CN117451625A - Method and device for detecting content of mixture components based on hyperspectrum and application of method and device - Google Patents
Method and device for detecting content of mixture components based on hyperspectrum and application of method and device Download PDFInfo
- Publication number
- CN117451625A CN117451625A CN202311381309.5A CN202311381309A CN117451625A CN 117451625 A CN117451625 A CN 117451625A CN 202311381309 A CN202311381309 A CN 202311381309A CN 117451625 A CN117451625 A CN 117451625A
- Authority
- CN
- China
- Prior art keywords
- spectrum
- mixture
- component
- detected
- content
- 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
- 239000000203 mixture Substances 0.000 title claims abstract description 106
- 238000000034 method Methods 0.000 title claims abstract description 97
- 238000001228 spectrum Methods 0.000 claims abstract description 120
- 239000002994 raw material Substances 0.000 claims abstract description 31
- 230000008569 process Effects 0.000 claims abstract description 27
- 238000002156 mixing Methods 0.000 claims abstract description 25
- 238000009826 distribution Methods 0.000 claims abstract description 23
- 238000004519 manufacturing process Methods 0.000 claims abstract description 22
- 238000003384 imaging method Methods 0.000 claims description 26
- 235000013305 food Nutrition 0.000 claims description 19
- 238000001514 detection method Methods 0.000 claims description 18
- 235000013599 spices Nutrition 0.000 claims description 17
- 239000000463 material Substances 0.000 claims description 15
- 238000004458 analytical method Methods 0.000 claims description 13
- 230000003595 spectral effect Effects 0.000 claims description 13
- 230000002596 correlated effect Effects 0.000 claims description 8
- 239000000126 substance Substances 0.000 claims description 8
- 239000006185 dispersion Substances 0.000 claims description 2
- 239000003814 drug Substances 0.000 claims description 2
- 230000008901 benefit Effects 0.000 abstract description 6
- 238000012544 monitoring process Methods 0.000 abstract description 6
- 238000011156 evaluation Methods 0.000 abstract description 5
- 239000002304 perfume Substances 0.000 description 26
- 239000000654 additive Substances 0.000 description 18
- 230000000875 corresponding effect Effects 0.000 description 16
- 230000000996 additive effect Effects 0.000 description 13
- 238000000701 chemical imaging Methods 0.000 description 11
- 238000010586 diagram Methods 0.000 description 10
- 238000003860 storage Methods 0.000 description 9
- 238000005507 spraying Methods 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 238000010606 normalization Methods 0.000 description 4
- 238000010521 absorption reaction Methods 0.000 description 3
- 238000000889 atomisation Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 239000003205 fragrance Substances 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000011158 quantitative evaluation Methods 0.000 description 3
- 230000005855 radiation Effects 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 238000003756 stirring Methods 0.000 description 3
- 241000208125 Nicotiana Species 0.000 description 2
- 235000002637 Nicotiana tabacum Nutrition 0.000 description 2
- 230000009471 action Effects 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 2
- 230000001276 controlling effect Effects 0.000 description 2
- 238000012937 correction Methods 0.000 description 2
- 238000011068 loading method Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- NIXOWILDQLNWCW-UHFFFAOYSA-N acrylic acid group Chemical group C(C=C)(=O)O NIXOWILDQLNWCW-UHFFFAOYSA-N 0.000 description 1
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 1
- 229910052782 aluminium Inorganic materials 0.000 description 1
- 230000003321 amplification Effects 0.000 description 1
- 239000011324 bead Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000000576 coating method Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000007599 discharging Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000005284 excitation Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000011049 filling Methods 0.000 description 1
- 239000000796 flavoring agent Substances 0.000 description 1
- 235000019634 flavors Nutrition 0.000 description 1
- 229910021389 graphene Inorganic materials 0.000 description 1
- 230000003760 hair shine Effects 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000003199 nucleic acid amplification method Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000003908 quality control method Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000000985 reflectance spectrum Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000004611 spectroscopical analysis Methods 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
- 238000003892 spreading Methods 0.000 description 1
- 230000007480 spreading Effects 0.000 description 1
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
- G01N21/01—Arrangements or apparatus for facilitating the optical investigation
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Abstract
The invention discloses a hyperspectral-based method and device for detecting the component content of a mixture and application thereof, wherein the method comprises the steps of obtaining hyperspectral images of a mixture sample, wherein the mixture at least comprises raw materials and components to be detected; respectively obtaining a spectrum alpha of a component to be detected and a spectrum beta of a raw material; carrying out spectrum unmixing treatment on the mixture spectrum of each pixel position in the hyperspectral image to obtain a first index, wherein the first index represents the relative content of the spectrum alpha of the component to be detected in the mixture spectrum of the current pixel position; and detecting the content value and/or the content distribution of the component to be detected in the mixture sample based on the first index of the plurality of pixel positions in the hyperspectral image. According to the method, the hyperspectral image of the mixture is obtained, the first index value of each pixel position is calculated, and the content value and distribution of the average component to be detected in the whole image range are measured. The method is suitable for monitoring the mixing process in industrial manufacturing, and can provide quantitative reference for benefit evaluation and fault early warning work of component adding process links.
Description
Technical Field
The application relates to the technical field of tobacco shred component detection, in particular to a hyperspectral-based method and device for detecting the content of a mixture component and application of the hyperspectral-based method and device.
