CN101059431A - Fusarium fungus infrared fingerprint quick identification method - Google Patents
Fusarium fungus infrared fingerprint quick identification method Download PDFInfo
- Publication number
- CN101059431A CN101059431A CN 200710040582 CN200710040582A CN101059431A CN 101059431 A CN101059431 A CN 101059431A CN 200710040582 CN200710040582 CN 200710040582 CN 200710040582 A CN200710040582 A CN 200710040582A CN 101059431 A CN101059431 A CN 101059431A
- Authority
- CN
- China
- Prior art keywords
- fusarium
- infrared fingerprint
- authentication method
- fingerprint authentication
- analysis
- 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
Images
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention relates to an infrared finger check method of Fusarium fungus, comprising that (1), collecting spectrum data, finding main absorption peak, (2), processing main component analysis of the binary derivative spectrum data which has been treated by base correction and normalization, to reduce variable dimension, (3), extracting front ten main components of the treated spectrum data as the nodes of artificial nerve network input layer, (4), coding the Fusarium, to build artificial nerve network model, calculating and comparing quantitatively. The inventive analysis method has small sample quantity, reasonable data pickup model, analysis method, and model analysis, with simple operation, high analysis speed, high accuracy, and low cost, to be used in disease prevention, biological prevention or the like.
Description
Technical field
The invention belongs to biotechnology and artificial intelligence field, particularly relate to the authentication method of a kind of fungi.
Background technology
The infrared fingerprint technology claims infra-red sepectrometry (FTIRS) again, is a kind of qualitative, the quantitative analysis method of the absorption characteristic of infrared radiation being set up with the research material.Infrared spectrum has the characteristic of height, and concerning the simplification compound, the infrared spectrum that does not almost have two compounds is on all four, and promptly each compound all has the infrared absorption spectrum of own feature.The infrared fingerprint technology applies to distinguish different microorganisms since the 1950's.The infrared spectroscopy signals of microorganism has very strong characteristic fingerprint feature, can differentiate fast, classify on the subspecies level of microorganism and screening on a large scale, mainly concentrates on bacterium, little algae and minority fungi in the microbiology field.At present, this method has expanded to other microorganisms except that bacterium, as the qualitative analysis of virus, fungi, saccharomycete even mammalian cell.In addition, also see similar report in the quality discriminating of Chinese crude drug, production control etc.
The infrared fingerprint The Application of Technology is worth and mainly contains:
1) agricultural protection: the sickle-like bacteria of Rapid identification harm industrial crops.As the fusarium moniliforme of the Fusarium oxysporum that causes the tung oil tree droop and the bakanae disease of rice that causes.
2) commercial production: the toxigenic sickle-like bacteria of metabolism in Rapid identification food or the feed.As the toxin that produces is the Fusarium graminearum and the Fusarium nivale that produces fusaric acid, dehydrofusarinic acid etc. of zearalenone.
3) health care: Rapid identification influences the sickle-like bacteria of human beings'health.As cause the fusarium semitectum of carcinogenic, teratogenesis, mutagenic potential hazard.
Fusarium fungus is distributed widely in the residual body of soil, waters, plant, animal and organism, the habitat complexity, and form is various.Many kinds are plant, animal and human pathogen, and the mycotoxin of its generation can jeopardize human and livestock health.
Though the principle of infrared fingerprint technology is similar, its sampling method, method for making sample, analytical approach all have nothing in common with each other.Along with the high speed development of computer technology and electronic technology, the fourier transform infrared spectroscopy of high-resolution, high light flux develops in the researchs such as quick discriminating of the Analysis and Identification of complex mixture, the damage monitoring of active somatic cell, the screening of bacterium, medicinal animal and plant true and false quality and many different detection meanss.Thereby caused the infrared fingerprint technology to be applied in the practice different gordian technique points is often arranged.
Because the technology comparative maturity of infrared spectrum equipment is so the key of fungi fingerprint identification is infrared data result's analytical approach.Close for interspecies relation, evolutionary relationship is nearer, and the microorganism that chemical information difference is little is adopted general multivariate statistical method to be difficult to realize the analysis of its infrared data, and needs to adopt artificial intelligence model to overcome these defectives.
" analytical chemistry " 2003 12 phases " Primary Study of the infrared spectrum method of discrimination of drug-resistant type candida albicans " disclose a kind of sensitive strain at candida albicans, persister, type strain and have carried out the infrared fingerprint analysis, utilize the spectrum cluster of former spectrum and, second derivative combination, persister and sensitive strain can be separated, but this preliminary information processing is unusual difficulty for the fungi of the such wide material sources of Fusarium, kind complexity.Reason is that the original spectrum of this method does not carry out homogenization to be handled, and therefore can not eliminate the SPECTRAL DIVERSITY that the sample size difference is brought, and repeatability reduces greatly, and science is not high; Infrared lamp is oven dry destructible fungal cell content down; This method adopts protein and nucleic acid SPECTRAL REGION to analyze, and quantity of information is limited, only is applicable to the fungi of protein and nucleic acid differences; This method adopts the judgement of the relative content (being protein and nucleic acid absorption peak ratio) between protein and the nucleic acid can produce defectives such as " taking a part for the whole ", interference from human factor are big.So this method is not suitable for the such filamentous fungi of Fusarium and analyzes.
Artificial neural network (Artificial Neural Networks is by a large amount of simple primary elements ANNs)---neuron is coupled to each other the self-adaptation nonlinear dynamic system that forms.Each neuronic 26S Proteasome Structure and Function is fairly simple, but the system action of a large amount of neuron combination results is very complicated.Artificial neural network has reflected some fundamental characteristics of human brain function, but is not the description true to nature of biosystem, just certain imitation, simplification and abstract.Compare with traditional digital operation, artificial neural network at aspects such as principle of compositionality and functional characteristics more near human brain, it is not to carry out computing length by length by given program, but can self conform, sums up rule, finishes certain computing, identification or process control.
The BP artificial neural network is the abbreviation of Back-Propagation Network (counterpropagation network), is the multilayer feedforward neural network that was at first proposed in 1986 by the scientific research group headed by Rumelhart and the Mccelland.Generally contain one or more hidden layers in the BP network, input signal to hidden layer, through the excitation function effect, propagates into output layer to the hidden layer output signal by the input layer propagated forward again, exports result (Fig. 1) at last.The learning process of BP neural network is made up of two parts: the forward transmission of information and the back transfer of error.In the forward-propagating process, input information successively calculates biography to output layer from input through hidden layer, and the neuronic state of each layer only influences the neuronic state of one deck down.If output layer does not obtain the output expected, then calculate the error changing value of output layer, turn to backpropagation then, by network error signal is gone back to revise weights that each layer neuron connect until reaching expectation target along original connecting path anti-pass.The advantageous property of BP artificial neural network mainly is reflected in aspects such as learning ability, adaptive ability, parallel processing capability, distributed storage ability and fault-tolerant ability, can handle complicated nonlinear characteristic object, abilities such as having stronger modeling, proofread and correct and estimate (Fischer MM LeungY.Geocomputational modelling:techniques and applications[M] .Springer, 2001).The BP algorithm is to control automatically at present to go up efficient algorithms most important, that application is maximum.
Usually with the major component (PCs) of principal component analysis (PCA) as the method for artificial neural network input layer be called the major component artificial neural network (Principal Component-Artificial NeuralNetworks, PC-ANNs).Principal component analysis (Principal Components Analysis, PCA) be that a large amount of variablees that will have correlativity each other are transformed into and are called main composition (PrincipalComponents, the mathematical method of the less and incoherent multi-stress of PCs) one group, and major component has kept most differences of raw data.Number that it is generally acknowledged major component influences neural network prediction, and the number that promptly reduces major component will shorten the neural network learning time, and the quantity of information that very few major component comprises very little.By analyzing different artificial neural network structures and learning parameter, the artificial nerve network model that can determine to optimize employing is as 10-15-1 or 390-15-1 etc.10-15-1 model researcher is abroad analyzed also report (AppliedSpectroscopy such as Udelhoven T, 2000,54 (10): 1471-1479) once of bacterium spectroscopic data.The major component artificial nerve network model is widely used in the damaged detection of arc welding, local tea variety evaluation, Pharmaceutical Analysis at present, determines many-sided research such as ozone concentration and metal ion content (Mirapeix J etc., NDT ﹠amp; E International, 2007,40 (4): 315-323.He Y etc., Journal of Food Engineering, 2007,79 (4): 1238-1242.Sousa SIV etc., Environmental Modelling ﹠amp; Software, 2007,22 (1): 97-103.Abbaspour A etc., Spectrochimica Acta Part A, 2006,64 (2): 477-482.Esseiva P etc., Talanta, 2005,67 (2): 360-367.)
Also do not have both at home and abroad and utilize the report of BP artificial neural network in conjunction with the infrared fingerprint method of principal component analysis (PCA) check and analysis simultaneously separate sources, different region, sickle-like bacteria that similarity is high.
Summary of the invention
Technical matters to be solved
Technical matters to be solved by this invention provides a kind of infrared fingerprint quick identification method of fusarium fungus, to overcome the analysis that general multivariate statistical method is difficult to realize complicated infrared data, can not detect the defective of the infrared fingerprint method of separate sources, different region, sickle-like bacteria that similarity is high simultaneously.
Technical scheme
Technical scheme of the present invention provides a kind of infrared fingerprint authentication method of fusarium fungus, comprises the steps:
1) spectrum data gathering belongs to main absorption peak;
2) will carry out principal component analysis (PCA) through baseline correction and normalized second derivative spectra data, reduction variable dimension;
3) spectroscopic data of analyzing through principal component analysis (PCA) is got preceding 10 major components as the artificial neural network input layer;
4) sickle-like bacteria is encoded, make up artificial nerve network model, and calculate quantitative comparison.
One of preferred version of the infrared fingerprint authentication method of above-mentioned fusarium fungus is that said artificial neural network is the BP artificial neural network.Preferred said BP artificial nerve network model η be 0.1 and α be 0.6.
Two of the preferred version of the infrared fingerprint authentication method of above-mentioned fusarium fungus is that said artificial nerve network model has 10 input layers, 15 hiding node layers and 1 output layer node.
Three of the preferred version of the infrared fingerprint authentication method of above-mentioned fusarium fungus is that said main absorption peak is the absorption peak of protein, polypeptide, glucosan, shitosan, nucleic acid, phosphatide, fatty acid, fat or its combination.
Four of the preferred version of the infrared fingerprint authentication method of above-mentioned fusarium fungus is that said infrared fingerprint is positioned at light spectral 3000-2800cm
-1And 1800-500cm
-1Preferably, said infrared fingerprint spectrum spectrum is distinguished and is respectively: 3000-2800cm
-1Fatty acid zone, 1800-1480cm
-1Protein acid amides I band and protein acid amides II band, 1480-1190cm from protein and polypeptide
-1The Mixed Zone of protein and fatty acid, 1190-900cm
-1Polysaccharide zone, 900-500cm
-1There are not the cellular component of ownership and the true finger-print region of functional group.
Five of the preferred version of the infrared fingerprint authentication method of above-mentioned fusarium fungus is that said fusarium fungus is fusarium semitectum, beading reaping hook, Fusarium graminearum, Fusarium oxysporum and Fusarium nivale.
Six of the preferred version of the infrared fingerprint authentication method of above-mentioned fusarium fungus is that said quantitative comparison result adopts square (R of Pearson's related coefficient
2) carry out check analysis with average relative standard deviation (R.S.E).
Beneficial effect
The infrared fingerprint quick identification method of fusarium fungus of the present invention has overcome the defective that general multivariate statistical method is difficult to realize the analysis of complicated infrared data.Prior art is utilized the spectrum cluster of former spectrum and, second derivative combination, persister and sensitive strain can be separated, but, this preliminary information processing is not carried out homogenization to original spectrum and is handled, therefore can not eliminate the SPECTRAL DIVERSITY that the sample size difference is brought, repeatable low, be very difficult for the fungi of analyzing the such wide material sources of Fusarium, kind complexity.The analysis and processing method of artificial intelligence model of the present invention makes the infrared fingerprint method can detect separate sources, different region, sickle-like bacteria that similarity is high simultaneously.
Prior art adopts protein and nucleic acid SPECTRAL REGION to analyze, and quantity of information is limited, only is applicable to the fungi of protein and nucleic acid differences.Present technique is selected full analysis of spectrum, and applicability is more extensive.
Prior art adopts the judgement of the relative content (being protein and nucleic acid absorption peak ratio) between protein and the nucleic acid can produce defectives such as " taking a part for the whole ", interference from human factor are big.Present technique has adopted mode construction that artificial neural network combines with principal component analysis (PCA) major component artificial neural network, utilize the good learning ability of artificial neural network, adaptive ability, parallel processing capability, distributed storage ability and fault-tolerant ability, handle complicated nonlinear characteristic object, abilities such as having powerful modeling, proofread and correct and estimate, it is slower to have overcome traditional artificial neural network speed of convergence, poor stability easily is absorbed in shortcomings such as local minimum.Therefore can set up fungi spectra database widely, practicality significantly strengthens.
The result shows that the major component artificial neural network is at R
2All being better than directly adopting raw data during value is added up with average relative standard deviation is the model of artificial neural network input layer, show good reliability and stability, and shortens the time of required analysis greatly.
Simultaneously, the present invention adopts the vacuum drying fungal cell, avoids oven dry, destructible fungal cell content under the infrared lamp, has fast, keeps advantages such as cell composition.
It is low that infrared fingerprint method of the present invention has overcome the accuracy of traditional form method, and defective such as waste time and energy.Only need the sickle-like bacteria of a small amount of biomass just can finish analysis, and spectral analysis speed is fast, has shortened qualification time greatly; The sickle-like bacteria that is used to analyze does not simultaneously need to destroy cell and adds any chemical reagent, has improved authenticity greatly and has reduced analysis cost; And simple to operate, those skilled in the art just can finish.
The present invention can be widely used in the sickle-like bacteria of identifying different regions or source, relates to fields such as agricultural, industry, medical science, food and animal doctor.Can utilize sampling and method for making sample, data acquisition scheme, analytical approach and the model set up, realize the evaluation fast and accurately of sickle-like bacteria.For in time carry out prevention and control of diseases reasonably, quality time and reasearch funds have been saved in work such as biological control.
Description of drawings
Fig. 1 is the working model that artificial neural network is handled spectroscopic data.
Fig. 2 is the infrared spectrum of Fusarium graminearum.Among the figure, horizontal ordinate is a wave number; As shown in the figure, the sickle-like bacteria infrared spectrum can be divided into 5 zones: [W1] 3000-2800cm
-1, the fatty acid zone; [W2] 1800-1480cm
-1, from the protein acid amides I band and the protein acid amides II band of protein and polypeptide; [W3] 1480-1190cm
-1, the Mixed Zone of protein and fatty acid; [W4] 1190-900cm
-1, the polysaccharide zone; [W5] 900-500cm
-1, great majority do not have the cellular component of ownership and the true finger-print region of functional group.
Fig. 3 is sickle-like bacteria second derivative spectra figure.
Fig. 4 is the curve fitting synoptic diagram of sickle-like bacteria spectrogram.Among the figure, horizontal ordinate position wave number, ordinate is an absorption value.
Fig. 5 is the principal component analysis (PCA) of sickle-like bacteria.Among the figure, horizontal ordinate is a major component one, and ordinate is a major component two; G: Fusarium graminearum, M: fusarium moniliforme, N: Fusarium nivale, S: fusarium semitectum, O: Fusarium oxysporum
Fig. 6 is an artificial nerve network model.
Fig. 7 is the hierarchial-cluster analysis result of sickle-like bacteria.
Embodiment
Below in conjunction with specific embodiment, further set forth the present invention.Should be understood that these embodiment only to be used to the present invention is described and be not used in and limit the scope of the invention.Should be understood that in addition those skilled in the art can make various changes or modifications the present invention after the content of having read the present invention's instruction, these equivalent form of values fall within the application's appended claims institute restricted portion equally.
Bacterial classification source: fusarium semitectum F.semitectum, fusarium moniliforme F.moniliforme, Fusarium graminearum F.graminearum, Fusarium oxysporum F.oxysporum and Fusarium nivale F.nivale, take from Yangtze river basin periphery field-crop root, stem is on the leaf; In the contaminated industrial products; In patient's focus, and identify through the molecules method; Nutrient culture media: potato sucrose nutrient culture media; Detecting instrument: NicoletAvatar-370 type Fourier transform infrared spectrometer (U.S. power ﹠ light company); The infrared light spectral: protein, carbohydrates, nucleic acid, lipid and other biomacromolecule spectrum district are specially 3000-2800cm
-1, 1800-500cm
-1
1) gathers sample, cultivate the back sample preparation and carry out spectrum data gathering, and belong to main protein, glucosan, shitosan, nucleic acid, phosphatide, fatty acid and fat absorption peak.The result shows spectrogram resolution height, favorable reproducibility (Fig. 2).
2) adopt second derivative spectra (Fig. 3) and curve fitting (Fig. 4) to characterize the structure of protein, carbohydrates, nucleic acid, lipid and other biomacromolecule in each dimorphic fungal cell.There is rich diversity between result's demonstration sickle-like bacteria.
3), the difference that constitutes the chemical substance of biology and the chemical substance that biosome is produced in life process is come comparison and difference species based on the sickle-like bacteria textural difference; Then, the series methods that adopts mathematical induction, deduction, inference etc. to set up is characterized as the basis with textural difference, utilizes a large amount of spectroscopic datas, presses digital model and metrology method (Fig. 5), carries out mathematical analysis.Concrete grammar is: at first, will carry out principal component analysis (PCA) through baseline correction and normalized second derivative spectra data, reduction variable dimension.Principal component analysis (PCA) is a kind of statistical analysis technique that a plurality of indexs is turned to the minority overall target, adopt major component (PCs) to see Goodacre etc. as the concrete grammar of artificial neural network input layer, FEMSMicrobiology Letters, 1996,140:233-239 and Goodacre etc., Microbiology, 1998,144:1157-1170. the original spectrum data are too many in the variable number, and exist certain correlativity each other, thereby make the data of being observed that the overlapping of information be arranged to a certain extent.When variable more for a long time, the regularity of distribution of research sample trouble more just in higher dimensional space.The method of dimensionality reduction is taked in principal component analysis (PCA), finds out several multi-stresses and represents original numerous variable, makes these multi-stresses reflect the quantity of information of primal variable as much as possible, and uncorrelated mutually each other, thereby reaches the purpose of simplification.Through the spectroscopic data that principal component analysis (PCA) is analyzed, get preceding 10 major components as the artificial neural network input layer, these major components are explained 99.93% of total variances altogether.Then, sickle-like bacteria is encoded, such as: with fusarium semitectum (F.semitectum), fusarium moniliforme (F.moniliforme), Fusarium graminearum (F.graminearum), Fusarium oxysporum (F.oxysporum) is encoded to 1.0000 respectively with Fusarium nivale (F.nivale), 2.0000,3.0000,4.0000 with 5.0000, method (Applied Spectroscopy such as Udelhoven T with reference to Udelhoven etc., 2000,54 (10): the 1471-1479) artificial nerve network model of structure 10-15-1, i.e. 10 input layers, 15 hiding node layers and 1 output layer node (Fig. 6).With reference to the method for Dou etc. (Dou Y etc., AnalyticalBiochemistry, 2006,351 (2): 174-180.), check different learning parameters, preferably determine BP artificial nerve network model η be 0.1 with α be 0.6.At last, (2006,63 (1): 99-108.), the result that the applying electronic computing machine draws makes organic quantitative comparison for Furusj E etc., Chemosphere with reference to the method for Furusj etc.Analysis result has reflected the sibship and the inner hereditary difference of its evolutionary process.(Fig. 7)
4) square (R by Pearson's related coefficient
2) presentation of results analyzed with average relative standard deviation (R.S.E), artificial neural network learning group experiment value and test group experiment value and each sickle-like bacteria encoded radio are very approaching, artificial neural network is successfully predicted by major component and has been discerned each sickle-like bacteria, makes the infrared fingerprint technology can be used to identify sickle-like bacteria accurately.The size (0-1) of Pearson's related coefficient square can be pointed out the level of intimate of two variable relations, and numerical value is high more, and then degree of accuracy is high more; Otherwise average relative standard deviation then.Square seeing the following form of Pearson's related coefficient with average relative standard deviation's analysis result:
| Grouping | The major component artificial neural network | |
| R 2 | R.S.E.(％) | |
| Study group | 0.9998 | 0.7375 |
| The checking group | 0.9997 | 0.8393 |
Claims (9)
1. the infrared fingerprint authentication method of a fusarium fungus comprises the steps:
1) spectrum data gathering belongs to main absorption peak;
2) will carry out principal component analysis (PCA) through baseline correction and normalized second derivative spectra data, reduction variable dimension;
3) spectroscopic data of analyzing through principal component analysis (PCA) is got preceding 10 major components as the artificial neural network input layer;
4) sickle-like bacteria is encoded, make up artificial nerve network model, and calculate quantitative comparison.
2. the infrared fingerprint authentication method of fusarium fungus according to claim 1 is characterized in that, said artificial neural network is the BP artificial neural network.
3. the infrared fingerprint authentication method of fusarium fungus according to claim 1 is characterized in that, said artificial nerve network model has 10 input layers, 15 hiding node layers and 1 output layer node.
4. the infrared fingerprint authentication method of fusarium fungus according to claim 1 is characterized in that, said main absorption peak is the absorption peak of protein, polypeptide, glucosan, shitosan, nucleic acid, phosphatide, fatty acid, fat or its combination.
5. the infrared fingerprint authentication method of fusarium fungus according to claim 1 is characterized in that, said infrared fingerprint is positioned at light spectral 3000-2800cm
-1And 1800-500cm
-1
6. the infrared fingerprint authentication method of fusarium fungus according to claim 5 is characterized in that, said infrared fingerprint spectrum spectrum district is respectively: 3000-2800cm
-1Fatty acid zone, 1800-1480cm
-1Protein acid amides I band and protein acid amides II band, 1480-1190cm from protein and polypeptide
-1The Mixed Zone of protein and fatty acid, 1190-900cm
-1Polysaccharide zone and 900-500cm
-1There are not the cellular component of ownership and the true finger-print region of functional group.
7. the infrared fingerprint authentication method of fusarium fungus according to claim 1 is characterized in that, said fusarium fungus is fusarium semitectum, beading reaping hook, Fusarium graminearum, Fusarium oxysporum and Fusarium nivale.
8. the infrared fingerprint authentication method of fusarium fungus according to claim 2 is characterized in that, said BP artificial nerve network model η be 0.1 and α be 0.6.
9. the infrared fingerprint authentication method of fusarium fungus according to claim 1 is characterized in that, said quantitative comparison result adopts square R of Pearson's related coefficient
2Carry out check analysis with average relative standard deviation R.S.E.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200710040582 CN101059431A (en) | 2007-05-11 | 2007-05-11 | Fusarium fungus infrared fingerprint quick identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200710040582 CN101059431A (en) | 2007-05-11 | 2007-05-11 | Fusarium fungus infrared fingerprint quick identification method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101059431A true CN101059431A (en) | 2007-10-24 |
Family
ID=38865650
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200710040582 Pending CN101059431A (en) | 2007-05-11 | 2007-05-11 | Fusarium fungus infrared fingerprint quick identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101059431A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101968438A (en) * | 2010-09-25 | 2011-02-09 | 西北农林科技大学 | Method for distinguishing water injection of raw material muscles quickly |
| CN102183467A (en) * | 2011-01-24 | 2011-09-14 | 中国科学院长春光学精密机械与物理研究所 | Modeling method for grading quality of Xinjiang red dates in near infrared range |
| CN105474013A (en) * | 2013-08-19 | 2016-04-06 | 西门子医学诊断产品有限责任公司 | Analysis method for supporting classification |
| CN105547927A (en) * | 2015-12-09 | 2016-05-04 | 辽宁工程技术大学 | Method for estimating wetting contact angle of coal dust based on BP artificial neural network |
-
2007
- 2007-05-11 CN CN 200710040582 patent/CN101059431A/en active Pending
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101968438A (en) * | 2010-09-25 | 2011-02-09 | 西北农林科技大学 | Method for distinguishing water injection of raw material muscles quickly |
| CN101968438B (en) * | 2010-09-25 | 2013-03-06 | 西北农林科技大学 | Method for distinguishing water injection of raw material muscles quickly |
| CN102183467A (en) * | 2011-01-24 | 2011-09-14 | 中国科学院长春光学精密机械与物理研究所 | Modeling method for grading quality of Xinjiang red dates in near infrared range |
| CN105474013A (en) * | 2013-08-19 | 2016-04-06 | 西门子医学诊断产品有限责任公司 | Analysis method for supporting classification |
| CN105474013B (en) * | 2013-08-19 | 2018-09-18 | 西门子医学诊断产品有限责任公司 | For the analysis method supported of classifying |
| US10401275B2 (en) | 2013-08-19 | 2019-09-03 | Siemens Healthcare Diagnostics Products Gmbh | Analysis method for supporting classification |
| CN105547927A (en) * | 2015-12-09 | 2016-05-04 | 辽宁工程技术大学 | Method for estimating wetting contact angle of coal dust based on BP artificial neural network |
| CN105547927B (en) * | 2015-12-09 | 2018-04-27 | 辽宁工程技术大学 | A kind of coal dust wetting contact angle evaluation method based on BP artificial neural networks |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109493287B (en) | Deep learning-based quantitative spectral data analysis processing method | |
| Huang et al. | Recent developments in hyperspectral imaging for assessment of food quality and safety | |
| Baek et al. | Rapid measurement of soybean seed viability using kernel-based multispectral image analysis | |
| Dey et al. | Comparative performance of four CNN-based deep learning variants in detecting Hispa pest, two fungal diseases, and NPK deficiency symptoms of rice (Oryza sativa) | |
| CN109001180B (en) | Raman spectrum combined artificial intelligence high-throughput single cell analysis and identification method | |
| Zhao et al. | Early detection of Aspergillus parasiticus infection in maize kernels using near-infrared hyperspectral imaging and multivariate data analysis | |
| Lin et al. | Identification of rice sheath blight through spectral responses using hyperspectral images | |
| Hu et al. | Detecting different pesticide residues on Hami melon surface using hyperspectral imaging combined with 1D-CNN and information fusion | |
| CN105424645A (en) | Method for fast identifying clinical pathogens based on principal component analysis and Fisher discriminance | |
| WO2004057524A1 (en) | Apparatus and method for removing non-discriminatory indices of an indexed dataset | |
| Yin et al. | Non-destructive detection of foreign contaminants in toast bread with near infrared spectroscopy and computer vision techniques | |
| CN101059431A (en) | Fusarium fungus infrared fingerprint quick identification method | |
| Tao et al. | Use of line-scan Raman hyperspectral imaging to identify corn kernels infected with Aspergillus flavus | |
| Jarolmasjed et al. | Hyperspectral imaging and spectrometry-derived spectral features for bitter pit detection in storage apples | |
| Xu et al. | Study on detection method of microplastics in farmland soil based on hyperspectral imaging technology | |
| Wang et al. | Identification of maize with different moldy levels based on catalase activity and data fusion of hyperspectral images | |
| Kumar et al. | A systematic ML based approach for quality analysis of fruits impudent | |
| Zhou et al. | Classification of aflatoxin B1 concentration of single maize kernel based on near-infrared hyperspectral imaging and feature selection | |
| Wang et al. | Detection of Insect-Damaged Maize Seed Using Hyperspectral Imaging and Hybrid 1D-CNN-BiLSTM Model | |
| Jin et al. | A Novel Tran_NAS Method for the Identification of Fe-and Mg-Deficient Pear Leaves from N-and P-Deficient Pear Leaf Data | |
| ALMEIDA et al. | Detecting Three Different Diseases of Plants by Using CNN Model and Image Processing | |
| CN115420703A (en) | Method for identifying pesticide residues on surfaces of Hami melons and identification model construction method | |
| Balytskyi et al. | Raman spectroscopy in open-world learning settings using the objectosphere approach | |
| Guo et al. | Spectral Detection of Peanut Southern Blight Severity Based on Continuous Wavelet Transform and Machine Learning | |
| Moharekar et al. | Detection and classification of plant leaf diseases using convolution neural networks and streamlit |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |