CN108444924B - Method for discriminating storage period of tea by applying hyperspectral image technology - Google Patents
Method for discriminating storage period of tea by applying hyperspectral image technology Download PDFInfo
- Publication number
- CN108444924B CN108444924B CN201810211952.6A CN201810211952A CN108444924B CN 108444924 B CN108444924 B CN 108444924B CN 201810211952 A CN201810211952 A CN 201810211952A CN 108444924 B CN108444924 B CN 108444924B
- Authority
- CN
- China
- Prior art keywords
- tea
- samples
- sample
- hyperspectral
- image
- 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.)
- Active
Links
- 238000003860 storage Methods 0.000 title claims abstract description 51
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000005516 engineering process Methods 0.000 title claims abstract description 21
- 241001122767 Theaceae Species 0.000 title abstract 3
- 238000012706 support-vector machine Methods 0.000 claims abstract description 12
- 244000269722 Thea sinensis Species 0.000 claims description 184
- 235000013616 tea Nutrition 0.000 claims description 182
- 238000012937 correction Methods 0.000 claims description 23
- 238000012360 testing method Methods 0.000 claims description 23
- 238000005070 sampling Methods 0.000 claims description 13
- 230000003595 spectral effect Effects 0.000 claims description 13
- 230000032683 aging Effects 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 7
- 238000001228 spectrum Methods 0.000 claims description 7
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 claims description 6
- 229910052782 aluminium Inorganic materials 0.000 claims description 6
- 239000011888 foil Substances 0.000 claims description 6
- 238000005303 weighing Methods 0.000 claims description 5
- 238000000605 extraction Methods 0.000 claims description 4
- 238000000513 principal component analysis Methods 0.000 claims description 4
- 238000005057 refrigeration Methods 0.000 claims description 4
- 238000007789 sealing Methods 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000007689 inspection Methods 0.000 claims description 3
- 235000006468 Thea sinensis Nutrition 0.000 claims description 2
- 235000020279 black tea Nutrition 0.000 claims description 2
- 238000012850 discrimination method Methods 0.000 claims description 2
- 239000011159 matrix material Substances 0.000 claims description 2
- 238000004451 qualitative analysis Methods 0.000 abstract description 4
- 238000001514 detection method Methods 0.000 description 4
- 238000010191 image analysis Methods 0.000 description 4
- 238000013441 quality evaluation Methods 0.000 description 3
- 239000000126 substance Substances 0.000 description 3
- 239000009137 wuling Substances 0.000 description 3
- 244000241257 Cucumis melo Species 0.000 description 2
- 235000015510 Cucumis melo subsp melo Nutrition 0.000 description 2
- FJJCIZWZNKZHII-UHFFFAOYSA-N [4,6-bis(cyanoamino)-1,3,5-triazin-2-yl]cyanamide Chemical compound N#CNC1=NC(NC#N)=NC(NC#N)=N1 FJJCIZWZNKZHII-UHFFFAOYSA-N 0.000 description 2
- 238000000701 chemical imaging Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000009659 non-destructive testing Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 235000009024 Ceanothus sanguineus Nutrition 0.000 description 1
- 240000003553 Leptospermum scoparium Species 0.000 description 1
- 235000015459 Lycium barbarum Nutrition 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 238000012271 agricultural production Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000009614 chemical analysis method Methods 0.000 description 1
- 239000003153 chemical reaction reagent Substances 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000035622 drinking Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 239000003205 fragrance Substances 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000002290 gas chromatography-mass spectrometry Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008092 positive effect Effects 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 235000014347 soups Nutrition 0.000 description 1
- 238000012306 spectroscopic technique Methods 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
- 239000013077 target material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/3563—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for analysing solids; Preparation of samples therefor
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/359—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light using near infrared light
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2201/00—Features of devices classified in G01N21/00
- G01N2201/12—Circuits of general importance; Signal processing
- G01N2201/129—Using chemometrical methods
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
Abstract
The invention discloses a method for judging the storage period of tea by applying a hyperspectral image technology, which is used for establishing a qualitative analysis model by combining the hyperspectral image technology with a support vector machine algorithm to realize the rapid and accurate judgment of the storage period of the tea.
Description
Technical Field
The invention relates to a method for judging the storage period of tea, in particular to a method for judging the storage period of tea by applying a hyperspectral image technology.
Background
The planting area and the tea yield of a tea garden in China are the first in the world, and tea is a traditional bulk export product in China. However, with the change of supply and demand conditions in international tea trade, the growth trend of tea export amount and trade volume in China falls back to some extent, the demand of domestic markets is close to saturation, the total production amount of tea far exceeds the market demand, and the phenomenon of tea lost is caused. In the process of drinking tea, the quality of fresh tea is better than that of old tea, because the appearance quality and taste substances of the tea are irreversibly changed along with the increase of storage time. The new tea has good quality and high price, and the new tea is just when the old tea leaves the market. In order to make profit, part of vendors use old tea to impersonate new tea, or mix old tea with new tea for sale. The inferior poor competition means not only affects the market order and endangers the tea brand, but also threatens the food safety by additives such as coloring agents used in the processing. In the process of purchasing, consumers can only judge the new tea and the old tea through the sense (looking at color, smelling fragrance and tasting tea soup) generally, and effective judgment is difficult to realize. Because of the lack of related methods and standards for tea storage period detection, it is particularly urgent to research a scientific, accurate, simple, convenient and quick tea storage period discrimination method.
At present, the tea quality evaluation is mainly carried out on sensory evaluation from two aspects of appearance and internal quality, and the evaluation result is easily influenced by subjective factors of tea evaluators. The chemical analysis methods such as gas chromatography-mass spectrometry, electronic nose and electronic tongue have expensive instruments and complex operation methods, and can not realize rapid and convenient discrimination of the storage period of the tea. Therefore, nondestructive testing means such as spectroscopic techniques and image analysis techniques have been drawing attention, but the ability of a single testing means to evaluate the quality of tea leaves is very limited. The Hyperspectral imaging (HSI) technology images a target material in a large number of subdivided spectral bands in visible and near-infrared regions of an electromagnetic spectrum, can simultaneously extract image characteristic parameters and more spectral characteristic parameters, and more accurately and comprehensively characterize a measured substance. As a green, rapid and efficient nondestructive testing technology, the hyperspectral image technology is the fusion of a spectrum technology and an image processing technology, has both spectrum analysis capability and image resolution capability, simultaneously tests the internal and external quality of tea, effectively improves the reliability and stability of the tea quality evaluation result, and can meet the quality evaluation requirement of scientific metering. In recent years, hyperspectral image technology is widely applied to agricultural production, and the application of hyperspectral image technology to tea production mainly focuses on tea tree cultivation management monitoring, tea grade division and tea identification. The hyperspectral image technology is applied to the judgment of the storage period of the tea, the image and the model of the tea sample to be detected are called, and compared with a chemical detection method, the hyperspectral image technology has the advantages of simplicity, convenience, accuracy and the like, can achieve the aim of scientifically and accurately judging the storage period of the tea, makes up the vacancy of related research fields, and has a positive effect on striking the phenomenon that old tea in the tea market is counterfeited with new tea and standardizing the market order.
Disclosure of Invention
The invention aims to provide a method for judging the storage period of tea, aiming at the current situation that old tea serves as new tea in the tea sale of part of enterprises at present and the conventional identification method is not easy to judge the storage period of the tea.
The invention is realized by the following technical scheme:
a method for discriminating the storage period of tea by applying a hyperspectral image technology is characterized by comprising the following steps:
(1) preparing an aged tea sample: after the initial moisture content of the purchased tea is measured, subpackaging the tea in aluminum foil bags and sealing the bags, and storing the tea under two conditions; storing a part of tea samples in a refrigeration house as fresh-keeping tea samples, setting the aging time as one year, respectively sampling at 0, 90, 180, 270 and 360 d, and taking five groups of tea samples, wherein each group of tea samples takes 30 samples, each sample takes 30 g, and 150 samples are obtained; storing the other part of tea samples at normal temperature, setting the aging time as one year, respectively sampling at 0, 90, 180, 270 and 360 d, and taking five groups of tea samples, wherein each group of tea samples takes 30 samples, and each sample takes 30 g to obtain 150 samples;
(2) collecting hyperspectral images, selecting a sample correction set and a test set: collecting hyperspectral images of samples by using a hyperspectral image system based on a visible/near-infrared spectrometer and matched collection and analysis software, wherein the scanning range of a visible/near-infrared camera is 370.38-1036.54 nm, the sampling interval is 0.55 nm, each spectrum has 1232 variables, scanning the tea samples in different storage periods prepared in the step (1) after setting the exposure time of the hyperspectral camera and a conveying device, scanning a standard white correction plate and a black correction image before collecting the images of the samples, correcting the images of the samples by using the black and white correction image to obtain the hyperspectral images of the tea samples in different storage periods, and selecting one sample as a test sample every two samples, namely 50 samples in 150 samples as an inspection set and the other 100 samples as correction sets;
(3) dimension reduction processing and characteristic value extraction: performing dimensionality reduction processing on original hyperspectral data by utilizing principal component analysis, extracting five characteristic wavelengths based on weight coefficients of first three principal component images, selecting the center of a 100 x 100 pixel region at the center of the image as an interested Region (ROI), extracting and storing spectral data of the ROI region of 5 characteristic wavelength images as spectral characteristic information, calculating the texture of the whole image of the tiled tea based on a Gray-level co-occurrence matrix (GLCM) according to the characteristics of the tea image, and extracting texture characteristic variables under the characteristic wavelengths as texture characteristic information;
(4) two models were established: selecting optimal parameters, and respectively establishing storage period discrimination models of a stored tea sample and a fresh-keeping tea sample based on a Support Vector Machine (SVM);
(5) and (3) testing the model: using the test set in the step (2) as an unknown sample to test the calibration set model;
(6) taking tea samples in an unknown storage period, performing hyperspectral collection according to the step (2), and analyzing hyperspectral images of the tea samples in the unknown storage period by utilizing the established SVM model to quickly and qualitatively judge the storage period of the tea.
When the hyperspectral image is collected in the step (2), 11.5-12.5g of tea leaves are weighed and evenly and flatly paved in a culture dish with the specification of phi 9 cm.
Before the hyperspectral image data are collected in the step (2), the collection parameters of the visible/near-infrared camera are set to be exposure time of 8.5 ms, and the speed of the conveying device is 1.15 mm/s.
The five characteristic wavelengths screened by adopting the main component analysis in the step (3) are 670.74 nm, 720.08 nm, 836.14 nm, 886.09 nm and 936.05 nm.
In the step (4), a model is established based on an SVM algorithm, a kernel function of the model is a Radial Basis Function (RBF), and an optimal solution is obtained by selecting an appropriate penalty factor and a kernel parameter for calculation.
The method for distinguishing the storage period of the tea by applying the hyperspectral image technology is suitable for various types of tea except partial black tea and compressed tea.
The application of an analysis model established by a tea storage period distinguishing method based on a hyperspectral image technology in distinguishing the storage period of tea is used for distinguishing the storage period of the tea.
The invention has the beneficial effects that:
the tea storage period judging method directly adopts the original sample detection, omits fussy pretreatment, and can analyze the tea storage period only by using the sample to be detected to acquire images through an instrument after a mathematical model is established, thereby meeting the requirement of online detection of the tea. The method has the characteristics of no need of sample weighing, no need of calculation, no need of sample pretreatment, no need of any chemical reagent, no sample damage, stability, quickness, greenness, environmental protection, easiness in operation, high accuracy and good repeatability.
Drawings
FIG. 1 shows hyperspectral original spectrograms of tea samples in different storage periods in the invention, (a) Liuan Guapian fresh-keeping tea, (b) Liuan Guapian fresh-keeping tea, (c) Huangshan Maofeng fresh-keeping tea, (d) Huangshan Maofeng stored tea.
FIG. 2 is a graph showing the weighting coefficients of the first three principal component images according to the present invention.
FIG. 3 shows an image of a fresh-keeping tea sample with six ampere of light sheets under 5 characteristic wavelengths in the invention.
FIG. 4 is a diagram showing an image of a six-ampere slide stored tea sample at 5 characteristic wavelengths in the present invention.
FIG. 5 is an image of Huangshan Maofeng tea sample at 5 characteristic wavelengths in the present invention.
FIG. 6 is an image showing the stored tea samples of Huangshan Maofeng tea at 5 characteristic wavelengths in the present invention.
Detailed Description
The following examples are given for the detailed implementation and specific operation of the present invention, but the scope of the present invention is not limited to the following examples.
Example 1
The method for discriminating the storage period of the tea by applying the qualitative analysis model established by combining the visible/near-infrared hyperspectral image analysis technology and the support vector machine algorithm comprises the following steps:
(1) preparing an aged tea sample: after measuring the initial water content of the purchased Liuan Guapian tea sample, subpackaging the Liuan Guapian tea sample in aluminum foil bags and sealing the aluminum foil bags, and storing the Liuan Guapian tea sample under two conditions; storing a part of tea samples in a refrigeration house as fresh-keeping tea samples, setting the aging time as one year, respectively sampling at 0, 90, 180, 270 and 360 d, and taking five groups of tea samples, wherein each group of tea samples takes 30 samples, each sample takes 30 g, and 150 samples are obtained; storing the other part of tea samples at normal temperature, setting the aging time as one year, respectively sampling at 0, 90, 180, 270 and 360 d, and taking five groups of tea samples, wherein each group of tea samples takes 30 samples, and each sample takes 30 g to obtain 150 samples; weighing 12 +/-0.5 g of tea leaves as a sample, uniformly and flatly paving the tea leaves in a culture dish with the specification of phi 9 cm, and collecting hyperspectral images;
(2) collecting hyperspectral images, selecting a sample correction set and a test set: collecting hyperspectral images of samples by using a hyperspectral image system based on a visible/near-infrared spectrometer and matched collection and analysis software of Taiwan Wuling optics corporation, wherein the scanning range of a visible/near-infrared camera is 370.38-1036.54 nm, the sampling interval is 0.55 nm, each spectrum has 1232 variables, and after setting the exposure time of a hyperspectral camera and a conveying device, the exposure time of the hyperspectral camera is predetermined to ensure the clarity of the images before collecting hyperspectral image data; the speed of the conveyor is determined to avoid image size and spatial resolution distortion, and the visible/near infrared camera acquisition parameters are set as follows: the exposure time is 8.5 ms, and the speed of the conveying device is 1.15 mm/s; scanning the tea samples in different storage periods prepared in the step (1), scanning a standard white correction plate and a black correction image before collecting sample images, correcting the sample images by using a black-white correction image to obtain hyperspectral images of the tea samples in different storage periods, selecting one sample as a test sample every two samples, namely 50 samples in 150 samples as an inspection set, and taking the other 100 samples as a correction set;
(3) dimension reduction processing and characteristic value extraction: performing dimensionality reduction on original hyperspectral data by using principal component analysis, extracting five characteristic wavelengths based on the weight coefficients of the first three principal component images, selecting the center of a 100 x 100 pixel region in the center of the image as an ROI region, selecting spectral images of the Liuan Guapian fresh-keeping tea sample and the stored tea sample as shown in (a) and (b) of a picture, selecting the weight coefficient results of the first three principal component images as shown in (2) and obtaining 5 characteristic wavelengths: 670.74, 720.08, 836.14, 886.09, and 936.05 nm; images of the preserved tea sample and the stored tea sample of the Liuan Guapian tea under 5 characteristic wavelengths are shown in figures 3 and 4, spectral data of ROI areas of the 5 characteristic wavelength images are extracted and stored as spectral characteristic information, textures of the whole image of the tiled tea are calculated based on GLCM according to the characteristics of the tea image, and texture characteristic variables under the characteristic wavelengths are extracted as texture characteristic information;
(4) establishing a model: selecting optimal parameters, establishing an SVM discrimination model of the storage period of the Liuan Guapian fresh-keeping tea sample and the stored tea sample introduced with a radial basis function, wherein the obtained optimal punishment factors c of the Liuan Guapian fresh-keeping tea sample and the stored tea sample are respectively 64 and 111.431, and the RBF nuclear parameter g is respectively 0.047 and 0.009; the discrimination rates of the correction sets of the Liuan melon slice fresh-keeping tea sample and the stored tea sample are respectively 100% and 99%, and the discrimination rate of the model correction set is higher;
(5) and (3) testing the model: using the test set in the step (2) as an unknown sample to test the calibration set model; the discrimination rates of the Liuan melon slice fresh-keeping tea sample and the stored tea sample are respectively 100% and 96%, and the accurate discrimination of the storage period can be realized.
Example 2
The method for discriminating the storage period of the tea by applying the qualitative analysis model established by combining the visible/near-infrared hyperspectral image analysis technology and the support vector machine algorithm comprises the following steps:
(1) preparing an aged tea sample: after measuring the initial water content of the purchased Huangshan Maofeng tea sample, subpackaging the Huangshan Maofeng tea sample in aluminum foil bags and sealing the aluminum foil bags, and storing the Huangshan Maofeng tea sample under two conditions; storing a part of tea samples in a refrigeration house as fresh-keeping tea samples, setting the aging time as one year, respectively sampling at 0, 90, 180, 270 and 360 d, and taking five groups of tea samples, wherein each group of tea samples takes 30 samples, each sample takes 30 g, and 150 samples are obtained; storing the other part of tea samples at normal temperature, setting the aging time as one year, respectively sampling at 0, 90, 180, 270 and 360 d, and taking five groups of tea samples, wherein each group of tea samples takes 30 samples, and each sample takes 30 g to obtain 150 samples; weighing 12 +/-0.5 g of tea leaves as a sample, uniformly and flatly paving the tea leaves in a culture dish with the specification of phi 9 cm, and collecting hyperspectral images;
(2) collecting hyperspectral images, selecting a sample correction set and a test set: collecting a hyperspectral image of a sample by using a hyperspectral image system based on a visible/near-infrared spectrometer and matched collection and analysis software of Taiwan Wuling optical corporation, wherein the scanning range of a visible/near-infrared camera is 370.38-1036.54 nm, the sampling interval is 0.55 nm, and each spectrum has 1232 variables; after setting the exposure time of the hyperspectral camera and the conveying device, and before acquiring hyperspectral image data, determining the exposure time of the hyperspectral camera in advance to ensure the clarity of the image; the speed of the conveyor is determined to avoid image size and spatial resolution distortion, and the visible/near infrared camera acquisition parameters are set as follows: the exposure time is 8.5 ms, and the speed of the conveying device is 1.15 mm/s; scanning the tea samples in different storage periods prepared in the step (1), scanning a standard white correction plate and a black correction image before collecting a sample image, and correcting the sample image by using the black and white correction image to obtain hyperspectral images of the tea samples in different storage periods. Selecting one sample as a test sample every two samples, namely 50 samples in 150 samples as a test set, and the rest 100 samples as a calibration set;
(3) dimension reduction processing and characteristic value extraction: carrying out dimensionality reduction on original hyperspectral data by using principal component analysis, extracting five characteristic wavelengths based on weight coefficients of the first three principal component images, selecting the center of a 100 x 100 pixel region at the center of the image as an ROI region, selecting spectral images of Huangshan Maofeng fresh-keeping tea samples and stored tea samples as shown in (c) (d) of a graph in figure 1, and selecting to obtain 5 characteristic wavelengths: 670.74, 720.08, 836.14, 886.09, and 936.05 nm; images of the Huangshan Maofeng fresh-keeping tea sample and the stored tea sample at 5 characteristic wavelengths are shown in FIGS. 5 and 6, and spectral data of ROI areas of the 5 characteristic wavelength images are extracted and stored as spectral characteristic information; calculating the texture of the whole image of the tiled tea based on GLCM according to the characteristics of the tea image, and extracting texture characteristic variables under characteristic wavelengths as texture characteristic information;
(4) establishing a model: selecting optimal parameters, establishing an SVM discrimination model of the storage period of the Huangshan Maofeng fresh-keeping tea sample and the stored tea sample introduced with a radial basis function, wherein the obtained optimal punishment factors c of the Huangshan Maofeng fresh-keeping tea sample and the stored tea sample are respectively 0.435 and 588.134, and the RBF nuclear parameter g is respectively 0.25 and 0.009; the discrimination rates of correction sets of the Huangshan Maofeng fresh-keeping tea sample and the stored tea sample reach 100 percent, and the discrimination rate of the model correction set is higher;
(5) and (3) testing the model: using the test set in the step (2) as an unknown sample to test the calibration set model; the discrimination rate of the Huangshan Maofeng fresh-keeping tea sample and the stored tea sample reaches 100 percent and 98 percent, and the accurate discrimination of the storage period can be realized.
Example 3
The method for rapidly judging the storage period of the tea by applying a qualitative analysis model established by combining a visible/near-infrared hyperspectral image analysis technology and a support vector machine algorithm comprises the following steps:
(1) pretreatment of unknown tea samples: taking an unknown sample tea sample, weighing 12 +/-0.5 g of tea leaves as a sample, and uniformly and flatly paving the tea leaves in a culture dish with the specification of phi 9 cm;
(2) collecting a hyperspectral image: collecting a hyperspectral image of an unknown sample by using a hyperspectral image system based on a visible/near-infrared spectrometer and matched collection and analysis software of Taiwan Wuling optics corporation, wherein a visible/near-infrared camera has a scanning range of 370.38-1036.54 nm, a sampling interval is 0.55 nm, and each spectrum has 1232 variables; after setting the exposure time of the hyperspectral camera and the conveying device, and before acquiring hyperspectral image data, determining the exposure time of the hyperspectral camera in advance to ensure the clarity of the image; the speed of the transport apparatus is determined to avoid distortion of image size and spatial resolution. The visible/near infrared camera acquisition parameters are set as follows: the exposure time is 8.5 ms, and the speed of the conveying device is 1.15 mm/s;
(3) determination of unknown samples: the established SVM model is called in Matlab software, the spectral information and the texture information of the storage samples are analyzed, and the storage period of the tea can be rapidly judged.
Claims (3)
1. A method for discriminating the storage period of tea by applying a hyperspectral image technology is characterized by comprising the following steps:
(1) preparing an aged tea sample: after the initial moisture content of the purchased tea is measured, subpackaging the tea in aluminum foil bags and sealing the bags, and storing the tea under two conditions; storing a part of tea samples at normal temperature, setting the aging time as one year, respectively sampling at 0, 90, 180, 270 and 360 d, and taking five groups of tea samples, wherein each group of tea samples takes 30 samples, and each sample takes 30 g to obtain 150 samples; storing the other part of tea samples in a refrigeration house as fresh-keeping tea samples, setting the aging time as one year, respectively sampling at 0, 90, 180, 270 and 360 d, and taking five groups of tea samples, wherein each group of tea samples takes 30 samples, each sample takes 30 g, and 150 samples are obtained;
(2) collecting hyperspectral images, selecting a sample correction set and a test set: collecting hyperspectral images of samples by using a hyperspectral image system based on a visible/near-infrared spectrometer and matched collection and analysis software, wherein the scanning range of a visible/near-infrared camera is 370.38-1036.54 nm, the sampling interval is 0.55 nm, each spectrum has 1232 variables, scanning the tea samples in different storage periods prepared in the step (1) after setting the exposure time of the hyperspectral camera and a conveying device, scanning a standard white correction plate and a black correction image before collecting the images of the samples, correcting the images of the samples by using the black and white correction image to obtain the hyperspectral images of the tea samples in different storage periods, and selecting one sample as a test sample every two samples, namely 50 samples in 150 samples as an inspection set and the other 100 samples as correction sets;
(3) dimension reduction processing and characteristic value extraction: performing dimensionality reduction on original hyperspectral data by utilizing principal component analysis, extracting five characteristic wavelengths based on weight coefficients of first three principal component images, selecting the center of a 100 x 100 pixel region at the center of an image as an interested region, extracting and storing spectral data of 5 characteristic wavelength image ROI regions as spectral characteristic information, calculating the texture of an overall image of tiled tea leaves based on a gray level co-occurrence matrix according to the characteristics of the tea leaf image, and extracting texture characteristic variables under the characteristic wavelengths as texture characteristic information;
(4) two models were established: selecting optimal parameters, and respectively establishing a storage period discrimination model of the stored tea sample and a storage period discrimination model of the fresh-keeping tea sample based on a support vector machine;
(5) and (3) testing the model: using the test set in the step (2) as an unknown sample to test the calibration set model;
(6) taking tea samples in an unknown storage period, performing hyperspectral collection according to the step (2), and analyzing hyperspectral images of the tea samples in the unknown storage period by utilizing the established SVM model to quickly and qualitatively judge the storage period of the tea;
when the hyperspectral image is acquired in the step (2), weighing 11.5-12.5g of tea leaves, and uniformly and flatly paving the tea leaves in a culture dish with the specification of phi 9 cm;
before the hyperspectral image data are collected in the step (2), the collection parameters of the visible/near-infrared camera are set to be exposure time of 8.5 ms, and the speed of the conveying device is 1.15 mm/s;
the five characteristic wavelengths screened by adopting the main component analysis in the step (3) are 670.74 nm, 720.08 nm, 836.14 nm, 886.09 nm and 936.05 nm.
2. The method for discriminating the storage period of tea leaves by applying the hyperspectral image technique according to claim 1, wherein the model is established in the step (4) based on an SVM algorithm, the kernel function of the model is a radial basis kernel function, and an optimal solution is obtained by selecting an appropriate penalty factor and a kernel parameter for calculation.
3. The method for discriminating the storage period of tea leaves by applying the hyperspectral image technique as claimed in claim 1, wherein the discrimination method is applied to various types of tea leaves except for partial black tea and compressed tea.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810211952.6A CN108444924B (en) | 2018-03-07 | 2018-03-07 | Method for discriminating storage period of tea by applying hyperspectral image technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810211952.6A CN108444924B (en) | 2018-03-07 | 2018-03-07 | Method for discriminating storage period of tea by applying hyperspectral image technology |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108444924A CN108444924A (en) | 2018-08-24 |
| CN108444924B true CN108444924B (en) | 2020-12-08 |
Family
ID=63195289
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810211952.6A Active CN108444924B (en) | 2018-03-07 | 2018-03-07 | Method for discriminating storage period of tea by applying hyperspectral image technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108444924B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109470639A (en) * | 2018-11-07 | 2019-03-15 | 浙江大学 | Diaphania larva based on visible light and near-infrared high light spectrum image-forming and its method for quickly identifying to mulberry leaf damage |
| CN110084788B (en) * | 2019-04-15 | 2022-12-20 | 虚境茶仓(宁波)网络科技有限公司 | Compressed tea appearance texture recording and identifying method |
| CN111524092B (en) * | 2020-03-16 | 2024-03-12 | 中国农业科学院茶叶研究所 | Nondestructive testing method for black tea withering degree index |
| CN112699756B (en) * | 2020-12-24 | 2023-08-25 | 中国农业科学院农业信息研究所 | Hyperspectral image-based tea origin identification method and system |
| CN113484309B (en) * | 2021-06-28 | 2022-08-09 | 安徽农业大学 | Method for monitoring black tea fermentation degree by hyperspectral coupling nano colorimetric sensor |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004061412A2 (en) * | 2002-12-23 | 2004-07-22 | Gentag, Inc. | Method and apparatus for wide area surveillance of a terrorist or personal threat |
| WO2006102415A3 (en) * | 2005-03-22 | 2009-04-16 | Applera Corp | Optical camera alignment |
| CN104897729A (en) * | 2015-06-04 | 2015-09-09 | 安徽农业大学 | Sorting method of storage time of melon slice tea by using electronic nose |
| CN107169407A (en) * | 2017-03-29 | 2017-09-15 | 广东工业大学 | Hyperspectral image classification method based on joint bilateral filtering and extreme learning machine |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101692052B (en) * | 2009-08-31 | 2012-05-23 | 江苏大学 | Hyperspectrum image technique-based method and hyperspectrum image technique-based device for identifying true and false famous tea |
| US8559719B2 (en) * | 2010-03-12 | 2013-10-15 | The United States Of America, As Represented By The Secretary Of The Navy | Spectral anomaly detection in deep shadows |
-
2018
- 2018-03-07 CN CN201810211952.6A patent/CN108444924B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004061412A2 (en) * | 2002-12-23 | 2004-07-22 | Gentag, Inc. | Method and apparatus for wide area surveillance of a terrorist or personal threat |
| WO2006102415A3 (en) * | 2005-03-22 | 2009-04-16 | Applera Corp | Optical camera alignment |
| CN104897729A (en) * | 2015-06-04 | 2015-09-09 | 安徽农业大学 | Sorting method of storage time of melon slice tea by using electronic nose |
| CN107169407A (en) * | 2017-03-29 | 2017-09-15 | 广东工业大学 | Hyperspectral image classification method based on joint bilateral filtering and extreme learning machine |
Non-Patent Citations (2)
| Title |
|---|
| Objects Detection in Hyperspectral Images Using Spectral Derivative;Kumar;《Journal of the Indian Society of Remote Sensing》;20171231;第45卷(第4期);第603-610页 * |
| 基于高光谱分析技术的机炒龙井茶等级识别方法;蒋帆;《农业工程学报》;20110731;第27卷(第7期);第46-50页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108444924A (en) | 2018-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108444924B (en) | Method for discriminating storage period of tea by applying hyperspectral image technology | |
| US11454594B2 (en) | Method for identifying frostbite condition of grain seeds using spectral feature wavebands of seed embryo hyperspectral images | |
| Afshari-Jouybari et al. | Evaluation of Photoshop software potential for food colorimetry | |
| CN110763698B (en) | Hyperspectral citrus leaf disease identification method based on characteristic wavelength | |
| Liu et al. | Prediction of nitrogen and phosphorus contents in citrus leaves based on hyperspectral imaging | |
| CN106290230B (en) | A kind of method of discrimination of the black tea withering degree of near infrared spectrum combination chemical component | |
| CN101692052B (en) | Hyperspectrum image technique-based method and hyperspectrum image technique-based device for identifying true and false famous tea | |
| CN107185850B (en) | Corn seed activity detection device based on hyperspectral imaging and electrical impedance | |
| CN105021562A (en) | Traceable method for rice origin and application thereof | |
| CN104977258B (en) | The quality detecting methods such as tealeaves/cosmetics based on two-dimensional correlation spectra | |
| CN103940748B (en) | Based on the prediction of oranges and tangerines canopy nitrogen content and the visualization method of hyperspectral technique | |
| CN113008817A (en) | Method for rapidly identifying authenticity and quality of bitter apricot kernels based on hyperspectral imaging technology | |
| CN105181606B (en) | The method that cane sugar content is distributed in peanut is detected based on high light spectrum image-forming technology | |
| CN101696935A (en) | Apple rigidity nondestructive testing method based on hyperspectral space scattering curve | |
| Laddi et al. | Significant physical attributes affecting quality of Indian black (CTC) tea | |
| Kuchekar et al. | Rice grain quality grading using digital image processing techniques | |
| Ren et al. | Evaluation of Dianhong black tea quality using near‐infrared hyperspectral imaging technology | |
| CN110108649A (en) | The fast non-destructive detection method of oil crops quality based on terahertz light spectral technology | |
| Yu et al. | Rapid and visual measurement of fat content in peanuts by using the hyperspectral imaging technique with chemometrics | |
| Torres et al. | Setting up a methodology to distinguish between green oranges and leaves using hyperspectral imaging | |
| Tang et al. | A nondestructive method for determination of green tea quality by hyperspectral imaging | |
| CN106940292A (en) | Bar denier wood raw material quick nondestructive discrimination method of damaging by worms based on multi-optical spectrum imaging technology | |
| Portalés et al. | An image-based system to preliminary assess the quality of grape harvest batches on arrival at the winery | |
| CN107356536B (en) | Wine quality detection method, server and spectral measurement equipment | |
| Li | Classification of black tea leaf water content based on hyperspectral imaging |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |