CN109978843A - Image partition method, system, computer equipment and the storage medium of wafer dopant - Google Patents
Image partition method, system, computer equipment and the storage medium of wafer dopant Download PDFInfo
- Publication number
- CN109978843A CN109978843A CN201910196050.4A CN201910196050A CN109978843A CN 109978843 A CN109978843 A CN 109978843A CN 201910196050 A CN201910196050 A CN 201910196050A CN 109978843 A CN109978843 A CN 109978843A
- Authority
- CN
- China
- Prior art keywords
- image
- evolution
- wafer
- indicate
- level set
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/0008—Industrial image inspection checking presence/absence
-
- 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
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30148—Semiconductor; IC; Wafer
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Image Processing (AREA)
- Testing Or Measuring Of Semiconductors Or The Like (AREA)
Abstract
The invention discloses image partition method, system, computer equipment and the storage mediums of a kind of wafer dopant, which comprises obtains wafer image;Construct domain type variation level set model;By wafer image input area type variation level set model, wafer image is split using the total energy functional of domain type variation level set model, output obtains the image of wafer dopant;The system comprises obtain module, building module and segmentation module.The present invention by the wafer image of noise effect by that will be inputted the domain type variation level set model of building, the wafer image is split using the total energy functional of domain type variation level set model, it can accurately be partitioned into wafer dopant from the wafer image, and be conducive to improve the low image segmentation precision of contrast.
Description
Technical field
The present invention relates to image partition method, system, computer equipment and the storage mediums of a kind of wafer dopant, belong to
Technical field of image segmentation.
Background technique
Wafer is the basic material for manufacturing semiconductor chip.It can be processed on a wafer and is fabricated to various circuit element knots
Structure, and become the IC product for having specific electrical functionality.According to SiC, GaN, the excellent properties of ZnO material device, the third generation is partly led
Body is rapidly growing, especially nitride material, they have big band gap (forbidden band usually 2eV or more), high breakdown electric field,
The advantages that high electronics saturation drift velocity, strong Radiation hardness, solves the problems, such as white-light illuminating, and be suitble to develop high
All kinds of electronic devices to work under temperature, high-power component and particular surroundings.
The wafer of manufacture chip is all through overdoping, by completing doping with thermal diffusion or ion implantation technology.Pass through
Doping techniques generate important P-N junction in semiconductor material, but introduce some new defects either compound simultaneously, cause
Epitaxial growth mode and surface topography are by Different Effects.SiC semiconductor material realizes conductive N-type material by doping techniques
Material.GaN semiconductor obtains reliable and stable N-type or P-type material by incorporation alms giver or acceptor atom.Mg atom is most suitable at present
Close the p-type doped chemical of GaN.Mg atom can generate some complex compounds in GaN growth, such as Mg-H complex compound, complex compound pair
Mg atom plays passivation, and Mg atom is made to be difficult to activate ionization.In addition, high Mg source flux can seriously affect epitaxial film surface shape
Looks make GaN rough surface, and under higher Mg doping, GaN can generate reversed farmland, change polarity.Si and Ge is that GaN is partly led at present
The n-type doping of body is easier the foreign atom realized.However, high silicon doping can significantly affect the surface topography of GaN epitaxy film,
Lead to rough surface, influence of the Ge doping to extension membrane strain is smaller.
Doped chemical and concentration all have an impact to oxidation growth rate.For example, the silicon face of high-dopant concentration is than low-mix
The silicon surface oxidation rate of miscellaneous concentration is fast.And the oxide layer on the silicon face of high-dopant concentration is than in other layers of upper oxygen grown
The density for changing layer is low.After the completion of oxidation, the distribution of foreign atom also has an impact to oxidation growth rate in silicon.For example, N-type is mixed
Sundries (P, As, Sb) they in silicon than there is higher solubility in silica.When oxide layer meets them, these are miscellaneous
Matter will enter in silicon, and N-type dopant is between silicon and silica than there is higher density (referred to as silica in crystal
Row's phosphorus effect).When dopant is boron (B) element of P-type material, opposite result will be generated.I.e. boron atom is drawn into
Silicon dioxide layer, causing to be run out in the silicon atom of SiO2 and Si intersection by boron atom, (referred to as the suction boron of silica is made
With).
In existing image capture device, component part includes industrial personal computer, microscope detection platform, the survey of zoom microscopy
Platform, XYZ axis precision movement platform and article carrying platform.Wherein, microscope detection platform is to utilize metallography microscope sem observation gold
Belong to the metallographic structure with opaque articles such as mineral.Illuminating bundle is mapped to observed object from object lens direction in metallographic microscope
Surface, returns again to object lens imaging after being reflected by object plane, detection accuracy can reach micro/nano level.This indirect illumination mode is used extensively
In the detection work of integrated circuit silicon chip.
Summary of the invention
The first purpose of this invention is to provide a kind of image partition method of wafer dopant, this method can from by
It is accurately partitioned into wafer dopant in the wafer image of noise effect, and is conducive to improve the low image segmentation essence of contrast
Degree.
Second object of the present invention is to provide a kind of image segmentation system of wafer dopant.
Third object of the present invention is to provide a kind of computer equipment.
Fourth object of the present invention is to provide a kind of storage medium.
The first purpose of this invention can be reached by adopting the following technical scheme that:
A kind of image partition method of wafer dopant, which comprises
Obtain wafer image;
Construct domain type variation level set model;
By wafer image input area type variation level set model, the total energy of domain type variation level set model is utilized
Functional is split wafer image, and output obtains the image of wafer dopant.
Further, described by wafer image input area type variation level set model, utilize domain type variation level set
The total energy functional of model is split wafer image, and output obtains the image of wafer dopant, specifically includes:
By wafer image input area type variation level set model;
According to wafer image, the parameters and clusters number and closure evolution extra curvature of setting total energy functional
The image entropy window size in portion;
Initialize level set function;
Calculate cluster centre point value, be closed the image entropy of evolution curved exterior, evolution curvilinear inner weight coefficient and
The weight coefficient of evolution curved exterior;
Calculate the image match value of evolution curvilinear inner match value and evolution curved exterior;
Update level set function;
Judge whether evolution curve is stable, if evolution curve is stablized, segmentation terminates, and output obtains the figure of wafer dopant
Picture recalculates the weight coefficient of cluster centre point value, local entropy information, evolution curvilinear inner if evolution curve is unstable
And the weight coefficient of evolution curved exterior, until evolution curve is stablized.
Further, the total energy functional, such as following formula:
Wherein, Ω indicates image space;φ (x, y) indicates level set function;u0(x, y) indicates image slices vegetarian refreshments;
Indicate evolution line smoothing item, u1Indicate the weight coefficient of the smooth item;
Indicate that region slickness keeps item, u2Indicate the weight coefficient of the holding item;
Indicate evolution curve fidelity term;
Indicate that regular terms, w indicate the weight coefficient of the regular terms.
Further, the H kept in item and fidelity termε(φ) indicates regularization smoothing procedure function, the δ in the smooth itemε
(φ) representative function HεThe derivative of (φ), wherein ε is constant, Hε(φ) and δε(φ) such as following formula:
Further, in the fidelity term,Indicate the fidelity term of evolution curvilinear inner,Indicate the fidelity term of evolution curved exterior, D1Indicate the weight system of evolution curvilinear inner
Number, D2Indicate the weight coefficient of evolution curved exterior;
In the fidelity term of the evolution curvilinear inner, c1iIt indicates evolution curvilinear inner match value, calculates such as following formula:
c1i=α1ki+(1-α1)u0 *
Wherein, kiIndicate that the i-th class clusters centerpoint value, i=1,2 ..., K, u0 *Indicate original image after mean filter
Image information, α1Indicate weight, value range 0-1;
In the fidelity term of the evolution curved exterior, c2It indicates the image match value of evolution curved exterior, calculates as follows
Formula:
Wherein, τoutIndicate the image entropy of closure evolution curved exterior, such as following formula:
Wherein, N indicates the tonal gradation in image;PiIt indicates the pixel percentage in the picture that gray value is i, takes
Value range is 0-1, α2Indicate image entropy coefficient, α2≥0。
Further, in evolution curvilinear inner, D1=max | u0(x,y)-m1|, m1In gray scale for evolution curvilinear inner
Value;In evolution curved exterior, D2=max | u0(x,y)-m2|, m2For the gray scale intermediate value of evolution curved exterior.
Further, the total energy functional substitutes into Euler's formula, and Introduction Time variable utilizes gradient descent method minimum
Change total energy functional, obtain such as sub-level set calculation formula:
Wherein, t is time variable.
Second object of the present invention can be reached by adopting the following technical scheme that:
A kind of image segmentation system of wafer dopant, the system comprises:
Module is obtained, for obtaining wafer image;
Module is constructed, domain type variation level set model is constructed;
Divide module, for utilizing domain type variation level set for wafer image input area type variation level set model
The total energy functional of model is split wafer image, and output obtains the image of wafer dopant.
Third object of the present invention can be reached by adopting the following technical scheme that:
A kind of computer equipment, including processor and for the memory of storage processor executable program, the place
When managing the program of device execution memory storage, above-mentioned image partition method is realized.
Fourth object of the present invention can be reached by adopting the following technical scheme that:
A kind of storage medium is stored with program, when described program is executed by processor, realizes above-mentioned image segmentation side
Method.
The present invention have compared with the existing technology it is following the utility model has the advantages that
1, the present invention is utilized by that will be inputted the domain type variation level set model of building by the wafer image of noise effect
The total energy functional of domain type variation level set model is split the wafer image, can be accurate from the wafer image
It is partitioned into wafer dopant, and is conducive to improve the low image segmentation precision of contrast.
2, the present invention can be according to maximum median absolute deviation automatic adjusument inside or outside of curve Evolution Rates, in image-region
Noise have higher steady degree.
3, the present invention utilizes cluster centre point value and filtered image weighted sum in evolution curvilinear inner, bent as developing
Match value inside line is conducive to improve the low image segmentation precision of contrast.
4. the present invention introduces supplement of the image entropy as grayscale information in evolution curved exterior, be conducive to improve contrast
Low image segmentation precision.
5, in evolutionary process, domain type variation level set model remains the slickness of evolution curve and puts down the present invention
Stability and gradient modulus value is always 1, avoids reinitializing, and improves and calculates the time.
Detailed description of the invention
In order to illustrate more clearly of technical solution of the present invention, attached drawing needed in embodiment will be made below
Simply introduce, it should be apparent that, the accompanying drawings in the following description is only some embodiments of the present invention, general for this field
For logical technical staff, without creative efforts, it is also possible to obtain other drawings based on these drawings.
Fig. 1 is the flow chart of the image partition method of the wafer dopant of the embodiment of the present invention 1.
Fig. 2 is the total energy functional using domain type variation level set model of the embodiment of the present invention 1 to wafer image
The flow chart being split.
Fig. 3 is the structural block diagram of the image segmentation system of the wafer dopant of the embodiment of the present invention 2.
Fig. 4 is the structural block diagram for dividing module in the image segmentation system of the wafer dopant of the embodiment of the present invention 2.
Fig. 5 is the computer equipment structural block diagram of the embodiment of the present invention 3.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Embodiment 1:
As shown in Figure 1, present embodiments providing a kind of image partition method of wafer dopant, this method includes following step
It is rapid:
S101, wafer image is obtained.
The present embodiment acquires wafer image using metallographic microscope amplification acquisition, according to dopant in the wafer image
Be of different shades, and wafer image gray scale is uneven and affected by noise, it is therefore desirable to be split.
S102, building domain type variation level set model.
The total energy functional of domain type variation level set model, such as following formula:
Wherein, Ω indicates image space;φ (x, y) indicates level set function, generally selection symbolic measurement;u0(x,
Y) image slices vegetarian refreshments is indicated;
Indicate that evolution line smoothing item, the smooth item make zero level collection as short as possible and smooth, u1
Indicate the weight coefficient of the smooth item;
It indicates that region slickness keeps item, acts on as holding area slickness, u2Indicate the holding item
Weight coefficient;
Indicate evolution curve fidelity term, the guarantor
True item makes the image after segmentation as close with original image as possible;
Indicate regular terms, the effect of the regular terms is to guarantee that level set can during curve evolvement
It is micro-, and remain | Δ φ |=1, avoid level set from reinitializing, w indicates the weight coefficient of the regular terms.
H in above-mentioned holding item and fidelity termε(φ) indicates regularization smoothing procedure function, the δ in above-mentioned smooth itemε(φ) is indicated
Function HεThe derivative of (φ), wherein ε is constant, Hε(φ) and δε(φ) such as following formula:
In above-mentioned fidelity term,Indicate the fidelity term of evolution curvilinear inner,Indicate the fidelity term of evolution curved exterior, D1Indicate the weight system of evolution curvilinear inner
Number, D2Indicate the weight coefficient of evolution curved exterior.
In the fidelity term of evolution curvilinear inner, c1iIt indicates evolution curvilinear inner match value, calculates such as following formula:
c1i=α1ki+(1-α1)u0 * (3)
Wherein, kiIndicate that the i-th class clusters centerpoint value, which obtains after can use K-means cluster,
I=1,2 ..., K, u0 *Indicate image information of the original image after mean filter, α1Indicate weight, value range 0-1.
From the foregoing it will be seen that the present embodiment using the image after cluster centre point value and mean filter weight and
It is low for wafer dopant picture contrast as evolution curvilinear inner match value, at it close to the isolated background dot of edge
There is very high similitude with target cluster centre point, evolution curvilinear inner match value combines the image after mean filter, has
Effect reduces model to the dependence of cluster result, and isolated background dot is prevented accidentally to be divided into the point in foreground image.
In the fidelity term of evolution curved exterior, c2It indicates the image match value of evolution curved exterior, calculates such as following formula:
Wherein, τoutIndicate the image entropy of closure evolution curved exterior, such as following formula:
Wherein, N indicates the tonal gradation in image;PiIt indicates the pixel percentage in the picture that gray value is i, takes
Value range is 0-1, α2Indicate image entropy coefficient, α2≥0。
From the foregoing it will be seen that the fidelity term of the evolution curved exterior of the present embodiment, introduces image entropy as original image
As the supplement of grayscale information, image entropy characterizes the distribution character and Texture complication of image, and image entropy is bigger, image texture information
It is abundanter, in the image that contrast reduces, be conducive to improve image segmentation precision.
Further, the present embodiment is using maximum median absolute deviation, it may be assumed that in evolution curvilinear inner, D1=max | u0(x,y)-
m1|, m1For the gray scale intermediate value of evolution curvilinear inner;In evolution curved exterior, D2=max | u0(x,y)-m2|, m2For evolution curve
External gray scale intermediate value;Compared with gray average, gray scale intermediate value is higher to the noise spot stationarity in region, D1、D2It can be adaptive
Adjust inside and outside curve evolvement speed.
Above-mentioned total energy functional substitutes into Euler's formula, and Introduction Time variable t minimizes total physical efficiency using gradient descent method
Functional is measured, is obtained such as sub-level set calculation formula:
S103, by wafer image input area type variation level set model, utilize the total of domain type variation level set model
Body energy functional is split wafer image, and output obtains the image of wafer dopant.
Step S103 is as shown in Fig. 2, specifically include:
S1031, by wafer image input area type variation level set model.
S1032, according to wafer image, the parameters of total energy functional, such as u are set1、u2、w、α1、α2, time step
The image entropy window size of long and clusters number and closure evolution curved exterior.
S1033, initialization level set function.
S1034, cluster centre point value is calculated, according to above formula (5), calculates the image entropy of closure evolution curved exterior, according to
Maximum median absolute deviation calculates the weight coefficient of evolution curvilinear inner and the weight coefficient of evolution curved exterior.
S1035, according to above formula (3), calculate evolution curvilinear inner match value;According to above formula (4), evolution curved exterior is calculated
Image match value.
S1036, level set function is updated.
S1037, judge whether evolution curve is stable.
Specifically, judge whether evolution curve is stable, can be by increasing the number of iterations, whether observation evolution curve occurs
Concussion, it is as unstable if shaking;If not shaking, that is, maintain at object boundary, it is as stable.
If evolution curve is stablized, segmentation terminates, and output obtains the image of wafer dopant, if evolution curve is unstable,
Then return step S1034.
It will be understood by those skilled in the art that realizing that all or part of the steps in the method for above-described embodiment can pass through
Program is completed to instruct relevant hardware, and corresponding program can store in computer readable storage medium.
It should be noted that this is not although describing the method operation of above-described embodiment in the accompanying drawings with particular order
It is required that hint must execute these operations in this particular order, could be real or have to carry out shown in whole operation
Existing desired result.On the contrary, the step of describing can change and execute sequence.Additionally or alternatively, it is convenient to omit certain steps,
Multiple steps are merged into a step to execute, and/or a step is decomposed into execution of multiple steps.
Embodiment 2:
As shown in figure 3, present embodiments providing a kind of image segmentation system of wafer dopant, which includes obtaining mould
Block 301, building module 302 and segmentation module 303, the concrete function of modules are as follows:
The acquisition module 301, for obtaining wafer image.
The building module 302 constructs domain type variation level set model.
The segmentation module 303, for utilizing domain type variation for wafer image input area type variation level set model
The total energy functional of Level Set Models is split wafer image, and output obtains the image of wafer dopant.
Further, segmentation module 303 is as shown in figure 4, specifically include:
Input unit 3031 is used for wafer image input area type variation level set model.
Setting unit 3032, for the parameters and clusters number of total energy functional to be arranged according to wafer image
With the image entropy window size of closure evolution curved exterior.
Initialization unit 3033, for initializing level set function.
First computing unit 3034, for calculating, cluster centre point value, the image entropy for being closed evolution curved exterior, develop song
The weight coefficient of weight coefficient and evolution curved exterior inside line.
Second computing unit 3035, the image for calculating evolution curvilinear inner match value and evolution curved exterior are fitted
Value.
Updating unit 3036, for updating level set function.
Judging unit 3037, for judging whether evolution curve is stable, if evolution curve is stablized, segmentation terminates, and exports
The image of wafer dopant is obtained, if evolution curve is unstable, cluster centre point value is recalculated, local entropy information, develops
The weight coefficient of curvilinear inner and the weight coefficient of evolution curved exterior, until evolution curve is stablized.
The specific implementation of modules may refer to above-described embodiment 1 in the present embodiment, and this is no longer going to repeat them;It needs
Illustrate, system provided in this embodiment only the example of the division of the above functional modules, in practical applications,
It can according to need and be completed by different functional modules above-mentioned function distribution, i.e., internal structure is divided into different functions
Module, to complete all or part of the functions described above.
It is appreciated that term " first ", " second " used in the system of the present embodiment etc. can be used for describing various lists
Member, but these units should not be limited by these terms.These terms are only used to distinguish first unit and another unit.Citing
For, without departing from the scope of the invention, the first computing unit can be known as the second computing unit, and similar
Second computing unit, can be known as the first computing unit by ground, and the first computing unit and the second computing unit both calculate list
Member, but it is not same computing unit.
Embodiment 3:
A kind of computer equipment is present embodiments provided, which can be the industry control in image capture device
The structure of machine, industrial personal computer is as shown in Figure 5 comprising processor 502, memory, the input unit connected by system bus 501
503, display 504 and network interface 505, which calculates for offer and control ability, the memory include non-volatile
Property storage medium 506 and built-in storage 507, the non-volatile memory medium 506 are stored with operating system, computer program sum number
According to library, which provides environment for the operation of operating system and computer program in non-volatile memory medium, place
When managing the computer program of the execution memory storage of device 502, the image partition method of above-described embodiment 1 is realized, as follows:
Obtain wafer image;
Construct domain type variation level set model;
By wafer image input area type variation level set model, the total energy of domain type variation level set model is utilized
Functional is split wafer image, and output obtains the image of wafer dopant.
Embodiment 4:
A kind of storage medium is present embodiments provided, which is computer readable storage medium, is stored with meter
Calculation machine program when described program is executed by processor, when processor executes the computer program of memory storage, realizes above-mentioned reality
The image partition method of example 1 is applied, as follows:
Obtain wafer image;
Construct domain type variation level set model;
By wafer image input area type variation level set model, the total energy of domain type variation level set model is utilized
Functional is split wafer image, and output obtains the image of wafer dopant.
Storage medium described in the present embodiment can be disk, CD, computer storage, read-only memory (ROM,
Read-Only Memory), random access memory (RAM, Random Access Memory), USB flash disk, mobile hard disk etc. be situated between
Matter.
In conclusion the present invention by the wafer image of noise effect by that will be inputted the domain type variation level set mould of building
Type is split the wafer image using the total energy functional of domain type variation level set model, can be from the wafer figure
It is accurately partitioned into wafer dopant as in, and is conducive to improve the low image segmentation precision of contrast.
The above, only the invention patent preferred embodiment, but the scope of protection of the patent of the present invention is not limited to
This, anyone skilled in the art is in the range disclosed in the invention patent, according to the present invention the skill of patent
Art scheme and its inventive concept are subject to equivalent substitution or change, belong to the scope of protection of the patent of the present invention.
Claims (10)
1. a kind of image partition method of wafer dopant, which is characterized in that the described method includes:
Obtain wafer image;
Construct domain type variation level set model;
By wafer image input area type variation level set model, the total energy functional of domain type variation level set model is utilized
Wafer image is split, output obtains the image of wafer dopant.
2. image partition method according to claim 1, which is characterized in that described by wafer image input area type variation
Level Set Models are split wafer image using the total energy functional of domain type variation level set model, and output obtains
The image of wafer dopant, specifically includes:
By wafer image input area type variation level set model;
According to wafer image, parameters and clusters number and the closure evolution curved exterior of setting total energy functional
Image entropy window size;
Initialize level set function;
Calculate cluster centre point value, the image entropy for being closed evolution curved exterior, the weight coefficient of evolution curvilinear inner and evolution
The weight coefficient of curved exterior;
Calculate the image match value of evolution curvilinear inner match value and evolution curved exterior;
Update level set function;
Judge whether evolution curve is stable, if evolution curve is stablized, segmentation terminates, and output obtains the image of wafer dopant,
If evolution curve is unstable, recalculate cluster centre point value, local entropy information, evolution curvilinear inner weight coefficient and
The weight coefficient of evolution curved exterior, until evolution curve is stablized.
3. -2 described in any item image partition methods according to claim 1, which is characterized in that the total energy functional, such as
Following formula:
Wherein, Ω indicates image space;φ (x, y) indicates level set function;u0(x, y) indicates image slices vegetarian refreshments;
Indicate evolution line smoothing item, u1Indicate the weight coefficient of the smooth item;
Indicate that region slickness keeps item, u2Indicate the weight coefficient of the holding item;
Indicate evolution curve fidelity term;
Indicate that regular terms, w indicate the weight coefficient of the regular terms.
4. image partition method according to claim 3, which is characterized in that the H kept in item and fidelity termε(φ)
Indicate regularization smoothing procedure function, the δ in the smooth itemε(φ) representative function HεThe derivative of (φ), wherein ε is constant, Hε(φ) and
δε(φ) such as following formula:
5. image partition method according to claim 3, which is characterized in that in the fidelity term,
Indicate the fidelity term of evolution curvilinear inner,Indicate the fidelity term of evolution curved exterior, D1
Indicate the weight coefficient of evolution curvilinear inner, D2Indicate the weight coefficient of evolution curved exterior;
In the fidelity term of the evolution curvilinear inner, c1iIt indicates evolution curvilinear inner match value, calculates such as following formula:
c1i=α1ki+(1-α1)u0 *
Wherein, kiIndicate that the i-th class clusters centerpoint value, i=1,2 ..., K, u0 *Indicate figure of the original image after mean filter
As information, α1Indicate weight, value range 0-1;
In the fidelity term of the evolution curved exterior, c2It indicates the image match value of evolution curved exterior, calculates such as following formula:
Wherein, τoutIndicate the image entropy of closure evolution curved exterior, such as following formula:
Wherein, N indicates the tonal gradation in image;PiIndicate the pixel percentage in the picture that gray value is i, value model
It encloses for 0-1, α2Indicate image entropy coefficient, α2≥0。
6. image partition method according to claim 5, which is characterized in that in evolution curvilinear inner, D1=max | u0(x,
y)-m1|, m1For the gray scale intermediate value of evolution curvilinear inner;In evolution curved exterior, D2=max | u0(x,y)-m2|, m2It is bent to develop
Gray scale intermediate value outside line.
7. -2 described in any item image partition methods according to claim 1, which is characterized in that the total energy functional substitutes into
Euler's formula, Introduction Time variable minimize total energy functional using gradient descent method, obtain calculating such as sub-level set public
Formula:
Wherein, t is time variable.
8. a kind of image segmentation system of wafer dopant, which is characterized in that the system comprises:
Module is obtained, for obtaining wafer image;
Module is constructed, domain type variation level set model is constructed;
Divide module, for utilizing domain type variation level set model for wafer image input area type variation level set model
Total energy functional wafer image is split, output obtain the image of wafer dopant.
9. a kind of computer equipment, including processor and for the memory of storage processor executable program, feature exists
In, when the processor executes the program of memory storage, the realization described in any item image partition methods of claim 1-7.
10. a kind of storage medium, is stored with program, which is characterized in that when described program is executed by processor, realize claim
The described in any item image partition methods of 1-7.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910196050.4A CN109978843B (en) | 2019-03-15 | 2019-03-15 | Image segmentation method and system for wafer dopant, computer equipment and storage medium |
PCT/CN2019/113458 WO2020186761A1 (en) | 2019-03-15 | 2019-10-25 | Image segmentation method and system for wafer dopant, computer device, and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910196050.4A CN109978843B (en) | 2019-03-15 | 2019-03-15 | Image segmentation method and system for wafer dopant, computer equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109978843A true CN109978843A (en) | 2019-07-05 |
CN109978843B CN109978843B (en) | 2020-11-24 |
Family
ID=67078991
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910196050.4A Active CN109978843B (en) | 2019-03-15 | 2019-03-15 | Image segmentation method and system for wafer dopant, computer equipment and storage medium |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN109978843B (en) |
WO (1) | WO2020186761A1 (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110363775A (en) * | 2019-06-21 | 2019-10-22 | 华南理工大学 | A kind of image partition method based on domain type variation level set |
WO2020186761A1 (en) * | 2019-03-15 | 2020-09-24 | 华南理工大学 | Image segmentation method and system for wafer dopant, computer device, and storage medium |
CN113218986A (en) * | 2021-04-09 | 2021-08-06 | 河南五建建设集团有限公司 | System and method for detecting compactness after prestressed grouting construction |
CN114667596A (en) * | 2019-11-21 | 2022-06-24 | 科磊股份有限公司 | Clustering secondary regions of interest based on noise characteristics |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2013023241A1 (en) * | 2011-08-12 | 2013-02-21 | Bt Imaging Pty Ltd | Photoluminescence imaging of doping variations in semiconductor wafers |
CN104732551A (en) * | 2015-04-08 | 2015-06-24 | 西安电子科技大学 | Level set image segmentation method based on superpixel and graph-cup optimizing |
CN107240108A (en) * | 2017-06-06 | 2017-10-10 | 衢州学院 | Movable contour model image partition method based on local Gaussian fitting of distribution and the poor energy driving of local symbol |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7177471B2 (en) * | 2002-02-01 | 2007-02-13 | Siemens Corporate Research, Inc. | Shape priors for level set representations |
CN104867143B (en) * | 2015-05-15 | 2018-09-04 | 东华理工大学 | The level set image segmentation method of energy model is fitted based on part guiding core |
CN107180433A (en) * | 2017-06-06 | 2017-09-19 | 衢州学院 | A kind of local cross-entropy measures the level set image segmentation algorithm of fuzzy C-mean algorithm |
CN109978843B (en) * | 2019-03-15 | 2020-11-24 | 华南理工大学 | Image segmentation method and system for wafer dopant, computer equipment and storage medium |
-
2019
- 2019-03-15 CN CN201910196050.4A patent/CN109978843B/en active Active
- 2019-10-25 WO PCT/CN2019/113458 patent/WO2020186761A1/en active Application Filing
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2013023241A1 (en) * | 2011-08-12 | 2013-02-21 | Bt Imaging Pty Ltd | Photoluminescence imaging of doping variations in semiconductor wafers |
CN104732551A (en) * | 2015-04-08 | 2015-06-24 | 西安电子科技大学 | Level set image segmentation method based on superpixel and graph-cup optimizing |
CN107240108A (en) * | 2017-06-06 | 2017-10-10 | 衢州学院 | Movable contour model image partition method based on local Gaussian fitting of distribution and the poor energy driving of local symbol |
Non-Patent Citations (2)
Title |
---|
T.F. CHAN AND L.A. VESE: "《Active contours without edges》", 《IEEE TRANSACTIONS ON IMAGE PROCESSING》 * |
叶丙刚、姜涛、罗朝晖、吴效明: "《结合图像熵的C-V模型分割算法》", 《计算机应用》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020186761A1 (en) * | 2019-03-15 | 2020-09-24 | 华南理工大学 | Image segmentation method and system for wafer dopant, computer device, and storage medium |
CN110363775A (en) * | 2019-06-21 | 2019-10-22 | 华南理工大学 | A kind of image partition method based on domain type variation level set |
CN114667596A (en) * | 2019-11-21 | 2022-06-24 | 科磊股份有限公司 | Clustering secondary regions of interest based on noise characteristics |
CN114667596B (en) * | 2019-11-21 | 2023-08-18 | 科磊股份有限公司 | Clustering regions of secondary interest based on noise characteristics |
CN113218986A (en) * | 2021-04-09 | 2021-08-06 | 河南五建建设集团有限公司 | System and method for detecting compactness after prestressed grouting construction |
CN113218986B (en) * | 2021-04-09 | 2023-05-12 | 河南五建建设集团有限公司 | System and method for detecting compactness after prestress grouting construction |
Also Published As
Publication number | Publication date |
---|---|
CN109978843B (en) | 2020-11-24 |
WO2020186761A1 (en) | 2020-09-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109978843A (en) | Image partition method, system, computer equipment and the storage medium of wafer dopant | |
CN107145908B (en) | A kind of small target detecting method based on R-FCN | |
CN105865462B (en) | The three-dimensional S LAM method based on event with depth enhancing visual sensor | |
Beres et al. | Effects of tropospheric wind shear on the spectrum of convectively generated gravity waves | |
Zhao et al. | Typhoon air‐sea drag coefficient in coastal regions | |
US11704841B2 (en) | Apparatus for estimating sameness of point cloud data and system for estimating sameness of point cloud data | |
Singh et al. | PI-LSTM: Physics-infused long short-term memory network | |
Singh et al. | An investigation on hybrid particle swarm optimization algorithms for parameter optimization of PV cells | |
Poterjoy et al. | Efficient assimilation of simulated observations in a high-dimensional geophysical system using a localized particle filter | |
CN109472819A (en) | A kind of binocular parallax estimation method based on cascade geometry context neural network | |
Ingebrigtsen et al. | Estimation of a non-stationary model for annual precipitation in southern Norway using replicates of the spatial field | |
Ramadhan et al. | Prim and floyd-warshall comparative algorithms in shortest path problem | |
CN107464244A (en) | A kind of image irradiation method of estimation based on neutral net | |
US20210133635A1 (en) | Material descriptor generation method, material descriptor generation device, recording medium storing material descriptor generation program, predictive model construction method, predictive model construction device, and recording medium storing predictive model construction program | |
Mahrt et al. | Formulation of turbulent fluxes in the stable boundary layer | |
Rao et al. | Mathematical programming on multivariate calibration estimation in stratified sampling | |
Agarwal et al. | Statistical field estimation for complex coastal regions and archipelagos | |
Premkumar et al. | A reliable optimization framework for parameter identification of single‐diode solar photovoltaic model using weighted velocity‐guided grey wolf optimization algorithm and Lambert‐W function | |
Montgomery et al. | Comments on “How does the boundary layer contribute to eyewall replacement cycles in axisymmetric tropical cyclones?” | |
Zhang et al. | Tropical cyclone intensity estimation using RVM and DADI based on infrared brightness temperature | |
Lu et al. | Tidal currents and mixing in the Gulf of St. Lawrence: an application of the incremental approach to data assimilation | |
CN109446339B (en) | Knowledge graph representation method based on multi-core Gaussian distribution | |
Ivakhnenko | The group method of data handling in long-range forecasting | |
CN109242786A (en) | A kind of automation morphologic filtering method suitable for urban area | |
Tymków et al. | 3D GIS for flood modelling in river valleys |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |