CN109089992A - A kind of classification method and system of the Estimation of The Fish Freshness based on machine vision - Google Patents
A kind of classification method and system of the Estimation of The Fish Freshness based on machine vision Download PDFInfo
- Publication number
- CN109089992A CN109089992A CN201810564825.4A CN201810564825A CN109089992A CN 109089992 A CN109089992 A CN 109089992A CN 201810564825 A CN201810564825 A CN 201810564825A CN 109089992 A CN109089992 A CN 109089992A
- Authority
- CN
- China
- Prior art keywords
- color
- fish
- predeterminable area
- image
- color feature
- 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
- 238000000034 method Methods 0.000 title claims abstract description 25
- 241000251468 Actinopterygii Species 0.000 claims abstract description 53
- 238000000926 separation method Methods 0.000 claims abstract description 16
- 239000000284 extract Substances 0.000 claims abstract description 14
- 238000007781 pre-processing Methods 0.000 claims abstract description 11
- 230000009466 transformation Effects 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 5
- 230000001960 triggered effect Effects 0.000 claims description 4
- 230000006378 damage Effects 0.000 abstract description 5
- 241001596950 Larimichthys crocea Species 0.000 description 42
- 210000003786 sclera Anatomy 0.000 description 13
- 238000001514 detection method Methods 0.000 description 11
- 238000006243 chemical reaction Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 210000005252 bulbus oculi Anatomy 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 239000003086 colorant Substances 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 238000004566 IR spectroscopy Methods 0.000 description 1
- 238000007792 addition Methods 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013145 classification model Methods 0.000 description 1
- 230000004456 color vision Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000012854 evaluation process Methods 0.000 description 1
- 210000003128 head Anatomy 0.000 description 1
- 238000013048 microbiological method Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
Classifications
-
- A—HUMAN NECESSITIES
- A01—AGRICULTURE; FORESTRY; ANIMAL HUSBANDRY; HUNTING; TRAPPING; FISHING
- A01K—ANIMAL HUSBANDRY; AVICULTURE; APICULTURE; PISCICULTURE; FISHING; REARING OR BREEDING ANIMALS, NOT OTHERWISE PROVIDED FOR; NEW BREEDS OF ANIMALS
- A01K61/00—Culture of aquatic animals
- A01K61/90—Sorting, grading, counting or marking live aquatic animals, e.g. sex determination
- A01K61/95—Sorting, grading, counting or marking live aquatic animals, e.g. sex determination specially adapted for fish
Landscapes
- Life Sciences & Earth Sciences (AREA)
- Zoology (AREA)
- Environmental Sciences (AREA)
- Marine Sciences & Fisheries (AREA)
- Animal Husbandry (AREA)
- Biodiversity & Conservation Biology (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides the classification methods and system of a kind of Estimation of The Fish Freshness based on machine vision, this method comprises: S10: the color feature value for receiving fish sample image classifies to fish sample image, establishes disaggregated model according to the color feature value；S20: acquiring the original image of fish, carries out image preprocessing to the original image, extracts the predeterminable area in the original image；S30: obtaining the color feature value of the predeterminable area, compares and matches with the color feature value of sample image, obtains the matched sample image of color feature value with the predeterminable area；S40: it according to the classification with the matched sample image of the color feature value of the predeterminable area, triggers corresponding separation controller and classifies to the fish.The color feature value of flake and fish body table section is acquired, the fish of different freshness can be quickly distinguished, the secondary damage of fish body when reducing manual sort improves classification effectiveness by establishing learning model using the present invention.
Description
Technical field
Planing machine vision technique of the present invention field more particularly to a kind of classification of the Estimation of The Fish Freshness based on machine vision
Method and system.
Background technique
For fish, freshness is always the important indicator that people judge its quality, at present to quality of fish
Ranking mainly based on subjective appreciation.Traditional subjective appreciation relies primarily on the personal experience of assessment personnel, subjective
Property it is strong, evaluation process can not quantitative analysis, and efficiency is lower；Microbiological method, chemical analysis, physical index evaluation etc. pair
The fish of detection have destruction, are easy to cause fish body in the detection process by secondary damage, reduce its quality；Near infrared light
The detection process such as spectrum, biosensor technology are cumbersome, time-consuming, the industrial requirements that inadaptable high-volume quickly detects.Due to fish
It is many kinds of, the detection of Estimation of The Fish Freshness must be in conjunction with specific product category, for example, when detecting to Larimichthys crocea, can
To judge its freshness situation according to flake color and fish body table color.
The patent of Publication No. CN104089906A discloses a kind of Larimichthys crocea freshness detection device and detection method, institute
Stating detection device includes laser array, spectrometer, luminous intensity regulating device, the light emitting for issuing incident light to sample
Head, bracket and optical receiver apparatus；The bracket includes bottom plate, the sample splint above bottom plate, on sample splint
Light-transmitting plate and the L shape fixed arm on bottom edge；The L shape fixed arm include level board above the sample splint and
Across sample splint and the vertical bar that is fixedly connected with sample splint, the position of level board being located above light-transmitting plate is equipped with to light transmission
The vertically extending arm that plate extends.The detection device and method are detected by infrared spectroscopy, and detection speed is slower, are unsatisfactory for extensive
Industrial detection demand.
Summary of the invention
The technical problem to be solved in the present invention is designed to provide a kind of classification of Estimation of The Fish Freshness based on machine vision
Method and system, low efficiency when solving the problems, such as different to freshness fish classification are easy to damage fish body.
To achieve the goals above, The technical solution adopted by the invention is as follows:
A kind of classification method of the Estimation of The Fish Freshness based on machine vision, comprising steps of
S10: the color feature value for receiving fish sample image establishes disaggregated model according to the color feature value；
S20: acquiring the original image of fish, carries out image preprocessing to the original image, extracts pre- in the original image
If region；
S30: obtaining the color feature value of the predeterminable area, compares and matches with the color feature value of sample image,
Obtain the matched sample image of color feature value with the predeterminable area；
S40: according to the classification with the matched sample image of the color feature value of the predeterminable area, corresponding classification is triggered
Controller classifies to the fish.
Further, described that original image progress image preprocessing is further comprised the steps of:
The original image is subjected to greyscale transformation, obtains grayscale image；
Threshold division is carried out to the grayscale image, obtains cut zone；
According to default circularity index, the border circular areas in the cut zone is screened out, target area is obtained；
The target area and the original image are compared, obtained corresponding with the target area in the original image
Region as predeterminable area.
Further, the step S30 is further comprised the steps of:
S31: receiving pre-set color model, according to the color data of the original image predeterminable area, obtains the default face
The color feature vector of color model
L=0.2126*R+0.7152*G+0.0722*B
A=1.4749* (0.2213*R03390*G+0.1177*B)+128
B=0.6245* (0.1949*R-0.6057*G+0.8006*B)+128
In above formula, R, G, B are respectively the color data in the original image predeterminable area；H, V, S, L, a, b are respectively institute
State the color feature vector of pre-set color model.
Further, the S30 is further comprised the steps of:
S301: according to the color feature vector, extracting the color feature value of the original image predeterminable area,
In above formula, mean is the mean value of predeterminable area color characteristic, and f (i, j) is predeterminable area i, the pixel value of the position j.
Further, predeterminable area includes fish body surface area and flake iris region.
A kind of categorizing system of the Estimation of The Fish Freshness based on machine vision, comprising:
Model building module: it for receiving the color feature value of fish sample image, according to the color feature value, establishes
Disaggregated model；
Image processing module: for acquiring the original image of fish, image preprocessing is carried out to the original image, described in extraction
Predeterminable area in original image；
Matching module: the color feature value for obtaining the color feature value of the predeterminable area, with the sample image
Matching is compared, the matched sample image of color feature value with the predeterminable area is obtained；
Categorization module: for according to the classification with the matched sample image of color feature value of the predeterminable area, triggering
Corresponding separation controller classifies to the fish.
Further, described image processing module further include:
Greyscale transformation unit: for the original image to be carried out greyscale transformation, grayscale image is obtained；
Thresholding unit: for carrying out threshold division to the grayscale image, cut zone is obtained；
Screen out unit: for the border circular areas in the cut zone being screened out, mesh is obtained according to circularity index being preset
Mark region；
Arithmetic element: for the target area and the original image to be compared, obtain in the original image with institute
The corresponding region in target area is stated as predeterminable area.
Further, the matching module further include:
Color description unit: it for receiving pre-set color model, according to the color data of the original image predeterminable area, obtains
Take the color feature vector of the pre-set color model
L=0.2126*R+0.7152*G+0.0722*B
A=1.4749* (0.2213*R-0.3390*G+0.1177*B)+128
B=0.6245* (0.1949*R-0.6057*G+0.8006*B)+128
In above formula, R, G, B are respectively the color data in the original image predeterminable area；H, V, S, L, a, b are respectively institute
State the color feature vector of pre-set color model.
Further, the matching module further include:
Characteristics extraction unit: for extracting the color of the original image predeterminable area according to the color feature vector
Characteristic value,
In above formula, mean is the mean value of predeterminable area color characteristic, and f (i, j) is predeterminable area i, the pixel value of the position j.
Further, the predeterminable area includes fish body surface area and flake iris region.
Using the present invention, color classification model is established, with machine vision, by the flake and body surface face that acquire fish
Color compares judgement, reduces the secondary damage in Estimation of The Fish Freshness detection process to fish body, improves to different freshness
Fish classification effectiveness, also ensure the accuracy of classification.
Detailed description of the invention
Fig. 1 is a kind of classification method stream for Estimation of The Fish Freshness based on machine vision that one embodiment of the present invention provides
Cheng Tu；
Fig. 2 is the RGB image for the Larimichthys crocea that one embodiment of the present invention provides；
Fig. 3 is the grayscale image after the Larimichthys crocea RGB image greyscale transformation that one embodiment of the present invention provides；
Fig. 4 is the flake area image that extracts in grayscale image that one embodiment of the present invention provides；
Fig. 5 is the flake iris image that screens in flake area image that one embodiment of the present invention provides；
Fig. 6 is the method that a kind of color feature value that one embodiment of the present invention provides is extracted；
Fig. 7 is a kind of categorizing system knot for Estimation of The Fish Freshness based on machine vision that one embodiment of the present invention provides
Composition.
Specific embodiment
Following is a specific embodiment of the present invention in conjunction with the accompanying drawings, technical scheme of the present invention will be further described,
However, the present invention is not limited to these examples.
In the present invention, fish include Larimichthys crocea, and the figure of different Larimichthys croceas has differences, but flake region and body surface
Colour-difference away from little, and each Larimichthys crocea is during storage, transport, and with the variation of time, freshness can be corresponding
Reduction, flake iris, the fish gill, fish body surface color corresponding change can also occur, can be sentenced according to the variation degree of color
The freshness of other Larimichthys crocea.
Embodiment one
With reference to Fig. 1, Fig. 1 is a kind of classification method flow chart of Estimation of The Fish Freshness based on machine vision, comprising steps of
S10: the color feature value for receiving fish sample image establishes disaggregated model according to the color feature value；
S20: acquiring the original image of fish, carries out image preprocessing to the original image, extracts pre- in the original image
If region；
S30: obtaining the color feature value of the predeterminable area, compares with the color feature value of the sample image
Match, obtains the matched sample image of color feature value with the predeterminable area；
S40: according to the classification with the matched sample image of the color feature value of the predeterminable area, corresponding classification is triggered
Controller classifies to the fish.
This method acquires the color characteristic of Larimichthys crocea flake sclera region and Larimichthys crocea body surface area, by the way that color is special
It levies and is compared with the color characteristic in sorting model, the different Larimichthys crocea individual of freshness can be identified, thus to freshness
Different Larimichthys croceas are classified.
In step S10, the color feature value of fish sample image is received, according to the color feature value, to fish sample
Image is classified, and disaggregated model is established.
When needing to establish disaggregated model, fish sample image is acquired by camera such as CCD camera, and by image
Reason software extracts color feature value from sample image, and color feature value is input in model, after carrying out learning training,
Obtain disaggregated model.
Disaggregated model uses BP neural network separation system, and BP neural network separation system can be divided into input layer, imply
Layer and output layer, giving anticipation error is 0.01, output layer number of nodes is set to 3, input layer inputs the face of great amount of samples image
The color feature value of sample image can be divided into three classes and exported by color characteristic value by model learning training.
In step S20, the original image of fish is acquired, image preprocessing is carried out to the original image, extracts the original image
In predeterminable area.
Disaggregated model foundation finishes, and starts the Larimichthys crocea different to freshness and classifies, passes through camera such as CCD camera
The original image for acquiring Larimichthys crocea carries out image preprocessing to original image, and original image is RGB image, including Larimichthys crocea flake region
And fish body table section, extract the flake sclera region in flake area image, predeterminable area include flake sclera region with
And fish body table section.
Referring to figs. 2 to Fig. 5, it illustrates carry out image to the original image of Larimichthys crocea in one embodiment of the present invention to locate in advance
Reason extracts the effect picture of flake sclera region.
Wherein, comprising steps of
The original image is subjected to greyscale transformation, obtains grayscale image；
Threshold division is carried out to the grayscale image, obtains cut zone；
According to default circularity index, the border circular areas in the cut zone is screened out, target area is obtained；
The target area and the original image are compared, obtained corresponding with the target area in the original image
Region as predeterminable area.
As shown in Fig. 2, the original image of Larimichthys crocea is RGB image, wherein R, G, B are three primary colors, and R, G, B is trichromatic to be taken
It is worth range between 0-255.Greyscale transformation is a kind of image processing mode, indicates original image with gray scale, can be obtained more
Clearly profile diagram, as shown in figure 3, Fig. 3 is the grayscale image that original image transformation obtains later；Threshold division is in image procossing
A kind of common image partition method, the image of different grey-scale range is occupied suitable for target and background, after greyscale transformation,
The gray level in flake region and the scale grade difference in other regions are obvious, can extract fish-eye area domain, flake region such as Fig. 4 institute
Show；Flake region includes flake sclera region and eyeball, then by preset circularity index, eyeball is sieved
It removes, obtains flake sclera region, as shown in Figure 5.In present embodiment, target area is the flake sclera region in Fig. 5.
Since Fig. 5 is gray level image at this time, need that flake sclera region in Fig. 5 and original image are carried out and operated, i.e., from
Interception obtains flake sclera region in original image, and flake sclera region is predeterminable area at this time.
In step S30, the color feature value of the predeterminable area is obtained, is carried out with the color feature value of the sample image
Comparison matching, obtains the matched sample image of color feature value with the predeterminable area.
In step slo, the color feature value of sample image and corresponding sample image are divided into three classes, step
It is extracted in S20 and obtains the color feature value of the sample image in the color feature value of predeterminable area and the disaggregated model of step S10
Matching is compared, confirms that the color feature value of predeterminable area is matched with which kind of in three classes, further obtains matched
Sample image.
In step S40, according to the classification with the matched sample image of the color feature value of the predeterminable area, triggering is corresponding
Separation controller classify to the fish.
Separation controller is also classified into three classes, corresponds to each other with three classes sample image, when predeterminable area color feature value with
When color feature value in three classes sample image matches, corresponding separation controller is triggered, Larimichthys crocea is divided into three kinds to reach
The different classes of purpose of freshness.
By acquiring the color feature value in Larimichthys crocea flake region, compared with the color feature value in disaggregated model
Match, to confirm the different Larimichthys crocea of freshness, classify to it, compare and manual sorting, improve classification accuracy,
Classification effectiveness.
Embodiment two
It is the method that a kind of color feature value provided in this embodiment is extracted with reference to Fig. 6, Fig. 6, comprising steps of
S31: receiving pre-set color model, according to the color data of the original image predeterminable area, obtains the default face
The color feature vector of color model
L=0.2126*R+0.7152*G+0.0722*B
A=1.4749* (0.2213*R03390*G+01177*B)+128
B=0.6245* (0.1949*R-0.6057*G+0.8006*B)+128
In above formula, R, G, B are respectively the color data in the original image predeterminable area；H, V, S, L, a, b are respectively institute
State the color feature vector of pre-set color model；
S301: according to the color feature vector, extracting the color feature value of the original image predeterminable area,
In above formula, mean is the mean value of predeterminable area color characteristic, and f (i, j) is predeterminable area i, the pixel value of the position j.
In the present embodiment, pre-set color model mainly has two kinds of color model of HSV, CIELAB, and HSV is compared to common
RGB color model, convenient to be split to designated color, three color parameters of HSV model are respectively H (color), S (pure
Degree), V (lightness), and three color parameters of CIELAB colour model are respectively L (brightness), a (green to red variation
Rate), the b change rate of yellow (blue arrive), hue angle and mankind's normal colour vision perception and acceptability are closest, are usually used in fish
The acquisition of class color data.
In step S31, the original image of Larimichthys crocea is RGB color model, R, G, three kinds of color datas of B between 0-255,
After original image is described with pre-set color model, color feature value can be more accurately and conveniently extracted.
In step S301, predeterminable area includes flake sclera region and fish body table section, the extraction of flake sclera region
It is had been described in upper one embodiment, and fish body table section then can be used conventional image procossing mode and extract, example
Fish body table section is such as obtained by background removal, gradation conversion, binary conversion treatment, contours extract.
Wherein, after carrying out color description to predeterminable area by pre-set color model, pass through calculation formulaEven if obtaining the color feature value of predeterminable area.
Larimichthys crocea original image is described by using pre-set color model, can guarantee the accuracy of color feature value, is improved
To the nicety of grading of the Larimichthys crocea of different freshness.
Embodiment three
It is the categorizing system structure for present embodiments providing a kind of Estimation of The Fish Freshness based on machine vision with reference to Fig. 7, Fig. 7
Figure, comprising:
Model building module 71: right according to the color feature value for receiving the color feature value of fish sample image
Fish sample image is classified, and disaggregated model is established；
Image processing module 72: for acquiring the original image of fish, image preprocessing is carried out to the original image, extracts institute
State the predeterminable area in original image；
Matching module 73: the color characteristic for obtaining the color feature value of the predeterminable area, with the sample image
Value compares matching, obtains the matched sample image of color feature value with the predeterminable area；
Categorization module 74: for according to the classification with the matched sample image of color feature value of the predeterminable area, touching
It sends out separation controller corresponding and classifies to the fish.
In the present embodiment, being sorted on industrial flow-line for the Larimichthys crocea of different freshness is completed, and all Larimichthys croceas are total
It is divided into three classes, one camera, such as CCD camera is set in the top of the front end of conveyer belt, acquire the original image of Larimichthys crocea, and pass
The model building module 71 in computer is transported to, computer according to the sample image of Larimichthys crocea and color feature value, is built first
A disaggregated model is found, which corresponds to the big of three classes difference freshness for the color feature value classification three classes of Larimichthys crocea
Yellow croaker.
It establishes and finishes when disaggregated model, can classify to Larimichthys crocea, Larimichthys crocea is acquired by the CCD camera of front end
Original image, and computer is sent an image to, the original image to Larimichthys crocea is carried out figure by the image processing module 72 in computer
As pretreatment, the predeterminable area in Larimichthys crocea original image is extracted, which includes flake sclera region and fish body surface area
Domain.
Matching module 73 is arranged in computer, and after image processing module 72 extracts predeterminable area, acquisition is default
The color feature value in region is matched with the color feature value in disaggregated model, determines the color feature value of current Larimichthys crocea
Belong to which kind of in three classes.
The separation controller of categorization module 74 and transmission end of tape establishes communication connection, and separation controller shares three classes, with
In correspondence with each other, categorization module 74 has determined the color feature value matched sample of Larimichthys crocea predeterminable area to the classification of three classes sample image
After one type in image three classes, corresponding separation controller will be controlled and sort out the Larimichthys crocea on conveyer belt, thus
Achieve the purpose that all Larimichthys croceas being divided into three classes.
For example, separation controller is divided into the first separation controller, the second separation controller, third separation controller, sample
The color feature value of image is divided into three classes, such as the first color characteristic class, the second color characteristic class, third color characteristic class, often
A kind of color feature value corresponds to the Larimichthys crocea of three classes difference freshness again.If the predeterminable area color feature value of current Larimichthys crocea with
The first color characteristic class matches in the color feature value of sample image, then categorization module 74 sends trigger signal to the first classification control
Device processed sorts out the Larimichthys crocea.
Wherein, image processing module 72 further include:
Gradation conversion unit 721: for the original image to be carried out gradation conversion, grayscale image is obtained；
Thresholding unit 722: for carrying out threshold division to the grayscale image, cut zone is obtained；
Screen out unit 723: for the border circular areas in the cut zone being screened out, is obtained according to circularity index being preset
To target area；
Arithmetic element 724: for the target area and the original image to be compared, obtain in the original image with
The corresponding region in the target area is as predeterminable area.
Matching module 73 further include:
Color description unit 730: for receiving pre-set color model, according to the number of colours of the original image predeterminable area
Value, obtains the color feature vector of the pre-set color model
Matching module 73 further include:
Characteristics extraction unit 731: for extracting the face of the original image predeterminable area according to the color feature vector
Color characteristic value.
Larimichthys crocea image is acquired by machine vision and identifies the color feature value in Larimichthys crocea image, which can
It quickly identifies the Larimichthys crocea of different freshness and classifies to it, be very suitable to large-scale industrial requirement, meanwhile, also can
Guarantee the accuracy of classification.
Specific embodiment described herein is only an example for the spirit of the invention.The neck of technology belonging to the present invention
The technical staff in domain can make various modifications or additions to the described embodiments or replace by a similar method
In generation, however, it does not deviate from the spirit of the invention or beyond the scope of the appended claims.
Claims (10)
1. a kind of classification method of the Estimation of The Fish Freshness based on machine vision, which is characterized in that comprising steps of
S10: the color feature value for receiving fish sample image establishes disaggregated model according to the color feature value；
S20: acquiring the original image of fish, carries out image preprocessing to the original image, extracts the preset areas in the original image
Domain；
S30: obtaining the color feature value of the predeterminable area, compares and matches with the color feature value of the sample image,
Obtain the matched sample image of color feature value with the predeterminable area；
S40: according to the classification with the matched sample image of the color feature value of the predeterminable area, corresponding classification control is triggered
Device classifies to the fish.
2. a kind of classification method of Estimation of The Fish Freshness based on machine vision according to claim 1, which is characterized in that institute
It states and original image progress image preprocessing is further comprised the steps of:
The original image is subjected to greyscale transformation, obtains grayscale image；
Threshold division is carried out to the grayscale image, obtains cut zone；
According to default circularity index, the border circular areas in the cut zone is screened out, target area is obtained；
The target area and the original image are compared, area corresponding with the target area in the original image is obtained
Domain is as predeterminable area.
3. a kind of classification method of Estimation of The Fish Freshness based on machine vision according to claim 1, which is characterized in that institute
Step S30 is stated to further comprise the steps of:
S31: pre-set color model is received according to the color data of the original image predeterminable area and obtains the pre-set color mould
The color feature vector of type
L=0.2126*R+0.7152*G+0.0722*B
A=1.4749* (0.2213*R-0.3390*G+0.1177*B)+128
B=0.6245* (0.1949*R-0.6057*G+0.8006*B)+128
In above formula, R, G, B are respectively the color data in the original image predeterminable area；H, V, S, L, a, b are respectively described pre-
If the color feature vector of color model.
4. a kind of classification method of Estimation of The Fish Freshness based on machine vision according to claim 1 or 3, feature exist
In the S30 is further comprised the steps of:
S301: according to the color feature vector, extracting the color feature value of the original image predeterminable area,
In above formula, mean is the mean value of predeterminable area color characteristic, and f (i, j) is predeterminable area i, the pixel value of the position j.
5. a kind of classification method of Estimation of The Fish Freshness based on machine vision according to claim 1, which is characterized in that institute
Stating predeterminable area includes fish body surface area and flake iris region.
6. a kind of categorizing system of the Estimation of The Fish Freshness based on machine vision characterized by comprising
Model building module: for receiving the color feature value of fish sample image, according to the color feature value, classification is established
Model；
Image processing module: for acquiring the original image of fish, image preprocessing is carried out to the original image, extracts the original image
Predeterminable area as in；
Matching module: it for obtaining the color feature value of the predeterminable area, is carried out with the color feature value of the sample image
Comparison matching, obtains the matched sample image of color feature value with the predeterminable area；
Categorization module: for according to the classification with the matched sample image of color feature value of the predeterminable area, triggering to be corresponding
Separation controller classify to the fish.
7. a kind of categorizing system of Estimation of The Fish Freshness based on machine vision according to claim 6, which is characterized in that institute
State image processing module further include:
Greyscale transformation unit: for the original image to be carried out greyscale transformation, grayscale image is obtained；
Thresholding unit: for carrying out threshold division to the grayscale image, cut zone is obtained；
Screen out unit: for the border circular areas in the cut zone being screened out, target area is obtained according to circularity index being preset
Domain；
Arithmetic element: for the target area and the original image to be compared, obtain in the original image with the mesh
The corresponding region in region is marked as predeterminable area.
8. a kind of categorizing system of Estimation of The Fish Freshness based on machine vision according to claim 6, which is characterized in that institute
State matching module further include:
Color description unit: for receiving pre-set color model, according to the color data of the original image predeterminable area, institute is obtained
State the color feature vector of pre-set color model
L=0.2126*R+0.7152*G+0.0722*B
A=1.4749* (0.2213*R-0.3390*G+0.1177*B)+128
B=0.6245* (0.1949*R-0.6057*G+0.8006*B)+128
In above formula, R, G, B are respectively the color data in the original image predeterminable area；H, V, S, L, a, b are respectively described pre-
If the color feature vector of color model.
9. a kind of categorizing system of the Estimation of The Fish Freshness based on machine vision, feature according to claim 6 or 8 exist
In the matching module further include:
Characteristics extraction unit: for extracting the color characteristic of the original image predeterminable area according to the color feature vector
Value,
In above formula, mean is the mean value of predeterminable area color characteristic, and f (i, j) is predeterminable area i, the pixel value of the position j.
10. a kind of categorizing system of Estimation of The Fish Freshness based on machine vision according to claim 6, which is characterized in that
The predeterminable area includes fish body surface area and flake iris region.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810564825.4A CN109089992A (en) | 2018-06-04 | 2018-06-04 | A kind of classification method and system of the Estimation of The Fish Freshness based on machine vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810564825.4A CN109089992A (en) | 2018-06-04 | 2018-06-04 | A kind of classification method and system of the Estimation of The Fish Freshness based on machine vision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109089992A true CN109089992A (en) | 2018-12-28 |
Family
ID=64796669
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810564825.4A Pending CN109089992A (en) | 2018-06-04 | 2018-06-04 | A kind of classification method and system of the Estimation of The Fish Freshness based on machine vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109089992A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109924147A (en) * | 2019-01-17 | 2019-06-25 | 广西壮族自治区水产引育种中心 | Information collection measurement system and measuring method in a kind of crucian hybrid seeding |
| CN111259926A (en) * | 2020-01-08 | 2020-06-09 | 珠海格力电器股份有限公司 | Meat freshness detection method and device, computing equipment and storage medium |
| CN111480606A (en) * | 2020-04-23 | 2020-08-04 | 舟山国家远洋渔业基地科技发展有限公司 | Marine product grading treatment system for ocean fishing ship |
| CN111768402A (en) * | 2020-07-08 | 2020-10-13 | 中国农业大学 | MU-SVM-based method for evaluating freshness of iced pomfret |
| CN112683899A (en) * | 2020-04-29 | 2021-04-20 | 海南远生渔业有限公司 | Aquatic product quality detection method |
| CN115561233A (en) * | 2022-09-20 | 2023-01-03 | 大连工业大学 | Method for visually and intelligently detecting freshness of meat based on hydrogel material |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103077396A (en) * | 2013-01-11 | 2013-05-01 | 上海电机学院 | Method and device for extracting vector space feature points of color image |
| CN105195442A (en) * | 2015-10-16 | 2015-12-30 | 中国水产科学研究院渔业机械仪器研究所 | Freshwater fish freshness grading system and method based on machine vision |
| CN106570855A (en) * | 2016-07-22 | 2017-04-19 | 北京农业信息技术研究中心 | Method and system for quickly judging pork freshness |
-
2018
- 2018-06-04 CN CN201810564825.4A patent/CN109089992A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103077396A (en) * | 2013-01-11 | 2013-05-01 | 上海电机学院 | Method and device for extracting vector space feature points of color image |
| CN105195442A (en) * | 2015-10-16 | 2015-12-30 | 中国水产科学研究院渔业机械仪器研究所 | Freshwater fish freshness grading system and method based on machine vision |
| CN106570855A (en) * | 2016-07-22 | 2017-04-19 | 北京农业信息技术研究中心 | Method and system for quickly judging pork freshness |
Non-Patent Citations (2)
| Title |
|---|
| 黄星奕等: "计算机视觉技术在鱼新鲜度检测中的应用研究", 《计算机工程与设计》 * |
| 黄福珍等: "《人脸检测》", 30 April 2006, 上海交通大学出版社 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109924147A (en) * | 2019-01-17 | 2019-06-25 | 广西壮族自治区水产引育种中心 | Information collection measurement system and measuring method in a kind of crucian hybrid seeding |
| CN111259926A (en) * | 2020-01-08 | 2020-06-09 | 珠海格力电器股份有限公司 | Meat freshness detection method and device, computing equipment and storage medium |
| CN111480606A (en) * | 2020-04-23 | 2020-08-04 | 舟山国家远洋渔业基地科技发展有限公司 | Marine product grading treatment system for ocean fishing ship |
| CN112683899A (en) * | 2020-04-29 | 2021-04-20 | 海南远生渔业有限公司 | Aquatic product quality detection method |
| CN111768402A (en) * | 2020-07-08 | 2020-10-13 | 中国农业大学 | MU-SVM-based method for evaluating freshness of iced pomfret |
| CN115561233A (en) * | 2022-09-20 | 2023-01-03 | 大连工业大学 | Method for visually and intelligently detecting freshness of meat based on hydrogel material |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109089992A (en) | A kind of classification method and system of the Estimation of The Fish Freshness based on machine vision | |
| WO2020221177A1 (en) | Method and device for recognizing image, storage medium and electronic device | |
| Liu et al. | Identification of rice seed varieties using neural network | |
| CN103761529B (en) | A kind of naked light detection method and system based on multicolour model and rectangular characteristic | |
| CN108181316B (en) | Bamboo strip defect detection method based on machine vision | |
| CN109269951A (en) | Floating tail-coal ash content, concentration, coarse granule detection method of content based on image | |
| CN109636824A (en) | A kind of multiple target method of counting based on image recognition technology | |
| CN106934386A (en) | A kind of natural scene character detecting method and system based on from heuristic strategies | |
| CN103076288A (en) | Automatic fish flesh grading device and method based on computer vision | |
| CN110070526A (en) | Defect inspection method based on the prediction of deep neural network temperature figure | |
| CN109063619A (en) | A kind of traffic lights detection method and system based on adaptive background suppression filter and combinations of directions histogram of gradients | |
| CN107505325B (en) | Omnibearing quality detection method for winter jujube fruits | |
| CN107330478A (en) | A kind of cherry tomato sorting technique, system and cherry tomato on-line sorting system | |
| CN103149214A (en) | Method for detecting flaw on surface of fruit | |
| CN110046565A (en) | A kind of method for detecting human face based on Adaboost algorithm | |
| Chao et al. | Color image classification systems for poultry viscera inspection | |
| CN109115775A (en) | A kind of betel nut level detection method based on machine vision | |
| Ignacio et al. | A YOLOv5-based deep learning model for in-situ detection and maturity grading of mango | |
| CN110110810B (en) | Squid quality grade identification and sorting method | |
| CN108460380A (en) | A kind of bamboo cane method for sorting colors and system based on domain color | |
| CN109937912B (en) | Egg classification system and method based on machine vision | |
| Meng et al. | Size characterisation of edible bird nest impurities: a preliminary study | |
| CN111860125A (en) | Method for identifying fruit maturity based on machine vision | |
| CN112816487B (en) | Machine vision-based preserved egg internal quality nondestructive testing method | |
| Ji et al. | Apple color automatic grading method based on machine vision |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181228 |
|
| RJ01 | Rejection of invention patent application after publication |