CN201392319Y - Meat online lossless detecting device based on the combination of images and spectrum information - Google Patents
Meat online lossless detecting device based on the combination of images and spectrum information Download PDFInfo
- Publication number
- CN201392319Y CN201392319Y CN200920117340U CN200920117340U CN201392319Y CN 201392319 Y CN201392319 Y CN 201392319Y CN 200920117340 U CN200920117340 U CN 200920117340U CN 200920117340 U CN200920117340 U CN 200920117340U CN 201392319 Y CN201392319 Y CN 201392319Y
- Authority
- CN
- China
- Prior art keywords
- meat
- video camera
- image
- quality
- information
- 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.)
- Expired - Fee Related
Links
Images
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The utility model discloses a meat online lossless detecting device based on the combination of images and spectrum information, which comprises a conveyor belt, a spectrum collecting chamber, an image collecting chamber, a camera hanging cross beam, an imaging light source, a camera, Y-shaped optical fiber, a spectrometer, a spectrum light source, a data processor PC, two photoelectric sensors, a motor and a conveyor belt support. The device collects images and spectrum information of meat on line, extracts the characteristic information presenting the quality of the meat, builds a characteristic layer data merging quantitative forecasted model based on the image characteristics, the spectrum characteristics and images and spectrums, utilizes the obtained quantitative forecasted model to combine with a quality estimating system, builds a decision layer data merging model based on the images and the spectrums, and detects the meat on line. The device utilizes the complementarity and the redundancy of the image information and the spectrum information according to the optical characteristics of the meat to realize grading raw materials, detecting the quality and improving the manufacturing efficiency during the meat processing procedures.
Description
Technical field
The utility model relates to a kind of on-line measuring device at agricultural and animal products, refers in particular to the online the cannot-harm-detection device of meat quality based on image and spectral information fusion.
Background technology
Meat and meat products are one of human main food sources.Along with the raising of the level of consumption, the quality of meat products has caused that the consumer extensively payes attention to; Meat packing enterprise obtains stronger competitive power in order to guarantee the quality of product in intense market competition, raw meat and processing semi-manufacture are carried out quality monitoring and control, and product is carried out by the qualitative classification price, with the economic worth of maximization product.
It is to adopt artificial sensory evaluation method and chemical analysis method that traditional meat quality detects.Artificial sensory evaluation need be through the evaluation personnel of professional training, and the evaluation result subjectivity is strong, and is repeatable poor, and it is big to carry out large batch of online detection labour intensity; Chemical analysis method is to carry out destructiveness through chemical method to detect, complex steps, and testing result relies on technical level of operators and skill level, is not suitable for the online detection in the actual production process.
Machine vision technique is realized information acquisition by various imaging systems, utilize image processing techniques to extract and explain the feature of acquisition target by computing machine, in conjunction with various algorithm for pattern recognitions, can carry out quantitatively object, describe qualitatively and classification, in the agricultural and animal products Quality Detection, obtain using widely.Utilize machine vision technique to detect indexs such as the color of meat, texture distribution, shape size, bones, further analyzing and processing can also detect indexs such as the pH value of meat, elasticity, tender degree.Spectral analysis technique is a kind of effective ways to the Non-Destructive Testing of agricultural and animal products quality, and it utilizes agricultural and animal products that the characteristics such as absorption, scattering, reflection and transmission of light are determined the agricultural and animal products quality; Utilize the spectrum detection technique can be to the chemical composition of meat products, the pH value, and index such as tender degree detects.
This class is fast, accurate based on the Dynamic Non-Destruction Measurement speed of agricultural and animal products optical characteristics, be not subjected to subjective man's activity, the researcher has launched a large amount of research both at home and abroad, but mainly be to utilize image information or meat product is carried out classification with single index, can not inside and outside a plurality of index of quality of meat products be detected with spectral information.Meat products quality complexity, and its quality is relevant with its purposes, should be by its inside and outside many index Comprehensive Assessment.Therefore utilize the image of sample and the information of spectrum two aspects, utilization has the width of the information expansion Quality Detection of complementary character, utilization has the precision of the information raising quality evaluation of redundant character, inside and outside index to meat products detects, utilize decision-making level's information fusion, in conjunction with the quality requirements system in the actual production, meat quality is carried out accurate comprehensively evaluation carry out classify and grading, for optimizing the meat products production technology, improve meat products production economic benefit, ensureing that consumption of resident rights and interests and food quality have very important practical sense.
Summary of the invention
The purpose of this utility model is to provide a kind of online the cannot-harm-detection device of meat quality based on image and spectral information fusion.Image information and spectral information by the online acquisition meat, utilize image processing techniques and spectral analysis technique to carry out feature extraction to the information that obtains, obtain to characterize the characteristic information of meat index, set up each testing index based on characteristics of image, based on quantitative forecast model spectral signature and that merge based on image and spectrum characteristics layer data, utilize the quantitative forecast result to carry out decision-making level's data fusion, meat is carried out online detection evaluation.
The technical solution adopted in the utility model is as follows:
Comprise that travelling belt, spectra collection chamber, image acquisition chamber, video camera articulate crossbeam, imaging source, video camera, y-type optical fiber, spectrometer, spectroscopic light source, data processing PC, two photoelectric sensors, motor, travelling belt support.The video camera that video camera is fixed on middle, image acquisition chamber articulates on the crossbeam, imaging source is installed around the video camera, first photoelectric sensor is fixed on the image acquisition chamber sidewall of video camera below, the travelling belt of placing meat is positioned at the video camera below and passes the image acquisition chamber, the motor-driven travelling belt rotates, the spectra collection chamber is installed in the travelling belt top, second photoelectric sensor is fixed on spectra collection chamber sidewall, meat enters the spectra collection chamber, second photoelectric sensor detects sample, send a signal to spectrometer, the spectrometer collection sends to data processing PC through the spectral information of y-type optical fiber transmission, meat transmits continuously and enters the image acquisition chamber, and the detected signal of first photoelectric sensor is given video camera, and the image information of gathering meat is connected to data processing PC.
A kind of method of the online Non-Destructive Testing of meat quality of merging based on image and spectral information:
Gather in real time meat image information and spectral information, carry out information fusion and handle, the integrated quality of meat is detected, its concrete steps are as follows:
1) set up online detection model:
According to the requirement of meat quality index, determine to detect the weight of index and each index, set up the grade estimation system; The image of online acquisition meat and spectral information utilize image processing techniques and spectral analysis technique that the information that obtains is carried out data processing, extract the characteristic information that characterizes index to be detected; Off-line carries out the physical and chemical testing of each index to meat; The spectrum that utilize to extract and each index actual value of image feature information and test, foundation based on image, based on quality quantitative forecast model spectrum or image and the fusion of spectral signature layer; In conjunction with the grade estimation system of setting up, carry out decision-making level's fusion treatment, set up the online detection model of meat quality based on image and spectral information fusion;
2) carry out online detection:
During online detection, meat image information to be measured and spectral information are gathered the back in real time and by the meat products Quality Detection model of setting up the image and the spectral information of input are handled, extract feature, predict by single meat quality quantitative forecast model, meat quality is carried out online quality grading evaluation by decision-making level's data fusion model.
The image information of described online acquisition meat, be image in ultraviolet, visible light, near infrared or the infrared wavelength range, Image Information Processing comprises that the compensation of moving image and correction, background segment, target area extract, the characteristics of image that extracts comprises that color, the correlated characteristic that the texture distribution is described, Fat Distribution reaches the index that requires with meat pH value, elasticity and tender degree actual production extract, and foundation is based on the quality prediction model of image information.
After described spectral information is gathered,, correction level and smooth and differential preprocess method in conjunction with spectrum, extract the characteristic information that characterizes meat chemical composition, pH value, color, elasticity and tender degree index, comprise that characteristic wavelength, spectrum get main gene through decomposing gained, set up the quality prediction model based on spectral information.
The image of described extraction and spectrum characteristics value utilize artificial neural network, support vector machine method to carry out the characteristic layer information fusion; Adopt D-S evidence theory, Bayesian network method to carry out decision-making level's information fusion.
Described meat quality forecast model comprises the attribute classification system in each stage of fresh meat, raw meat or meat products meat products and production.
The beneficial effect that the utlity model has is:
1, the utility model is according to the optical characteristics of meat, organically combine image processing techniques and spectral analysis technique, adopt the information fusion disposal route, utilize the complementarity and the redundancy of image information and spectral information, combine with the appraisement system of the actual production processing meat quality that requires, meat quality is evaluated.
2, the utility model online raw material classification of being used for the meat products process of manufacture is selected, process quality control, meat products end product quality detect, auxiliary and replace professional testing staff, the liberation labour gets rid of artificial subjective factor, enhances productivity.
The utility model utilizes the optical characteristics of meat, image information and spectral information are carried out Data Fusion, inside and outside a plurality of index of quality to meat are measured, classify according to the requirement to each quality in the actual production, be used for the online detection of meat products process of manufacture, intelligent classification processing, the raw meat in the meat products process of realizing cold fresh meat quality classified, the quality of monitoring raw meat changes, can optimize technology for processing meat food, strict control product quality, the economic worth of raising meat and meat products.
Description of drawings
Fig. 1 is the utility model pick-up unit structural representation.
Among the figure: 1. travelling belt, 2. spectra collection chamber, 3. image acquisition chamber, 4. video camera articulates crossbeam, 5. imaging source, 6. video camera, 7.Y type optical fiber, 8. spectra collection instrument, 9. spectroscopic light source, 10. data processing PC, 11. photoelectric sensor, 12. photoelectric sensors, 13. motors, 14. travelling belt supports.
Fig. 2 is the utility model embodiment (fresh pork flavor evaluation) technology path synoptic diagram.
Embodiment
As shown in Figure 1, the utility model comprises that travelling belt 1, spectra collection chamber 2, image acquisition chamber 3, video camera articulate crossbeam 4, imaging source 5, video camera 6, y-type optical fiber 7, spectrometer 8, spectroscopic light source 9, data processing PC10, two photoelectric sensors 11,12, motor 13, travelling belt support 14; The video camera that video camera 6 is fixed on 3 middles, image acquisition chamber articulates on the crossbeam 4, imaging source 5 is installed around the video camera 6, first photoelectric sensor 11 is fixed on image acquisition chamber 3 sidewalls of video camera 6 belows, the travelling belt 1 of placing meat is positioned at video camera 6 belows and passes image acquisition chamber 3, motor 13 drives travelling belt 1 and rotates, spectra collection chamber 2 is installed in travelling belt 1 top, second photoelectric sensor 12 is fixed on spectra collection chamber 2 sidewalls, meat enters spectra collection chamber 2, second photoelectric sensor 12 detects sample, send a signal to spectrometer 8, the spectral information that spectrometer 8 is gathered through y-type optical fiber 7 transmission sends to data processing PC10, meat transmits continuously and enters image acquisition chamber 3, the first photoelectric sensors, 11 detected signals to video camera 6, and the image information of gathering meat is connected to data processing PC10.
Be example with hierarchical detection in the present embodiment to edible fresh meat, longissimus dorsi muscle with the pork trunk is an object, detect PSE (pale, soft, exudative) meat in the fresh meat and DFD (dry, hard, look dark) meat, and intramuscular fat content is estimated as an edible quality index.
As shown in Figure 2, the implementation process of present embodiment is as follows: according to the characteristics of PSE meat and DFD meat, determining to detect index is yellowish pink (adopting CIE L*a*b* colour system), elasticity, tender degree, seepage of water, pH value, fat content (2%～4% for best), sets up the fresh pork appraisement system; Gather PSE meat, DFD meat, when the image information of normal fresh meat sample and spectral information, sample levels is placed on the travelling belt 1 that is driven by motor 13, enter spectra collection chamber 2, be detected during through second photoelectric sensor 12, second photoelectric sensor 12 sends a signal to spectrometer 8, the sample spectral signals that spectrometer 8 is gathered by y-type optical fiber 2 transmission, pass to data processing PC 10, sample transmits continuously and enters image acquisition chamber 3, be detected during through first photoelectric sensor 11, the first photoelectric sensor device 11 sends a signal to video camera 6, and the image information of video camera 6 collected specimens passes to data processing PC 10; Off-line carries out code test to each index physical and chemical indexs such as sample CIE L*a*b* color, elasticity, tender degree, seepage of water, pH value, fat contents and measures; The image information that obtains is carried out pre-service, comprise image is carried out motion compensation and correction, extract the longissimus dorsi muscle zone; Calculate the target area average RGB value of extracting and marble grain distribution situation, and analyze the correlativity of itself and other each index, searching can characterize the characteristics of image of pH value, elasticity, tender degree, seepage of water, fat content; To the spectrum that obtains proofread and correct, smoothly, multiple preprocess method such as differential, adopt progressively linear regression, principal component analysis (PCA), partial least square method, successive projection is sought definite features that can characterize each index such as track algorithm, genetic algorithm, comprise major component, the offset minimum binary main gene, characteristic wavelength etc., the forecast model of a foundation index; Utilize characteristics of image, spectral signature and spectrum and characteristics of image layer information fusion to handle, set up the forecast model of each index; The bond quality overall evaluation system carrying out decision-making level's information fusion processing, is judged its quality to each index prediction result, sets up the meat products quality on-line detection system; During online detection, detected sample spectrum is with the spectra collection process of above-mentioned modeling sample, after spectrum and the image acquisition, import data processing PC (10) into, data processing PC carries out pre-service, feature extraction according to the quality on-line detection system of setting up to raw data, the fusion of characteristic layer and data Layer, judge that revise product are PSE (pale, soft, exudative) meat or DFD (dry, hard, look dark) meat, still are normal meat, and provide its fat content.
Claims (1)
1, a kind of online the cannot-harm-detection device of meat based on image and spectral information fusion, it is characterized in that: comprise that travelling belt (1), spectra collection chamber (2), image acquisition chamber (3), video camera articulate crossbeam (4), imaging source (5), video camera (6), y-type optical fiber (7), spectrometer (8), spectroscopic light source (9), data processing PC (10), two photoelectric sensors (11,12), motor (13), travelling belt support (14); The video camera that video camera (6) is fixed on middle, image acquisition chamber (3) articulates on the crossbeam (4), imaging source (5) is installed around the video camera (6), first photoelectric sensor (11) is fixed on image acquisition chamber (3) sidewall of video camera (6) below, the travelling belt (1) of placing meat is positioned at video camera (6) below and passes image acquisition chamber (3), motor (13) drives travelling belt (1) and rotates, spectra collection chamber (2) is installed in travelling belt (1) top, second photoelectric sensor (12) is fixed on spectra collection chamber (2) sidewall, meat enters spectra collection chamber (2), second photoelectric sensor (12) detects sample, send a signal to spectrometer (8), the spectral information that spectrometer (8) is gathered through y-type optical fiber (7) transmission sends to data processing PC (10), meat transmits continuously and enters image acquisition chamber (3), the detected signal of first photoelectric sensor (11) is given video camera (6), and the image information of gathering meat is connected to data processing PC (10).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200920117340U CN201392319Y (en) | 2009-04-09 | 2009-04-09 | Meat online lossless detecting device based on the combination of images and spectrum information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200920117340U CN201392319Y (en) | 2009-04-09 | 2009-04-09 | Meat online lossless detecting device based on the combination of images and spectrum information |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN201392319Y true CN201392319Y (en) | 2010-01-27 |
Family
ID=41599140
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200920117340U Expired - Fee Related CN201392319Y (en) | 2009-04-09 | 2009-04-09 | Meat online lossless detecting device based on the combination of images and spectrum information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN201392319Y (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102519962A (en) * | 2011-12-07 | 2012-06-27 | 中国农业大学 | Pork quality grade non-destructive detection system |
| CN102621079A (en) * | 2012-03-19 | 2012-08-01 | 齐齐哈尔大学 | Method and device for detecting acrylamide content in barbecued meat product on line |
| CN102707024A (en) * | 2012-05-04 | 2012-10-03 | 中国水产科学研究院南海水产研究所 | Construction method of tilapia fillet quality evaluation model based on proteins and enzymes in muscle |
| CN103063813A (en) * | 2012-12-26 | 2013-04-24 | 江苏大学 | Rapid nondestructive on-line detection method and apparatus for freshness of fishes |
| CN106841540A (en) * | 2016-12-29 | 2017-06-13 | 上海遥薇(集团)有限公司 | A kind of food materials safety detection device and method |
| WO2018209383A1 (en) * | 2017-05-16 | 2018-11-22 | MEQ Probe Pty Ltd | Methods and systems for assessing quality of a meat product |
| CN110470804A (en) * | 2019-06-30 | 2019-11-19 | 安徽宇艳智能机械科技有限公司 | A kind of intelligent improved device of the improvement meat quality based on hyperfrequency acoustic wave vibrations |
| CN110987826A (en) * | 2019-12-16 | 2020-04-10 | 江苏大学 | Meat quality detection device |
| CN113057598A (en) * | 2021-04-20 | 2021-07-02 | 中国农业大学 | Meat quality grading method and system for animal living body slaughterless |
-
2009
- 2009-04-09 CN CN200920117340U patent/CN201392319Y/en not_active Expired - Fee Related
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102519962A (en) * | 2011-12-07 | 2012-06-27 | 中国农业大学 | Pork quality grade non-destructive detection system |
| CN102621079A (en) * | 2012-03-19 | 2012-08-01 | 齐齐哈尔大学 | Method and device for detecting acrylamide content in barbecued meat product on line |
| CN102707024A (en) * | 2012-05-04 | 2012-10-03 | 中国水产科学研究院南海水产研究所 | Construction method of tilapia fillet quality evaluation model based on proteins and enzymes in muscle |
| CN102707024B (en) * | 2012-05-04 | 2015-07-22 | 中国水产科学研究院南海水产研究所 | Construction method of tilapia fillet quality evaluation model based on proteins and enzymes in muscle |
| CN103063813A (en) * | 2012-12-26 | 2013-04-24 | 江苏大学 | Rapid nondestructive on-line detection method and apparatus for freshness of fishes |
| CN103063813B (en) * | 2012-12-26 | 2014-12-24 | 江苏大学 | Rapid nondestructive on-line detection method and apparatus for freshness of fishes |
| CN106841540A (en) * | 2016-12-29 | 2017-06-13 | 上海遥薇(集团)有限公司 | A kind of food materials safety detection device and method |
| WO2018209383A1 (en) * | 2017-05-16 | 2018-11-22 | MEQ Probe Pty Ltd | Methods and systems for assessing quality of a meat product |
| US11519862B2 (en) | 2017-05-16 | 2022-12-06 | MEQ Probe Pty Ltd | Methods and systems for assessing quality of a meat product |
| AU2018271137B2 (en) * | 2017-05-16 | 2023-11-30 | MEQ Probe Pty Ltd | Methods and systems for assessing quality of a meat product |
| CN110470804A (en) * | 2019-06-30 | 2019-11-19 | 安徽宇艳智能机械科技有限公司 | A kind of intelligent improved device of the improvement meat quality based on hyperfrequency acoustic wave vibrations |
| CN110987826A (en) * | 2019-12-16 | 2020-04-10 | 江苏大学 | Meat quality detection device |
| CN110987826B (en) * | 2019-12-16 | 2024-05-17 | 江苏大学 | Meat quality detection device |
| CN113057598A (en) * | 2021-04-20 | 2021-07-02 | 中国农业大学 | Meat quality grading method and system for animal living body slaughterless |
| CN113057598B (en) * | 2021-04-20 | 2022-02-11 | 中国农业大学 | Meat quality grading method and system for animal living body slaughterless |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101551341B (en) | Meat online non-destructive testing method and apparatus based on integration of image and spectrum information | |
| CN201392319Y (en) | Meat online lossless detecting device based on the combination of images and spectrum information | |
| CN1995987B (en) | Non-destructive detection method for agricultural and animal products based on hyperspectral image technology | |
| Tao et al. | A method for nondestructive prediction of pork meat quality and safety attributes by hyperspectral imaging technique | |
| CN101251526B (en) | Method and apparatus for nondestructively testing food synthetic quality | |
| Yu et al. | Nondestructive freshness discriminating of shrimp using visible/near-infrared hyperspectral imaging technique and deep learning algorithm | |
| Barbin et al. | Non-destructive determination of chemical composition in intact and minced pork using near-infrared hyperspectral imaging | |
| CN108663339B (en) | On-line detection method for mildewed corn based on spectrum and image information fusion | |
| ElMasry et al. | Non-destructive determination of water-holding capacity in fresh beef by using NIR hyperspectral imaging | |
| Qiao et al. | Pork quality and marbling level assessment using a hyperspectral imaging system | |
| Tao et al. | Simultaneous determination of tenderness and Escherichia coli contamination of pork using hyperspectral scattering technique | |
| Xing et al. | Combination of chemometric tools and image processing for bruise detection on apples | |
| AU2010203357B2 (en) | System and method for analyzing properties of meat using multispectral imaging | |
| Zhang et al. | Rapid evaluation of texture parameters of Tan mutton using hyperspectral imaging with optimization algorithms | |
| CN104048939A (en) | Near infrared rapid detection method for blood sugar content in live pig blood | |
| Liu et al. | Predicting pork loin intramuscular fat using computer vision system | |
| CN204287031U (en) | A kind of online the cannot-harm-detection device of fish freshness based on high light spectrum image-forming technology | |
| Baek et al. | Shortwave infrared hyperspectral imaging system coupled with multivariable method for TVB-N measurement in pork | |
| Chao et al. | On-line inspection of poultry carcasses by a dual-camera system | |
| Beghi et al. | Rapid evaluation of grape phytosanitary status directly at the check point station entering the winery by using visible/near infrared spectroscopy | |
| Sugiyama et al. | NIR spectral imaging with discriminant analysis for detecting foreign materials among blueberries | |
| Shao et al. | Identification of pesticide varieties by detecting characteristics of Chlorella pyrenoidosa using Visible/Near infrared hyperspectral imaging and Raman microspectroscopy technology | |
| CN102628794A (en) | Method for quickly measuring total quantity of livestock meat bacteria based on hyperspectral imaging technology | |
| Furtado et al. | Prediction of pH and color in pork meat using VIS-NIR near-infrared spectroscopy (NIRS) | |
| CN201156043Y (en) | Non-destruction detector for synthetic quality of food |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20100127 Termination date: 20110409 |