CN109191470A - Image partition method and device suitable for big data image - Google Patents
Image partition method and device suitable for big data image Download PDFInfo
- Publication number
- CN109191470A CN109191470A CN201810943197.0A CN201810943197A CN109191470A CN 109191470 A CN109191470 A CN 109191470A CN 201810943197 A CN201810943197 A CN 201810943197A CN 109191470 A CN109191470 A CN 109191470A
- Authority
- CN
- China
- Prior art keywords
- image
- segmentation
- cell
- skewback
- cell image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/155—Segmentation; Edge detection involving morphological operators
-
- 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
- G06T2207/30004—Biomedical image processing
- G06T2207/30096—Tumor; Lesion
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
Present disclose provides a kind of image partition methods suitable for big data image, divide to received multiple cell images；According to maximum between-cluster variance algorithm, coarse segmentation operation is carried out for the area-of-interest in multiple cell images after division；It is suitable for the result obtained after coarse segmentation operation the image segmentation operations of cell image by shape test completion.This method is directed to cervical exfoliated cell image, carries out coarse segmentation to cell image using maximum between-cluster variance algorithm, obtains individual cells, population of cells and impurity block image；Then shape test is carried out to the result divided above, if test condition passes through, then area-of-interest is cell image, is finally completed feature extraction.The above method efficiently, precisely, quickly can realize image segmentation strategy for cervical exfoliated cell image, and have ease for use and applicability.The disclosure additionally provides a kind of image segmentation device suitable for cell image.
Description
Technical field
This disclosure relates to which field of computer technology and image identification technical field, are suitable for big in particular to one kind
The image partition method and device of data image.
Background technique
In the prior art, the cell in cell image is accurately and rapidly divided, obtains ROI (Region Of
Interest, area-of-interest), it is the important foundation differentiated to cell.Especially to the cell in cervical exfoliated cell image into
Row is accurately and rapidly divided, and obtains area-of-interest, is the more vital basis differentiated to cervical cancer cell in the later period.By
During smear production is with Image Acquisition, inevitably it is mingled with some impurity, interference etc. in image, so that image
In have the problems such as noise, fuzzy, gray scale is uneven, increase difficulty and challenge accurately divides to image.
Summary of the invention
In order to solve in the prior art due to during smear production is with Image Acquisition, in image inevitably
It is mingled with some impurity, interference etc., so that there is the problems such as noise, fuzzy, gray scale is uneven in image, image is accurately divided not enough
Accurately with quick problem, the embodiment of the present disclosure provides a kind of image partition method and device suitable for big data image,
For cervical exfoliated cell image, coarse segmentation is carried out to cell image using maximum between-cluster variance algorithm, obtains individual cells, thin
Born of the same parents group and impurity block image；Then shape test is carried out to the result divided above, it is if test condition passes through, then interested
Region is cell image, into characteristic extracting module；Test unacceptable area-of-interest, that is, be likely to be population of cells or
Impurity block group then carries out secondary splitting to area-of-interest based on gradient vector flow-active contour model dividing method, so
Shape test is carried out to secondary splitting result afterwards.
In a first aspect, the embodiment of the present disclosure provides a kind of image partition method suitable for big data image, including with
Lower step: received multiple cell images are divided；According to maximum between-cluster variance algorithm, for the multiple after division
Area-of-interest in cell image carries out coarse segmentation operation；The result obtained after coarse segmentation operation is tested by shape and is completed
Image segmentation operations suitable for cell image.
In one of the embodiments, it is described to received multiple cell images carry out divide include: based on training sample
Cell image data statistics rule, selection segmentation channel；The segmentation threshold in the segmentation channel is chosen, and to described
Cell image carries out foreground and background segmentation；According to the foreground pixel and background pixel progress connection regional analysis after segmentation, obtain
Take qualified cell compartment.
In one of the embodiments, further include: to obtain qualified cell compartment with N row N column division mode into
The division of the row cell image sub-block, wherein the N row is classified as equal numerical value with the N.
The result obtained after the operation to coarse segmentation in one of the embodiments, is completed to be suitable for by shape test
The figure cutting operation of cell image includes: to test test by area to the result obtained after coarse segmentation operation to complete to be suitable for
The figure cutting operation of cell image, wherein area test be judge the area-of-interest pixel number whether
Meet the pixel threshold interval of default normal cell area.
The result obtained after the operation to coarse segmentation in one of the embodiments, is completed to be suitable for by shape test
The figure cutting operation of cell image includes: to be completed by the test of deformity degree suitable for thin to the result obtained after coarse segmentation operation
The figure cutting operation of born of the same parents' image, wherein the deformity degree test is to pass through simple deformity degree calculation formula γ=l/NpIt calculates
The lopsided degree of the area-of-interest, wherein l is the perimeter of the area-of-interest, NpFor the pixel of the area-of-interest
Point number.
In one of the embodiments, further include: preset lopsided degree high threshold γT；As γ≤γTWhen, determine to thick
The result obtained after cutting operation is tested by deformity degree.
In one of the embodiments, further include: as γ > γTWhen, based on gradient vector flow-active contour model point
Segmentation method carries out secondary coarse segmentation operation to area-of-interest；The result obtained after the operation of secondary coarse segmentation is tested by shape
Complete the image segmentation operations for being suitable for cell image.
Second aspect, the embodiment of the present disclosure provide a kind of computer readable storage medium, are stored thereon with computer journey
The step of sequence, which realizes above-mentioned method when being executed by processor.
The third aspect, the embodiment of the present disclosure provide a kind of computer equipment, including memory, processor and are stored in
On reservoir and the computer program that can run on a processor, the processor realize above-mentioned method when executing described program
Step.
Fourth aspect, the embodiment of the present disclosure provide a kind of image segmentation device suitable for cell image, described device
It include: division module, for being divided to received multiple cell images；Coarse segmentation module, for according to side between maximum kind
Difference algorithm carries out coarse segmentation operation for the area-of-interest in the multiple cell image after division；Divide module, is used for
It is suitable for the result obtained after coarse segmentation operation the image segmentation operations of cell image by shape test completion.
A kind of image partition method and device suitable for big data image provided by the invention, to received multiple cells
Image is divided；According to maximum between-cluster variance algorithm, carried out for the area-of-interest in multiple cell images after division
Coarse segmentation operation；The result obtained after coarse segmentation operation is grasped by the image segmentation that shape test completes to be suitable for cell image
Make.This method is directed to cervical exfoliated cell image, carries out coarse segmentation to cell image using maximum between-cluster variance algorithm, obtains single
A cell, population of cells and impurity block image；Then shape test is carried out to the result divided above, as test condition is logical
It crosses, then area-of-interest is cell image, into characteristic extracting module；Unacceptable area-of-interest is tested, that is, is likely to be
Population of cells or impurity block group, then based on gradient vector flow-active contour model dividing method to area-of-interest into
Then row secondary splitting carries out shape test to secondary splitting result.Thus, it is possible to efficiently, precisely, quickly be taken off for uterine neck
It falls cell image and realizes image segmentation strategy, and there is ease for use and applicability.
Detailed description of the invention
In order to illustrate more clearly of the technical solution of the embodiment of the present disclosure, below to needed in embodiment description
Attached drawing is briefly described:
Fig. 1 is the step process for the image partition method that one of one embodiment of the invention is suitable for big data image
Schematic diagram；
The step of Fig. 2 is image partition method of one of the another embodiment of the present invention suitable for big data image is flowed
Journey schematic diagram；And
Fig. 3 is the structural representation for the image segmentation device that one of one embodiment of the invention is suitable for cell image
Figure.
Fig. 4 is master of one of the one embodiment of the invention using the intelligent apparatus of the image partition method of cell image
View；
Fig. 5 is the top view of Fig. 4；
Fig. 6 uses the intelligent apparatus of the image partition method of cell image for the another kind in one embodiment of the invention
Main view；
Fig. 7 is the top view of Fig. 6；
Fig. 8 is the motion state variation diagram of Fig. 7.
Specific embodiment
The application is further discussed in detail with reference to the accompanying drawings and examples.
In following introductions, term " first ", " second " only for descriptive purposes, and should not be understood as instruction or dark
Show relative importance.Following introductions provide multiple embodiments of the disclosure, can replace or merge between different embodiments
Combination, therefore the application is it is also contemplated that all possible combinations comprising documented identical and/or different embodiments.Thus, such as
Fruit one embodiment include feature A, B, C, another embodiment include feature B, D, then the application also should be regarded as include containing
A, the every other possible combined embodiment of one or more of B, C, D, although the embodiment may be in the following contents
In have specific literature record.
In order to make the objectives, technical solutions, and advantages of the present invention clearer, by the following examples, it and combines attached
Figure carries out further specifically the specific embodiment of image partition method and device that the present invention is suitable for big data image
It is bright.It should be appreciated that described herein, specific examples are only used to explain the present invention, is not intended to limit the present invention.
Studies have pointed out that accurately and rapidly being divided to the cell in cervical exfoliated cell image, ROI is obtained
(Region Of Interest, area-of-interest) is the important foundation differentiated to cervical cancer cell in the later period.Due in smear system
During making with Image Acquisition, some impurity, interference etc. are inevitably mingled in image so that have in image noise,
Fuzzy, the problems such as gray scale is uneven, increase difficulty and challenge accurately divides to image.This allows for efficient, accurate, quickly needle
Image segmentation strategy, which becomes the work that an existing learning value has realistic meaning again, to be realized to cervical exfoliated cell image.
As shown in Figure 1, showing for the process that one of one embodiment is suitable for the image partition method of big data image
It is intended to, specifically includes the following steps:
Step 102, received multiple cell images are divided.
In one embodiment, divide to received multiple cell images includes: the cytological map based on training sample
As the statistics rule of data, selection segmentation channel；The segmentation threshold in segmentation channel is chosen, and prospect is carried out to cell image
And background segment；According to the foreground pixel and background pixel progress connection regional analysis after segmentation, qualified cell is obtained
Region.
It should be noted that the statistics rule of the cell image data based on training sample, selection segmentation channel include:
Cell image statistical data rule based on training sample, obtains the distribution situation of image value in different color channel；From point
The maximum color channels of image value variance are obtained in cloth situation, form segmentation channel.In addition it is also necessary to which explanation, chooses and divides
The segmentation threshold in channel is cut, and carrying out foreground and background segmentation to cell image includes: by maximum between-cluster variance algorithm
Minimum algorithm, obtain segmentation threshold；Obtain the image pixel value of cell image；According to image pixel value and segmentation threshold
Dichotomy segmentation is carried out, foreground and background is obtained.Wherein multiple cell images are divided, that is, can be regarded as counting greatly
It is divided according to image, so that being divided using same software, Same Way for multiple images.
Further, it should be noted that dichotomy segmentation is carried out according to image pixel value and segmentation threshold, before acquisition
Scape and background include: the region for obtaining image pixel value and being higher than segmentation threshold, as prospect；Image pixel value is obtained to be lower than or wait
In the region of segmentation threshold, as background.
Further, it according to the foreground pixel and background pixel progress connection regional analysis after segmentation, obtains and meets item
The cell compartment of part include: to after segmentation foreground pixel and background pixel cluster, formed connection region；Select connection area
Size is maximum in domain and meets the region of a priori location information, forms qualified cell compartment, and to qualified thin
Born of the same parents region exports.
In addition, in one embodiment, this disclosure relates to the image partition method suitable for big data image further include:
The division of cell image sub-block is carried out with N row N column division mode to the qualified cell compartment of acquisition, wherein N row and N are arranged
For equal numerical value.The agility for subsequent image segmentation provides necessary data basis as a result,.
Step 104, according to maximum between-cluster variance algorithm, for the area-of-interest in multiple cell images after division into
The operation of row coarse segmentation.Wherein, it should be noted that maximum between-cluster variance algorithm is that original image is divided into prospect using threshold value, is carried on the back
Two images of scape.Specifically, prospect: using n1, csum, m1 indicate the points of the prospect under present threshold value, and moment of mass is average
Gray scale；Background: using n2, and sum-csum, m2 indicate the points of the background under present threshold value, moment of mass, average gray.When taking
When optimal threshold, background should be maximum with prospect difference, and key is the standard for how selecting to measure difference, and measures difference
Standard is exactly maximum between-cluster variance, and inter-class variance indicates that maximum between-cluster variance is indicated with fmax with sb.Further, about most
Variance algorithm is sensitive to noise and target sizes between major class, it is only that unimodal image generates preferable segmentation effect to inter-class variance
Fruit.When the size great disparity of target and background, bimodal or multimodal may be presented in inter-class variance criterion function, and effect is not at this time
It is good, but Ostu method is that the used time is least.Further, the derivation of equation about maximum between-cluster variance hair are as follows: remember that t is
The segmentation threshold of prospect and background, it is w0, average gray u0 that prospect points, which account for image scaled,；Background points account for image scaled and are
W1, average gray u1.The then overall average gray scale of image are as follows: u=w0*u0+w1*u1.The variance of foreground and background image can be with
It is expressed by following formula: g=w0* (u0-u) * (u0-u)+w1* (u1-u) * (u1-u)=w0*w1* (u0-u1) * (u0-
u1).It should be noted that above-mentioned formula is formula of variance.It can refer to the formula i.e. sb as described below of the g of probability theory
Expression formula.When variance g maximum, it is believed that foreground and background difference is maximum at this time, and gray scale t at this time is optimal threshold sb=
w0*w1*(u1-u0)*(u0-u1)。
Step 106, the image for being suitable for cell image is completed by shape test to the result obtained after coarse segmentation operation
Cutting operation.
In one embodiment, the result obtained after coarse segmentation operation is completed to be suitable for cell image by shape test
Figure cutting operation include: to be completed by area test test suitable for cell image to the result that obtains after coarse segmentation operation
Figure cutting operation, wherein area test is judges whether the pixel number of area-of-interest meets default normal cell
The pixel threshold interval of area.In addition, completing to be suitable for cell by shape test to the result obtained after coarse segmentation operation
The figure cutting operation of image includes: to complete to be suitable for cytological map by the test of deformity degree to the result obtained after coarse segmentation operation
The figure cutting operation of picture, wherein lopsided degree test is to pass through simple deformity degree calculation formula γ=l/NpCalculate region of interest
The lopsided degree in domain, wherein l is the perimeter of the area-of-interest, NpFor the pixel number of area-of-interest.
Further, this disclosure relates to the image partition method suitable for big data image further include: preset abnormal
Shape degree high threshold γT；As γ≤γTWhen, determine to test the result obtained after coarse segmentation operation by deformity degree: as γ > γT
When, secondary coarse segmentation operation is carried out to area-of-interest based on gradient vector flow-active contour model dividing method；To secondary
The result obtained after coarse segmentation operation completes the image segmentation operations for being suitable for cell image by shape test.
In order to it is more clear with accurately understand and apply this disclosure relates to a kind of image suitable for big data image
Dividing method carries out following example in conjunction with Fig. 2, it should be noted that the range that the disclosure is protected is not limited to following example.
Specifically, step 201 to step 208 successively are as follows: receive multiple images；N*N sub-block is divided to image, by most
Variance algorithm carries out coarse segmentation operation between major class, judges whether area-of-interest meets cell grown form, if area-of-interest
Meet cell grown form, then send the feature for completing cell image to preset characteristic model to mention the image of area-of-interest
It takes；If area-of-interest does not meet cell grown form, secondary splitting behaviour is carried out based on gradient vector flow-active contour model
After work, then judge whether area-of-interest meets cell grown form, it, will sense if area-of-interest meets cell grown form
The image in interest region send to preset characteristic model the feature extraction for completing cell image；If area-of-interest does not meet cell
Grown form then carries out rejecting operation to the impurity in cell image.
It is understood that being divided to received multiple cell images；According to maximum between-cluster variance algorithm, for drawing
The area-of-interest in multiple cell images after point carries out coarse segmentation operation；Shape is passed through to the result obtained after coarse segmentation operation
The image segmentation operations for being suitable for cell image are completed in shape test.Specifically, being directed to cervical exfoliated cell image, the disclosure is first
Coarse segmentation is carried out to cell image using maximum between-cluster variance algorithm, obtains individual cells, population of cells and impurity block image；
Then shape test is carried out to the result divided above.
It should be noted that test condition are as follows: area test.ROI's (Region Of Interest, area-of-interest)
Pixel number Np, i.e. the ROI area range [N that whether meets normal cell areamin,Nmax] within；Lopsided degree test.Pass through
Simple deformity degree calculation formula γ=l/NpThe lopsided degree of ROI region is calculated, l is the perimeter of ROI in formula, is equipped with the high threshold of deformity degree
Value γT, as γ≤γTWhen test pass through.Further, it is assumed that test condition passes through, then ROI region is cell image, into spy
Levy extraction module；If the unacceptable ROI region of test condition, that is, it is likely to be population of cells or impurity block group, then is based on
Gradient vector flow-active contour model dividing method carries out secondary splitting to ROI region, then carries out to secondary splitting result
Shape test, test condition is as described above.Wherein, those skilled in the art is not it is understood that test passes through then ROI
For impurity, directly abandon；Testing the ROI region that passes through is cell image, into default characteristic extracting module to cell image into
Row feature extraction.
A kind of image partition method suitable for big data image provided by the invention, to received multiple cell images into
Row divides；According to maximum between-cluster variance algorithm, coarse segmentation is carried out for the area-of-interest in multiple cell images after division
Operation；It is suitable for the result obtained after coarse segmentation operation the image segmentation operations of cell image by shape test completion.It should
Method is directed to cervical exfoliated cell image, carries out coarse segmentation to cell image using maximum between-cluster variance algorithm, obtains single thin
Born of the same parents, population of cells and impurity block image；Then shape test is carried out to the result divided above, if test condition passes through, then
Area-of-interest is cell image, into characteristic extracting module；Unacceptable area-of-interest is tested, that is, is likely to be cell mass
It falls or impurity block group, then area-of-interest is carried out based on gradient vector flow-active contour model dividing method secondary
Then segmentation carries out shape test to secondary splitting result, efficiently, precisely, quickly realize for cervical exfoliated cell image
Image segmentation strategy, and there is ease for use and applicability.
Based on the same inventive concept, a kind of image segmentation device suitable for cell image is additionally provided.Due to this device
The principle solved the problems, such as is similar to a kind of aforementioned image partition method suitable for big data image, therefore, the implementation of the device
It can realize that overlaps will not be repeated according to the specific steps of preceding method.
As shown in figure 3, being suitable for the structural representation of the image segmentation device of cell image for one of one embodiment
Figure.The image segmentation device 10 for being suitable for cell image includes: division module 200, coarse segmentation module 400 and segmentation module
600。
Wherein, division module 200 is for dividing received multiple cell images；Coarse segmentation module 400 is used for root
According to maximum between-cluster variance algorithm, coarse segmentation operation is carried out for the area-of-interest in the multiple cell image after division；
Segmentation module 600 is used to complete the result obtained after coarse segmentation operation by shape test the image point for being suitable for cell image
Cut operation.
A kind of image segmentation device suitable for cell image provided by the invention, division module is to received multiple cells
Image is divided；Coarse segmentation module is emerging for the sense in multiple cell images after division according to maximum between-cluster variance algorithm
Interesting region carries out coarse segmentation operation；Segmentation module is completed to be suitable for thin to the result obtained after coarse segmentation operation by shape test
The image segmentation operations of born of the same parents' image.The device is directed to cervical exfoliated cell image, using maximum between-cluster variance algorithm to cytological map
As carrying out coarse segmentation, individual cells, population of cells and impurity block image are obtained；Then shape is carried out to the result divided above
Test, if test condition passes through, then area-of-interest is cell image, into characteristic extracting module；It is emerging to test unacceptable sense
Interesting region is likely to be population of cells or impurity block group, then based on gradient vector flow-active contour model segmentation side
Method carries out secondary splitting to area-of-interest, then carries out shape test to secondary splitting result, efficiently, precisely, quickly needle
Image segmentation strategy is realized to cervical exfoliated cell image, and there is ease for use and applicability.
A kind of intelligent apparatus of the present invention, is stored thereon with computer program, wherein the realization when program is executed by processor
The step of any one of claim 1-7 the method；
Wherein, the intelligent apparatus further includes Touch Screen 701, shell 702, spring 703, oval hemisphere block 704, first
Skewback 705, the second skewback 706, first axle 707, range finder module 708, the first permanent magnetic iron block 711, the second permanent magnetic iron block 712, institute
The surface of the lower surface and oval hemisphere block 704 of stating Touch Screen 701 is fixed, and the positive decentralization of the ellipse hemisphere block 704 is equipped with
First skewback 705, the second skewback 706, first skewback 705, the second skewback 706 one end pass through bearing and be mounted on first
On axis 707, first skewback 705, the second skewback 706 the other end be located at the underface at oval 704 center of hemisphere block,
The first axle 707 and the downside of the inside of shell 702 are fixed, the other end point of first skewback 705, the second skewback 706
It is not fixed with the first permanent magnetic iron block 711, the second permanent magnetic iron block 712, wherein first permanent magnetic iron block 711, the second permanent magnetic iron block
712 is mutually attracted, and two opposite side surfaces of the inside of the shell 702 are respectively fixed with range finder module 708, described in two
Range finder module 708 be respectively used to detect first skewback 705, the second skewback 706 moving distance a, b and be transmitted to processing
Device, the processor convert described a, b to according to database the angle of departure between first skewback 705, the second skewback 706
C is spent, and determines the longitudinal movement distance d of the oval hemisphere block 704 according to the database according to the separation angle c；
Wherein, the processor is equipped with first distance section, second distance section, third apart from section for d,
When the d is first distance section, then the Touch Screen 701 shows archaeocyte image；
When the d is second distance section, then the Touch Screen 701 shows the cell image after coarse segmentation operation；
When the d be third apart from section when, then the Touch Screen 701 shows that shape test is completed to be suitable for cytological map
The cell image of picture；
Wherein, the first interval be [0, h), second interval be [h, 2h), 3rd interval be [2h, 3h), the processing
Device according to screen size R, intelligent apparatus weight M, 701 weight N of Touch Screen, export h as follows:
Wherein, the unit of the screen size R is inch；
The intelligent apparatus weight M, 701 weight N of Touch Screen unit be gram；
The unit of longitudinal movement distance d, h is 0.1 millimeter.
The present invention can show in above procedure the pressing size of Touch Screen 701 based on user through the above way
Whole process, also, can according to the concrete condition of intelligent apparatus it is customized pressing Touch Screen 701 compression distance, wherein
Since the present invention is the pressing distance resetted according to the rebound of spring, then can illustrate, pressing distance of the invention is also
Reflect pressing dynamics, the present invention is pressing dynamics based on user or pressing distance gradually to show each of above-mentioned steps
Image facilitates experimental study so that user clearly be allowed to see all images before and after treatment.
Wherein, the separation being equipped in the database between the first skewback 705 corresponding with a, b, the second skewback 706
Angle c.Also, there is d corresponding with the c in the database.
Wherein, the oval hemisphere block 704 is an oval sphere or rugby shape sphere, is cut along long axis direction
It is one of at two pieces.
Wherein, it is triangle, 1/4 ellipse, fan-shaped, inclined-plane that first skewback 705, the second skewback 706, which can be section,
Recessed triangle, so that deflecting, and above-mentioned ellipse when the oval face contact of its inclined-plane and oval hemisphere block 704
Hemisphere block 704 compared to taper, circle, block and above-mentioned first skewback 705 of triangle, the second skewback 706 contact when, can
When oval hemisphere block 704 is not vertical downward movement, for two the first skewbacks 705 unevenly moved, the second skewbacks 706
Certain movement compensation is carried out, so that distance separated between the first skewback 705, the second skewback 706 be made to meet certain pressing dynamics
Rule.For example, oval hemisphere block 704 of the invention by the power of 20N be pressed and drive vertically downward the first skewback 705,
The angle X and oval hemisphere block 704 that second skewback 706 symmetrically opens outward are by 20N and non-perpendicular, inclined quilt downwards
Press and drive the first skewback 705, the outside asymmetrical opening angle Y of the second skewback 706 similar or identical, i.e. X is equal to or about
Equal to Y.Compared to taper, the oval hemisphere block 704 of triangle, the gap between the X and Y can become larger.
Wherein the first skewback 705 of 1/4 ellipse, the section of the second skewback 706 can be understood as the convex triangle in inclined-plane
Shape.
The present invention driven by above-mentioned oval hemisphere block 704 first skewback 705, the second skewback 706 separately can compensate for by
Seam bring Touch Screen 701 between Touch Screen 701, shell 702 can be since user be by pressure when moving down
The angle driving of power with the lower surface of shell 702 and non-perpendicular have an effect and the problem of non-perpendicular movement.
Wherein, the intelligent apparatus may include desktop computer, laptop, tablet computer, mobile phone, smartwatch.
More than, according to the image partition method and device suitable for big data image of the embodiment of the present disclosure, and calculate
Machine readable storage medium storing program for executing carries out coarse segmentation to cell image using maximum between-cluster variance algorithm for cervical exfoliated cell image,
Obtain individual cells, population of cells and impurity block image；Then shape test, such as test-strips are carried out to the result divided above
Part passes through, then area-of-interest is cell image, into characteristic extracting module；Unacceptable area-of-interest is tested, that is, having can
It can be population of cells or impurity block group, then based on gradient vector flow-active contour model dividing method to region of interest
Domain carries out secondary splitting, then carries out shape test to secondary splitting result.Thus, it is possible to efficiently, precisely, quickly be directed to palace
Neck cast-off cells image realizes image segmentation strategy, and has ease for use and applicability.The embodiment of the invention also provides one kind
Computer readable storage medium is stored with computer program on the computer readable storage medium, and the program is by processor in Fig. 1
It executes.
The embodiment of the invention also provides a kind of computer program products comprising instruction.When the computer program product exists
When being run on computer, so that the method that computer executes above-mentioned Fig. 1.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the program can be stored in a computer-readable storage medium
In, the program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein, the storage medium can be magnetic
Dish, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access
Memory, RAM) etc..
Each technical characteristic of embodiment described above can be combined arbitrarily, for simplicity of description, not to above-mentioned reality
It applies all possible combination of each technical characteristic in example to be all described, as long as however, the combination of these technical characteristics is not deposited
In contradiction, all should be considered as described in this specification.
The embodiments described above only express several embodiments of the present invention, and the description thereof is more specific and detailed, but simultaneously
Limitations on the scope of the patent of the present invention therefore cannot be interpreted as.It should be pointed out that for those of ordinary skill in the art
For, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to guarantor of the invention
Protect range.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
The basic principle of the disclosure is described in conjunction with specific embodiments above, however, it is desirable to, it is noted that in the disclosure
The advantages of referring to, advantage, effect etc. are only exemplary rather than limitation, must not believe that these advantages, advantage, effect etc. are the disclosure
Each embodiment is prerequisite.In addition, detail disclosed above is merely to exemplary effect and the work being easy to understand
With, rather than limit, it is that must be realized using above-mentioned concrete details that above-mentioned details, which is not intended to limit the disclosure,.
Device involved in the disclosure, device, equipment, system block diagram only as illustrative example and being not intended to
It is required that or hint must be attached in such a way that box illustrates, arrange, configure.As those skilled in the art will appreciate that
, it can be connected by any way, arrange, configure these devices, device, equipment, system.Such as "include", "comprise", " tool
" etc. word be open vocabulary, refer to " including but not limited to ", and can be used interchangeably with it.Vocabulary used herein above
"or" and "and" refer to vocabulary "and/or", and can be used interchangeably with it, unless it is not such that context, which is explicitly indicated,.Here made
Vocabulary " such as " refers to phrase " such as, but not limited to ", and can be used interchangeably with it.
In addition, as used herein, the "or" instruction separation used in the enumerating of the item started with "at least one"
It enumerates, such as enumerating for " at least one of A, B or C " means A or B or C or AB or AC or BC or ABC (i.e. A and B and C).
In addition, wording " exemplary " does not mean that the example of description is preferred or more preferable than other examples.
It may also be noted that in the system and method for the disclosure, each component or each step are can to decompose and/or again
Combination nova.These decompose and/or reconfigure the equivalent scheme that should be regarded as the disclosure.
The technology instructed defined by the appended claims can not departed from and carried out to the various of technology described herein
Change, replace and changes.In addition, the scope of the claims of the disclosure is not limited to process described above, machine, manufacture, thing
Composition, means, method and the specific aspect of movement of part.Can use carried out to corresponding aspect described herein it is essentially identical
Function or realize essentially identical result there is currently or later to be developed processing, machine, manufacture, event group
At, means, method or movement.Thus, appended claims include such processing, machine, manufacture, event within its scope
Composition, means, method or movement.
The above description of disclosed aspect is provided so that any person skilled in the art can make or use this
It is open.Various modifications in terms of these are readily apparent to those skilled in the art, and are defined herein
General Principle can be applied to other aspect without departing from the scope of the present disclosure.Therefore, the disclosure is not intended to be limited to
Aspect shown in this, but according to principle disclosed herein and the consistent widest range of novel feature.
Above description is had been presented for for purposes of illustration and description.In addition, this description is not intended to the reality of the disclosure
It applies example and is restricted to form disclosed herein.Although already discussed above multiple exemplary aspects and embodiment, this field skill
Its certain modifications, modification, change, addition and sub-portfolio will be recognized in art personnel.
Claims (10)
1. a kind of image partition method suitable for big data image, which comprises the following steps:
Received multiple cell images are divided；
According to maximum between-cluster variance algorithm, coarse segmentation is carried out for the area-of-interest in the multiple cell image after division
Operation；
It is suitable for the result obtained after coarse segmentation operation the image segmentation operations of cell image by shape test completion.
2. the image partition method according to claim 1 suitable for big data image, which is characterized in that described pair of reception
Multiple cell images carry out divide include:
The statistics rule of cell image data based on training sample, selection segmentation channel；
The segmentation threshold in the segmentation channel is chosen, and foreground and background segmentation is carried out to the cell image；
According to the foreground pixel and background pixel progress connection regional analysis after segmentation, qualified cell compartment is obtained.
3. the image partition method according to claim 2 suitable for big data image, which is characterized in that further include:
The division of the cell image sub-block is carried out with N row N column division mode to the qualified cell compartment of acquisition, wherein
The N row is classified as equal numerical value with the N.
4. the image partition method according to claim 1 suitable for big data image, which is characterized in that described to rough segmentation
It includes: to grasp to coarse segmentation that the result obtained after operation, which is cut, by the figure cutting operation that shape test is completed to be suitable for cell image
The result obtained after work tests test by area and completes the figure cutting operation for being suitable for cell image, wherein the area
Test is to judge whether the pixel number of the area-of-interest meets the pixel threshold interval of default normal cell area.
5. the image partition method according to claim 1 suitable for big data image, which is characterized in that described to rough segmentation
It includes: to grasp to coarse segmentation that the result obtained after operation, which is cut, by the figure cutting operation that shape test is completed to be suitable for cell image
The result obtained after work completes the figure cutting operation for being suitable for cell image by the test of deformity degree, wherein the deformity degree
Test is to pass through simple deformity degree calculation formula γ=l/NpCalculate the lopsided degree of the area-of-interest, wherein l is the sense
The perimeter in interest region, NpFor the pixel number of the area-of-interest.
6. the image partition method according to claim 5 suitable for big data image, which is characterized in that further include: it is pre-
Deformity degree high threshold γ is first setT；
As γ≤γTWhen, determine to test the result obtained after coarse segmentation operation by deformity degree.
7. the image partition method according to claim 6 suitable for big data image, which is characterized in that further include: when
γ > γTWhen, secondary coarse segmentation behaviour is carried out to area-of-interest based on gradient vector flow-active contour model dividing method
Make；
It is suitable for the result obtained after the operation of secondary coarse segmentation the image segmentation operations of cell image by shape test completion.
8. a kind of intelligent apparatus, is stored thereon with computer program, which is characterized in that the program realizes institute when being executed by processor
The step of stating any one of claim 1-7 the method；
Wherein, the intelligent apparatus further include Touch Screen (701), shell (702), spring (703), oval hemisphere block (704),
First skewback (705), the second skewback (706), first axle (707), range finder module (708), the first permanent magnetic iron block (711), second
Permanent magnetic iron block (712), the lower surface of the Touch Screen (701) and the surface of oval hemisphere block (704) are fixed, and described oval half
The positive decentralization of ball block (704) is equipped with the first skewback (705), the second skewback (706), first skewback (705), the second skewback
(706) one end passes through bearing and is mounted on first axle (707), first skewback (705), the second skewback (706) it is another
One end is located at the underface at described oval hemisphere block (704) center, under the inside of the first axle (707) and shell (702)
Side is fixed, first skewback (705), the second skewback (706) the other end be respectively fixed with the first permanent magnetic iron block (711),
Two permanent magnetic iron blocks (712), wherein mutual attracted, the shell of first permanent magnetic iron block (711), the second permanent magnetic iron block (712)
(702) two opposite side surfaces of inside are respectively fixed with range finder module (708), two range finder modules (708) point
Yong Yu not detect first skewback (705), the second skewback (706) moving distance a, b and be transmitted to processor, the processing
Device converts the separation angle c between first skewback (705), the second skewback (706) for described a, b according to database, and
The longitudinal movement distance d of the oval hemisphere block (704) is determined according to the database according to the separation angle c；
Wherein, the processor is equipped with first distance section, second distance section, third apart from section for d,
When the d is first distance section, then the Touch Screen (701) shows archaeocyte image；
When the d is second distance section, then the Touch Screen (701) shows the cell image after coarse segmentation operation；
When the d be third apart from section when, then the Touch Screen (701) shows that shape test is completed to be suitable for cell image
Cell image；
Wherein, the first interval be [0, h), second interval be [h, 2h), 3rd interval be [2h, 3h), the processor root
According to screen size R, intelligent apparatus weight M, Touch Screen (701) weight N, h is exported as follows:
Wherein, the unit of the screen size R is inch；
The intelligent apparatus weight M, Touch Screen (701) weight N unit be gram；
The unit of longitudinal movement distance d, h is 0.1 millimeter.
9. a kind of computer equipment including memory, processor and stores the meter that can be run on a memory and on a processor
Calculation machine program, which is characterized in that the processor realizes side described in any one of described claim 1-7 when executing described program
The step of method.
10. a kind of image segmentation device suitable for cell image, which is characterized in that described device includes:
Division module, for being divided to received multiple cell images；
Coarse segmentation module is used for according to maximum between-cluster variance algorithm, emerging for the sense in the multiple cell image after division
Interesting region carries out coarse segmentation operation；
Divide module, the result for obtaining after operating to coarse segmentation completes the image for being suitable for cell image by shape test
Cutting operation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810943197.0A CN109191470A (en) | 2018-08-18 | 2018-08-18 | Image partition method and device suitable for big data image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810943197.0A CN109191470A (en) | 2018-08-18 | 2018-08-18 | Image partition method and device suitable for big data image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109191470A true CN109191470A (en) | 2019-01-11 |
Family
ID=64918300
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810943197.0A Pending CN109191470A (en) | 2018-08-18 | 2018-08-18 | Image partition method and device suitable for big data image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109191470A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111458269A (en) * | 2020-05-07 | 2020-07-28 | 厦门汉舒捷医疗科技有限公司 | Artificial intelligent identification method for peripheral blood lymph micronucleus cell image |
| CN112378837A (en) * | 2020-09-15 | 2021-02-19 | 深圳市华中生物药械有限公司 | Cervical exfoliated cell detection method and related device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100014755A1 (en) * | 2008-07-21 | 2010-01-21 | Charles Lee Wilson | System and method for grid-based image segmentation and matching |
| CN101877074A (en) * | 2009-11-23 | 2010-11-03 | 常州达奇信息科技有限公司 | Tubercle bacillus target recognizing and counting algorithm based on diverse characteristics |
| CN102831607A (en) * | 2012-08-08 | 2012-12-19 | 深圳市迈科龙生物技术有限公司 | Method for segmenting cervix uteri liquid base cell image |
| CN105183359A (en) * | 2015-09-09 | 2015-12-23 | 魅族科技（中国）有限公司 | Zooming-in and zooming-out method and device and terminal |
| CN107527028A (en) * | 2017-08-18 | 2017-12-29 | 深圳乐普智能医疗器械有限公司 | Target cell recognition methods, device and terminal |
| CN107808381A (en) * | 2017-09-25 | 2018-03-16 | 哈尔滨理工大学 | A kind of unicellular image partition method |
-
2018
- 2018-08-18 CN CN201810943197.0A patent/CN109191470A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100014755A1 (en) * | 2008-07-21 | 2010-01-21 | Charles Lee Wilson | System and method for grid-based image segmentation and matching |
| CN101877074A (en) * | 2009-11-23 | 2010-11-03 | 常州达奇信息科技有限公司 | Tubercle bacillus target recognizing and counting algorithm based on diverse characteristics |
| CN102831607A (en) * | 2012-08-08 | 2012-12-19 | 深圳市迈科龙生物技术有限公司 | Method for segmenting cervix uteri liquid base cell image |
| CN105183359A (en) * | 2015-09-09 | 2015-12-23 | 魅族科技（中国）有限公司 | Zooming-in and zooming-out method and device and terminal |
| CN107527028A (en) * | 2017-08-18 | 2017-12-29 | 深圳乐普智能医疗器械有限公司 | Target cell recognition methods, device and terminal |
| CN107808381A (en) * | 2017-09-25 | 2018-03-16 | 哈尔滨理工大学 | A kind of unicellular image partition method |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111458269A (en) * | 2020-05-07 | 2020-07-28 | 厦门汉舒捷医疗科技有限公司 | Artificial intelligent identification method for peripheral blood lymph micronucleus cell image |
| CN112378837A (en) * | 2020-09-15 | 2021-02-19 | 深圳市华中生物药械有限公司 | Cervical exfoliated cell detection method and related device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106056155B (en) | Superpixel segmentation method based on boundary information fusion | |
| CN105744256B (en) | Based on the significant objective evaluation method for quality of stereo images of collection of illustrative plates vision | |
| CN105160317B (en) | One kind being based on area dividing pedestrian gender identification method | |
| CN110827193B (en) | Panoramic video significance detection method based on multichannel characteristics | |
| CN103295025B (en) | A kind of automatic selecting method of three-dimensional model optimal view | |
| CN102509104B (en) | Confidence map-based method for distinguishing and detecting virtual object of augmented reality scene | |
| CN110428432A (en) | The deep neural network algorithm of colon body of gland Image Automatic Segmentation | |
| CN108416266A (en) | A kind of video behavior method for quickly identifying extracting moving target using light stream | |
| CN110147721A (en) | A kind of three-dimensional face identification method, model training method and device | |
| CN105389554A (en) | Face-identification-based living body determination method and equipment | |
| CN104361313A (en) | Gesture recognition method based on multi-kernel learning heterogeneous feature fusion | |
| CN113762138B (en) | Identification method, device, computer equipment and storage medium for fake face pictures | |
| CN103605986A (en) | Human motion recognition method based on local features | |
| CN103996018A (en) | Human-face identification method based on 4DLBP | |
| CN102147852A (en) | Method for detecting hair area | |
| CN106156798B (en) | Scene image classification method based on annular space pyramid and Multiple Kernel Learning | |
| CN101655913A (en) | Computer generated image passive detection method based on fractal dimension | |
| CN108985200A (en) | A kind of In vivo detection algorithm of the non-formula based on terminal device | |
| CN106056631A (en) | Pedestrian detection method based on motion region | |
| CN104834909B (en) | A kind of new image representation method based on Gabor comprehensive characteristics | |
| CN109191470A (en) | Image partition method and device suitable for big data image | |
| CN105069745A (en) | face-changing system based on common image sensor and enhanced augmented reality technology and method | |
| CN115240119A (en) | Pedestrian small target detection method in video monitoring based on deep learning | |
| CN106407916A (en) | Distributed face recognition method, apparatus and system | |
| CN105740787A (en) | Face recognition method based on multi-kernel authentication color space |
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: 20190111 |
|
| RJ01 | Rejection of invention patent application after publication |