CN113237836A - Flue-cured tobacco leaf moisture content estimation method based on hyperspectral image - Google Patents
Flue-cured tobacco leaf moisture content estimation method based on hyperspectral image Download PDFInfo
- Publication number
- CN113237836A CN113237836A CN202110504440.0A CN202110504440A CN113237836A CN 113237836 A CN113237836 A CN 113237836A CN 202110504440 A CN202110504440 A CN 202110504440A CN 113237836 A CN113237836 A CN 113237836A
- Authority
- CN
- China
- Prior art keywords
- flue
- tobacco leaf
- moisture content
- cured tobacco
- data set
- 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
- 235000002637 Nicotiana tabacum Nutrition 0.000 title claims abstract description 126
- 238000000034 method Methods 0.000 title claims abstract description 46
- 241000208125 Nicotiana Species 0.000 title claims abstract 44
- 238000001228 spectrum Methods 0.000 claims abstract description 20
- 238000012549 training Methods 0.000 claims abstract description 14
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 14
- 238000012545 processing Methods 0.000 claims abstract description 12
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 244000061176 Nicotiana tabacum Species 0.000 claims description 82
- 230000003595 spectral effect Effects 0.000 claims description 23
- 238000012360 testing method Methods 0.000 claims description 8
- 238000000605 extraction Methods 0.000 claims description 7
- 230000003044 adaptive effect Effects 0.000 claims description 6
- 230000002860 competitive effect Effects 0.000 claims description 6
- 238000002790 cross-validation Methods 0.000 claims description 6
- 238000005303 weighing Methods 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 4
- 102000004190 Enzymes Human genes 0.000 claims description 3
- 108090000790 Enzymes Proteins 0.000 claims description 3
- 229910052736 halogen Inorganic materials 0.000 claims description 3
- 150000002367 halogens Chemical class 0.000 claims description 3
- 238000003709 image segmentation Methods 0.000 claims description 3
- 230000000877 morphologic effect Effects 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000002310 reflectometry Methods 0.000 claims description 3
- 238000010200 validation analysis Methods 0.000 claims description 3
- 238000000643 oven drying Methods 0.000 claims description 2
- 238000012795 verification Methods 0.000 claims description 2
- 238000001514 detection method Methods 0.000 abstract description 6
- 239000000126 substance Substances 0.000 abstract description 6
- 238000000691 measurement method Methods 0.000 abstract description 2
- 238000012544 monitoring process Methods 0.000 description 4
- 238000001035 drying Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000000701 chemical imaging Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 244000141359 Malus pumila Species 0.000 description 1
- 240000007594 Oryza sativa Species 0.000 description 1
- 235000007164 Oryza sativa Nutrition 0.000 description 1
- 235000021307 Triticum Nutrition 0.000 description 1
- 244000098338 Triticum aestivum Species 0.000 description 1
- 240000008042 Zea mays Species 0.000 description 1
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 description 1
- 235000002017 Zea mays subsp mays Nutrition 0.000 description 1
- 235000021016 apples Nutrition 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 235000005822 corn Nutrition 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000003306 harvesting Methods 0.000 description 1
- 230000000873 masking effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 235000009566 rice Nutrition 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
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
- G06T5/30—Erosion or dilatation, e.g. thinning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/155—Segmentation; Edge detection involving morphological operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Chemical & Material Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
- Manufacture Of Tobacco Products (AREA)
Abstract
The invention provides a flue-cured tobacco leaf moisture content estimation method based on hyperspectral images, which is used for solving the problems that the existing moisture measurement method is long in time consumption and damages the chemical components and the tissue structure of tobacco leaves. The method comprises the following steps: collecting hyperspectral images of tobacco leaves of the flue-cured tobacco, determining the moisture content by an oven method and establishing a tobacco moisture content data set; preprocessing the hyperspectral image and extracting an interested area; extracting characteristic wavelengths of water related spectrums in the region of interest by using a CARS algorithm; performing Mask calibration on the image corresponding to the characteristic wavelength to obtain a tobacco leaf spectrum data set; and (3) adopting a support vector regression algorithm to construct a flue-cured tobacco leaf moisture estimation model, training the model, and verifying according to the tobacco leaf moisture content data set. According to the method, the sample is not required to be processed through image data processing, the integrity of the sample can be ensured, the operation is simple, the detection result is rapid and accurate, and the moisture detection efficiency of the flue-cured tobacco leaves is effectively improved.
Description
Technical Field
The invention relates to the technical field of flue-cured tobacco growth monitoring, in particular to a flue-cured tobacco moisture content estimation method based on hyperspectral images.
Background
The flue-cured tobacco is used as an economic crop needing water, the moisture content of the flue-cured tobacco leaf directly influences the cell activity, the organ and tissue functions and the physical characteristics in the tobacco leaf, certain influence is generated on the chemical components, the aroma components and the quality of the tobacco leaf after curing, and the change of the moisture content of the tobacco leaf in the mature period and after harvesting is an important basis for adjusting the curing process.
At present, the method for measuring the moisture of the flue-cured tobacco leaves is mainly an oven method, the tobacco leaves with known weight are put into an oven, the water is removed for 20 minutes at 105 ℃, the tobacco leaves are dried to constant weight at 70 ℃, and the moisture content of the tobacco leaves is calculated according to the weight difference of the tobacco leaves before and after drying. The following problems are involved in the moisture measurement using the oven method: (1) cannot be directly measured; (2) the measuring process takes longer time; (3) in the process of de-enzyming and drying tobacco leaves, the chemical components and the tissue structure of the tobacco leaves can be damaged in the high-temperature environment.
The hyperspectral imaging technology is derived from the development of remote sensing technology at the end of the last century, and is gradually applied to the fields of biological parameters, chemical component prediction and the like due to the characteristics of rapid and lossless detection and simultaneous display of spectral information and image information of a sample. At present, hyperspectral imaging technology is applied to different degrees in monitoring of crops such as wheat, corn, rice, apples and the like, a good effect is achieved, the research of a hyperspectral monitoring model in the field of tobacco is still in a primary stage, and no moisture monitoring model suitable for flue-cured tobacco exists.
Disclosure of Invention
The invention provides a flue-cured tobacco leaf moisture content estimation method based on a hyperspectral image, aiming at solving the problems of long time consumption and self-loss of tobacco leaves in the moisture measurement of the tobacco leaves, aiming at solving the technical problems that the existing flue-cured tobacco leaf moisture measurement method consumes long time and damages the chemical components and the tissue structure of the tobacco leaves in a high-temperature environment.
In order to achieve the purpose, the technical scheme of the invention is realized as follows: a flue-cured tobacco leaf moisture content estimation method based on hyperspectral images comprises the following steps:
s1, sample collection: collecting hyperspectral images of tobacco leaves of flue-cured tobaccos, determining the moisture content of the tobacco leaves by an oven method, and establishing a tobacco moisture content data set;
s2, data processing: preprocessing the spectral data of the hyperspectral image acquired in the step S1, and extracting an interested area of the preprocessed hyperspectral image;
s3, characteristic wave band extraction: extracting characteristic wavelengths of the moisture correlation spectrum in the region of interest in the step S2 by adopting a competitive adaptive re-weighting algorithm;
s4, performing Mask calibration on the image corresponding to the characteristic wavelength extracted in the step S3 to obtain a tobacco leaf spectrum data set corresponding to the characteristic wavelength;
s5, constructing a moisture prediction model: and (3) constructing a flue-cured tobacco leaf moisture estimation model by adopting a support vector regression algorithm, training the flue-cured tobacco leaf moisture estimation model by utilizing the tobacco leaf moisture content data set and the tobacco leaf spectrum data set, and verifying the flue-cured tobacco leaf moisture estimation model according to the tobacco leaf moisture content data set.
The flue-cured tobacco leaves collected in the step S1 are from tobacco plants of different varieties, different parts and different maturity, wherein the maturity comprises under-mature, proper-mature and over-mature, and the parts comprise upper leaves, middle leaves and lower leaves.
The parameters of the hyperspectral image acquisition system for acquiring the hyperspectral image in the step S1 are as follows: the object distance is 1m, the wavelength range is 400-1000 nm, the spectrum distance is 2.34nm, and the light source is a halogen lamp light source; the hyperspectral image acquisition system corrects the black board and the white board before acquiring each time.
The method for establishing the tobacco leaf moisture content data set comprises the following steps:
collecting flue-cured tobacco leaf samples of different varieties, different parts and different maturity, weighing on a balance to obtain the fresh weight of the leaves, and accurately obtaining the result to two positions after a decimal point;
deactivating enzyme in oven at 105 deg.C for 20 min, oven drying at 70 deg.C to constant weight, weighing and recording dry weight of leaf, and collecting fresh weight and dry weight of leafDifference, calculating the water content of the tobacco leaf
And obtaining a tobacco leaf water content data set.
The preprocessing method in step S2 is any one or more of a first derivative algorithm, a second derivative algorithm, a normalization transform, a detrending algorithm, a standard normal transform, or a multivariate scattering correction.
The extraction method of the region of interest comprises the following steps: carrying out gray level processing on the preprocessed hyperspectral image to obtain a gray level image; determining a threshold value according to the gray value, and performing image segmentation: corroding and expanding through morphological processing, eliminating noise, and realizing binarization of a gray level image, wherein the gray level value of a background is 0, and the gray level value of a tobacco leaf area is 255; and positioning the leaf area according to the gray value of the pixel point, and selecting the area with the gray value of 255 as the interested area.
In the step S3, the competitive adaptive re-weighting algorithm uses the regression coefficient of each wavelength as a reference index, obtains a wavelength with a large absolute value of the regression coefficient, eliminates a wavelength with a small absolute value of the regression coefficient, and screens the characteristic wavelength.
The Mask calibration method comprises the following steps: performing mask calibration on the extracted image of each characteristic wavelength based on the region of interest through an OpenCV library function, removing the influence of a background image, and obtaining a mask image; and calculating the average spectral reflectivity of the region of interest corresponding to each characteristic wavelength, and establishing a tobacco leaf spectral data set.
The method for constructing the flue-cured tobacco leaf moisture estimation model comprises the following steps: and (3) fusing the tobacco leaf moisture content data set obtained in the step (S1) and the tobacco leaf spectrum data set obtained in the step (S4) into a sample data set, randomly dividing the sample data set into a training set and a testing set according to the proportion of 80% to 20%, using the screened characteristic wavelength as an input variable of a support vector regression model, using the tobacco leaf moisture content as a target variable, establishing the support vector regression model, and inputting the training set into the support vector regression model to obtain a flue-cured tobacco leaf piece moisture estimation model.
The flue-cured tobacco leaf moisture estimation modelThe verification method comprises the following steps: adopting a KFold function in a sklern library to carry out five-fold cross validation, improving the validation precision of the model, evaluating the model precision by using the result of the five-fold cross validation, and simultaneously respectively calculating the decision coefficients R of a test set and a training set according to the water content predicted value and the actual water content of the model2And root mean square error RSME.
Compared with the prior art, the invention has the beneficial effects that: and extracting moisture related spectral characteristic wave bands from tobacco leaf hyperspectral image data, constructing a moisture prediction model, and reflecting the moisture content of tobacco leaves. According to the method, the sample is not required to be processed through image data processing during detection, the integrity of the sample can be ensured, the operation is simple, the detection result is rapid and accurate, and the moisture detection efficiency of the flue-cured tobacco leaves is effectively improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a schematic flow chart of the present invention.
Fig. 2 is a schematic diagram of region of interest extraction according to the present invention.
FIG. 3 is a hyperspectral image corresponding to the first five optimal characteristic bands after Mask masking is performed.
FIG. 4 is a test diagram of a support vector regression model of spectral features of a principal component image of tobacco leaves according to the invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without inventive effort based on the embodiments of the present invention, are within the scope of the present invention.
As shown in fig. 1, an embodiment of the present invention provides a method for estimating moisture content of flue-cured tobacco lamina based on hyperspectral images, which specifically includes the following steps:
s1, sample collection: collecting hyperspectral images of tobacco leaves of the flue-cured tobacco, determining the moisture content of a tobacco sample by an oven method, and establishing a tobacco moisture content data set.
In order to ensure the reliability of the model and the difference of sample data, the collected flue-cured tobacco leaves come from tobacco plants of different varieties, different parts and different maturity. The maturity includes under-mature, still mature, proper mature and over mature, and the parts include upper leaves, middle leaves and lower leaves. The parameters for collecting the hyperspectral image are as follows: the object distance is 1m, the wavelength range is 400-1000 nm, the spectrum distance is 2.34nm, and the light source is a halogen lamp light source.
In the specific implementation, 100 tobacco leaf samples of different varieties, different parts and different maturity are collected, the fresh weight of the leaves is obtained by weighing on a balance, and the result is accurate to two decimal places (unit g).
Gather the high spectral image, put the tobacco leaf sample and gather spectral data under the high spectral image collection system and be the high spectral image, for preventing to gather in-process collection equipment or light intensity and produce noise influence to the high spectral image, carry out black blank correction before the survey at every turn, the optimal system parameter that the debugging obtained is: the object distance is 1m, the wavelength range is 400-1000 nm, 256 wave bands are totally formed, the spectrum interval is 2.34nm, and the exposure time is 10 ms.
Determining the moisture content of the tobacco leaf sample by an oven method, deactivating enzyme in the oven at 105 ℃ for 20 minutes, drying at 70 ℃ to constant weight, weighing and recording the dry weight of the leaves, and calculating the moisture content of the tobacco leaves according to the difference between the fresh weight of the leaves and the dry weight of the leaves. The relative water content (FMC) of the flue-cured tobacco leaf slices is calculated by the following formula:
s2, data processing: and (4) preprocessing the spectral data of the hyperspectral image acquired in the step (S1) and extracting the region of interest of the preprocessed hyperspectral image.
In order to reduce or eliminate the influence of factors such as collecting instruments, collecting backgrounds and ambient stray light in original spectral information, extract effective spectral information and improve the prediction accuracy of a model, the original spectral data needs to be preprocessed. The method for preprocessing the original spectral data is any one or more of a first derivative algorithm, a second derivative algorithm, a normalization transformation, a detrending algorithm, a standard normal transformation or a multivariate scattering correction.
In specific implementation, the first derivative processing is performed on the original spectrum data, so that baseline drift can be removed, overlapped information is amplified and separated, and spectrum information is enhanced.
As background and noise can affect the hyperspectral image, the tobacco leaves and the background need to be segmented, and the tobacco leaf area is an interested area.
The extraction method of the region of interest comprises the following steps: firstly, the gray level processing is carried out on the preprocessed flue-cured tobacco leaf image to obtain a gray level image, and as shown in fig. 2, a threshold value is determined according to the gray level value to carry out image segmentation. And corroding and expanding through morphological processing, eliminating noise, and realizing binarization of a gray level image, wherein the gray level value of a background is 0, and the gray level value of a tobacco leaf area is 255. And positioning the leaf area according to the gray value of the pixel point, and selecting the area with the gray value of 255 as the interested area. Corresponding spectral information is extracted based on the region of interest, a three-dimensional data block containing image information of all wavelengths and spectral information of all pixel points of the image can be obtained, and external features, physical results and internal chemical components of the sample can be reflected.
S3, characteristic wave band extraction: the characteristic wavelengths of the moisture-related spectrum in the region of interest in step S2 are extracted using a competitive adaptive re-weighting algorithm (CARS). And (3) taking the regression coefficient of each wavelength as a reference index, obtaining the wavelength with a large regression coefficient absolute value, eliminating the wavelength with a small regression coefficient absolute value, and screening the characteristic wavelength.
The characteristic wave band is extracted through a competitive adaptive re-weighting algorithm (CARS). Setting the sampling frequency as 50 times, gradually reducing the reserved number of the wavelength along with the increase of the sampling frequency, and finally obtaining 20 variables through the CARS algorithm screening, wherein the characteristic wavelengths corresponding to the 20 variables are as follows: 408.32nm, 448.45nm, 480.37nm, 504.48nm, 529.46nm, 543.95nm, 578.63nm, 634.72nm, 669.32nm, 698.75nm, 702.67nm, 736.41nm, 767.96nm, 795.48nm, 812.82nm, 835.99nm, 867.77nm, 879.01nm, 910.23nm and 986.14 nm.
And S4, performing Mask calibration on the image corresponding to the characteristic wavelength extracted by the CARS algorithm in the step S3, and extracting a tobacco leaf spectrum data set corresponding to the characteristic wavelength.
The Mask calibration method comprises the following steps: through an OpenCV library function, the image of each characteristic wavelength extracted in step S3 is subjected to mask calibration based on the region of interest in step S2, the influence of the background image is removed, and a mask image is obtained, as shown in fig. 3. And calculating the average spectral reflectivity of the region of interest corresponding to each characteristic wavelength, and establishing a tobacco leaf spectral data set.
S5, constructing a moisture prediction model: and constructing a flue-cured tobacco leaf moisture estimation model by adopting a Support Vector Regression (SVR) algorithm, training the flue-cured tobacco leaf moisture estimation model by utilizing the tobacco leaf moisture content data set and the tobacco leaf spectrum data set, and verifying the flue-cured tobacco leaf moisture estimation model.
Fusing the tobacco leaf moisture content data set and the spectrum data set respectively obtained in the step S1 and the step S4 into a sample data set, and randomly dividing the sample data set into a training set R according to the proportion of 80% to 20%cAnd test set RvAnd establishing a support vector regression model by taking the characteristic wavelength screened by the CARS algorithm as an input variable of the support vector regression model and taking the moisture content of the tobacco leaves as a target variable, and inputting the training set into the support vector regression model to obtain a moisture estimation model of the flue-cured tobacco leaves. The method for verifying the moisture estimation model of the flue-cured tobacco leaves comprises the following steps: adopting a KFold function in a sklern library to carry out five-fold cross validation, improving the validation precision of the model, evaluating the precision of the model by using the result of the five-fold cross validation, and calculating a decision coefficient R of a test set and a training set according to the predicted value of the water content of the model and the actual water content value2And root mean square error RSME.
Calling matplotliAnd b, drawing a graph library, and visualizing the model prediction result, as shown in fig. 4. Training set decision coefficient R of flue-cured tobacco leaf slice moisture estimation modelc 2And root mean square error RSMEc0.947 and 0.221, respectively, the determination coefficient R of the test set2 vAnd root mean square error RSMEvRespectively 0.926 and 0.269. Therefore, the method for identifying the moisture in the flue-cured tobacco leaves has the characteristics of no damage to samples, high accuracy and high efficiency.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (10)
1. A flue-cured tobacco leaf moisture content estimation method based on hyperspectral images is characterized by comprising the following steps:
s1, sample collection: collecting hyperspectral images of tobacco leaves of flue-cured tobaccos, determining the moisture content of the tobacco leaves by an oven method, and establishing a tobacco moisture content data set;
s2, data processing: preprocessing the spectral data of the hyperspectral image acquired in the step S1, and extracting an interested area of the preprocessed hyperspectral image;
s3, characteristic wave band extraction: extracting characteristic wavelengths of the moisture correlation spectrum in the region of interest in the step S2 by adopting a competitive adaptive re-weighting algorithm;
s4, performing Mask calibration on the image corresponding to the characteristic wavelength extracted in the step S3 to obtain a tobacco leaf spectrum data set corresponding to the characteristic wavelength;
s5, constructing a moisture prediction model: and (3) constructing a flue-cured tobacco leaf moisture estimation model by adopting a support vector regression algorithm, training the flue-cured tobacco leaf moisture estimation model by utilizing the tobacco leaf moisture content data set and the tobacco leaf spectrum data set, and verifying the flue-cured tobacco leaf moisture estimation model according to the tobacco leaf moisture content data set.
2. The method for estimating the moisture content of the flue-cured tobacco leaf based on the hyperspectral image as claimed in claim 1, wherein the flue-cured tobacco leaf collected in the step S1 is from tobacco plants of different varieties, different parts and different maturity, wherein the maturity comprises under-mature, proper-mature and over-mature, and the parts comprise upper leaf, middle leaf and lower leaf.
3. The flue-cured tobacco moisture content estimation method based on hyperspectral image according to claim 1, wherein the parameters of the hyperspectral image acquisition system for acquiring the hyperspectral image in the step S1 are as follows: the object distance is 1m, the wavelength range is 400-1000 nm, the spectrum distance is 2.34nm, and the light source is a halogen lamp light source; the hyperspectral image acquisition system corrects the black board and the white board before acquiring each time.
4. The estimation method for the moisture content of the flue-cured tobacco leaves based on the hyperspectral images according to claim 2 is characterized in that the establishment method of the tobacco leaf moisture content data set is as follows:
collecting flue-cured tobacco leaf samples of different varieties, different parts and different maturity, weighing on a balance to obtain the fresh weight of the leaves, and accurately obtaining the result to two positions after a decimal point;
deactivating enzyme in oven at 105 deg.C for 20 min, oven drying at 70 deg.C to constant weight, weighing and recording dry weight of leaf, and calculating water content of tobacco leaf according to difference between fresh weight and dry weight of leaf
And obtaining a tobacco leaf water content data set.
5. The flue-cured tobacco moisture content estimation method based on hyperspectral image according to claim 1 or 4, wherein the preprocessing method in the step S2 is any one or more of a first derivative algorithm, a second derivative algorithm, a normalization transform, a detrending algorithm, a standard normal transform or a multivariate scattering correction.
6. The estimation method of moisture content of flue-cured tobacco leaves based on hyperspectral images according to claim 5 is characterized in that the extraction method of the region of interest is as follows: carrying out gray level processing on the preprocessed hyperspectral image to obtain a gray level image; determining a threshold value according to the gray value, and performing image segmentation: corroding and expanding through morphological processing, eliminating noise, and realizing binarization of a gray level image, wherein the gray level value of a background is 0, and the gray level value of a tobacco leaf area is 255; and positioning the leaf area according to the gray value of the pixel point, and selecting the area with the gray value of 255 as the interested area.
7. The estimation method of moisture content of flue-cured tobacco leaves based on hyperspectral images according to claim 1 or 5 is characterized in that the competitive adaptive re-weighting algorithm in step S3 takes the regression coefficient of each wavelength as a reference index, obtains the wavelength with large absolute value of the regression coefficient, eliminates the wavelength with small absolute value of the regression coefficient, and screens characteristic wavelengths.
8. The estimation method of moisture content of flue-cured tobacco leaves based on hyperspectral images according to claim 7 is characterized in that the Mask calibration method comprises the following steps: performing mask calibration on the extracted image of each characteristic wavelength based on the region of interest through an OpenCV library function, removing the influence of a background image, and obtaining a mask image; and calculating the average spectral reflectivity of the region of interest corresponding to each characteristic wavelength, and establishing a tobacco leaf spectral data set.
9. The flue-cured tobacco leaf moisture content estimation method based on the hyperspectral image according to claim 1 or 8 is characterized in that the flue-cured tobacco leaf moisture estimation model is constructed by the following method: and (3) fusing the tobacco leaf moisture content data set obtained in the step (S1) and the tobacco leaf spectrum data set obtained in the step (S4) into a sample data set, randomly dividing the sample data set into a training set and a testing set according to the proportion of 80% to 20%, using the screened characteristic wavelength as an input variable of a support vector regression model, using the tobacco leaf moisture content as a target variable, establishing the support vector regression model, and inputting the training set into the support vector regression model to obtain a flue-cured tobacco leaf piece moisture estimation model.
10. The flue-cured tobacco leaf moisture content estimation method based on the hyperspectral image according to claim 9 is characterized in that the flue-cured tobacco leaf moisture estimation model verification method comprises the following steps: adopting a KFold function in a sklern library to carry out five-fold cross validation, improving the validation precision of the model, evaluating the model precision by using the result of the five-fold cross validation, and simultaneously respectively calculating the decision coefficients R of a test set and a training set according to the water content predicted value and the actual water content of the model2And root mean square error RSME.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110504440.0A CN113237836A (en) | 2021-05-10 | 2021-05-10 | Flue-cured tobacco leaf moisture content estimation method based on hyperspectral image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110504440.0A CN113237836A (en) | 2021-05-10 | 2021-05-10 | Flue-cured tobacco leaf moisture content estimation method based on hyperspectral image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113237836A true CN113237836A (en) | 2021-08-10 |
Family
ID=77132904
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110504440.0A Pending CN113237836A (en) | 2021-05-10 | 2021-05-10 | Flue-cured tobacco leaf moisture content estimation method based on hyperspectral image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113237836A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114397297A (en) * | 2022-01-19 | 2022-04-26 | 河南中烟工业有限责任公司 | Rapid nondestructive testing method for starch content of flue-cured tobacco |
| CN115049902A (en) * | 2022-05-11 | 2022-09-13 | 华南农业大学 | Citrus leaf water content visual prediction method, system, equipment and storage medium |
| CN115349654A (en) * | 2022-08-23 | 2022-11-18 | 中国烟草总公司郑州烟草研究院 | Calibration method of tobacco leaf baking parameters |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102033043A (en) * | 2010-10-19 | 2011-04-27 | 浙江大学 | Grain moisture content detecting method based on hyperspectral image technology |
| CN102495005A (en) * | 2011-11-17 | 2012-06-13 | 江苏大学 | Method for diagnosing crop water deficit through hyperspectral image technology |
| CN104949927A (en) * | 2015-06-01 | 2015-09-30 | 浙江大学 | Hyperspectral detection method of dried scallop moisture content |
| CN110441248A (en) * | 2019-09-10 | 2019-11-12 | 四川轻化工大学 | The method that pit mud Water Prediction Models based on hyper-spectral image technique are established |
| CN110542658A (en) * | 2019-09-10 | 2019-12-06 | 中国烟草总公司郑州烟草研究院 | tobacco non-smoke substance classification method based on hyperspectral imaging technology |
-
2021
- 2021-05-10 CN CN202110504440.0A patent/CN113237836A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102033043A (en) * | 2010-10-19 | 2011-04-27 | 浙江大学 | Grain moisture content detecting method based on hyperspectral image technology |
| CN102495005A (en) * | 2011-11-17 | 2012-06-13 | 江苏大学 | Method for diagnosing crop water deficit through hyperspectral image technology |
| CN104949927A (en) * | 2015-06-01 | 2015-09-30 | 浙江大学 | Hyperspectral detection method of dried scallop moisture content |
| CN110441248A (en) * | 2019-09-10 | 2019-11-12 | 四川轻化工大学 | The method that pit mud Water Prediction Models based on hyper-spectral image technique are established |
| CN110542658A (en) * | 2019-09-10 | 2019-12-06 | 中国烟草总公司郑州烟草研究院 | tobacco non-smoke substance classification method based on hyperspectral imaging technology |
Non-Patent Citations (3)
| Title |
|---|
| 于旭峰 等: "基于近红外光谱技术的马铃薯叶片含水率高效预测", 光学仪器, vol. 42, no. 4, 31 August 2020 (2020-08-31), pages 7 - 13 * |
| 李玉鹏 等: "烤烟叶片含水量的光谱检测模型研究", 中国农学通报, vol. 29, no. 34, 5 December 2013 (2013-12-05), pages 158 - 161 * |
| 李红 等: "基于高光谱成像技术的生菜冠层含水率检测", 农业机械学报, vol. 52, no. 2, 28 February 2021 (2021-02-28), pages 211 - 217 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114397297A (en) * | 2022-01-19 | 2022-04-26 | 河南中烟工业有限责任公司 | Rapid nondestructive testing method for starch content of flue-cured tobacco |
| CN114397297B (en) * | 2022-01-19 | 2024-01-23 | 河南中烟工业有限责任公司 | Rapid nondestructive testing method for starch content of flue-cured tobacco |
| CN115049902A (en) * | 2022-05-11 | 2022-09-13 | 华南农业大学 | Citrus leaf water content visual prediction method, system, equipment and storage medium |
| CN115049902B (en) * | 2022-05-11 | 2024-05-10 | 华南农业大学 | Visual prediction method, system, equipment and storage medium for water content of citrus blade |
| CN115349654A (en) * | 2022-08-23 | 2022-11-18 | 中国烟草总公司郑州烟草研究院 | Calibration method of tobacco leaf baking parameters |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113237836A (en) | Flue-cured tobacco leaf moisture content estimation method based on hyperspectral image | |
| CN113192117B (en) | Method for improving estimation of leaf area index of early growth stage of wheat based on Sentinel-2 satellite image red edge band | |
| Cogdill et al. | Single-kernel maize analysis by near-infrared hyperspectral imaging | |
| CN110132862B (en) | Construction method and application of disease index special for wheat scab detection | |
| CN109100323B (en) | Nondestructive quantitative evaluation method for transmission spectrum of apple water core disease | |
| Reddy et al. | Accurate histopathology from low signal-to-noise ratio spectroscopic imaging data | |
| CN111062251A (en) | Monitoring method of farmland cotton aphid pest grade model based on unmanned aerial vehicle imaging | |
| CN112348812A (en) | Forest stand age information measuring method and device | |
| CN112881327B (en) | Tobacco leaf SPAD value estimation method based on novel vegetation index | |
| CN110779875B (en) | Method for detecting moisture content of winter wheat ear based on hyperspectral technology | |
| Wang et al. | Determination of moisture content of single maize seed by using long-wave near-infrared hyperspectral imaging (LWNIR) Coupled with UVE-SPA combination variable selection method | |
| Zhaoyong et al. | Detection of moldy core in apples and its symptom types using transmittance spectroscopy | |
| CN112129709A (en) | Apple tree canopy scale nitrogen content diagnosis method | |
| Xu et al. | Nondestructive detection of internal flavor in ‘Shatian’pomelo fruit based on visible/near infrared spectroscopy | |
| Tian et al. | Research on apple origin classification based on variable iterative space shrinkage approach with stepwise regression–support vector machine algorithm and visible‐near infrared hyperspectral imaging | |
| CN114527082A (en) | Sweet corn seed germination prediction method and device | |
| Zhang et al. | Uninformative Biological Variability Elimination in Apple Soluble Solids Content Inspection by Using Fourier Transform Near‐Infrared Spectroscopy Combined with Multivariate Analysis and Wavelength Selection Algorithm | |
| Liu et al. | Estimation of chlorophyll content in maize canopy using wavelet denoising and SVR method | |
| CN113267458A (en) | Method for establishing quantitative prediction model of soluble protein content of sweet potatoes | |
| CN116994126A (en) | Crop leaf area index obtaining method and device based on canopy reflectivity spectrum | |
| Wen et al. | Visualization of chlorophyll content distribution in apple leaves based on hyperspectral imaging technology | |
| CN117541835A (en) | Citrus leaf water stress detection method and system based on hyperspectral imaging and deep learning | |
| CN111751295A (en) | Modeling method and application of wheat powdery mildew severity detection model based on imaging hyperspectral data | |
| Zhou et al. | Hyperspectral imaging technology for detection of moisture content of tomato leaves | |
| CN115049902B (en) | Visual prediction method, system, equipment and storage medium for water content of citrus blade |
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 |