Background
In the flavoring process of tobacco shreds, adding additives such as essence is a common process, and for the process quality control of added products, the conventional sampling and chemical detection processes are generally adopted, so that the added quality cannot be fully reflected.
For example, a common additive application mode is a combination of atomization spraying and stirring, and under this framework, the method of controlling the application proportion is as follows: the method is characterized in that an induction electronic scale for calculating the instantaneous material flow is arranged on a conveying belt, if the instantaneous material flow sensed by the electronic scale at the moment T is recorded as M (kg/h), the additive content regulated in the industrial process standard is K, the instantaneous additive flow pumped into an additive tank at the moment should be M.K (kg/h), the instantaneous additive flow actually detected by a flowmeter in the additive tank is S (kg/h), the additive proportion error fed back to the system is epsilon= |S-M.K|/(M.K), obviously, the index epsilon only reflects the difference between the extracted additive mass and the preset mass, and the steps of atomizing spraying, cylinder charging stirring, heating and moisture discharging and the like are also carried out after the additive is charged, and the influence factors such as uneven spraying, volatilization of the additive and the like contained in the steps can lead the content of the additive actually absorbed by the material to be lower than K.
The hyperspectral imaging technique is a high-precision imaging technique based on fine spectroscopy. When mixed light in an imaging environment irradiates a beam splitter inside a hyperspectral camera, the mixed light is decomposed into monochromatic light with different wavelengths, and the final imaging result contains both planar image information (shape, size, position) and spectral information (absorption or reflection characteristics of light with different wavelengths) of a photographed object. The spectrum information has high correlation with the component types and specific contents of substances, and hyperspectral imaging provides rich spectrum information, can be used for distinguishing targets with similar shapes and colors and different materials, and can accurately calculate the contents of specific components, which is not realized by common RGB (red-green-blue) three-color imaging and single-color imaging using specific wave bands. Hyperspectral imaging techniques have therefore also been used for sample components in various industrial inspection fields.
The typical hyperspectral detection method is to select a region of interest, select a characteristic wavelength, acquire a spectrum of the region of interest, perform pretreatment, and verify the content of a sample through a model, wherein a standard sample with a known concentration is required to be in one-to-one correspondence with the spectrum, so that an effective prediction model can be established.
In the actual production process, the additives may volatilize or not be sufficiently mixed or adhered, so that the content of the additives cannot be accurately calculated, and thus, a standard sample with known concentration cannot be obtained, and a prediction model cannot be built. Meanwhile, based on the confidentiality requirement of the formula, a prediction model can not be established on the premise of not knowing the specific components of the mixture.
Disclosure of Invention
In order to solve the above problems, the embodiments of the present application provide a method and an apparatus for detecting component content of a mixture based on hyperspectral, and applications thereof, which combine hyperspectral imaging, machine vision and spectroscopic substance component analysis technologies to calculate a first index value of each pixel position and measure the average component content value and distribution to be detected in the whole graph range, so that the method and the apparatus are suitable for monitoring the mixing process in industrial manufacturing, and can provide a quantization reference for benefit evaluation and fault early warning in component adding process links.
To overcome the above-mentioned drawbacks of the prior art, the first aspect of the present invention provides a method for detecting the content of components in a mixture based on hyperspectrum, comprising the steps of:
obtaining a hyperspectral image of a mixture sample, wherein the mixture at least comprises a raw material and a component to be detected;
respectively obtaining a spectrum alpha of a component to be detected and a spectrum beta of a raw material;
performing spectrum unmixing treatment on the mixture spectrum of each pixel position in the hyperspectral image to obtain a first index, wherein the first index represents the relative content of spectrum alpha of a component to be detected in the mixture spectrum of the current pixel position;
and detecting the content value and/or the content distribution of the component to be detected in the mixture sample based on the first indexes of the pixel positions in the hyperspectral image.
Preferably, the step of obtaining the first index specifically includes the steps of:
obtaining abundance estimation values of spectrum alpha of to-be-detected components in the mixture spectrum and spectrum beta of raw materials according to the spectrum unmixing treatmentAnd->；
Based on abundance estimatesAnd->The difference builds a first index.
Preferably, the first index of pixel position pWherein->Is a preset parameter and，/>mixing coefficient estimates for the spectrum alpha and the spectrum beta, respectively, in the mixture spectrum of the current pixel position p>And->Is included in the standard of the coefficient of (c).
Preferably, the first index of pixel position pWhereinMixing coefficient estimates for the spectrum alpha and the spectrum beta, respectively, in the mixture spectrum of the current pixel position p>And->Is included in the standard of the coefficient of (c).
Preferably, the method further comprises the step of obtaining a second index for detecting the content value of the component to be detected in the mixture sample, wherein the value of the second index is positively correlated with the average value of the first index of the plurality of pixel positions in the hyperspectral image.
Preferably, the method further comprises the step of obtaining a third index for detecting the content distribution of the component to be detected in the mixture sample, wherein the value of the third index is positively correlated with the degree of dispersion of the first index at a plurality of pixel positions in the hyperspectral image.
