CN107154049A - A kind of vegetable leaf scab dividing method and server based on colouring information - Google Patents
A kind of vegetable leaf scab dividing method and server based on colouring information Download PDFInfo
- Publication number
- CN107154049A CN107154049A CN201710358725.1A CN201710358725A CN107154049A CN 107154049 A CN107154049 A CN 107154049A CN 201710358725 A CN201710358725 A CN 201710358725A CN 107154049 A CN107154049 A CN 107154049A
- Authority
- CN
- China
- Prior art keywords
- color
- scab
- leaf
- ccf
- leaf image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30188—Vegetation; Agriculture
Abstract
The present invention discloses a kind of vegetable leaf scab dividing method and server based on colouring information.Wherein, methods described includes：The scab leaf image of vegetables is obtained, and the scab leaf image is pre-processed；Color characteristic of the pretreated scab leaf image in different colours space is extracted respectively；The filtering of Color feature detection is built according to the color characteristic in the different colours space；Filtered according to the Color feature detection, obtain Color characteristic pattern corresponding with the scab leaf image；According to the Color characteristic pattern, the scab in the scab leaf image is split.The server is used to perform the above method.The vegetable leaf scab dividing method and server based on colouring information that the present invention is provided, improve the accuracy split to scab in scab leaf image.
Description
Technical field
The present invention relates to technical field of image processing, and in particular to a kind of vegetable leaf scab segmentation based on colouring information
Method and server.
Background technology
In greenhouse vegetable plantation, disease is to cause one of principal element of quality of vegetable reduction, and accurately disease is carried out
Identifying and diagnosing, pair with greenhouse vegetable plant it is significant.
The key accurately identified to disease is the accurate method for obtaining scab information, being split based on image, from field
Between extract scab image information in the leaf image that gathers, be one of current main approach for obtaining defect information.Color is believed
Breath is to discriminate between the most direct information of scab blade and normal blade, but colouring information is highly prone to the influence of illumination condition.Field
The image gathered in practical situations both, the interference such as background, uneven illumination by IMAQ be even, causes the leaf image of collection
The identification degree reduction of middle scab, causes the difficulty judged scab image.
Therefore, a kind of method how is proposed, can be illumination condition be uneven and the feelings such as leaf image collection background complexity
Under condition, the scab leaf image collected is handled, the accuracy split to scab in scab leaf image is improved, turns into
Industry urgent problem to be solved.
The content of the invention
For defect of the prior art, the present invention provides a kind of vegetable leaf scab dividing method based on colouring information
And server.
On the one hand, the present invention proposes a kind of vegetable leaf scab dividing method based on colouring information, including：
The scab leaf image of vegetables is obtained, and the scab leaf image is pre-processed；
Color characteristic of the pretreated scab leaf image in different colours space is extracted respectively；
The filtering of Color feature detection is built according to the color characteristic in the different colours space；
Filtered according to the Color feature detection, obtain Color feature corresponding with the scab leaf image
Figure；
According to the Color characteristic pattern, the scab in the scab leaf image is split.
On the other hand, the present invention provides a kind of server, including：
Image pre-processing module, the scab leaf image for obtaining vegetables, and the scab leaf image is carried out pre-
Processing；
Characteristic extracting module, for extracting face of the pretreated scab leaf image in different colours space respectively
Color characteristic；
Filtering builds module, is filtered for building Color feature detection according to the color characteristic in the different colours space
Ripple；
Characteristic pattern obtains module, for being filtered according to the Color feature detection, obtains and the scab blade figure
As corresponding Color characteristic pattern；
Scab splits module, for according to the Color characteristic pattern, entering to the scab in the scab leaf image
Row segmentation.
The vegetable leaf scab dividing method and server based on colouring information that the present invention is provided, due to can be by right
The scab leaf image of acquisition is pre-processed, and extracts color of the pretreated scab leaf image in different colours space
Feature, builds the filtering of Color feature detection, obtains Color characteristic pattern corresponding with scab leaf image, realize to disease
Scab in variegated leaf picture is split, and improves the accuracy split to scab in scab leaf image.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the accompanying drawing used required in technology description to be briefly described, it should be apparent that, drawings in the following description are this hairs
Some bright embodiments, for those of ordinary skill in the art, on the premise of not paying creative work, can be with root
Other accompanying drawings are obtained according to these accompanying drawings.
Fig. 1 is the schematic flow sheet of vegetable leaf scab dividing method of the one embodiment of the invention based on colouring information；
Fig. 2 is the ExR parameters and CCF ratios of vegetable leaf scab dividing method of the one embodiment of the invention based on colouring information
The graph of a relation of rate；
Fig. 3 is the structural representation of one embodiment of the invention server.
Embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached in the embodiment of the present invention
Figure, the technical scheme in the embodiment of the present invention is explicitly described, it is clear that described embodiment is a part of the invention
Embodiment, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not making wound
The every other embodiment obtained under the premise of the property made work, belongs to the scope of protection of the invention.
Fig. 1 is the schematic flow sheet of vegetable leaf scab dividing method of the one embodiment of the invention based on colouring information, such as
Shown in Fig. 1, the vegetable leaf scab dividing method based on colouring information that the present invention is provided, including：
S101, the scab leaf image for obtaining vegetables, and the scab leaf image is pre-processed；
Specifically, server can get the scab leaf image for the vegetables being planted in greenhouse by video camera, so
The scab leaf image got is pre-processed afterwards, the pretreatment includes image denoising and adjusts the scab leaf
The size of picture, for example, be adjusted to 800x600 pixels.Wherein, the scab blade is the leaf that blade surface includes scab
Piece.
S102, color characteristic of the pretreated scab leaf image in different colours space is extracted respectively；
Specifically, the server carries out face in different color spaces to the pretreated scab leaf image
Color characteristic is extracted.For example, when illumination condition is relatively uniform, color characteristic, such as the super green feature in RGB color
(Excess Green Index, hereinafter referred to as ExG) and exceedingly popular feature (Excess Red Index, hereinafter referred to as ExR),
H components in hsv color space and the b* components in L*a*b* color spaces can realize the area of scab and normal blade
Point.And when illumination condition is uneven, ExG and ExR are easily affected, and H components and institute in the hsv color space
The separating capacity to scab and normal blade can be kept by stating the b* components in L*a*b* color spaces.The server can be carried
The pretreated scab leaf image is taken in the exceedingly popular feature of RGB color, H components in hsv color space and
In the b* components of L*a*b* color spaces.
S103, Color feature detection is built according to the color characteristic in the different colours space filtered；
Specifically, color characteristic of the server based on the different colours space, builds Color feature detection
Filtering.The Color feature detection filtering that the present invention is built includes filtering (Difference of using difference of Gaussian
Gaussian, hereinafter referred to as DoG) and border circular areas mean filter, because RGB color is easily by the shadow of illumination condition
Ring, ExR parameters can be introduced in the Color feature detection filtering, so as to reduce the uneven situation of illumination condition to scab
Split the influence of quality.
For example, based on the pretreated scab leaf image in the exceedingly popular feature of RGB color, HSV face
The H components of the colour space and the b* components in L*a*b* color spaces, the Color feature detection rate ripple is by formula Build, wherein, I is institute
State the pretreated leaf image, IExRFor the exceedingly popular characteristic image of RGB color, Ib*For L*a*b* color spaces
B* component images, IHFor the H component images in hsv color space, pb(r)The border circular areas mean filter for being r for radius,It is (σ for standard deviationH, σL) Difference of Gaussian filter, α be RGB color exceedingly popular characteristic parameter, value
Scope for (0,1], * be two-dimensional discrete convolution operation.
S104, filtered according to the Color feature detection, obtain corresponding with the scab leaf image comprehensive face
Color characteristic figure；
Specifically, the server is filtered according to the Color feature detection of structure, is obtained and the scab leaf
The corresponding Color characteristic pattern of picture.
For example, based on the pretreated scab leaf image in the exceedingly popular feature of RGB color, HSV face
The H components of the colour space and the b* components in L*a*b* color spaces, the Color characteristic pattern is according to formula：CCF=exp (-
β|f(I：R, σH, σL, αR) |) obtain, wherein, CCF represents the Color characteristic pattern, and β is fall off rate parameter, αRFor R values
Corresponding α values when minimum, R is CCF ratios, by formulaCalculate and obtain, M is by illumination effect
The pixel count of leaf area, N is the pixel count of normal leaf area, CCF (xi, yi) and CCF (xj, yj) it is the Color
The CCF values of pixel on characteristic pattern, i and j are positive integer.Wherein, M and N can be obtained by handmarking；It will be appreciated that
In view of α value for (0,1], in order to rapidly obtain αRValue, can (0,1] between equably select predetermined number α,
So as to calculate the R values of the predetermined number, the R values of minimum are selected from the R values of the predetermined number, with the minimum R
It is α to be worth corresponding α valuesR.Fig. 2 is the ExR of vegetable leaf scab dividing method of the one embodiment of the invention based on colouring information
The graph of a relation of parameter and CCF ratios, as shown in Fig. 2 CCF ratio Rs indicate the influence leaf area CCF values by illumination condition
With the CCF value degrees of closeness of normal leaf area, R value is smaller, show influence leaf area by illumination condition with just
The CCF values of normal leaf area are closer to the influence of illumination condition is just smaller.Wherein, the predetermined number can be according to actual feelings
Condition is set, and the embodiment of the present invention is not limited.
S105, according to the Color characteristic pattern, the scab in the scab leaf image is split.
Specifically, the server selectes initial seed point described comprehensive according to the Color characteristic pattern of acquisition
The growth position in characteristic pattern is closed, using region growing method, the scab in the scab leaf image is split.Its
In, growth position of the initial seed point in the comprehensive characteristics figure can be by manually demarcating；To the scab blade
During scab in image is split, the result that scab is split can be optimized using combination form operation,
The operation of the combination form is Wherein,For dilation operation
Symbol,Accorded with for erosion operation, bw is bianry image, S1And S2For the structural element in morphology.
The vegetable leaf scab dividing method based on colouring information that the present invention is provided, due to the disease to acquisition can be passed through
Variegated leaf picture is pre-processed, and extracts pretreated scab leaf image in the color characteristic in different colours space, structure
The filtering of Color feature detection is built, Color characteristic pattern corresponding with scab leaf image is obtained, to scab leaf image
In scab split, improve the accuracy that scab in scab leaf image is split.
It is further, described to extract the pretreated scab blade figure respectively on the basis of the various embodiments described above
The color characteristic in different colours space of picture includes：
Exceedingly popular feature of the pretreated scab leaf image in RGB color is extracted respectively, in HSV face
The H components of the colour space and the b* components in L*a*b* color spaces.
Specifically, when illumination condition is relatively uniform, super green feature and exceedingly popular feature in RGB color, in HSV face
H components in the colour space and the b* components in L*a*b* color spaces can realize the area of scab blade and normal blade
Point.And when illumination condition is uneven, ExG and ExR are easily affected, and H components and institute in the hsv color space
The separating capacity to scab blade and normal blade can be kept by stating the b* components in L*a*b* color spaces.The server can
To extract ExR of the pretreated scab leaf image in RGB color, H components in hsv color space and
The b* components of L*a*b* color spaces.ExR is according to formula IExR(x, y)=1.3R (x, y)-G (x, y) is extracted, wherein, (x, y) is
Pixel coordinate, R (x, y), G (x, y) are the color component value of RGB color (x, y)；H component roots in hsv color space
According to formula H(x, y)=IH(x, y) is extracted, wherein, (x, y) is pixel coordinate, IH(x, y) is H points of (x, y) in hsv color space
Measure value；B* components in L*a*b* color spaces are according to formula b*(x, y)=Ib*(x, y) is extracted, wherein, (x, y) sits for pixel
Mark, Ib*(x, y) is the b* component values of (x, y) in L*a*b* color spaces.
On the basis of the various embodiments described above, further, the color characteristic structure according to the different colours space
Building the filtering of Color feature detection includes：
According to formula
The filtering of Color feature detection is built, wherein, I is the pretreated leaf image, IExRFor RGB color
Exceedingly popular characteristic image, Ib*For the b* component images of L*a*b* color spaces, IHFor the H component images in hsv color space, pb(r)
The border circular areas mean filter for being r for radius,It is (σ for standard deviationH, σL) Difference of Gaussian filter, α is
The exceedingly popular characteristic parameter of RGB color, span for (0,1], * be two-dimensional discrete convolution operation.
Specifically, the server based on the pretreated scab leaf image in the super of RGB color
Red feature, the H components in hsv color space and the b* components in L*a*b* color spaces, according to formula The Color feature detection rate ripple is built, its
In, I is the pretreated leaf image, IExRFor the exceedingly popular characteristic image of RGB color, Ib*For L*a*b* face
The b* component images of the colour space, IHFor the H component images in hsv color space, pb(r)The border circular areas mean filter for being r for radius
Device,It is (σ for standard deviationH, σL) Difference of Gaussian filter, α be RGB color exceedingly popular characteristic parameter,
Span for (0,1], * be two-dimensional discrete convolution operation.Wherein, single pixel in the pretreated leaf image
Point can be expressed as I (x, y), and the Color feature detection filtering of the single pixel point can be expressed as
Wherein, IExR(x, y) is (x, y) of RGB color exceedingly popular characteristic component value.
It is further, described to be filtered according to the Color feature detection on the basis of the various embodiments described above, obtain
Color characteristic pattern corresponding with the scab leaf image includes：
According to formula：CCF=exp (- β | f (I：R, σH, σL, αR) |) the Color characteristic pattern is obtained, wherein, CCF
The Color characteristic pattern is represented, β is fall off rate parameter, αRCorresponding α values during for R values minimum, R is CCF ratios, by public affairs
Formula Calculate and obtain, M is the pixel count by illumination effect leaf area, N is normal leaf area
Pixel count, CCF (xi, yi) and CCF (xj, yj) for the CCF values of pixel on the Color characteristic pattern, i and j are just whole
Number.
Specifically, the server based on the pretreated scab leaf image in the super of RGB color
Red feature, the H components in hsv color space and the b* components in L*a*b* color spaces, according to formula：CCF=exp (- β | f (I：
R, σH, σL, αR) |) the Color characteristic pattern is obtained, wherein, CCF represents the Color characteristic pattern, and β is fall off rate
Parameter, can be obtained, α by testingRCorresponding α values during for R values minimum, R is CCF ratios, by formulaCalculate and obtain, M is the pixel count by illumination effect leaf area, N is normal leaf area
Pixel count, CCF (xi, yi) and CCF (xj, yj) for the CCF values of pixel on the Color characteristic pattern, CCF (xi, yi)=
exp(-β|f(I(xi, yi)：R, σH, σL, αR) |), CCF (xj, yj)=exp (- β | f (I (xj, yj)：R, σH, σL, αR) |), i and j
For positive integer.Wherein, M and N can be obtained by handmarking；It will be understood that, it is contemplated that α value for (0,1], in order to fast
α is obtained fastlyRValue, can (0,1] between equably select predetermined number α, so as to calculate the R of the predetermined number
Value, selects the R values of minimum from the R values of the predetermined number, and α values corresponding with the minimum R values are αR,.Fig. 2 is this
The ExR parameters and the graph of a relation of CCF ratios of vegetable leaf scab dividing method of the embodiment based on colouring information are invented, is such as schemed
Shown in 2, CCF ratio Rs are indicated is influenceed leaf area CCF values and the CCF values of normal leaf area to approach by illumination condition
Degree, R value is smaller, shows to be influenceed the CCF values of leaf area and normal leaf area closer to light by illumination condition
Influence according to condition is just smaller.Wherein, the predetermined number can be set according to actual conditions, and the embodiment of the present invention is not done
Limit.
It is further, described according to the Color characteristic pattern on the basis of the various embodiments described above, to the scab
Scab in leaf image, which carries out segmentation, to be included：
Obtain in the Color characteristic pattern, the growth position of initial seed point；
According to the growth position of the initial seed point, using region growing method to the disease in the scab leaf image
Spot is split.
Specifically, the server is obtained in the Color characteristic pattern, the growth position of initial seed point, described first
The growth position of beginning seed point can have multiple；Wherein, the growth position of the seed point can be by manually demarcating.The clothes
Device be engaged according to the growth position of the initial seed point, using region growing method, to the scab in the scab leaf image
Split.
Fig. 3 is the structural representation of one embodiment of the invention server, as shown in figure 3, the server bag that the present invention is provided
Include image pre-processing module 301, characteristic extracting module 302, filtering structure module 303, characteristic pattern and obtain module 304 and scab point
Module 305 is cut, wherein：
Image pre-processing module 301 is used for the scab leaf image for obtaining vegetables, and the scab leaf image is carried out
Pretreatment；Characteristic extracting module 302 is used to extract the pretreated scab leaf image respectively in different colours space
Color characteristic；Filtering, which builds module 303, to be used to build the inspection of Color feature according to the color characteristic in the different colours space
Survey filtering；Characteristic pattern, which obtains module 304, to be used to be filtered according to the Color feature detection, is obtained and the scab blade figure
As corresponding Color characteristic pattern；Scab segmentation module 305 is used for according to the Color characteristic pattern, to the scab leaf
Scab in picture is split.
Specifically, image pre-processing module 301 can get the scab for the vegetables being planted in greenhouse by video camera
Leaf image, is then pre-processed to the scab leaf image got, and the pretreatment includes image denoising and tune
The size of the whole scab leaf image, for example, be adjusted to 800x600 pixels.Wherein, the scab blade is blade table bread
Blade containing scab.
Characteristic extracting module 302, to the pretreated scab leaf image, carries out color in different color spaces
Feature extraction.For example, when illumination condition is relatively uniform, color characteristic, such as super green feature in RGB color and super
Red feature, the H components in hsv color space and the b* components in L*a*b* color spaces can realize scab blade with
The differentiation of normal blade.And when illumination condition is uneven, ExG and ExR are easily affected, and the hsv color space
In H components and the L*a*b* color spaces in b* components can keep the differentiation energy to scab blade and normal blade
Power.The server can extract exceedingly popular feature of the pretreated scab leaf image in RGB color, in HSV
H components in color space and the b* components in L*a*b* color spaces.
Filtering builds color characteristic of the module 303 based on the different colours space, builds the filter of Color feature detection
Ripple.The Color feature detection filtering that the present invention is built is included using difference of Gaussian filtering and the filter of border circular areas average
Ripple, because RGB color is easily influenceed by illumination condition, can be introduced in the Color feature detection filtering
ExR parameters, so as to reduce the influence that the uneven situation of illumination condition splits quality to scab.
Characteristic pattern obtains module 304 and filtered according to the Color feature detection of structure, obtains and the scab leaf
The corresponding Color characteristic pattern of picture.
Scab splits the Color characteristic pattern of the module 305 according to acquisition, selectes initial seed point in the synthesis
Growth position in characteristic pattern, using region growing method, splits to the scab in the scab leaf image.Wherein,
Growth position of the initial seed point in the comprehensive characteristics figure can be by manually demarcating；To the scab leaf image
In scab split during, can using combination form operation, to scab segmentation result optimize, it is described
The operation of combination form is Wherein,Accorded with for dilation operation,
Accorded with for erosion operation, bw is bianry image, S1And S2For the structural element in morphology.
The server that the present invention is provided, due to that can be pre-processed by the scab leaf image to acquisition, and is extracted
Pretreated scab leaf image builds the filtering of Color feature detection, obtained in the color characteristic in different colours space
Color characteristic pattern corresponding with scab leaf image, realization is split to the scab in scab leaf image, improves
The accuracy split to scab in scab leaf image.
On the basis of above-described embodiment, further, characteristic extracting module 302 specifically for：
The pretreated scab leaf image is extracted respectively in the exceedingly popular feature of RGB color, hsv color
The H components in space and the b* components in L*a*b* color spaces.
Specifically, when illumination condition is relatively uniform, super green feature and exceedingly popular feature in RGB color, in HSV face
H components in the colour space and the b* components in L*a*b* color spaces can realize the area of scab blade and normal blade
Point.And when illumination condition is uneven, ExG and ExR are easily affected, and H components and institute in the hsv color space
The separating capacity to scab blade and normal blade can be kept by stating the b* components in L*a*b* color spaces.Characteristic extracting module
302 can extract ExR of the pretreated scab leaf image in RGB color, the H in hsv color space points
Measure and in the b* components of L*a*b* color spaces.ExR is according to formula IExR(x, y)=1.3R (x, y)-G (x, y) is extracted, wherein,
(x, y) is pixel coordinate, and R (x, y), G (x, y) is the color component value of RGB color (x, y)；H in hsv color space
Component is according to formula H(x, y)=IH(x, y) is extracted, wherein, (x, y) is pixel coordinate, IH(x, y) be hsv color space in (x,
Y) H component values；B* components in L*a*b* color spaces are according to formula b*(x, y)=Ib*(x, y) is extracted, wherein, (x, y)
For pixel coordinate, Ib*(x, y) is the b* component values of (x, y) in L*a*b* color spaces.
On the basis of the various embodiments described above, further, filtering build module 303 specifically for：
According to formula f (I：R, σH, σL, α) and=α (pb(r)*IExR)+DoG(σH, σL)*IH+pb(r)*Ib*Build Color
Feature detection is filtered, wherein, I is the pretreated leaf image, IExRFor the exceedingly popular characteristic pattern of RGB color
Picture, Ib*For the b* component images of L*a*b* color spaces, IHFor the H component images in hsv color space, pb(r)It is r's for radius
Border circular areas mean filter,It is for standard deviationDifference of Gaussian filter, α is that RGB color is empty
Between exceedingly popular characteristic parameter, span for (0,1], * be two-dimensional discrete convolution operation.
Specifically, filtering builds module 303 and is based on the pretreated scab leaf image in RGB color
Exceedingly popular feature, the H components in hsv color space and the b* components in L*a*b* color spaces, according to formula Build the comprehensive face
Color characteristic verification and measurement ratio ripple, wherein, I is the pretreated leaf image, IExRFor the exceedingly popular feature of RGB color
Image, Ib*For the b* component images of L*a*b* color spaces, IHFor the H component images in hsv color space, pb(r)It is r for radius
Border circular areas mean filter,It is (σ for standard deviationH, σL) Difference of Gaussian filter, α is that RGB color is empty
Between exceedingly popular characteristic parameter, span for (0,1], * be two-dimensional discrete convolution operation.Wherein, it is described pretreated described
Single pixel point can be expressed as I (x, y), the Color feature detection filtering of the single pixel point in leaf image
It can be expressed asIH(x, y)+
pb(r)*Ib*(x, y), wherein, IExR(x, y) is (x, y) of RGB color exceedingly popular characteristic component value.
On the basis of the various embodiments described above, further, characteristic pattern obtain module 304 specifically for：
According to formula：CCF=exp (- β | f (I：R, σH, σL, αR) |) the Color characteristic pattern is obtained, wherein, CCF
The Color characteristic pattern is represented, β is fall off rate parameter, αRCorresponding α values during for R values minimum, R is CCF ratios, by public affairs
Formula Calculate and obtain, M is the pixel count by illumination effect leaf area, N is normal leaf area
Pixel count, CCF (xi, yi) and CCF (xj, yj) for the CCF values of pixel on the Color characteristic pattern, i and j are just whole
Number.
Specifically, characteristic pattern obtains module 304 and is based on the pretreated scab leaf image in RGB color sky
Between exceedingly popular feature, the H components in hsv color space and the b* components in L*a*b* color spaces, according to formula：CCF=exp (-
β|f(I：R, σH, σL, αR) |) the Color characteristic pattern is obtained, wherein, CCF represents the Color characteristic pattern, under β is
Rate parameter is dropped, can be obtained by testing, αRCorresponding α values during for R values minimum, R is CCF ratios, by formulaCalculate and obtain, M is the pixel count by illumination effect leaf area, N is normal leaf area
Pixel count, CCF (xi, yi) and CCF (xj, yj) for the CCF values of pixel on the Color characteristic pattern, CCF (xi, yi)=
exp(-β|f(I(xi, yi)：R, σH, σL, αR) |), CCF (xj, yj)=exp (- β | f (I (xj, yj)：R, σH, σL, αR) |), i and j
For positive integer.Wherein, M and N can be obtained by handmarking；It will be understood that, it is contemplated that α value for (0,1], in order to fast
α is obtained fastlyRValue, can (0,1] between equably select predetermined number α, so as to calculate the R of the predetermined number
Value, selects the R values of minimum from the R values of the predetermined number, and α values corresponding with the minimum R values are αR.Fig. 2 is this
The ExR parameters and the graph of a relation of CCF ratios of vegetable leaf scab dividing method of the embodiment based on colouring information are invented, is such as schemed
Shown in 2, CCF ratio Rs are indicated is influenceed leaf area CCF values and the CCF values of normal leaf area to approach by illumination condition
Degree, R value is smaller, shows to be influenceed the CCF values of leaf area and normal leaf area closer to light by illumination condition
Influence according to condition is just smaller.Wherein, the predetermined number can be set according to actual conditions, and the embodiment of the present invention is not done
Limit.
On the basis of the various embodiments described above, further, scab segmentation module 305 includes acquiring unit and segmentation is single
Member, wherein：
The acquiring unit is used for the growth position for obtaining initial seed point in the Color characteristic pattern；The segmentation
Unit is used for the growth position according to the initial seed point, using region growing method to the disease in the scab leaf image
Spot is split.
Specifically, the acquiring unit is obtained in the Color characteristic pattern, and the growth position of initial seed point is described
The growth position of initial seed point can have multiple；Wherein, the growth position of the seed point can be by manually demarcating.It is described
Cutting unit is according to the growth position of the initial seed point, using region growing method, in the scab leaf image
Scab is split.
The embodiment for the server that the present invention is provided specifically can be used for the handling process for performing above-mentioned each method embodiment,
Its function will not be repeated here, and be referred to the detailed description of above method embodiment.
Server example described above is only schematical, wherein the unit illustrated as separating component
It can be or may not be physically separate, the part shown as unit can be or may not be physics list
Member, you can with positioned at a place, or can also be distributed on multiple NEs.It can be selected according to the actual needs
In some or all of module realize the purpose of this embodiment scheme.Those of ordinary skill in the art are not paying creativeness
Work in the case of, you can to understand and implement.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
Realized by the mode of software plus required general hardware platform, naturally it is also possible to pass through hardware.Understood based on such, on
The part that technical scheme substantially in other words contributes to prior art is stated to embody in the form of software product, should
Computer software product can be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disc, CD, including some fingers
Order is to cause a computer equipment (can be personal computer, server, or network equipment etc.) to perform each implementation
Method described in some parts of example or embodiment.
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. a kind of vegetable leaf scab dividing method based on colouring information, it is characterised in that including：
The scab leaf image of vegetables is obtained, and the scab leaf image is pre-processed；
Color characteristic of the pretreated scab leaf image in different colours space is extracted respectively；
The filtering of Color feature detection is built according to the color characteristic in the different colours space；
Filtered according to the Color feature detection, obtain Color characteristic pattern corresponding with the scab leaf image；
According to the Color characteristic pattern, the scab in the scab leaf image is split.
2. according to the method described in claim 1, it is characterised in that described to extract the pretreated scab blade figure respectively
The color characteristic in different colours space of picture includes：
Exceedingly popular feature of the pretreated scab leaf image in RGB color is extracted respectively, it is empty in hsv color
Between H components and the b* components in L*a*b* color spaces.
3. according to the method described in claim 1, it is characterised in that the color characteristic structure according to the different colours space
Building the filtering of Color feature detection includes：
According to formula Structure
The filtering of Color feature detection is built, wherein, I is the pretreated leaf image, IExRFor RGB color
Exceedingly popular characteristic image, Ib*For the b* component images of L*a*b* color spaces, IHFor the H component images in hsv color space, pb(r)For
Radius is r border circular areas mean filter,It is (σ for standard deviationH,σL) Difference of Gaussian filter, α is
The exceedingly popular characteristic parameter of RGB color, span for (0,1], * be two-dimensional discrete convolution operation.
4. method according to claim 3, it is characterised in that described to be filtered according to the Color feature detection, is obtained
Obtaining Color characteristic pattern corresponding with the scab leaf image includes：
According to formula：CCF=exp (- β | f (I:r,σH,σL,αR) |) the Color characteristic pattern is obtained, wherein, CCF is represented
The Color characteristic pattern, β is fall off rate parameter, αRCorresponding α values during for R values minimum, R is CCF ratios, by formula Calculate and obtain, M is the pixel count by illumination effect leaf area, N is normal leaf area
Pixel count, CCF (xi,yi) and CCF (xj,yj) for the CCF values of pixel on the Color characteristic pattern, i and j are positive integer.
5. according to the method described in claim 1, it is characterised in that described according to the Color characteristic pattern, to the disease
Scab in variegated leaf picture, which carries out segmentation, to be included：
Obtain the growth position of initial seed point in the Color characteristic pattern；
According to the growth position of the initial seed point, the scab in the scab leaf image is entered using region growing method
Row segmentation.
6. a kind of server, it is characterised in that including：
Image pre-processing module, the scab leaf image for obtaining vegetables, and the scab leaf image is pre-processed；
Characteristic extracting module, it is special for extracting color of the pretreated scab leaf image in different colours space respectively
Levy；
Filtering builds module, is filtered for building Color feature detection according to the color characteristic in the different colours space；
Characteristic pattern obtains module, for being filtered according to the Color feature detection, obtains and the scab leaf image pair
The Color characteristic pattern answered；
Scab splits module, for according to the Color characteristic pattern, being divided the scab in the scab leaf image
Cut.
7. server according to claim 1, it is characterised in that the characteristic extracting module specifically for：
Exceedingly popular feature of the pretreated scab leaf image in RGB color is extracted respectively, it is empty in hsv color
Between H components and the b* components in L*a*b* color spaces.
8. server according to claim 1, it is characterised in that the filtering build module specifically for：
According to formula Structure
The filtering of Color feature detection is built, wherein, I is the pretreated leaf image, IExRFor RGB color
Exceedingly popular characteristic image, Ib*For the b* component images of L*a*b* color spaces, IHFor the H component images in hsv color space, pb(r)For
Radius is r border circular areas mean filter,It is (σ for standard deviationH,σL) Difference of Gaussian filter, α is
The exceedingly popular characteristic parameter of RGB color, span for (0,1], * be two-dimensional discrete convolution operation.
9. server according to claim 8, it is characterised in that the characteristic pattern obtain module specifically for：
According to formula：CCF=exp (- β | f (I:r,σH,σL,αR) |) the Color characteristic pattern is obtained, wherein, CCF is represented
The Color characteristic pattern, β is fall off rate parameter, αRCorresponding α values during for R values minimum, R is CCF ratios, by formula Calculate and obtain, M is the pixel count by illumination effect leaf area, N is normal leaf area
Pixel count, CCF (xi,yi) and CCF (xj,yj) for the CCF values of pixel on the Color characteristic pattern, i and j are positive integer.
10. server according to claim 1, it is characterised in that the scab segmentation module includes acquiring unit and divided
Unit is cut, wherein：
The acquiring unit, the growth position for obtaining initial seed point in the Color characteristic pattern；
The cutting unit, for the growth position according to the initial seed point, using region growing method to the scab
Scab in leaf image is split.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710358725.1A CN107154049B (en) | 2017-05-19 | 2017-05-19 | Vegetable leaf scab segmentation method based on color information and server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710358725.1A CN107154049B (en) | 2017-05-19 | 2017-05-19 | Vegetable leaf scab segmentation method based on color information and server |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107154049A true CN107154049A (en) | 2017-09-12 |
| CN107154049B CN107154049B (en) | 2020-12-25 |
Family
ID=59793052
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710358725.1A Active CN107154049B (en) | 2017-05-19 | 2017-05-19 | Vegetable leaf scab segmentation method based on color information and server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107154049B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111292231A (en) * | 2018-11-21 | 2020-06-16 | 杭州职业技术学院 | High-throughput plant phenotypic characteristic extraction method based on mosaic image |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102184392A (en) * | 2011-05-25 | 2011-09-14 | 中国水稻研究所 | DSP (Digital Signal Processor)-based rice weed recognition system and method |
| CN103247059A (en) * | 2013-05-27 | 2013-08-14 | 北京师范大学 | Remote sensing image region of interest detection method based on integer wavelets and visual features |
| CN103559511A (en) * | 2013-11-20 | 2014-02-05 | 天津农学院 | Automatic identification method of foliar disease image of greenhouse vegetable |
| CN103996041A (en) * | 2014-05-15 | 2014-08-20 | 武汉睿智视讯科技有限公司 | Vehicle color identification method and system based on matching |
| CN103996185A (en) * | 2014-04-29 | 2014-08-20 | 重庆大学 | Image segmentation method based on attention TD-BU mechanism |
| CN104021544A (en) * | 2014-05-07 | 2014-09-03 | 中国农业大学 | Greenhouse vegetable disease surveillance video key frame extracting method and extracting system |
| CN104063686A (en) * | 2014-06-17 | 2014-09-24 | 中国科学院合肥物质科学研究院 | System and method for performing interactive diagnosis on crop leaf segment disease images |
| CN104598908A (en) * | 2014-09-26 | 2015-05-06 | 浙江理工大学 | Method for recognizing diseases of crop leaves |
| CN106023159A (en) * | 2016-05-10 | 2016-10-12 | 中国农业大学 | Disease spot image segmentation method and system for greenhouse vegetable leaf |
| CN106250895A (en) * | 2016-08-15 | 2016-12-21 | 北京理工大学 | A kind of remote sensing image region of interest area detecting method |
-
2017
- 2017-05-19 CN CN201710358725.1A patent/CN107154049B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102184392A (en) * | 2011-05-25 | 2011-09-14 | 中国水稻研究所 | DSP (Digital Signal Processor)-based rice weed recognition system and method |
| CN103247059A (en) * | 2013-05-27 | 2013-08-14 | 北京师范大学 | Remote sensing image region of interest detection method based on integer wavelets and visual features |
| CN103559511A (en) * | 2013-11-20 | 2014-02-05 | 天津农学院 | Automatic identification method of foliar disease image of greenhouse vegetable |
| CN103996185A (en) * | 2014-04-29 | 2014-08-20 | 重庆大学 | Image segmentation method based on attention TD-BU mechanism |
| CN104021544A (en) * | 2014-05-07 | 2014-09-03 | 中国农业大学 | Greenhouse vegetable disease surveillance video key frame extracting method and extracting system |
| CN103996041A (en) * | 2014-05-15 | 2014-08-20 | 武汉睿智视讯科技有限公司 | Vehicle color identification method and system based on matching |
| CN104063686A (en) * | 2014-06-17 | 2014-09-24 | 中国科学院合肥物质科学研究院 | System and method for performing interactive diagnosis on crop leaf segment disease images |
| CN104598908A (en) * | 2014-09-26 | 2015-05-06 | 浙江理工大学 | Method for recognizing diseases of crop leaves |
| CN106023159A (en) * | 2016-05-10 | 2016-10-12 | 中国农业大学 | Disease spot image segmentation method and system for greenhouse vegetable leaf |
| CN106250895A (en) * | 2016-08-15 | 2016-12-21 | 北京理工大学 | A kind of remote sensing image region of interest area detecting method |
Non-Patent Citations (7)
| Title |
|---|
| MASSIMO MINERVINI等: "Image-based plant phenotyping with incremental learning and active contours", 《ECOLOGICAL INFORMATICS》 * |
| 刁智华等: "基于颜色和形状特征的棉花害螨图像分割方法", 《农机化研究》 * |
| 汤慧: "多颜色空间融合的颜色特征提取方法及应用", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 王津京等: "采摘机器人基于支持向量机苹果识别方法", 《农业机械学报》 * |
| 王玉德等: "复杂背景下甜瓜果实分割算法", 《农业工程学报》 * |
| 赵钦佩等: "基于颜色信息与区域生长的图像分割新算法", 《上海交通大学学报》 * |
| 马浚诚等: "基于图像处理的温室黄瓜霜霉病诊断系统", 《农业机械学报》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111292231A (en) * | 2018-11-21 | 2020-06-16 | 杭州职业技术学院 | High-throughput plant phenotypic characteristic extraction method based on mosaic image |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107154049B (en) | 2020-12-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP1977371B1 (en) | Method and system for identifying illumination fields in an image | |
| Payne et al. | Estimation of mango crop yield using image analysis–segmentation method | |
| Diago et al. | Assessment of flower number per inflorescence in grapevine by image analysis under field conditions | |
| CN108206917B (en) | Image processing method and device, storage medium and electronic device | |
| CN102214306B (en) | Leaf disease spot identification method and device | |
| CN110569786B (en) | Fruit tree identification and quantity monitoring method and system based on unmanned aerial vehicle data acquisition | |
| CN104636759B (en) | A kind of method and picture filter information recommendation system for obtaining picture and recommending filter information | |
| CN106250896B (en) | The recognition methods of the front and back sides of online tobacco leaf based on image collecting device | |
| CN109740721B (en) | Wheat ear counting method and device | |
| CN108564021B (en) | Method for extracting vegetation coverage of desert based on digital photos | |
| CN109325431A (en) | The detection method and its device of vegetation coverage in Crazing in grassland sheep feeding path | |
| Bhookya et al. | Yield estimation of Chilli crop using image processing techniques | |
| CN109472788A (en) | A kind of scar detection method on airplane riveting surface | |
| CN110428374B (en) | Small-size pest automatic detection method and system | |
| CN113313692B (en) | Automatic banana young plant identification and counting method based on aerial visible light image | |
| CN105654088B (en) | Offal extracting method based on Local Symmetric relationship | |
| CN107154049A (en) | A kind of vegetable leaf scab dividing method and server based on colouring information | |
| CN111768456A (en) | Feature extraction method based on wood color space | |
| Yao et al. | Study on detection method of external defects of potato image in visible light environment | |
| CN106504258B (en) | A kind of leaf image extracting method and device | |
| Di et al. | The research on the feature extraction of sunflower leaf rust characteristics based on color and texture feature | |
| EP1038267A1 (en) | Method and device for detecting and processing images of biological tissue | |
| CN115170862A (en) | Machine vision-based fresh tobacco part identification method | |
| CN108634241A (en) | Select halogen method and its brine-adding system | |
| CN114581660A (en) | Plant leaf segmentation identification method and system |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |