa Nutrition 0.000 claims description 26
- 235000009566 rice Nutrition 0.000 claims description 26
- 238000004611 spectroscopical analysis Methods 0.000 claims description 19
- 230000008569 process Effects 0.000 claims description 11
- 230000001066 destructive effect Effects 0.000 claims description 10
- 239000011159 matrix material Substances 0.000 claims description 8
- 239000000284 extract Substances 0.000 claims description 6
- 238000002372 labelling Methods 0.000 claims description 5
- 238000004519 manufacturing process Methods 0.000 claims description 3
- 240000008042 Zea mays Species 0.000 claims description 2
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 claims description 2
- 235000002017 Zea mays subsp mays Nutrition 0.000 claims description 2
- 238000013528 artificial neural network Methods 0.000 claims description 2
- 235000005822 corn Nutrition 0.000 claims description 2
- 238000012937 correction Methods 0.000 claims description 2
- 238000012850 discrimination method Methods 0.000 claims description 2
- 238000009499 grossing Methods 0.000 claims description 2
- 238000012417 linear regression Methods 0.000 claims description 2
- 229940057995 liquid paraffin Drugs 0.000 claims description 2
- 238000012628 principal component regression Methods 0.000 claims description 2
- 238000005070 sampling Methods 0.000 claims description 2
- 238000001228 spectrum Methods 0.000 abstract description 6
- 238000010586 diagram Methods 0.000 abstract description 4
- 230000008901 benefit Effects 0.000 abstract description 2
- 238000001845 vibrational spectrum Methods 0.000 description 15
- 238000011156 evaluation Methods 0.000 description 9
- 238000002329 infrared spectrum Methods 0.000 description 5
- 238000011161 development Methods 0.000 description 4
- 238000012271 agricultural production Methods 0.000 description 3
- 230000000052 comparative effect Effects 0.000 description 3
- 238000009776 industrial production Methods 0.000 description 3
- 239000000463 material Substances 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 238000012827 research and development Methods 0.000 description 3
- 238000010521 absorption reaction Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 235000016709 nutrition Nutrition 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- 239000004215 Carbon black (E152) Substances 0.000 description 1
- 238000001237 Raman spectrum Methods 0.000 description 1
- 150000001338 aliphatic hydrocarbons Chemical class 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 150000004945 aromatic hydrocarbons Chemical class 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 239000003153 chemical reaction reagent Substances 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 239000000796 flavoring agent Substances 0.000 description 1
- 235000019634 flavors Nutrition 0.000 description 1
- 238000002795 fluorescence method Methods 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 229930195733 hydrocarbon Natural products 0.000 description 1
- 150000002430 hydrocarbons Chemical class 0.000 description 1
- 239000002398 materia medica Substances 0.000 description 1
- 238000004476 mid-IR spectroscopy Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000035764 nutrition Effects 0.000 description 1
- 239000003348 petrochemical agent Substances 0.000 description 1
- 238000005120 petroleum cracking Methods 0.000 description 1
- 230000035479 physiological effects, processes and functions Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000005610 quantum mechanics Effects 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/3563—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for analysing solids; Preparation of samples therefor
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/359—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light using near infrared light
Abstract
The invention provides a method for quickly and nondestructively identifying polished grains. The method includes steps of constructing training sets by the aid of grain samples under known polished conditions, acquiring near-infrared absorption spectra in characteristic wave bands of various samples in the training sets, respectively assigning different numerical values to the polished samples and the unpolished samples in the training sets, then extracting two groups of characteristic vectors, solving corresponding two groups of score values, drawing two-dimensional score scatter diagrams of the samples in the training sets by the aid of the score values used as horizontal and vertical coordinates and dividing the two-dimensional score scatter diagrams into polished regions and unpolished regions; scanning spectra of to-be-detected samples, solving two groups of score values of the to-be-detected samples, drawing corresponding to-be-detected points in the scatter diagrams and identifying the to-be-detected samples according to the regions where the to-be-detected points are located so as to determine whether the to-be-detected samples are the polished grains or not. The method has the advantages that the grains polished by the aid of mineral oil can be quickly and nondestructively detected, the method is easy to implement, free of pollution, intelligentization and informatization can be facilitated, and important technical means and support can be provided for guaranteeing the quality safety of agricultural products, researching and developing quick and nondestructive detection instruments and the like.
Description
Technical field
The present invention relates to agricultural product quality and safety field, specifically, relate to a kind of mineral oil based on vibrational spectrum and polish grain fast non-destructive detection method.
Background technology
Agricultural product are the food important sources that the mankind depend on for existence up to now. Along with development and the social progress of the productivity, the concern of agricultural product quality and safety is improved day by day. Agricultural product quality and safety comprises the implication of two aspects, first quality, and namely whether the content of nutritional labeling is up to standard, whether local flavor is good; It two is safety, namely whether there is harm or the possible factor of the physics of potential hazard human health, chemistry, biology and other each side. The second best in quality agricultural product can provide the energy of high-quality, source of nutrition to the mankind, to ensure human survival, multiply significant. But in recent years, agricultural product inferior are adulterated by illegal retailer, the behaviors such as false making of faking are of common occurrence. For rice, have been reported the illegal retailer of title and polish outmoded rice with mineral oil so that it is " glossy reflect ", to pretend to be high quality white rice to be sold to consumer. Being there is this great potential safety hazard by edible above-mentioned mineral oil polishing rice in people, even directly harm consumer is healthy. Further, rice is the large agricultural product of China, has huge consumer groups and consumption market. Therefore, the Fast nondestructive evaluation for mineral oil polishing rice becomes problem demanding prompt solution.
Mineral oil is petrochemicals, based on aliphatic hydrocarbon. Owing to being wherein practically free of aromatic hydrocarbon, traditional fluorescence method is therefore adopted to be difficult to it is detected. Traditional sensory evaluation method is adopted then to have stronger subjective factors, it is easy to form erroneous judgement. Although Modern Large-scale Instruments method can obtain testing result accurately, but complex operation, the professional of testing staff is required that high, detection process needs to use chemical reagent, detection cycle length, detection efficiency is low, be difficult in adapt to the quick detection demand of large agricultural product.
Vibrational spectrum is a kind of spectrum that electromagnetic wave produces absorption based on the vibration of material molecule, rotational energy level transition. According to principle of quantum mechanics, when electromagnetic energy and the vibration of material molecule, rotational energy level transition energy difference are equal, and when meeting vibrational spectrum occurrence condition, vibrational spectrum can be produced and absorb. Common vibrational spectrum has: near infrared spectrum, middle infrared spectrum, Raman spectrum, Terahertz wave spectrum etc. Vibrational spectrum technology has the features such as sample rate is fast, analysis efficiency is high, green non-pollution, is widely applied at numerous areas such as life sciences, medical science and physiology, materia medica, agronomy, chemistry at present.
The present invention proposes a kind of mineral oil based on vibrational spectrum and polishes grain fast non-destructive detection method.The method has the features such as quick, accurate, efficient, green non-pollution, mineral oil can be polished grain and carry out quick nondestructive discriminating, and the aspects such as guarantee agricultural product quality and safety, the research and development improving industrial and agricultural production efficiency, promotion market for farm products sound development, Fast nondestructive evaluation instrument are had positive role.
Summary of the invention
The present invention is directed to the quick nondestructive of mineral oil polishing grain and differentiate a difficult problem, it is provided that a kind of mineral oil based on vibrational spectrum polishes grain fast non-destructive detection method, its objective is that mineral oil is polished grain carries out Fast nondestructive evaluation and discriminating.
A kind of fast non-destructive detection method polishing grain provided by the invention, the flow process of described method is referred to shown in Fig. 1.
Specifically, said method comprising the steps of:
(1) with known polishing grain with not polish grain be sample, build sample and always organize the training set that number is n; In described training set, n >=50, the ratio of polishing sample and the group number not polishing sample is 1～3.5:1;
(2) gather in described training set and often organize sample at 8900cm-1～12000cm-1Near-infrared absorption spectrum in wave-length coverage, often organizes the near-infrared absorption spectrum of sample with 1cm-1～3cm-1M spectroscopic data is included for interval;
(3) with the near-infrared absorption spectrum data construct n × m rank matrix of samples whole in described training set, to the polishing sample in described training set with do not polish sample and give different numerical value respectively, extract some stack features vector of described matrix according to described numerical value regression algorithm and obtain and described characteristic vector eigenvalue one to one, vectorial as the 1st characteristic vector and the 2nd characteristic vector using two stack features that eigenvalue is the highest;
(4) according to the near-infrared absorption spectrum data often organizing sample in training set, obtain weighted value corresponding with described 1st characteristic vector and the 2nd characteristic vector respectively, score value 1 and score value 2 as each training set sample, and with its respectively horizontal, vertical coordinate, draw the two-dimentional score scatterplot (point in one group of sample corresponding diagram) of training set sample; Labelling one polishing sample and the demarcation line not polishing sample, both sides, described demarcation line respectively polishing area and non-polishing area on this figure;
(5) the near-infrared absorption spectrum data of testing sample are gathered by the method identical with step (2), data obtain weighted value corresponding with described 1st characteristic vector and the 2nd characteristic vector respectively accordingly, as score value 1 and the score value 2 of testing sample;
(6) with the score value 1 of testing sample and score value 2 respectively horizontal, vertical coordinate, described two dimension score scatterplot is drawn the tested point of correspondence, the region according to described tested point place, differentiate whether testing sample is polishing grain.
Polishing grain of the present invention refers to the grain after mineral oil polishing, and the described grain that do not polish refers to the grain without any mineral oil polishing. Specifically, the mineral oil content not polished in sample of the present invention is 0; Mineral oil content in described polishing sample is 0.4～0.6%.
Described mineral oil refers to the fat hydrocarbon material in petroleum cracking composition, it is preferable that liquid paraffin.
Described grain is for being preferably corn, more preferably rice.
In the method for the invention, the unit of sample is group, and namely training set includes n group sample. The sampling method often organizing sample is: take the grain of same breed, the place of production and collecting time no less than 3kg, adopts heap four points of division methods of cone to obtain one group of sample, and the quality often organizing sample is 200 ± 5g.
In described step (1), polishing sample can be identical with the group number not polishing sample, it is also possible to different; Specifically, described polishing sample size is 1～3.5:1 with the ratio not polishing number of sets of sample. In order to ensure the accuracy of testing result, in described training set, it is preferable that described polishing sample and do not polish the group number of sample and be all not less than 40. In the method for the invention, in order to ensure the objectivity of testing result, when choosing polishing sample and do not polish sample, it should be ensured that it is representative; Described representativeness specifically refers to the kind of sample, the place of production and collecting time and covers objective circumstances comprehensively.
The present invention is 1cm for gathering the spectrogrph resolution of described near-infrared absorption spectrum data-1～64cm-1, it is preferred to 4cm-1～16cm-1。
In described step (2), often the interval between group spectroscopic data contained by sample near-infrared absorption spectrum is equal, is specially 1cm-1～3cm-1, it is preferred to 2cm-1. Described step (2) can also include described near-infrared absorption spectrum data are processed, so that testing result is more accurate. Described process includes smoothing, differential, baseline correction, one or more in data standard normal state; It is preferably data standard normal state.
Specifically, described data standard normal stateization processes, and is that the standard normalization conversion that namely each spectroscopic data does self, shown in its computing formula such as formula (I) for the near-infrared absorption spectrum data often organizing sample.
In described formula (I), X is spectroscopic data before treatment, XSNVIt it is the spectroscopic data after processing; Often organize in the near-infrared absorption spectrum data of sample, namely in each spectroscopic data, comprise m data, after treatment, obtain the data after m standard normalization processes, be i.e. a spectroscopic data after SNV process.
Wherein, XMIt is this spectroscopic data meansigma methods before treatment, shown in its computing formula such as formula (II); S is this spectroscopic data standard deviation before treatment, shown in its computing formula (III).
Step of the present invention (3) extracts the homing method of described characteristic vector selected from multiple linear regression, principal component regression, PLS, artificial neural network recurrence, Support vector regression; It is preferably PLS.
Under PLS algorithm, from training set data, ask for some stack features vector and obtain and described characteristic vector eigenvalue one to one, vectorial as the 1st characteristic vector and the 2nd characteristic vector using two stack features that eigenvalue is the highest, i.e. P1、P2; According to the near-infrared absorption spectrum data often organizing sample in training set, obtain weighted value corresponding with described 1st characteristic vector and the 2nd characteristic vector respectively, as score value 1 and the score value 2 of each training set sample, i.e. T1、T2. Wherein, T1、T2、P1、P2And XSNVBetween relation be:
XSNV=T1P1+T2P2+E(IV)
In formula (IV), E is spectrum residual error.
For testing sample, under same experimental conditions, gather the near infrared spectrum data of testing sample, through identical data prediction, i.e. the pretreated near infrared spectrum data X of gained testing sample after SNV processP_SNV, ask for XP_SNVAt P1、P2Weighted value, i.e. the score value 1 (T of testing sampleP_1) and the score value 2 (T of testing sampleP_2):
TP_1=XP_SNVP1
TP_2=XP_SNVP2
More than being the routine data processing method in linear algebra field, the present invention does not do particular determination.
For training set sample, respectively with T1、T2For horizontal, vertical coordinate, drafting two dimension score scatterplot. For each testing sample, respectively with TP_1And TP_2For horizontal, vertical coordinate, two dimension score scatterplot is drawn the tested point of correspondence, the region according to described tested point place, differentiates whether testing sample is polishing grain.
As a preferred embodiment of the present invention, in described step (3), to polishing sample assignment 0.1～5, it is preferred to 1, to not polishing sample assignment-5～-0.1, it is preferred to-1; Step (4) described demarcation line is be 135 °～170 °, be preferably the straight line of 150 ° through coordinate axes initial point and with abscissa positive direction, and described polishing area is positioned at described marginal upper right side.
A kind of mineral oil based on vibrational spectrum that the present invention proposes polishes grain fast non-destructive detection method, extends vibrational spectrum purposes in agricultural product quality and safety Fast nondestructive evaluation discriminating. The method has the features such as quick, accurate, work efficiency is high, green non-pollution, it is possible not only to provide powerful guarantee and technical support for the Fast nondestructive evaluation discriminating of mineral oil polishing agricultural product, and the aspects such as guarantee agricultural product quality and safety, the research and development improving industrial and agricultural production efficiency, promotion market for farm products sound development, Fast nondestructive evaluation instrument are had positive role.
Accompanying drawing explanation
Fig. 1 is that a kind of mineral oil based on vibrational spectrum of the present invention polishes grain fast non-destructive detection method operational flowchart;
Fig. 2 is the near-infrared absorption spectrum figure of mineral oil of the present invention polishing rice and non-polished rice;
Fig. 3 is the scatterplot that a kind of mineral oil based on vibrational spectrum of the present invention polishes grain fast non-destructive detection method embodiment 1;
Fig. 4 is the scatterplot that a kind of mineral oil based on vibrational spectrum of the present invention polishes grain fast non-destructive detection method comparative example 1.
Detailed description of the invention
For making the purpose of the embodiment of the present invention, technical scheme and advantage clearly, below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described. Obviously, described embodiment is a part of embodiment of the present invention, rather than whole embodiments. Based on the embodiment in the present invention, the every other embodiment that those of ordinary skill in the art obtain under the premise not making creative work, broadly fall into the scope of protection of the invention.
In following embodiment, instrument is: FT-mid-IR fiber optics spectroscopy instrument, model: Spectrum400, PerkinElmer company of the U.S.; Adnexa: integrating sphere diffuse-reflectance adnexa;
Parameter: spectral resolution: 8cm-1, spectral region: 12800cm-1～4000cm-1, accumulative frequency: 64 times, make reference with integrating sphere sky light path.
Sample: do not polish rice, mineral oil polishing rice.
Data process: adopt data standard normal state preprocess method that every spectrum is carried out data prediction. So-called data standard normal state, is that the standard normalization conversion that namely each spectroscopic data does self, shown in its computing formula such as formula (I) for the near-infrared absorption spectrum data often organizing sample.
In formula (I), X is the spectroscopic data before pretreatment, XSNVIt is pretreated spectroscopic data, XMBeing the meansigma methods before this spectroscopic data pretreatment, S is the standard deviation before this spectroscopic data pretreatment. XMWith the computing formula of S respectively as shown in formula (II) and formula (III).
Embodiment 1
Known mineral oil polishes rice sample 40 groups (being numbered 1#～40#) and does not polish rice sample 40 groups (being numbered 61#～100#), take 1# sample and 61# sample respectively, adopt identical method to gather 1# sample and 61# sample at 4000cm-1～12800cm-1Near infrared light spectrogram under wavelength, as shown in Figure 2.From Figure 2 it can be seen that the near infrared absorption of two groups of samples is all concentrated mainly on 4000cm-1～8900cm-1Wave band, and mineral oil polishing rice and not polish the spectrum shape of near-infrared absorption spectrum of rice, peak position basically identical, naked eyes cannot observe the difference of the two, that is only mineral oil cannot polish rice with perusal spectrogram and carry out detection and identification.
In order to realize that the quick nondestructive of polishing rice is differentiated, the present embodiment is adopted and is detected with the following method:
(1) with known polishing sample 1#～40# with not polish sample 61#～100# structure total sample number be the training set of 80;
(2) gather in described training set and often organize sample at 8900cm-1～12000cm-1Near-infrared absorption spectrum in wave-length coverage, often organizes the near-infrared absorption spectrum of sample with 2cm-11551 spectroscopic datas are included for spacing; Adopt above-mentioned data standard normal state preprocess method that the near-infrared absorption spectrum data often organizing sample are processed;
(3) with near-infrared absorption spectrum data construct 80 × 1551 rank matrix of whole samples after data standard normal stateization processes in described training set, it is 1 to the polishing sample assignment in described training set, it is-1 to not polishing sample assignment, extract ten stack features vectors of described matrix according to described numerical value regression algorithm and obtain and described characteristic vector eigenvalue one to one, vectorial as the 1st characteristic vector and the 2nd characteristic vector using two stack features that eigenvalue is the highest;
(4) according to the near-infrared absorption spectrum data often organizing sample in training set, obtain weighted value corresponding with described 1st characteristic vector and the 2nd characteristic vector respectively, score value 1 and score value 2 as each training set sample, and with its respectively horizontal, vertical coordinate, draw the two-dimentional score scatterplot (as shown in Figure 3) of training set sample; On this figure labelling one polishing sample with do not polish the demarcation line of sample, described demarcation line is through coordinate axes initial point and with abscissa positive direction be the straight line at 150 ° of angles, and described marginal upper right side is polishing area, lower left is non-polishing area;
(5) the near-infrared absorption spectrum data of testing sample are gathered by the method identical with step (2), data obtain weighted value corresponding with described 1st characteristic vector and the 2nd characteristic vector respectively accordingly, as score value 1 and the score value 2 of testing sample;
(6) with the score value 1 of testing sample and score value 2 respectively horizontal, vertical coordinate, described two dimension score scatterplot is drawn the tested point of correspondence, as shown in Figure 3, the region according to described tested point place, differentiate whether testing sample is polishing rice.
Adopt said method detection testing sample (being numbered 41#～60# and 101#～120#), result shows, testing sample 41#～tested point corresponding for 60# is all distributed in characteristic straight line upper right side, namely testing sample 41#～60# is polishing rice, testing sample 101#～tested point corresponding for 120# is all distributed in characteristic straight line lower left, and namely testing sample 101#～120# is not for polish rice; After the source of above-mentioned testing sample and the practical situation that whether polishes are confirmed, it is judged that above-mentioned testing result is all consistent with practical situation, and accuracy rate is 100%. , it is possible to whether rice is carried out Fast nondestructive evaluation by mineral oil polishing, it is not necessary to complicated sample pretreatment, and environmental friendliness, easy and simple to handle, work efficiency is high, detection accuracy is high as can be seen here, according to the method described in the present invention.
Comparative example 1
Compared with embodiment 1, differing only in, this comparative example is at 4000cm-1～12800cm-1All-wave length under gather the near-infrared absorption spectrum of training set and testing sample, gained gathers the scatterplot of training set sample as shown in Figure 4.Even if by Fig. 4 it will be seen that other experiment conditions, data processing method are all consistent with embodiment 1, at 4000cm-1～12800cm-1The training set sample scatterplot that the near infrared spectrum data gathered under wavelength obtains, it is difficult to draw out the characteristic demarcation line clearly distinguished of rice samples different for two kinds of polishing situations, it is impossible to obtain and analyze result accurately.
A kind of mineral oil based on vibrational spectrum that the present invention proposes polishes grain fast non-destructive detection method and extends vibrational spectrum purposes in agricultural product quality and safety quick nondestructive discriminating. The method has the features such as quick, accurate, work efficiency is high, green non-pollution, it is possible not only to provide powerful guarantee and technical support for the discriminating of deterioration agricultural product quick nondestructive, and the aspects such as guarantee agricultural product quality and safety, the research and development improving industrial and agricultural production efficiency, promotion market for farm products sound development, Fast nondestructive evaluation instrument are had positive role.
Although, above use generality explanation, detailed description of the invention and test, the present invention is described in detail, but on basis of the present invention, it is possible to it is made some modifications or improvements, and this will be apparent to those skilled in the art. Therefore, these modifications or improvements without departing from theon the basis of the spirit of the present invention, belong to the scope of protection of present invention.
Claims (10)
1. the quick nondestructive discrimination method polishing grain, it is characterised in that said method comprising the steps of:
(1) with known polishing grain with not polish grain be sample, build sample and always organize the training set that number is n; In described training set, n >=50, the ratio of polishing sample and the group number not polishing sample is 1～3.5:1;
(2) gather in described training set and often organize sample at 8900cm-1～12000cm-1Near-infrared absorption spectrum in wave-length coverage, often organizes the near-infrared absorption spectrum of sample with 1cm-1～3cm-1M spectroscopic data is included for interval;
(3) with the near-infrared absorption spectrum data construct n × m rank matrix of samples whole in described training set, to the polishing sample in described training set with do not polish sample and give different numerical value respectively, extract some stack features vector of described matrix according to described numerical value regression algorithm and obtain and described characteristic vector eigenvalue one to one, vectorial as the 1st characteristic vector and the 2nd characteristic vector using two stack features that eigenvalue is the highest;
(4) according to the near-infrared absorption spectrum data often organizing sample in training set, obtain weighted value corresponding with described 1st characteristic vector and the 2nd characteristic vector respectively, score value 1 and score value 2 as each training set sample, and with its respectively horizontal, vertical coordinate, draw the two-dimentional score scatterplot of training set sample; Labelling one polishing sample and the demarcation line not polishing sample, both sides, described demarcation line respectively polishing area and non-polishing area on this figure;
(5) the near-infrared absorption spectrum data of testing sample are gathered by the method identical with step (2), data obtain weighted value corresponding with described 1st characteristic vector and the 2nd characteristic vector respectively accordingly, as score value 1 and the score value 2 of testing sample;
(6) with the score value 1 of testing sample and score value 2 respectively horizontal, vertical coordinate, described two dimension score scatterplot is drawn the tested point of correspondence, the region according to described tested point place, differentiate whether testing sample is polishing grain.
2. method according to claim 1, it is characterised in that the sampling method often organizing sample is: take the grain of same breed, the place of production and collecting time no less than 3kg, adopts heap four points of division methods of cone to obtain one group of sample, and the quality often organizing sample is 200 ± 5g.
3. method according to claim 1 and 2, it is characterised in that described grain is corn, it is preferred to rice;
Described mineral oil is liquid paraffin.
4. the method according to claims 1 to 3 any one, it is characterised in that in described training set, polishing sample and do not polish the quantity of sample and be all not less than 40.
5. the method according to Claims 1 to 4 any one, it is characterised in that the described mineral oil content not polishing sample is 0; Mineral oil content in described polishing sample is 0.4～0.6%.
6. the method according to Claims 1 to 5 any one, it is characterised in that the spectrogrph resolution being used for gathering described near-infrared absorption spectrum data is 1cm-1～64cm-1, it is preferred to 4cm-1～16cm-1。
7. the method according to claim 1～6 any one, it is characterised in that step (2) also includes described near-infrared absorption spectrum data are processed;
Described process includes smoothing, differential, baseline correction, one or more in data standard normal state; It is preferably data standard normal state.
8. the method according to claim 1～7 any one, it is characterized in that, step (3) extracts the homing method of described characteristic vector selected from multiple linear regression, principal component regression, PLS, artificial neural network recurrence, Support vector regression; It is preferably PLS.
9. method according to claim 1, it is characterised in that in described step (3), to polishing sample assignment 0.1～5, to not polishing sample assignment-5～-0.1;
Step (4) described demarcation line is through coordinate axes initial point and with abscissa positive direction be the straight line of 135 °～170 °, and polishing area is positioned at described marginal upper right side, and non-polishing area is positioned at described marginal lower left.
10. the fast non-destructive detection method polishing rice, it is characterised in that said method comprising the steps of:
(1) with known be whether that to build sample group number be the training set of n for the sample of polishing rice; In described training set, n=80～100, polishing sample and do not polish the group number of sample and be all not less than 40;
(2) gather in described training set and often organize sample at 8900cm-1～12000cm-1Near-infrared absorption spectrum in wave-length coverage, often organizes the near-infrared absorption spectrum of sample with 2cm-1M spectroscopic data is included for interval; Described spectroscopic data carries out data standard normal stateization process;
(3) with the near-infrared absorption spectrum data construct n × m rank matrix of whole samples after data standard normal stateization processes in described training set, it is 1 to the polishing sample assignment in described training set, it is-1 to not polishing sample assignment, extract some stack features vector of described matrix according to described numerical value regression algorithm and obtain and described characteristic vector eigenvalue one to one, vectorial as the 1st characteristic vector and the 2nd characteristic vector using two stack features that eigenvalue is the highest;
(4) according to the near-infrared absorption spectrum data often organizing sample in training set, obtain weighted value corresponding with described 1st characteristic vector and the 2nd characteristic vector respectively, score value 1 and score value 2 as each training set sample, and with its respectively horizontal, vertical coordinate, draw the two-dimentional score scatterplot of training set sample; On this figure labelling one polishing sample with do not polish the demarcation line of sample, described demarcation line is through coordinate axes initial point and with abscissa positive direction be the straight line at 150 ° of angles, and described marginal upper right side is polishing area, lower left is non-polishing area;
(5) the near-infrared absorption spectrum data of testing sample are gathered by the method identical with step (2), data obtain weighted value corresponding with described 1st characteristic vector and the 2nd characteristic vector respectively accordingly, as score value 1 and the score value 2 of testing sample;
(6) with the score value 1 of testing sample and score value 2 respectively horizontal, vertical coordinate, described two dimension score scatterplot is drawn the tested point of correspondence, the region according to described tested point place, differentiate whether testing sample is polishing rice.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610180371.1A CN105675534A (en) | 2016-03-25 | 2016-03-25 | Method for quickly and nondestructively identifying polished grains |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610180371.1A CN105675534A (en) | 2016-03-25 | 2016-03-25 | Method for quickly and nondestructively identifying polished grains |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105675534A true CN105675534A (en) | 2016-06-15 |
Family
ID=56225333
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610180371.1A Pending CN105675534A (en) | 2016-03-25 | 2016-03-25 | Method for quickly and nondestructively identifying polished grains |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105675534A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106706596A (en) * | 2017-01-09 | 2017-05-24 | 河南工业大学 | Raman spectrum detection method for rice freshness |
| CN110068544A (en) * | 2019-05-08 | 2019-07-30 | 广东工业大学 | Material identification network model training method and tera-hertz spectra substance identification |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH03160345A (en) * | 1989-11-17 | 1991-07-10 | Satake Eng Co Ltd | Method of evaluating quality of boiled rice |
| JP2005069930A (en) * | 2003-08-26 | 2005-03-17 | Jfe Steel Kk | Manufacturing method for directional electromagnetic steel plate having excellent coating adhesiveness |
| CN101655454A (en) * | 2009-09-15 | 2010-02-24 | 北京市农林科学院 | Rapid determination method for evaluation of storage quality of grain |
| CN101936895A (en) * | 2010-09-02 | 2011-01-05 | 中南林业科技大学 | Near infrared spectroscopy analysis rapid detection method of rice freshness |
| CN102087212A (en) * | 2010-11-25 | 2011-06-08 | 西南大学 | Pueraria lobata starch adulteration identification method based on principal component analysis |
| CN102564993A (en) * | 2011-12-31 | 2012-07-11 | 江南大学 | Method for identifying rice varieties by using Fourier transform infrared spectrum and application of method |
| CN102768196A (en) * | 2012-08-13 | 2012-11-07 | 中国计量学院 | Method for identifying different transgenic rice |
| CN103344602A (en) * | 2013-07-04 | 2013-10-09 | 中国科学院合肥物质科学研究院 | Nondestructive testing method for rice idioplasm authenticity based on near infrared spectrum |
| CN103575694A (en) * | 2013-11-11 | 2014-02-12 | 东北农业大学 | Quick detection method for aflatoxin B1 of rice based on near infrared spectroscopy |
| CN103592256A (en) * | 2013-11-29 | 2014-02-19 | 重庆市计量质量检测研究院 | Mid-infrared spectroscopic method for distinguishing normal edible vegetable oil from refined hogwash oil based on Fourier transform |
| 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 |
| CN105424636A (en) * | 2015-12-29 | 2016-03-23 | 南京财经大学 | Rapid detection method for grain fungal pollution cases and applications thereof |
-
2016
- 2016-03-25 CN CN201610180371.1A patent/CN105675534A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH03160345A (en) * | 1989-11-17 | 1991-07-10 | Satake Eng Co Ltd | Method of evaluating quality of boiled rice |
| JP2005069930A (en) * | 2003-08-26 | 2005-03-17 | Jfe Steel Kk | Manufacturing method for directional electromagnetic steel plate having excellent coating adhesiveness |
| CN101655454A (en) * | 2009-09-15 | 2010-02-24 | 北京市农林科学院 | Rapid determination method for evaluation of storage quality of grain |
| CN101936895A (en) * | 2010-09-02 | 2011-01-05 | 中南林业科技大学 | Near infrared spectroscopy analysis rapid detection method of rice freshness |
| CN102087212A (en) * | 2010-11-25 | 2011-06-08 | 西南大学 | Pueraria lobata starch adulteration identification method based on principal component analysis |
| CN102564993A (en) * | 2011-12-31 | 2012-07-11 | 江南大学 | Method for identifying rice varieties by using Fourier transform infrared spectrum and application of method |
| CN102768196A (en) * | 2012-08-13 | 2012-11-07 | 中国计量学院 | Method for identifying different transgenic rice |
| CN103344602A (en) * | 2013-07-04 | 2013-10-09 | 中国科学院合肥物质科学研究院 | Nondestructive testing method for rice idioplasm authenticity based on near infrared spectrum |
| CN103575694A (en) * | 2013-11-11 | 2014-02-12 | 东北农业大学 | Quick detection method for aflatoxin B1 of rice based on near infrared spectroscopy |
| CN103592256A (en) * | 2013-11-29 | 2014-02-19 | 重庆市计量质量检测研究院 | Mid-infrared spectroscopic method for distinguishing normal edible vegetable oil from refined hogwash oil based on Fourier transform |
| 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 |
| CN105424636A (en) * | 2015-12-29 | 2016-03-23 | 南京财经大学 | Rapid detection method for grain fungal pollution cases and applications thereof |
Non-Patent Citations (2)
| Title |
|---|
| 张耀武等: "红外光谱法测定涂油大米中的矿物油", 《粮食食品科技》 * |
| 范璐等: "涂石蜡大米傅立叶变换红外吸收光谱识别分析", 《分析实验室》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106706596A (en) * | 2017-01-09 | 2017-05-24 | 河南工业大学 | Raman spectrum detection method for rice freshness |
| CN110068544A (en) * | 2019-05-08 | 2019-07-30 | 广东工业大学 | Material identification network model training method and tera-hertz spectra substance identification |
| CN110068544B (en) * | 2019-05-08 | 2021-09-17 | 广东工业大学 | Substance identification network model training method and terahertz spectrum substance identification method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Feng et al. | Preliminary study on classification of rice and detection of paraffin in the adulterated samples by Raman spectroscopy combined with multivariate analysis | |
| CN104990894B (en) | A kind of gasoline property detection method based on weighting absorbance and similar sample | |
| CN107064047A (en) | A kind of Fuji apple quality damage-free detection method based near infrared spectrum | |
| CN105181642A (en) | Near-infrared detection method for peanut quality and application | |
| CN106124445A (en) | A kind of quick, Undamaged determination genetically engineered soybean method | |
| CN106841083A (en) | Sesame oil quality detecting method based on near-infrared spectrum technique | |
| CN109669023A (en) | A kind of soil attribute prediction technique based on Multi-sensor Fusion | |
| CN103674864A (en) | Fish water content distribution detection method based on hyperspectral imaging technology | |
| CN109540836A (en) | Near infrared spectrum pol detection method and system based on BP artificial neural network | |
| CN102735642A (en) | Method for quickly and losslessly identifying virgin olive oil and olive-residue oil | |
| CN106198447A (en) | Chemical Mixed Fertilizer main component harmless quantitative detection method based on near-infrared spectrum technique | |
| CN109374548A (en) | A method of quickly measuring nutritional ingredient in rice using near-infrared | |
| Si et al. | Quality assessment of fruits and vegetables based on spatially resolved spectroscopy: A review | |
| CN104655585B (en) | A kind of PSE meat screening technique based near infrared spectrum | |
| CN106018321A (en) | Establishment method for single corn seed protein detection model and application of establishment method | |
| CN105675534A (en) | Method for quickly and nondestructively identifying polished grains | |
| He et al. | Study on the identification of resistance of rice blast based on near infrared spectroscopy | |
| Wang et al. | Grading detection of “Red Fuji” apple in Luochuan based on machine vision and near-infrared spectroscopy | |
| CN106483095A (en) | A kind of method of each component oil content in quick, accurate quantitative analysis quaternary ready-mixed oil | |
| CN109406447A (en) | A kind of near infrared detection method of tannin in sorghum | |
| CN111398208A (en) | Method for rapidly identifying traditional art jadeite by utilizing near-infrared technology | |
| CN107037039B (en) | A kind of Xinjiang walnut place of production is traced to the source research method | |
| CN105158178A (en) | Rapid modeling method for detecting sugar content of navel orange based on spectral peak area in high spectral transmission technology | |
| CN104914068A (en) | Spectrum rapid detection method of trans-fatty acid content in grease | |
| van Wyngaard et al. | Infrared spectroscopy investigation of fresh grapevine (Vitis vinifera) shoots, leaves, and berries using novel chemometric applications for viticultural data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160615 |
|
| RJ01 | Rejection of invention patent application after publication |