Preferably, the spectral unmixed processing specifically includes the steps of:
constructing the spectrum of the mixture into a linear combination relation of the spectrum alpha of the component to be detected and the spectrum beta of the raw material;
calculating abundance estimation values of spectrum alpha and spectrum beta in the linear combination relation by using a numerical analysis methodAnd->。
In a second aspect, the present invention provides a hyperspectral-based device for detecting the content of components in a mixture, the device comprising:
and an imaging module: the method comprises the steps of configuring a hyperspectral image for acquiring a mixture sample, wherein the mixture at least comprises a raw material and a component to be detected;
and a spectrum acquisition module: the method comprises the steps of configuring a spectrum alpha for a component to be detected and a spectrum beta for a raw material to be detected;
and a numerical analysis module: the method comprises the steps of configuring a spectrum unmixing process for a mixture spectrum of each pixel position in the hyperspectral image to obtain a first index, wherein the first index represents the relative content of a spectrum alpha of a component to be detected in the mixture spectrum of the current pixel position;
the quality detection module is as follows: is configured to detect a content value and/or a content distribution of a component to be detected in the mixture sample based on a first index of a plurality of pixel positions in the hyperspectral image.
Preferably, the numerical analysis module is further configured to obtain the first index by: obtaining abundance estimation values of spectrum alpha of to-be-detected components in the mixture spectrum and spectrum beta of raw materials according to the spectrum unmixing treatmentAnd->The method comprises the steps of carrying out a first treatment on the surface of the Based on the abundance estimate->And->The difference builds a first index.
Preferably, the imaging module comprises a darkroom, a hyperspectral camera, a light source and a sample box, wherein a light-shielding space is formed in the darkroom, and a lens of the hyperspectral camera, the light source and the sample box are accommodated in the light-shielding space.
The third aspect of the present invention proposes the use of the hyperspectral-based method for detecting the content of a component in a mixture of a raw material and a component to be measured in a manufacturing process of chemical industry, food or medicine according to any one of the first aspect.
Preferably, the method is applied to the detection of the content of the spice added to the food raw materials in the food manufacturing process.
The beneficial effects of the invention are as follows:
according to the method, a hyperspectral image of the mixture is obtained, a first index value of each pixel position is calculated, and the content value and distribution of the average component to be detected in the whole image range are measured; the method is suitable for monitoring the mixing process in industrial manufacturing, and can provide quantitative reference for benefit evaluation and fault early warning work of component adding process links.
The invention fully utilizes the special performance of the hyperspectral camera for cooperatively acquiring the space information and the spectrum information, and the spectrum data obtained from each pixel position in the space dimension is obtained from one-time independent spectrum measurement, so that the obtained hyperspectral image has extremely high information abundance, and the content index of the additive is acquired from each data point by a mode identification and statistics method.
The method is particularly suitable for monitoring the mixing process in industrial manufacturing, the algorithm is input into hyperspectral images of the mixture, the hyperspectral images are output into first index values of all pixel positions and are used for measuring the content values of the components to be measured in the whole image range, and accurate and stable quantitative references can be provided for benefit evaluation and fault early warning work of component adding process links. The first index is a dimensionless numerical index, and can achieve the purpose of detecting and evaluating the operation state and the effect of the adding process by only comparing the historical data and the real-time measurement value of the first index without depending on externally introduced measurement data.
Meanwhile, the steps of sensing and collecting the characteristic data of the materials can be completed only through non-contact imaging, the materials cannot be damaged, and the method has traceability and repeatability. The hardware equipment supported by the invention is only composed of a camera and a darkroom (which can be made manually), has the advantages of portability, stability and small influence by workshop temperature and humidity environment and external illumination, and is easy to carry, arrange and maintain flexibly according to the specific requirements of detection tasks of different production lines.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram showing the steps of a method for detecting the content of a mixture component based on hyperspectral spectrum according to an embodiment of the present invention;
FIG. 2 is a schematic diagram showing a perfume content detection process of a mixture sample at the outlet of a perfuming barrel of a production line according to another embodiment of the present invention;
FIG. 3 is a diagram of a hyperspectral image obtained and a detection process according to another embodiment of the present invention;
FIG. 4 is a schematic structural diagram of a hyperspectral-based device for detecting the component content of a mixture according to another embodiment of the present invention;
fig. 5 is a schematic diagram of an apparatus structure of a hyperspectral imaging module according to another embodiment of the present invention.
In the figure: 1. a darkroom bracket; 2. a movable door; 3. a storage box; 4. a hyperspectral camera; 5. a light source.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application.
In the following description, the terms "first," "second," and "first," are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. The following description provides various embodiments of the present application, and various embodiments may be substituted or combined, so that the present application is also intended to encompass all possible combinations of the same and/or different embodiments described. Thus, if one embodiment includes feature A, B, C and another embodiment includes feature B, D, then the present application should also be considered to include embodiments that include one or more of all other possible combinations including A, B, C, D, although such an embodiment may not be explicitly recited in the following.
The following description provides examples and does not limit the scope, applicability, or examples set forth in the claims. Changes may be made in the function and arrangement of elements described without departing from the scope of the application. Various examples may omit, replace, or add various procedures or components as appropriate. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Furthermore, features described with respect to some examples may be combined into other examples.
The invention provides a method for detecting the content of a mixture component based on hyperspectral images. FIG. 1 is a schematic diagram showing steps of a method for detecting the content of components in a mixture based on hyperspectral, which specifically includes:
s1, acquiring hyperspectral images of a mixture sample, wherein the mixture at least comprises raw materials and components to be detected.
Specifically, a hyperspectral camera is used to shoot hyperspectral images of a mixture sample containing components to be detected, and the imaging light can be transmitted to the camera through the sample in a way that the sample emits light, such as fluorescence excited by an ultraviolet LED lamp, or reflected exogenous light, including but not limited to ultraviolet light, visible light, near infrared light and the like.
After the hyperspectral image is obtained, preprocessing is needed, including but not limited to the steps of reducing black frame noise, correcting brightness, normalizing spectrum and the like: if the sample emits light by external excitation, the result of spectrum standardization is a radiation brightness spectrum; if the sample shines in a manner that reflects exogenous light, the result of spectral normalization is a reflectance spectrum.
S2, respectively acquiring a spectrum alpha of the component to be detected and a spectrum beta of the raw material.
The mixture at least comprises two types of materials, and the two types of materials are divided into a raw material and a component to be detected, wherein the component to be detected is an object of interest during detection. The component to be measured can be a trace addition component or a component with higher content. The hyperspectral images of the raw materials and the components to be detected are obtained by the same method as S1, and the images are subjected to region selection and spectrum extraction by combining an image segmentation technology to obtain the spectrum alpha of the components to be detected and the spectrum beta of the raw materials.
S3, carrying out spectrum unmixing treatment on the mixture spectrum of each pixel position in the hyperspectral image to obtain a first index, wherein the first index is used for representing the relative content of the spectrum beta of the raw material in the mixture spectrum of the current pixel position, namely the content of the component to be detected in the mixture of the current pixel position.
In a preferred embodiment, linear unmixing may be used to obtain an estimate of the abundance of the spectrum α of the component to be measured in the mixture spectrum and the spectrum β of the feedstockAnd->. For any planar pixel position in the image of the sample to be detected, the mixture spectrum corresponding to the position can be regarded as the linear combination of the two sample spectrums of the raw material and the component to be detected, and the optimal solution of the mixing coefficient of the linear combination under the constraint condition of the corresponding method can be given through numerical analysis methods including but not limited to a least square method, an iterative solution and the like suitable for solving an overdetermined linear equation system, so as to obtain the abundance estimation value of the spectrum alpha of the component to be detected and the spectrum beta of the raw material>And->。
The calculation of the first index constructed in the present embodiment is based on another coefficient distributed Pixel by Pixel (Pixel-wise) within the imaging plane; the first index represents the content of the component to be detected in the mixture at the current pixel position, and the content of the component to be detected in the mixture is positively correlated with the coefficient of the corresponding component to be detected in the optimal solution of the mixing coefficient. The first index value set, the mixing coefficient set corresponding to the component to be detected and the pixel coordinate set in the image plane have a mapping relation which corresponds one by one.
Preferably, the estimation is based on the mixing coefficientAnd->The difference establishes a first index, namely, the difference of the ratio of the two components in the mixture is adopted to characterize the content of the component to be detected, and the difference is converted into a reasonable range (such as 0-100) so as to be more consistent with the sense of concentration or content.
In a preferred embodiment, a first index of pixel position p may be usedDesigned as a mixed coefficient estimation with spectrum alpha, spectrum beta +.>And->Normalized coefficient of>The difference correlation may be in the form of a linear correlation, an exponential correlation, etc., depending on the actual, for example,
wherein epsilon is a preset parameter and is respectively the estimated value of the mixing coefficient of the spectrum alpha and the spectrum beta in the spectrum of the mixture of the current pixel position pAnd->Is a normalized coefficient of (2);
or,
wherein epsilon is a preset parameter,mixing coefficient estimates for the spectrum alpha and the spectrum beta, respectively, in the mixture spectrum of the current pixel position p>And->Is included in the standard of the coefficient of (c).
S4, detecting the content value and/or the content distribution of the component to be detected in the mixture sample based on the first index of the plurality of pixel positions in the hyperspectral image.
On the basis of the first index, a second index which is positively correlated with the average value of the first index in the imaging range can be further obtained and used for detecting the content value of the additive in the mixture sample. It will be appreciated that if the total amount of the component to be measured in the sample within the imaging range is known, the magnitude of the second index is positively correlated with the total amount of the component to be measured.
In a specific embodiment, for N sample control groups (N. Gtoreq.20) before and after adding the component to be measured, the second index is used to represent the content of the component to be measured in each sample, the second index value set of the sample not containing the component to be measured and the second index value set of the sample containing the component to be measured are assumed to be independent samples taken from two different distributions, and the test proves that the assumption can pass the assumption test of α=0.05 or more strictly, whether the single control group is composed of the state before and after adding the component to be measured in the same sample in the simulation experiment or composed of the irrelevant samples actually taken before and after mixing the components to be measured in the actual production line.
On the basis of the first index, a third index positively correlated with the discrete degree of the first index of the plurality of pixel positions in the hyperspectral image can be further obtained and used for detecting the content distribution of the components to be detected in the mixture sample, namely, evaluating the mixing uniformity of the components to be detected. For any sample to be detected, the variables included in the process of calculating the third index are only from hyperspectral image data generated by the sample through the hyperspectral imaging equipment of the current model, and are irrelevant to hyperspectral image data of other samples or data acquired by the same sample through other sensors; it can be understood that, for two components to be tested with the same quality and the same brand, two mixture samples are obtained under the simulation condition of manually controlling uniform mixing and the actual condition of mixing in a production line respectively, and under the premise that the total amount of the perfume absorbed by the two samples in the imaging range is the same, the value of a third index corresponding to the sample obtained by the simulation mixing experiment should be higher than that of the sample obtained by sampling in the field.
In the following, the embodiment of the present invention will be described by taking the quality detection of the flavor added during the manufacture of food products as an example, it is understood that the detection of the component content of similar additives or mixtures during the manufacture of other products can be realized based on the disclosed embodiments of the present invention.
In this embodiment, in combination with hyperspectral imaging, machine vision and spectral substance component analysis techniques, a method applicable to detection of trace perfume additives is provided, and by directly performing imaging analysis on the mixed state of perfume and food, the perfume content of the corresponding area of each pixel position is quantitatively calculated according to spectral characteristics, so that quantitative evaluation can be performed on the total amount of perfume and the uniformity of perfume distribution in the food to be detected.
In this example, a first index, the relative fragrance content index (Related Spice Index, RSI), was designed for qualitative assessment of fragrance content at each pixel location in a hyperspectral image of a mixture sample. Specifically, the mixture at the outlet of the perfuming roller is directly subjected to spectral analysis by a hyperspectral imaging technology, the relative proportions of different components are calculated according to the spectral characteristics of the mixture, and the relative spice content index RSI of the physical quantity reflecting the perfuming condition is calculated, so that quantitative evaluation is carried out on the absorption quantity and plane distribution uniformity of the spice, and when the conditions of insufficient spice adding proportion and uneven spraying occur, problems can be found in time and an alarm can be given to a control system.
In a specific embodiment, the spectral unmixing process may employ linear unmixing, in particularFor a hyperspectral image with m rows, n columns, k bands, the mixture spectrum for pixel position p is noted asThe spectrum of the perfume standard sample is +.>The spectrum of the standard sample without perfume is +.>The relationship between the three variables can be described by a system of linear equations as follows:
the above unknowns areWherein->Respectively representing the mixing coefficient (mixing efficiency) corresponding to the alpha and beta components at the pixel position p, and at k>2, the above formula is a binary overdetermined linear equation set, and can be given by using a least square method, an iteration method or other numerical analysis methods suitable for solving the overdetermined linear equation setThe optimal solution under the corresponding constraint condition is marked as +.>The method comprises the steps of carrying out a first treatment on the surface of the The single-point RSI value corresponding to pixel position p is noted +.>Which reflects the amount of fragrance in the mixture at the actual position corresponding to the p-point.
In a specific embodiment, the first indexCan be used forThe construction is as follows:
wherein the method comprises the steps ofThe calculation mode of the standardized coefficient is as follows:
wherein the method comprises the steps of、/>Is a preset parameter.
In another embodiment, the first indexIt can also be constructed as follows:
based on the single pointThe second index can be further calculated, and quantitative evaluation is carried out on the perfume content in the mixture in the unit area corresponding to the pixel position; based on an m-row n-column matrix formed by the distribution of single-point RSI in an image plane, a third index can be obtained, and the average perfume content and the perfume distribution uniformity of all the mixtures in the imaging range are quantitatively evaluated.
Fig. 2 is a schematic diagram of a process for detecting the perfume content of a mixture sample at the outlet of a perfuming barrel of a production line according to another embodiment, and fig. 3 is a hyperspectral image obtained according to the process of the present embodiment, wherein the detected sample is a food sample to which perfume is added, and the process specifically includes the following steps:
(1) taking a mixture sample at the outlet of a perfuming barrel of a production line, and loading the mixture sample into a hyperspectral imaging device;
(2) acquiring a spectral image of a sample to be identified by using a hyperspectral camera, and recording asWherein m, n, k respectively represent the total number of lines, total number of columns and total number of spectral bands of the image
(3) From theSubtracting the pre-stored camera black frame noise image +.>And performing pretreatment such as correction to obtain standard hyperspectral image +.>。
Standard spectrum of k wave band corresponding to each plane pixel position pIn this example, the spectrum α of the component to be tested is the standard spectrum of the perfume +.>Spectrum beta of raw material, i.e. spectrum of food without perfumeIn the known standard spectrum +.>And->On the premise of->Denoted as->And->Is a linear combination of (a):
wherein the method comprises the steps of、/>Respectively representing the mixing coefficients of the two components alpha and beta at the pixel position p>And->The estimated value of the mixing coefficient can be obtained by least square method, jacobian iterative method, gaussian-Saidel iterative method, etc., and respectively recorded as +.>、/>The distribution diagram along with the planar pixel position is marked +.>、/>；
(4) Pair of、/>Coefficient value corresponding to pixel position p in each plane +.>、P=1, 2, …, mxn, single-point normalization (Local generalization) was performed as shown below, and the normalized coefficient was recorded as +.>：
Normalized coefficient distribution is plotted as、/>，/>For a preset parameter greater than 1, in this embodiment +.>；
(5) By normalizing the coefficients for each pixel position pCombining to obtain a combination of +.> Positive correlation value ∈>The relative perfume content index (RSI) is represented at the value of pixel position p:
the distribution diagram along with the planar pixel position is marked as a matrix +.>，/>For a preset parameter greater than 1, in this embodiment +.>=2.7；
(6) Pair ofPerforming global normalization (Global generalization) calculation to obtain a value range of [0,1]]Is referred to as the global normalized relative perfume content index (Globally Generalized Related Spice Index), a second index, of all mixtures in the imaging range, noted GGRSI:
all single point RSI values within the imaging range are first normalized to within the range of values [0,1] by:
based on all single points within the imaging rangeValue determination GGRSI:
(7) counting according to the historical data of the same food stored in the database, and calculating a preset threshold valueIf the GGRSI value is not less than the threshold value, the total perfume content is considered to be acceptable
(8) Calculation ofAll positions within the imaging rangeStandard deviation of values>Ratio to GGRSI value, i.e. third index:
and is matched with a preset threshold value set based on the statistics of the historical data of the same foodIn contrast, if delta is not greater than +.>Namely, the atomization nozzle is considered to work normally, and the situation of uneven flavoring does not occur;
(9) and (3) the step (6) is performedAnd (3) outputting the matrix, the GGRSI value calculated in the step (7) and the two Boolean values obtained in the step (8) and the step (9) to a production line central control system.
The algorithm provided in this embodiment shows that after the analyzed hyperspectral image is imaged after the food leaves the perfuming barrel, the food and spice mixture in this stage has undergone spice loss caused by various environmental factors in the process of adding, and the food raw materials in the mixture also complete the absorption of the undamaged spice, so that the gap between the calculated value and the actual effective adding amount in this stage is minimum, and the technology for detecting the spice adding amount in the industry can only monitor the stage of spice entering the pot, and the monitoring of the subsequent spraying and stirring processes is in a blank state and cannot reach the effective absorbed spice content formed after the process link is finished.
In another embodiment, the perfume content of the mixture sample at the outlet of the perfuming barrel of the production line is subjected to a detection analysis according to the following steps:
(1) taking about 10g of mixture sample from the outlet of the perfuming barrel of the production line, uniformly spreading the mixture sample on the bottom of the storage box, pushing the storage box into a darkroom, closing a movable door to enable a hyperspectral camera to image from the vertical upper part of the mixture, and recording the imaging result asFrom the slaveSubtracting the pre-stored camera black frame noise image +.>And performing pretreatment such as radiation brightness correction to obtain standard radiation brightness image +.>；
(2) Loading pre-stored perfume standard spectrumAnd standard spectrum of non-aroma food->；
(3) Pair ofK-band spectrum +.>Based on->Performing linear decomposition of the mixed spectrum to obtain a mixing coefficient corresponding to position p>、/>Optimal estimate of +.>. Taking the least square method as an example, the solving method is as follows:
recording device，/>The spectral linear mixture model is then written as:
，
according to the least square method, the following steps are obtained:
；
(4) corresponding to each plane position pThe values are respectively single-point normalized (Local generalization), and the relative perfume content index corresponding to the pixel position p is obtained based on the normalized coefficients>Distribution map with plane position->；
(5) Global normalization (Global generalization) is carried out on all RSI values in the effective imaging range to obtain global normalized relative spice content index GGRSI;
(6) comparing GGRSI with a preset thresholdThe threshold value can be obtained based on historical numerical statistics of the same food GGRSI accumulated by an algorithm database, and the calculation method selected in the embodiment is as follows:
wherein the method comprises the steps ofGGRSI history number sets representing pre-and post-addition mixes, respectively, ++>Representing the ranking of the values in set X from small to large and finding the demarcation value of front (gamma.times.100)% and rear ((1-gamma). Times.100)% from the ranking, e.g.. +.>Namely, it means that among all the history values of GGRSI of the foods before addition, it is equal to or less than +.>The number of (2) is 95%, similarlyIndicating that in all the mixtures after the addition of perfume, less than or equal to +.>The larger one of the two is taken as the judgment threshold value +.>If the GGRSI value obtained by the sample sampled at this time is not lower than the threshold value, the perfume content is qualified, otherwise, the perfume is judged to be insufficient;
(7) calculating all positions within the imaging rangeStandard deviation of values>Ratio to GGRSI valueAnd is equal to a preset threshold->In contrast, a set available for setting +.>As an example:
wherein the method comprises the steps ofThe method is characterized in that the artificial control of the uniform spraying of the perfume is performed by a simulated perfuming experiment to obtain the delta index estimated value +.>The value set after repeated experiments is marked as +.>Get the collection->Mode of (2)And multiplying the obtained product by an amplification factor theta (θ=2 in this embodiment) larger than 1 to obtain a flavoring uniformity determination threshold value +.>Is a value of (a). If the delta value of the sample to be tested is not greater than +.>The atomization nozzle is considered to work normally, and the condition of uneven flavoring does not occur.
Fig. 4 is a schematic structural diagram of a hyperspectral-based mixture component content detecting apparatus 400 according to another embodiment of the present invention, which includes:
an imaging module 401 configured to obtain a hyperspectral image of a sample of a mixture, the mixture comprising at least a raw material and a component to be measured;
a spectrum acquisition module 402 configured to acquire a spectrum α of a component to be measured and a spectrum β of a raw material, respectively;
the numerical analysis module 403 is configured to perform spectral unmixing processing on the mixture spectrum at each pixel position in the hyperspectral image, so as to obtain a first index, where the first index represents the relative content of the spectrum α of the raw material in the mixture spectrum at the current pixel position, and the first index is used to represent the content of the component to be detected in the mixture at the current pixel position;
the quality detection module 404 is configured to detect a content value and/or a content distribution of the component to be detected in the mixture sample based on the first index of the plurality of pixel positions in the hyperspectral image.
Fig. 5 is a schematic view of an apparatus structure of a hyperspectral imaging module according to an embodiment of the present invention, the apparatus includes a darkroom bracket 1, a movable door 2, a storage box 3, a hyperspectral camera 4, and a light source 5. The darkroom bracket 1 is used for accommodating a storage box with a shooting target into the darkroom environment, and fixing the hyperspectral camera 4 and the light source 5 above the storage box 3 by using a beam structure, so that an imaging plane of the camera is parallel to the bottom surface of the storage box, and the bracket and the beam structure in the bracket can be made of aluminum profiles, PVC pipes or other materials with strength enough to bear the weight of the camera and the light source; the darkroom environment is formed by filling shading materials (such as black matte acrylic plates, graphene coatings and the like) on each surface of a hexahedron formed by the brackets, and the rest surface area is completely sealed by the shading materials except for leaving a gap extending into the darkroom for a lens on one surface where a camera is positioned; the movable door 2 is used for taking in and out the storage box from the darkroom; the shooting target is placed at the bottom of the storage box 3, and the materials which can be placed are foods before the spice is added, spices, mixtures after the spice is added and the like; a hyperspectral camera 4 whose usable imaging wavelength range should cover the detection wavelength and have an interface design such as screw holes or the like at the junction of the camera housing and the darkroom mount 1 that enables it to be fixed at the junction; the light source 5 may use an LED bulb or other type of lamp, and if there are more than 1 light beads in the light source, the arrangement should be regular and symmetrical.
The method provided by the embodiment combines hyperspectral imaging, machine vision and spectral substance component analysis technologies to analyze the content and the content distribution condition of the added components in the mixed materials, can be widely applied to the monitoring of the mixed materials in the manufacturing process of the industries such as food, chemical industry and the like, and provides accurate and stable quantitative references for the benefit evaluation and fault early warning work of the component adding process links.
The foregoing is a preferred embodiment of the present application, and it will be apparent to those skilled in the art that various modifications and changes to the embodiment of the present application may be made without departing from the spirit and scope of the present application. In this manner, the present application is also intended to cover such modifications and changes as fall within the scope of the claims of the application and the equivalents thereof.
It should be noted that, for simplicity of description, the foregoing method embodiments are all expressed as a series of action combinations, but it should be understood by those skilled in the art that the present application is not limited by the order of actions described, as some steps may be performed in other order or simultaneously in accordance with the present application. Further, those skilled in the art will also appreciate that the embodiments described in the specification are all preferred embodiments, and that the acts and modules referred to are not necessarily required in the present application.
In the foregoing embodiments, the descriptions of the embodiments are emphasized, and for parts of one embodiment that are not described in detail, reference may be made to related descriptions of other embodiments.
The foregoing is merely exemplary embodiments of the present disclosure and is not intended to limit the scope of the present disclosure. That is, equivalent changes and modifications are contemplated by the teachings of this disclosure, which fall within the scope of the present disclosure. Embodiments of the present disclosure will be readily apparent to those skilled in the art from consideration of the specification and practice of the disclosure herein. This application is intended to cover any adaptations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a scope and spirit of the disclosure being indicated by the claims.
Claims (10)
1. The method for detecting the content of the components of the mixture based on hyperspectrum is characterized by comprising the following steps:
obtaining a hyperspectral image of a mixture sample, wherein the mixture at least comprises a raw material and a component to be detected;
respectively obtaining a spectrum alpha of a component to be detected and a spectrum beta of a raw material;
performing spectrum unmixing treatment on the mixture spectrum of each pixel position in the hyperspectral image to obtain a first index, wherein the first index represents the relative content of spectrum alpha of a component to be detected in the mixture spectrum of the current pixel position;
and detecting the content value and/or the content distribution of the component to be detected in the mixture sample based on the first indexes of the pixel positions in the hyperspectral image.
2. The method according to claim 1, wherein the obtaining the first index specifically comprises the steps of:
obtaining abundance estimation values of spectrum alpha of to-be-detected components in the mixture spectrum and spectrum beta of raw materials according to the spectrum unmixing treatmentAnd->；
Based on abundance estimatesAnd->The difference builds a first index.
3. The method of claim 2, wherein the first indicator of pixel position pWherein->Is a preset parameter and->，/>Mixing coefficient estimates for the spectrum alpha and the spectrum beta, respectively, in the mixture spectrum of the current pixel position p>And->Is included in the standard of the coefficient of (c).
4. The method of claim 2, wherein the first indicator of pixel position pWherein->Mixing coefficient estimates for the spectrum alpha and the spectrum beta, respectively, in the mixture spectrum of the current pixel position p>And->Is included in the standard of the coefficient of (c).
5. The method of claim 1, further comprising obtaining a second indicator for detecting a value of the content of the component to be measured in the sample mixture, the value of the second indicator being positively correlated with an average of the first indicator for the plurality of pixel locations in the hyperspectral image.
6. The method of claim 1, further comprising obtaining a third indicator for detecting a content distribution of the component to be measured in the mixture sample, the value of the third indicator being positively correlated with the degree of dispersion of the first indicator for the plurality of pixel locations in the hyperspectral image.
7. The method according to claim 1, characterized in that said spectral unmixing process comprises in particular the steps of:
constructing the spectrum of the mixture into a linear combination relation of the spectrum alpha of the component to be detected and the spectrum beta of the raw material;
calculating abundance estimation values of spectrum alpha and spectrum beta in the linear combination relation by using a numerical analysis methodAnd->。
8. A hyperspectral-based device for detecting the component content of a mixture, comprising:
and an imaging module: the method comprises the steps of configuring a hyperspectral image for acquiring a mixture sample, wherein the mixture at least comprises a raw material and a component to be detected;
and a spectrum acquisition module: the method comprises the steps of configuring a spectrum alpha for a component to be detected and a spectrum beta for a raw material to be detected;
and a numerical analysis module: the method comprises the steps of configuring a spectrum unmixing process for a mixture spectrum of each pixel position in the hyperspectral image to obtain a first index, wherein the first index represents the relative content of a spectrum alpha of a component to be detected in the mixture spectrum of the current pixel position;
the quality detection module is as follows: is configured to detect a content value and/or a content distribution of a component to be detected in the mixture sample based on a first index of a plurality of pixel positions in the hyperspectral image.
9. The use of the method for detecting the content of a component in a hyperspectral based mixture according to any one of claims 1 to 7, which is applied to the detection of the content of a component to be detected in a mixture of a raw material and the component to be detected after mixing in the manufacturing process of chemical industry, food or medicine.
10. The use of the hyperspectral based mixture component content detecting method as claimed in claim 9, wherein the method is applied to the content detection of spices added to food materials in the manufacturing process of food.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311381309.5A CN117451625A (en) | 2023-10-24 | 2023-10-24 | Method and device for detecting content of mixture components based on hyperspectrum and application of method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311381309.5A CN117451625A (en) | 2023-10-24 | 2023-10-24 | Method and device for detecting content of mixture components based on hyperspectrum and application of method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117451625A true CN117451625A (en) | 2024-01-26 |
Family
ID=89588489
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311381309.5A Pending CN117451625A (en) | 2023-10-24 | 2023-10-24 | Method and device for detecting content of mixture components based on hyperspectrum and application of method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117451625A (en) |
-
2023
- 2023-10-24 CN CN202311381309.5A patent/CN117451625A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Recent advances in nondestructive analytical techniques for determining the total soluble solids in fruits: a review | |
| Sanaeifar et al. | Prediction of banana quality indices from color features using support vector regression | |
| Taghizadeh et al. | Comparison of hyperspectral imaging with conventional RGB imaging for quality evaluation of Agaricus bisporus mushrooms | |
| Tang et al. | Multispectral imaging for predicting sugar content of ‘Fuji’apples | |
| CN109799195B (en) | High-precision quantitative analysis method for laser-induced breakdown spectroscopy | |
| Burger et al. | Hyperspectral NIR image regression part II: dataset preprocessing diagnostics | |
| EP3236245B1 (en) | Sample-analyzing system | |
| CN112508070B (en) | Calibration modeling method for online detection of dehydration amount in drying process based on near infrared spectrum | |
| CN108444924B (en) | Method for discriminating storage period of tea by applying hyperspectral image technology | |
| US20190369013A1 (en) | Method and apparatus for spectroscopic analysis, implementing infrared and fluorescence multichannel processing of spectral data | |
| CN108760647A (en) | A kind of wheat content of molds line detecting method based on Vis/NIR technology | |
| Duan et al. | Sensitive variables extraction, non-destructive detection and visualization of total viable count (TVC) and pH in vacuum packaged lamb using hyperspectral imaging | |
| US20050279924A1 (en) | Method and a spectrometer for quantitative determination of a constituent in a sample | |
| Yu et al. | Rapid and visual measurement of fat content in peanuts by using the hyperspectral imaging technique with chemometrics | |
| CN106841167A (en) | The lossless detection method of garden stuff pesticide residue | |
| Shirai et al. | Detection of fluorescence signals from ATP in the second derivative excitation–emission matrix of a pork meat surface for cleanliness evaluation | |
| Shei et al. | An optical automatic measurement method for the moisture content of rough rice using image processing techniques | |
| Brouckaert et al. | Calibration transfer of a Raman spectroscopic quantification method for the assessment of liquid detergent compositions from at-line laboratory to in-line industrial scale | |
| Sun et al. | Surface gloss evaluation of apples based on computer vision and support vector machine method | |
| CN111122635B (en) | Online evaluation method and system for edible fungus drying state | |
| CN117451625A (en) | Method and device for detecting content of mixture components based on hyperspectrum and application of method and device | |
| Kaswati et al. | Examination system of chicken meat quality based on hyperspectral imaging | |
| US8213011B2 (en) | Spatial imaging/imaging spectroscopy system and method | |
| JP2022527850A (en) | Methods for configuring a spectroscopic measuring device | |
| CN113740275B (en) | Haematococcus pluvialis astaxanthin content detection method based on snapshot-type multispectral imaging |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |