CN112504977A - Tea water content detection method and model construction method, medium and equipment thereof - Google Patents
Tea water content detection method and model construction method, medium and equipment thereof Download PDFInfo
- Publication number
- CN112504977A CN112504977A CN202011283346.9A CN202011283346A CN112504977A CN 112504977 A CN112504977 A CN 112504977A CN 202011283346 A CN202011283346 A CN 202011283346A CN 112504977 A CN112504977 A CN 112504977A
- Authority
- CN
- China
- Prior art keywords
- tea
- tea leaf
- hyperspectral data
- regression model
- leaf sample
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/27—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands using photo-electric detection ; circuits for computing concentration
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N5/00—Analysing materials by weighing, e.g. weighing small particles separated from a gas or liquid
- G01N5/04—Analysing materials by weighing, e.g. weighing small particles separated from a gas or liquid by removing a component, e.g. by evaporation, and weighing the remainder
- G01N5/045—Analysing materials by weighing, e.g. weighing small particles separated from a gas or liquid by removing a component, e.g. by evaporation, and weighing the remainder for determining moisture content
Abstract
The embodiment discloses a tea leaf moisture content detection method and a model construction method, medium and equipment thereof, wherein hyperspectral data of tea leaves serving as samples are obtained firstly, and moisture content obtained by processing the hyperspectral data through a drying method is obtained; respectively establishing a first regression model MA, a second regression model MB, a third regression model MC and a fourth regression model MD according to the corresponding relation between the hyperspectral data of the front surface and the back surface of each tea leaf and the moisture content of each tea leaf; verifying the accuracy of the first regression model MA, the second regression model MB, the third regression model MC and the fourth regression model MD, and taking the model with the highest accuracy as a tea water content detection model. The invention can simply, rapidly, accurately and nondestructively detect the moisture content of the tea and has the advantage of good reproducibility.
Description
Technical Field
The invention relates to the technical field of tea processing detection, in particular to a tea moisture content detection method, a model construction method, a medium and equipment thereof.
Background
The water content of the tea leaves is closely related to the quality of the tea leaves, when the water content is between 6 and 7 percent, the quality of the tea leaves is relatively stable, when the water content of the tea leaves exceeds 8 percent, the tea leaves are easy to age, and when the water content of the tea leaves exceeds 12 percent, the tea leaves are easy to mildew. Therefore, after the tea processing is finished, the detection of the moisture content of the tea is very important.
The detection of tea moisture content among the prior art includes at present:
(1) the moisture content of the tea leaves is judged according to the sound of the tea leaves by artificially touching and applying force to the tea leaves. This method requires a great deal of experience from the tester, and the accuracy of the moisture content of the tea leaves obtained is low, since it is determined by the sense of the sense.
(2) The measurement is carried out by a drying method, the quality of the tea leaves before drying is recorded, the direct current of the dried tea leaves is recorded after the tea leaves are dried, and the tax rate of the tea leaves before drying is calculated according to the direct current of the dried tea leaves before and after drying. In this way, the accuracy of the moisture content test is greatly improved compared with the former method, but the tea leaves need to be dried, and the treated tea leaves may not meet the relevant requirements, which is a test which is harmful to the tea leaves.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a tea moisture content detection model method which can simply, quickly, accurately and nondestructively detect the moisture content of tea and has the advantage of good reproducibility.
The second purpose of the invention is to provide a tea moisture content detection model construction device.
The third purpose of the invention is to provide a method for detecting the moisture content of tea.
The fourth purpose of the invention is to provide a tea moisture content detection system.
A fifth object of the present invention is to provide a storage medium.
It is a sixth object of the invention to provide a computing device.
The first purpose of the invention is realized by the following technical scheme: a method for constructing a tea moisture content detection model comprises the following steps:
collecting hyperspectral data of tea leaves serving as samples by a hyperspectral image collection system;
acquiring hyperspectral data of the front surface and the back surface of each tea leaf sample;
for each tea leaf sample, acquiring the water content obtained by drying treatment;
fitting and establishing a first regression model MA through the hyperspectral data of the front side of each tea leaf sample and the water content of each tea leaf sample;
fitting and establishing a second regression model MB through the hyperspectral data of the reverse side of each tea leaf sample and the water content of each tea leaf sample;
aiming at each tea leaf sample, calculating average hyperspectral data through the front hyperspectral data and the back hyperspectral data of the tea leaf sample, and fitting and establishing a third regression model MC through the average hyperspectral data of each tea leaf sample and the water content of each tea leaf sample;
fitting and establishing a fourth regression model MD through the hyperspectral data of the front side of each tea leaf sample, the water content of each tea leaf sample, the hyperspectral data of the back side of each tea leaf sample and the water content of each tea leaf sample;
verifying the accuracy of the first regression model MA, the second regression model MB, the third regression model MC and the fourth regression model MD, and taking the model with the highest accuracy as a tea water content detection model.
Preferably, when a certain number of tea leaf samples are arranged in the sample pool, hyperspectral data is acquired by a hyperspectral image acquisition system;
the hyperspectral data acquired by the hyperspectral image acquisition system each time is ax×ay×azIn which axRepresenting camera level in a hyperspectral image acquisition systemNumber of pixels of direction, ayRepresents the total number of points of the wavelength 363-1025 nm, azAnd the total scanning line number of the camera in the hyperspectral image acquisition system is shown.
Furthermore, extract the high spectral data of every tealeaves leaf sample openly or reverse side in the sample cell from the high spectral data that high spectral image collection system gathered at every turn, wherein when tealeaves leaf sample openly up, what extract is the positive high spectral data of tealeaves leaf sample, when tealeaves leaf sample reverse side up, what extract is the high spectral data of tealeaves leaf sample reverse side, specifically as follows:
SA, selecting a spectral image with the wavelength of 580nm from three-dimensional hyperspectral data;
SB, converting the spectral image with the wavelength of 580nm into a binary image, and acquiring the edge space position information of each tea leaf blade discharged in the sample cell through the binary image;
SC, correspondingly extracting hyperspectral data of each tea leaf sample from the three-dimensional hyperspectral data according to the edge space position information of each tea leaf sample;
and SD, averaging the acquired hyperspectral data of each tea leaf sample to obtain average spectral data serving as final hyperspectral data of the tea leaf sample.
Preferably, the method further comprises the following steps: performing black-and-white plate correction processing and hyperspectral data preprocessing on hyperspectral data acquired by a hyperspectral image acquisition system;
wherein:
performing black and white plate correction on the hyperspectral data by the following formula:
in the formula I0The hyperspectral data after the white board correction is acquired, I is the original hyperspectral data acquired by the original spectrum hyperspectral image acquisition system, IWAverage spectral data for white board, IDAverage spectral data for the blackboard;
the post-correction hyperspectral data is preprocessed by decentralization, standard normal transformation, multivariate scatter correction, first derivative application of Savitzky-Golay smoothing, or second derivative application of Savitzky-Golay smoothing.
Preferably, for each sample of tea leaves, the process of obtaining the moisture content by the drying method is as follows:
sa, firstly, placing a clean bottle without a cover in a blast drier at 105 ℃ for drying for 1.0 h;
sb, measuring and recording the mass of the dried front piece of the tea leaf sample by using an AL analytical balance with the precision of 0.001g, putting the tea leaf sample into a bottle with a corresponding label, drying the tea leaf sample in a blast drier at 105 ℃ for 3.0h, and recording the mass of the tea leaf after cooling for 0.5 h;
sc, repeating the step Sb for multiple times, and entering a step Sd if the difference between the quality of the tea leaf sample measured after the step Sb is executed for the current time and the quality of the tea leaf sample measured after the step Sb is executed for the last time is not more than 2 mg;
sd, calculating the water content of the tea leaf sample by the following formula
The water content is (m 1-m 2)/m1 multiplied by 100 percent;
m1 is the mass of the tea leaves before drying treatment, and m2 is the mass of the tea leaf sample after the last drying.
The second purpose of the invention is realized by the following technical scheme: a tea moisture content detection model construction device comprises:
the system comprises a first acquisition module, a second acquisition module and a control module, wherein the first acquisition module is used for acquiring hyperspectral data of tea leaves serving as samples through a hyperspectral image acquisition system; acquiring hyperspectral data of the front surface and the back surface of each tea leaf sample;
the second acquisition module is used for acquiring the moisture content of each tea leaf sample obtained by drying treatment;
the first regression model establishing module is used for fitting and establishing a first regression model MA through the hyperspectral data and the water content of the front side of each tea leaf;
the second regression model establishing module is used for fitting and establishing a second regression model MB through the hyperspectral data of the reverse side of each tea leaf sample and the water content of each tea leaf sample;
the third regression model establishing module is used for calculating average hyperspectral data according to the front hyperspectral data and the back hyperspectral data of each tea leaf sample, and fitting and establishing a third regression model MC according to the average hyperspectral data of each tea leaf sample and the water content of each tea leaf sample;
the fourth regression model establishing module is used for fitting and establishing a fourth regression model MD through the hyperspectral data of the front surface of each tea leaf sample and the moisture content of each tea leaf sample, the hyperspectral data of the front surface of each tea leaf sample and the moisture content of each tea leaf sample;
and the verification module is used for verifying the accuracy of the first regression model MA, the second regression model MB, the third regression model MC and the fourth regression model MD and taking the model with the highest accuracy as the tea water content detection model.
The third purpose of the invention is realized by the following technical scheme: a method for detecting the moisture content of tea comprises the following steps:
acquiring a tea moisture content detection model acquired by the tea moisture content detection model construction method of the first object of the invention;
if the tea moisture content detection model is a first regression model MA, then:
the method comprises the steps that hyperspectral data of the front side of tea leaves to be detected are obtained for the tea leaves to be detected, the hyperspectral data are input into a first regression model MA, and the moisture content of the tea leaves to be detected is obtained through calculation of the first regression model MA;
if the tea moisture content detection model is a second regression model MB, then:
acquiring hyperspectral data of the reverse side of the tea to be detected aiming at the tea leaves to be detected, inputting the hyperspectral data into a second regression model MB, and calculating by using the second regression model MB to obtain the water content of the tea to be detected;
if the tea moisture content detection model is a third regression model MC, then:
for the tea leaves to be detected, acquiring the average value of the front hyperspectral data and the back hyperspectral data of the tea leaves to be detected, inputting the average value into a third regression model MC, and calculating through the third regression model MC to obtain the water content of the tea leaves to be detected;
if the tea moisture content detection model is a fourth regression model MD, then:
and aiming at the tea leaves to be detected, randomly acquiring the hyperspectral data of the front side or the hyperspectral data of the back side of the tea leaves to be detected, inputting the data into a fourth regression model MD, and calculating through the fourth regression model MD to obtain the water content of the tea leaves to be detected.
The fourth purpose of the invention is realized by the following technical scheme: a tea moisture content detection system comprises an image acquisition device and an upper computer; wherein:
the image acquisition equipment is used for acquiring hyperspectral data of the tea leaves;
and the upper computer is used for executing the tea moisture content detection model construction method for the first purpose of the invention and/or the tea moisture content detection method for the third purpose of the invention.
The fifth purpose of the invention is realized by the following technical scheme: a storage medium storing a program for implementing the method for constructing a tea water content detection model according to the first aspect of the present invention and/or the method for detecting a tea water content according to the third aspect of the present invention when the program is executed by a processor.
The sixth purpose of the invention is realized by the following technical scheme: a computing device comprising a processor and a memory for storing a program executable by the processor, wherein the processor executes the program stored in the memory to implement the tea moisture content detection model construction method according to the first object of the present invention and/or to implement the tea moisture content detection method according to the third object of the present invention.
Compared with the prior art, the invention has the following advantages and effects:
(1) the method for constructing the tea water content detection model comprises the steps of firstly, acquiring hyperspectral data of tea leaves serving as samples, and acquiring the water content of the tea leaves obtained through processing by a drying method; performing correlation analysis through hyperspectral data and water content of the front surface of each tea leaf to establish a first regression model MA; fitting and establishing a second regression model MB through the hyperspectral data of the reverse side of each tea leaf sample and the water content of each tea leaf sample; fitting and establishing a third regression model MC through average hyperspectral data of the front side and the back side of each tea leaf sample and the water content of each tea leaf sample; fitting and establishing a fourth regression model MD through the hyperspectral data of the front side or the back side of each tea leaf sample, the water content of each tea leaf sample, the hyperspectral data of the back side of each tea leaf sample and the water content of each tea leaf sample; verifying the accuracy of the first regression model MA, the second regression model MB, the third regression model MC and the fourth regression model MD, and taking the model with the highest accuracy as a tea water content detection model. According to the invention, the tea leaf moisture content detection model is constructed based on the tea leaf sample hyperspectral data and the moisture content of the tea leaf sample, the model constructed by the invention can detect the moisture content of the tea leaf without damaging the tea leaf, and the moisture content detection is realized by combining the tea leaf hyperspectral data and a mathematical formula, can simply, quickly, accurately and nondestructively detect the moisture content of the tea leaf, and has the advantage of good reproducibility.
(2) According to the method for constructing the tea leaf water content detection model, the hyperspectral data of a plurality of tea leaves can be acquired at one time through the sample pool capable of containing the plurality of leaves, and the hyperspectral data of each tea leaf sample can be extracted from the once acquired spectral data based on the edge position information of each tea leaf in one picture.
(3) According to the tea water content detection model construction method, black and white board correction processing and hyperspectral data preprocessing are carried out on hyperspectral data acquired by a hyperspectral image acquisition system, and the accuracy of tea water content detection can be further improved by the tea water content detection model constructed based on the hyperspectral data after correction processing and preprocessing.
(4) According to the tea leaf moisture content detection model construction method, the model accuracy constructed by hyperspectral data acquired from different faces of tea leaves is compared, so that the optimal tea leaf moisture content detection model is determined, and the optimal tea leaf moisture content detection model can be acquired to the greatest extent.
Drawings
FIG. 1 is a flow chart of a method for constructing a tea moisture content detection model according to the present invention.
FIG. 2 is a block diagram of the structure of a tea moisture content detection model construction device according to the present invention.
FIG. 3 is a block diagram of the structure of the system for detecting the moisture content of tea leaves according to the present invention.
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but the present invention is not limited thereto.
Example 1
The embodiment discloses a tea moisture content detection model building method, and the tea moisture content detection model built by the method can quickly and accurately detect the tea moisture content without damaging tea leaves. As shown in fig. 1, the method includes:
s1, randomly obtaining a plurality of tea leaf samples, in this embodiment, the obtained samples may include 10 tea leaf samples in total, such as longjing green tea, bi rochun green tea, huangshan maofeng green tea, lushan yunwu green tea, taiping kui green tea, Pingyu bead green tea, sunshine black tea, yunnan red tea, mu lan black tea, and phoenix simplex cluster, and randomly selecting 50 tea leaf blades from each sample, wherein the total number of the tea leaf samples is 500.
And S2, selecting hyperspectral image acquisition equipment. In the embodiment, a VNIR-HIS-B1621 type visible-near infrared imaging system can be selected as the hyperspectral image acquisition equipment, and the equipment comprises a CCD camera, a 363-1025 nm spectral imager, an IT 3900150W halogen light source, a mobile platform, a dark box and the like.
S3, selecting a sample cell for loading green tea, wherein the sample cell can be a plastic tray with a length and a width of 10 cm. In this embodiment, a reference white board is provided, and a 17mm wide white electrical insulating tape is attached to the top of the sample cell in advance as the reference white board, so that not only the spectral information of the sample can be obtained each time the sample is scanned, but also the information of the customized reference white board can be obtained.
And S4, arranging a certain number of tea leaf samples in the sample cell each time, and particularly, placing 10 tea leaves in the sample cell each time. The CCD camera in the hyperspectral image acquisition equipment uses a 35n lens, when shooting is carried out, the distance between the lens and the sample pool is 33cm, the exposure time is 21ms, the speed of the mobile platform of the hyperspectral image acquisition equipment is 0.792mm/s, and the image resolution is 1632 × 1232.
In this embodiment, hyperspectral image data is collected by a hyperspectral image collection system, and the size of each hyperspectral data collected is ax×ay×azIn which axRepresenting the number of pixels in the horizontal direction of the camera in a hyperspectral image acquisition system, ayRepresents the total number of points of the wavelength 363-1025 nm, azRepresenting the total scanning line number of the camera in the hyperspectral image acquisition system; in this embodiment ayIt can be set to 1623, which means that the wavelength 363 ~ 1025 nm is divided into 1623 points.
And S5, collecting hyperspectral data of the tea leaves serving as the samples through a hyperspectral image collection system. In the sample pool, the tea sample is placed into the sample pool with the front side, the tea in the sample pool is arranged according to 5 rows and 2 columns, 10 pieces of tea high-spectrum data can be obtained through one-time spectrum scanning, the process is repeated for 50 times, the high-spectrum data of 500 tea leaf samples can be obtained, then the tea sample is placed into the sample pool with the B surface, and the high-spectrum data of 1000 tea leaves can be obtained through the repeated process.
For the 10 tea leaf high spectral data that above-mentioned each time obtained, follow this high spectral data and draw out the high spectral data of every tea leaf sample front or reverse side, wherein when the tea leaf sample openly up, what draw is the positive high spectral data of tea leaf sample, when the tea leaf sample reverse side up, what draw is the high spectral data of tea leaf sample reverse side, specifically as follows:
SA, selecting a spectral image with the wavelength of 580nm from three-dimensional hyperspectral data;
SB, converting the spectral image with the wavelength of 580nm into a binary image, and acquiring the edge space position information of each tea leaf blade discharged in the sample cell through the binary image;
SC, correspondingly extracting hyperspectral data of each tea leaf sample from the three-dimensional hyperspectral data according to the edge space position information of each tea leaf sample;
and SD, averaging the acquired hyperspectral data of each tea leaf sample to obtain average spectral data serving as final hyperspectral data of the tea leaf sample.
In this embodiment, the black-and-white board correction processing and the hyperspectral data preprocessing are performed on the acquired hyperspectral data, so that the preprocessed hyperspectral data is used in the following steps.
Wherein:
performing black and white plate correction on the hyperspectral data by the following formula:
in the formula I0The hyperspectral data after the white board correction is acquired, I is the original hyperspectral data acquired by the original spectrum hyperspectral image acquisition system, IWAverage spectral data for white board, IDAverage spectral data for the blackboard;
the post-correction hyperspectral data is preprocessed by decentralizing Mean, standard normal transformation SNV, multivariate scatter correction MSC, first derivative Savitzky-Golay smoothing SAVG1 or second derivative savvvg 2.
In this embodiment, after different preprocessing methods, the new spectral data will have different systemsAnd (6) measuring the characteristics. If the original spectrum is subjected to decentralized processing, the mean value of the new spectrum data is 0; after SNV prediction processing, the new spectral mean is 0 and the variance is 1. MSC pre-processing is beneficial to eliminate spectral scattering effects. SAVG1 or SAVG2 pretreatments correspond to modeling with difference spectra. And respectively comparing the prediction results of the pretreatment (None) with Mean, SNV, MSC + Mean, SAVG1+ Mean, SAVG1+ SNV and SAVG2+ SNV on the water content of the tea. According to correction set root mean square error (RM-SEC), prediction set Root Mean Square Error (RMSEP), correction set correlation coefficient (R)C) Prediction set correlation coefficient (R)P) To evaluate the predictive performance of the model. Under different Principal Components (PC), the smaller the RMSEP value, the larger the value, which shows that the model has better prediction performance.
S6, aiming at each tea leaf sample, obtaining the moisture content obtained by drying treatment; in this embodiment, for each tea leaf sample, the process of obtaining the moisture content by the drying method is as follows:
sa, firstly, placing a clean bottle without a cover in a blast drier at 105 ℃ for drying for 1.0 h;
sb, measuring and recording the mass of the dried front piece of the tea leaf sample by using an AL analytical balance with the precision of 0.001g, putting the tea leaf sample into a bottle with a corresponding label, drying the tea leaf sample in a blast drier at 105 ℃ for 3.0h, and recording the mass of the tea leaf after cooling for 0.5 h;
sc, repeating the step Sb for multiple times, and entering a step Sd if the difference between the quality of the tea leaf sample measured after the step Sb is executed for the current time and the quality of the tea leaf sample measured after the step Sb is executed for the last time is not more than 2 mg;
sd, calculating the water content of the tea leaf sample by the following formula
The water content is (m 1-m 2)/m1 multiplied by 100 percent;
m1 is the mass of the tea leaves before drying treatment, and m2 is the mass of the tea leaf sample after the last drying.
S7, fitting and establishing a first regression model MA through the hyperspectral data and the water content of the front side of each tea leaf;
fitting and establishing a first regression model MA through the hyperspectral data of the front side of each tea leaf sample and the water content of each tea leaf sample;
fitting and establishing a second regression model MB through the hyperspectral data of the reverse side of each tea leaf sample and the water content of each tea leaf sample;
aiming at each tea leaf sample, calculating average hyperspectral data through the front hyperspectral data and the back hyperspectral data of the tea leaf sample, and fitting and establishing a third regression model MC through the average hyperspectral data of each tea leaf sample and the water content of each tea leaf sample;
fitting and establishing a fourth regression model MD through the hyperspectral data of the front side of each tea leaf sample, the water content of each tea leaf sample, the hyperspectral data of the back side of each tea leaf sample and the water content of each tea leaf sample;
in this embodiment, the first regression model MA, the second regression model MB, the third regression model MC, and the fourth regression model MD may be modeled by multiple linear regression, principal component regression, Partial Least Squares Regression (PLSR), and the like.
S5, verifying the accuracy of the first regression model MA, the second regression model MB, the third regression model MC and the fourth regression model MD, and taking the model with the highest accuracy as a tea water content detection model. In this embodiment, the accuracy of each model can be detected through the obtained tea leaf sample and the known moisture content, and the smaller the RMSEP value is, the larger the value is, the better the prediction performance of the model is.
The embodiment constructs the tea moisture content detection model based on the tea leaf sample hyperspectral data and the moisture content of the tea leaf sample, and the model obtained by the embodiment can detect the tea moisture content without damaging the tea leaves, so that the moisture content detection realized by combining the tea leaf hyperspectral data and the mathematical formula can detect the tea moisture content simply, quickly, accurately and nondestructively, and has the advantage of good reproducibility.
Those skilled in the art will appreciate that all or part of the steps in the method according to the present embodiment may be implemented by a program to instruct the relevant hardware, and the corresponding program may be stored in a computer-readable storage medium. It should be noted that although the method operations of embodiment 1 are described in the foregoing description and in the accompanying drawings in a particular order, this does not require or imply that these operations must be performed in this particular order, or that all of the illustrated operations must be performed, to achieve desirable results. Rather, the depicted steps may change the order of execution, and some steps may be executed concurrently. Additionally or alternatively, certain steps may be omitted, multiple steps combined into one step execution, and/or one step broken down into multiple step executions.
Example 2
The embodiment discloses a tea moisture content detection model building device, as shown in fig. 2, including a first obtaining module, a second obtaining module, a first regression model building module, a second regression model building module, a third regression model building module, a fourth regression model building module and a verification module, the specific functions of each module are as follows:
the first acquisition module is used for acquiring hyperspectral data of the tea leaves serving as the samples through a hyperspectral image acquisition system. Acquiring hyperspectral data of the front surface and the back surface of each tea leaf sample;
the second acquisition module is used for acquiring the moisture content of each tea leaf sample obtained by drying treatment;
the first regression model establishing module is used for fitting and establishing a first regression model MA through the hyperspectral data and the water content of the front side of each tea leaf;
the second regression model establishing module is used for fitting and establishing a second regression model MB through the hyperspectral data of the reverse side of each tea leaf sample and the water content of each tea leaf sample;
the third regression model establishing module is used for calculating average hyperspectral data according to the front hyperspectral data and the back hyperspectral data of each tea leaf sample, and fitting and establishing a third regression model MC according to the average hyperspectral data of each tea leaf sample and the water content of each tea leaf sample;
the fourth regression model establishing module is used for fitting and establishing a fourth regression model MD through the hyperspectral data of the front surface of each tea leaf sample and the moisture content of each tea leaf sample, the hyperspectral data of the front surface of each tea leaf sample and the moisture content of each tea leaf sample;
and the verification module is used for verifying the accuracy of the first regression model MA, the second regression model MB, the third regression model MC and the fourth regression model MD and taking the model with the highest accuracy as the tea water content detection model.
For specific implementation of each module in this embodiment, reference may be made to embodiment 1, and details are not described here. It should be noted that, the apparatus provided in this embodiment is only illustrated by dividing the functional modules, and in practical applications, the functions may be distributed by different functional modules according to needs, that is, the internal structure is divided into different functional modules to complete all or part of the functions described above.
Example 3
The embodiment discloses a tea moisture content detection method, which comprises the following steps:
(1) and obtaining the tea water content detection model obtained by the tea water content detection model construction method in the embodiment 1.
(2) If the tea moisture content detection model is a first regression model MA, then:
the method comprises the steps that hyperspectral data of the front side of tea leaves to be detected are obtained for the tea leaves to be detected, the hyperspectral data are input into a first regression model MA, and the moisture content of the tea leaves to be detected is obtained through calculation of the first regression model MA;
if the tea moisture content detection model is a second regression model MB, then:
acquiring hyperspectral data of the reverse side of the tea to be detected aiming at the tea leaves to be detected, inputting the hyperspectral data into a second regression model MB, and calculating by using the second regression model MB to obtain the water content of the tea to be detected;
if the tea moisture content detection model is a third regression model MC, then:
for the tea leaves to be detected, acquiring the average value of the front hyperspectral data and the back hyperspectral data of the tea leaves to be detected, inputting the average value into a third regression model MC, and calculating through the third regression model MC to obtain the water content of the tea leaves to be detected;
if the tea moisture content detection model is a fourth regression model MD, then:
and aiming at the tea leaves to be detected, randomly acquiring the hyperspectral data of the front side or the hyperspectral data of the back side of the tea leaves to be detected, inputting the data into a fourth regression model MD, and calculating through the fourth regression model MD to obtain the water content of the tea leaves to be detected.
Example 4
The embodiment discloses a tea moisture content detection system, as shown in fig. 3, comprising an image acquisition device and an upper computer; wherein:
the image acquisition equipment is used for acquiring hyperspectral data of the tea leaves; the model of the image acquisition device and the manner of acquiring the hyperspectral data may be referred to in embodiment 1, and are not described herein again.
And the upper computer is used for executing the tea water content detection model construction method in the embodiment 1 and/or executing the tea water content detection method in the embodiment 3.
Wherein:
the tea moisture content detection model construction method described in embodiment 1 is executed as follows:
collecting hyperspectral data of tea leaves serving as samples by a hyperspectral image collection system;
acquiring hyperspectral data of the front surface and the back surface of each tea leaf sample;
for each tea leaf sample, acquiring the water content obtained by drying treatment;
fitting and establishing a first regression model MA through the hyperspectral data of the front side of each tea leaf sample and the water content of each tea leaf sample;
fitting and establishing a second regression model MB through the hyperspectral data of the reverse side of each tea leaf sample and the water content of each tea leaf sample;
aiming at each tea leaf sample, calculating average hyperspectral data through the front hyperspectral data and the back hyperspectral data of the tea leaf sample, and fitting and establishing a third regression model MC through the average hyperspectral data of each tea leaf sample and the water content of each tea leaf sample;
fitting and establishing a fourth regression model MD through the hyperspectral data of the front side of each tea leaf sample, the water content of each tea leaf sample, the hyperspectral data of the back side of each tea leaf sample and the water content of each tea leaf sample;
verifying the accuracy of the first regression model MA, the second regression model MB, the third regression model MC and the fourth regression model MD, and taking the model with the highest accuracy as a tea water content detection model.
The method for detecting the moisture content of the tea leaves in the embodiment 3 is implemented as follows:
acquiring a tea water content detection model acquired by the tea water content detection model construction method in the embodiment 1;
if the tea moisture content detection model is a first regression model MA, then:
the method comprises the steps that hyperspectral data of the front side of tea leaves to be detected are obtained for the tea leaves to be detected, the hyperspectral data are input into a first regression model MA, and the moisture content of the tea leaves to be detected is obtained through calculation of the first regression model MA;
if the tea moisture content detection model is a second regression model MB, then:
acquiring hyperspectral data of the reverse side of the tea to be detected aiming at the tea leaves to be detected, inputting the hyperspectral data into a second regression model MB, and calculating by using the second regression model MB to obtain the water content of the tea to be detected;
if the tea moisture content detection model is a third regression model MC, then:
for the tea leaves to be detected, acquiring the average value of the front hyperspectral data and the back hyperspectral data of the tea leaves to be detected, inputting the average value into a third regression model MC, and calculating through the third regression model MC to obtain the water content of the tea leaves to be detected;
if the tea moisture content detection model is a fourth regression model MD, then:
and aiming at the tea leaves to be detected, randomly acquiring the hyperspectral data of the front side or the hyperspectral data of the back side of the tea leaves to be detected, inputting the data into a fourth regression model MD, and calculating through the fourth regression model MD to obtain the water content of the tea leaves to be detected.
In this embodiment, the upper computer may be a computer, a server, or other devices.
Example 5
This embodiment discloses a storage medium storing a program which, when executed by a processor, implements the tea moisture content detection model construction method described in embodiment 1 and/or implements the tea moisture content detection method described in embodiment 3.
Wherein:
the method for constructing the tea moisture content detection model according to embodiment 1 is implemented as follows:
collecting hyperspectral data of tea leaves serving as samples by a hyperspectral image collection system;
acquiring hyperspectral data of the front surface and the back surface of each tea leaf sample;
for each tea leaf sample, acquiring the water content obtained by drying treatment;
fitting and establishing a first regression model MA through the hyperspectral data of the front side of each tea leaf sample and the water content of each tea leaf sample;
fitting and establishing a second regression model MB through the hyperspectral data of the reverse side of each tea leaf sample and the water content of each tea leaf sample;
aiming at each tea leaf sample, calculating average hyperspectral data through the front hyperspectral data and the back hyperspectral data of the tea leaf sample, and fitting and establishing a third regression model MC through the average hyperspectral data of each tea leaf sample and the water content of each tea leaf sample;
fitting and establishing a fourth regression model MD through the hyperspectral data of the front side of each tea leaf sample, the water content of each tea leaf sample, the hyperspectral data of the back side of each tea leaf sample and the water content of each tea leaf sample;
verifying the accuracy of the first regression model MA, the second regression model MB, the third regression model MC and the fourth regression model MD, and taking the model with the highest accuracy as a tea water content detection model.
The method for detecting the moisture content of the tea leaves in the embodiment 3 is realized as follows:
acquiring a tea water content detection model acquired by the tea water content detection model construction method in the embodiment 1;
if the tea moisture content detection model is a first regression model MA, then:
the method comprises the steps that hyperspectral data of the front side of tea leaves to be detected are obtained for the tea leaves to be detected, the hyperspectral data are input into a first regression model MA, and the moisture content of the tea leaves to be detected is obtained through calculation of the first regression model MA;
if the tea moisture content detection model is a second regression model MB, then:
acquiring hyperspectral data of the reverse side of the tea to be detected aiming at the tea leaves to be detected, inputting the hyperspectral data into a second regression model MB, and calculating by using the second regression model MB to obtain the water content of the tea to be detected;
if the tea moisture content detection model is a third regression model MC, then:
for the tea leaves to be detected, acquiring the average value of the front hyperspectral data and the back hyperspectral data of the tea leaves to be detected, inputting the average value into a third regression model MC, and calculating through the third regression model MC to obtain the water content of the tea leaves to be detected;
if the tea moisture content detection model is a fourth regression model MD, then:
and aiming at the tea leaves to be detected, randomly acquiring the hyperspectral data of the front side or the hyperspectral data of the back side of the tea leaves to be detected, inputting the data into a fourth regression model MD, and calculating through the fourth regression model MD to obtain the water content of the tea leaves to be detected.
In this embodiment, the storage medium may be a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a Random Access Memory (RAM), a usb disk, a removable hard disk, or other media.
Example 6
The embodiment discloses a computing device, which comprises a processor and a memory for storing a program executable by the processor, and is characterized in that when the processor executes the program stored in the memory, the method for constructing the tea moisture content detection model in any one of embodiment 1 is implemented, and/or the method for detecting the tea moisture content in embodiment 3 is implemented.
The following were used:
wherein:
the tea moisture content detection model construction method described in embodiment 1 is executed as follows:
collecting hyperspectral data of tea leaves serving as samples by a hyperspectral image collection system;
acquiring hyperspectral data of the front surface and the back surface of each tea leaf sample;
for each tea leaf sample, acquiring the water content obtained by drying treatment;
fitting and establishing a first regression model MA through the hyperspectral data of the front side of each tea leaf sample and the water content of each tea leaf sample;
fitting and establishing a second regression model MB through the hyperspectral data of the reverse side of each tea leaf sample and the water content of each tea leaf sample;
aiming at each tea leaf sample, calculating average hyperspectral data through the front hyperspectral data and the back hyperspectral data of the tea leaf sample, and fitting and establishing a third regression model MC through the average hyperspectral data of each tea leaf sample and the water content of each tea leaf sample;
fitting and establishing a fourth regression model MD through the hyperspectral data of the front side of each tea leaf sample, the water content of each tea leaf sample, the hyperspectral data of the back side of each tea leaf sample and the water content of each tea leaf sample;
verifying the accuracy of the first regression model MA, the second regression model MB, the third regression model MC and the fourth regression model MD, and taking the model with the highest accuracy as a tea water content detection model.
The method for detecting the moisture content of the tea leaves in the embodiment 3 is implemented as follows:
acquiring a tea water content detection model acquired by the tea water content detection model construction method in the embodiment 1;
if the tea moisture content detection model is a first regression model MA, then:
the method comprises the steps that hyperspectral data of the front side of tea leaves to be detected are obtained for the tea leaves to be detected, the hyperspectral data are input into a first regression model MA, and the moisture content of the tea leaves to be detected is obtained through calculation of the first regression model MA;
if the tea moisture content detection model is a second regression model MB, then:
acquiring hyperspectral data of the reverse side of the tea to be detected aiming at the tea leaves to be detected, inputting the hyperspectral data into a second regression model MB, and calculating by using the second regression model MB to obtain the water content of the tea to be detected;
if the tea moisture content detection model is a third regression model MC, then:
for the tea leaves to be detected, acquiring the average value of the front hyperspectral data and the back hyperspectral data of the tea leaves to be detected, inputting the average value into a third regression model MC, and calculating through the third regression model MC to obtain the water content of the tea leaves to be detected;
if the tea moisture content detection model is a fourth regression model MD, then:
and aiming at the tea leaves to be detected, randomly acquiring the hyperspectral data of the front side or the hyperspectral data of the back side of the tea leaves to be detected, inputting the data into a fourth regression model MD, and calculating through the fourth regression model MD to obtain the water content of the tea leaves to be detected.
In this embodiment, the computing device may be a desktop computer, a notebook computer, a smart phone, a PDA handheld terminal, a tablet computer, or other terminal devices.
The above embodiments are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents thereof, and all such changes, modifications, substitutions, combinations, and simplifications are intended to be included in the scope of the present invention.
Claims (10)
1. A tea moisture content detection model construction method is characterized by comprising the following steps:
collecting hyperspectral data of tea leaves serving as samples by a hyperspectral image collection system;
acquiring hyperspectral data of the front surface and the back surface of each tea leaf sample;
for each tea leaf sample, acquiring the water content obtained by drying treatment;
fitting and establishing a first regression model MA through the hyperspectral data of the front side of each tea leaf sample and the water content of each tea leaf sample;
fitting and establishing a second regression model MB through the hyperspectral data of the reverse side of each tea leaf sample and the water content of each tea leaf sample;
aiming at each tea leaf sample, calculating average hyperspectral data through the front hyperspectral data and the back hyperspectral data of the tea leaf sample, and fitting and establishing a third regression model MC through the average hyperspectral data of each tea leaf sample and the water content of each tea leaf sample;
fitting and establishing a fourth regression model MD through the hyperspectral data of the front side of each tea leaf sample, the water content of each tea leaf sample, the hyperspectral data of the back side of each tea leaf sample and the water content of each tea leaf sample;
verifying the accuracy of the first regression model MA, the second regression model MB, the third regression model MC and the fourth regression model MD, and taking the model with the highest accuracy as a tea water content detection model.
2. The tea leaf moisture content detection model building method according to claim 1, characterized by collecting hyperspectral data by a hyperspectral image collection system when a certain number of tea leaf samples are arranged in a sample pool;
the hyperspectral data acquired by the hyperspectral image acquisition system each time is ax×ay×azIn which axIndicating high lightNumber of pixels in horizontal direction of camera in spectral image acquisition system, ayRepresents the total number of points of the wavelength 363-1025 nm, azAnd the total scanning line number of the camera in the hyperspectral image acquisition system is shown.
3. The tea leaf moisture content detection model building method according to claim 2, characterized in that the hyperspectral data of the front side or the back side of each tea leaf sample in the sample pool is extracted from the hyperspectral data acquired by the hyperspectral image acquisition system each time, wherein when the tea leaf sample is right side up, the hyperspectral data of the front side of the tea leaf sample is extracted, and when the tea leaf sample is reverse side up, the hyperspectral data of the back side of the tea leaf sample is extracted, specifically as follows:
SA, selecting a spectral image with the wavelength of 580nm from three-dimensional hyperspectral data;
SB, converting the spectral image with the wavelength of 580nm into a binary image, and acquiring the edge space position information of each tea leaf blade discharged in the sample cell through the binary image;
SC, correspondingly extracting hyperspectral data of each tea leaf sample from the three-dimensional hyperspectral data according to the edge space position information of each tea leaf sample;
and SD, averaging the acquired hyperspectral data of each tea leaf sample to obtain average spectral data serving as final hyperspectral data of the tea leaf sample.
4. The tea leaf moisture content detection model building method according to claim 1, further comprising: performing black-and-white plate correction processing and hyperspectral data preprocessing on hyperspectral data acquired by a hyperspectral image acquisition system;
wherein:
performing black and white plate correction on the hyperspectral data by the following formula:
in the formula I0The hyperspectral data after the white board correction is acquired, I is the original hyperspectral data acquired by the original spectrum hyperspectral image acquisition system, IWAverage spectral data for white board, IDAverage spectral data for the blackboard;
the post-correction hyperspectral data is preprocessed by decentralization, standard normal transformation, multivariate scatter correction, first derivative application of Savitzky-Golay smoothing, or second derivative application of Savitzky-Golay smoothing.
5. The method for constructing a tea water content detection model according to claim 1, wherein for each tea leaf sample, the process of obtaining the water content by a drying method is as follows:
sa, firstly, placing a clean bottle without a cover in a blast drier at 105 ℃ for drying for 1.0 h;
sb, measuring and recording the mass of the dried front piece of the tea leaf sample by using an AL analytical balance with the precision of 0.001g, putting the tea leaf sample into a bottle with a corresponding label, drying the tea leaf sample in a blast drier at 105 ℃ for 3.0h, and recording the mass of the tea leaf after cooling for 0.5 h;
sc, repeating the step Sb for multiple times, and entering a step Sd if the difference between the quality of the tea leaf sample measured after the step Sb is executed for the current time and the quality of the tea leaf sample measured after the step Sb is executed for the last time is not more than 2 mg;
sd, calculating the water content of the tea leaf sample by the following formula
The water content is (m 1-m 2)/m1 multiplied by 100 percent;
m1 is the mass of the tea leaves before drying treatment, and m2 is the mass of the tea leaf sample after the last drying.
6. The utility model provides a tealeaves moisture content detects model construction equipment which characterized in that includes:
the system comprises a first acquisition module, a second acquisition module and a control module, wherein the first acquisition module is used for acquiring hyperspectral data of tea leaves serving as samples through a hyperspectral image acquisition system; acquiring hyperspectral data of the front surface and the back surface of each tea leaf sample;
the second acquisition module is used for acquiring the moisture content of each tea leaf sample obtained by drying treatment;
the first regression model establishing module is used for fitting and establishing a first regression model MA through the hyperspectral data and the water content of the front side of each tea leaf;
the second regression model establishing module is used for fitting and establishing a second regression model MB through the hyperspectral data of the reverse side of each tea leaf sample and the water content of each tea leaf sample;
the third regression model establishing module is used for calculating average hyperspectral data according to the front hyperspectral data and the back hyperspectral data of each tea leaf sample, and fitting and establishing a third regression model MC according to the average hyperspectral data of each tea leaf sample and the water content of each tea leaf sample;
the fourth regression model establishing module is used for fitting and establishing a fourth regression model MD through the hyperspectral data of the front surface of each tea leaf sample and the moisture content of each tea leaf sample, the hyperspectral data of the front surface of each tea leaf sample and the moisture content of each tea leaf sample;
and the verification module is used for verifying the accuracy of the first regression model MA, the second regression model MB, the third regression model MC and the fourth regression model MD and taking the model with the highest accuracy as the tea water content detection model.
7. A method for detecting the moisture content of tea is characterized by comprising the following steps:
obtaining a tea water content detection model obtained by the tea water content detection model construction method according to claims 1-4;
if the tea moisture content detection model is a first regression model MA, then:
the method comprises the steps that hyperspectral data of the front side of tea leaves to be detected are obtained for the tea leaves to be detected, the hyperspectral data are input into a first regression model MA, and the moisture content of the tea leaves to be detected is obtained through calculation of the first regression model MA;
if the tea moisture content detection model is a second regression model MB, then:
acquiring hyperspectral data of the reverse side of the tea to be detected aiming at the tea leaves to be detected, inputting the hyperspectral data into a second regression model MB, and calculating by using the second regression model MB to obtain the water content of the tea to be detected;
if the tea moisture content detection model is a third regression model MC, then:
for the tea leaves to be detected, acquiring the average value of the front hyperspectral data and the back hyperspectral data of the tea leaves to be detected, inputting the average value into a third regression model MC, and calculating through the third regression model MC to obtain the water content of the tea leaves to be detected;
if the tea moisture content detection model is a fourth regression model MD, then:
and aiming at the tea leaves to be detected, randomly acquiring the hyperspectral data of the front side or the hyperspectral data of the back side of the tea leaves to be detected, inputting the data into a fourth regression model MD, and calculating through the fourth regression model MD to obtain the water content of the tea leaves to be detected.
8. A tea moisture content detection system is characterized by comprising image acquisition equipment and an upper computer; wherein:
the image acquisition equipment is used for acquiring hyperspectral data of the tea leaves;
the upper computer is used for executing the tea water content detection model construction method of any one of claims 1-4 and/or the tea water content detection method of claim 7.
9. A storage medium storing a program for implementing the tea water content detection model construction method according to any one of claims 1 to 4 and/or the tea water content detection method according to claim 7 when the program is executed by a processor.
10. A computing device comprising a processor and a memory for storing a program executable by the processor, wherein the processor implements the tea moisture content detection model building method according to any one of claims 1 to 4 and/or implements the tea moisture content detection method according to claim 7 when executing the program stored in the memory.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011283346.9A CN112504977A (en) | 2020-11-17 | 2020-11-17 | Tea water content detection method and model construction method, medium and equipment thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011283346.9A CN112504977A (en) | 2020-11-17 | 2020-11-17 | Tea water content detection method and model construction method, medium and equipment thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112504977A true CN112504977A (en) | 2021-03-16 |
Family
ID=74956429
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011283346.9A Pending CN112504977A (en) | 2020-11-17 | 2020-11-17 | Tea water content detection method and model construction method, medium and equipment thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112504977A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113096747A (en) * | 2021-04-12 | 2021-07-09 | 江苏丰尚智能科技有限公司 | Method and device for predicting moisture content of discharged material of dryer and computer equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102721651A (en) * | 2012-06-18 | 2012-10-10 | 浙江大学 | Detection method and system of water content in plant leaf based on multispectral image |
| CN103344575A (en) * | 2013-07-01 | 2013-10-09 | 江南大学 | Hyperspectral-image-technology-based multi-quality nondestructive testing method for dried green soybeans |
| CN103389255A (en) * | 2013-08-05 | 2013-11-13 | 浙江大学 | Method for predicting water content of tea on basis of near-infrared hyperspectral textural feature modeling |
| CN103744777A (en) * | 2013-12-26 | 2014-04-23 | 浙江大学 | Detection method and purpose for detecting water content of tea by same |
| CN110108648A (en) * | 2019-04-30 | 2019-08-09 | 深圳市太赫兹科技创新研究院有限公司 | A kind of discrimination method and identification system of dried orange peel |
| CN110320164A (en) * | 2019-06-28 | 2019-10-11 | 华南农业大学 | A kind of method for building up of romaine lettuce total nitrogen content EO-1 hyperion inverse model and its application |
| CN110333195A (en) * | 2019-07-15 | 2019-10-15 | 北华航天工业学院 | Water content in plant leaf detection method and device |
-
2020
- 2020-11-17 CN CN202011283346.9A patent/CN112504977A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102721651A (en) * | 2012-06-18 | 2012-10-10 | 浙江大学 | Detection method and system of water content in plant leaf based on multispectral image |
| CN103344575A (en) * | 2013-07-01 | 2013-10-09 | 江南大学 | Hyperspectral-image-technology-based multi-quality nondestructive testing method for dried green soybeans |
| CN103389255A (en) * | 2013-08-05 | 2013-11-13 | 浙江大学 | Method for predicting water content of tea on basis of near-infrared hyperspectral textural feature modeling |
| CN103744777A (en) * | 2013-12-26 | 2014-04-23 | 浙江大学 | Detection method and purpose for detecting water content of tea by same |
| CN110108648A (en) * | 2019-04-30 | 2019-08-09 | 深圳市太赫兹科技创新研究院有限公司 | A kind of discrimination method and identification system of dried orange peel |
| CN110320164A (en) * | 2019-06-28 | 2019-10-11 | 华南农业大学 | A kind of method for building up of romaine lettuce total nitrogen content EO-1 hyperion inverse model and its application |
| CN110333195A (en) * | 2019-07-15 | 2019-10-15 | 北华航天工业学院 | Water content in plant leaf detection method and device |
Non-Patent Citations (2)
| Title |
|---|
| 刘震等: "《水土保持监测技术》", 31 July 2004, 中国大地出版社 * |
| 吴伟斌等: "基于高光谱的茶叶含水量检测模型建立与试验研究", 《河南农业大学学报》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113096747A (en) * | 2021-04-12 | 2021-07-09 | 江苏丰尚智能科技有限公司 | Method and device for predicting moisture content of discharged material of dryer and computer equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP2511680A2 (en) | Optimized orthonormal system and method for reducing dimensionality of hyperspectral Images | |
| CN100357725C (en) | Method and device for rapidly detecting tenderness of beef utilizing near infrared technology | |
| CN112504977A (en) | Tea water content detection method and model construction method, medium and equipment thereof | |
| CN112974303B (en) | Hyperspectrum-based fruit quality detection method, device and medium | |
| CN108537106A (en) | Fingerprint detection method and circuit thereof | |
| CN112697984A (en) | Fruit defect nondestructive testing method based on neural network and fruit grading method | |
| RU2458397C1 (en) | Method of searching for and recognising objects on digital images | |
| CN111879709A (en) | Method and device for detecting spectral reflectivity of lake water body | |
| CN113390799B (en) | Method for identifying and detecting stems in tobacco leaves | |
| CN114002203A (en) | Method and device for analyzing content of wood components based on Raman spectrum | |
| CN111898314B (en) | Lake water parameter inspection method and device, electronic equipment and storage medium | |
| CN111669575B (en) | Method, system, electronic device, medium and terminal for testing image processing effect | |
| CN112964719B (en) | Hyperspectrum-based food fructose detection method and device | |
| CN115406860A (en) | Rapid yellow dragon disease detection device and method based on modeling comparison | |
| WO2021195817A1 (en) | Method for extracting spectral information of object to be detected | |
| CN115170548A (en) | Leather defect automatic detection method and device based on unsupervised learning | |
| CN111256609B (en) | Method and device for detecting USB interface depth | |
| Liu et al. | Multispectral LiDAR point cloud highlight removal based on color information | |
| CN112415014A (en) | Copper foil defect detection method and medium | |
| CN111914741A (en) | House property certificate identification method, device and equipment | |
| CN101354357A (en) | Method for analyzing micro-density image of tree annual ring | |
| CN117409011B (en) | High-voltage sleeve surface pollution monitoring method and system based on target identification | |
| CN114813631B (en) | Wheat variety purity detection method and device and electronic equipment | |
| CN117054372B (en) | Tea quality grade detection method and system based on NIRS and CV | |
| CN117540146A (en) | Sampler for investigating foreign invasive species and control method thereof |
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: 20210316 |
|
| RJ01 | Rejection of invention patent application after publication |