CN113466156A - Wood validity verification method based on visible-near infrared spectrum - Google Patents
Wood validity verification method based on visible-near infrared spectrum Download PDFInfo
- Publication number
- CN113466156A CN113466156A CN202110873658.3A CN202110873658A CN113466156A CN 113466156 A CN113466156 A CN 113466156A CN 202110873658 A CN202110873658 A CN 202110873658A CN 113466156 A CN113466156 A CN 113466156A
- Authority
- CN
- China
- Prior art keywords
- wood
- model
- visible
- near infrared
- spectrum
- 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
- 239000002023 wood Substances 0.000 title claims abstract description 198
- 238000002329 infrared spectrum Methods 0.000 title claims abstract description 42
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000012795 verification Methods 0.000 title claims abstract description 36
- 238000001228 spectrum Methods 0.000 claims description 66
- 239000002245 particle Substances 0.000 claims description 17
- 238000012706 support-vector machine Methods 0.000 claims description 17
- 238000004519 manufacturing process Methods 0.000 claims description 13
- 238000005070 sampling Methods 0.000 claims description 12
- 230000004044 response Effects 0.000 claims description 11
- 238000012360 testing method Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 9
- 238000004458 analytical method Methods 0.000 claims description 5
- 238000011156 evaluation Methods 0.000 claims description 3
- 241000894007 species Species 0.000 description 29
- 239000000523 sample Substances 0.000 description 16
- 241000183024 Populus tremula Species 0.000 description 9
- 230000000694 effects Effects 0.000 description 9
- 238000012545 processing Methods 0.000 description 8
- 238000012937 correction Methods 0.000 description 7
- 238000001514 detection method Methods 0.000 description 5
- 241000544657 Larix gmelinii Species 0.000 description 4
- 241001106462 Ulmus Species 0.000 description 4
- 238000004468 VIS-NIR spectroscopy Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 238000000354 decomposition reaction Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 239000000126 substance Substances 0.000 description 3
- 235000005205 Pinus Nutrition 0.000 description 2
- 241000218602 Pinus <genus> Species 0.000 description 2
- 238000005452 bending Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000002790 cross-validation Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 229920002488 Hemicellulose Polymers 0.000 description 1
- 241000218652 Larix Species 0.000 description 1
- 235000005590 Larix decidua Nutrition 0.000 description 1
- 241001443170 Larix gmelinii var. gmelinii Species 0.000 description 1
- 235000000405 Pinus densiflora Nutrition 0.000 description 1
- 240000008670 Pinus densiflora Species 0.000 description 1
- 241000219000 Populus Species 0.000 description 1
- 238000002835 absorbance Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229920002678 cellulose Polymers 0.000 description 1
- 239000001913 cellulose Substances 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000001066 destructive effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 229920005610 lignin Polymers 0.000 description 1
- 210000001724 microfibril Anatomy 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 150000002894 organic compounds Chemical class 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000010561 standard procedure Methods 0.000 description 1
- 238000005211 surface analysis Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/3103—Atomic absorption analysis
-
- 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
- G01N21/84—Systems specially adapted for particular applications
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a wood validity verification method based on a visible-near infrared spectrum, which comprises the following steps: acquiring visible-near infrared spectrum data of wood to be verified; denoising the visible-near infrared spectrum data by adopting second-generation wavelet-lifting wavelet transform; inputting the visible-near infrared spectrum data or the denoised visible-near infrared spectrum data into a legality verification model; the validity verification model comprises: a wood property prediction model, a wood producing area tracing model and a wood tree species identification model; and outputting the wood property prediction result, the wood origin tracing data and the wood species identification data of the verified wood. The verification method has high prediction precision and strong applicability, and can simultaneously predict wood property information and identify the wood producing area and the tree species type, thereby realizing the authenticity identification of a large amount of wood property information and reducing the illegal transaction rate of wood.
Description
Technical Field
The invention relates to the technical field of visible-near infrared spectroscopy, in particular to a wood validity verification method based on visible-near infrared spectroscopy.
Background
The related data show that more than 50% of the forests around the world are threatened by illegal felling; every year, illegal wood trade accounts for 15% -30% of the total wood trade, so that the price of forest products is greatly reduced, and serious social and ecological problems are caused. At present, an intelligent, efficient and green wood validity verification system is urgently needed to be constructed, so that the international wood validity requirement is met, and the international trade of forest products and the national forestry legislation process are promoted.
The essence of wood legality is to ensure the legality of the raw material sources of wood and forest products, and the verification of the wood legality is mainly carried out through the wood producing area, the tree species and the wood legality information. For example, wood imported from russia, southeast asia, africa, etc. is partly considered "illegal wood";
as a complex natural organic compound, related enterprises can hardly judge the wood property information such as the production area, the tree species, the physical and chemical properties and the like of wood simultaneously, rapidly, accurately and massively. In addition, although the classical traditional detection method meets the requirement of test precision, the following defects still exist:
1. high destructiveness and incapability of realizing batch sample detection
In the traditional laboratory detection method, wood is required to be prepared into a standard sample piece during detection, and then a related national standard method is adopted for determination. The method has the advantages of complex operation, time and labor waste and large destructiveness, and the obtained experimental result is only suitable for the determined sample piece and cannot be used for rapid, accurate and large-batch nondestructive prediction of the material property information of the same unknown sample.
2. The accuracy of the data cannot be ensured
In the aspect of wood species identification, related workers mainly adopt a sensory method for identification, and the judgment result has strong subjectivity. In the aspect of tracing the origin of the wood production area, the method completely depends on circulation information from the upstream to the downstream of an industrial chain, data is uniformly managed by an operator, but a user cannot participate in the whole circulation process, and the wood production area information is easily tampered maliciously. Driven by benefits, part of lawbreakers replace high-price wood with low-price wood with extremely similar appearance; or illegal felling is carried out on the forest areas where felling is forbidden, the forest areas are claimed to be legal woods for authorized felling, and related workers are difficult to quickly and accurately judge the origin of the large amount of woods and the tree species according to own experiences. Therefore, in order to ensure the legality of the traded wood, a method for rapidly, accurately and nondestructively predicting the production area, the tree species and the wood property information of the mass wood is urgently needed.
However, the current research is only directed to single tree species identification, such as public number: CN 109870421A 'an incremental wood tree species classification and identification method based on visible light/near infrared spectrum analysis', and can not realize the origin tracing of wood, the tree species identification and the simultaneous accurate prediction of wood property information.
Disclosure of Invention
The invention aims to solve the defects of the prior art, provides a wood validity verification method based on a visible-near infrared spectrum, which can simultaneously identify the wood producing area, the tree type and the predicted wood property information, has high prediction precision and wide applicability, can realize the authenticity identification of a large amount of wood property information, and reduces the illegal transaction rate of wood.
In order to achieve the purpose, the invention adopts the technical scheme that:
in a first aspect, an embodiment of the present invention provides a method for verifying wood validity based on visible-near infrared spectroscopy, including:
s1, acquiring visible-near infrared spectrum data of the wood to be verified;
s2, denoising the visible-near infrared spectrum data by adopting second-generation wavelet-lifting wavelet transform;
s3, inputting the visible-near infrared spectrum data of the step S1 or the denoised visible-near infrared spectrum data of the step S2 into a validity verification model; the validity verification model comprises: a wood property prediction model, a wood producing area tracing model and a wood tree species identification model;
and S4, outputting the wood property prediction result, the wood origin tracing data and the wood species identification data of the verified wood.
Further, the process of constructing the validity verification model in step S3 includes:
s301, collecting various types of wood samples belonging to different producing areas; respectively scanning the samples for multiple times, and randomly selecting the average spectrum of a preset number of sampling points as the original spectrum of the sample;
s302, denoising the original spectrum by adopting second-generation wavelet-lifting wavelet transform; obtaining wavelet coefficients and a reconstructed spectrum;
s303, inputting the wavelet coefficient and the reconstructed spectrum into a particle swarm-support vector machine model as independent variables; optimizing the particle swarm-support vector machine model by adopting a response surface method to obtain a wood property prediction model;
s304, inputting the original spectrums and the wavelet coefficients belonging to different producing areas into a particle swarm-support vector machine model to obtain a wood producing area tracing model;
s305, inputting the original spectrums and the wavelet coefficients which are distinguished by various types into a particle swarm-support vector machine model to obtain a wood tree species identification model.
Further, in the denoising processing in step S2 and/or step S302, a partial least squares model statistical parameter is used as an evaluation index for denoising; the statistical parameters include: coefficient of determination, root mean square error, and relative analysis error.
Further, the step S301 includes:
collecting various types of wood belonging to different producing areas, and manufacturing disc test pieces with preset lengths at intervals of preset distances from bottom to top along a trunk;
collecting the spectrums of all the disc test pieces by adopting a spectrum scanner, wherein the spectrum collection range is 350-2500 nm; the scanning times are X times, and X is more than or equal to 30;
and randomly selecting a preset number of sampling points for each disc test piece, and taking the average spectrum of the preset number of sampling points as the original spectrum of the sample.
Compared with the prior art, the invention has the following beneficial effects:
the wood validity verification method based on the visible-near infrared spectrum provided by the embodiment of the invention comprises the following steps: acquiring visible-near infrared spectrum data of wood to be verified; denoising the visible-near infrared spectrum data by adopting second-generation wavelet-lifting wavelet transform; inputting the visible-near infrared spectrum data or the denoised visible-near infrared spectrum data into a legality verification model; the validity verification model comprises: a wood property prediction model, a wood producing area tracing model and a wood tree species identification model; and outputting the wood property prediction result, the wood origin tracing data and the wood species identification data of the verified wood. The verification method has high prediction precision and strong applicability, and can simultaneously predict wood property information and identify the wood producing area and the tree species type, thereby realizing the authenticity identification of a large amount of wood property information and reducing the illegal transaction rate of wood.
Drawings
Fig. 1 is a flow chart of a wood validity verification method based on visible-near infrared spectrum provided by an embodiment of the invention.
Fig. 2 is a flowchart of steps for constructing a wood validity verification model according to an embodiment of the present invention.
FIG. 3 is a schematic diagram of a visible-near infrared spectrum of four types of wood collected according to an embodiment of the present invention.
Fig. 4a is a graph of wood origin legitimacy identification results based on the original spectrum according to the embodiment of the present invention.
Fig. 4b is a graph of the wood origin legitimacy identification result based on lifting wavelet coefficients according to the embodiment of the present invention.
Detailed Description
In order to make the technical means, the creation characteristics, the achievement purposes and the effects of the invention easy to understand, the invention is further described with the specific embodiments.
In the description of the present invention, it should be noted that the terms "upper", "lower", "inner", "outer", "front", "rear", "both ends", "one end", "the other end", and the like indicate orientations or positional relationships based on those shown in the drawings, and are only for convenience of description and simplicity of description, but do not indicate or imply that the referred device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and thus, should not be construed as limiting the present invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In the description of the present invention, it is to be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "disposed," "connected," and the like are to be construed broadly, such as "connected," which may be fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
Referring to fig. 1, an embodiment of the present invention provides a method for verifying wood validity based on visible-near infrared spectroscopy, including:
s1, acquiring visible-near infrared spectrum data of the wood to be verified;
s2, denoising the visible-near infrared spectrum data by adopting second-generation wavelet-lifting wavelet transform;
s3, inputting the visible-near infrared spectrum data of the step S1 or the denoised visible-near infrared spectrum data of the step S2 into a validity verification model; the validity verification model comprises: a wood property prediction model, a wood producing area tracing model and a wood tree species identification model;
and S4, outputting the wood property prediction result, the wood origin tracing data and the wood species identification data of the verified wood.
In this embodiment, in step S1, the visible-near infrared spectrum data of the wood to be verified may be obtained by the spectrum collecting instrument; for example, the spectrum collection instrument is a LabSpec Pro portable spectrum scanner produced by American ASD company, the spectrum collection range is 350-2500 nm, the spectrum of the wood to be verified is collected by adopting a strong light probe, the wood to be verified can be scanned for multiple times, then, for example, 3 sampling points can be randomly selected, and the average spectrum of the 3 sampling points is taken as the original spectrum of the wood to be verified.
In step S2, denoising the visible-near infrared spectrum data using a second generation wavelet-lifting wavelet transform. The collected spectrum often contains a certain amount of noise due to interference from the external environment and the background of the instrument. In order to eliminate the influence of noise, the second generation wavelet-lifting wavelet transform is adopted to carry out denoising processing on the collected visible-near infrared spectrum data.
In the steps S3-S4, the visible-near infrared spectrum data (when no noise exists) or the visible-near infrared spectrum data are subjected to lifting wavelet denoising treatment and then input into a legality verification model to obtain a wood property prediction result, wood origin tracing data and wood tree species identification data. The method realizes simultaneous prediction of wood property information and identification of wood producing areas and tree species types based on visible-near infrared spectrum and a constructed validity verification model, thereby realizing authenticity identification of large quantities of wood property information and reducing illegal transaction rate of wood.
Wherein the wood property prediction result comprises: physical composition of wood, chemical composition of wood, mechanical properties of wood and microstructure parameters of wood; and the physical components of wood include: density, moisture content, wood color, etc.; the chemical components of the wood comprise lignin, cellulose, hemicellulose, holocellulose, ash and other data; the mechanical properties of wood include: bending strength, bending modulus of elasticity, etc.; wood microstructure parameters include: microfibril angle, tracheid length, etc.
In the following, the construction process of the validity verification model in step S3 is described in detail with two producing areas of four kinds of wood, which have legal and illegal data, and in order to improve the verification capability of the model, the more original sample data is used, the better the result is.
Referring to fig. 2, steps for constructing a wood validity verification model based on the visible-near infrared spectrum technology are shown. The method comprises the following specific steps:
the method comprises the following steps:
7 aspen standard trees and 11 elm standard trees are respectively collected in a scenic mountain experimental forest farm in Jingyu county, mountain city, Jilin province, 8 pinus dahuricus standard trees and 7 aspen standard trees are collected in a Dayang forest farm of Pinus densiflora Linn forest Burn, Black Longjiang province, and 5cm circular disc test pieces are manufactured every 2m from bottom to top along a trunk, and 280 samples are obtained in total. The wood density is determined according to GB/T1933-2009 Wood Density determination Standard. The spectrum acquisition instrument is a LabSpec Pro portable spectrum scanner produced by American ASD company, the spectrum acquisition range is 350-2500 nm, the spectrum of each disc test piece is acquired by adopting a strong light probe, and the sample scanning frequency is 30; and randomly selecting 3 sampling points for each disc test piece, and taking the average spectrum of the 3 sampling points as the original spectrum of the sample. Suppose that the wood in the producing area of Heilongjiang province is legal wood and the wood in the producing area of Jilin province is illegal wood. The same kind of wood (aspen) is collected at two production places simultaneously to verify whether a merchant replaces the same kind of legal wood with the wood of an illegal production place. The wood properties prediction is exemplified by wood density. The obtained spectra of the respective trees are shown in FIG. 3, and the abscissa represents the wavelength and the ordinate represents the absorbance.
Step two:
random sampling is adopted, such as the sampling rate of 3: 1, dividing a correction set and a prediction set in proportion, wherein the correction set is used for constructing a visible-near infrared model, and the prediction set is used for external verification of the model. Statistical parameters for the wood density of the correction set and the prediction set are shown in table 1.
TABLE 1 statistical parameters for wood density
As can be seen from table 1, the maximum and minimum wood density values of the prediction set are within the calibration set range for the four woods, and the quantitative analysis shows that the selected sample set has better representativeness.
The collected visible-near infrared spectrum of wood is interfered by external environment and instrument background in the process of collecting the visible-near infrared spectrum of woodThe spectrum often contains a certain amount of noise. In order to eliminate the influence of noise, the second generation wavelet-lifting wavelet transform is adopted to carry out denoising processing on the original spectrum. Different parameters and different denoising effects of the spectrum. Using biorNrNdMother wavelet (N)rFor reconstructing wavelet order, NdDecomposing wavelet order) to carry out 1-7-layer decomposition on the original spectrum of the wood, and comparing and analyzing the wavelet coefficients under different wavelet orders and decomposing layers and the denoising effect of the reconstructed spectrum. The method adopts Partial Least Squares (PLS) model statistical parameters as the denoising evaluation indexes, and the statistical parameters comprise: determining the coefficient (R)2) Root Mean Square Error (RMSE), and relative analytical error (RPD); in general, R2And the larger the RPD value is, the smaller the RMSE value is, the better the model prediction effect is, and the best denoising parameter corresponding to the spectrum used for modeling is. The optimal denoising parameters for each tree are shown in Table 2. The accuracy is for the discrimination of tree species and origin. Generally, among the modeling parameters of the prediction set, the coefficient (R) is determined2)>0.75, the model effect is acceptable. Of course, R2The larger the value, the smaller the error value, the better the modeling effect.
TABLE 2 four optimal lifting wavelet de-noising parameters for wood
As can be seen from table 2, the wavelet coefficients and the optimal wavelet functions and the number of decomposition layers for denoising the reconstructed spectrum are different for the four kinds of wood; the wavelet function bior2.8 has a good denoising effect on four kinds of wood, and for poplars in Heilongjiang, the optimal denoising functions of the wavelet coefficient and the reconstruction spectrum are bior2.8, but the optimal decomposition layers of the wavelet coefficient and the reconstruction spectrum are different.
Step three
Taking the wood material property information density as an example, the following description is given:
and establishing a wood property prediction model (prediction density), a wood producing area tracing model and a wood tree species identification model based on the spectrum of each wood sample after the optimal lifting wavelet parameter denoising.
Constructing a wood property prediction model: the collected visible-near infrared spectrum of the wood is input into a partial least square and particle swarm-support vector machine model as independent variables respectively based on the wavelet coefficient and the reconstructed spectrum after the wavelet parameter denoising is optimally promoted by various trees, the modeling effects of the original spectrum and the wavelet coefficient are contrastively analyzed, and the statistical parameters of the model are shown in a table 3.
TABLE 3 Wood Density PLS model prediction accuracy
In Table 3, LVs represents the optimum number of principal components, Rc 2Determination coefficient, RMSE, representing correction modelCRoot mean square error, RPD, representing the corrected modelcRepresenting the relative analytical error of the correction model. As can be seen from Table 3, the decision coefficient of the model built by the wavelet coefficient is greater than the modeling parameter of the original spectrum of the corresponding wood; the error value is less than or equal to the model built by the original spectrum of the corresponding wood; comparing the wavelet coefficient with the reconstructed spectrum, except that the determined coefficients of the dahurian larch and the aspen wood (Jilin) are similar, the modeling parameters of the wavelet coefficients of other tree species are superior to the reconstructed spectrum, and quantitatively showing that the prediction precision of the wood density is improved after the original spectrum of the wood is subjected to lifting wavelet processing; the wavelet coefficients after lifting wavelet processing are input into a PSO-SVM model (particle swarm-support vector machine model), parameters of the PSO-SVM model are optimized by a response surface method, the optimized model parameters are interactive verification times, maximum iteration times and population numbers respectively, and model statistical results are shown in table 4.
TABLE 4 Wood Density PSO-SVM model prediction accuracy
Comparing table 3 and table 4, it can be seen that: for four kinds of wood, the prediction precision of a PSO-SVM model established by default parameters is superior to that of a PLS model of a corresponding tree species; after response surface optimization, the prediction precision of the three wood density correction models except the aspen produced in Heilongjiang province reaches over 0.95.
In the step, a response surface method is adopted to optimize the optimization process of the particle swarm-support vector machine model:
design the influencing factor of particle swarm-support vector machine model (PSO-SVM), namely cross validation fraction (X) by adopting classic design method Box-Behnken in Response Surface Method (RSM)1) Maximum number of iterations (X)2) And population number (X)3). The experimental level settings for each factor are shown in table 5.
TABLE 5 variable levels and codes based on Box-Behnken design
A validation set mean square error (CVmse) is used as a response value. Taking dahurian larch as an example, table 6 is the variance analysis result of the dahurian larch response surface model. As can be seen from the table, the F value of the model is 3343.02, and the P value is less than 0.0001, which shows that the Xingan larch model is remarkable, the fitting effect is good, and response surface analysis can be carried out. Factor X1The mean square error of the verification set is obviously influenced, and the influence of the three factors on the mean square error of the verification set is X from large to small1>X3>X2I.e. cross validation of the fold number>Number of groups>The maximum number of iterations. Model correction decision coefficient R20.9998, that is, the model can explain the variability of independent variable factors of 99.98 percent, and the model has better precision and reliability.
TABLE 6 variance analysis results of Larix dahurica response surface model
The optimal parameter combination values of the various RSM-PSO-SVM models are shown in Table 7.
TABLE 7 optimal parameter combination values for various RSM-PSO-SVM models
Step four
The model is used for predicting unknown wood samples, and the model prediction accuracy is shown in table 8.
TABLE 8 Wood Density prediction accuracy
As can be seen from table 8, the particle swarm-support vector machine model optimized for the response surface can be used to predict the density values of four woods. However, as can be seen from Table 1, the densities of the legitimate aspen wood and the illegitimate elm wood are similar, and the average densities are 0.757 and 0.761g cm, respectively-3Therefore, the validity of the wood cannot be judged only by the wood density index.
Step five
Considering the deficiency of judging the validity of the wood by the parameters of single wood components, it is necessary to accurately trace the origin of the wood. In this embodiment, the four kinds of wood original spectra and the wavelet coefficients after lifting wavelet processing are respectively input into the particle swarm-support vector machine model. Wherein the legal production area wood and the illegal production area wood are respectively represented by labels 0 and 1. The results of the legitimacy of the wood origin for the two input variables are shown in FIGS. 4a-4 b.
As can be seen from FIGS. 4a-4b, the judgment of the legality of the producing areas of the four kinds of wood is not misjudged, the correct judgment rates are all 100%, and the PSO-SVM model can verify whether the merchant replaces the same kind of legal wood with illegal producing area wood quantitatively.
In this step, original spectra belonging to different producing areas (for example, original spectra of two producing areas, which are sample original spectra corresponding to Heilongjiang and Jilin, respectively) and wavelet coefficients are input into a particle swarm-support vector machine model, so as to obtain a wood producing area traceability model.
Step six
In order to prevent lawless persons from replacing legitimate wood with illegitimate wood of similar growth, it is necessary to make a quick, accurate and non-destructive determination of the wood species. The collected wood spectra and the wavelet coefficients after lifting wavelet processing are input into a particle swarm-support vector machine model, and the obtained wood tree species identification results are shown in table 9:
TABLE 9 false judgment of wood species
As can be seen from table 9, after the wavelet transformation processing is performed, the correct rate of the tree species discrimination is greater than that of the original spectrum, and is 98.61%, and one of the woods is misjudged, that is, the elm wood is misjudged as pinus dahurica.
In this step, the original spectra (for example, the original spectra of three wood types, which are respectively the sample original spectra corresponding to dahurian larch, elm and aspen) and the wavelet coefficients, which are distinguished by various types, are input into a particle swarm-support vector machine model, and a wood species identification model is obtained. The method comprises the steps of respectively marking and inputting aspen in legal wood producing areas and aspen in illegal wood producing areas into a model so as to judge whether merchants replace wood in the legal producing areas with wood in the illegal producing areas. And the experimental result shows that: the wood in the legal production place and the illegal production place can be correctly distinguished.
In conclusion, the PSO-SVM model can be used for judging the wood property (density), the wood production place legality and the tree species, and the prediction precision of the PSO-SVM model is superior to that of a linear partial least square model.
In this embodiment, the construction process of the validity verification model has simpler steps, and can be used for simultaneously predicting wood property information and identifying wood origin and tree species types, so that authenticity identification of a large amount of wood property information can be realized, and qualitative and quantitative identification results can be realized. The method is simple to operate, and rapid, accurate and large-batch nondestructive prediction of unknown sample material property information is realized; and the detection result has high accuracy.
The foregoing shows and describes the general principles and broad features of the present invention and advantages thereof. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (4)
1. A wood validity verification method based on a visible-near infrared spectrum is characterized by comprising the following steps:
s1, acquiring visible-near infrared spectrum data of the wood to be verified;
s2, denoising the visible-near infrared spectrum data by adopting second-generation wavelet-lifting wavelet transform;
s3, inputting the visible-near infrared spectrum data of the step S1 or the denoised visible-near infrared spectrum data of the step S2 into a validity verification model; the validity verification model comprises: a wood property prediction model, a wood producing area tracing model and a wood tree species identification model;
and S4, outputting the wood property prediction result, the wood origin tracing data and the wood species identification data of the verified wood.
2. The method for verifying wood validity based on visible-near infrared spectrum of claim 1, wherein the process of constructing the validity verification model in step S3 includes:
s301, collecting various types of wood samples belonging to different producing areas; respectively scanning the samples for multiple times, and randomly selecting the average spectrum of a preset number of sampling points as the original spectrum of the sample;
s302, denoising the original spectrum by adopting second-generation wavelet-lifting wavelet transform; obtaining wavelet coefficients and a reconstructed spectrum;
s303, inputting the wavelet coefficient and the reconstructed spectrum into a particle swarm-support vector machine model as independent variables; optimizing the particle swarm-support vector machine model by adopting a response surface method to obtain a wood property prediction model;
s304, inputting the original spectrums and the wavelet coefficients belonging to different producing areas into a particle swarm-support vector machine model to obtain a wood producing area tracing model;
s305, inputting the original spectrums and the wavelet coefficients which are distinguished by various types into a particle swarm-support vector machine model to obtain a wood tree species identification model.
3. The visible-near infrared spectrum-based wood legality verification method as claimed in claim 1 or2, wherein the denoising process in step S2 and/or step S302 adopts a partial least squares model statistical parameter as an evaluation index of denoising; the statistical parameters include: coefficient of determination, root mean square error, and relative analysis error.
4. The method for verifying wood validity based on visible-near infrared spectrum according to claim 2, wherein the step S301 comprises:
collecting various types of wood belonging to different producing areas, and manufacturing disc test pieces with preset lengths at intervals of preset distances from bottom to top along a trunk;
collecting the spectrums of all the disc test pieces by adopting a spectrum scanner, wherein the spectrum collection range is 350-2500 nm; the scanning times are X times, and X is more than or equal to 30;
and randomly selecting a preset number of sampling points for each disc test piece, and taking the average spectrum of the preset number of sampling points as the original spectrum of the sample.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110873658.3A CN113466156A (en) | 2021-07-30 | 2021-07-30 | Wood validity verification method based on visible-near infrared spectrum |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110873658.3A CN113466156A (en) | 2021-07-30 | 2021-07-30 | Wood validity verification method based on visible-near infrared spectrum |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113466156A true CN113466156A (en) | 2021-10-01 |
Family
ID=77883468
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110873658.3A Pending CN113466156A (en) | 2021-07-30 | 2021-07-30 | Wood validity verification method based on visible-near infrared spectrum |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113466156A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101393122A (en) * | 2008-10-31 | 2009-03-25 | 中国农业大学 | Honey quality rapid detection method |
| CN106770003A (en) * | 2016-11-21 | 2017-05-31 | 无锡迅杰光远科技有限公司 | Wood Identification Method and system based on near-infrared spectrum technique |
| CN112945881A (en) * | 2021-03-11 | 2021-06-11 | 内蒙古农业大学 | Method for monitoring water content of potato leaves based on hyperspectral characteristic parameters |
| CN113033066A (en) * | 2021-04-07 | 2021-06-25 | 温州大学 | Method for establishing near infrared spectrum identification model of sargassum fusiforme production area, strain and cultivation mode and identification method |
-
2021
- 2021-07-30 CN CN202110873658.3A patent/CN113466156A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101393122A (en) * | 2008-10-31 | 2009-03-25 | 中国农业大学 | Honey quality rapid detection method |
| CN106770003A (en) * | 2016-11-21 | 2017-05-31 | 无锡迅杰光远科技有限公司 | Wood Identification Method and system based on near-infrared spectrum technique |
| CN112945881A (en) * | 2021-03-11 | 2021-06-11 | 内蒙古农业大学 | Method for monitoring water content of potato leaves based on hyperspectral characteristic parameters |
| CN113033066A (en) * | 2021-04-07 | 2021-06-25 | 温州大学 | Method for establishing near infrared spectrum identification model of sargassum fusiforme production area, strain and cultivation mode and identification method |
Non-Patent Citations (7)
| Title |
|---|
| LI Y 等: "Visible-near infrared spectroscopy and chemometric methods for wood density prediction and origin/species identification", 《FORESTS》 * |
| 尹世逵 等: "基于不同预处理的椴树木材基本密度近红外光谱估测及模型优化", 《中南林业科技大学学报》 * |
| 李颖: "基于Vis-NIR光谱的木材树种鉴别及密度模型研究", 《中国优秀博士学位论文全文数据库工程科技Ⅰ辑》 * |
| 李颖: "基于Vis-NIR光谱的木材树种鉴别及密度模型研究", 中国优秀博士学位论文全文数据库工程科技Ⅰ辑 * |
| 李颖等: "基于小波变换的木材近红外光谱模型优化及质量追溯", 《光谱学与光谱分析》 * |
| 汪紫阳 等: "基于可见/近红外光谱识别东北地区常见木材", 《浙江农林大学学报》 * |
| 赵鹏 等: "可见光/近红外光谱分析的木材树种与密度同时预测", 《光谱学与光谱分析》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2018010352A1 (en) | Qualitative and quantitative combined method for constructing near infrared quantitative model | |
| Kuang et al. | Influence of the number of samples on prediction error of visible and near infrared spectroscopy of selected soil properties at the farm scale | |
| CN112766075B (en) | Hyperspectral remote sensing black and odorous water body grading method based on semi-supervised learning strategy | |
| CN106918567B (en) | A kind of method and apparatus measuring trace metal ion concentration | |
| CN110376153B (en) | Method for tracing origin of market saffron by combining ATR-FTIR with RBF neural network | |
| Li et al. | Lifting wavelet transform for Vis-NIR spectral data optimization to predict wood density | |
| CN109211829A (en) | A method of moisture content in the near infrared spectroscopy measurement rice based on SiPLS | |
| Yang et al. | Preliminary investigation into the identification of wood species from different locations by near infrared spectroscopy | |
| CN105717066A (en) | Near-infrared spectrum recognition model based on weighting association coefficients | |
| CN111523587A (en) | Woody plant species spectrum identification method based on machine learning | |
| CN105223140A (en) | The method for quickly identifying of homology material | |
| CN103743705A (en) | Rapid detection method for sorghum halepense and similar species | |
| Fernández‐Getino et al. | Exploratory analysis of the structural variability of forest soil humic acids based on multivariate processing of infrared spectral data | |
| CN113903407A (en) | Component identification method, component identification device, electronic equipment and storage medium | |
| Liu et al. | Research on the online rapid sensing method of moisture content in famous green tea spreading | |
| CN117250169A (en) | Method for nondestructively and rapidly distinguishing wild cordyceps sinensis based on model transfer technology | |
| CN113484278A (en) | Tomato comprehensive quality nondestructive testing method based on spectrum and principal component analysis | |
| CN113466156A (en) | Wood validity verification method based on visible-near infrared spectrum | |
| CN117491293A (en) | High-steep bank slope carbonate rock corrosion rapid evaluation method based on hyperspectrum | |
| CN112179870A (en) | Cigarette classification recognition model construction method based on near infrared spectrum and OPLS-DA | |
| CN116429724A (en) | Method for detecting pollution degree of composite insulator material based on infrared hyperspectral analysis | |
| CN114550843B (en) | Prediction model of monosaccharide composition and content in traditional Chinese medicine polysaccharide, construction method and application thereof | |
| Li et al. | Spectrometric prediction of wood basic density by comparison of different grain angles and variable selection methods | |
| CN115452758A (en) | Rice category identification model, construction method thereof and method for identifying rice category | |
| NL2029376B1 (en) | Wood legality verifying method based on visible-near infrared spectroscopy |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20211001 |