CN106127226A - Grain grain and the flexible grain quality detection method of grain grain test sample - Google Patents

Grain grain and the flexible grain quality detection method of grain grain test sample Download PDF

Info

Publication number
CN106127226A
CN106127226A CN201610424371.1A CN201610424371A CN106127226A CN 106127226 A CN106127226 A CN 106127226A CN 201610424371 A CN201610424371 A CN 201610424371A CN 106127226 A CN106127226 A CN 106127226A
Authority
CN
China
Prior art keywords
grain
quality
sample
cloud model
concept
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.)
Granted
Application number
CN201610424371.1A
Other languages
Chinese (zh)
Other versions
CN106127226B (en
Inventor
范艳峰
杨志晓
张庆辉
孙丽君
张翔宇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Henan University of Technology
Original Assignee
Henan University of Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Henan University of Technology filed Critical Henan University of Technology
Priority to CN201610424371.1A priority Critical patent/CN106127226B/en
Publication of CN106127226A publication Critical patent/CN106127226A/en
Application granted granted Critical
Publication of CN106127226B publication Critical patent/CN106127226B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting

Landscapes

  • Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses the flexible grain quality detection method of a kind of grain grain and grain grain test sample, including 1) choose m metric attribute feature；2) if choosing solid food grain training sample, machine vision method is used to extract metric attribute eigenvalue；3) use reverse cloud algorithm, estimate the Multidimensional Cloud Model numerical characteristic of grain sample qualities, set up grain quality Multidimensional Cloud Model；4) to tested grain grain, utilize machine vision to extract metric attribute eigenvalue, calculate degree of certainty μ respectively for K Evolvement of quality concept, determine the Evolvement of quality concept of tested grain grain according to degree of certainty, and define foreign peoples's wheat grain accounting, determine therefrom that the Evolvement of quality concept of grain grain test sample.The present invention represents the ambiguity of grain quality, randomness with the method for expressing of qualitativing concept, grain grain quality carries out flexible classification, and judges that tested sample is the unusualst according to the regularity of distribution of grain quality attribute.

Description

Grain grain and the flexible grain quality detection method of grain grain test sample
Technical field
The present invention relates to artificial intelligence and application, grain quality detection technique field, particularly relate to a kind of grain grain and The flexible grain quality detection method of grain grain test sample.
Background technology
In current grain quality detection, use deterministic quantitative criteria, seed product different in tested grain samples Matter may belong to different quality ranks；Even if same seed, owing to property value measurement result exists certain with chance error Difference, homogeneous result may not belong to different quality rank.Grain quality is carried out just by existing method according to existing quantizating index Property classification, it is impossible to effectively process these uncertain, it is impossible to effectively judge that tested sample is the unusualst.
In recent years, the modem technologies such as embedded, computer vision, artificial intelligence, pattern recognition are utilized, from grain The attributes such as certain physical characteristics, geometric properties, texture are started with, carry out grain quality automatic, quickly detect, obtained increasingly Many concerns.At artificial intelligence field, there has been proposed many sorting techniques, as decision tree, Bayes, artificial neural network, K-neighbour, support vector machine, rule learning, rough set, fuzzy logic, etc..For Optimum Classification device performance, also introduce The optimized algorithm such as ant colony, population.Bayes method, fuzzy logic, D-S evidence theory are also specifically designed to uncertainty and push away Reason.These sorting techniques can obtain the classifying quality of excellence when being applied to specific field.But, the output of these graders is Rigidity.Identical input is obtained well-determined classification results.This is to belong to exact method in itself due to grader.As Probability density function in Bayesian inference, the membership function of fuzzy reasoning, D-S evidence theory degree of belief function once really Fixed, it is all that there is accurately output, repeatably result.And certain methods depends on the subjective experience of people.Obviously these classification sides Method can not process the border uncertain problem of grain quality class.
Grain is typical bulk solid.Grain seed carries abundant information.Even the grain of same kind, it is not No matter belong in geometric shape, color and luster, weight, hardness, water content, each index of quality, various physics, chemical property etc. with seed Property, all there is random difference.This determines grain quality classification in itself and can not impose uniformity without examining individual cases from each attribute dimension division.Bag Including including grain quality, uncertainty is only the dialectical reflection of real world.A grain quality detection inherently classification Problem.I.e. it is divided into tested such as classification (referred to hereinafter as grains such as qualified, defective, qualified X grades by one or more attributes Quality class, or it is called for short quality class).According to clear and definite quantitative criteria, grain qualities different in tested grain samples may adhere to separately In different classifications；Even if same seed, owing to property value measurement result exists certain random error, homogeneous result can not Can belong to different classes of.As can be seen here, there is not boundary clearly between grain quality class.How this processes grain quality Uncertainty in detection？
Summary of the invention
For these problems, the present invention proposes the detection of a kind of flexible grain quality based on grain quality property distribution feature Method, extracts the regularity of distribution of grain quality attribute, represents the ambiguity of grain quality, random with the method for expressing of qualitativing concept Property, grain grain quality is carried out flexible classification, and judges that tested sample is the unusualst according to the regularity of distribution of grain quality attribute.? When sample is normal, there is qualities belonging to the grain grain that quantity is many and degree of certainty is big in tested belonging to.
The present invention is achieved through the following technical solutions:
The flexible grain quality detection method of a kind of grain grain, comprises the steps:
Step 1, chooses m the metric attribute feature based on image of grain grain, sets up m and ties up cloud model, and m is whole more than 1 Number；
Step 2, if choosing solid food grain training sample, every kind of grain grain training sample comprises grain grain quantity >=N, uses machine to regard Feel method is extracted the m of N grain grain grain in training sample and is tieed up metric attribute eigenvalue；
Step 3, to the training sample property value obtained, uses reverse cloud algorithm, estimates the multidimensional cloud of grain sample qualities Model numerical characteristic, sets up K grain quality Multidimensional Cloud Model；
Step 4, to tested grain grain, utilizes machine vision to extract metric attribute eigenvalue, for K Evolvement of quality concept respectively Calculate degree of certainty μ, tested grain grain is belonged to the Evolvement of quality concept making degree of certainty μ obtain maximum.
In described step 1, metric attribute feature includes the grain long and short axial length of grain length axle, axial ratio, texture and color and luster.
In described step 3, reverse cloud algorithm right to use revaluation each sample qualities Multidimensional Cloud Model parameter, sets up multidimensional and adds Power cloud model.
The degree of certainty method calculating grain grain in described step 4 is:
The property value of note jth grain grain is xj={ xji| i=1,2 ..., m}, j=1,2 ..., n；To K quality cloud mould Type, is utilized respectively formula (1) and K degree of certainty μ (x of formula (2) calculating grain grainj)={ μ (xjk) | k=1,2 ..., K}:
μ ( x j k ) = e - Σ i = 1 m ( x j i - Ex k i ) 2 2 En k i ′ 2 - - - ( 1 )
Enki'=Norm (Enki, Heki 2) (2)
Norm represents normal distribution, Exki, EnkiAnd HekiIt is respectively the phase of kth quality cloud model ith attribute feature Prestige, entropy and super entropy.
The flexible grain quality detection method of a kind of grain grain test sample, comprises the steps:
Step 1, chooses m the metric attribute feature based on image of grain grain, sets up m and ties up cloud model, and m is whole more than 1 Number；
Step 2, if choosing solid food grain training sample, every kind of grain grain training sample comprises grain grain quantity >=N, uses machine to regard Feel method is extracted the m of N grain grain grain in training sample and is tieed up metric attribute eigenvalue；
Step 3, to the training sample property value obtained, uses reverse cloud algorithm, estimates the multidimensional cloud of grain sample qualities Model numerical characteristic, sets up K grain quality Multidimensional Cloud Model；
Step 4, if the grain grain quantity of test sample is n, utilizes machine vision to carry each grain grain of test sample respectively Take metric attribute eigenvalue, calculate degree of certainty μ respectively for K Evolvement of quality concept, belong to each grain grain make μ obtain maximum Evolvement of quality concept；
Step 5, calculates foreign peoples's grain grain accounting r according to formula (3),
R=(n-nmax)/n (3), n in formulamaxThe grain grain quantity most for belonging to same Evolvement of quality concept；
Step 6, sets threshold value r0, as r≤r0Time, it is believed that sample is normal, and test sample is belonged to nmaxIndividual grain grain is returned The Evolvement of quality concept belonged to；Otherwise it is assumed that sample is unusual, refusal judges its quality.
In described step 1, metric attribute feature includes the grain long and short axial length of grain length axle, axial ratio, texture and color and luster.
In described step 3, reverse cloud algorithm right to use revaluation each sample qualities Multidimensional Cloud Model parameter, sets up multidimensional and adds Power cloud model.
The present invention compared with prior art, has a following obvious advantage:
(1) grain quality is considered as qualitativing concept, between each quality, need not necessarily exist boundary clearly
Owing to grain quality is comprehensively embodied by a large amount of grain grains, and metric attribute has random difference in the realization of each grain grain Different.The method defining grain quality according to set quantizating index rigidity accordingly, with respect to tradition, is regarded as having uncertainty Qualitativing concept more particularly suitable.Grain quality is considered as qualitativing concept by the present invention, estimates grain quality concept according to grain grain sample Uncertain feature, use cloud model to portray its ambiguity, randomness, use grain grain quality degree of certainty instruction grain grain quality The degree of grain quality can be represented.Fine limit, the attribute of energy objective description grain quality is there is not between grain quality concept Distribution characteristics.
(2) grain grain attribute classification utensil has flexible output
Grain grain quality affiliation method is designed based on grain grain quality degree of certainty.Belong to grain grain quality cloud space makes it take Obtain the maximum grain quality determining angle value.Owing to degree of certainty is the random number with steady tendency, therefore its classification results is not Deterministic.Even if to same grain grain, homogeneous does not uses water dust ownership algorithm, will obtain different classification results.Grain grain depends on Probability belongs to several concepts in grain quality space.Therefore, the output of grain grain attribute classification device has flexibility.It is especially suitable for Process the quality value of borderline region.This feature has no effect on the result of determination of sample overall quality.On the contrary, if borderline region Foreign peoples's quantity increases, and can differentiate the singularity of sample on the contrary, plays the effect strengthening grain quality discriminant information.
Accompanying drawing explanation
Fig. 1 is the flexible grain quality detection method schematic diagram of the present invention.
Fig. 2 is the wheat quality cloud exemplary plot of the present invention.
Fig. 3 is the water dust ownership probability distribution result plane exemplary plot of the present invention.
Detailed description of the invention
Fig. 1 is the flexibility qualities sorting technique population structure schematic diagram in the grain quality detection that the present invention announces.Mainly Detailed description is as follows.
(1) grain quality is considered as qualitativing concept.
Grain quality property value has diversity in the distribution of each grain grain.Different grain grains may belong to different quality, The not homogeneous testing result of same grain grain is likely to be divided into different quality.Fine limit is there is not, at this between grain quality Qualitativing concept is belonged in matter.Its uncertainty includes ambiguity and randomness.Difference between each grain grain metric attribute value is the least, Then Evolvement of quality concept is the most clear；Otherwise it is the fuzzyyest.Randomness features the distribution in domain of the grain grain metric attribute.
(2) the basic research instrument with normal cloud model as qualitativing concept
The ambiguity of qualitativing concept can be described by cloud model with randomness unified quantization.Grain quality can be considered qualitative generally Read.Grain grain quality is the Stochastic implementation of grain quality, can be considered water dust.Therefore, represent that grain quality is most suitable not with cloud model Cross.The parameter of cloud model C (Ex, En, He) estimates have objectivity from sample.
(3) using attributes such as grain grain geometric properties, texture, color and lusters as metric attribute collection
With the grain long and short axial length of grain length axle, axial ratio, texture, color and luster, as grain grain quality candidate attribute collection.
(4) choose grain grain sample, use machine vision method, respectively extraction attribute character value
As a example by Semen Tritici aestivi.Choose more than 10 kinds common wheat breeds.Each sample wheat grain is not less than 500, as training Sample.Machine vision technique means are utilized to obtain each sample grain grain property value respectively.
(5) use reverse cloud algorithm, estimate the Multidimensional Cloud Model numerical characteristic of grain sample qualities, set up grain quality many Dimension normal cloud model
To the training sample property value obtained, use reverse cloud algorithm respectively, and consider whether to use weight, estimate various kinds This quality Multidimensional Cloud Model parameter, sets up their Multidimensional Cloud Model or multidimensional weighting cloud model.
(6) design is based on a determination that the grain grain quality flexibility ownership algorithm of degree μ, investigates grain grain and belongs to the probability of different quality Feature
Water dust degree of certainty μ is the random number having steady tendency.Utilize this feature can realize the flexible classification of grain grain quality.
Algorithm design basic thought: the grain grain to test sample, utilizes machine vision to extract metric attribute eigenvalue, respectively Calculate they associating degree of certainty μ in each attribute dimension in all Evolvement of quality concept spaces, it is belonged to the product making μ obtain maximum Matter concept.Due to the randomness of μ, classification results has uncertainty.Water dust belongs to multiple concept according to probability.
Whole grain grains to any test sample, repeat to use algorithm, investigate grain grain and belong to the probability distribution of each quality Feature.Fig. 3 is a water dust ownership probability distribution results exemplary plot along L-S=0.7mm planar interception.In Fig. 3, P is for estimating The probability belonging to certain cloud model, L is the long axial length of wheat grain (unit: millimeter), and S is the short axial length of wheat grain (unit: millimeter). C1, C2, C3 respectively corresponding C{ (1.8,0.05,0.02), (1.1,0.05,0.02) }, C{ (2.0,0.05,0.02), (1.3, 0.05,0.02)}、C{(2.2,0.05,0.02),(1.5,0.05,0.02)}。
(7) based on the grain grain metric attribute regularity of distribution, the overall quality fusion method of design test sample
According to above-mentioned sorting technique, the grain grain of borderline region can belong to multiple Evolvement of quality concept.Divide according to grain grain quality Cloth rule, further away from the expectation of Evolvement of quality concept, the degree of certainty that quality value can represent this Evolvement of quality concept is the lowest.Its probability occurred is also Ying Yue little.According to this principle, the singularity method of discrimination of design grain grain attribute classification.To normal sample, its quality is belonged to Quality belonging to and grain grain that degree of certainty is big many in occurrence number.
Such as, it is detected as object of study with wheat quality.
(1) quantization means of wheat quality
Remember that wheat quality attribute character based on image has m, cloud model C (X can be tieed up with m1, X2..., Xm) represent little Wheat quality.That is: C{ (Exi, Eni, Hei) | i=1,2 ..., m}, wherein Exi, Eni, HeiEstimation self-training sample, respectively i-th The expectation of individual attribute character, entropy and super entropy.Choose the sample of different quality, corresponding wheat quality cloud model can be set up.Assuming that The wheat quality cloud model set up is K.Then K wheat quality cloud is designated as Ck{(Exki, Enki, Heki) | i=1,2 ..., m}, k =1,2 ..., K.Wheat quality cloud example is as in figure 2 it is shown, in Fig. 2, μ is degree of certainty, and L is the long axial length of wheat grain (unit: millimeter), S For the short axial length of wheat grain (unit: millimeter).The parameter of three wheat grain major axis length axial length two-dimension cloud model be respectively C{ (1.8, 0.05,0.02),(1.1,0.05,0.02)}、C{(2.0,0.05,0.02),(1.3,0.05,0.02)}、C{(2.2,0.05, 0.02), (1.5,0.05,0.02) }, there is not fine limit between different qualities as can be seen from Figure 2.
(2) the flexible ownership algorithm of wheat grain quality
N grain wheat grain to tested sample, utilizes machine vision to extract the m dimension attribute eigenvalue of every wheat grain.Jth grain wheat grain Property value be designated as xj={ xji| i=1,2 ..., m}, j=1,2 ..., n.To K quality cloud model, it is utilized respectively following formula and calculates xiBelong to their degree of certainty:
μ ( x j k ) = e - Σ i = 1 m ( x j i - Ex k i ) 2 2 En k i ′ 2 - - - ( 1 )
Enki'=Norm (Enki, Heki 2) (2)
In above formula, Norm represents normal distribution.Available K degree of certainty μ (xj)={ μ (xjk) | k=1,2 ..., K}.If μI =max{ μ (xjk) | k=1,2 ..., K}, it is the maximum in K degree of certainty, then by water dust xjBelonging to quality cloud I, I is K The subscript k value of the maximum degree of certainty in individual degree of certainty.
From formula (1) and (2) it can be seen that due to Enki' meet normal distribution, μ (xjk) also it is random.Therefore to water dust xj, each result of calculation may make it belong to different Evolvement of quality concepts.But μ (xjk) there is steady tendency.Thus, water dust xj Some wheat quality cloud concepts will be belonged to according to probability, and be no longer rigidity be divided into certain quality grade.Wheat grain quality Belong to the probability distribution example such as figure (3) of different quality cloud.
(3) tested sample singularity method of discrimination
Above-mentioned wheat grain can belong to the flexible sorting algorithm of different Evolvement of quality concept, it appears that the overall quality for tested sample is sentenced Surely conflict is caused.It practice, this method not only will not cause contradiction, the effect of information enhancement can be played on the contrary.Can from Fig. 2 Going out, away from desired quality value, the degree of certainty that it can represent this quality is the least, and the number of times occurred is the lowest.According to this rule, May determine that sample is the unusualst.If the wheat grain quantity belonging to different quality is close, then illustrate that this sample is unusual.And Only belong to the wheat grain quantity of a certain quality much larger than when belonging to the wheat grain quantity of foreign peoples's quality, could illustrate that sample is just Normal.Principle accordingly, to there being the sample of n wheat grain, defines foreign peoples's wheat grain accounting
R=(n-nmax)/n (3)
N in formulamaxThe wheat grain quantity most for belonging to same Evolvement of quality concept.Set threshold value r0.As r≤r0Time, it is believed that sample This is normal.Otherwise it is assumed that sample is unusual.
(4) tested Semen Tritici aestivi sample qualities judges
To unusual sample, refusal judges its quality.When sample is normal, it is n by its quality judgingmaxIndividual wheat grain institute is common That Evolvement of quality concept of ownership.
Each attribute of grain quality shows as the comprehensive of a large amount of grain grain respective attributes.Seeing in turn, each seed is this One Stochastic implementation of grain, the property value of each seed is a Stochastic implementation of its grain quality respective attributes.Even if Being the kind that integrity attribute is close, the property value distribution characteristics that it realizes on each seed may be different.Due to grain quality At the distributional difference of each grain grain, it has ambiguity and random nature.Ambiguity characterizes metric attribute value in domain Intensity.Grain grain property value is more concentrated, and overall quality is the most clear；Otherwise it is the fuzzyyest.Randomness then characterizes each grain The difference degree of grain metric attribute value.Therefore, grain quality belongs to qualitativing concept in itself.From grain attribute at each seed Distribution characteristics start with, process the uncertain problem in grain quality detection according to the expression of qualitativing concept and sorting technique, More reasonably result can be obtained.
Grain quality determines it in the diversity that each grain grain realizes and belongs to qualitativing concept.The metric attribute of each grain grain Value is a Stochastic implementation of grain quality respective attributes.Its degree of certainty μ indicates it can represent this overall quality attribute Degree.Due to steady tendency and the randomness of μ, relate to grader based on it, it is possible to the flexibility realizing grain quality classification is drawn Point.
Technological means disclosed in the present invention program is not limited only to the technological means disclosed in above-mentioned embodiment, also includes The technical scheme being made up of above technical characteristic combination in any.It should be pointed out that, for those skilled in the art For, under the premise without departing from the principles of the invention, it is also possible to make some improvements and modifications, these improvements and modifications are also considered as Protection scope of the present invention.

Claims (7)

1. the flexible grain quality detection method of a grain grain, it is characterised in that: comprise the steps:
Step 1, chooses m metric attribute feature based on image of grain grain, sets up m and tie up cloud model, and m is the integer more than 1；
Step 2, if choosing solid food grain training sample, every kind of grain grain training sample comprises grain grain quantity >=N, uses machine vision side Method is extracted the m of N grain grain grain in training sample and is tieed up metric attribute eigenvalue；
Step 3, to the training sample property value obtained, uses reverse cloud algorithm, estimates the Multidimensional Cloud Model of grain sample qualities Numerical characteristic, sets up K grain quality Multidimensional Cloud Model；
Step 4, to tested grain grain, utilizes machine vision to extract metric attribute eigenvalue, calculates respectively for K Evolvement of quality concept Degree of certainty μ, belongs to the Evolvement of quality concept making degree of certainty μ obtain maximum by tested grain grain.
The flexible grain quality detection method of grain grain the most according to claim 1, it is characterised in that: product in described step 1 Matter attribute character includes the grain long and short axial length of grain length axle, axial ratio, texture and color and luster.
The flexible grain quality detection method of grain grain the most according to claim 1, it is characterised in that: inverse in described step 3 To cloud algorithm right to use revaluation each sample qualities Multidimensional Cloud Model parameter, set up multidimensional weighting cloud model.
The flexible grain quality detection method of grain grain the most according to claim 1, it is characterised in that: described step 4 is fallen into a trap The degree of certainty method calculating grain grain is:
The property value of note jth grain grain is xj={ xji| i=1,2 ..., m}, j=1,2 ..., n；To K quality cloud model, point Formula (1) and formula (2) is not utilized to calculate K degree of certainty μ (x of grain grainj)={ μ (xjk) | k=1,2 ..., K}:
μ ( x j k ) = e - Σ i = 1 m ( x j i - Ex k i ) 2 2 En k i ′ 2 - - - ( 1 )
Enki'=Norm (Enki, Heki 2) (2)
Norm represents normal distribution, Exki, EnkiAnd HekiIt is respectively the expectation of kth quality cloud model ith attribute feature, entropy With super entropy.
5. the flexible grain quality detection method of a grain grain test sample, it is characterised in that: comprise the steps:
Step 1, chooses m metric attribute feature based on image of grain grain, sets up m and tie up cloud model, and m is the integer more than 1；
Step 2, if choosing solid food grain training sample, every kind of grain grain training sample comprises grain grain quantity >=N, uses machine vision side Method is extracted the m of N grain grain grain in training sample and is tieed up metric attribute eigenvalue；
Step 3, to the training sample property value obtained, uses reverse cloud algorithm, estimates the Multidimensional Cloud Model of grain sample qualities Numerical characteristic, sets up K grain quality Multidimensional Cloud Model；
Step 4, if the grain grain quantity of test sample is n, each grain grain to test sample utilizes machine vision to extract product respectively Matter attribute character value, calculates degree of certainty μ respectively for K Evolvement of quality concept, and each grain grain belongs to the product making μ obtain maximum Matter concept；
Step 5, calculates foreign peoples's grain grain accounting r according to formula (3),
R=(n-nmax)/n (3), n in formulamaxThe grain grain quantity most for belonging to same Evolvement of quality concept；
Step 6, sets threshold value r0, as r≤r0Time, it is believed that sample is normal, and test sample is belonged to nmaxIndividual grain grain is belonged to Evolvement of quality concept；Otherwise it is assumed that sample is unusual, refusal judges its quality.
The flexible grain quality detection method of grain grain test sample the most according to claim 5, it is characterised in that: described step In rapid 1, metric attribute feature includes the grain long and short axial length of grain length axle, axial ratio, texture and color and luster.
The flexible grain quality detection method of grain grain test sample the most according to claim 5, it is characterised in that: described step Reverse cloud algorithm right to use revaluation each sample qualities Multidimensional Cloud Model parameter in rapid 3, sets up multidimensional weighting cloud model.
CN201610424371.1A 2016-06-14 2016-06-14 The flexible grain quality detection method of grain grain and grain grain test sample Expired - Fee Related CN106127226B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610424371.1A CN106127226B (en) 2016-06-14 2016-06-14 The flexible grain quality detection method of grain grain and grain grain test sample

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610424371.1A CN106127226B (en) 2016-06-14 2016-06-14 The flexible grain quality detection method of grain grain and grain grain test sample

Publications (2)

Publication Number Publication Date
CN106127226A true CN106127226A (en) 2016-11-16
CN106127226B CN106127226B (en) 2019-09-03

Family

ID=57469438

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610424371.1A Expired - Fee Related CN106127226B (en) 2016-06-14 2016-06-14 The flexible grain quality detection method of grain grain and grain grain test sample

Country Status (1)

Country Link
CN (1) CN106127226B (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106768077A (en) * 2017-02-18 2017-05-31 塔里木大学 A kind of jujube quality detection system
CN108171432A (en) * 2018-01-04 2018-06-15 南京大学 Ecological risk evaluating method based on Multidimensional Cloud Model-fuzzy support vector machine
CN108241290A (en) * 2017-12-19 2018-07-03 中国农业生产资料流通协会 A kind of warehouse recirculation fumigation control method based on fuzzy-adaptation PID control
CN110879955A (en) * 2018-09-06 2020-03-13 埃森哲环球解决方案有限公司 Digital quality control using computer vision with deep learning

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3763818B2 (en) * 2003-01-06 2006-04-05 株式会社クボタ Grain inspection equipment
CN101251526A (en) * 2008-02-26 2008-08-27 浙江大学 Method and apparatus for nondestructively testing food synthetic quality
CN102589470A (en) * 2012-02-14 2012-07-18 大闽食品（漳州）有限公司 Fuzzy-neural-network-based tea leaf appearance quality quantification method
CN103489006A (en) * 2013-10-11 2014-01-01 河南城建学院 Computer vision-based rice disease, pest and weed diagnostic method
CN104155312A (en) * 2014-08-11 2014-11-19 华北水利水电大学 Method for detecting pests in food grains based on near infrared machine vision, and apparatus thereof
CN104535646A (en) * 2014-12-17 2015-04-22 河南工业大学 Method for detecting imperfection of food grains
CN104749070A (en) * 2015-04-15 2015-07-01 河南工业大学 Method for rapidly evaluating wheat quality

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3763818B2 (en) * 2003-01-06 2006-04-05 株式会社クボタ Grain inspection equipment
CN101251526A (en) * 2008-02-26 2008-08-27 浙江大学 Method and apparatus for nondestructively testing food synthetic quality
CN102589470A (en) * 2012-02-14 2012-07-18 大闽食品（漳州）有限公司 Fuzzy-neural-network-based tea leaf appearance quality quantification method
CN103489006A (en) * 2013-10-11 2014-01-01 河南城建学院 Computer vision-based rice disease, pest and weed diagnostic method
CN104155312A (en) * 2014-08-11 2014-11-19 华北水利水电大学 Method for detecting pests in food grains based on near infrared machine vision, and apparatus thereof
CN104535646A (en) * 2014-12-17 2015-04-22 河南工业大学 Method for detecting imperfection of food grains
CN104749070A (en) * 2015-04-15 2015-07-01 河南工业大学 Method for rapidly evaluating wheat quality

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
范艳峰 等: "基于云理论的粮食品质检测不确定性研究", 《河南工业大学学报(自然科学版)》 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106768077A (en) * 2017-02-18 2017-05-31 塔里木大学 A kind of jujube quality detection system
CN108241290A (en) * 2017-12-19 2018-07-03 中国农业生产资料流通协会 A kind of warehouse recirculation fumigation control method based on fuzzy-adaptation PID control
CN108171432A (en) * 2018-01-04 2018-06-15 南京大学 Ecological risk evaluating method based on Multidimensional Cloud Model-fuzzy support vector machine
CN110879955A (en) * 2018-09-06 2020-03-13 埃森哲环球解决方案有限公司 Digital quality control using computer vision with deep learning
CN110879955B (en) * 2018-09-06 2024-05-14 埃森哲环球解决方案有限公司 Digital quality control using computer vision with deep learning

Also Published As

Publication number Publication date
CN106127226B (en) 2019-09-03

Similar Documents

Publication Publication Date Title
Rostami et al. A fast hypervolume driven selection mechanism for many-objective optimisation problems
Tesfahun et al. Intrusion detection using random forests classifier with SMOTE and feature reduction
CN106127226A (en) Grain grain and the flexible grain quality detection method of grain grain test sample
CN110135167B (en) Edge computing terminal security level evaluation method for random forest
CN111835707B (en) Malicious program identification method based on improved support vector machine
CN108229550A (en) A kind of cloud atlas sorting technique that network of forests network is cascaded based on more granularities
CN108304316A (en) A kind of Software Defects Predict Methods based on collaboration migration
Grzymala-Busse et al. Handling missing attribute values in preterm birth data sets
CN101251896B (en) Object detecting system and method based on multiple classifiers
CN109948726A (en) A kind of Power Quality Disturbance Classification Method based on depth forest
CN117174313B (en) Method and system for establishing cerebral hemorrhage patient neural function prognosis prediction model
CN111079283A (en) Method for processing information saturation unbalanced data
CN108171432B (en) Ecological risk evaluation method based on multi-dimensional cloud model-fuzzy support vector machine
CN117197591B (en) Data classification method based on machine learning
Abu-Nimeh et al. Bayesian additive regression trees-based spam detection for enhanced email privacy
Abolkarlou et al. Ensemble imbalance classification: Using data preprocessing, clustering algorithm and genetic algorithm
Lobel et al. Preliminary results on social learning with partial observations
Thanh et al. An approach to reduce data dimension in building effective network intrusion detection systems
CN105975994A (en) SAR target discrimination method based on non-similarity transformation A-type SVM model
Li et al. Study on the Prediction of Imbalanced Bank Customer Churn Based on Generative Adversarial Network
Kong et al. Popularity prediction based on interactions of online contents
Li et al. Social network change detection using a genetic algorithm based back propagation neural network model
Cattelan et al. On the performance of uncertainty estimation methods for deep-learning based image classification models
Liu et al. A united classification system of X-ray image based on fuzzy rule and neural networks
Zhang et al. The nearest neighbor algorithm of filling missing data based on cluster analysis

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20190903

Termination date: 20200614