CN104008545B - One kind human oocytes in optical microphotograph injection process split implementation method - Google Patents
One kind human oocytes in optical microphotograph injection process split implementation method Download PDFInfo
- Publication number
- CN104008545B CN104008545B CN201410214660.XA CN201410214660A CN104008545B CN 104008545 B CN104008545 B CN 104008545B CN 201410214660 A CN201410214660 A CN 201410214660A CN 104008545 B CN104008545 B CN 104008545B
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- munderover
- mfrac
- msubsup
- 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.)
- Expired - Fee Related
Links
Landscapes
- Image Analysis (AREA)
- Peptides Or Proteins (AREA)
Abstract
The invention discloses one kind, human oocytes split implementation method in optical microphotograph injection process, comprise the following steps：1) according to the human oocytes shape difference opposite sex, using matrix morphology operator extraction interesting target ROI；2) improved fuzzy clustering method is used, rough sort is carried out to the human oocytes of obscurity boundary and noise according to local gray level and local Variance feature；3) level set initialized location and Evolution Control parameter are set on the basis of fuzzy clustering, realizes the segmentation of cytoplasm, polar body and injection needle；4) intracytoplasmic partial syringe pin image is directed to, using being directly fitted, marks complete injection needle image.During ICSI is Intracytoplasmic Sperm Injection, the segmentation problem of the cytoplasm of human oocytes, polar body and injection needle can be effectively realized with the inventive method, contribute to automation to realize the raising with success rate, and reduce to manually-operated highly difficult requirement and dependence.
Description
Technical field
The present invention relates to Medical Image Processing and applied technical field, refers in particular to one kind in optical microphotograph injection process
Human oocytes split implementation method.
Background technology
Cell microinjection is the important technical in biological gene engineering.The present invention is mainly studied in ICSI (Intra
Cytoplasmic Sperm Injection) it is the cell of human egg during Intracytoplasmic Sperm Injection
The segmentation problem of matter, polar body and injection needle.Cytoplasm and polar body are partitioned into from egg mother cell, is not only able to help analysis expert
Oocyte maturation degree (such as diameter, granularity, the thickness of oolemma, polar body diameter, perivitelline size), and contribute to
ICSI automations are realized and the raising of success rate, and reduce to manually-operated highly difficult requirement and dependence.
The scene of ICSI injections is as shown in Figure 6.It is broadly divided into egg cell, injection needle and absorption three objects of pin.Wherein, ovum
Cell periphery fuzzy region is corona radiata, and corona radiata inner side black curve band is oolemma, and the material that oolemma is surrounded is thin
Kytoplasm, cytoplasmic upper rounded particulate material are the polar body of cell；The absorption pin in left side is used for adsorbing and fixing egg cell；
The injection needle of the long fine strip shape in the right realizes the transport and injection of sperm.It is thin to be necessary to ensure that injection needle is correctly pierced into during ICSI
The position of born of the same parents, while the polar body of cell is not damaged, and judge whether really to pierce through cytoplasm.So realize above-mentioned segmentation most
It is important.
Find that the automation for realizing cell microinjection in recent years is current emphasis through the literature search to prior art
Research direction.In the research and development of experimental provision, have scholar both at home and abroad and deploy basic research work Ladjal H et al. in paper
《Micro-to-Nano biomechanical modeling for assisted biological cell injection》
(IEEE Transactions on Biomedical Engineering,2013,(60)9:2461-2471), and Tian Guizhong
Et al. in paper《Cell pose regulation technology and experimental study in microinjection》(China Mechanical Engineering, 2009,20 (4):500-
503) in, the injection platform based on optical microphotograph built is described respectively.
And in terms of the selection of egg mother cell, Manna et al. is in paper《Artificial intelligence
techniques for embryo and oocyte classification》(Reproductive biomedicine
online,2013,26(1):In 42-49), the classification and selection to egg mother cell are realized using textural characteristics, in next step
ICSI improves success rate and prepared.Basile et al. is in paper《Atexture-based image processing
approach for the description of human oocyte cytoplasm》(Instrumentation and
Measurement,IEEE Transactions on,2010,59(10):In 2591-2601), by image preprocessing from thin
Kytoplasm center hews out one piece of rectangular area, and multiple Statistic Textures of egg mother cell are extracted after Har wavelet transformations, then uses
FCM (Fuzzy C-Means) clusters realize the egg mother cell classification of different degrees of granularity.Caponetti et al. is being discussed respectively
Text《Multiresolution texture analysis for human oocyte cytoplasm description》
(Medical Measurements and Applications,2009.MeMeA2009.IEEE International
Workshop on.IEEE,2009:150-155) and paper《Fuzzy mathematical morphology for
biological image segmentation》(Applied Intelligence.2014:In 1-11), fuzzy mathematics is first used
Method realizes segmentation, and realizes that egg mother cell is classified with reference to textural characteristics.But Basile and Caponetti pretreatment is used
Hough transform realizes the pre-segmentation of round cell matter, can not be applied to the situation of cells deformation, and is not partitioned into cell
Polar body.Et al. in paper《Elevated active contour with global image energy based
on electrostatic force》(Zeszyty Naukowe Politechniki
Informatyka,2010:A kind of global active contour snake model of basic exterior static power driving is proposed in 5-21), although
Cytoplasmic profile has been partitioned into, but has wanted manual initiation evolution profile, and has been not carried out the segmentation of cell polar body.
Motion control, Zhang et al. for injection needle is in paper《Controlled aspiration and
positioning of biological cells in a micropipette》(IEEE Transactions on
Biomedical Engineering,2012,59(4):In 1032-1040), it have studied based on micro-vision and injection needle moved
Closed loop control method, while realize the positioning and tracking of the sperm motility in injection needle.
In summary, the signature analysis of ICSI various experimental provisions and egg mother cell has attracted the concern of scholar, still
Segmentation and orientation problem rare research for egg mother cell during cell microinjection, and this is cell microinjection
Automate key point.Therefore cytoplasm, injection needle and the segmentation of egg cell polar body during this paper primary studies ICSI.
The content of the invention
It is an object of the invention to overcome the deficiencies in the prior art and defect, there is provided one kind is effective, reliably shows in optics
Human oocytes split implementation method during microinjection, can realize that human oocytes are split under the conditions of common hardware.
To achieve the above object, technical scheme provided by the present invention is：One kind mankind in optical microphotograph injection process
Egg mother cell splits implementation method, comprises the following steps：
1) according to the human oocytes shape difference opposite sex, using matrix morphology operator extraction interesting target ROI；
2) use improved fuzzy clustering method, according to local gray level and local Variance feature to obscurity boundary and noise compared with
Big pretreatment image carries out rough sort；
3) level set initialized location and Evolution Control parameter are set on the basis of fuzzy clustering, realizes cytoplasm, pole
The segmentation of body and injection needle；
4) intracytoplasmic partial syringe pin image is directed to, using fitting a straight line, marks complete injection needle image.
In step 1), according to the human oocytes shape difference opposite sex, using matrix morphology operator extraction mesh interested
ROI is marked, its process is：First, complete bianry image is obtained with OTSU methods, then does dilation operation, then through two-value skeletonizing,
The profile of cytoplasm, the polar body of egg mother cell and absorption pin is intactly extracted, its profile width is a pixel, while will absorption
Pin skeleton turns to single pixel straight line, and it is connection that the image of skeletonizing, which only has cytoplasm, is easy to fill and marks；Secondly, pass through
Area filling and mark, then using regular filter, the cytoplasm of egg mother cell is divided from the object background on periphery and opened out
Come；Finally, interesting target ROI is extracted.
In step 2), the improved fuzzy clustering method is to add local gray level and local Variance feature so that when
The classification of preceding central point pixel is together decided on by the key words sorting of the pixel in its neighbour domain, so as to realize cytoplasm, polar body and
The rough sort of endocellular injection pin, its process are as follows：
Assuming that the real image y collected is made up of true picture x and offset-domain b, and b is slowly to become in subrange
Change, image y, which amounts to, N number of pixel, then has：
So, the FCM objective cost functions rectification to standard is as follows：
Wherein, p is fuzzy membership Weighted Index, influences the fog-level of final classification；NkRepresent with xkCentered on it is near
Neighborhood；NRRepresent NkThe number of middle pixel；α is the control parameter in neighbour domain, and value is inversely proportional to the signal to noise ratio of image, by above formula (1)
Substitution has：
The optimal problem of above formula is sought using Lagrange multipliers：
Wherein, Dik=| | yk-bk-vi||2,
Respectively to uik、vikIt is 0 to seek first-order partial derivative and make result, can obtain degree of membershipAnd cluster centreSimilarly, it is right
Offset-domain bkEstimation, it is 0 to seek first-order partial derivative and make its result, can be in the hope of：
In step 3), with reference to fuzzy clustering result, realize the initialization of level set position and set level set movements to join
Number, direction and the speed parameter that the result of fuzzy clustering is used as developing without the variation level set of initialization so that in profile
Border can automatically determine evolution direction, accurately track boundary position, so as to realize the correct segmentation of target；Variation equation is such as
Under：
Wherein, φ is level set function, and three, the right is respectively penalty term, zero level collection length of curve regular terms and acceleration
The area item of curve evolvement；Coefficient is respectively μ, λ and ν；G is edge indicator function, and δ (φ) is single argument Dirac functions, and H is
Odd symmetry Heaviside functions；By the Euler-Lagrange calculus of variations, energy in the gradient descent method solving model of standard is used
Flow function minimizes；Obtain level set Evolution Equation be：
Wherein, the positive and negative decision initial profile of coefficient ν values does contraction or expansion evolution relative to target to be split, and
The speed of evolution, for actual segmentation problem, when it is expected away from object boundary, Evolution Rates are fast, conversely, it is slow to develop；Work as profile
During across object boundary, evolution direction can be changed automatically, rather than it is unalterable；Therefore, adaptive change coefficient ν values are more
It is adapted to actual segmentation demand, make the image of fuzzy clustering degree of membership so has for R：
νk=G (Rk)=1-2Rk,Rk∈(0,1) (10)
The Evolution Equation of level set is updated to：
The selection of other parameters：
Level set is initialized and its developed as follows：
It is assumed that the image of selected fuzzy clustering degree of membership is R, an adjustable threshold value b is set0∈ (0,1), obtains one
Individual bianry image Bk：
So, level set initialization function, which reaches, is：
φ0=-4 ε (0.5-Bk), ε=1.5 (13).
The present invention compared with prior art, has the following advantages that and beneficial effect：
1st, during ICSI is Intracytoplasmic Sperm Injection, mankind's ovum mother can effectively be realized with the inventive method
The segmentation problem of the cytoplasm of cell, polar body and injection needle, contribute to automation to realize and the raising of success rate, and reduce to people
The highly difficult requirement and dependence of work operation；
2nd, the problem of traditional FCM is to noise-sensitive is improved using improved fuzzy clustering algorithm, adds local gray level drawn game
Portion's Variance feature so that the classification of Current central point pixel is together decided on by the key words sorting of the pixel in its neighbour domain, so as to
The rough sort of cytoplasm, polar body and endocellular injection pin is realized, internal substantially smooth, seldom particle and empty bubble, border is relative
Clearly；
3rd, the result that fuzzy clustering obtains can be used for the initial position that lead-out level collection develops, and reduce curve in-depth computing
Amount, segmentation precision is improved, direction and the speed parameter that the result of cluster is used as developing without the variation level set of initialization, made
Evolution direction can be automatically determined on profile border by obtaining, and accurately track boundary position.
Brief description of the drawings
Fig. 1 is the flow chart that human oocytes of the present invention split implementation method.
Fig. 2 a are the original image before image preprocessing.
Skeletal extraction figure when Fig. 2 b are image preprocessing.
Fig. 2 c are the pre-segmentation result figure after image preprocessing.
Fig. 3 a are that ROI image uses the pre-segmentation image before improved fuzzy clustering method.
Fig. 3 b are the cluster result figure of egg mother cell matter and polar body.
Fig. 3 c are the image of partial syringe pin in ROI image.
The segmentation figure picture of cytoplasm and polar body when Fig. 4 a are level-set segmentation.
Fig. 4 b are level set movements 3 d effect graph.
Fig. 4 c are the segmentation result figure of one section of injection needle inside egg mother cell matter.
Fig. 5 a are injection needle segmentation figure when injection needle is fitted.
Fig. 5 b are intracellular rectilinear when injection needle is fitted.
Fig. 5 c are the complete pin image after injection needle fitting.
Fig. 6 is that the ICSI in background technology injects scene figure；
In figure：1 is absorption pin, and 2 be cytoplasm, and 3 be polar body, and 4 be cell week gap, and 5 be oolemma, and 6 be injection needle.
Embodiment
With reference to specific embodiment, the invention will be further described.
As shown in figure 1, the segmentation implementation method of the human oocytes in optical microphotograph injection process described in the present embodiment,
Comprise the following steps：
1) according to the human oocytes shape difference opposite sex, using matrix morphology operator extraction interesting target ROI, specifically
Polar body, one section of injection needle and cytoplasmic pretreatment image are included for extraction；
2) pretreatment image intensity inequality is considered and by noise pollution, using improved fuzzy clustering method, according to office
Portion's gray scale and the local Variance feature pretreatment image larger to obscurity boundary and noise carry out rough sort, that is, realize cytoplasm,
The rough sort of polar body and endocellular injection pin；
3) level set initialized location and Evolution Control parameter are set on the basis of fuzzy clustering, realizes cytoplasm, pole
The correct segmentation of body and injection needle；
4) intracytoplasmic partial syringe pin image is directed to, using fitting a straight line, the position of injection needle is fitted, marks
Complete injection needle image.
In step 1), according to the human oocytes shape difference opposite sex, using matrix morphology operator extraction mesh interested
Mark ROI, it is contemplated that the transparent knowledge for carrying two pixel wides, combining form operator of the present invention and skeletal extraction of cytoplasm,
Realize the coarse segmentation of target.Object can be reduced to lines by bianry image skeletonizing, but not change the basic structure of image, together
When retain the elementary contour of image well.
First, complete bianry image is obtained with OTSU methods, then does dilation operation, then through two-value skeletonizing, intactly
The profile of cytoplasm, the polar body of egg mother cell and absorption pin is extracted, its profile width is a pixel, while will adsorb pin skeleton
Single pixel straight line is turned to, it is connection that the image of skeletonizing, which only has cytoplasm, is easy to fill and marks；Secondly, filled out by region
Fill and mark, then using regular filter, the cytoplasm of egg mother cell is separately come out from the object background on periphery；Most
Afterwards, interesting target ROI is extracted.
It is to add local gray level and local Variance feature using improved fuzzy clustering method in step 2) so that when
The classification of preceding central point pixel is together decided on by the key words sorting of the pixel in its neighbour domain, so as to realize cytoplasm, polar body and
The rough sort of endocellular injection pin.
The ROI image extracted for image preprocessing, there are three cutting objects：One in cytoplasm, polar body and cytoplasm
Section injection needle.Because ICSI view data collects under light microscope, uneven illumination and noise pollution be present, and pass
The FCM of system is a kind of minimum cost function of global mean square deviation, when there is abnormity point, there is poor cluster result.And noise
Unavoidably occur in real Biomedical Image.Usual resolving ideas is to consider to add local characteristicses so that Current central
The classification of point pixel is together decided on by the key words sorting of the pixel in its neighbour domain, and then improves traditional FCM to noise-sensitive
Problem.
Assuming that the real image y collected is made up of true picture x and offset-domain b, and b is slowly to become in subrange
Change, image y, which amounts to, N number of pixel, then has：
So, the FCM objective cost functions rectification to standard is as follows：
Wherein, p is fuzzy membership Weighted Index, influences the fog-level of final classification；NkRepresent with xkCentered on it is near
Neighborhood；NRRepresent NkThe number of middle pixel；α is the control parameter in neighbour domain, and value is inversely proportional to the signal to noise ratio of image, by above formula (1)
Substitution has：
The optimal problem of above formula is sought using Lagrange multipliers：
Wherein, Dik=| | yk-bk-vi||2,
Respectively to uik、vikIt is 0 to seek first-order partial derivative and make result, can obtain degree of membershipAnd cluster centreSimilarly, it is right
Offset-domain bkEstimation, it is 0 to seek first-order partial derivative and make its result, can be in the hope of：
In step 3), with reference to fuzzy clustering result, realize the initialization of level set position and set level set movements to join
Number, direction and the speed parameter that the result of fuzzy clustering is used as developing without the variation level set of initialization so that in profile
Border can automatically determine evolution direction, accurately track boundary position, so as to realize the correct segmentation of target.
The result that fuzzy clustering obtains is a rough ideal sort, cluster result lead-out level collection can be used to develop first
Beginning position, so as to reduce curve in-depth operand, improve segmentation precision.Level Set Method be easy to combine priori guiding or
Drive the evolution of profile.The variation level diversity method without initialization that the priori such as Li propose, variation equation are as follows：
Wherein, φ is level set function, and three, the right is respectively penalty term, zero level collection length of curve regular terms and acceleration
The area item of curve evolvement；Coefficient is respectively μ, λ and ν；G is edge indicator function, and δ (φ) is single argument Dirac functions, and H is
Odd symmetry Heaviside functions；By the Euler-Lagrange calculus of variations, energy in the gradient descent method solving model of standard is used
Flow function minimizes；Obtain level set Evolution Equation be：
Wherein, the positive and negative decision initial profile of coefficient ν values does contraction or expansion evolution relative to target to be split, and
The speed of evolution, for actual segmentation problem, when it is expected away from object boundary, Evolution Rates are fast, conversely, it is slow to develop；Work as profile
During across object boundary, evolution direction can be changed automatically, rather than it is unalterable；Therefore, adaptive change coefficient ν values are more
It is adapted to actual segmentation demand, make the image of fuzzy clustering degree of membership so has for R：
νk=G (Rk)=1-2Rk,Rk∈(0,1) (10)
The Evolution Equation of level set is updated to：
The selection of other parameters：
Level set is initialized and its developed as follows：
Fuzzy clustering obtains a coarse classification results, the initial position that can develop from cluster result lead-out level collection,
So as to reduce operand.It is assumed that the image of selected fuzzy clustering degree of membership is R, an adjustable threshold value b is set0∈ (0,1),
Obtain a bianry image Bk：
So, level set initialization function, which reaches, is：
φ0=-4 ε (0.5-Bk), ε=1.5 (13).
With reference to Fig. 2 a to Fig. 5 c, implementation method is split to human egg of the present invention and is specifically described, its
Situation is as follows：
Experiment porch is PentiumIV 3.0G, 1G RAM, and video card is NVIDIA Quadro FX 1400.
Original image obtains complete bianry image as shown in Figure 2 a, with OTSU methods, and skeleton is extracted after expanded operation, such as schemes
Shown in 2b.Skeletonizing intactly remains the profile of cytoplasm, polar body, absorption pin and injection needle, and width is single pixel, injection
Pin skeleton turns to single pixel straight line, and cytoplasm is connection in the image of skeletonizing；Secondly, by area filling and mark, and
Regular filter is used afterwards, and the cytoplasm of egg mother cell is separately come out from the object background on periphery；Finally, it is emerging to extract sense
Interesting target ROI, as shown in Figure 2 c.
Cluster feature chooses each pixel local mean value and local variance, according to FCM sorting procedures, fuzzy clustering result
As shown in Fig. 3 a to Fig. 3 c, Fig. 3 a are divided into two classes, Fig. 3 b to pre-process to obtain ROI image in place with improved fuzzy clustering algorithm
It is the cluster result of ooplasm and polar body, cytoplasm inside is substantially smooth, and border is relatively clear, the Main Boundaries region of polar body
It is able to separate with cytoplasm；Fig. 3 c are the image of partial syringe pin in ROI image, and the cell outline on periphery can be in subsequent operation
Middle removal.
The solution that quantizes is realized using finite difference calculus, time local derviation quantizes using forward difference, the number of space local derviation
Value then uses centered difference.On the basis of dendrogram picture, the cluster result and injection needle of egg mother cell matter are based respectively on
Cluster result, realize and split with reference to level set equation.
Fig. 4 a are the segmentation figure picture of cytoplasm and polar body, smoother and develop boundary profile exactly；Fig. 4 b are water
Flat collection evolution 3 d effect graph, middle empty region is the position of injection needle, and a cell of bottom right independence is polar body；Fig. 4 c are
The segmentation result of one section of injection needle inside egg mother cell matter.In view of the linear feature of injection needle, on the basis of binaryzation, and
With reference to Fig. 4 c, straight-line detection is realized.Fig. 5 a to Fig. 5 c illustrate injection needle full segmentation process.
Embodiment described above is only present pre-ferred embodiments, and the practical range of the present invention is not limited with this, therefore
The change that all shape, principles according to the present invention are made, it all should cover within the scope of the present invention.
Claims (3)
1. one kind human oocytes in optical microphotograph injection process split implementation method, it is characterised in that including following step
Suddenly：
1) according to the human oocytes shape difference opposite sex, using matrix morphology operator extraction interesting target ROI；
2) improved fuzzy clustering method is used, it is larger to obscurity boundary and noise according to local gray level and local Variance feature
Pretreatment image carries out rough sort；Wherein, the improved fuzzy clustering method is to add local gray level and local Variance feature,
So that the classification of Current central point pixel is together decided on by the key words sorting of the pixel in its neighbour domain, so as to realize cytoplasm,
The rough sort of polar body and endocellular injection pin, its process are as follows：
Assuming that the real image y collected is made up of true picture x and offset-domain b, and b is slowly varying in subrange
, image y, which amounts to, N number of pixel, then has：
<mrow>
<msub>
<mi>y</mi>
<mi>k</mi>
</msub>
<mo>=</mo>
<msub>
<mi>x</mi>
<mi>k</mi>
</msub>
<mo>+</mo>
<msub>
<mi>b</mi>
<mi>k</mi>
</msub>
<mo>,</mo>
<mo>&ForAll;</mo>
<mi>k</mi>
<mo>&Element;</mo>
<mo>{</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>,</mo>
<mo>...</mo>
<mo>,</mo>
<mi>N</mi>
<mo>}</mo>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
So, the FCM objective cost functions rectification to standard is as follows：
<mrow>
<msub>
<mi>J</mi>
<mi>m</mi>
</msub>
<mo>=</mo>
<munderover>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>c</mi>
</munderover>
<munderover>
<mi>&Sigma;</mi>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<msubsup>
<mi>u</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mi>p</mi>
</msubsup>
<mo>|</mo>
<mo>|</mo>
<msub>
<mi>x</mi>
<mi>k</mi>
</msub>
<mo>-</mo>
<msub>
<mi>v</mi>
<mi>i</mi>
</msub>
<mo>|</mo>
<msup>
<mo>|</mo>
<mn>2</mn>
</msup>
<mo>+</mo>
<mfrac>
<mi>&alpha;</mi>
<msub>
<mi>N</mi>
<mi>R</mi>
</msub>
</mfrac>
<munderover>
<mi>&Sigma;</mi>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>c</mi>
</munderover>
<munderover>
<mi>&Sigma;</mi>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<msubsup>
<mi>u</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mi>p</mi>
</msubsup>
<mrow>
<mo>(</mo>
<munder>
<mi>&Sigma;</mi>
<mrow>
<mi>x</mi>
<mo>&Element;</mo>
<msub>
<mi>N</mi>
<mi>k</mi>
</msub>
</mrow>
</munder>
<mo>|</mo>
<mo>|</mo>
<msub>
<mi>x</mi>
<mi>r</mi>
</msub>
<mo>-</mo>
<msub>
<mi>v</mi>
<mi>i</mi>
</msub>
<mo>|</mo>
<msup>
<mo>|</mo>
<mn>2</mn>
</msup>
<mo>)</mo>
</mrow>
<mo>,</mo>
<munderover>
<mi>&Sigma;</mi>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>c</mi>
</munderover>
<msub>
<mi>u</mi>
<mrow>
<mi>j</mi>
<mi>k</mi>
</mrow>
</msub>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mo>&ForAll;</mo>
<mi>k</mi>
<mo>&Element;</mo>
<mo>{</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mi>N</mi>
<mo>}</mo>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>2</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, p is fuzzy membership Weighted Index, influences the fog-level of final classification；NkRepresent with xkCentered on neighbour
Domain；NRRepresent NkThe number of middle pixel；α is the control parameter in neighbour domain, and value is inversely proportional to the signal to noise ratio of image, by above formula (1) generation
Enter to have：
<mrow>
<msub>
<mi>J</mi>
<mi>m</mi>
</msub>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>c</mi>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<msubsup>
<mi>u</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mi>p</mi>
</msubsup>
<mo>|</mo>
<mo>|</mo>
<msub>
<mi>y</mi>
<mi>k</mi>
</msub>
<mo>-</mo>
<msub>
<mi>b</mi>
<mi>k</mi>
</msub>
<mo>-</mo>
<msub>
<mi>v</mi>
<mi>i</mi>
</msub>
<mo>|</mo>
<msup>
<mo>|</mo>
<mn>2</mn>
</msup>
<mo>+</mo>
<mfrac>
<mi>&alpha;</mi>
<msub>
<mi>N</mi>
<mi>R</mi>
</msub>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>c</mi>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<msubsup>
<mi>u</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mi>p</mi>
</msubsup>
<mrow>
<mo>(</mo>
<munder>
<mo>&Sigma;</mo>
<mrow>
<msub>
<mi>y</mi>
<mi>r</mi>
</msub>
<mo>&Element;</mo>
<msub>
<mi>N</mi>
<mi>k</mi>
</msub>
</mrow>
</munder>
<mo>|</mo>
<mo>|</mo>
<msub>
<mi>y</mi>
<mi>r</mi>
</msub>
<mo>-</mo>
<msub>
<mi>b</mi>
<mi>r</mi>
</msub>
<mo>-</mo>
<msub>
<mi>v</mi>
<mi>i</mi>
</msub>
<mo>|</mo>
<msup>
<mo>|</mo>
<mn>2</mn>
</msup>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>3</mn>
<mo>)</mo>
</mrow>
</mrow>
The optimal problem of above formula is sought using Lagrange multipliers：
<mrow>
<msub>
<mi>F</mi>
<mi>m</mi>
</msub>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>c</mi>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<mrow>
<mo>(</mo>
<msubsup>
<mi>u</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mi>p</mi>
</msubsup>
<msub>
<mi>D</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
</msub>
<mo>+</mo>
<mfrac>
<mi>&alpha;</mi>
<msub>
<mi>N</mi>
<mi>R</mi>
</msub>
</mfrac>
<msubsup>
<mi>u</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mi>p</mi>
</msubsup>
<msub>
<mi>&gamma;</mi>
<mi>i</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>+</mo>
<mi>&lambda;</mi>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>-</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>c</mi>
</munderover>
<msubsup>
<mi>u</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mi>p</mi>
</msubsup>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>4</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, Dik=| | yk-bk-vi||2,
Respectively to uik、vikIt is 0 to seek first-order partial derivative and make result, obtains degree of membershipAnd cluster centreSimilarly, to offset-domain
bkEstimation, it is 0 to seek first-order partial derivative and make its result, is tried to achieve：
<mrow>
<msubsup>
<mi>u</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mo>*</mo>
</msubsup>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>c</mi>
</munderover>
<msup>
<mrow>
<mo>(</mo>
<mfrac>
<mrow>
<msub>
<mi>D</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
</msub>
<mo>+</mo>
<mfrac>
<mi>&alpha;</mi>
<msub>
<mi>N</mi>
<mi>R</mi>
</msub>
</mfrac>
<msub>
<mi>&gamma;</mi>
<mi>i</mi>
</msub>
</mrow>
<mrow>
<mo>(</mo>
<msub>
<mi>D</mi>
<mrow>
<mi>j</mi>
<mi>k</mi>
</mrow>
</msub>
<mo>+</mo>
<mfrac>
<mi>&alpha;</mi>
<msub>
<mi>N</mi>
<mi>R</mi>
</msub>
</mfrac>
<msub>
<mi>&gamma;</mi>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mrow>
<mo>-</mo>
<mn>1</mn>
<mo>/</mo>
<mrow>
<mo>(</mo>
<mi>p</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
</msup>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>5</mn>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msubsup>
<mi>v</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mo>*</mo>
</msubsup>
<mo>=</mo>
<mfrac>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<msubsup>
<mi>u</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mi>p</mi>
</msubsup>
<mrow>
<mo>(</mo>
<mo>(</mo>
<mrow>
<msub>
<mi>y</mi>
<mi>k</mi>
</msub>
<mo>-</mo>
<msub>
<mi>b</mi>
<mi>k</mi>
</msub>
</mrow>
<mo>)</mo>
<mo>+</mo>
<mfrac>
<mi>&alpha;</mi>
<msub>
<mi>N</mi>
<mi>R</mi>
</msub>
</mfrac>
<munder>
<mo>&Sigma;</mo>
<mrow>
<msub>
<mi>y</mi>
<mi>r</mi>
</msub>
<mo>&Element;</mo>
<msub>
<mi>N</mi>
<mi>k</mi>
</msub>
</mrow>
</munder>
<mo>(</mo>
<mrow>
<msub>
<mi>y</mi>
<mi>r</mi>
</msub>
<mo>-</mo>
<msub>
<mi>b</mi>
<mi>r</mi>
</msub>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>+</mo>
<mi>&alpha;</mi>
<mo>)</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<msubsup>
<mi>u</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mi>p</mi>
</msubsup>
</mrow>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>6</mn>
<mo>)</mo>
</mrow>
</mrow>
1
<mrow>
<msubsup>
<mi>b</mi>
<mi>k</mi>
<mo>*</mo>
</msubsup>
<mo>=</mo>
<msub>
<mi>y</mi>
<mi>k</mi>
</msub>
<mo>-</mo>
<mfrac>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>c</mi>
</munderover>
<msubsup>
<mi>u</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mi>p</mi>
</msubsup>
<msub>
<mi>v</mi>
<mi>i</mi>
</msub>
</mrow>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>c</mi>
</munderover>
<msubsup>
<mi>u</mi>
<mrow>
<mi>i</mi>
<mi>k</mi>
</mrow>
<mi>p</mi>
</msubsup>
</mrow>
</mfrac>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>7</mn>
<mo>)</mo>
</mrow>
<mo>;</mo>
</mrow>
3) level set initialized location and Evolution Control parameter are set on the basis of fuzzy clustering, realize cytoplasm, polar body and
The segmentation of injection needle；
4) intracytoplasmic partial syringe pin image is directed to, using fitting a straight line, marks complete injection needle image.
2. one kind according to claim 1 human oocytes in optical microphotograph injection process split implementation method, its
It is characterised by：In step 1), according to the human oocytes shape difference opposite sex, using matrix morphology operator extraction mesh interested
ROI is marked, its process is：First, complete bianry image is obtained with OTSU methods, then does dilation operation, then through two-value skeletonizing,
The profile of cytoplasm, the polar body of egg mother cell and absorption pin is intactly extracted, its profile width is a pixel, while will absorption
Pin skeleton turns to single pixel straight line, and it is connection that the image of skeletonizing, which only has cytoplasm, is easy to fill and marks；Secondly, pass through
Area filling and mark, then using regular filter, the cytoplasm of egg mother cell is divided from the object background on periphery and opened out
Come；Finally, interesting target ROI is extracted.
3. one kind according to claim 1 human oocytes in optical microphotograph injection process split implementation method, its
It is characterised by：In step 3), with reference to fuzzy clustering result, realize the initialization of level set position and set level set movements to join
Number, direction and the speed parameter that the result of fuzzy clustering is used as developing without the variation level set of initialization so that in profile
Border automatically determines evolution direction, accurately tracks boundary position, so as to realize the correct segmentation of target；Variation equation is as follows：
<mrow>
<msub>
<mi>E</mi>
<mn>0</mn>
</msub>
<mrow>
<mo>(</mo>
<mi>&phi;</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mi>&mu;</mi>
<munder>
<mrow>
<mo>&Integral;</mo>
<mo>&Integral;</mo>
</mrow>
<mi>&Omega;</mi>
</munder>
<mfrac>
<mn>1</mn>
<mn>2</mn>
</mfrac>
<msup>
<mrow>
<mo>(</mo>
<mo>|</mo>
<mrow>
<mo>&dtri;</mo>
<mi>&phi;</mi>
</mrow>
<mo>|</mo>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mi>d</mi>
<mi>x</mi>
<mi>d</mi>
<mi>y</mi>
<mo>+</mo>
<mi>&lambda;</mi>
<munder>
<mrow>
<mo>&Integral;</mo>
<mo>&Integral;</mo>
</mrow>
<mi>&Omega;</mi>
</munder>
<mi>g</mi>
<mi>&delta;</mi>
<mrow>
<mo>(</mo>
<mi>&phi;</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
<mrow>
<mo>&dtri;</mo>
<mi>&phi;</mi>
</mrow>
<mo>|</mo>
<mi>d</mi>
<mi>x</mi>
<mi>d</mi>
<mi>y</mi>
<mo>+</mo>
<mi>v</mi>
<munder>
<mrow>
<mo>&Integral;</mo>
<mo>&Integral;</mo>
</mrow>
<mi>&Omega;</mi>
</munder>
<mi>g</mi>
<mi>H</mi>
<mrow>
<mo>(</mo>
<mo>-</mo>
<mi>&phi;</mi>
<mo>)</mo>
</mrow>
<mi>d</mi>
<mi>x</mi>
<mi>d</mi>
<mi>y</mi>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>8</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, φ is level set function, and three, the right is respectively penalty term, zero level collection length of curve regular terms and acceleration curve
The area item of evolution；Coefficient is respectively μ, λ and ν；G is edge indicator function, and δ (φ) is single argument Dirac functions, and H is strange right
Claim Heaviside functions；By the Euler-Lagrange calculus of variations, energy letter in the gradient descent method solving model of standard is used
Number minimizes；Obtain level set Evolution Equation be：
<mrow>
<mfrac>
<mrow>
<mo>&part;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>&part;</mo>
<mi>t</mi>
</mrow>
</mfrac>
<mo>=</mo>
<mi>&mu;</mi>
<mo>&lsqb;</mo>
<mi>&Delta;</mi>
<mi>&phi;</mi>
<mo>-</mo>
<mi>d</mi>
<mi>i</mi>
<mi>v</mi>
<mrow>
<mo>(</mo>
<mfrac>
<mrow>
<mo>&dtri;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>|</mo>
<mrow>
<mo>&dtri;</mo>
<mi>&phi;</mi>
</mrow>
<mo>|</mo>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mo>&rsqb;</mo>
<mo>+</mo>
<mi>&delta;</mi>
<mrow>
<mo>(</mo>
<mi>&phi;</mi>
<mo>)</mo>
</mrow>
<mo>&lsqb;</mo>
<mi>&lambda;</mi>
<mi>d</mi>
<mi>i</mi>
<mi>v</mi>
<mrow>
<mo>(</mo>
<mi>g</mi>
<mfrac>
<mrow>
<mo>&dtri;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>|</mo>
<mrow>
<mo>&dtri;</mo>
<mi>&phi;</mi>
</mrow>
<mo>|</mo>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mo>+</mo>
<mi>v</mi>
<mi>g</mi>
<mo>&rsqb;</mo>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>9</mn>
<mo>)</mo>
</mrow>
</mrow>
Wherein, the positive and negative decision initial profile of coefficient ν values is contraction or expansion evolution, Yi Jiyan relative to human oocytes
The speed of change, for actual segmentation problem, when it is expected away from object boundary, Evolution Rates are fast, conversely, it is slow to develop；When profile across
When crossing object boundary, change evolution direction automatically, rather than it is unalterable；Therefore, adaptive change coefficient ν values are more suitable for reality
Demand is split on border, makes the image of fuzzy clustering degree of membership and so has for R：
νk=G (Rk)=1-2Rk,Rk∈(0,1) (10)
The Evolution Equation of level set is updated to：
<mrow>
<mfrac>
<mrow>
<mo>&part;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>&part;</mo>
<mi>t</mi>
</mrow>
</mfrac>
<mo>=</mo>
<mi>&mu;</mi>
<mo>&lsqb;</mo>
<mi>&Delta;</mi>
<mi>&phi;</mi>
<mo>-</mo>
<mi>d</mi>
<mi>i</mi>
<mi>v</mi>
<mrow>
<mo>(</mo>
<mfrac>
<mrow>
<mo>&dtri;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>|</mo>
<mrow>
<mo>&dtri;</mo>
<mi>&phi;</mi>
</mrow>
<mo>|</mo>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mo>&rsqb;</mo>
<mo>+</mo>
<mi>&delta;</mi>
<mrow>
<mo>(</mo>
<mi>&phi;</mi>
<mo>)</mo>
</mrow>
<mo>&lsqb;</mo>
<mi>&lambda;</mi>
<mi>d</mi>
<mi>i</mi>
<mi>v</mi>
<mrow>
<mo>(</mo>
<mi>g</mi>
<mfrac>
<mrow>
<mo>&dtri;</mo>
<mi>&phi;</mi>
</mrow>
<mrow>
<mo>|</mo>
<mrow>
<mo>&dtri;</mo>
<mi>&phi;</mi>
</mrow>
<mo>|</mo>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mo>+</mo>
<mi>G</mi>
<mi>g</mi>
<mo>&rsqb;</mo>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>11</mn>
<mo>)</mo>
</mrow>
</mrow>
The selection of other parameters：
Level set is initialized and its developed as follows：
It is assumed that the image of selected fuzzy clustering degree of membership is R, an adjustable threshold value b is set0∈ (0,1), obtain a two-value
Image Bk：
<mrow>
<msub>
<mi>B</mi>
<mi>k</mi>
</msub>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mn>1</mn>
<mo>,</mo>
<msub>
<mi>R</mi>
<mi>k</mi>
</msub>
<mo>&GreaterEqual;</mo>
<msub>
<mi>b</mi>
<mn>0</mn>
</msub>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mn>0</mn>
<mo>,</mo>
<msub>
<mi>R</mi>
<mi>k</mi>
</msub>
<mo><</mo>
<msub>
<mi>b</mi>
<mn>0</mn>
</msub>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>12</mn>
<mo>)</mo>
</mrow>
</mrow>
So, level set initialization function, which reaches, is：
φ0=-4 ε (0.5-Bk), ε=1.5 (13).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410214660.XA CN104008545B (en) | 2014-05-20 | 2014-05-20 | One kind human oocytes in optical microphotograph injection process split implementation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410214660.XA CN104008545B (en) | 2014-05-20 | 2014-05-20 | One kind human oocytes in optical microphotograph injection process split implementation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104008545A CN104008545A (en) | 2014-08-27 |
| CN104008545B true CN104008545B (en) | 2017-12-01 |
Family
ID=51369185
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410214660.XA Expired - Fee Related CN104008545B (en) | 2014-05-20 | 2014-05-20 | One kind human oocytes in optical microphotograph injection process split implementation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104008545B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110428430B (en) * | 2019-07-03 | 2023-06-27 | 广州微林软件有限公司 | Linear positioning fitting method in region and application |
| CN111104894B (en) * | 2019-12-17 | 2021-01-12 | 推想医疗科技股份有限公司 | Sperm labeling method and device, electronic equipment and storage medium |
| CN116402775B (en) * | 2023-03-29 | 2023-12-22 | 浙江大学 | Cell deformation control method based on multitasking sensing network |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103400370A (en) * | 2013-06-25 | 2013-11-20 | 西安电子科技大学 | Adaptive fuzzy C-means image segmentation method based on potential function |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE112011103687B4 (en) * | 2010-11-05 | 2017-01-05 | Cytognomix, Inc. | Centromere detector and method for determining radiation exposure by chromosomal abnormalities |
-
2014
- 2014-05-20 CN CN201410214660.XA patent/CN104008545B/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103400370A (en) * | 2013-06-25 | 2013-11-20 | 西安电子科技大学 | Adaptive fuzzy C-means image segmentation method based on potential function |
Non-Patent Citations (3)
| Title |
|---|
| A Modified Fuzzy C-Means Algorithm for Bias Field Estimation and Segmentation of MRI Data;Mohamed N.Ahmed等;《IEEE Transactions on Medical Imaging》;20020807;第21卷(第3期);第193-199页 * |
| Integrating spatial fuzzy clustering with level set methods for automated medical image segmentation;Bing Nan Li等;《Computers in Biology and Medicine》;20110131;第41卷(第1期);第1-10页 * |
| 基于变分水平集的显微注射过程细胞分割;秦传波等;《华南理工大学学报》;20140315;第42卷(第3期);第15-20,34页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104008545A (en) | 2014-08-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109034210B (en) | Target detection method based on super-feature fusion and multi-scale pyramid network | |
| Oztel et al. | Mitochondria segmentation in electron microscopy volumes using deep convolutional neural network | |
| CN106446930A (en) | Deep convolutional neural network-based robot working scene identification method | |
| CN103996018A (en) | Human-face identification method based on 4DLBP | |
| Wang et al. | Morphological analysis of dendrites and spines by hybridization of ridge detection with twin support vector machine | |
| CN105469111A (en) | Small sample set object classification method on basis of improved MFA and transfer learning | |
| CN104008545B (en) | One kind human oocytes in optical microphotograph injection process split implementation method | |
| CN107103608B (en) | A kind of conspicuousness detection method based on region candidate samples selection | |
| CN109325451B (en) | High-spatial-resolution cultivated land block full-automatic extraction method based on deep learning | |
| CN104268600A (en) | Mineral flotation froth image texture analysis and working condition identification method based on Minkowski distance | |
| CN107680099A (en) | A kind of fusion IFOA and F ISODATA image partition method | |
| CN105426924A (en) | Scene classification method based on middle level features of images | |
| CN103150562B (en) | A kind of ant group many cells tracker based on premise of image background extraction | |
| Li et al. | ADVANCING ORE PROCESSING THROUGH DEEP LEARNING-POWERED IMAGE RECOGNITION: A COMPREHENSIVE OVERVIEW AND FUTURE PERSPECTIVES | |
| Ma et al. | Feature selection module for CNN based object detector | |
| CN115661123B (en) | Industrial product surface defect position detection method based on weak supervision target detection | |
| CN105528791A (en) | Quality evaluation device facing touch screen hand-drawn image and evaluation method thereof | |
| CN113205526B (en) | Distribution line accurate semantic segmentation method based on multi-source information fusion | |
| Li et al. | Identification and localization of grape diseased leaf images captured by UAV based on CNN | |
| Ge et al. | Real-time object detection algorithm for Underwater Robots | |
| Oztel et al. | Deep learning approaches in electron microscopy imaging for mitochondria segmentation | |
| Zhang et al. | Multiple Objects Detection based on Improved Faster R-CNN | |
| CN106778831A (en) | The online tagsort of rigid-object and tracking based on gauss hybrid models | |
| Cao et al. | A hybrid method for identifying the feeding behavior of tilapia | |
| CN110751638A (en) | Galvanized sheet surface defect segmentation method based on matrix decomposition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20171201 |