CN106971393A - The phenotype measuring method and system of a kind of corn kernel - Google Patents
The phenotype measuring method and system of a kind of corn kernel Download PDFInfo
- Publication number
- CN106971393A CN106971393A CN201710099982.8A CN201710099982A CN106971393A CN 106971393 A CN106971393 A CN 106971393A CN 201710099982 A CN201710099982 A CN 201710099982A CN 106971393 A CN106971393 A CN 106971393A
- Authority
- CN
- China
- Prior art keywords
- corn kernel
- information
- image
- inscribed circle
- division
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Geometry (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the present invention discloses the phenotype measuring method and system of a kind of corn kernel, and methods described includes：The image information of corn kernel is obtained, described image information is subjected to image procossing, the pre-segmentation processing of the corn kernel is carried out, obtains the colourity binary map of the corn kernel；According to the colourity binary map, the division contour images information and inscribed circle image information of the corn kernel are obtained；The division contour images information and the inscribed circle image information are subjected to convergence analysis, the phenotypic information of the corn kernel is obtained, wherein the phenotypic information includes 100-grain weight information, grain length information and the wide information of grain of the corn kernel.The system is used to perform the above method.The embodiment of the present invention realizes the automation of the phenotype measurement of corn kernel, improves the efficiency and accuracy of the phenotype measurement of corn kernel.
Description
Technical field
The present embodiments relate to technical field of image processing, and in particular to a kind of phenotype measuring method of corn kernel and
System.
Background technology
Corn is the grain and forage crop that China plants extensively.In corn variety test, cultivar identification, quality estimating
Deng in agri-scientific research, the phenotypic character of quick, accurate measurement seed is most important.The phenotype measurement key index bag of corn kernel
Include that grain length, grain be wide and 100-grain weight etc..Wherein, 100-grain weight index is counted with seed and overall weight is relevant, typically need to be to up to a hundred
Seed is weighed, and accurately statistics seed number obtains to calculate, the length and width measurement of corn kernel, then is needed to a large amount of typical cases
Statistical analysis is obtained after seed measurement.
In the prior art, the phenotype measuring method of corn kernel is to arrange seed along length or width, is surveyed respectively
Overall length or width are measured, seed average length and width is then calculated.In traditional corn seed species test, seed number, seed
Long, grain is wide, weight depends on manual measurement, with the problems such as efficiency is low, error is big, cost is high, standard type is poor.In the prior art
The phenotype measurement for carrying out corn kernel is primarily present problems with：Seed measuring apparatus to the adaptability of different size seeds not
Foot, detects that the quantity of seed and the computational accuracy of seed are difficult to adjust；Precisely detection is difficult to the seed of complex overlapping and recognized,
Seed count accuracy has much room for improvement；Length and width to corn kernel rely on simple oriented bounding box judgement, it is difficult to handle some
Seed has that width is more than length.
Therefore, how to propose a kind of scheme, it is possible to increase the phenotype seed measurement efficiency of corn kernel and the degree of accuracy, turn into
Urgent problem to be solved.
The content of the invention
For defect of the prior art, the embodiment of the present invention provides a kind of phenotype measuring method of corn kernel and is
System.
On the one hand, the embodiment of the present invention proposes a kind of phenotype measuring method of corn kernel, including：
The image information of corn kernel is obtained, described image information is subjected to image procossing, the corn kernel is carried out
Pre-segmentation is handled, and obtains the colourity binary map of the corn kernel；
According to the colourity binary map, the division contour images information and inscribe circular image letter of the corn kernel are obtained
Breath；
The division contour images information and the inscribed circle image information are subjected to convergence analysis, the Corn Seeds are obtained
The phenotypic information of grain, wherein the phenotypic information includes 100-grain weight information, grain length information and the wide information of grain of the corn kernel.
On the other hand, the embodiment of the present invention provides a kind of phenotype measuring system of corn kernel, including：The figure of interconnection
As harvester and image processing apparatus, described image harvester is used for the image information for obtaining corn kernel, described image
Processing unit is used to described image information carrying out image procossing, and wherein described image processing unit includes：
Image chroma processing module, for described image information to be carried out into image procossing, carries out the pre- of the corn kernel
Dividing processing, obtains the colourity binary map of the corn kernel；
Image outline processing module, splits and base for carrying out the seed based on progressive threshold value according to the colourity binary map
In the inscribe loop truss of initial profile, the division contour images information and inscribe circular image letter of the corn kernel are obtained respectively
Breath；
Image fusion processing module, for the division contour images information and the inscribed circle image information to be melted
Analysis is closed, the phenotypic information of the corn kernel is obtained, wherein the 100-grain weight that the phenotypic information includes the corn kernel is believed
Breath, grain length information and the wide information of grain.
The phenotype measuring method and system of corn kernel provided in an embodiment of the present invention, pass through the corn kernel to collecting
Image information carry out image procossing, obtain corn kernel division contour images information and inscribed circle image information, will point
Split contour images information and inscribed circle image information carries out convergence analysis, obtain the phenotypic information of corn kernel.Directly to jade
The image information of rice seed is handled and analyzed, and is realized the automation of the phenotype measurement of corn kernel, is improved Corn Seeds
The efficiency and accuracy of the phenotype measurement of grain.
Brief description of the drawings
Fig. 1 is the schematic flow sheet of the phenotype measuring method of corn kernel in the embodiment of the present invention；
Fig. 2 is the structural representation of the phenotype measuring system of corn kernel in the embodiment of the present invention；
Fig. 3 is the structural representation of the phenotype measuring system of another corn kernel in the embodiment of the present invention；
Fig. 4 is the schematic flow sheet of the phenotype measuring method of another corn kernel in the embodiment of the present invention.
Embodiment
To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below in conjunction with the embodiment of the present invention
In accompanying drawing, the technical scheme in the embodiment of the present invention is explicitly described, it is clear that described embodiment be the present invention
A part of embodiment, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not having
The every other embodiment obtained under the premise of creative work is made, the scope of protection of the invention is belonged to.
Fig. 1 is the schematic flow sheet of the phenotype measuring method of corn kernel in the embodiment of the present invention, as shown in figure 1, this hair
The phenotype measuring method for the corn kernel that bright embodiment is provided includes：
R1, the image information for obtaining corn kernel, image procossing is carried out by described image information, carries out the corn kernel
Pre-segmentation processing, obtain the colourity binary map of the corn kernel；
Specifically, using the image information of many corn kernels to be measured of image acquisition device, to what is collected
Image information is handled, and the embodiment of the present invention, which is mainly, is first split the image of corn kernel and background image, to jade
Rice seed makees preliminary pre-segmentation, and gets the colourity binary map of corn kernel.The image that will be collected carries out being based on color
The binary conversion treatment of degree, intercepts the image-region of corn kernel, is easy to the image procossing in later stage.
R2, according to the colourity binary map, obtain the division contour images information and inscribe circular image of the corn kernel
Information；
Specifically, get after the binary map of corn kernel, because some corn kernels can be sticked together, it is impossible to accurate
The grain length and grain for obtaining corn kernel are wide.Because the border in seed adhesion region exists substantially transitional special in color and gray scale
Levy (color is dark), the image information got is carried out image procossing by the embodiment of the present invention, utilizes progressive threshold segmentation method
Adhesion seed is split, the division profile of each corn kernel is obtained, that is, obtains the division profile information of corn kernel.Because
The division profile of obtained corn kernel have modified the real border of seed, and the later stage can deposit when calculating the geometry of corn kernel
In certain error, the embodiment of the present invention adds the inscribe loop truss of corn kernel initial profile, i.e., in single corn kernel
Initial profile in make inscribed circle, get the inscribed circle image information of corn kernel.I.e. single corn kernel can be with one
Row inscribed circle is represented, by being fitted the convex closure polygon of these inscribed circles, obtains the general shape of corn kernel.
R3, the division contour images information and the inscribed circle image information is subjected to convergence analysis, obtains the jade
The phenotypic information of rice seed, wherein 100-grain weight information, grain length information and grain that the phenotypic information includes the corn kernel are wide
Information.
Specifically, get after the division contour images information and inscribed circle image information of corn kernel, profile will be divided
Image information and inscribed circle image information carry out convergence analysis, that is, consider the division contour images information of corn kernel and interior
Circle of contact image information, the image information to corn kernel is analyzed, and obtains the phenotypic information of corn kernel.Wherein corn kernel
Phenotypic information include 100-grain weight information, grain length information and the wide information of grain of corn kernel, other information can also be included certainly
Such as garbled information, the embodiment of the present invention is not especially limited.
The phenotype measuring method of corn kernel provided in an embodiment of the present invention, passes through the image of the corn kernel to collecting
Information carries out image procossing, obtains the division contour images information and inscribed circle image information of corn kernel, will divide profile
Image information and inscribed circle image information carry out convergence analysis, obtain the phenotypic information of corn kernel.Directly to corn kernel
Image information handled and analyzed, realize corn kernel phenotype measurement automation, improve the table of corn kernel
The efficiency and accuracy of type measurement.
It is described that described image information is subjected to image procossing on the basis of above-described embodiment, carry out the corn kernel
Pre-segmentation processing, obtain the colourity binary map of the corn kernel, including：The effective coverage in described image information is obtained,
The image that luminance threshold scope and Chroma threshold scope are met in the effective coverage is extracted, the colourity binary map is obtained.
Specifically, the embodiment of the present invention is that corn kernel is put into after pallet to carry out IMAQ, is carrying out image procossing
When, the interference region beyond pallet area is removed first, the effective coverage of pallet is obtained, to the corresponding image in effective coverage
Information is handled, the corresponding image-region of crawl corn kernel.Obtaining the method for the corresponding effective coverage of pallet can pass through
Image is negated into color, maximum white connected region is obtained, noise is eliminated using morphology operations, the image letter collected is obtained
Maximum independent black region in breath, regard the convex closure of black region as the effective coverage of pallet.Get the effective of pallet
Behind region, by the image collected from RGB (red, Green, Blue) color space conversions into HSV (Hue, Saturation,
Value) color space, in the effective coverage of pallet, by setting in luminance threshold scope (V passages) filtering image information
Black region, by setting Chroma threshold scope (H passages) to obtain the colourity binary map of corn kernel.Obtain and meet luminance threshold
It is worth the image of scope and Chroma threshold scope, wherein luminance threshold scope and Chroma threshold scope can be carried out according to actual conditions
Set, luminance threshold scope is set to 46 by the embodiment of the present invention<V<255, Chroma threshold scope is set to 11<H<77, i.e. V<
46 pixel is considered black picture element, 11<H<<77 pixel is considered seed prime area.
The phenotype measuring method of corn kernel provided in an embodiment of the present invention, by setting luminance threshold scope and colourity threshold
It is worth scope, the pre-segmentation that corn kernel is carried out gets the image-region of corn kernel, is the image for doing corn kernel the later stage
Processing, the phenotypic information for obtaining corn kernel prepares.
It is described according to the colourity binary map on the basis of above-described embodiment, obtain the division wheel of the corn kernel
Wide image information, including：The colourity binary map is subjected to image procossing, the gray level image of the corn kernel is obtained；
Gradually increase brightness value, obtain the area that area threshold is met in the corresponding gray level image of each brightness value
Domain, the division profile of the corn kernel is obtained according to the region.
Specifically, the colourity binary map of the corn kernel got is subjected to image mask processing, gets corn kernel
Gray level image, because there is obvious transition feature (color in color and gray scale in the border in the adhesion region of corn kernel
It is dark), the embodiment of the present invention is split adhesion seed using progressive threshold segmentation method, obtains the division wheel of each seed
It is wide.Particular by the gray level image by corn kernel is only included, gradually increased brightness value, gray level image is divided
Cut, the independent communication region in segmentation result carries out areal analysis, judges whether the independent communication region meets area threshold
Value, if meeting, using the convex closure in the independent communication region as the division profile of single corn kernel, and is preserved.By preservation
Independent communication region is deleted from gray image, and increases brightness value, and gray image is split again, until brightness value reaches 255.
Area threshold includes maximum area threshold value and minimum area threshold value, and wherein maximum area threshold value refers to if split
The area of target area is more than the value, then it is assumed that the region is also needed to continue to split, and the value is traditionally arranged to be：More than single seed is most
Large area, the area less than two seeds and.Minimum area threshold value refers to if segmentation object area is less than the value, then it is assumed that should
Region can not be split.The size of maximum area threshold value and minimum area threshold value can be configured according to actual conditions, this
Inventive embodiments are not especially limited.
For example：The image that brightness value is less than 46 is rejected, obtained when carrying out corn kernel pre-segmentation by the embodiment of the present invention
Brightness value is more than or equal to the image information of 46 corn kernels, then carries out the seed based on progressive threshold value and split.It will obtain first
The colourity binary map that the brightness value arrived is more than or equal to 46 corn kernel carries out image mask processing, obtains and only includes corn kernel
Gray level image.Since brightness value is 46, gray level image is split, the region of independent communication in gray level image is obtained.
Judge whether the region is more than minimum area threshold value and is less than maximum area threshold value, if, then it is assumed that the region is single Corn Seeds
The corresponding image-region of grain, the division profile of the region convex closure as corn kernel is preserved.By the region from gray level image
Delete, get new gray level image, increase brightness value, specifically brightness value can be increased by 1, i.e., brightness value is set to 47,
The gray level image newly obtained is split, above-mentioned action is repeated, when acquisition brightness value is 47, area threshold is met in gray level image
The region of value as corn kernel division profile.Gradually increase brightness value, and obtain the corresponding Corn Seeds of each brightness value
The division profile of grain, until brightness value reaches 255.
The phenotype measuring method of corn kernel provided in an embodiment of the present invention, by gradually in increase brightness value, obtaining every
Meet the independent communication region of area threshold in the corresponding gray level image of one brightness value, and using the convex closure in the region as single
The division profile of corn kernel.The fractionation of adhesion corn kernel is realized, is follow-up progress corn kernel quantity and length and width
Calculating is prepared, and improves the accuracy of the phenotype measurement of corn kernel.
It is described according to the colourity binary map on the basis of above-described embodiment, obtain the inscribed circle of the corn kernel
Image information, including：
S1, by the colourity binary map carry out range conversion, obtain distance map；
S2, the maximum range value obtained in the distance map, if judging to know that the maximum distance value is more than radius threshold
Value, then the maximum range value is made into the inscribed circle of the corresponding profile of the corn kernel as radius, by the inscribed circle from
Deleted in the colourity binary map, obtain new colourity binary map；
S3, repeat the above steps S1, S2, until the maximum range value is less than the radius threshold, obtains the inscribe
Circular image information.
Specifically, the division profile of the corn kernel got due to the progressive segmentation figure picture based on brightness value, is continuous
By increasing brightness so that seed edge Stepwize Shrink is completely separated until the seed being connected, but have modified Corn Seeds
The real border of grain, if the geometry for directly calculating corn kernel using the division profile can have certain error.The present invention
Embodiment carries out the detection of inscribed circle by the initial profile to corn kernel, to realize the amendment to the error and make up.Will
The binary map of the corn kernel got carries out range conversion, obtains distance map, each pixel correspondence wherein in distance map
Distance value the distance between be the pixel to zero pixel closest with the pixel.Obtain maximum in the distance map
Distance value, judge the distance value whether be more than default radius threshold, if being more than, using the maximum range value as radius,
Make the inscribed circle of the corresponding profile of corn kernel.The inscribed circle is rejected from colourity binary map, new colourity two is got
Value figure, and new colourity binary map is subjected to range conversion, obtain new distance map.Judge the ultimate range in new distance map
Whether value is more than default radius threshold, if being more than, makees the initial profile of corn kernel using the maximum range value as radius
Inscribed circle, until maximum range value be less than default radius threshold.The filling of inscribed circle is carried out to corn kernel, will be single
Corn kernel is represented with a series of inscribed circles, gets the inscribed circle information of corn kernel.By being fitted the convex of these inscribed circles
Bag polygon, can obtain the general shape of seed；, can be to corn kernel by setting up the annexation of these inscribed circles
Positive and negative and long cross direction are judged.
Wherein default radius threshold illustrates the precision that corn kernel shape is described using inscribed circle, and the radius threshold is got over
Small, the circle for filling corn kernel is more, but computational efficiency is lower, therefore the setting of radius threshold can be according to actual conditions
Fixed, the embodiment of the present invention is not especially limited, and the radius threshold minimum of this in the embodiment of the present invention could be arranged to 3 pixels.
The phenotype measuring method of corn kernel provided in an embodiment of the present invention, by a series of inscribe of single corn kernel
Circle is indicated, and the shape details of corn kernel is described by different size of inscribed circle, by being fitted these inscribed circles
Convex closure polygon, can obtain the general shape of seed；, can be to corn kernel by setting up the annexation of these inscribed circles
Positive and negative and long cross direction judged.Improve the accuracy and efficiency of the phenotype measurement of corn kernel.
It is described to enter the division contour images information and the inscribed circle image information on the basis of above-described embodiment
Row convergence analysis, including：
According to the division contour images information and the inscribed circle image information, the division contour images information is obtained
In division profile and the inscribed circle in the inscribed circle image information position shape corresponding relation, position shape correspondence
Relation include it is intersecting and comprising；
The division profile and the inscribed circle are classified according to the position shape corresponding relation, by the division
Profile type corresponding with its, inscribed circle type corresponding with its are stored in division profile-inscribed circle look-up table.
Specifically, division profile is the result that (in order to split adhesion seed) is shunk to seed real border, and
Each seed shrinkage degree is simultaneously differed, therefore division profile actually illustrates the region after seed shrinks.However, progressive
The single seed in adhesion region may be split into two parts in Threshold segmentation, especially when the germinal layer region of adjacent seed is close
When together, close part overall brightness is higher so that this adhesion region is likely to be identified as independent seed.It is interior
The circle of contact is obtained based on seed initial segmentation result, unrelated with seed internal color texture, seed can be depicted original outer
Border and shape.Therefore, the inscribed circle image information of corn kernel is combined with division contour images information, it is possible to achieve
Validation checking and judgement to dividing profile.
The embodiment of the present invention determines the only of corn kernel using dividing profile using the feature of profile and inscribed circle is divided, i.e.,
Vertical property, seed real border and internal relations are defined using inscribed circle, so as to according to the position between inscribed circle and division profile
Put shape relation and set up division profile-inscribed circle look-up table.With specific reference to division contour images information and the inscribe circular image
Information, according to the shape and position feature of division profile, obtains the position shape corresponding relation of division profile and inscribed circle, position
Shape corresponding relation includes intersecting and comprising can also include other corresponding relations certainly, the embodiment of the present invention does not make specific limit
It is fixed.According to the division profile and the position shape corresponding relation of inscribed circle got, division profile and inscribed circle are classified,
And profile and its corresponding type will be divided, inscribed circle and its corresponding type are stored in division profile-inscribed circle look-up table.
The phenotype measuring method of corn kernel provided in an embodiment of the present invention, it is special according to the position of division profile and inscribed circle
Seek peace shape facility, set up division profile-inscribed circle look-up table, be that the phenotype measurement and calculating for carrying out corn kernel the later stage are done
Prepare, improve the accuracy of the phenotype measurement of corn kernel.
It is described that profile is divided and described by described according to the position shape corresponding relation on the basis of above-described embodiment
Inscribed circle is classified, including：
The inscribed circle is divided into zero matching inscribed circle, a matching inscribe by the quantity of the division profile matched according to inscribed circle
Circle and many matching inscribed circles；
The division profile is divided into zero and included comprising division profile, one by the quantity of the inscribed circle included according to division profile
Divide profile and many comprising division profile.
Specifically, the quantity for the division profile that each inscribed circle is matched is obtained according to division profile-inscribed circle look-up table, and
The inscribed circle is divided into zero matching inscribed circle, a matching inscribed circle and many by the quantity of the division profile matched according to inscribed circle
With inscribed circle.The quantity of the inscribed circle that each division profile is included in division profile-inscribed circle look-up table is obtained, and according to division
Division profile is divided into zero comprising division profile, one comprising division profile and more comprising division by the quantity for the inscribed circle that profile is included
Profile.
Wherein, zero matching inscribed circle refers to that the inscribed circle does not match any division profile, represents that the inscribed circle is corresponding
Region carry out seed fractionation after disappear, the division profile not matched, the region be generally seed borderline region (radius compared with
Small, brightness is relatively low), radius and brightness to all inscribed circles carry out clustering and obtain the mean radius of inscribed circle and be averaged bright
Degree, thus judges whether the inscribed circle of the type can be classified as an effective seed.One matching inscribed circle refers to have matched one
Divide the inscribed circle of profile, be the type of common inscribed circle.Many matching inscribed circles refer to that have matched more than 2 divides profile
Inscribed circle, represents that the inscribed circle is located on the intersecting area of adjacent division profile, correspond to the adhesion border between corn kernel, should
Inscribed circle radius is larger, brightness is higher.
Wherein, zero refers in the division profile not comprising any inscribed circle comprising division profile；One includes division profile
Refer to contain an inscribed circle in the division profile, represent that division profile and the inscribe toroidal goodness of fit are higher, corresponding jade
Rice seed is not contacted typically and the close circle of its shape with other seeds；It is many to refer to the division profile bag comprising division profile
More than 2 inscribed circles are contained, in this case, the embodiment of the present invention further judges that the inscribed circle that the division profile is included is
It is no to match inscribed circles for more, if it is, the shape that the inscribed circle will not include the corn kernel is calculated.
The phenotype measuring method of corn kernel provided in an embodiment of the present invention, it is special according to the position of division profile and inscribed circle
Seek peace shape facility, obtain the corresponding relation of division profile and inscribed circle, the quantity pair of the inscribed circle included according to division profile
Division profile is classified, and the quantity of the division profile matched according to inscribed circle is classified to inscribed circle, further according to each point
The measurement and calculating of the phenotype of the type progress corn kernel of profile and inscribed circle are split, the phenotype measurement of corn kernel is improved
Accuracy.
On the basis of above-described embodiment, the phenotypic information for obtaining the corn kernel, including：
According to the division profile-inscribed circle look-up table and formula N=C0+C1+C2+R0-R2Calculate the corn kernel
In quantity, formula：C0Represent described zero quantity for including division profile, C1Represent described one quantity for including division profile, C2Represent
The quantity for including division profile, R more0Represent the quantity of the zero matching inscribed circle, R2Represent the zero matching inscribed circle
Quantity；
The quantity of the corn kernel obtained according to calculating, obtains the 100-grain weight information of the corn kernel.
Specifically, inscribed circle and division profile be classified and stored in after division profile-inscribed circle look-up table, according to
Divide profile-inscribed circle look-up table and obtain zero quantity for including division profile, one includes the quantity of division profile, include division more
The quantity of the quantity of profile, the quantity of zero matching inscribed circle and zero matching inscribed circle, and according to formula N=C0+C1+C2+R0-R2Meter
Calculate the quantity of the corn kernel in the image information collected.In formula：C0Zero quantity for including division profile is represented, generally
0；C1The quantity for including division profile is represented, that is, includes the division outlines of 1 inscribed circle；C2Represent many comprising division wheel
Wide quantity, i.e. the division outlines comprising multiple inscribed circles；R0Zero quantity for matching inscribed circle is represented, that is, is not belonging to any
Divide profile and its radius and luminance mean value meet the inscribed circle quantity for clustering threshold value；R2Zero quantity for matching inscribed circle is represented,
The inscribed circle quantity intersected simultaneously with more than 2 division profiles.Corn Seeds are obtained according to the quantity of the corn kernel got
The 100-grain weight information of grain, specifically can be by obtaining the gross weights of all corn kernels in image information, by the corn got
The quantity of seed divided by the gross weight are used as the 100-grain weight information of corn kernel multiplied by with the value of 100 gained.Above-mentioned formula can be with
Find out, the quantity statistics of corn kernel utilizes the location and shape attribute of inscribed circle to carry out seed to divide based on profile calculates
Quantity amendment.
The phenotype measuring method of corn kernel provided in an embodiment of the present invention, according to different types of division outline-color quantity
And the quantity of different types of inscribed circle, the total quantity of corn kernel is calculated, the quantity calculating of corn kernel is improved
Accuracy, further increases the accuracy of the phenotype measurement of corn kernel.
On the basis of above-described embodiment, the phenotypic information for obtaining the corn kernel includes：
Inscribed circle according to the division profile and with the division profile existence position shape corresponding relation, constructs pixel
Point set；
The convex closure polygon of all pixels point in the pixel point set is obtained, the jade is obtained according to the convex closure polygon
The oriented bounding box of rice seed；
The grain length information and the wide information of grain of the corn kernel are obtained according to the length and width of the oriented bounding box.
Specifically, after the division contour images information and inscribed circle image information that obtain corn kernel, profile diagram will be divided
As information and inscribed circle image information progress convergence analysis, each position shape divided between profile and inscribed circle of acquisition is corresponding to close
System.Pixel point set is constructed with the corresponding division profile of corn kernel and the inscribed circle related to the division profile, pixel is calculated
The convex closure polygon of all pixels point in point set, obtains the corresponding direction of corn kernel according to the convex closure polygon of acquisition and surrounds
Box.Using the length of direction bounding box and wide length and width as corn kernel, in conjunction with the positive and negative according to corn kernel
Deng information acquisition corn kernel length direction of principal axis, the grain length information and the wide information of grain of corn kernel are obtained.Specifically can be in corn
In effective coverage in the image information of seed, corn kernel surface can be divided into according to color by two main regions, i.e. seed
Aleurone (yellow etc.) region and germinal layer (white) region.Wherein, seed front refers to the larger one side of germinal layer area,
And reverse side is then the less one side of germinal layer area.Therefore the pixel in the effective coverage in the image information of corn kernel is carried out
Clustering based on colourity, is divided into two big regions by corn kernel, then counts the quantity of seed crown areas pixel
Accounting, and then the positive and negative of seed captured by according to germinal layer area accounting determining image.Base is being carried out to seed interior pixels
In the aleurone region and germinal layer region that are obtained after the clustering of colourity, the centroid in two regions is then calculated respectively,
The line of centroid is the long axis direction for having defined seed, with reference to the oriented bounding box of the corn kernel of acquisition, to seed
Length and width be modified.
It is wide to the grain length and grain of corn kernel respectively after the grain length and grain of the single corn kernel of the effective seed of acquisition are wide
Be ranked up, be provided for count grain length grain width seed quantitative proportion, such as 30%, then can take centre from ranking results
30% corn kernel counts the average length and width of seed, obtains the grain length information and the wide information of grain of corn kernel.Wherein carry out
The method that statistics obtains the average length and width of corn kernel, can also be other method, the embodiment of the present invention is not especially limited.
The phenotype measuring method of corn kernel provided in an embodiment of the present invention, it is special according to the position of division profile and inscribed circle
Seek peace shape facility, fit the real border of corn kernel, obtain the oriented bounding box of corn kernel, further obtain corn
The wide information of grain length information and grain of seed.Improve the accuracy and efficiency of the phenotype measurement of corn kernel.
Fig. 2 is the structural representation of the phenotype measuring system of corn kernel in the embodiment of the present invention, as shown in Fig. 2 corn
The phenotype measuring system of seed includes：The image collecting device 20 and image processing apparatus 21 of interconnection, image collecting device
20 image information for obtaining corn kernel, image processing apparatus 21 is used to described image information carrying out image procossing, its
Middle image processing apparatus 21 includes：
Image chroma processing module 211, for described image information to be carried out into image procossing, carries out the corn kernel
Pre-segmentation is handled, and obtains the colourity binary map of the corn kernel；
Image outline processing module 212, splits for carrying out the seed based on progressive threshold value according to the colourity binary map
With the inscribe loop truss based on initial profile, the division contour images information and inscribe circular image of the corn kernel are obtained respectively
Information；
Image fusion processing module 213, for the division contour images information and the inscribed circle image information to be entered
Row convergence analysis, obtains the phenotypic information of the corn kernel, wherein the phenotypic information includes hundred of the corn kernel
Weight information, grain length information and the wide information of grain.
Specifically, the phenotype measuring system of corn kernel provided in an embodiment of the present invention, including image collecting device 20, is used
In the image information of collection corn kernel, its can be specifically a camera or video camera or including camera or
Video camera etc. is used for other devices for gathering image information, and the embodiment of the present invention is not especially limited.Image collecting device 20 with
Image processing apparatus 21 is connected, and the image letter collected is delivered into image processing apparatus 21,21 pairs of receptions of image processing apparatus
The image information arrived carries out image procossing, and obtains the wide information isophenous of 100-grain weight information, grain length information and grain of corn kernel
Information.Wherein image processing apparatus 21 includes image chroma processing module 211, image outline processing module 212 and image co-registration
The image collected is carried out the binary conversion treatment based on colourity by processing module 213, image chroma processing module 211, and interception is beautiful
The image-region of rice seed, makees preliminary pre-segmentation to corn kernel, and get the colourity binary map of corn kernel.Image wheel
Wide processing module 212 is split adhesion seed using progressive threshold segmentation method, obtains the division profile of each corn kernel,
The division profile information of corn kernel is obtained, and makees inscribed circle in the initial profile of single corn kernel, corn is got
The inscribed circle image information of seed.I.e. image outline processing module 212 represents single corn kernel with a series of inscribed circles,
By being fitted the convex closure polygon of these inscribed circles, the general shape of corn kernel is obtained.Image fusion processing module 213 will divide
Split contour images information and inscribed circle image information carries out convergence analysis, that is, consider the division contour images letter of corn kernel
Breath and inscribed circle image information, the image information to corn kernel are analyzed, and obtain the phenotypic information of corn kernel such as：Hundred
Weight information, grain length information and the wide information of grain etc..
It should be noted that the modules in image processing apparatus are used to perform the above method, embodiment is same
Above-described embodiment is consistent, and here is omitted.
The phenotype measuring system of corn kernel provided in an embodiment of the present invention, passes through the image of the corn kernel to collecting
Information carries out image procossing, obtains the division contour images information and inscribed circle image information of corn kernel, will divide profile
Image information and inscribed circle image information carry out convergence analysis, obtain the phenotypic information of corn kernel.Directly to corn kernel
Image information handled and analyzed, realize corn kernel phenotype measurement automation, improve the table of corn kernel
The efficiency and accuracy of type measurement.
Fig. 3 is the structural representation of the phenotype measuring system of another corn kernel in the embodiment of the present invention, as shown in figure 3,
On the basis of above-described embodiment, image collecting device 10 includes：Band is provided with species test folding stand 101, species test folding stand 101
There is the electronic balance 102 of pallet；
One end of species test folding stand 101 is fixedly connected with lift in height frame 103, lift in height frame 103 and is provided with level
Displacement frame 104, for moving up and down and moving in the horizontal direction along lift in height frame 103；
Light source 105 and image capture module 106 are provided with horizontal displacement frame 104, for gathering the corn kernel
Image information；
Image capture module 106 and electronic balance 102 are connected with image processing apparatus 11 respectively, are respectively used to collect
Described image information and the weight information of corn kernel send to image processing apparatus 11.
Specifically, as shown in figure 3, the image collector of the phenotype measuring system of corn kernel provided in an embodiment of the present invention
Putting includes the species test folding stand 101 as bottom plate, and the electronic balance of pallet is provided with the upper surface of species test folding stand 101
102, electronic balance 102 can fix or be placed on bottom plate i.e. species test folding stand 101 that there is provided a variety of range specifications；Electronics day
Seed pallet 1021 on flat 102 can be black matte material, it is possible to be multiple dimensions, to accommodate varying number seed
Grain (100,200,300 etc.), is having white standard correcting block 1022, for certainly close to the corner location of seed pallet 1021
Dynamic correcting image distortion and image resolution ratio.Disposal, which is fixed, as one end of species test folding stand 101 of bottom plate is connected with lift in height
Frame 103, horizontal displacement frame 104 is fixed on lift in height frame 103 with sliding component, lift in height frame 103 and horizontal displacement frame
104 could be arranged to folding structure, be convenient for carrying.Image capture module 106 such as can be that camera and light source 105 are used respectively
Sliding component is fixed on horizontal displacement frame 104, can be moved along horizontal level frame 104, wherein the light of image capture module 106
Axle vertically downward, the lamp holder of light source 105 be pointed into pallet, bright and dark light is adjustable with direction, adapt to varying environment illumination condition.Image is adopted
Collection module 106 and electronic balance 102 connect data processing apparatus 11, and such as computer or single-chip microcomputer can be used for processing data
Device, the image information and weight data synchronous acquisition of the control seed of data processing equipment 11 and call corn kernel phenotype
Computing module performs the phenotypic analysis of corn kernel.
The phenotype measuring system of corn kernel provided in an embodiment of the present invention, image collecting device therein, IMAQ
Module and light source liftable, the simple mechanism of adjustable angle, imaging region size is controlled by adjusting shooting grease head highness, is realized
High-volume corn kernel is quickly measured, and can also realize the high resolution detection of small lot seed, improves the phenotype of corn kernel
The efficiency of measurement.
Fig. 4 is the schematic flow sheet of the phenotype measuring method of another corn kernel in the embodiment of the present invention, as shown in figure 4,
The embodiment of the present invention carries out the phenotype measurement of corn kernel using the phenotype measuring system of the corn kernel shown in above-mentioned Fig. 3
Method includes：
T1, collection seed image information.According to seed quantity to be measured from the seed pallet for being adapted to size, electronics day is placed in
On flat；Connection image capture module and weight sensor are the data wire of electronic balance, open light source, camera, weight sensor,
Regulation lift in height frame and horizontal displacement frame enable the visual field of image capture module to cover pallet area as far as possible, adjust light source
Brightness and angle cause the seed coat color shot close to eye-observation, that is, to complete system startup work.Trigger software image collection
Operation, shoots current seed image, obtains the image information of corn kernel.
T2, the real-time weight data of acquisition.While the head image information of corn kernel is obtained, the real-time of electronic balance is read
Weight data (has removed pallet weight), and the 100-grain weight information for calculating corn kernel is prepared.
T3, image rectification simultaneously calculate image resolution ratio.Detect the circular correction of white on pallet in the image information collected
Block, carries out pattern distortion analysis, and calculate pixel resolution.
Wherein, the circular correcting block of white is detected, is to utilize Hough transform (to be a kind of parameter Estimation using voting principle
Technology) it is circular in detection image, be then by what the radius threshold values of correcting block filtered out that invalid circle obtains representing correcting block
Row circle, calculates round radius respectively, carries out distortion analysis.Wherein correcting block radius threshold values is image capture module in highest
Position and extreme lower position photographs correcting block image calculate obtained minimum and maximum radius of circle.Pattern distortion is analyzed, and is basis
The serial radius of circle that is detected from image judges whether pallet is correctly put.If detecting two or more correction in image
Block, then compare whether the error between the radius of circle of these correcting blocks can such as set value error as 3% within the specific limits, if
More than the error, then judge that seed pallet putting position is unreasonable, i.e., seed pallet may not be laid flat, tray center and camera light
Axle unique deviation is larger, provides early warning and prompting；If without correcting block, acquiescence corrected result using last time, unless by hand
Correction.
Because calculating the 100-grain weight of corn kernel, grain length and during the information such as grain is wide, the unit used is pixel, is finally needed
It is converted into unit of weight or long measure, it is therefore desirable to calculate the pixel resolution of corn kernel image.Specific method can be with
It is by the radius of known standard correction block divided by the radius of circle detected, that is, to obtain the pixel resolution (cm/ pixels) of image.
The value is highly directly related with camera imaging, once camera crane is have adjusted, it is necessary to re-calibrate pixel resolution.
T4, the seed segmentation controlled based on colourity.Interference region beyond pallet area is removed, the effective of pallet is obtained
Region, is handled to the corresponding image information in effective coverage, the corresponding image-region of crawl corn kernel, by setting color
Threshold range is spent, the colourity binary map of corn kernel is obtained.After the colourity binary map for obtaining corn kernel, step T5 is performed respectively
And T6.
T5, the seed based on progressive threshold value are split.The colourity binary map of the corn kernel got is subjected to image mask
Processing, gets the gray level image of corn kernel, by the gray level image by corn kernel is only included, is gradually increased brightness
Value, gray level image is split, the independent communication that area threshold is met in the corresponding gray level image of each brightness value is obtained
Region, and using the convex closure in the region as single corn kernel division profile.
T6, the inscribe loop truss based on initial profile.The binary map of the corn kernel got is subjected to range conversion, obtained
To distance map, make the inscribed circle of the initial profile of corn kernel according to distance map.The filling of inscribed circle is carried out to corn kernel,
Single corn kernel is represented with a series of inscribed circles, the inscribed circle information of corn kernel is got.
T7, division profile and inscribed circle convergence analysis.According to the division contour images information and the inscribe circular image
Information, obtains the position shape corresponding relation of inscribed circle and division profile, sets up division profile-inscribed circle look-up table.Further according to
Divide the phenotypic information that profile-inscribed circle look-up table obtains corn kernel, perform step T8 and T10.
T8, seed quantity statistical analysis.According to the division profile in division profile-inscribed circle look-up table in Corn Seeds with it is interior
The position shape corresponding relation of the circle of contact, classifies to division profile with inscribed circle, and according to all types of division profiles and inscribe
Round quantity, calculates the total quantity of the corn kernel in image information.
T9, calculating 100-grain weight.With reference to the gross weight of the step T2 corn kernels obtained, according to the total of the corn kernel of acquisition
Quantity, calculates the 100-grain weight of corn kernel.
T10, corn kernel length and width statistics.The division contour images information and inscribed circle image information of corn kernel are entered
Row convergence analysis, obtains the corresponding oriented bounding box of corn kernel, in conjunction with information acquisitions such as the positive and negatives according to corn kernel
Corn kernel length direction of principal axis, obtains the grain length information and the wide information of grain of corn kernel.By the single corn kernel of effective seed
Grain length and grain is wide is ranked up statistics, obtain the average length and width of corn kernel.
The phenotype measuring method and system of corn kernel provided in an embodiment of the present invention, structure can be simple, convenient operation and
Carry, and the method based on division profile and inscribed circle, realize that corn kernel is precisely counted and shape measure, break through complicated seed
The technological difficulties such as grain adhesion, grain length grain width judgement, accurately calculate the important seed phenotypic parameter such as the wide, 100-grain weight of grain length, grain, are
The species test of corn kernel batch provides low cost, high efficiency, flexible species test equipment.
Finally it should be noted that：The above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although
The present invention is described in detail with reference to the foregoing embodiments, it will be understood by those within the art that：It still may be used
To be modified to the technical scheme described in foregoing embodiments, or equivalent substitution is carried out to which part technical characteristic；
And these modification or replace, do not make appropriate technical solution essence depart from various embodiments of the present invention technical scheme spirit and
Scope.
Claims (10)
1. the phenotype measuring method of a kind of corn kernel, it is characterised in that methods described includes：
The image information of corn kernel is obtained, described image information is subjected to image procossing, pre- point of the corn kernel is carried out
Processing is cut, the colourity binary map of the corn kernel is obtained；
According to the colourity binary map, the division contour images information and inscribed circle image information of the corn kernel are obtained；
The division contour images information and the inscribed circle image information are subjected to convergence analysis, the corn kernel is obtained
Phenotypic information, wherein the phenotypic information includes 100-grain weight information, grain length information and the wide information of grain of the corn kernel.
2. according to the method described in claim 1, it is characterised in that described that described image information is subjected to image procossing, carry out
The pre-segmentation processing of the corn kernel, obtains the colourity binary map of the corn kernel, including：Obtain in described image information
Effective coverage, extract the image that luminance threshold scope and Chroma threshold scope are met in the effective coverage, obtain the color
Spend binary map.
3. according to the method described in claim 1, it is characterised in that described according to the colourity binary map, obtain the corn
The division contour images information of seed, including：
The colourity binary map is subjected to image procossing, the gray level image of the corn kernel is obtained；
Gradually increase brightness value, obtain the region that area threshold is met in the corresponding gray level image of each brightness value, root
The division profile of the corn kernel is obtained according to the region.
4. according to the method described in claim 1, it is characterised in that described according to the colourity binary map, obtain the corn
The inscribed circle image information of seed, including：
S1, by the colourity binary map carry out range conversion, obtain distance map；
S2, the maximum range value obtained in the distance map, if judging to know that the maximum distance value is more than radius threshold,
Using the maximum range value as radius, make the inscribed circle of the corresponding profile of the corn kernel, by the inscribed circle from described
Deleted in colourity binary map, obtain new colourity binary map；
S3, repeat the above steps S1, S2, until the maximum range value is less than the radius threshold, obtains the inscribe circle diagram
As information.
5. according to the method described in claim 1, it is characterised in that described by the division contour images information and the inscribe
Circular image information carries out convergence analysis, including：
According to the division contour images information and the inscribed circle image information, obtain in the division contour images information
Divide the position shape corresponding relation of profile and the inscribed circle in the inscribed circle image information, the position shape corresponding relation
Including it is intersecting and comprising；
The division profile and the inscribed circle are classified according to the position shape corresponding relation, by the division profile
Type corresponding with its, the inscribed circle type corresponding with its are stored in division profile-inscribed circle look-up table.
6. method according to claim 5, it is characterised in that described described to be divided according to the position shape corresponding relation
Split profile and the inscribed circle is classified, including：
The quantity of the division profile matched according to inscribed circle by the inscribed circle be divided into zero matching inscribed circle, one matching inscribed circle and
Many matching inscribed circles；
The division profile is divided into zero comprising division profile, one and includes division by the quantity of inscribed circle included according to division profile
Profile and include divide profile more.
7. method according to claim 6, it is characterised in that the phenotypic information of the acquisition corn kernel, including：
According to the division profile-inscribed circle look-up table and formula N=C0+C1+C2+R0-R2The quantity of the corn kernel is calculated,
In formula：C0Represent described zero quantity for including division profile, C1Represent described one quantity for including division profile, C2Represent described
The quantity for including division profile, R more0Represent the zero matching inscribed circle and radius and brightness are satisfied by clustering the inscribe of threshold value
Round quantity, R2Represent the quantity of the zero matching inscribed circle；
The quantity of the corn kernel obtained according to calculating, obtains the 100-grain weight information of the corn kernel.
8. method according to claim 5, it is characterised in that the phenotypic information of the acquisition corn kernel includes：
Inscribed circle according to the division profile and with the division profile existence position shape corresponding relation, constructs pixel
Collection；
The convex closure polygon of all pixels point in the pixel point set is obtained, the Corn Seeds are obtained according to the convex closure polygon
The oriented bounding box of grain；
The grain length information and the wide information of grain of the corn kernel are obtained according to the length and width of the oriented bounding box.
9. the phenotype measuring system of a kind of corn kernel, it is characterised in that the system includes：The image collector of interconnection
Put and image processing apparatus, described image harvester is used for the image information for obtaining corn kernel, described image processing unit
For described image information to be carried out into image procossing, wherein described image processing unit includes：
Image chroma processing module, for described image information to be carried out into image procossing, carries out the pre-segmentation of the corn kernel
Processing, obtains the colourity binary map of the corn kernel；
Image outline processing module, splits and based on first for carrying out the seed based on progressive threshold value according to the colourity binary map
The inscribe loop truss of beginning profile, obtains the division contour images information and inscribed circle image information of the corn kernel respectively；
Image fusion processing module, for the division contour images information and the inscribed circle image information to be carried out into fusion point
Analysis, obtains the phenotypic information of the corn kernel, wherein the phenotypic information includes the 100-grain weight information of the corn kernel, grain
Long message and the wide information of grain.
10. system according to claim 9, it is characterised in that described image harvester includes：Species test folding stand, institute
State and the electronic balance with pallet is provided with species test folding stand；
One end fixed vertical of the species test folding stand is connected with lift in height frame, the lift in height frame and is provided with horizontal position
Frame is moved, for moving up and down and moving in the horizontal direction along the lift in height frame；
Light source and image capture module are provided with the horizontal displacement frame, the image information for gathering the corn kernel；
Described image acquisition module and the electronic balance are connected with described image processing unit respectively, are respectively used to collect
Described image information and the weight information of corn kernel send to described image processing unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710099982.8A CN106971393B (en) | 2017-02-23 | 2017-02-23 | A kind of the phenotype measurement method and system of corn kernel |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710099982.8A CN106971393B (en) | 2017-02-23 | 2017-02-23 | A kind of the phenotype measurement method and system of corn kernel |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106971393A true CN106971393A (en) | 2017-07-21 |
| CN106971393B CN106971393B (en) | 2019-09-10 |
Family
ID=59328749
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710099982.8A Active CN106971393B (en) | 2017-02-23 | 2017-02-23 | A kind of the phenotype measurement method and system of corn kernel |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106971393B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108872028A (en) * | 2018-08-27 | 2018-11-23 | 武汉古奥基因科技有限公司 | A kind of soybean beans Phenotypic examination device and method |
| CN110490868A (en) * | 2019-08-23 | 2019-11-22 | 陈�峰 | A kind of lossless method of counting based on computer vision corn grain number per spike |
| CN111435428A (en) * | 2019-01-14 | 2020-07-21 | 珠海格力电器股份有限公司 | Rice identification method and device based on chromaticity detection |
| CN111742641A (en) * | 2020-07-28 | 2020-10-09 | 上海市农业科学院 | High-flux seed testing device and method |
| CN112819745A (en) * | 2019-10-31 | 2021-05-18 | 合肥美亚光电技术股份有限公司 | Nut kernel center worm-eating defect detection method and device |
| CN115205319A (en) * | 2022-09-19 | 2022-10-18 | 山东省金春雨种业科技有限公司 | Seed feature extraction and classification method used in seed selection process |
| CN115375694A (en) * | 2022-10-27 | 2022-11-22 | 浙江托普云农科技股份有限公司 | Portable rice whole ear measuring method based on image recognition and application thereof |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040241311A1 (en) * | 2003-05-16 | 2004-12-02 | Baianu Ion C. | Methods and apparatus for treating plant products using electromagnetic fields |
| CN101281112A (en) * | 2008-04-30 | 2008-10-08 | 浙江理工大学 | Image type automatic analysis method for mesh adhesion rice corn |
| CN102013102A (en) * | 2010-12-01 | 2011-04-13 | 华中科技大学 | Method for segmenting adhesion cells in image |
| CN103020970A (en) * | 2012-12-25 | 2013-04-03 | 北京农业信息技术研究中心 | Corn ear image grain segmentation method |
| CN106169186A (en) * | 2016-07-01 | 2016-11-30 | 西安电子科技大学 | Based on the method obtaining initial profile in level set moving object detection |
| CN106296644A (en) * | 2015-06-10 | 2017-01-04 | 浙江托普云农科技股份有限公司 | Method is analyzed in a kind of corn kernel species test based on image procossing |
-
2017
- 2017-02-23 CN CN201710099982.8A patent/CN106971393B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040241311A1 (en) * | 2003-05-16 | 2004-12-02 | Baianu Ion C. | Methods and apparatus for treating plant products using electromagnetic fields |
| CN101281112A (en) * | 2008-04-30 | 2008-10-08 | 浙江理工大学 | Image type automatic analysis method for mesh adhesion rice corn |
| CN102013102A (en) * | 2010-12-01 | 2011-04-13 | 华中科技大学 | Method for segmenting adhesion cells in image |
| CN103020970A (en) * | 2012-12-25 | 2013-04-03 | 北京农业信息技术研究中心 | Corn ear image grain segmentation method |
| CN106296644A (en) * | 2015-06-10 | 2017-01-04 | 浙江托普云农科技股份有限公司 | Method is analyzed in a kind of corn kernel species test based on image procossing |
| CN106169186A (en) * | 2016-07-01 | 2016-11-30 | 西安电子科技大学 | Based on the method obtaining initial profile in level set moving object detection |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108872028A (en) * | 2018-08-27 | 2018-11-23 | 武汉古奥基因科技有限公司 | A kind of soybean beans Phenotypic examination device and method |
| CN111435428A (en) * | 2019-01-14 | 2020-07-21 | 珠海格力电器股份有限公司 | Rice identification method and device based on chromaticity detection |
| CN111435428B (en) * | 2019-01-14 | 2023-10-31 | 珠海格力电器股份有限公司 | Rice identification method and device based on chromaticity detection |
| CN110490868A (en) * | 2019-08-23 | 2019-11-22 | 陈�峰 | A kind of lossless method of counting based on computer vision corn grain number per spike |
| CN110490868B (en) * | 2019-08-23 | 2023-05-30 | 陈�峰 | Nondestructive counting method based on computer vision corn cob grain number |
| CN112819745A (en) * | 2019-10-31 | 2021-05-18 | 合肥美亚光电技术股份有限公司 | Nut kernel center worm-eating defect detection method and device |
| CN112819745B (en) * | 2019-10-31 | 2023-02-28 | 合肥美亚光电技术股份有限公司 | Nut kernel center worm-eating defect detection method and device |
| CN111742641A (en) * | 2020-07-28 | 2020-10-09 | 上海市农业科学院 | High-flux seed testing device and method |
| CN111742641B (en) * | 2020-07-28 | 2021-11-09 | 上海市农业科学院 | High-flux seed testing device and method |
| CN115205319A (en) * | 2022-09-19 | 2022-10-18 | 山东省金春雨种业科技有限公司 | Seed feature extraction and classification method used in seed selection process |
| CN115205319B (en) * | 2022-09-19 | 2023-01-10 | 山东省金春雨种业科技有限公司 | Seed feature extraction and classification method used in seed selection process |
| CN115375694A (en) * | 2022-10-27 | 2022-11-22 | 浙江托普云农科技股份有限公司 | Portable rice whole ear measuring method based on image recognition and application thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106971393B (en) | 2019-09-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106971393A (en) | The phenotype measuring method and system of a kind of corn kernel | |
| US8606002B2 (en) | Apparatus, method and article for evaluating a stack of objects in an image | |
| CN100559826C (en) | Image processing equipment and method thereof | |
| CN110728225B (en) | High-speed face searching method for attendance checking | |
| CN109191520B (en) | Plant leaf area measuring method and system based on color calibration | |
| CN106384117B (en) | A kind of vehicle color identification method and device | |
| CN108765433A (en) | One kind is for carrying high-precision leafy area measurement method | |
| CN111915704A (en) | Apple hierarchical identification method based on deep learning | |
| CN102974551A (en) | Machine vision-based method for detecting and sorting polycrystalline silicon solar energy | |
| AU2020100915A4 (en) | A Method For Quickly Counting Rice And Wheat Grains | |
| CN107895362A (en) | A kind of machine vision method of miniature binding post quality testing | |
| CN111398287A (en) | Battery pole piece scratch detection system and detection method | |
| CN108805144A (en) | Shell hole recognition methods based on morphology correction and system, indication of shots equipment | |
| CN107609600A (en) | A kind of greenhouse insect-sticking plate insect automatic recognition classification method and system | |
| CN111458345A (en) | Visual detection mechanism for defects of mask | |
| CN114286014A (en) | Image database acquisition method | |
| CN112819017B (en) | High-precision color cast image identification method based on histogram | |
| CN116229856B (en) | Automatic control screen detection system and method for computer | |
| CN116486212A (en) | Water gauge identification method, system and storage medium based on computer vision | |
| CN111127494B (en) | Highway truck weight limiting identification method based on image processing | |
| CN112098415A (en) | Nondestructive testing method for quality of waxberries | |
| CN107243461B (en) | Automatic grading device | |
| CN106652898A (en) | LED display point-to-point correction method | |
| CN116152220A (en) | Seed counting and size measuring method based on machine vision | |
| JPH07198714A (en) | Method and device for discriminating activity of cell |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |