CN108593589A - Application of the near-infrared high light spectrum image-forming technology in chicken weight quickly detection - Google Patents
Application of the near-infrared high light spectrum image-forming technology in chicken weight quickly detection Download PDFInfo
- Publication number
- CN108593589A CN108593589A CN201810228627.0A CN201810228627A CN108593589A CN 108593589 A CN108593589 A CN 108593589A CN 201810228627 A CN201810228627 A CN 201810228627A CN 108593589 A CN108593589 A CN 108593589A
- Authority
- CN
- China
- Prior art keywords
- spectrum
- image
- weight
- spectrum image
- sample
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001228 spectrum Methods 0.000 title claims abstract description 31
- 241000287828 Gallus gallus Species 0.000 title claims abstract description 16
- 238000005516 engineering process Methods 0.000 title claims abstract description 10
- 238000001514 detection method Methods 0.000 title claims abstract description 9
- 210000000481 breast Anatomy 0.000 claims abstract description 19
- 230000003595 spectral effect Effects 0.000 claims description 6
- 241000272201 Columbiformes Species 0.000 claims description 4
- 238000012937 correction Methods 0.000 claims description 4
- 238000002310 reflectometry Methods 0.000 claims description 4
- 235000013330 chicken meat Nutrition 0.000 abstract description 17
- 238000004611 spectroscopical analysis Methods 0.000 abstract description 6
- 238000005303 weighing Methods 0.000 abstract description 5
- 238000013178 mathematical model Methods 0.000 abstract description 2
- 238000000034 method Methods 0.000 description 9
- 235000013372 meat Nutrition 0.000 description 5
- 238000002790 cross-validation Methods 0.000 description 3
- HVYWMOMLDIMFJA-DPAQBDIFSA-N cholesterol Chemical compound C1C=C2C[C@@H](O)CC[C@]2(C)[C@@H]2[C@@H]1[C@@H]1CC[C@H]([C@H](C)CCCC(C)C)[C@@]1(C)CC2 HVYWMOMLDIMFJA-DPAQBDIFSA-N 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 235000012000 cholesterol Nutrition 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 239000000796 flavoring agent Substances 0.000 description 1
- 235000019634 flavors Nutrition 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 235000004213 low-fat Nutrition 0.000 description 1
- 238000009659 non-destructive testing Methods 0.000 description 1
- 230000000050 nutritive effect Effects 0.000 description 1
- 235000013613 poultry product Nutrition 0.000 description 1
- 102000004169 proteins and genes Human genes 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 238000005057 refrigeration Methods 0.000 description 1
- 238000012795 verification Methods 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
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/3563—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for analysing solids; Preparation of samples therefor
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01G—WEIGHING
- G01G17/00—Apparatus for or methods of weighing material of special form or property
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/359—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light using near infrared light
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses application of the near-infrared high light spectrum image-forming technology in chicken weight quickly detection.The present invention only needs to obtain the spectroscopic data of sample, reflectance value under most optimum wavelengths is brought directly to can be obtained the weight of Fresh Grade Breast sample in the prediction model of mathematical model, the efficiency of weighing of sample is substantially increased, it can be achieved that the extensive online weighing of chicken meat sample detects.
Description
Technical field
The present invention relates to field of food detection, relate generally to near-infrared high light spectrum image-forming technology and are quickly detected in chicken weight
In application.
Background technology
With the continuous improvement of living standards, the meat consumption concept of consumer is also gradually changing, and chicken is because with low
Fat, low cholesterol, high protein, it is easy to digest the advantages that, liked by consumers in general.The consumption figure of China's chicken year by year on
It rises, what the quality of chicken had become consumer pays close attention to object.Chilled fresh chicken can keep Meat Flavor (matter to the maximum extent
Ground softness, elastic good, tasty mouthfeel) and nutritive value, have become the mainstream of fresh meat consumption in China big and medium-sized cities.However
In daily life, consumer is to weigh its price according to weight, weight more high price lattice are higher when buying chicken.So
And in daily life, it is one of most common problem that consumer encounters to give short weight, because this directly compromises consumer
Interests, therefore consumer increasingly payes attention to the authenticity of chicken weight.Now, the weight of chicken is claimed using balance
Weight, but the method is cumbersome, inefficiency, can not meet the large-scale online weighing of current meat.High light spectrum image-forming technology
Traditional image technique and spectral technique are merged, can obtain the spatial information of determinand can also provide the spectrum letter of determinand
Breath, and have the characteristics that quick, lossless, the research in terms of quality of agricultural and poultry products non-destructive testing and security control in recent years
It is more, to produce many achievements.But the research report in terms of chicken weight is less.
Invention content
The present invention provides a kind of near-infrared high light spectrum image-forming technology easily operated, that detection speed is fast is existing to make up with this
There is the defects of technical operation is cumbersome, inefficiency, to detect the weight of chicken.
The technical scheme is that：Provide near-infrared high light spectrum image-forming technology answering in chicken weight quickly detection
With.
Further improvement of the present invention includes：
The application, its step are as follows：Step 1 acquires pigeon breast using the near-infrared Hyperspectral imager debugged
The high spectrum image of meat；The high spectrum image of step 2, acquisition is pre-processed, to obtain the reflectance value of spectrum；Step 3,
The data of acquisition are substituted into following formula：
YW=8.393+116.94X928.551nm-43.965X938.431nm+53.572X964.774nm-87.349X1025.669nm+
80.612X1070.091nm-89.666X1134.246nm+97.1X1155.631mm-61.053X1216.498nm+39.673X1308.658nm-
46.681X1343.236nm+36.06X1364.648nm+29.746X1435.516nm-62.815X1681.829nm+74.677X1686.804nm,
Wherein YWFor the weight of pigeon breast, unit g, X928.551nm、X938.431nm、X964.774nm、X1025.669nm、
X1070.091nm、X1134.246nm、X1155.631nm、X1216.498nm、X1308.658nm、X1343.236nm、X1364.648nm、X1435.516nm、
X1681.829nm、X1686.804nm, respectively wavelength 928.551nm, 938.431nm, 964.774nm, 1025.669nm,
1070.091nm、1134.246nm、1155.631nm、1216.498nm、1308.658nm、1343.236nm、1364.648nm、
Spectral reflectance values at 1435.516nm, 1681.829nm, 1686.804nm, coefficient R=0.965 of above formula, just
Root error RMSE=0.626.
The spectrum of Hyperspectral imager is opened handle after 30min is preheated by the application in advance before detection starts
System mode is modulated to most preferably, i.e., sweep speed is 6.54mm/s, time for exposure 4.65ms.
The application, the spectrum picture pretreatment is to be carried out black and white board correction to original image, to remove the external world
For environment to the influence caused by spectrum picture, updating formula is as follows：
Wherein C is the image after correction, and R is original spectrum image；B is blackboard image, and reflectivity 0%, W is blank
Image, reflectivity 99.9%.
The invention has the advantages that：The present invention only needs to obtain the spectroscopic data of sample, most optimum wavelengths
Under reflectance value be brought directly to can be obtained the weight of Fresh Grade Breast sample in the prediction model of mathematical model, substantially increase by
The efficiency of weighing of sample is, it can be achieved that the extensive online weighing of chicken meat sample detects.
Description of the drawings
Fig. 1 is all band spectral signature figure of 89 Fresh Grade Breast samples；
Fig. 2 is extraction of the regression coefficient method to Fresh Grade Breast most optimum wavelengths；
Fig. 3 is the relationship between the predicted value and measured value of Fresh Grade Breast weight.
Specific implementation mode
It elaborates to the present invention with reference to embodiment.
Embodiment
(1) fresh grade breast in the present embodiment now killed fresh morning purchased from the local market of farm produce on the day of being
Fresh Grade Breast.The fresh pigeon breast of purchase is divided into the sample of 3cm*3cm*1cm (long * wide * high) in laboratory, obtains 89 samples altogether
Product, then 7 parts are being divided into, it is individually placed in the disposable plastic box with lid, by the one piece of placement of box carry sample
It is stored in 4 DEG C of refrigerator, was taken out at 0,1,2,3,4,5,6 day carry out next step experiment respectively；
(2) before on-test, the spectrum of Hyperspectral imager is opened 30min in advance and is preheated, after system
System state is modulated to most preferably, i.e., sweep speed is 6.54mm/s, time for exposure 4.65ms, this near-infrared Hyperspectral imager
Detectable wave-length coverage is in 900-1700nm；
(3) EO-1 hyperion of Fresh Grade Breast sample during different refrigerations is acquired using the near-infrared Hyperspectral imager debugged
Image；
(4) it is weighed immediately to it using position balance very much to the sample for acquiring high spectrum image, records its weight
(g)；
The weight of 89 samples according to being ranked sequentially from small to large, data statistics such as the following table 1：
The weight data of 189 samples of table counts
(5) high spectrum image of acquisition is pre-processed, i.e., original image is carried out black and white board correction, to remove the external world
For environment to the influence caused by spectrum picture, updating formula is as follows：
To obtain the spectral reflectance values of chicken sample；
(6) spectroscopic data in spectrum picture in area-of-interest (ROI) is extracted, extracts all band of 89 samples respectively
Spectroscopic data；As a result such as Fig. 1：
(7) offset minimum binary (PLSR) method associated steps (4) and the weight of step (6) calibration set sample and spectrum number are used
The inner link of all band (486 wavelength) between, that is, establish full wave PLSR prediction models.Use coefficient R
The precision and stability of institute's established model is evaluated with root-mean-square error RMSE, when R is got over hour closer to 1, RMSE, model
Precision it is more higher more stable, and cross validation collection is also a kind of inspection to built calibration set model, when the related coefficient of the two
With root-mean-square error closer to when, show that the model of calibration set is better.As a result such as table 2：
2 all band of table Fresh Grade Breast weight PLSR models
The coefficient R for the PLSR models that calibration set is established as can be drawn from Table 2 is 0.978, and root-mean-square error is
0.494, and the related coefficient of cross validation collection is 0.947, root-mean-square error 0.770 shows built calibration set model not
But precision is high and relatively stablizes.
(8) it in order to optimize the PLSR prediction models obtained by (7) step, is carried out of 486 all bands using regression coefficient method
14 most optimum wavelengths are taken out, as a result such as Fig. 2：
14 most optimum wavelengths are extracted out of all band using regression coefficient method as can be drawn from Figure 2, respectively
928.551nm、938.431nm、964.774nm、1025.669nm、1070.091nm、1134.246nm、1155.631nm、
1216.498nm、1308.658nm、1343.236nm、1364.648nm、1435.516nm、1681.829nm、1686.804nm。
(9) weight of the calibration set chicken meat sample obtained come establishment step (4) using offset minimum binary (PLSR) method and
Prediction model between 14 most optimum wavelengths that step (8) is extracted, result such as table 3：
The PLSR models for the prediction Fresh Grade Breast weight that 3 most optimum wavelengths of table are established
Calibration set PLSR models coefficient R=0.965 established using most optimum wavelengths number can be obtained from table, just
The related coefficient of root error RMSEC=0.626, cross validation collection are 0.953, root-mean-square error 0.738, then calibration set and friendship
The related coefficient and root-mean-square error of the model of fork verification collection are all very close to therefore the precision and stability of calibration set model is all very
It is good.
(10) the PLSR calibration models of the most optimum wavelengths obtained are as follows：
YW=8.393+116.94X928.551nm-43.965X938.431nm+53.572X964.774nm-87.349X1025.669nm+
80.612X1070.091nm-89.666X1134.246nm+97.1X1155.631nm-61.053X1216.498nm+39.673X1308.658nm-
46.681X1343.236nm+36.06X1364.648nm+29.746X1435.516nm-62.815X1681.829nm+74.677X1686.804nm, wherein
YWFor the weight (g) of Fresh Grade Breast sample, X928.551nm、X938.431nm、X964.774nm、X1025.669nm、X1070.091nm、X1134.246nm、
X1155.631nm、X1216.498nm、X1308.658nm、X1343.236nm、X1364.648nm、X1435.516nm、X1681.829nm、X1686.804nm, respectively wave
Grow 928.551nm, 938.431nm, 964.774nm, 1025.669nm, 1070.091nm, 1134.246nm,
1155.631nm、1216.498nm、1308.658nm、1343.236nm、1364.648nm、1435.516nm、1681.829nm、
Spectral reflectance values at 1686.804nm.
(11) it tests：
1. acquiring the near-infrared high spectrum image of 28 pieces of Fresh Grade Breast samples to be measured, spectrum picture is corrected, it is interested
The knowledge in region and the extraction of spectroscopic data, obtain the spectroscopic data in 28 sample all bands；
2. the reflectance value under 14 characteristic wavelengths of each sample to be tested of gained is brought into step (10) to be obtained
Most optimum wavelengths PLSR calibration models in, the prediction gravimetric value of 28 Fresh Grade Breast to be measured can be obtained, as a result such as table 4：
The weight predicted value of the PLSR calibration models of 4 28 samples to be tested of table
3. the weight predicted value of obtained 28 samples is associated with using the gravimetric value measured by position balance very much,
Its related coefficient is up to 0.932, and root-mean-square error 0.943 is related fine between predicted value and measured value, as a result as schemed
3。
It can be obtained from Fig. 3, the difference very little between method of the invention and actually measured Fresh Grade Breast weight shows this
Invention has prodigious feasibility.
The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention.The technology of the industry
Personnel are it should be appreciated that the present invention is not limited to the above embodiments, and the above embodiments and description only describe this
The principle of invention, without departing from the spirit and scope of the present invention, various changes and improvements may be made to the invention, these changes
Change and improvement all fall within the protetion scope of the claimed invention.The claimed scope of the invention by appended claims and its
Equivalent thereof.
Claims (4)
1. application of the near-infrared high light spectrum image-forming technology in chicken weight quickly detection.
2. application according to claim 1, which is characterized in that its step are as follows：
Step 1 acquires the high spectrum image of Fresh Grade Breast using the near-infrared Hyperspectral imager debugged；
The high spectrum image of step 2, acquisition is pre-processed, to obtain the reflectance value of spectrum；
The data of acquisition are substituted into following formula by step 3：
YW=8.393+116.94X928.551nm-43.965X938.431nm+53.572X964.774nm-87.349X1025.669nm+
80.612X1070.091nm-89.666X1134.246nm+97.1X1155.631nm-61.053X1216.498nm+39.673X1308.658nm-
46.681X1343.236nm+36.06X1364.648nm+29.746X1435.516nm-62.815X1681.829nm+74.677X1686.804nm,
Wherein YWFor the weight of pigeon breast, unit g, X928.551nm、X938.431nm、X964.774nm、X1025.669nm、X1070.091nm、
X1134.246nm、X1155.631nm、X1216.498nm、X1308.658nm、X1343.236nm、X1364.648nm、X1435.516nm、X1681.829nm、
X1686.804nm, respectively wavelength 928.551nm, 938.431nm, 964.774nm, 1025.669nm, 1070.091nm,
1134.246nm、1155.631nm、1216.498nm、1308.658nm、1343.236nm、1364.648nm、1435.516nm、
Spectral reflectance values at 1681.829nm, 1686.804nm, coefficient R=0.965 of above formula, root-mean-square error RMSE=
0.626。
3. application according to claim 1, which is characterized in that before detection starts, the spectrum of Hyperspectral imager is carried
Front opening 30min is modulated to system mode most preferably after being preheated, i.e., sweep speed is 6.54mm/s, and the time for exposure is
4.65ms。
4. application according to claim 1, which is characterized in that the spectrum picture pretreatment is to be carried out to original image
Black and white plate corrects, and to remove external environment to the influence caused by spectrum picture, updating formula is as follows：
Wherein C is the image after correction, and R is original spectrum image；B is blackboard image, and reflectivity 0%, W is blank figure
Picture, reflectivity 99.9%.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810228627.0A CN108593589A (en) | 2018-03-19 | 2018-03-19 | Application of the near-infrared high light spectrum image-forming technology in chicken weight quickly detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810228627.0A CN108593589A (en) | 2018-03-19 | 2018-03-19 | Application of the near-infrared high light spectrum image-forming technology in chicken weight quickly detection |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108593589A true CN108593589A (en) | 2018-09-28 |
Family
ID=63626826
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810228627.0A Pending CN108593589A (en) | 2018-03-19 | 2018-03-19 | Application of the near-infrared high light spectrum image-forming technology in chicken weight quickly detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108593589A (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103674864A (en) * | 2013-11-12 | 2014-03-26 | 浙江大学 | Fish water content distribution detection method based on hyperspectral imaging technology |
-
2018
- 2018-03-19 CN CN201810228627.0A patent/CN108593589A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103674864A (en) * | 2013-11-12 | 2014-03-26 | 浙江大学 | Fish water content distribution detection method based on hyperspectral imaging technology |
Non-Patent Citations (2)
| Title |
|---|
| 熊振杰: "基于高光谱成像技术的鸡肉品质快速无损检测", 《中国优秀硕士学位论文全文数据库》 * |
| 程国首 等: "基于高光谱图像技术的新疆红富士苹果重量预测", 《新疆农业大学学报》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| ElMasry et al. | Recent applications of multispectral imaging in seed phenotyping and quality monitoring—An overview | |
| Sunli et al. | Non‐destructive detection for mold colonies in rice based on hyperspectra and GWO‐SVR | |
| CN103439285B (en) | A kind of fillet Noninvasive Measuring Method of Freshness based on high light spectrum image-forming | |
| Weng et al. | Non-destructive detection of strawberry quality using multi-features of hyperspectral imaging and multivariate methods | |
| An et al. | Advances in infrared spectroscopy and hyperspectral imaging combined with artificial intelligence for the detection of cereals quality | |
| CN108444924B (en) | Method for discriminating storage period of tea by applying hyperspectral image technology | |
| Mortensen et al. | The use of multispectral imaging and single seed and bulk near-infrared spectroscopy to characterize seed covering structures: Methods and applications in seed testing and research | |
| Aznan et al. | Computer vision and machine learning analysis of commercial rice grains: A potential digital approach for consumer perception studies | |
| CN105136737A (en) | Method for fast measuring content of potato flour in steamed buns based on near infrared spectrums | |
| Yang et al. | Assessment of grain harvest moisture content using machine learning on smartphone images for optimal harvest timing | |
| Sun et al. | Identification of eggs from different production systems based on hyperspectra and CS-SVM | |
| Mu et al. | Non‐destructive detection of blueberry skin pigments and intrinsic fruit qualities based on deep learning | |
| Zhou et al. | Machine learning modeling and prediction of peanut protein content based on spectral images and stoichiometry | |
| Tian et al. | Research on apple origin classification based on variable iterative space shrinkage approach with stepwise regression–support vector machine algorithm and visible‐near infrared hyperspectral imaging | |
| Wang et al. | Evaluating the nutritional properties of food: a scoping review | |
| Punalekar et al. | Assessing suitability of Sentinel-2 bands for monitoring of nutrient concentration of pastures with a range of species compositions | |
| Ebadi et al. | Accurate prediction of nutritional value of sorghum grain using image analysis | |
| CN106018292A (en) | Non-destructive testing device for protein conformation in egg white and method of non-destructive testing device | |
| Zhang et al. | Rapid determination of the oil and moisture contents in Camellia gauchowensis Chang and Camellia semiserrata Chi seeds kernels by near-infrared reflectance spectroscopy | |
| Song et al. | Non-destructive detection of moisture and fatty acid content in rice using hyperspectral imaging and chemometrics | |
| Forte et al. | Quality Evaluation of Fair-Trade Cocoa Beans from Different Origins Using Portable Near-Infrared Spectroscopy (NIRS) | |
| Zhang et al. | Non-destructive hyperspectral imaging for rapid determination of catalase activity and ageing visualization of wheat stored for different durations | |
| CN108593589A (en) | Application of the near-infrared high light spectrum image-forming technology in chicken weight quickly detection | |
| CN107643255A (en) | A kind of method of Non-Destructive Testing hatching egg middle and later periods addled egg | |
| CN110609011A (en) | Near-infrared hyperspectral detection method and system for starch content of single-kernel corn seeds |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180928 |
|
| WD01 | Invention patent application deemed withdrawn after publication |