CN110389200A - A kind of tea aroma of the same race differentiation detection method of different brands grade - Google Patents
A kind of tea aroma of the same race differentiation detection method of different brands grade Download PDFInfo
- Publication number
- CN110389200A CN110389200A CN201910637871.7A CN201910637871A CN110389200A CN 110389200 A CN110389200 A CN 110389200A CN 201910637871 A CN201910637871 A CN 201910637871A CN 110389200 A CN110389200 A CN 110389200A
- Authority
- CN
- China
- Prior art keywords
- sample
- tea
- fragrance
- data
- different brands
- 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
- 241001122767 Theaceae Species 0.000 title claims abstract description 69
- 238000001514 detection method Methods 0.000 title claims abstract description 14
- 230000004069 differentiation Effects 0.000 title description 2
- 239000003205 fragrance Substances 0.000 claims abstract description 38
- 238000000034 method Methods 0.000 claims abstract description 23
- 238000004364 calculation method Methods 0.000 claims abstract description 7
- 238000003066 decision tree Methods 0.000 claims abstract description 6
- 239000004615 ingredient Substances 0.000 claims abstract description 3
- 230000035945 sensitivity Effects 0.000 claims abstract description 3
- 238000005070 sampling Methods 0.000 claims description 6
- 238000012549 training Methods 0.000 claims description 6
- 239000012634 fragment Substances 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 4
- 244000062793 Sorghum vulgare Species 0.000 claims description 3
- 238000009835 boiling Methods 0.000 claims description 3
- 239000000284 extract Substances 0.000 claims description 3
- 235000019713 millet Nutrition 0.000 claims description 3
- 238000007637 random forest analysis Methods 0.000 claims description 3
- 238000007789 sealing Methods 0.000 claims description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 2
- 238000012360 testing method Methods 0.000 claims description 2
- 235000019568 aromas Nutrition 0.000 abstract description 2
- 239000002304 perfume Substances 0.000 description 8
- 238000012545 processing Methods 0.000 description 5
- 239000000796 flavoring agent Substances 0.000 description 2
- 235000019634 flavors Nutrition 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- FGUUSXIOTUKUDN-IBGZPJMESA-N C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 Chemical compound C1(=CC=CC=C1)N1C2=C(NC([C@H](C1)NC=1OC(=NN=1)C1=CC=CC=C1)=O)C=CC=C2 FGUUSXIOTUKUDN-IBGZPJMESA-N 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000003754 machining Methods 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 235000013599 spices Nutrition 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0027—General constructional details of gas analysers, e.g. portable test equipment concerning the detector
- G01N33/0031—General constructional details of gas analysers, e.g. portable test equipment concerning the detector comprising two or more sensors, e.g. a sensor array
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0027—General constructional details of gas analysers, e.g. portable test equipment concerning the detector
- G01N33/0036—General constructional details of gas analysers, e.g. portable test equipment concerning the detector specially adapted to detect a particular component
- G01N33/0047—Organic compounds
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0062—General constructional details of gas analysers, e.g. portable test equipment concerning the measuring method or the display, e.g. intermittent measurement or digital display
- G01N33/0068—General constructional details of gas analysers, e.g. portable test equipment concerning the measuring method or the display, e.g. intermittent measurement or digital display using a computer specifically programmed
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Immunology (AREA)
- Combustion & Propulsion (AREA)
- General Health & Medical Sciences (AREA)
- Analytical Chemistry (AREA)
- Medicinal Chemistry (AREA)
- Pathology (AREA)
- Food Science & Technology (AREA)
- Biochemistry (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Tea And Coffee (AREA)
Abstract
The invention discloses a kind of tea aromas of the same race of different brands grade to distinguish detection method, method includes the following steps: the correlation calculations and sensor first with sensor performance filter out 8 sensors for the sensitivity of tea aroma ingredient, building electric nasus system is detected for fragrance.Secondly acquisition different brands and different grades of Longjing tea sample, are detected with tea aroma of the electric nasus system to sample, obtain the fragrance data of sample, counted using fragrance data as initial data, obtain fragrance database.Finally decision-tree model CART is constructed using initial data, carry out Split Attribute using Gini coefficient, the optimal solution that Split Attribute is obtained by the Gini coefficient after division carries out data classification to fragrance database, obtains the tea aroma classification results of the same race of different brands grade.The present invention has higher classification accuracy to the tealeaves of the same race of different brands grade.
Description
Technical field
The present invention relates to data analysis technique field more particularly to a kind of tea aroma data of the same race of different brands grade
Classification method.
Background technique
China belongs to tealeaves big country, and the type and grade of tealeaves are thousands of, and also there is the well-known tea of different brands in each area
With characteristic tea.But being gradually expanded with the market demand, the tea quality of commercial type are irregular.The processing side of tealeaves
There are mainly two types of methods, manual processing and machining.General large automatic tea processing factory, raw material picking and process have bright
True standard.But the tealeaves of manual frying, according to different frying master workers, processing technology more or less has difference.Consumption
Person, due to lacking the relevant professional knowledge of tealeaves, faces numerous and complicated tealeaves brand and grade during buying tealeaves
Classification, consumer are difficult to select suitable tealeaves.For this demand, it is mainly the following solution at present:
The first, engages tealeaves expert manually to distinguish, is commented by the color of tea, fragrance, flavour and the experience of many years
Sentence；
Second, the electric nasus system and taste sensor constituted in conjunction with fragrance sensor constitutes electronic tongue system, to perfume (or spice)
Gas and flavour data are combined analysis, obtain Classification of Tea result.But the higher cost of such methods consumption；
The third, detects tea aroma data using fragrance sensor building electric nasus system, divides fragrance data
Class processing.
Summary of the invention
It is an object of the invention to the problems more difficult to tea grades type classification on the market for consumer, mention
For a kind of tea aroma data classification method of the same race of different brands grade.
The purpose of the present invention is achieved through the following technical solutions: a kind of tea aroma of the same race of different brands grade
Detection method is distinguished, method includes the following steps:
(1) it is filtered out using the correlation calculations of sensor performance and sensor for the sensitivity of tea aroma ingredient
8 sensors, building electric nasus system are detected for fragrance；
(2) different brands and different grades of Longjing tea sample are acquired, with the electric nasus system of step (1) to the tea of sample
Leaf fragrance is detected, and the fragrance data of sample are obtained, and is counted using fragrance data as initial data, is obtained fragrance data
Library；
(3) decision-tree model CART is constructed using initial data, carrys out Split Attribute using Gini coefficient
Wherein D is fragrance database, and c is sample type quantity, piA possibility that being every kind of sample type；If CART is by sample
This type A be considered as can split vertexes, then corresponding child node be DLAnd DR；Then the Gini coefficient after division is
The optimal solution that Split Attribute is obtained by the Gini coefficient after division carries out data classification to fragrance database, obtains
To the tea aroma classification results of the same race of different brands grade.
Further, in the step (1), 8 sensor model numbers of screening are TGS813, TGS822, TGS2602,
TGS2620, TGS2600, MQ-138, MQ-135, MQ-6.
Further, the step (2) specifically:
(2.1) use 7 kinds of different brands and different grades of Longjing tea sample, every kind of sample randomly select 25g be used as to
Test sample sheet；
(2.2) sample in step (2.1) is divided into 5g/ glasss, and is brewed with 250ml boiling water；
(2.3) sealed soak after five minutes, outwells millet paste, filters out tealeaves；
(2.4) by the tea sealing filtered out in step (2.3) in cup, 45 minutes is stood and is cooled, guarantee room temperature 25 ± 1
DEG C, and indoor relative humidity is 80 ± 2%；
(2.5) tea aroma in step (2.4) in cup is drawn into electric nasus system；
(2.6) 8 sensors, 80 seconds data are counted, then each cup there are 640 data, and 22400 data are constituted in total
Fragrance database.
Further, the step (3) specifically:
(3.1) will in total 22400 initial data as fragrance database D；
(3.2) n sample of random sampling with replacement from D, and k stochastical sampling is done, obtain k training set；
(3.3) k categorised decision tree-model CART is constructed using k training set；
(3.4) it is directed to each CART, carrys out Split Attribute using Gini coefficient, the Gini coefficient after being divided；
(3.5) Gini coefficient represents more greatly the uncertain also bigger of sample, therefore the smallest Gini coefficient is division
The optimal solution of attribute；
(3.6) continue Split Attribute until sample is divided into same category according to above-mentioned steps；
(3.7) all CART models constitute a random forest, and the final classification result of tea aroma sample is exactly institute
There are the voting results of CART decision tree.
Further, after obtaining fragrance database, fixed slice parameter t can be used and carry out fragment segmentation, when by being based on
The tea aroma data of domain segment classify to the tea aroma of the same race of different brands grade, the specific steps are as follows:
A. the fragrance data sample of sensor 80s is acquired for calculating；
B. fragment segmentation is carried out using fixed slice parameter t, respectively by the fragrance data of 80s be divided into mono- section of t=10s or
It is mono- section of t=20s；
C. the data slot in step b is re-started into classification and Detection using the method for step (3), and extracts 3 at random
The classification results of time-domain snapshots；
D. classification results average value is the classification results of tea aroma sample.
The beneficial effects of the present invention are: the present invention is by taking Longjing tea as an example, to 7 kinds of different grades of tea aromas of different brands
Data are classified, and the results show classifies success rate higher than 95%.Illustrate the tea fragrant destiny based on electric nasus system
According to, tealeaves can accurately be distinguished, for consumer choose tealeaves have higher reference value.
Detailed description of the invention
The flow chart of the tea aroma data classification method of the same race of Fig. 1 different brands grade of the present invention；
The present invention is based on the tea perfume data dividing methods of time-domain snapshots by Fig. 2.
Specific embodiment
Invention is further described in detail in the following with reference to the drawings and specific embodiments.
A kind of different brands and different grades of tea aroma of the same race distinguish detection method, comprising the following steps:
(1) 8 sensors are filtered out using correlation calculations and sensor sensing degree, building electric nasus system is for perfume
Gas detection, 8 sensor model numbers and sensitive materials filtered out are as shown in table 1；
1 sensor model number of table and sensitive materials table
(2) tea aroma is detected with the electric nasus system of step 1, is with the different grades of Longjing tea of different brands
Example, fragrance data are counted to obtain fragrance database；
(2.1) 7 kinds of different brands and different grades of Longjing tea sample are purchased, as shown in table 2, every kind of sample randomly selects
25g sample to be tested；
2 Longjing tea experiment sample data source table of table
| Brand grade | Price ($/500g) | Brand |
| Xihu Longjing Tea-superfine | 500 | It is bought at the tea grower of Hangzhou native country |
| Xihu Longjing Tea-level-one | 260 | It is bought at the tea grower of Hangzhou native country |
| Xihu Longjing Tea-second level | 100 | It is bought at the tea grower of Hangzhou native country |
| Poly- is in Dragon Well tea-level-one | 850 | Poly- is in Xihu Longjing Tea |
| Poly- is in Dragon Well tea-second level | 336 | Poly- is in Xihu Longjing Tea |
| Pendant Yun Longjing-level-one | 690 | Pendant cloud Xihu Longjing Tea |
| Pendant Yun Longjing-second level | 345 | Pendant cloud Xihu Longjing Tea |
(2.2) sample in step 2.1 is divided into 5g/ glasss, and is brewed with 250ml boiling water；
(2.3) sealed soak after five minutes, outwells millet paste, filters out tealeaves；
(2.4) by the tea sealing filtered out in step 2.3 in cup, standing cools for 45 minutes (guarantees room temperature 25 ± 1
DEG C, and indoor humidity is 80 ± 2% or so)；
(2.5) the tea perfume gas in step 2.4 in cup is drawn into electric nasus system；
(2.6) 8 sensors, 80 seconds data are counted, then each cup there are 640 data (80*8)；
(3) data classification is carried out to fragrance database using machine learning method, it is therefore intended that it is different to distinguish different brands
The Dragon Well tea tea perfume of grade, specific method process are as shown in Figure 1.
(3.1) (7 kinds * 8 sensor * 80 seconds * 5 glasss) a initial data is considered as tea perfume data set D by 22400；
(3.2) n sample of random sampling with replacement from D, and k stochastical sampling is done, obtain k training set；
(3.3) k categorised decision tree-model (CART) is constructed using k training set；
(3.4) it is directed to each CART, carrys out Split Attribute using Gini coefficient
Wherein c is sample type quantity, piA possibility that being every kind of sample type；
(3.5) if CART by sample type A be considered as can split vertexes, corresponding child node be DLAnd DR。
Then the Gini coefficient after division is
(3.6) Gini coefficient represents more greatly the uncertain also bigger of sample, therefore the smallest Gini coefficient is division
The optimal solution of attribute；
(3.7) continue Split Attribute until sample is divided into same category according to above-mentioned steps；
(3.8) all CART models constitute a random forest, and the final classification result of tea aroma sample is exactly institute
There are the voting results of CART decision tree.
Propose a kind of tea perfume data classification simplified calculation method based on time-domain snapshots, the specific steps are as follows:
A. in general, the data that tea aroma needs to acquire 60s or more just can be carried out precise classification, this method acquires 80s
Fragrance data sample for calculating；
B. the tea perfume data classification simplified calculation method based on time-domain snapshots need to provide fixed slice parameter t for segment point
It cuts；
C. the present invention demonstrates t=10,20 two kinds of parameters, respectively by the fragrance data of 80s be divided into mono- section of 10s or
Mono- section of 20s, specific dividing method is as shown in Figure 2；
D. the data slot in step c is re-started into classification and Detection using the method for step (3)；
E. result proves that the result of each segment is approximate with the result that overall data in step (3) is classified；
F. in order to reduce the calculation amount of classification method, and utmostly ensure classification accuracy, the present invention extracts 3 at random
The classification results of time-domain snapshots；
G. the classification results average value in step f is the classification results of tea aroma sample.
The different grades of Longjing tea fragrance of 7 kinds of different brands is done using the method for the present invention and is classified, wherein classifying quality is such as
Shown in table 3.It compared the classifying quality of 3 kinds of other machine learning methods: linear discrimination classification (LDA), multilayer perceptron classification
(MLP), support vector cassification (SVM).It can be found that the method for the present invention accuracy rate is all 95% or more, classifying quality is preferable.
3 the method for the present invention of table and other algorithm classification effect accuracy rate comparison sheets
| RF | LDA | MLP | SVM | |
| Xihu Longjing Tea-superfine | 1.0000 | 0.9375 | 0.7663 | 0.9725 |
| Xihu Longjing Tea-level-one | 0.9975 | 0.9925 | 0.8713 | 0.9925 |
| Xihu Longjing Tea-second level | 1.0000 | 1.0000 | 1.0000 | 1.0000 |
| Poly- is in Dragon Well tea-second level | 1.0000 | 0.8388 | 0.8213 | 0.7550 |
| Poly- is in Dragon Well tea-level-one | 0.9937 | 0.6588 | 0.4500 | 0.3438 |
| Pendant Yun Longjing-second level | 0.9937 | 0.6513 | 0.4613 | 0.4700 |
| Pendant Yun Longjing-level-one | 0.9962 | 0.5413 | 0.3563 | 0.7575 |
| Average value | 0.9973 | 0.8029 | 0.6752 | 0.7559 |
Claims (5)
1. a kind of tea aroma of the same race of different brands grade distinguishes detection method, which is characterized in that this method includes following step
It is rapid:
(1) 8 are filtered out for the sensitivity of tea aroma ingredient using the correlation calculations of sensor performance and sensor
Sensor, building electric nasus system are detected for fragrance.
(2) different brands and different grades of Longjing tea sample are acquired, with the electric nasus system of step (1) to the tea fragrant of sample
Gas is detected, and the fragrance data of sample are obtained, and is counted using fragrance data as initial data, is obtained fragrance database.
(3) decision-tree model CART is constructed using initial data, carrys out Split Attribute using Gini coefficient
Wherein D is fragrance database, and c is sample type quantity, piA possibility that being every kind of sample type；If CART is by sample class
Type A be considered as can split vertexes, then corresponding child node be DLAnd DR；Then the Gini coefficient after division is
The optimal solution that Split Attribute is obtained by the Gini coefficient after division carries out data classification to fragrance database, obtains not
With the tea aroma classification results of the same race of brand grade.
2. a kind of tea aroma of the same race of different brands grade according to claim 1 distinguishes detection method, feature exists
In, in the step (1), 8 sensor model numbers of screening are TGS813, TGS822, TGS2602, TGS2620, TGS2600,
MQ-138, MQ-135, MQ-6.
3. a kind of tea aroma of the same race of different brands grade according to claim 1 distinguishes detection method, feature exists
In the step (2) specifically:
(2.1) 7 kinds of different brands and different grades of Longjing tea sample are used, every kind of sample randomly selects 25g and is used as to test sample
This；
(2.2) sample in step (2.1) is divided into 5g/ glasss, and is brewed with 250ml boiling water；
(2.3) sealed soak after five minutes, outwells millet paste, filters out tealeaves；
(2.4) by the tea sealing filtered out in step (2.3) in cup, stand and cool for 45 minutes, guarantee room temperature at 25 ± 1 DEG C,
And indoor relative humidity is 80 ± 2%；
(2.5) tea aroma in step (2.4) in cup is drawn into electric nasus system；
(2.6) 8 sensors, 80 seconds data are counted, then each cup there are 640 data, and 22400 data constitute fragrance in total
Database.
4. a kind of tea aroma of the same race of different brands grade according to claim 1 distinguishes detection method, feature exists
In the step (3) specifically:
(3.1) will in total 22400 initial data as fragrance database D；
(3.2) n sample of random sampling with replacement from D, and k stochastical sampling is done, obtain k training set；
(3.3) k categorised decision tree-model CART is constructed using k training set；
(3.4) it is directed to each CART, carrys out Split Attribute using Gini coefficient, the Gini coefficient after being divided；
(3.5) Gini coefficient represents more greatly the uncertain also bigger of sample, therefore the smallest Gini coefficient is Split Attribute
Optimal solution；
(3.6) continue Split Attribute until sample is divided into same category according to above-mentioned steps；
(3.7) all CART models constitute a random forest, and the final classification result of tea aroma sample is exactly all
The voting results of CART decision tree.
5. a kind of tea aroma of the same race of different brands grade according to claim 4 distinguishes detection method, feature exists
In fixed slice parameter t can be used and carry out fragment segmentation, pass through the tea fragrant based on time-domain snapshots after obtaining fragrance database
Destiny evidence classifies to the tea aroma of the same race of different brands grade, the specific steps are as follows:
A. the fragrance data sample of sensor 80s is acquired for calculating.
B. fragment segmentation is carried out using fixed slice parameter t, the fragrance data of 80s is divided into mono- section of t=10s or t respectively
Mono- section of=20s.
C. the data slot in step b is re-started into classification and Detection using the method for step (3), and extracts 3 time domains at random
The classification results of segment.
D. classification results average value is the classification results of tea aroma sample.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910637871.7A CN110389200A (en) | 2019-07-15 | 2019-07-15 | A kind of tea aroma of the same race differentiation detection method of different brands grade |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910637871.7A CN110389200A (en) | 2019-07-15 | 2019-07-15 | A kind of tea aroma of the same race differentiation detection method of different brands grade |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110389200A true CN110389200A (en) | 2019-10-29 |
Family
ID=68285067
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910637871.7A Pending CN110389200A (en) | 2019-07-15 | 2019-07-15 | A kind of tea aroma of the same race differentiation detection method of different brands grade |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110389200A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111814846A (en) * | 2020-06-19 | 2020-10-23 | 浙江大华技术股份有限公司 | Training method and recognition method of attribute recognition model and related equipment |
| CN112033911A (en) * | 2020-07-29 | 2020-12-04 | 浙江大学 | Method for rapidly identifying grade of tea based on chromatic aberration and ultraviolet spectrum |
| CN112434646A (en) * | 2020-12-08 | 2021-03-02 | 浙江大学 | Finished tea quality identification method based on transfer learning and computer vision technology |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106529584A (en) * | 2016-10-25 | 2017-03-22 | 福建农林大学 | Flue-cured tobacco aroma type and quality judgment intelligent evaluation method |
-
2019
- 2019-07-15 CN CN201910637871.7A patent/CN110389200A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106529584A (en) * | 2016-10-25 | 2017-03-22 | 福建农林大学 | Flue-cured tobacco aroma type and quality judgment intelligent evaluation method |
Non-Patent Citations (2)
| Title |
|---|
| 常美茁: "茶香气电子鼻系统中气敏传感器阵列筛选方法研究", 《中国优秀硕士学位论文全文数据库 工程科技I辑》 * |
| 蒋云志等编著: "《遥感时间序列分析—以西南退耕还林监测系统为例》", 30 November 2014 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111814846A (en) * | 2020-06-19 | 2020-10-23 | 浙江大华技术股份有限公司 | Training method and recognition method of attribute recognition model and related equipment |
| CN111814846B (en) * | 2020-06-19 | 2023-08-01 | 浙江大华技术股份有限公司 | Training method and recognition method of attribute recognition model and related equipment |
| CN112033911A (en) * | 2020-07-29 | 2020-12-04 | 浙江大学 | Method for rapidly identifying grade of tea based on chromatic aberration and ultraviolet spectrum |
| CN112434646A (en) * | 2020-12-08 | 2021-03-02 | 浙江大学 | Finished tea quality identification method based on transfer learning and computer vision technology |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110389200A (en) | A kind of tea aroma of the same race differentiation detection method of different brands grade | |
| CN101419212B (en) | Bird egg freshness detection device and method by utilizing volatile matter | |
| CN110349153B (en) | Image method tobacco peduncle content detection method based on multi-feature fusion | |
| CN108230085A (en) | A kind of commodity evaluation system and method based on user comment | |
| CN110646416B (en) | Colorimetric card for evaluating smoked chicken quality and application | |
| CN110133050B (en) | Method for qualitatively and quantitatively detecting tea quality based on multi-sensor fingerprint spectrum | |
| Huang et al. | Evaluating quality of tomato during storage using fusion information of computer vision and electronic nose | |
| CN111815573B (en) | Coupling outer wall detection method and system based on deep learning | |
| CN106295802A (en) | A kind of Folium Camelliae sinensis based on particle cluster algorithm Optimized BP Neural Network storage method chronological classification | |
| CN116485418A (en) | Tracing method and system for tea refining production | |
| CN109295159A (en) | Sausage quality Intelligent detecting method | |
| CN111652319A (en) | Cloth defect detection method and device | |
| CN105699397B (en) | A kind of apple surface glossiness detection method | |
| CN107273421B (en) | High-accuracy mode identification and detection method for aroma type and quality of tea | |
| CN106918817B (en) | A kind of preferred method for the gas sensor that can detect tea type and quality | |
| Gu et al. | Surface defect detection of cabbage based on curvature features of 3D point cloud | |
| Macías-Macías et al. | Mask R-CNN for quality control of table olives | |
| CN114493700A (en) | Alliance platform alliance merchant data analysis processing system based on cloud computing | |
| CN113449792A (en) | Method for nondestructive rapid detection of food quality | |
| US20230029413A1 (en) | System and method for non-destructive rapid food profiling using artificial intelligence | |
| KR102495868B1 (en) | Fashion-related customized perfume recommendation system using ai | |
| CN114549485A (en) | Stem detection method based on X-ray vision | |
| Xu et al. | An electronic nose system for the monitoring of water cane shoots quality with swarm clustering algorithm | |
| Rasooli Sharabiani et al. | Recognition and classification of pure and adulterated rice using the electronic nose | |
| Akhtar et al. | Shape and texture based classification of citrus using principal component analysis |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191029 |
|
| RJ01 | Rejection of invention patent application after publication |