CN106875365A - Spill image partition method based on GSA - Google Patents
Spill image partition method based on GSA Download PDFInfo
- Publication number
- CN106875365A CN106875365A CN201710115436.9A CN201710115436A CN106875365A CN 106875365 A CN106875365 A CN 106875365A CN 201710115436 A CN201710115436 A CN 201710115436A CN 106875365 A CN106875365 A CN 106875365A
- Authority
- CN
- China
- Prior art keywords
- snake
- image
- force
- point
- pixel
- 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
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000005192 partition Methods 0.000 title claims abstract description 11
- 241000270295 Serpentes Species 0.000 claims abstract description 132
- 238000005452 bending Methods 0.000 claims abstract description 15
- 238000000205 computational method Methods 0.000 claims description 11
- 230000000694 effects Effects 0.000 claims description 10
- 238000003709 image segmentation Methods 0.000 claims description 5
- 241001074037 Virginia Species 0.000 claims description 4
- 238000006073 displacement reaction Methods 0.000 claims description 4
- 238000009827 uniform distribution Methods 0.000 claims description 4
- 241000208340 Araliaceae Species 0.000 claims 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims 1
- 235000003140 Panax quinquefolius Nutrition 0.000 claims 1
- 235000008434 ginseng Nutrition 0.000 claims 1
- 238000000605 extraction Methods 0.000 abstract description 2
- 238000001514 detection method Methods 0.000 description 4
- 238000009795 derivation Methods 0.000 description 3
- 230000007812 deficiency Effects 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
- G06T2207/20116—Active contour; Active surface; Snakes
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of spill image partition method based on GSA, gray proces are carried out to original image first, the convergence of initial snake element is then completed using GSA models, the final convergence of snake element is completed finally by greedy algorithm；The method of the present invention is normalized to image, balanced continuous force, bending force and image force, on the premise of snake element original series are ensured, allow snake element to enter rapidly and uniformly to be fitted to around objective contour, particularly it is fitted in the concave regions of image, while remaining GSA models high efficiency calculating advantage, the precision of contours extract is improve, the method for particularly being proposed has processed spill contour extraction of objects problem well.
Description
Technical field
The present invention relates to a kind of spill image partition method based on GSA
Background technology
In the research of computer vision, traditional target detection is mainly by a bottom-up process to image
Split and rim detection.This process is less to consider the feature that target has in itself, therefore the deficient validity.Snake moulds
Type is also called active contour model, proposes that the method is in the fields such as image segmentation, detection, identification by Kass et al. earliest
It is widely used.Different from traditional images dividing method, Snake models take full advantage of target image speciality in itself, figure
The information such as size, position, the shape of picture are combined with the feature such as gray scale, gradient of image.At present, Snake models have turned into figure
As the most commonly used method of target detection.
Traditional Snake models have that such as external force catching range is smaller, difficult, topological into sunk area
It is constant etc..Particularly when spill figure is processed, basic Snake models cannot generally converge to sunk area.It is above-mentioned to make up
Deficiency, some innovatory algorithms are proposed, wherein being more successfully that Xu et al. proposes GVF models, had been applied to many in succession
The practical problem of spill image segmentation, but the model spill figure larger for some curvature of caving in, effect are still undesirable.This
Outward, GVF models are to be improved to obtain by basic Snake models, the energy function of basic Snake models need with compared with
Many first derivation, second order derivation and integral operation, therefore there is a problem of that computationally intensive, computational efficiency is not high, GVF models
Equally exist this respect problem.
Greedy snake algorithm (Greedy Snake Algorithm, GSA), is that a kind of improvement is larger and more successful
Snake models.GSA algorithms have used for reference the external force of Snake models, internal force comprehensive function in the basic thought of snake vegetarian refreshments, but use
Substantial amounts of derivation need not be completed different from the energy function of traditional Snake models, the particularly energy function of GSA and integration is transported
Calculate, therefore with computational efficiency higher.In traditional GSA algorithms, the span of image force is excessive in energy function, and then
So that resultant force is almost influenceed by image force completely, the deficiency of GSA models is the profile that cannot extract spill image.
The content of the invention
The technical problem to be solved in the present invention is：For Snake model computational efficiencies it is low, treatment spill image effect pay no attention to
The problem thought, designs a kind of spill image partition method based on GSA, and the method is normalized to image, and equilibrium is even
Continuous power, bending force and image force, on the premise of snake element original series are ensured, allow snake element to enter rapidly and uniformly to be fitted to target
Around profile, particularly it is fitted in the concave regions of image.
The present invention solve technical problem technical scheme be：Gray proces are carried out to original image first, one is then used
The convergence that improved GSA models complete initial snake element is planted, the final convergence of snake element is completed finally by a kind of greedy algorithm；Including
Following specific steps：
Step 1：Gray scale pretreatment is carried out to image；
Step 2：Determine that object boundary is initial profile point in the picture, i.e. snake element；
Step 3：By GSA algorithms, by external force and internal force, deformed using the curve to being formed by connecting by snake element,
All snakes element energy sums are minimized, initial snake element is finally displaced into objective contour marginal position, and ensured between snake element
In order and be uniformly distributed, complete snake element initial convergence；
Step 4：The final convergence of snake element is completed using greedy strategy, all of snake element locks onto image target to be split
On border, the particularly sunk area of image；
Step 5：Be linked in sequence snake element, forms objective contour, completes image segmentation.
More specifically, in the step 1, gray scale pretreatment is carried out to image according to following calculating process：
The formula of gray scale pretreatment is：
Gray(v(i),j)'=(Δ1'R(v(i),j)+Δ2'G(v(i),j)+Δ3'B(v(i),j)) ＞ ＞ k
In above formula, Gray(v(i),j)Represent the gray value of certain pixel v (i), formula R(v(i),j)、G(v(i),j)、B(v(i),j)
Red, green, blue channel value, △ are represented respectively1、△2、△3Used as weight coefficient, k represents displacement coefficient, wherein, △1=
38, △2=75, △3=15, k=7.
More specifically, in the step 3, the initial convergence of snake element is completed according to following calculating process；Assuming that snake prime sequences
It is made up of N number of snake vegetarian refreshments, comprehensive force functionIt is as follows：
V (i) represents i-th snake element, and j is i snakes element neighborhood GpiJ-th interior pixel, value and a step of snake element of j are moved
It is dynamic relevant apart from δ；One moved further computational methods are：J=i ± δk, δ ∈ { 0,1,2 ... }, δ represent the moved further of snake vegetarian refreshments one away from
From；EcontRepresent continuous force, EcurvRepresent bending force, EimageRepresent image force；α, β, γ are warp parameters for controlling each group
Into the weight relationship of part；Entirely the calculating target of GSA is：In the neighborhood Gp of snake vegetarian refreshments iiIn calculate a pixel make it is comprehensive
With joint efforts E (v (i), j) value minimize, and as the snake vegetarian refreshments of next moved further；
EcontThe computational methods of continuous force are as follows：
Wherein, dv(i,j)-v(i+1)Represent i-th snake element neighborhood GpiIn j-th pixel straight line plain with i+1 snake away from
From；davrThe average distance between all snake vegetarian refreshments is represented, computational methods areContinuous force Econt
Calculating is aimed at：It was found that GpiIn with i+1 snakes element distance closest to average distance a pixel, effect be control
Uniform distribution of forces of the snake vegetarian refreshments in profile space；D after each round iterationavrNeeds are recalculated；
Bending force EcurvEnergy function it is as follows:
In above formula, | v (i-1) -2v (i, j)+v (i+1) | is represented to i-1, i+1 snake element and GpiIn pixel j carry out and
Vector is calculated；Bending force EcurvCalculating aim at：It was found that GpiIn with former and later two snake vegetarian refreshments to constitute vector angle minimum
Pixel, effect is the smoothness of snake vegetarian refreshments；
The specific energy function E of image forceimageIt is as follows:
(v (i) j) represents Gp to wherein IiIn j-th Grad of pixel；(v (i) j) represents Gp to miniComprising picture
The minimal gradient value of element；Image force EimageCalculating aim at：It was found that GpiMiddle Grad maximum pixel point.
More specifically, in the step 4, the final convergence of snake element is completed by greedy algorithm according to following calculating process：
Step (4.1):Initial profile is regarded as a N sides shape, then the center of the N sides shape is registered as reference point p(x,y)；
If reference point is in contour area, step (4.4) is jumped to；If reference point is outside contour area, order performs step
(4.2)；
Step (4.2):With reference point p(x,y)It is end points, N bars ray L (i), i=1 is constructed by each snake element v (i),
2,…,N；
Step (4.3):The ray L intersected with profile for everyp, p=1,2,3 ... n find its intersecting point coordinate and according to
Secondary record, then this N number of data point be expressed as p(xi,yi)(i=1,2,3 ..., n), and calculate snake vegetarian refreshments and wheel on every ray
The middle point coordinates M of wide intersection point(xi,yi)(i=1,2,3 ..., n)；
The middle point coordinates M being made up of this N/2(xi,yi)(i=1,2,3 ..., n) calculating last reference point is：
Step (4.4):With reference point p(x,y)As this current snakelike model energy function EimageCenter, construct n bars
Ray Lj(j=1,2 ..., n), from current reference point p(x,y)It is issued to the contour edge after this iteration i times；Now, two it is adjacent
The angle of straight line is 2 π/n；
Step (4.5):With reference point p(x,y)As this current snakelike model energy function EimageCenter, construct n bars
Ray Lj(j=1,2 ..., n), from current reference point p(x,y)It is issued to the contour edge after this iteration i times；Now, two it is adjacent
The angle of straight line is 2 π/n；
Step (4.6):For every straight line Lj(j=1,2 ..., n), we define Pv(i,j)It is this straight line LjOn i-th
Point, calculates the gradient G of each pointv(i,j), choose maximum of gradients max { G on this linev(i,j)As convergent maximal end point, i.e.,
The final position of snake element.
The present invention has advantages below：
(1) in some image segmentation problems, original image often has that spot is coarse, noise；The present invention
Gray scale pretreatment cause that the feature of original image is more obvious, gray proces make image border be strengthened, denoising, sharpening
Degree is also correspondingly improved.
(2) during whole energy function carries out minimization, continuous force energy function is uniformly distributed in snake element ought
Preceding objective contour, bending force energy function makes its profile smooth as far as possible, and image force makes snake element rest on high gradient as much as possible
Pixel；Three power are acted on simultaneously, and snake element is constantly moved in the presence of external force to true specific profile, and internal force is being protected
Hold snake element topological it is preferable while constantly vary with the movement of snake vegetarian refreshments, final internal force and external force mutually restrict to be formed
One preferable initial profile.
(3) computing formula of original GSA algorithm synthesis power is that the plus-minus mixing of three power is calculated, by improving Eimage's
Computational methods so that EimageThe span of power (0,1], three power are added and cause Econt、EcurvAnd EimageThree power are more
Balance, such change makes comprehensive force function controllability more preferable.
(4) greedy algorithm take into account reference point in the inner and outer two kinds of situations of initial profile, according to the gradient of image itself
Feature makes snake element finally converge to recessed profile zone.
(5) present invention makes improvement on the basis of original GSA models, and introduces greedy algorithm, gray proces etc.
Strategy, the method for being proposed improves the precision of contours extract while GSA models high efficiency calculating advantage is remained, special
It is not that proposed method processes spill contour extraction of objects problem well.
Brief description of the drawings
Fig. 1 is overall algorithm flow of the invention；
Fig. 2 is the neighborhood of snake element；
Fig. 3 is the overall procedure of greedy algorithm；
Fig. 4 is that non-recessed image reference point is chosen；
Fig. 5 is that spill image reference point is chosen.
Specific embodiment
With reference to the accompanying drawings and examples, technical scheme is described in detail, but is should not be understood as
Limitation to technical scheme.In the following description, it is more thorough to the present invention to provide to give a large amount of concrete details
Understanding.However, to those skilled in the art, the present invention can be able to reality without one or more of these details
Apply.In other examples, in order to avoid obscuring with the present invention, do not carried out for some technical characteristics well known in the art
Description.
Fig. 1 gives the overall procedure of multiple target point path planning method of the invention；Fig. 1 is referred to, here is other side
The detailed description of each step in method：
Step S101：Gray scale pretreatment is carried out to whole image, it is therefore an objective to so that the feature of original image is more obvious；
The formula of gray scale pretreatment is：
Gray(v(i),j)'=(Δ1'R(v(i),j)+Δ2'G(v(i),j)+Δ3'B(v(i),j)) ＞ ＞ k (1)
In above formula, Gray(v(i),j)Represent the gray value of certain pixel v (i), formula R(v(i),j)、G(v(i),j)、B(v(i),j)
Red, green, blue channel value, △ are represented respectively1、△2、△3Used as weight coefficient, k represents displacement coefficient；Wherein, △1=
38, △2=75, △3=15, k=7, compare by test of many times, and display above-mentioned parameter allocative efficiency is higher；
Step S102：It is N to set snake element scale, and determines the initial position of each snake element, total iteration time of set algorithm
Number M, and current iteration number of times is recorded with iterations accumulator m, m is initialized as 1；
Step S103：The value for setting snake element counter i is 1, represents and is calculated since the snake element that numbering is 1；
Step S104：As shown in Fig. 2 the position according to i snakes element 100 obtains the (Gp of neighborhood 200i), 9 pictures are had in neighborhood
Element 300, calculates continuous force, bending force and the image force of each pixel 300, and then calculate the resultant force of each pixel；
EcontThe computational methods of continuous force are as follows：
Wherein, dv(i,j)-v(i+1)Represent i-th snake element neighborhood GpiIn j-th pixel straight line plain with i+1 snake away from
From；davrThe average distance between all snake vegetarian refreshments is represented, computational methods areContinuous force Econt
Main calculating aims at：It was found that GpiIn with i+1 snakes element distance closest to average distance a pixel, effect is
Uniform distribution of forces of the snake vegetarian refreshments in profile space；D after each round iterationavrNeeds are recalculated；
Bending force EcurvEnergy function it is as follows:
In above formula, | v (i-1) -2v (i, j)+v (i+1) | is represented to i-1, i+1 snake element and GpiIn pixel j carry out and
Vector is calculated；Bending force EcurvCalculating aim at：It was found that GpiIn with former and later two snake vegetarian refreshments to constitute vector angle minimum
Pixel, effect is the smoothness of snake vegetarian refreshments；
The specific energy function E of image forceimageIt is as follows:
(v (i) j) represents Gp to wherein IiIn j-th Grad of pixel；(v (i) j) represents Gp to miniComprising picture
The minimal gradient value of element；Image force EimageCalculating aim at：It was found that GpiMiddle Grad maximum pixel point；
Comprehensive force functionIt is as follows：
Step S105：The minimum pixel of selection resultant force replaces current i snakes element 100；
Step S106：Snake element counter adds one, represents that algorithm sequentially calculates next snake element；
Step S107：If i≤N, algorithm jumps to S104；Otherwise, algorithm performs S108, now from snake element 1 to snake
Plain N is calculated one time；
Step S108：Iterations accumulator m adds 1；
Step S109：If iterations is not reaching to total iterations, step S103 is jumped to；Otherwise, step is performed
Rapid S110；
Step S110：The final convergence of snake element is completed according to greedy strategy；
Step S111：Snake element is sequentially connected, final image object profile is formed.
Fig. 3 gives the overall procedure of step S110 greedy algorithms；Fig. 3 is referred to, here is to each step in algorithm
Detailed description；
Step S201:As shown in figure 4, initial profile is regarded as into a N sides shape, then the center of the N sides shape is referred to as reference
Point 400, is recorded as：p(x,y)；
Step S202:If reference point is in contour area, step S206 is jumped to；If reference point is outside contour area,
Then order performs step S203；
Step S203:As shown in figure 5, with reference point 400 as end points, N bar rays are constructed by each snake element v (i)
500, it is recorded as：L (i), i=1,2 ..., N；
Step S204:The ray L intersected with profile for everyp, p=1,2,3 ... n find its intersecting point coordinate and according to
Secondary record, then this N number of data point be expressed as p(xi,yi)(i=1,2,3 ..., n), and calculate snake element and profile on every ray
The middle point coordinates 600 of intersection point, is recorded as：M(xi,yi)(i=1,2,3 ..., n)；
Step S205:The middle point coordinates M being made up of this N/2(xi,yi)(i=1,2,3 ..., n) calculate last reference point
For：
S206:With reference point p(x,y)As this current snakelike model energy function EimageCenter, construct n bar rays Lj
(j=1,2 ..., n), from current reference point p(x,y)It is issued to the contour edge after this iteration i times；Now, two adjacent straight lines
Angle is 2 π/n；
S207:For every straight line Lj(j=1,2 ..., n), we define Pv(i,j)It is this straight line LjOn i-th point, meter
Calculate the gradient G of each pointv(i,j), choose maximum of gradients max { G on this linev(i,j)As convergent maximal end point, i.e. snake element
Final position.
Claims (6)
1. the spill image partition method of GSA is based on, it is characterized in that：Gray proces are carried out to original image first, is then used
GSA models complete the convergence of initial snake element, and the final convergence of snake element is completed finally by greedy algorithm；It comprises the following steps：
Step 1：Gray scale pretreatment is carried out to image；
Step 2：Determine that object boundary is initial profile point in the picture, i.e. snake element；
Step 3：By GSA algorithms, by external force and internal force, deform using to the curve being formed by connecting by snake element, make institute
There is snake element energy sum to minimize, initial snake element is finally displaced into objective contour marginal position, and ensure that snake is orderly between plain
And be uniformly distributed, complete the initial convergence of snake element；
Step 4：The final convergence of snake element is completed using greedy strategy, all of snake element locks onto the border of image target to be split
On, the particularly sunk area of image；
Step 5：Be linked in sequence snake element, forms objective contour, completes image segmentation.
2. the spill image partition method based on GSA according to claim 1, it is characterized in that：In the step 1, according to
Following calculating process carry out gray scale pretreatment to image：
The formula of gray scale pretreatment is：
Gray(v(i),j)'=(Δ1'R(v(i),j)+Δ2'G(v(i),j)+Δ3'B(v(i),j)) ＞ ＞ k
In above formula, Gray(v(i),j)Represent the gray value of certain pixel v (i), formula R(v(i),j)、G(v(i),j)、B(v(i),j)Respectively
Represent red, green, blue channel value, △1、△2、△3Used as weight coefficient, k represents displacement coefficient, wherein, △1=38, △2
=75, △3=15, k=7.
3. the spill image partition method based on GSA according to claim 1, it is characterized in that：In the step 3, according to
Following calculating process complete the initial convergence of snake element；Assuming that snake prime sequences are made up of N number of snake vegetarian refreshments, comprehensive force functionIt is as follows：
V (i) represents i-th snake element, and j is i snakes element neighborhood GpiJ-th interior pixel, the value of j and a moved further distance of snake element
δ is relevant；One moved further computational methods are：J=i ± δk, δ ∈ 0,1,2 ... }, δ represents the moved further distance of snake vegetarian refreshments one；Econt
Represent continuous force, EcurvRepresent bending force, EimageRepresent image force；α, β, γ are warp parameters for controlling each part
Weight relationship；Entirely the calculating target of GSA is：In the neighborhood Gp of snake vegetarian refreshments iiIn calculate a pixel and make resultant force E (v
(i), j) value minimum, and as the snake vegetarian refreshments of next moved further；
EcontThe computational methods of continuous force are as follows：
Wherein, dv(i,j)-v(i+1)Represent i-th snake element neighborhood GpiIn j-th pixel air line distance plain with i+1 snake；
davrThe average distance between all snake vegetarian refreshments is represented, computational methods areContinuous force EcontMeter
Aim at：It was found that GpiIn with i+1 snakes element distance closest to average distance a pixel, effect be control snake
Uniform distribution of forces of the vegetarian refreshments in profile space；D after each round iterationavrNeeds are recalculated；
Bending force EcurvEnergy function it is as follows:
In above formula, | v (i-1) -2v (i, j)+v (i+1) | is represented to i-1, i+1 snake element and GpiIn pixel j carry out and vector
Calculate；Bending force EcurvCalculating aim at：It was found that GpiIn constitute the minimum pixel of vector angle with former and later two snake vegetarian refreshments
Point, effect is the smoothness of snake vegetarian refreshments；
The specific energy function E of image forceimageIt is as follows:
(v (i) j) represents Gp to wherein IiIn j-th Grad of pixel；(v (i) j) represents Gp to miniComprising pixel
Minimal gradient value；Image force EimageCalculating aim at：It was found that GpiMiddle Grad maximum pixel point.
4. the spill image partition method based on GSA according to claim 1, it is characterized in that：In the step 4, according to
Following calculating process complete the final convergence of snake element by greedy algorithm：
Step (4.1):Initial profile is regarded as a N sides shape, then the center of the N sides shape is registered as reference point p(x,y)；If ginseng
Examination point jumps to step (4.4) in contour area；If reference point is outside contour area, order performs step (4.2)；
Step (4.2):With reference point p(x,y)It is end points, N bars ray L (i), i=1,2 is constructed by each snake element v (i) ...,
N；
Step (4.3):The ray L intersected with profile for everyp, p=1,2,3 ... n find its intersecting point coordinate and remember successively
Record, then this N number of data point is expressed as p(xi,yi)(i=1,2,3 ..., n), and calculate snake vegetarian refreshments and profile friendship on every ray
The middle point coordinates M of point(xi,yi)(i=1,2,3 ..., n)；
The middle point coordinates M being made up of this N/2(xi,yi)(i=1,2,3 ..., n) calculating last reference point is：
Step (4.4):With reference point p(x,y)As this current snakelike model energy function EimageCenter, construct n bar rays
Lj(j=1,2 ..., n), from current reference point p(x,y)It is issued to the contour edge after this iteration i times；Now, two adjacent straight lines
Angle be 2 π/n；
Step (4.5):With reference point p(x,y)As this current snakelike model energy function EimageCenter, construct n bar rays
Lj(j=1,2 ..., n), from current reference point p(x,y)It is issued to the contour edge after this iteration i times；Now, two adjacent straight lines
Angle be 2 π/n；
Step (4.6):For every straight line Lj(j=1,2 ..., n), we define Pv(i,j)It is this straight line LjOn i-th point,
Calculate the gradient G of each pointv(i,j), choose maximum of gradients max { G on this linev(i,j)As convergent maximal end point, i.e. snake element
Final position.
5. the spill image partition method based on GSA according to claim 1, it is characterized in that it includes following specific step
Suddenly：
Step S101：Gray scale pretreatment is carried out to whole image, it is therefore an objective to so that the feature of original image is more obvious；
The formula of gray scale pretreatment is：
Gray(v(i),j)'=(Δ1'R(v(i),j)+Δ2'G(v(i),j)+Δ3'B(v(i),j)) ＞ ＞ k (1)
In above formula, Gray(v(i),j)Represent the gray value of certain pixel v (i), formula R(v(i),j)、G(v(i),j)、B(v(i),j)Respectively
Represent red, green, blue channel value, △1、△2、△3Used as weight coefficient, k represents displacement coefficient；Wherein, △1=38, △2
=75, △3=15, k=7；
Step S102：It is N to set snake element scale, and determines the initial position of each snake element, total iterations M of set algorithm,
And current iteration number of times is recorded with iterations accumulator m, m is initialized as 1；
Step S103：The value for setting snake element counter i is 1, represents and is calculated since the snake element that numbering is 1；
Step S104：Position according to i snakes element 100 obtains the (Gp of neighborhood 200i), 9 pixels 300 are had in neighborhood, calculate every
The continuous force of individual pixel 300, bending force and image force, and then calculate the resultant force of each pixel；
EcontThe computational methods of continuous force are as follows：
Wherein, dv(i,j)-v(i+1)Represent i-th snake element neighborhood GpiIn j-th pixel air line distance plain with i+1 snake；
davrThe average distance between all snake vegetarian refreshments is represented, computational methods areContinuous force EcontIt is main
Calculate and aim at：It was found that GpiIn with i+1 snakes element distance closest to average distance a pixel, effect is snake
Uniform distribution of forces of the vegetarian refreshments in profile space；D after each round iterationavrNeeds are recalculated；
Bending force EcurvEnergy function it is as follows:
In above formula, | v (i-1) -2v (i, j)+v (i+1) | is represented to i-1, i+1 snake element and GpiIn pixel j carry out and vector
Calculate；Bending force EcurvCalculating aim at：It was found that GpiIn constitute the minimum pixel of vector angle with former and later two snake vegetarian refreshments
Point, effect is the smoothness of snake vegetarian refreshments；
The specific energy function E of image forceimageIt is as follows:
(v (i) j) represents Gp to wherein IiIn j-th Grad of pixel；(v (i) j) represents Gp to miniComprising pixel
Minimal gradient value；Image force EimageCalculating aim at：It was found that GpiMiddle Grad maximum pixel point；
Comprehensive force functionIt is as follows：
Step S105：The minimum pixel of selection resultant force replaces current i snakes element 100；
Step S106：Snake element counter adds one, represents that algorithm sequentially calculates next snake element；
Step S107：If i≤N, algorithm jumps to S104；Otherwise, algorithm performs S108, now element N quilts from snake element 1 to snake
Calculate one time；
Step S108：Iterations accumulator m adds 1；
Step S109：If iterations is not reaching to total iterations, step S103 is jumped to；Otherwise, step is performed
S110；
Step S110：The final convergence of snake element is completed according to greedy strategy；
Step S111：Snake element is sequentially connected, final image object profile is formed.
6. the spill image partition method based on GSA according to claim 5, it is characterized in that：Step S110 greedy algorithms
Overall procedure in each step it is as follows：
Step S201:Initial profile is regarded as a N sides shape, then the center of the N sides shape is referred to as reference point 400, is recorded as：
p(x,y)；
Step S202:If reference point is in contour area, step S206 is jumped to；It is suitable if reference point is outside contour area
Sequence performs step S203；
Step S203:With reference point 400 as end points, N bars ray 500 is constructed by each snake element v (i), be recorded as：L (i), i
=1,2 ..., N；
Step S204:The ray L intersected with profile for everyp, p=1,2,3 ... n find its intersecting point coordinate and remember successively
Record, then this N number of data point is expressed as p(xi,yi)(i=1,2,3 ..., n), and calculate snake element and profile intersection point on every ray
Middle point coordinates 600, be recorded as：M(xi,yi)(i=1,2,3 ..., n)；
Step S205:The middle point coordinates M being made up of this N/2(xi,yi)(i=1,2,3 ..., n) calculating last reference point is：
S206:With reference point p(x,y)As this current snakelike model energy function EimageCenter, construct n bar rays Lj(j=
1,2 ..., n), from current reference point p(x,y)It is issued to the contour edge after this iteration i times；Now, two angles of adjacent straight line
It is 2 π/n；
S207:For every straight line Lj(j=1,2 ..., n), we define Pv(i,j)It is this straight line LjOn i-th point, calculate every
The gradient G of individual pointv(i,j), choose maximum of gradients max { G on this linev(i,j)Used as convergent maximal end point, i.e., snake element is final
Position.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710115436.9A CN106875365B (en) | 2017-03-01 | 2017-03-01 | Spill image partition method based on GSA |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710115436.9A CN106875365B (en) | 2017-03-01 | 2017-03-01 | Spill image partition method based on GSA |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106875365A true CN106875365A (en) | 2017-06-20 |
| CN106875365B CN106875365B (en) | 2019-06-11 |
Family
ID=59168245
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710115436.9A Active CN106875365B (en) | 2017-03-01 | 2017-03-01 | Spill image partition method based on GSA |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106875365B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109489554A (en) * | 2018-12-29 | 2019-03-19 | 浙江科技学院 | A kind of each layer parameter intelligent detecting method of full automatic cable and device |
| CN113077484A (en) * | 2021-03-30 | 2021-07-06 | 中国人民解放军战略支援部队信息工程大学 | Image instance segmentation method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6484099B1 (en) * | 1999-07-16 | 2002-11-19 | Deutsches Zentrum Fur Luft -Und Raumfahrt E.V. | Process for correcting atmospheric influences in multispectral optical remote sensing data |
| CN101408983A (en) * | 2008-10-29 | 2009-04-15 | 南京邮电大学 | Multi-object tracking method based on particle filtering and movable contour model |
| CN104680498A (en) * | 2015-03-24 | 2015-06-03 | 江南大学 | Medical image segmentation method based on improved gradient vector flow model |
-
2017
- 2017-03-01 CN CN201710115436.9A patent/CN106875365B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6484099B1 (en) * | 1999-07-16 | 2002-11-19 | Deutsches Zentrum Fur Luft -Und Raumfahrt E.V. | Process for correcting atmospheric influences in multispectral optical remote sensing data |
| CN101408983A (en) * | 2008-10-29 | 2009-04-15 | 南京邮电大学 | Multi-object tracking method based on particle filtering and movable contour model |
| CN104680498A (en) * | 2015-03-24 | 2015-06-03 | 江南大学 | Medical image segmentation method based on improved gradient vector flow model |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109489554A (en) * | 2018-12-29 | 2019-03-19 | 浙江科技学院 | A kind of each layer parameter intelligent detecting method of full automatic cable and device |
| CN109489554B (en) * | 2018-12-29 | 2024-02-02 | 浙江科技学院 | Intelligent detection method and device for parameters of all layers of full-automatic cable |
| CN113077484A (en) * | 2021-03-30 | 2021-07-06 | 中国人民解放军战略支援部队信息工程大学 | Image instance segmentation method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106875365B (en) | 2019-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107146228B (en) | A method for supervoxel generation of brain magnetic resonance images based on prior knowledge | |
| CN106683139A (en) | Fisheye-camera calibration system based on genetic algorithm and image distortion correction method thereof | |
| CN103353987A (en) | Superpixel segmentation method based on fuzzy theory | |
| CN107578430B (en) | Stereo matching method based on self-adaptive weight and local entropy | |
| CN106651897B (en) | Parallax correction method based on super-pixel segmentation | |
| CN102663367B (en) | Three-dimensional face identification method on basis of simulated annealing algorithm | |
| CN108876769B (en) | A method for segmentation of left atrial appendage CT images | |
| CN102354396A (en) | Method for segmenting image with non-uniform gray scale based on level set function | |
| CN110853064B (en) | An Image Cooperative Segmentation Method Based on Minimum Fuzzy Divergence | |
| CN113409332B (en) | Building plane segmentation method based on three-dimensional point cloud | |
| CN102446358A (en) | Multi-modal medical image registration method based on edge features and CS information | |
| CN108022261B (en) | Non-rigid image registration method based on improved optical flow field model | |
| CN105046643A (en) | Image splicing method based on brightness adaptive registration | |
| CN106548478B (en) | Active contour image partition method based on local fit image | |
| CN104143186A (en) | A SLIC Superpixel Block Optimization Method | |
| CN103955945A (en) | Self-adaption color image segmentation method based on binocular parallax and movable outline | |
| CN107871321A (en) | Image segmentation method and device | |
| CN111429447A (en) | Focal region detection method, device, equipment and storage medium | |
| CN106875365A (en) | Spill image partition method based on GSA | |
| US9251600B2 (en) | Method and apparatus for determining an alpha value | |
| CN106355607A (en) | Wide-baseline color image template matching method | |
| CN104899592A (en) | Road semi-automatic extraction method and system based on circular template | |
| CN107194351A (en) | Face recognition features' extraction algorithm based on weber Local Symmetric graph structure | |
| CN110728718B (en) | Method for improving camera calibration parameters | |
| TW201128568A (en) | Image processing method for feature retention and the system of the same |
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 | ||
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20170620 Assignee: Huaian Jinyu Technology Co.,Ltd. Assignor: Jiangsu vocationnal college of electronics and information Contract record no.: X2020320000067 Denomination of invention: Concave image segmentation method based on GSA Granted publication date: 20190611 License type: Common License Record date: 20200828 |
|
| CP01 | Change in the name or title of a patent holder |
Address after: No.3, Meicheng Road, higher education park, Huai'an City, Jiangsu Province 223005 Patentee after: Jiangsu electronic information Vocational College Address before: No.3, Meicheng Road, higher education park, Huai'an City, Jiangsu Province 223005 Patentee before: Jiangsu vocationnal college of electronics and information |
|
| CP01 | Change in the name or title of a patent holder | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230420 Address after: Room 2920, Building B, No. 9 Hisense Tianchen Road, Jinan City, Shandong Province, 250000 Patentee after: Zhuofan (Jinan) Technology Innovation Co.,Ltd. Address before: No.3, Meicheng Road, higher education park, Huai'an City, Jiangsu Province 223005 Patentee before: Jiangsu electronic information Vocational College |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230621 Address after: 3427, 4th Floor, Building 3, Yard 33, Zique Road, Haidian District, Beijing, 100000 Patentee after: Beijing Lingyi times Communication Technology Co.,Ltd. Address before: Room 2920, Building B, No. 9 Hisense Tianchen Road, Jinan City, Shandong Province, 250000 Patentee before: Zhuofan (Jinan) Technology Innovation Co.,Ltd. |
|
| TR01 | Transfer of patent right |