CN108765433A - One kind is for carrying high-precision leafy area measurement method - Google Patents
One kind is for carrying high-precision leafy area measurement method Download PDFInfo
- Publication number
- CN108765433A CN108765433A CN201810552477.9A CN201810552477A CN108765433A CN 108765433 A CN108765433 A CN 108765433A CN 201810552477 A CN201810552477 A CN 201810552477A CN 108765433 A CN108765433 A CN 108765433A
- Authority
- CN
- China
- Prior art keywords
- image
- objective table
- leaf
- pixel
- formula
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000691 measurement method Methods 0.000 title claims abstract description 17
- 238000012937 correction Methods 0.000 claims abstract description 18
- 238000003709 image segmentation Methods 0.000 claims abstract description 16
- 238000005259 measurement Methods 0.000 claims abstract description 13
- 238000012545 processing Methods 0.000 claims abstract description 7
- 238000002360 preparation method Methods 0.000 claims abstract description 5
- 238000000605 extraction Methods 0.000 claims description 11
- 238000013461 design Methods 0.000 claims description 9
- 239000011521 glass Substances 0.000 claims description 7
- 230000011218 segmentation Effects 0.000 claims description 7
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 238000003708 edge detection Methods 0.000 claims description 6
- 238000013507 mapping Methods 0.000 claims description 5
- 238000013519 translation Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 3
- 101150036841 minJ gene Proteins 0.000 claims description 3
- 230000002093 peripheral effect Effects 0.000 claims description 3
- 238000000926 separation method Methods 0.000 claims description 3
- 230000004075 alteration Effects 0.000 claims 1
- 238000000034 method Methods 0.000 abstract description 22
- 239000000284 extract Substances 0.000 abstract description 5
- 230000000694 effects Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 3
- 230000008635 plant growth Effects 0.000 description 3
- 241000196324 Embryophyta Species 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 241000894007 species Species 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 241000748095 Hymenopappus filifolius Species 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000007598 dipping method Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 230000000243 photosynthetic effect Effects 0.000 description 1
- 230000008121 plant development Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
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
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/002—Measuring arrangements characterised by the use of optical techniques for measuring two or more coordinates
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B11/00—Measuring arrangements characterised by the use of optical techniques
- G01B11/28—Measuring arrangements characterised by the use of optical techniques for measuring areas
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/80—Geometric correction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of leafy area measurement methods of high-precision, include the following steps:Advance preparation of taking pictures;Image Acquisition;The geometric correction of imagery;Image segmentation and binaryzation;Sub-pixel edge extracts;Calculate objective table and leaf valid pixel;Noise is isolated in removal;Leaf area calculates.Present invention employs the measurement accuracy that image sub-pixel edge extractive technique improves leaf area;The present invention carries out the calculating of leaf area using image processing techniques to improve measurement efficiency automatically.
Description
Technical field
The invention belongs to plant growth information measurement technical fields, specifically, being related to a kind of high-precision more for carrying
Leaf area measurement method.
Background technology
Blade is that the important photosynthetic organs of plant and plant carry out rising main place, and size is to plant
Growth and development, crop yield and cultivation management all have very great influence.Therefore, the leaf of accurate crops is obtained
Area is important one of the Task in ecological agriculture infomation detection field, for assessing the growing way of crop and establishing plant growth
Model has very important research significance.
The method that leaf area measures conventional comparative maturity has graph paper method, duplicates weight method, measures the foundation of blade length and width
Regression equation method, leaf area instrument measuring method etc..Graph paper method and duplicating Weight needs manually draw or cut leaf edges,
Heavy workload, it is complicated for operation, using wideless, error is larger if leaf edges are complex-shaped.Blade length and width are measured to establish back
Return equation method to need to extract leaf-shaped parameter from a large amount of blade, establish and calculate leaf area regression equation, this method all needs
Great amount of samples is artificially collected, carries out hand dipping, program is cumbersome and precision is not high.Leaf area instrument mensuration speed is fast, but its
Expensive, repair is inconvenient, and measurement accuracy is affected by crop groups structure, and is unsuitable for long wide blade face
Product measures.
Invention content
In view of this, being used to carry high-precision leafy area measurement method the present invention provides one kind, the measurement method packet
Containing specially designed objective table and image processing algorithm software, the blade on mobile phone or digital camera shooting objective table, warp are utilized
The image processing algorithms such as the geometric correction of imagery, the extraction of image sub-pixel edge and image segmentation are crossed, the effective of leaf is respectively obtained
The valid pixel number of pixel number, objective table backlight, the area of objective table backlight is accurately known, according to green leaves and loading
The ratio of the number of pixels of platform backlight calculates the area of green blade.
In order to solve the above-mentioned technical problem, the invention discloses one kind for carrying high-precision leafy area measurement method,
Include the following steps:
Step 1, preparation of taking pictures in advance:Multiple blades are placed on objective table, leaf is covered with glass cover-plate, connects and carry
The power supply of object platform source of parallel light, entire loading deck plate show white high bright background, and the color of image of leaf shows green;
Step 2, Image Acquisition:Loading deck plate is faced with mobile phone or digital camera so that viewing field of camera can cover
Entire objective table adjusts focal length and takes pictures, and image request is clear, fuzzy without there is shake;
Step 3, the geometric correction of imagery:With mobile terminal image procossing APP softwares or the ends PC image processing software to figure of taking pictures
As being handled, 4 vertex of objective table and center identification point corresponding position on the image are identified first, then to image
Carry out geometric correction;
Step 4, image segmentation and binaryzation:With mobile terminal image procossing APP softwares or the ends PC image processing software, to figure
Picture is split and binary conversion treatment so that background and target separation;
Step 5, sub-pixel edge extraction:With mobile terminal image procossing APP softwares or the ends PC image processing software to image
It is handled, Pixel-level coarse positioning is carried out to image border with canny operators first, then carry out sub-pixel edge extraction;
Noise is isolated in step 6, removal:Different matrixes is set according to the size of Different Crop and picture resolution;
The calculating of step 7, objective table and leaf valid pixel:When carrying out image segmentation first by objective table peripheral images
It gets rid of, leaves the image of objective table part, wherein green leaves image is included in inside objective table image;According to image point
Cut, the extraction of binaryzation and sub-pixel edge is as a result, objective table and leaf image are divided into the region being independently closed, pass through calculating
Machine counts, and calculates constitute objective table independence enclosed region valid pixel number P respectivelyZWith leaf independence enclosed region valid pixel number
PL;
Step 8, leaf area calculate:The area of known objective table is S, and the valid pixel number of leaf is obtained by image procossing
It is PL, the valid pixel number that objective table is obtained by image procossing is PZ, then the calculation formula of leaf area be:
Optionally, the stage design requires as follows:It is installed additional in the backlight of collimated backlight light source onesize
Glass cover-plate, to flatten leaf;The shell of objective table uses standard module shell, compact-sized;The objective table of design is flat
The size of the backlight of row back light is 200mm х 200mm squares.
Optionally, carrying out geometric correction to image is specially:It is practical with them according to 5 pixels in known fault image
Image 5 corresponding points solve geometric correction equation;This 5 points are 4 vertex of square objective table, objective table center respectively
Identification point;It is (x to be located at and correspond to the coordinate of this 5 points on fault image1,y1)、(x2,y2)、(x3,y3)、(x4,y4)、(x5,y5),
Actual coordinate is (X1,Y1)、(X2,Y2)、(X3,Y3)、(X4,Y4)、(X5,Y5), mapping relations such as formula (1) institute between them
Show,
In formula, (ai,bi) it is equation coefficient (i=1,2,3,4) respectively,For scaling, wherein r1It is to carry
The ratio of object platform actual (tube) length and fault image length of taking pictures, r2It is the practical wide ratio with distortion figure image width of taking pictures of objective table;B1For water
Flat translation distance, B2For vertical translation distance;In formula, a total of 10 unknown number and 10 equations find out this by least square
10 parameters.
Optionally, described image segmentation and binaryzation are specially:
It cuts:The objective table front that the present invention designs is white light background, and the target of measurement is green leaves, so background and mesh
It is big to mark color distortion.First, image is converted to gray level image using formula (2), R, G, B respectively represent the red point of image in formula
Amount, green component and blue component:
After carrying out gradation conversion to image, if the size of image is M × N, gray level L, gray value is the number of pixels of i
Use niIt indicates, total number of pixels is n, wherein n=n0+n1+…+nL.Use piIndicate the pixel that gray value is i in gray level image
The probability of appearance, then have
Wherein
Assuming that t is background D0With target D1Segmentation threshold, i.e. D0={ 0,1 ..., t }, D1={ t+1, t+2 ..., L-1 },
Then D0And D1Respective prior probability P0(t) and P1(t) it is respectively
Wherein, P0(t)+P1(t)=1
D0And D1The mean value being respectively distributed is μ0(t) and μ1(t), respectively
Wherein,
WithRespectively D0And D1The variance being respectively distributed is shown in formula (7),
According to minimal error classificating thought between class, error classification object function J (t) between infima species is established herein
Optimal threshold t*It is obtained when J (t) is minimized, i.e.,
t*=arg { minJ (t) } 0≤t≤L-1 (10)
The gray value of image is indicated with g (x, y), then finally obtained binary segmentation image is
Optionally, the sub-pixel edge, which extracts, is specially:If curvilinear regression of second degree equation is:
In formula, k1、k2、k3It is the coefficient of conic section regression equation, when the value of x is x1、x2、…、xnWhen, it is corresponding to return
The equation is returned to be:
If regressand valueWith actually detected value yiDifference σiIt indicates, has:
Formula (14) indicates marginal point (xi,yi) with the departure degree of regressand value.If T is whole edge detection values and regressand value
Departure degree, then have:
Keep whole edge detection values and regressand value closest, it is desirable that the departure degree of the two is minimum, according to extremum conditions
HaveWithIt can be in the hope of:
Wherein,
It is enabled in formula (16):
With least square, then can find out K is:
K=(DTD)-1C (17)。
Optionally, the removal is isolated in noise, and different squares is arranged according to the size of Different Crop and picture resolution
Battle array be specially:5 х, 5 matrixes are set, that is, analyze the pixel value of some 24 point in pixel periphery, are connected to if it is greater than 10 points,
It is then calculated as noise, is otherwise considered as Efficient leaf area.
Compared with prior art, the present invention can be obtained including following technique effect:
1) present invention employs image sub-pixel edge extractive techniques so that and leaf image segmentation precision greatly improves, from
And improve the measurement accuracy of leaf area.
2) present invention carries out the calculating of leaf area using image processing techniques automatically, need not be artificial manual located
Reason, to improve measurement efficiency.
Certainly, it implements any of the products of the present invention it is not absolutely required to while reaching all the above technique effect.
Description of the drawings
Attached drawing described herein is used to provide further understanding of the present invention, and constitutes the part of the present invention, this hair
Bright illustrative embodiments and their description are not constituted improper limitations of the present invention for explaining the present invention.In the accompanying drawings:
Fig. 1 is the flow chart of the leafy area measurement method of high-precision of the invention;
Fig. 2 is that the present invention is used for piecture geometry fault calibration result figure;Wherein, a is geometric distortion image, and b is that geometry is abnormal
Become image after correcting;
Fig. 3 is that the present invention is used for multiple-blade image segmentation and sub-pixel edge extraction effect figure;Wherein, a is artwork, and b is
Image after segmentation and binaryzation, c are image after sub-pixel edge extraction;
Fig. 4 is case of the present invention and verification the verifying results figure, wherein a is artwork, and b is image after processing.
Specific implementation mode
Carry out the embodiment that the present invention will be described in detail below in conjunction with embodiment, thereby to the present invention how application technology hand
Section solves technical problem and reaches the realization process of technical effect to fully understand and implement.
The invention discloses a kind of leafy area measurement methods of high-precision, as shown in Figure 1, including the following steps:
1) advance preparation of taking pictures:Multiple blades are placed on objective table, leaf is covered with glass cover-plate, connects objective table
The power supply of source of parallel light, entire loading deck plate show white high bright background, and the color of image of leaf shows green;
In order to ensure leaf edges image quality, objective table is designed using white collimated backlight light source so that leaf side
Edge is clear-cut, reduces empty side phenomenon, improves the precision of measurement.It is installed additional in the backlight of collimated backlight light source an equal amount of
Glass cover-plate, to flatten leaf.The shell of objective table uses standard module shell, compact-sized.The objective table of design is parallel
The size of the backlight of back light is 200mm х 200mm squares.
2) Image Acquisition:Loading deck plate is faced with mobile phone or digital camera so that viewing field of camera can cover entirely
Objective table adjusts focal length and takes pictures, and image request is clear, fuzzy without there is shake;
3) geometric correction of image:With mobile terminal image procossing APP softwares or the ends PC image processing software to image of taking pictures
Handled, identify 4 vertex of objective table and center identification point corresponding position on the image first, then to image into
Row geometric correction;
Carrying out geometric correction to image is specially:Since mobile phone or digital camera tilt, to cause the image of shooting will produce several
What distorts, and the main thought of geometric correction is the Function Mapping relationship for establishing fault image and real image, to fault image institute
There is pixel to be corrected according to mapping relations, then gray-level interpolation is carried out to vacancy pixel, to realize geometric correction.The present invention
It proposes and solves geometric correction equation with 5 corresponding points of their real images according to 5 pixels in known fault image.This
5 points are 4 vertex of square objective table, objective table center identification point respectively.It is located at and corresponds to this 5 points on fault image
Coordinate is (x1,y1)、(x2,y2)、(x3,y3)、(x4,y4)、(x5,y5), actual coordinate is (X1,Y1)、(X2,Y2)、(X3,Y3)、
(X4,Y4)、(X5,Y5), shown in the mapping relations such as formula (1) between them,
In formula, (ai,bi) it is equation coefficient (i=1,2,3,4) respectively,For scaling, wherein r1It is to carry
The ratio of object platform actual (tube) length and fault image length of taking pictures, r2It is the practical wide ratio with distortion figure image width of taking pictures of objective table.B1For water
Flat translation distance (pixel), B2For vertical translation distance (pixel).In formula, a total of 10 unknown number and 10 equations, by most
Small two multiply and can calculate;Geometric distortion image is shown in Fig. 2 a, and image is shown in Fig. 2 b after geometric distortion correction.
4) image segmentation:The objective table front that the present invention designs is white light background, and the target of measurement is green leaves, so
Background and color of object difference are big.First, image is converted to gray level image using formula (2), R, G, B respectively represent image in formula
Red component, green component and blue component:
After carrying out gradation conversion to image, if the size of image is M × N, gray level L, gray value is the number of pixels of i
Use niIt indicates, total number of pixels is n, wherein n=n0+n1+…+nL.Use piIndicate the pixel that gray value is i in gray level image
The probability of appearance, then have
Wherein
Assuming that t is background D0With target D1Segmentation threshold, i.e. D0={ 0,1 ..., t }, D1={ t+1, t+2 ..., L-1 },
Then D0And D1Respective prior probability P0(t) and P1(t) it is respectively
Wherein, P0(t)+P1(t)=1
D0And D1The mean value being respectively distributed is μ0(t) and μ1(t), respectively
Wherein,
WithRespectively D0And D1The variance being respectively distributed is shown in formula (7),
According to minimal error classificating thought between class, error classification object function J (t) between infima species is established herein
Optimal threshold t*It is obtained when J (t) is minimized, i.e.,
t*=arg { minJ (t) } 0≤t≤L-1 (10)
The gray value of image is indicated with g (x, y), then finally obtained binary segmentation image is
Image is shown in Fig. 3 b after segmentation and binaryzation.
(5) sub-pixel edge extracts:On the basis of image segmentation and binaryzation, in order to improve the measurement essence of leaf area
Degree, it is necessary to extract the sub-pixel edge of image, and then can just obtain high-precision leaf area.First, it is carried out with canny operators
Pixel edge positioning extraction, if the coordinate of extraction edge pixel is (x1,y1)、(x2,y2)、…、(xn,yn), leaf area edge
It is curve in most cases, so being fitted leaf sub-pixel edge with conic section.If curvilinear regression of second degree equation
For:
In formula, k1、k2、k3It is the coefficient of conic section regression equation, when the value of x is x1、x2、…、xnWhen, it is corresponding to return
The equation is returned to be:
If regressand valueWith actually detected value yiDifference σiIt indicates, has:
Formula (14) indicates marginal point (xi,yi) with the departure degree of regressand value.If T is whole edge detection values and regressand value
Departure degree, then have:
Keep whole edge detection values and regressand value closest, it is desirable that the departure degree of the two is minimum, according to extremum conditions
HaveWithIt can be in the hope of:
Wherein,
It is enabled in formula (16):
With least square, then can find out K is:
K=(DTD)-1C (17)
It is verified, using formula (12) carry out sub-pixel edge fitting when, carry out piecewise fitting when effect it is more preferable, the present invention in
Piecewise fitting is carried out with 50 points.
6) calculating of objective table and leaf valid pixel:The objective table front that the present invention designs is highlighted white light background,
First get rid of objective table peripheral images when image segmentation, leaves the image of objective table part, wherein green leaf subgraph
As being included in inside objective table image;It is extracted as a result, objective table and leaf figure according to image segmentation, binaryzation and sub-pixel edge
As being divided into the region being independently closed, is counted by computer, calculate constitute the effective picture of objective table independence enclosed region respectively
Prime PZWith leaf independence enclosed region valid pixel number PL;
7) it removes and isolates noise:Different matrixes is set according to the size of Different Crop and picture resolution;The present invention is set
5 х, 5 matrixes are set, that is, analyze the pixel value of some 24 point in pixel periphery, is connected to if it is greater than 10 points, is then calculated as noise,
Otherwise it is considered as Efficient leaf area.
8) leaf area calculates:The area of known objective table is S, preferably S=200mm × 200mm, is obtained by image procossing
Valid pixel number to leaf is PL, the valid pixel number that objective table is obtained by image procossing is PZ, then the calculating of leaf area is public
Formula is:
Embodiment 1
Advance preparation of taking pictures:Multiple blades are placed on objective table, leaf is covered with glass cover-plate, it is flat to connect objective table
The power supply of line light source, entire loading deck plate show white high bright background, and the color of image of leaf shows green.
Image Acquisition:Loading deck plate is faced with mobile phone or digital camera so that viewing field of camera can cover entire load
Object platform adjusts focal length and takes pictures, and image request is clear, fuzzy without there is shake.
The geometric correction of imagery:Image of taking pictures is carried out with mobile terminal image procossing APP softwares or the ends PC image processing software
Processing identifies 4 vertex of objective table and center identification point corresponding position on the image, is then carried out to image several first
What is corrected, and geometric correction algorithm sees above described.
Image segmentation and binaryzation:With mobile terminal image procossing APP softwares or the ends PC image processing software, image is carried out
Segmentation and binary conversion treatment so that background and target separation, image segmentation and Binarization methods see above described.
Noise is isolated in removal:Digital picture can be regarded as the set of pixel, exist between pixel interconnected or mutually adjacent
The relationship connect.It for binaryzation digital picture, connects on 2 points of path, the value of pixel is constant, then it is connection to claim at this 2 points.
And the leaf image of binaryzation is exactly a big unicom set, herein by analyzing the pixel values of multiple points around some pixel
To analyze whether the point is contained within the scope of blade.According to Different Crop and the size of picture resolution, different squares can be set
Battle array.5 х, 5 matrixes are arranged in this experiment, that is, analyze the pixel value of some 24 point in pixel periphery, are connected to if it is greater than 10 points,
It is then calculated as noise, is otherwise considered as Efficient leaf area.
Sub-pixel edge extracts:With mobile terminal image procossing APP softwares or the ends PC image processing software to image at
Reason, Pixel-level coarse positioning is carried out to image border with canny operators first, then with sub-pixel positioning algorithm mentioned above into
Row sub-pixel edge extracts, as shown in Figure 3.
Leaf area calculates:The area of known objective table is S, and the valid pixel number that leaf is obtained by image procossing is PL,
The valid pixel number that objective table is obtained by image procossing is PZ, then the calculation formula of leaf area be:
In order to verify the validity of the method for the present invention, respectively with graph paper method, duplicate weight method, three kinds of leaves of potoshop methods
Area measurement method is compared, and the index compared is measurement accuracy and time of measuring.Choose 4 sizes, shape, length and width
It than different blades, is numbered respectively by 1,2,3,4,4 leaf images of implementation steps pair according to the invention are handled
It is calculated with leaf area, processing result image is as shown in figure 4, four kinds of measurement method results contrasts are as shown in table 1.It can from table 1
Go out, the method for the present invention is better than other three kinds of methods from measurement accuracy and time of measuring.
Table 1
Above description has shown and described several preferred embodiments of invention, but as previously described, it should be understood that invention is not
It is confined to form disclosed herein, is not to be taken as excluding other embodiments, and can be used for various other combinations, modification
And environment, and can be carried out by the above teachings or related fields of technology or knowledge in the scope of the invention is set forth herein
Change.And changes and modifications made by those skilled in the art do not depart from the spirit and scope of invention, then should all be weighed appended by invention
In the protection domain that profit requires.
Claims (6)
1. one kind is for carrying high-precision leafy area measurement method, which is characterized in that include the following steps:
Step 1, preparation of taking pictures in advance:Multiple blades are placed on objective table, leaf is covered with glass cover-plate, connects objective table
The power supply of source of parallel light, entire loading deck plate show white high bright background, and the color of image of leaf shows green;
Step 2, Image Acquisition:Loading deck plate is faced with mobile phone or digital camera so that viewing field of camera can cover entirely
Objective table adjusts focal length and takes pictures, and image request is clear, fuzzy without there is shake;
Step 3, the geometric correction of imagery:With mobile terminal image procossing APP softwares or the ends PC image processing software to take pictures image into
Row processing identifies 4 vertex of objective table and center identification point corresponding position on the image, is then carried out to image first
Geometric correction;
Step 4, image segmentation and binaryzation:With mobile terminal image procossing APP softwares or the ends PC image processing software, to image into
Row segmentation and binary conversion treatment so that background and target separation;
Step 5, sub-pixel edge extraction:Image is carried out with mobile terminal image procossing APP softwares or the ends PC image processing software
Processing, carries out Pixel-level coarse positioning with canny operators to image border first, then carries out sub-pixel edge extraction;
Noise is isolated in step 6, removal:Different matrixes is set according to the size of Different Crop and picture resolution;
The calculating of step 7, objective table and leaf valid pixel:Objective table peripheral images are removed first when carrying out image segmentation
Fall, leave the image of objective table part, wherein green leaves image is included in inside objective table image;According to image segmentation, two
Value and sub-pixel edge extraction pass through computer meter as a result, objective table and leaf image are divided into the region being independently closed
Number calculates constitute objective table independence enclosed region valid pixel number P respectivelyZWith leaf independence enclosed region valid pixel number PL;
Step 8, leaf area calculate:The area of known objective table is S, and the valid pixel number that leaf is obtained by image procossing is PL,
The valid pixel number that objective table is obtained by image procossing is PZ, then the calculation formula of leaf area be:
2. according to claim 1 for carrying high-precision leafy area measurement method, which is characterized in that the loading
Platform design requirement is as follows:An equal amount of glass cover-plate is installed additional in the backlight of collimated backlight light source, to flatten leaf;It carries
The shell of object platform uses standard module shell, compact-sized;The size of the backlight of the objective table collimated backlight light source of design is
200mm х 200mm squares.
3. according to claim 1 for carrying high-precision leafy area measurement method, which is characterized in that carried out to image
Geometric correction is specially:According to 5 pixels in known fault image geometry school is solved with 5 corresponding points of their real images
Positive equation;This 5 points are 4 vertex of square objective table, objective table center identification point respectively;It is located on fault image corresponding
The coordinate of this 5 points is (x1,y1)、(x2,y2)、(x3,y3)、(x4,y4)、(x5,y5), actual coordinate is (X1,Y1)、(X2,Y2)、
(X3,Y3)、(X4,Y4)、(X5,Y5), shown in the mapping relations such as formula (1) between them,
In formula, (ai,bi) it is equation coefficient (i=1,2,3,4) respectively,For scaling, wherein r1It is objective table
The ratio of actual (tube) length and fault image length of taking pictures, r2It is the practical wide ratio with distortion figure image width of taking pictures of objective table;B1It is flat for level
Move distance, B2For vertical translation distance;In formula, a total of 10 unknown number and 10 equations find out this 10 by least square
Parameter.
4. according to claim 1 for carrying high-precision leafy area measurement method, which is characterized in that described image point
It cuts and is specially with binaryzation:
It cuts:The objective table front that the present invention designs is white light background, and the target of measurement is green leaves, so background and target face
Aberration is different big.First, convert image to gray level image using formula (2), in formula R, G, B respectively represent image red component,
Green component and blue component:
After carrying out gradation conversion to image, if the size of image is M × N, gray level L, gray value is the number of pixels n of ii
It indicates, total number of pixels is n, wherein n=n0+n1+…+nL.Use piIndicate that gray value is that the pixel of i is pointed out in gray level image
Existing probability, then have
Wherein
Assuming that t is background D0With target D1Segmentation threshold, i.e. D0={ 0,1 ..., t }, D1={ t+1, t+2 ..., L-1 }, then D0
And D1Respective prior probability P0(t) and P1(t) it is respectively
Wherein, P0(t)+P1(t)=1
D0And D1The mean value being respectively distributed is μ0(t) and μ1(t), respectively
Wherein,
WithRespectively D0And D1The variance being respectively distributed is shown in formula (7),
According to minimal error classificating thought between class, error classification object function J (t) between infima species is established herein
Optimal threshold t*It is obtained when J (t) is minimized, i.e.,
t*=arg { minJ (t) } 0≤t≤L-1 (10)
The gray value of image is indicated with g (x, y), then finally obtained binary segmentation image is
5. according to claim 1 for carrying high-precision leafy area measurement method, which is characterized in that the sub- picture
Plain edge extracting is specially:If curvilinear regression of second degree equation is:
In formula, k1、k2、k3It is the coefficient of conic section regression equation, when the value of x is x1、x2、…、xnWhen, corresponding recurrence side
Cheng Wei:
If regressand valueWith actually detected value yiDifference σiIt indicates, has:
Formula (14) indicates marginal point (xi,yi) with the departure degree of regressand value.If T is the inclined of whole edge detection values and regressand value
From degree, then have:
Keep whole edge detection values and regressand value closest, it is desirable that the departure degree of the two is minimum, is had according to extremum conditionsWithIt can be in the hope of:
Wherein,
It is enabled in formula (16):
With least square, then can find out K is:
K=(DTD)-1C (17)。
6. according to claim 1 for carrying high-precision leafy area measurement method, which is characterized in that the removal
Different matrixes, which is arranged, according to the size of Different Crop and picture resolution in isolated noise is specially:5 х, 5 matrixes are set, i.e.,
The pixel value for analyzing some 24 point in pixel periphery is connected to if it is greater than 10 points, is then calculated as noise, is otherwise considered as effective leaf
Area.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810552477.9A CN108765433A (en) | 2018-05-31 | 2018-05-31 | One kind is for carrying high-precision leafy area measurement method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810552477.9A CN108765433A (en) | 2018-05-31 | 2018-05-31 | One kind is for carrying high-precision leafy area measurement method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108765433A true CN108765433A (en) | 2018-11-06 |
Family
ID=64001467
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810552477.9A Pending CN108765433A (en) | 2018-05-31 | 2018-05-31 | One kind is for carrying high-precision leafy area measurement method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108765433A (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109724569A (en) * | 2019-02-02 | 2019-05-07 | 西安卡拉光电科技有限公司 | A method of object plane size is measured using mobile phone camera |
CN110118530A (en) * | 2019-04-22 | 2019-08-13 | 西京学院 | A kind of big workpiece high-precision optical vision measuring method |
CN110349200A (en) * | 2019-07-01 | 2019-10-18 | 西京学院 | A kind of high-precision wheat head length measurement method based on monocular camera |
CN110363807A (en) * | 2019-06-28 | 2019-10-22 | 湖北省农业科学院经济作物研究所 | Based on the high-throughput lossless method for obtaining field cotton Efficient leaf area of RGB image |
CN110953993A (en) * | 2019-11-14 | 2020-04-03 | 国网山西省电力公司大同供电公司 | Detection device and method for sag and distance limit of power transmission line |
CN111223138A (en) * | 2019-12-09 | 2020-06-02 | 天津市气候中心(天津市生态气象和卫星遥感中心、天津市农业气象中心) | Method for calibrating and extracting blade form |
CN111521128A (en) * | 2020-04-15 | 2020-08-11 | 中国科学院海洋研究所 | Shellfish external form automatic measurement method based on optical projection |
CN112465893A (en) * | 2020-11-05 | 2021-03-09 | 内蒙古工业大学 | Vegetation leaf area measuring method and device, terminal and storage medium |
CN113298776A (en) * | 2021-05-21 | 2021-08-24 | 佛山职业技术学院 | Method for detecting appearance defects of metal closed water pump impeller |
CN113624165A (en) * | 2021-08-09 | 2021-11-09 | 海南青峰生物科技有限公司 | Rice leaf area measuring device and method based on 5G and image recognition |
CN114241033A (en) * | 2021-12-17 | 2022-03-25 | 贵州省山地资源研究所 | Ectopic acquisition method for plant leaf area vertical structure |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101813469A (en) * | 2010-04-23 | 2010-08-25 | 中国科学院成都生物研究所 | Method for measuring leaf area by manually shooting |
CN102506772A (en) * | 2011-11-01 | 2012-06-20 | 西北农林科技大学 | Method and device for quickly detecting area of leaf blade based on mobile phone |
CN103530889A (en) * | 2013-11-04 | 2014-01-22 | 天津工业大学 | Method for decoding coding mark points of measuring stick based on skeleton extraction |
CN103968772A (en) * | 2014-05-19 | 2014-08-06 | 南京信息工程大学 | Piston ring detection method based on monocular vision detection |
CN104282018A (en) * | 2014-09-09 | 2015-01-14 | 苏州科力迪软件技术有限公司 | Method for on-line detection of overall diameter of industrial product based on machine vision |
CN105865344A (en) * | 2016-06-13 | 2016-08-17 | 长春工业大学 | Workpiece dimension measuring method and device based on machine vision |
CN106468543A (en) * | 2015-08-21 | 2017-03-01 | 浙江托普云农科技股份有限公司 | A kind of method for measuring leaf area based on image procossing |
CN106468536A (en) * | 2016-11-22 | 2017-03-01 | 天津市气候中心 | A kind of leaf area measurement method |
CN107024187A (en) * | 2017-04-18 | 2017-08-08 | 广西壮族自治区农业科学院农业资源与环境研究所 | A kind of leaf area measurement method based on digital camera images |
CN107154036A (en) * | 2017-03-24 | 2017-09-12 | 中南大学 | Blood vessel segmentation method and its segmenting system based on sub-pix |
CN107478656A (en) * | 2017-08-09 | 2017-12-15 | 齐鲁工业大学 | Paper pulp mixing effect method of determination and evaluation based on machine vision, device, system |
CN108007388A (en) * | 2017-06-30 | 2018-05-08 | 长沙湘计海盾科技有限公司 | A kind of turntable angle high precision online measuring method based on machine vision |
-
2018
- 2018-05-31 CN CN201810552477.9A patent/CN108765433A/en active Pending
Patent Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101813469A (en) * | 2010-04-23 | 2010-08-25 | 中国科学院成都生物研究所 | Method for measuring leaf area by manually shooting |
CN102506772A (en) * | 2011-11-01 | 2012-06-20 | 西北农林科技大学 | Method and device for quickly detecting area of leaf blade based on mobile phone |
CN103530889A (en) * | 2013-11-04 | 2014-01-22 | 天津工业大学 | Method for decoding coding mark points of measuring stick based on skeleton extraction |
CN103968772A (en) * | 2014-05-19 | 2014-08-06 | 南京信息工程大学 | Piston ring detection method based on monocular vision detection |
CN104282018A (en) * | 2014-09-09 | 2015-01-14 | 苏州科力迪软件技术有限公司 | Method for on-line detection of overall diameter of industrial product based on machine vision |
CN106468543A (en) * | 2015-08-21 | 2017-03-01 | 浙江托普云农科技股份有限公司 | A kind of method for measuring leaf area based on image procossing |
CN105865344A (en) * | 2016-06-13 | 2016-08-17 | 长春工业大学 | Workpiece dimension measuring method and device based on machine vision |
CN106468536A (en) * | 2016-11-22 | 2017-03-01 | 天津市气候中心 | A kind of leaf area measurement method |
CN107154036A (en) * | 2017-03-24 | 2017-09-12 | 中南大学 | Blood vessel segmentation method and its segmenting system based on sub-pix |
CN107024187A (en) * | 2017-04-18 | 2017-08-08 | 广西壮族自治区农业科学院农业资源与环境研究所 | A kind of leaf area measurement method based on digital camera images |
CN108007388A (en) * | 2017-06-30 | 2018-05-08 | 长沙湘计海盾科技有限公司 | A kind of turntable angle high precision online measuring method based on machine vision |
CN107478656A (en) * | 2017-08-09 | 2017-12-15 | 齐鲁工业大学 | Paper pulp mixing effect method of determination and evaluation based on machine vision, device, system |
Non-Patent Citations (4)
Title |
---|
CHAOHUI LÜ: "Leaf Area Measurement Based on Image Processing", 《2010 INTERNATIONAL CONFERENCE ON MEASURING TECHNOLOGY AND MECHATRONICS AUTOMATION》 * |
DERONG ZHANG等: "A New Method and Instrument for Measurement of Plant Leaf Area", 《2010 IEEE INTERNATIONAL CONFERENCE ON INTELLIGENT SYSTEMS AND KNOWLEDGE ENGINEERING》 * |
吴大伟等: "基于图像处理技术的植物叶片相关参数测量研究", 《科技信息》 * |
谭峰等: "基于图像的植物叶面积无损测量方法研究", 《农业工程学报》 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109724569A (en) * | 2019-02-02 | 2019-05-07 | 西安卡拉光电科技有限公司 | A method of object plane size is measured using mobile phone camera |
CN110118530A (en) * | 2019-04-22 | 2019-08-13 | 西京学院 | A kind of big workpiece high-precision optical vision measuring method |
CN110363807A (en) * | 2019-06-28 | 2019-10-22 | 湖北省农业科学院经济作物研究所 | Based on the high-throughput lossless method for obtaining field cotton Efficient leaf area of RGB image |
CN110349200A (en) * | 2019-07-01 | 2019-10-18 | 西京学院 | A kind of high-precision wheat head length measurement method based on monocular camera |
CN110953993A (en) * | 2019-11-14 | 2020-04-03 | 国网山西省电力公司大同供电公司 | Detection device and method for sag and distance limit of power transmission line |
CN111223138B (en) * | 2019-12-09 | 2023-05-09 | 天津市气候中心(天津市生态气象和卫星遥感中心、天津市农业气象中心) | Blade form calibration and extraction method |
CN111223138A (en) * | 2019-12-09 | 2020-06-02 | 天津市气候中心(天津市生态气象和卫星遥感中心、天津市农业气象中心) | Method for calibrating and extracting blade form |
CN111521128A (en) * | 2020-04-15 | 2020-08-11 | 中国科学院海洋研究所 | Shellfish external form automatic measurement method based on optical projection |
CN112465893A (en) * | 2020-11-05 | 2021-03-09 | 内蒙古工业大学 | Vegetation leaf area measuring method and device, terminal and storage medium |
CN113298776A (en) * | 2021-05-21 | 2021-08-24 | 佛山职业技术学院 | Method for detecting appearance defects of metal closed water pump impeller |
CN113298776B (en) * | 2021-05-21 | 2023-01-24 | 佛山职业技术学院 | Method for detecting appearance defects of metal closed water pump impeller |
CN113624165A (en) * | 2021-08-09 | 2021-11-09 | 海南青峰生物科技有限公司 | Rice leaf area measuring device and method based on 5G and image recognition |
CN114241033A (en) * | 2021-12-17 | 2022-03-25 | 贵州省山地资源研究所 | Ectopic acquisition method for plant leaf area vertical structure |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108765433A (en) | One kind is for carrying high-precision leafy area measurement method | |
CN108764257B (en) | Multi-view pointer instrument identification method | |
CN103808263B (en) | The high-flux detection method of Grain rice shape parameter | |
Makanza et al. | High-throughput method for ear phenotyping and kernel weight estimation in maize using ear digital imaging | |
CN106468536B (en) | A kind of leaf area measurement method | |
CN108195736B (en) | Method for extracting vegetation canopy clearance rate through three-dimensional laser point cloud | |
CN111879735B (en) | Rice appearance quality detection method based on image | |
CN109859101B (en) | Crop canopy thermal infrared image identification method and system | |
CN107727010A (en) | A kind of method for measuring corps leaf surface product index | |
CN112465893A (en) | Vegetation leaf area measuring method and device, terminal and storage medium | |
CN109242787A (en) | It paints in a kind of assessment of middle and primary schools' art input method | |
CN108564021A (en) | A method of deserta cover degree is extracted based on digital photo | |
Constantino et al. | Towards an automated plant height measurement and tiller segmentation of rice crops using image processing | |
CN115687850A (en) | Method and device for calculating irrigation water demand of farmland | |
CN109118484A (en) | A method of acquisition vegetation coverage and leaf area index based on machine vision | |
CN106023235B (en) | A kind of method that the effective seed number of crops measures | |
CN105574516B (en) | The ornamental pine apple chlorophyll detection method returned based on logistic in visible images | |
CN109030488B (en) | Algae biomass detection method and device | |
CN103322946B (en) | A kind of method obtaining porosity of maize canopy | |
CN116721344A (en) | Vegetation detection method, device and equipment based on aerial photographing equipment | |
CN109166127B (en) | Wearable plant phenotype sensing system | |
CN116258844A (en) | Rapid and accurate identification method for phenotype character of cotton leaf | |
CN109409402A (en) | A kind of image contamination detection method and system based on dark channel prior histogram | |
CN107243461B (en) | Automatic grading device | |
Tu et al. | 2D in situ method for measuring plant leaf area with camera correction and background color calibration |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181106 |
|
RJ01 | Rejection of invention patent application after publication |