CN108663367A - A kind of egg quality lossless detection method based on egg unit weight - Google Patents
A kind of egg quality lossless detection method based on egg unit weight Download PDFInfo
- Publication number
- CN108663367A CN108663367A CN201810296375.5A CN201810296375A CN108663367A CN 108663367 A CN108663367 A CN 108663367A CN 201810296375 A CN201810296375 A CN 201810296375A CN 108663367 A CN108663367 A CN 108663367A
- Authority
- CN
- China
- Prior art keywords
- egg
- volume
- unit weight
- freshness
- image
- 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
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
- G01N21/84—Systems specially adapted for particular applications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/02—Food
- G01N33/08—Eggs, e.g. by candling
Landscapes
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Food Science & Technology (AREA)
- Engineering & Computer Science (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Medicinal Chemistry (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a kind of egg quality lossless detection methods based on egg unit weight of quality of agricultural product technical field of nondestructive testing.The present invention passes through to egg Image Quick Collection and processing, set up the linear relationship between egg area and volume, fast and easy calculates egg unit weight, then establishes the linear relationship between egg unit weight and egg freshness common counter Hough unit, predicts the freshness of egg.When the unit weight ρ >=1.067, egg has high freshness, is suitble to consumer edible；In 1.046~1.067 value range, egg can be eaten the unit weight ρ by consumer；For the unit weight ρ in 0.996~1.046 value range, egg freshness is poor, and it is edible for consumer to be not suitable as shell egg；When the unit weight ρ≤0.996, egg cannot be eaten.As long as the egg image input system that will be obtained, you can the freshness for quickly and easily learning egg is classified egg.
Description
Technical field
The invention belongs to quality of agricultural product technical field of nondestructive testing, and in particular to a kind of egg product based on egg unit weight
Matter lossless detection method.
Background technology
Egg is one of wholefood, and due to many factors such as breeding, feeding environment, heredity, quality is also different.Due to big
Most fresh egg productions ignore processes, the eggshell surfaces such as disinfection, classification, inspection and usually carry a large amount of microorganisms, shorten the goods of fresh egg
The frame phase, edible quality is directly influenced, reduce integral product quality and the market competitiveness, or even is unfavorable for the health of human body.
Egg freshness is to weigh a principal economic indicators of egg quality, and influence the major reason of egg sale.So
In egg storage, processing and circulation, the detection of egg freshness is just particularly important, and also it is necessary to be carried out to egg freshness
Further research.If in egg circulation and sales process, egg can be carried out according to indexs such as the freshness of egg
Classification and Quality Detection, not only contribute to increase economic efficiency, and advantageously improve feeding and management method, improve egg product
Matter.But all the time, the non-destructive testing of egg freshness is always a problem, and people is still used in production Shang Rengyou enterprises
Work detects the quality of egg according to egg, the physical methods such as weigh, and is as a result affected by subjective factor, production efficiency is low.It is close
Year, in terms of the non-destructive testing of egg freshness, there are many new results, are broadly divided into following a few classes.
(1) Machine Vision Detection egg quality is utilized
By using machine, the fluoroscopy images of egg can be obtained, by the analysis to picture, mathematical modulo can be obtained
Type, to predict freshness.
Ding Youchun (learn by algorithm [J] the Hua Zhong Agriculture University that duck's egg dynamic image pre-processed and obtained Color characteristics parameters
Report, 2005,24 (5):It 512-515.) proposes the pretreatment of duck's egg dynamic image and obtains the algorithm of Color characteristics parameters, utilize figure
The red component R of picture identifies egg image outline to judge the boundary point of egg image with this, and searches egg image boundary and meter
Calculate in egg pixel and, while finding out egg core color characteristic region with round automatic Searching Method and extracting characteristics of image color and joining
Number.Wang Qiaohua (egg freshness lossless detection method [J] agricultural mechanical journals based on BP neural network, 2006,37 (1):
Research [J] the Hua Zhong Agriculture University journal of 104-106. Detection System of Egg Fresh Degree Based on Neural Network, 2005,24 (6):630-
632.) the egg freshness lossless detection method based on BP neural network is proposed, using brightness of image (I) component in HIS models
Eggshell color information is extracted, white shell and brown shell egg are effectively classified, color of egg center and coloration (H) is established, brightness (I), satisfies
Relational model between degree (S) component, and utilize the freshness of neural network detection egg.(the egg reflected light such as Wang Qiaohua
Characteristic and its relationship [J] Hua Zhong Agriculture University journal with freshness, 2008,27 (1):140-143.) to be investigated egg anti-
Penetrate the relationship between light characteristic and its freshness.(egg freshness integrates Nondestructive Testing Model and experiment [J] agriculturals to Wei little Biao
Engineering journal, 2009,25 (3):Brightness, the egg shape index that egg color 242-247.) is obtained with machine vision device, use sound
The power area under spectrum of harvester acquisition sound, the barycenter of formant frequency, X-direction, obtain egg freshness and its image
Iptimum relationship between characteristic parameter and sound characteristic parameter, institute's established model differentiate that the accuracy of egg freshness is 92%.It is beautiful
State University of Georgia Patel (Color computer vision and artificial neural networks for
the detection of defects in poultry eggs[J].Artificial Intelligence Review,
1998,12:163-176.) propose based on the color image acquisition system of computer vision and neural network detection have crackle,
The egg of blood cake, stain trains neural network using color histogram, and Billy is more superior with grey level histogram training, especially exists
Fault-tolerant ability is had more in terms of crack detection, the verification and measurement ratio of blood cake is up to 92.8%, and stain is 85%, crackle 87.8%, but
The system is only capable of detecting white shell egg.Zheng Limin (the egg freshness non-destructive testings based on computer vision of China Agricultural University
[J] Journal of Agricultural Engineering, 2009,25 (2):2 high rate industrial digital pick-up lens respectively 335-339.) is used, it is automatic every time
25 egg images are captured, the characteristics of image of egg is analyzed based on computer vision technique, utilizes the characteristics of image (yolk of egg
Index, gas chamber etc.) with actual characteristic correlation is established, predict the freshness of egg.All equality (egg bodies based on machine vision
Product and surface area calculation method [J] agricultural mechanical journals, 2010, (05):168-171+208.) machine vision technique is utilized, it is false
If ideal egg image is symmetrical about vertical diameter, proposes pixel volume and Vp and Pixel surface product and Sp, be given in digital picture
Computational methods.Finally establish the relational model between egg volume V, surface area S and Vp, Sp.Verification experimental verification shows：Egg
Volume predictions model related coefficient is 0.965, rate of accuracy reached 92% in measurement error ± 1cm3；Egg surface accumulates prediction model phase
Relationship number is 0.971, rate of accuracy reached 88% in measurement error ± 1cm2.
(2) optical characteristics is utilized to detect egg quality
Detect egg quality using optical characteristics, mainly utilize the transmission of light, refraction, principle of reflection and egg inside
Quality establishes a kind of relationship, its method for detecting egg quality is established by mathematical model.Near infrared spectrum (NIR) technology is 20
One of with fastest developing speed since the nineties in century, most noticeable spectral analysis technique has analyze speed fast, efficient, easy
In realize lossless on-line analysis the features such as, it is widely used to the quality determination of the agricultural product such as meat, fruits and vegetables.
The country, early in 1989, Wu kept first-class (optical non-destructive detection of egg freshness and classification [J] agricultural engineerings
Report, 1989, (04):Pre-test 64-70.) has been carried out to egg freshness optical detecting method, has been had studied in the transmission and egg of light
Correlativity between component matter, establishes the freshness factor of egg, and has obtained corresponding ranking score dividing value, be egg without
It damages examination criteria and design egg freshness grading plant provides reference.(light characteristic of egg inside quality is lossless by Fang Ruming etc.
Detection [J] Journal of Agricultural Engineering, 1993, (03):102-107.) in order to improve egg inside quality light characteristic non-destructive testing
Precision establishes the optical model of egg, finds out the relationship between whole egg, content, eggshell three's transmissison characteristic.Chen Bin (chickens
Research [J] Jiangsu Science ＆ Engineering Univ. journal of Egg Quality Photoelectric Detection, 1996, (06):1-5.) study the light of egg main component
Transmissison characteristic is composed, analyzes the variation tendency between their spectral transmission curve and resting period, and done largely to egg
Tracking test explores the new method of the spectral-transmission characteristics evaluation egg quality by measurement egg, complete certainly for further design
Dynamic egg quality detection device provides valuable reference data.Zhao Hongxia (the phases of egg Ultra-weak Luminescence and its freshness
It closes and analyzes [J] Journal of Agricultural Engineering, 2004, (02):177-180.) egg is had studied using Ultra-weak Luminescence image detection system
Luminous situation in storage period, it was demonstrated that overall target of the weak light as egg biochemical reactions can be used for measuring egg
Freshness, and have many advantages, such as that method is simple, not damaged, high sensitivity.
Foreign countries, (the Development of a rapid method based on front face such as Kemps
fluorescence spectroscopy for the monitoring of egg freshness:1-evolution of
thick and thin egg albumens[J].European Food Research and Technology,2006,223
(3):The freshness for 303-312.) utilizing visible light/near-infrared reflection technology (VIS/NIR) detection egg, obtains measurement egg
The pH value of content is more accurate than measuring Hough unit.(the Non-destructive freshness such as Alessandro
assessment of shell eggs using FT-NIR spectroscopy[J].Journal of Food
Engineering,2008,89(2):The freshness for 142-148.) using Fourier transformation near infrared detection egg, uses main composition
Prediction model is established in the methods of analysis, Partial Least Squares Regression and offset minimum binary differentiation, and to egg freshness index, (gas chamber is high
Degree, yolk index and Hough unit) prediction related coefficient R2 be respectively 0.72,0.78 and 0.67, and can accurately differentiate chicken
The storage number of days of egg.
(3) acoustic impulse Characteristics Detection egg quality is utilized
Acoustic impulse testing principle is to do spectrum analysis according to ping vibration caused by percussion egg to study egg
Quality characteristic.
Wang Shucai (correlation [J] the Hua Zhong Agriculture University journal of egg percussion response characteristic and its freshness, 2009,
(03):Birds, beasts and eggs detection and hierarchical intelligence robot research 373-376.) are devised, is adopted by tapping voice signal to eggshell
Collection, sampling do Fourier transform, obtain the power spectrum of frequency domain, use using DSP as the sound collection of core, signal processing and fuzzy
Pattern-recognition breakage egg, the Detection accuracy to normal egg are 90%, and the Detection accuracy to damaged egg is 95%；And it can basis
Egg is divided into 3 qualified grades and 1 unqualified grade by Hough unit-sized.(one kind being based on magnetostrictive technology to Lu Wei etc.
Quality of poultry eggs non-destructive testing device and its method [P] China:201410485660.3,2014-09-23.) in existing acoustics
On the basis of characteristic achievement in research, current research detects egg cracks using the method that frequency sweep vibration and support vector machines combine, and comes
The difference for enhancing the vibration signal of lossless egg and crack egg, carries out Model Parameter Optimization by way of cross validation, and identification is accurate
True rate reaches 98%, and solves the problems, such as in the past through the more difficult detection eggshell fine crack of acoustic characteristic.(the Eggshell such as Lin
crack detection based on acoustic impulse response and supervised pattern
recognition[J].Czech J Food Sci,2009,27(6):393-402.) research and utilization acoustic wave vibrations detection eggshell
Intensity, the frequency response that eggshell is excited is struck by a slight percussion system by measuring, uses offset minimum binary data point
Analysis technology establishes eggshell strength prediction model, and the prediction related coefficient of model is 0.75 or more.
(4) dielectric property is utilized to detect egg quality
The change for detecting the node dielectric property of egg is turned to a kind of detection that detection means changes applied to egg quality.
Dong Jianping and Shen Linsheng etc. (differentiate research [J] Journal of Agricultural Engineering of the detection device of fertile egg with bioelectricity,
1996,(03):167-170.) research egg bioelectricity and fertilization relationship, and embryo's electrograph during kind of egg hatching is had studied,
It was found that fertile egg brooding time is related to the frequency of its electrical activity and amplitude, and develop a kind of energy bioelectricity Undamaged determination by
The detection device of smart egg.Flower bud etc. (variation [J] chemurgies of egg storage intermediary's electrical characteristics and fresh quality are studied,
2008,4(4):146-154.) dielectric parameter of the frequency of use between 100k~1MHz and fresh quality with storage time change
Law, the results showed that quality comparison has significant correlation between dielectric constant values corresponding thereto during egg stores.Liu Xi etc.
(using conductivity meter detect egg freshness [J] Food Sciences, 1991, (10):45-47.) conductivity meter is used to detect egg
Freshness.(egg freshness non-destructive testing [J] the agriculture projects based on dielectric property Yu yolk index regression model such as Sun Jun
Journal, 2016, (21):Parallel plate electrode method 290-295.) is used to measure different freshness eggs in temperature as 20 DEG C, relative humidity
Dielectric property parameter under being 1~200kHz for 72%~89%, frequency, analyzes the changing rule of egg dielectric property, and builds
Vertical mathematical model between egg dielectric property and yolk index predicts egg freshness.
(5) odor characteristics are utilized to detect egg quality
In egg liquid decay process, protein generates amine, hydrogen sulfide, methane etc., and fat generates low molecule resin acid alcohol etc., sugar
Class generates the substances such as lower fatty acid, carbon dioxide, methane, hydrogen.Electronic nose can be used to be monitored in real time.
(the Non-destructive egg freshness determination such as Dutta：an electronic
nose based approach[J].Measurement Science and Technology,2003,14(2):190-
191.) people has detected 4 groups of eggs in 20~40d with the gas sensor array of 4 cheap commercial zinc oxide sensor compositions
Freshness in the case of storage is used in combination multivariate statistical method, especially neural network to delimit 3 egg areas of different freshness
Domain, precision 95%.Liu Ming etc. (electronic nose detection egg shelf life freshness variation [J] Journal of Agricultural Engineering, 2010,
(04):317-321.) people is detected the index of quality of Fresh Egg using electronic nose and electronic nose data analysis, obtains chicken
Egg inside quality deteriorates the basic source that the amino-oxide, alkane, alcohols generated is complete egg odor variation, can be sentenced according to it
Disconnected egg freshness.
Evaluating common index to egg freshness at present mainly has Hough unit, proportion, air room height etc..But these refer to
All there are some drawbacks in object detection method.From this, the non-destructive testing technology of egg quality will be an important research
Direction.
Since the process for commonly using egg freshness detection method is cumbersome, egg producer is caused to purchase chicken to egg raiser
The freshness of egg can not be rapidly and accurately held when egg.Pass through non-destructive testing technology, so that it may with quickly to the egg collected into
Row freshness is classified, and effectively improves production efficiency.And the index that unit weight can be used for reflecting egg freshness as one, it can
One side as non-destructive testing research.Conventional method, process is cumbersome, and especially drainage surveys egg volume, needs chicken
Egg impregnates in water, reads front and back volume differences.Range estimation reading easily generates larger error, even Fresh Egg generally holds
Weight is also no more than 1.10, if volume predictions differ 1-2cm3, unit weight will differ 0.02~0.04 or so, resulting to chicken
It influences be huge caused by the judgement of egg freshness.The egg for impregnating water is also faster rotten compared with other eggs corrupt,
Bring loss.Cumbersome operation also will produce certain cost of labor.So a kind of side that can quickly measure egg unit weight of research
Method, which just seems, to be of practical significance.
Invention content
The purpose of the present invention is to propose to a kind of egg quality lossless detection methods based on egg unit weight.Specific technical solution
It is as follows：
A kind of egg quality lossless detection method based on egg unit weight, includes the following steps：
(1) egg image obtains：Obtain egg front view and lateral-view image；
(2) image procossing：Image procossing is carried out using image processing software Matlab R2016a, steps are as follows：
Step a：The extraction of egg contour feature；
Step b：Egg image is corrected using connection domain method；
Step c：Calculate egg image area；
The number for the occupied pixel of egg view section product being partitioned into is counted, as its area, unit are pixel
Point；
(3) drainage surveys egg volume；
(4) egg area-volume-based model is established：
The prediction of egg is calculated using the method for analogy ellipse area calculation formula and spheroid volume calculation formula
Then volume x establishes egg prediction volume x and the direct linear relationship of egg actual volume；
Egg predicts that volume parameter calculation formula is：
Wherein, f is to face the area of pictural surface, and r is the side view area of pictural surface,
By being measured to a collection of egg, the area of two views of egg and the volume of drainage are obtained, to data
It is fitted, obtains a linear equation；
(5) calculating of egg unit weight, formula are：
In formula, ρ is the unit weight of egg, unit g/cm3；M is the quality of egg, unit g；V is egg volume, unit
For cm3；With the increase of storage number of days, egg unit weight declines, and the freshness of egg reduces；
(6) according to the unit weight of surveyed egg, judge the freshness of the egg.
The step of egg image obtains be：Camera position is fixed and ajusted, computer is connected, it is soft to open Image Acquisition
Part chooses positive industry camera, adjusts resolution ratio, removes lens cap, egg to be measured is positioned on egg tray, adjusts light
Source is until be suitble to obtain image；It rotates camera lens and adjusts focal length, until egg edge clear in view, and fixed lens focal length, with
After can obtain front view picture；It chooses the industry camera of side, step to be same as above, obtains lateral-view image；The light source is adopted
The LED light for being 5W with power, intensity of illumination reach 10000Lx.
The drainage surveys the step of egg volume and includes：Small beaker is placed under spilling water cup smallmouth；It is noted in spilling water cup
Full water makes liquid level not have smallmouth, extra water to be flowed into small beaker from spilling water cup smallmouth；Wait for liquid level stabilizing, when water stops flowing,
Small beaker is placed on electronic balance and is weighed, weight m1 is obtained；Then egg to be measured is slowly put into spilling water cup, waits for that liquid level is steady
It is fixed, when water stops flowing, small beaker is weighed again, obtains m2；The weight that difference is water corresponding to egg volume, i.e. m are weighed twice
=m2-m1；The actual volume of egg is obtained by calculating.
The a linear equation is：
V=4.7872857 × 10-7x-1.1920186
In formula, v is egg actual volume, unit cm3；X is that egg predicts volume.
When the unit weight ρ >=1.067, egg has high freshness, is suitble to consumer edible；The unit weight ρ 1.046~
When 1.067 value range, egg can be eaten by consumer；The unit weight ρ is in 0.996~1.046 value range, egg freshness
It is edible for consumer to be not suitable as shell egg for difference；When the unit weight ρ≤0.996, egg cannot be eaten.
Beneficial effects of the present invention are：
The present invention passes through to egg Image Quick Collection and processing, it is established that the linear pass between egg area and volume
System, fast and easy calculate egg unit weight, then establish the line between egg unit weight and egg freshness common counter Hough unit
The freshness of egg is predicted in sexual intercourse.As long as the egg image input system that will be obtained, you can quickly and easily learn egg
Freshness, the production firm of egg can carry out quick nondestructive inspection to egg, to judge egg if appropriate for being processed,
Egg is classified.
Description of the drawings
Fig. 1 is that the present invention is based on the egg quality lossless detection method flow charts of area-volume regression model.
Fig. 2 is egg image collecting device top view illustration, wherein 1- bottom plates, 2- background baffles, 3- are placed in egg tray
On egg, 4- cameras, 5- data lines.
Fig. 3 is egg image procossing and correction map, wherein figure a is extraction red component figure, and figure b is extraction blue component
Figure, figure c are the differential chart of two components, and figure d is the egg contour feature figure being partitioned into, and figure e is to be stayed inside filling egg image
Sky, figure f are revised egg image.
Fig. 4 is to calculate egg image area schematic diagram, wherein figure a- keys in program name in command window, and figure b- is obtained
As a result.
Fig. 5 is the schematic diagram that drainage surveys egg volume.
Fig. 6 is oval and spheroid schematic diagram, wherein figure e- ellipse schematic diagrames scheme f- spheroid schematic diagrames.
Fig. 7 is the scatter plot of egg area-volume regression model.
Fig. 8 is the relationship of egg bracket number of days and unit weight.
Fig. 9 is the relationship of egg bracket number of days and Hough unit.
Figure 10 is the relationship of egg bracket number of days and yolk index.
Figure 11 is the relationship of egg unit weight and Hough unit.
Figure 12 is the relationship of egg unit weight and yolk index.
Figure 13 is the relationship of egg Hough unit and yolk index.
Specific implementation mode
Following embodiment facilitates a better understanding of the present invention.
The present invention is based on the flow chart of the egg quality lossless detection method of area-volume regression model is as shown in Figure 1.
Embodiment 1：The structure of egg area-volume regression model
Experiment obtains an egg image information since third day of laying eggs, every 3d.Experimental procedure is as follows each time：
1. image obtains
Image obtains the XG200 COMS USB digital cameras 2 that camera uses the micro- Seiko in Shenzhen in the present embodiment
Platform, resolution ratio are up to 1600 × 1200, and camera lens uses the tight shot (focal length 8mm) of 3,000,000 pixels, camera to pass through
USB line connects computer.Light source uses power for the LED light of 5W, and intensity of illumination can reach 10000Lx.It is faced to obtain egg
Figure and side view, are individually fixed in appropriate location on tablet so that camera is putting position apart from egg by camera and egg tray
Setting a front surface and a side surface has certain distance, and angle is right angle.Light source is placed on the both sides of camera.Egg image collecting device is bowed
View schematic diagram is shown in Fig. 2.Acquisition image software is S-EYE, can adjust the ginsengs such as resolution ratio, exposure, white balance, contrast immediately
Number, the image collected are JPG formats, and resolution ratio is 1600 × 1200.
Image capture flow is as follows：Fixed camera ajusts camera position, connects computer.S-EYE softwares are opened, are chosen
Positive industry camera, adjustment resolution ratio is 1600 × 1200, remaining parameter depends on the circumstances (general default parameters).It removes
Egg to be measured is positioned on egg tray by lens cap, and adjustment light source is until be suitble to obtain image.It rotates camera lens and adjusts focal length, directly
To egg edge clear in view, and fixed lens focal length.Front view picture can be then obtained, and more by the image got
Entitled " f.jpg ".Choose the industry camera of side, step to be same as above in S-EYE softwares, obtain lateral-view image, and by its
It is renamed as " r.jpg ".
2. image procossing
The present embodiment system uses computer processor for Intel (R) Core (TM) 2.50GHz, physical memory 8GB,
Hard disk is 250GB, video memory 4166MB.Used image processing software is Matlab R2016a.Image processing step is such as
Under：
(1) extraction of egg contour feature
Image f.jpg is imported in Matlab.Sentence：
I=imread (' f.jpg ')
The red component and blue component in image are extracted respectively, and seek their difference.It can be seen in the figure that at this time
Obvious egg.Sentence：
Ib=I (:,:,1)；Ir=I (:,:,3)；II=Ib-Ir
It is binary image, sentence by greyscale image transitions：
Im=im2bw (Im)
Egg profile can be obviously told at this time, but also has blank inside egg, and surrounding also has some interference ranges
Domain needs further to fill, remove.
(2) egg image is corrected using connection domain method
Pixel in the matrix of this size of Im, if any of which pixel its have an other pixel to the upper, lower, left, or right,
Think that they are connections.This measure can delete the single pixel point spread in image.At the same time screening extraction area is more than
Or the connected domain equal to threshold value, threshold value are pixel number.After handling twice in succession, just obtain that be that egg is revised regard
Figure.Sentence：
L=bwlabeln (Im, 4)；S=regionprops (L, ' Area ')；Im2=ismember (L, find
([S.Area]>=n1))
L=bwlabeln (Im2,4)；S=regionprops (L, ' Area ')；Im3=ismember (L, find
([S.Area]>=n2))
N1 and n2 is pixel number in sentence, is depended on the circumstances, acquiescence 100000.
The processing procedure of r.jpg is same as above.
Egg image procossing and amendment are as shown in Figure 3.
(3) egg image area is calculated
As shown in figure 4, the number for the occupied pixel of egg view section product that statistics is partitioned into, as its area,
Unit is pixel.Sentence：
Area_front=length (find (Im3=0))；Area_right=length (find (Fig3=0)).
3. drainage surveys egg volume
Prepare one, 500ml spilling waters cup and 200ml small beakers one, small beaker is placed under spilling water cup smallmouth, such as Fig. 5 institutes
Show.Water is filled in spilling water cup, liquid level is made not have smallmouth, extra water that can be flowed into small beaker from spilling water cup smallmouth.Wait for liquid level
Stablize, when water stops flowing, small beaker is placed on electronic balance and is weighed, weight m1 is obtained.Then egg to be measured is slowly put
Enter in spilling water cup.Wait for that liquid level stabilizing weighs small beaker, obtain m2 again when water stops flowing.It is exactly egg body to weigh difference twice
The weight of the corresponding water of product, i.e. m=m2-m1.In formula, the unit of m, m1, m2 are g.Due to the liquid in spilling water cup be from
Sub- water, density 1.00, so can quickly obtain the corresponding volume of egg by calculating.
4. establishing egg area-volume-based model
In view of the shape of egg is closer to spheroid, so using analogy ellipse area calculation formula and spheroid
The method of volume calculation formula obtains a parameter x, then establishes parameter x and the direct linear relationship of egg actual volume.
Oval and spheroid schematic diagram is respectively as shown in Fig. 6-e and Fig. 6-f.
Assuming that egg is the spheroid of rule, there are three semiaxis a, b and c for tool, and wherein a and b have embodiment, b in front view
There is embodiment in side view with c.The side view of egg is close to round, b and c approximately equals, then
Wherein, r is the side view area of pictural surface, and π is pi.
The front view of egg is close to ellipse, according to elliptical calculation formula S=π ab, then
Wherein, f is to face the area of pictural surface, and π is pi.
Spheroid cubature formula isThree and half axial length a, b and c obtained above are brought into and can be obtained：
Wherein, x is parameter, and k is constant,
By being measured to a collection of egg, the area of two views of egg and the volume of drainage are obtained.To data
It is fitted, obtains the equation that shape is v=ax+b.In formula, v is egg volume, and x is above mentioned parameter, and a and b are normal
Number.
After some measurement, damaged egg is removed, 50 groups of data availables are shared.It is calculated, is built by OriginPro 8
One-variable linear regression mould between vertical egg actual volume v (volume of egg drainage), egg lateral area treated parameter x
Type.Egg area-volume regression model expression formula is：
V=4.7872857 × 10-7x-1.1920186
In formula, v is egg volume, unit cm3；X is parameter, unit cm3。
Scatter plot such as Fig. 7 of egg area-volume regression model.
Correlation and significant result are as shown in table 1.
1 egg area of table-volume analysis of regression model
As it can be seen from table 1 the total sum of squares of the regression model is 230.236, and error sum of squares 13.672, phase relation
Number is 0.972, this shows that model has very high correlation.F distribution tables are looked into, calculated F values are 774.659, are much larger than Fp
(n1, n2), p ＜ 0.001, this shows that the regression model has high conspicuousness.
5. verifying model
For obtained model, to be verified.Need to prepare again egg of the same race several, adopt and measure with the aforedescribed process
The area and displacement of volume of two views of egg.It brings the area of two views of egg into equation and obtains egg prediction volume, it will
Prediction volume is compared with displacement of volume, calculates error, and testing model is used to predict the accuracy of egg volume.
It tests to 60 pieces of eggs, verification egg area-volume regression model, the results are shown in Table 2.The results show that
The regression model can predict the volume of egg by the area of the front view of egg and side view well.
2 egg area of table-volume regression model verification result
Embodiment 2：Practical application of the egg area-volume regression model on prediction egg freshness
1. the egg unit weight of different phase, Hough unit and yolk index
Experiment obtains an egg image information since third day of laying eggs, every 3d, and surveys egg size and Hough unit, egg
Yellow index (being freshness), until 21d.It surveys altogether 7 times, 4 eggs, amount to 28 eggs every time.
Hough unit calculation formula：HU=100log (H-1.7m0.37+7.57)
In formula, H is albumen height, unit mm；M is egg quality, unit g.
Yolk index calculation formula：
In formula, h is yolk height, unit mm；W is yolk diameter, unit mm.
Using egg image information and formula 2-11 calculating parameter x, then egg area-volume regression model is utilized to express
Formula：V=4.7872857 × 10-7X-1.1920186 calculates the volume (prediction volume) of egg.
The calculation formula of egg unit weight is：
In formula, ρ is the unit weight of egg, unit g/cm3；M is the quality of egg, unit g；V is the (prediction of egg volume
Volume), unit cm3.The quality of egg uses balance for one thousandth electronic balance, and producer is New Jersey Ao Haosi instrument
Device Co., Ltd.
Through measurement after a period of time, this collection of egg is obtained in the unit weight of different phase, Hough unit and yolk index,
Such as table 3.Then data are analyzed using OriginPro 8, studies existing relationship between them.
Egg unit weight, Hough unit and the yolk index of 3 different phase of table
2. the relationship of egg unit weight and freshness
To obtaining table 4 after 3 data preparation of table.
4 egg of table with the resting period extend unit weight, Hough unit yolk index variation
According to data in table, is drawn out with OriginPro 8 and increased with number of days, unit weight, Hough unit and yolk index
Change curve, respectively Fig. 8,9 and 10.This three matched curves are analyzed successively, obtain egg bracket number of days and egg
Unit weight, Hough unit and yolk index relationship such as table 5.
The analysis of table 5 egg bracket number of days and egg unit weight, Hough unit and yolk index relationship
In formula, ρ is unit weight, unit g/cm3；Hu is Hough unit；YI is yolk index；D is the time, and unit is day.
Pass through the F that tables look-up to obtain0.001(1,5)=47.18, the F values in table are all higher than this value, illustrate p ＜ 0.001.Thus can recognize
There is very strong correlativities between storage number of days and unit weight, Hough unit and yolk index, have significance difference anisotropic.
As the resting period increases, unit weight, Hough unit and the yolk index of egg all constantly decline therewith.
Equally, unit weight and Hough unit, the relation curve of unit weight and yolk index are drawn out, respectively such as Figure 11 and Figure 12.
This two matched curves are analyzed successively, obtain egg unit weight and egg Hough unit and yolk index relationship such as table 6.
The analysis of table 6 egg unit weight and egg Hough unit and yolk index relationship
Pass through the F that tables look-up to obtain0.001(1,5)=47.18, the F values in table are all higher than this value, illustrate p ＜ 0.001.Thus can recognize
There is very strong correlativities between egg unit weight and Hough unit, yolk index, have significance difference anisotropic.Egg holds
The decline of weight, it is meant that Hough unit and yolk index all decrease.Hough unit and yolk index are new as egg is judged
The index of freshness, establishes the relationship of unit weight and they, is also equal to establish the relationship of unit weight and egg freshness.
Using same method, the relation curve of egg Hough unit and yolk index, such as Figure 13 are drawn out.To curve into
Egg Hough unit is obtained after row regression analysis and yolk index relationship is as shown in table 7.
The analysis of table 7 egg Hough unit and yolk index relationship
Pass through the F that tables look-up to obtain0.001(1,5)=47.12, the F values in upper table are 160.8391 to be much larger than this value, illustrate p ＜
0.001.Thus it is believed that there is very strong correlativities between egg Hough unit and yolk index, there is significant difference
Property.It is concluded that, it can assert that Hough unit and yolk index can be used in the judgement of egg freshness in conjunction with above, and
It can convert mutually.It is generally acknowledged that egg can be divided into 4 classification standards, as shown in table 8.
8 egg of table graduation standard
The egg classification that experiment obtains and the relationship between egg unit weight, freshness are as shown in table 9.Utilize egg area-
Volume regression model effectively can estimate the Hough unit and yolk index of egg by calculating unit weight, so that it is determined that egg is new
Freshness is classified egg.
Relationship between the classification of 9 egg of table, unit weight and freshness
Claims (5)
1. a kind of egg quality lossless detection method based on egg unit weight, which is characterized in that include the following steps：
(1) egg image obtains：Obtain egg front view and lateral-view image；
(2) image procossing：Image procossing is carried out using image processing software Matlab R2016a, steps are as follows：
Step a：The extraction of egg contour feature；
Step b：Egg image is corrected using connection domain method；
Step c：Calculate egg image area；
The number for the occupied pixel of egg view section product being partitioned into is counted, as its area, unit are pixel；
(3) drainage surveys egg volume；
(4) egg area-volume-based model is established：
The prediction volume of egg is calculated using the method for analogy ellipse area calculation formula and spheroid volume calculation formula
Then x establishes egg prediction volume x and the direct linear relationship of egg actual volume；
Egg predicts that volume parameter calculation formula is：
Wherein, f is to face the area of pictural surface, and r is the side view area of pictural surface,
By being measured to a collection of egg, the area of two views of egg and the volume of drainage are obtained, data are carried out
Fitting, obtains a linear equation；
(5) calculating of egg unit weight, formula are：
In formula, ρ is the unit weight of egg, unit g/cm3；M is the quality of egg, unit g；V is egg volume, and unit is
cm3；With the increase of storage number of days, egg unit weight declines, and the freshness of egg reduces；
(6) according to the unit weight of surveyed egg, judge the freshness of the egg.
2. egg quality lossless detection method according to claim 1, which is characterized in that the step that the egg image obtains
Suddenly it is：Camera position is fixed and ajusted, computer is connected, image capture software is opened, chooses positive industry camera, is adjusted
Resolution ratio removes lens cap, and egg to be measured is positioned on egg tray, and adjustment light source is until be suitble to obtain image；Rotate camera lens
Focal length is adjusted, until egg edge clear in view, and fixed lens focal length, it can then obtain front view picture；Choose side
The industry camera in face, step are same as above, and obtain lateral-view image；The light source uses power for the LED light of 5W, and intensity of illumination reaches
To 10000Lx.
3. egg quality lossless detection method according to claim 1, which is characterized in that the drainage surveys egg volume
The step of include：Small beaker is placed under spilling water cup smallmouth；Water is filled in spilling water cup, liquid level is made not have smallmouth, extra water
It is flowed into small beaker from spilling water cup smallmouth；Wait for that small beaker is placed on electronic balance and weighs when water stops flowing by liquid level stabilizing,
Obtain weight m1；Then egg to be measured is slowly put into spilling water cup, waits for that liquid level stabilizing weighs small again when water stops flowing
Beaker obtains m2；The weight that difference is water corresponding to egg volume, i.e. m=m2-m1 are weighed twice；Egg is obtained by calculating
Actual volume.
4. egg quality lossless detection method according to claim 1, which is characterized in that a linear equation
For：
V=4.7872857 × 10-7x-1.1920186
In formula, v is egg actual volume, unit cm3；X is that egg predicts volume.
5. egg quality lossless detection method according to claim 1, which is characterized in that when the unit weight ρ >=1.067,
Egg has high freshness, is suitble to consumer edible；In 1.046~1.067 value range, egg can be disappeared the unit weight ρ
The person of expense is edible；For the unit weight ρ in 0.996~1.046 value range, egg freshness is poor, is not suitable as shell egg for consumer
It is edible；When the unit weight ρ≤0.996, egg cannot be eaten.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810296375.5A CN108663367A (en) | 2018-03-30 | 2018-03-30 | A kind of egg quality lossless detection method based on egg unit weight |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810296375.5A CN108663367A (en) | 2018-03-30 | 2018-03-30 | A kind of egg quality lossless detection method based on egg unit weight |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108663367A true CN108663367A (en) | 2018-10-16 |
Family
ID=63783114
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810296375.5A Pending CN108663367A (en) | 2018-03-30 | 2018-03-30 | A kind of egg quality lossless detection method based on egg unit weight |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108663367A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109813494A (en) * | 2019-02-27 | 2019-05-28 | 西北农林科技大学 | A method of based on measurement mass center non-destructive testing egg freshness |
| CN109937912A (en) * | 2019-04-08 | 2019-06-28 | 武汉科技大学 | A kind of egg categorizing system and method based on machine vision |
| CN110333160A (en) * | 2019-07-18 | 2019-10-15 | 华中农业大学 | Egg freshness non-destructive testing device and its method based on drainage |
| CN110411957A (en) * | 2019-08-28 | 2019-11-05 | 北京农业质量标准与检测技术研究中心 | The lossless method for quick predicting and device of fruit shelf life and freshness |
| CN111136027A (en) * | 2020-01-14 | 2020-05-12 | 广东技术师范大学 | Salted duck egg quality sorting device and method based on convolutional neural network |
| CN113575467A (en) * | 2021-07-29 | 2021-11-02 | 福州木鸡郎智能科技有限公司 | Block chain algorithm-based egg positioning method suitable for free-range chickens |
| CN113598090A (en) * | 2021-07-09 | 2021-11-05 | 临沂大学 | Detection device and detection method for egg appearance quality |
| CN114578002A (en) * | 2022-03-09 | 2022-06-03 | 武汉轻工大学 | Egg quality nondestructive testing method based on egg volume weight |
| CN116026829A (en) * | 2023-03-30 | 2023-04-28 | 华中农业大学 | Rapid nondestructive testing device and method for freshness of eggs |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1896736A (en) * | 2006-06-20 | 2007-01-17 | 长春吉大科学仪器设备有限公司 | Cereal grade analyzer |
| CN102147402A (en) * | 2011-03-08 | 2011-08-10 | 江苏大学 | Machine vision technology based method for rapidly detecting egg freshness |
| CN103307978A (en) * | 2013-05-23 | 2013-09-18 | 浙江大学 | Method for measuring volume and surface area of egg |
| KR20130107620A (en) * | 2012-03-22 | 2013-10-02 | 박창하 | Omega-3 polyunsaturated fatty acids composition method to enhance the breeding of poultry |
| KR101508257B1 (en) * | 2012-12-13 | 2015-04-14 | 성균관대학교산학협력단 | Apparatuses for measureing egg quality and methods thereof |
| CN104535588A (en) * | 2014-12-24 | 2015-04-22 | 河海大学常州校区 | Egg freshness nondestructive testing system based on Android system and egg freshness nondestructive testing method |
| CN105352555A (en) * | 2015-12-11 | 2016-02-24 | 江苏大学 | Portable detection device for quickly identifying storage time of fowl eggs, and application method thereof |
| CN107084941A (en) * | 2017-05-16 | 2017-08-22 | 中国农业大学 | The detection method and detection means of ω 3PUFA contents in a kind of egg |
| CN107102086A (en) * | 2017-06-24 | 2017-08-29 | 深圳源广安智能科技有限公司 | A kind of Beer Brewage material quality evaluation system |
-
2018
- 2018-03-30 CN CN201810296375.5A patent/CN108663367A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1896736A (en) * | 2006-06-20 | 2007-01-17 | 长春吉大科学仪器设备有限公司 | Cereal grade analyzer |
| CN102147402A (en) * | 2011-03-08 | 2011-08-10 | 江苏大学 | Machine vision technology based method for rapidly detecting egg freshness |
| KR20130107620A (en) * | 2012-03-22 | 2013-10-02 | 박창하 | Omega-3 polyunsaturated fatty acids composition method to enhance the breeding of poultry |
| KR101508257B1 (en) * | 2012-12-13 | 2015-04-14 | 성균관대학교산학협력단 | Apparatuses for measureing egg quality and methods thereof |
| CN103307978A (en) * | 2013-05-23 | 2013-09-18 | 浙江大学 | Method for measuring volume and surface area of egg |
| CN104535588A (en) * | 2014-12-24 | 2015-04-22 | 河海大学常州校区 | Egg freshness nondestructive testing system based on Android system and egg freshness nondestructive testing method |
| CN105352555A (en) * | 2015-12-11 | 2016-02-24 | 江苏大学 | Portable detection device for quickly identifying storage time of fowl eggs, and application method thereof |
| CN107084941A (en) * | 2017-05-16 | 2017-08-22 | 中国农业大学 | The detection method and detection means of ω 3PUFA contents in a kind of egg |
| CN107102086A (en) * | 2017-06-24 | 2017-08-29 | 深圳源广安智能科技有限公司 | A kind of Beer Brewage material quality evaluation system |
Non-Patent Citations (3)
| Title |
|---|
| ROMDHANE KAROUI 等: ""Methods to evaluate egg freshness in research and industry：A review"", 《EUR FOOD RES TECHNOL》 * |
| 刘雪 等: ""基于BP神经网络的鲜鸡蛋货架期预测模型"", 《农业机械学报》 * |
| 王巧华 等: ""基于机器视觉的鸡蛋品质无损检测研究"", 《中国家禽》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11262341B2 (en) | 2019-02-27 | 2022-03-01 | Northwest Agriculture And Forestry University | Method for non-destructive detection of egg freshness based on centroid measurement |
| WO2020173072A1 (en) * | 2019-02-27 | 2020-09-03 | 西北农林科技大学 | Method for egg freshness non-destructive detection based on centroid measurement |
| CN109813494A (en) * | 2019-02-27 | 2019-05-28 | 西北农林科技大学 | A method of based on measurement mass center non-destructive testing egg freshness |
| CN109937912A (en) * | 2019-04-08 | 2019-06-28 | 武汉科技大学 | A kind of egg categorizing system and method based on machine vision |
| CN109937912B (en) * | 2019-04-08 | 2021-06-15 | 武汉科技大学 | Egg classification system and method based on machine vision |
| CN110333160A (en) * | 2019-07-18 | 2019-10-15 | 华中农业大学 | Egg freshness non-destructive testing device and its method based on drainage |
| CN110411957A (en) * | 2019-08-28 | 2019-11-05 | 北京农业质量标准与检测技术研究中心 | The lossless method for quick predicting and device of fruit shelf life and freshness |
| CN110411957B (en) * | 2019-08-28 | 2021-11-19 | 北京农业质量标准与检测技术研究中心 | Nondestructive rapid prediction method and device for shelf life and freshness of fruits |
| CN111136027A (en) * | 2020-01-14 | 2020-05-12 | 广东技术师范大学 | Salted duck egg quality sorting device and method based on convolutional neural network |
| CN111136027B (en) * | 2020-01-14 | 2024-04-12 | 广东技术师范大学 | Salted duck egg quality sorting device and method based on convolutional neural network |
| CN113598090A (en) * | 2021-07-09 | 2021-11-05 | 临沂大学 | Detection device and detection method for egg appearance quality |
| CN113575467B (en) * | 2021-07-29 | 2022-09-09 | 福州木鸡郎智能科技有限公司 | Block chain algorithm-based egg positioning method suitable for free-range chickens |
| CN113575467A (en) * | 2021-07-29 | 2021-11-02 | 福州木鸡郎智能科技有限公司 | Block chain algorithm-based egg positioning method suitable for free-range chickens |
| CN114578002A (en) * | 2022-03-09 | 2022-06-03 | 武汉轻工大学 | Egg quality nondestructive testing method based on egg volume weight |
| CN116026829A (en) * | 2023-03-30 | 2023-04-28 | 华中农业大学 | Rapid nondestructive testing device and method for freshness of eggs |
| CN116026829B (en) * | 2023-03-30 | 2023-08-11 | 华中农业大学 | Rapid nondestructive testing device and method for freshness of eggs |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108663367A (en) | A kind of egg quality lossless detection method based on egg unit weight | |
| CN104764744B (en) | Visual inspection device and method for inspecting freshness of poultry eggs | |
| CN108663339B (en) | On-line detection method for mildewed corn based on spectrum and image information fusion | |
| CN103278464B (en) | Flesh of fish detection method and device | |
| Pan et al. | Eggshell crack detection based on computer vision and acoustic response by means of back-propagation artificial neural network | |
| CN101419212B (en) | Bird egg freshness detection device and method by utilizing volatile matter | |
| CN104535588B (en) | Egg freshness nondestructive testing system based on Android system and egg freshness nondestructive testing method | |
| CN102147402A (en) | Machine vision technology based method for rapidly detecting egg freshness | |
| Jia et al. | Hyperspectral Imaging (HSI) for meat quality evaluation across the supply chain: Current and future trends | |
| CN107564000B (en) | Hericium erinaceus non-destructive testing stage division based on computer vision | |
| CN105548028A (en) | Fowl egg freshness optical fiber spectroscopic grading detection device and method | |
| CN110702637B (en) | Near-infrared online fusion rapid discrimination method for hot fresh mutton and cold fresh mutton | |
| Zhihui et al. | Nondestructive detection of infertile hatching eggs based on spectral and imaging information | |
| CN104237235A (en) | Rapid detection method based on near-infrared imaging technology for food-borne pathogens | |
| Rafiq et al. | Application of computer vision system in food processing | |
| CN108489927A (en) | Fish origin tracing method, electronic equipment, storage medium and device | |
| Guo et al. | Bruise detection and classification of strawberries based on thermal images | |
| Xu et al. | A novel hyperspectral microscopic imaging system for evaluating fresh degree of pork | |
| CN112816487B (en) | Machine vision-based preserved egg internal quality nondestructive testing method | |
| Zhang et al. | Development of a hyperspectral imaging system for the early detection of apple rottenness caused by P enicillium | |
| Chen et al. | Double yolk nondestructive identification system based on Raspberry Pi and computer vision | |
| CN106018292A (en) | Non-destructive testing device for protein conformation in egg white and method of non-destructive testing device | |
| Cheng et al. | Transmission spectral analysis models for the assessment of white-shell eggs and brown-shell eggs freshness | |
| CN107643255A (en) | A kind of method of Non-Destructive Testing hatching egg middle and later periods addled egg | |
| CN110456010A (en) | A kind of method of lossless quick detection egg freshness |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181016 |