CN110378924A - Level set image segmentation method based on local entropy - Google Patents
Level set image segmentation method based on local entropy Download PDFInfo
- Publication number
- CN110378924A CN110378924A CN201910654340.9A CN201910654340A CN110378924A CN 110378924 A CN110378924 A CN 110378924A CN 201910654340 A CN201910654340 A CN 201910654340A CN 110378924 A CN110378924 A CN 110378924A
- Authority
- CN
- China
- Prior art keywords
- image
- segmentation
- local
- level set
- local entropy
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/11—Complex mathematical operations for solving equations, e.g. nonlinear equations, general mathematical optimization problems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/149—Segmentation; Edge detection involving deformable models, e.g. active contour models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20016—Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Computational Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Operations Research (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses the level set image segmentation methods based on local entropy, it is related to technical field of image processing, this method includes choosing original image, calculate the local entropy of original image, obtain pretreatment image, it carries out thresholding to pretreatment image to handle to obtain the coarse segmentation of the pretreatment image, resulting result is as initial level set profile；Utilize local entropy combination Local Binary Fitting (LBF) Construction of A Model image segmentation energy functional, it divides the image into energy functional and Local Chan-Vese (LCV) model linear combination obtains the global and local movable contour model based on local entropy, and then obtain EVOLUTION EQUATION；EVOLUTION EQUATION is solved using Hermite differential operator, fine segmentation is carried out to coarse segmentation image.The present invention can effectively eliminate the influence of noise on image segmentation, and this method has the advantages that CV model and LBF model by reference local entropy.
Description
Technical field
The present invention relates to technical field of image processing, more particularly to the level set image segmentation method based on local entropy.
Background technique
With the development of internet technology, image procossing arrives image understanding to image analysis again and has been deep into each research
And application field.Image procossing under complex scene is one important basic research problems of computer vision field, and is
A key technology in many related applications of artificial intelligence, application prospect is very extensive, therefore has important researching value.
In image analysis and understanding process, people's usually interested not instead of whole image, some in image or several areas
Domain (referred to as prospect), they usually have different nature and characteristics from background, belong to different things.In order to identify and analyze mesh
Mark, needs that separation and Extraction comes out from image by them.As the basis of image analysis and image understanding, image segmentation is to calculate
One of problem most basic in machine identification field and most difficult.Complex scene image segmentation be still one it is challenging
Work.
Image segmentation is the basis of image processing techniques, its main purpose is by region of interest area image and background point
From.Since there may be many noises and intensity profile inequalities in image, while target will receive such as hardware in image
Condition, viewing angle, lighting condition, complex background such as block at the influence of factors, this can make processing become more difficult.So
People are difficult to extract information related with target identification from image.Even if useful information, solution procedure can be extracted
Also there is unreliability.
Summary of the invention
When to solve image segmentation, due to noise in image is more, intensity profile is unequal and image in target expression
The problem of will receive the influence of many factors, and processing difficulty caused to increase, the present invention provides based on local entropy by slightly to
Thin improvement Chan-Vese level set image segmentation method.
Level set image segmentation method based on local entropy, method includes the following steps:
Step 1: choosing original image；
Step 2: calculating the local entropy of original image, obtain pretreatment image, thresholding processing is carried out to pretreatment image
The coarse segmentation of the pretreatment image is obtained, resulting result is as initial level set profile；
Step 3: utilize local entropy combination LBF Construction of A Model image segmentation energy functional, divide the image into energy functional and
LCV model linear combination obtains the global and local movable contour model based on local entropy, and then obtains EVOLUTION EQUATION；
Step 4: solving EVOLUTION EQUATION using Hermite differential operator, fine segmentation is carried out to coarse segmentation image.
2, as described in claim 1 based on the level set image segmentation method of local entropy, which is characterized in that the rough segmentation
The step of cutting are as follows:
Step 1: the neighborhood centered on each pixel i is 9 × 9, calculates the local entropy matrix E of original image；
Step 2: the cinetosis in the upper left corner and the upper right corner of detection local entropy matrix E, then removing the cinetosis makes E (i)=C
The vignetting in vignetting region, wherein C is constant；
Step 3: local entropy matrix E being converted into gray level image, generates local entropy image Eim, intermediate value E (i) in 0 He
Linear transformation is carried out between 255；
Step 4: coarse segmentation result being obtained as Eim is split to local entropy diagram using OTSU threshold method.
3, as described in claim 1 based on the level set image segmentation method of local entropy, which is characterized in that described fine
The step of segmentation are as follows:
Step 1: obtaining initialization profile using coarse segmentation result, and using the initialization profile as level set function
φ；
Step 2: the parameters in initialization level set movements equation, the parameter are σ, r, Δ t, ε, μ and υ；It is described
Level set movements equation are as follows:
In formula:
Wherein, σ is the standard deviation of kernel function；R is the window size of entropy function；Δ t is sliced time；ω is weight, 0≤
ω≤1；δ is Dirac function；ε is the regularization parameter of δ；μ=λ × 2552, λ ∈ [0,1]；The weighting constant that υ is positive；KσFor
Gaussian kernel function；ErFor local entropy；λ1And λ2It is the constant that value is positive；c1And c2Respectively develop the original image of inside or outside of curve
The average gray value in region；gkIt indicates that the average convolution kernel for local message detection, k indicate window size, is used for control office
Portion controls the sensibility of noise；I is image to be split；d1,d2After expression original image and convolved image operation inside curve C
With external gray value；
Step 3: on the basis of step 2, according to level set movements equation evolution level set function φ；
Step 4: the evolutionary process of analysis level set function φ, if the evolution difference of level set function φ is less than given threshold
Value, then termination of developing extract the zero level collection of function phi (x) as final segmentation result, and export and obtain segmented image；Instead
It, then return step 3 continues to develop to level set function φ.
Beneficial effects of the present invention: compared with prior art, the present invention by local entropy obtain image coarse segmentation as a result,
A formula similar with the energy theorem of LBF model is redefined further according to local entropy, and obtains one kind with LCV models coupling
Image partition method from thick to thin.This new improved model not only can handle the non-uniform image of gray scale, also enhance
Robustness of the model to noise.Local entropy allow us to prejudge out target object as initial profile where position i.e.
The general profile of object had both avoided interference of the background information to target in this way, but also the number that profile develops is able to substantially
It reduces, to improve the effect and efficiency of traditional activity skeleton pattern.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is model realization flow chart of the invention；
Fig. 2 is the segmentation result figure of the uneven image of gray scale；
Fig. 3 is the segmentation result figure with noise image；
Fig. 4 is to compare segmentation result figure with LBF with noise image；
Fig. 5 compares segmentation result figure with CV for the non-uniform image of gray scale；
Fig. 6 is the segmentation result figure of true picture.
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.
As shown in Figure 1, providing the level set image segmentation method based on local entropy in the embodiment of the present invention, comprising:
Step 1: choosing original image；
Step 2: calculating the local entropy of original image, obtain pretreatment image, thresholding processing is carried out to pretreatment image
The coarse segmentation of the pretreatment image is obtained, resulting result is as initial level set profile；
Step 3: utilize local entropy combination LBF Construction of A Model image segmentation energy functional, divide the image into energy functional and
LCV model linear combination obtains the global and local movable contour model based on local entropy, and then obtains EVOLUTION EQUATION；
Step 4: solving EVOLUTION EQUATION using Hermite differential operator, fine segmentation is carried out to coarse segmentation image.
Image coarse segmentation
Image segmentation can be regarded as all image pixels being divided into the different clusters with similar features, choose first former
Beginning image carries out coarse segmentation to original image, the step of coarse segmentation are as follows:
Step 1: the neighborhood centered on each pixel i is 9 × 9, calculates the local entropy matrix E of given image；
Step 2: the cinetosis in the upper left corner and the upper right corner of detection local entropy matrix E, then removing the cinetosis makes E (i)=C
The vignetting in vignetting region, wherein C is constant；
Step 3: local entropy matrix E being converted into gray level image, generates local entropy image Eim, intermediate value E (i) in 0 He
Linear transformation is carried out between 255；
Step 4: coarse segmentation result being obtained as Eim is split to local entropy diagram using OTSU threshold method.
Image fine segmentation
The step of image fine segmentation are as follows:
Step 1: obtaining initialization profile using coarse segmentation result, and using the initialization profile as level set function；
Step 2: the parameters in initialization (1.20) formula, the parameter are σ, r, Δ t, ε, μ and υ；
Step 3: on the basis of step 2, according to (1.20) formula evolution level set function φ；
Step 4: the evolutionary process of analysis level set function φ, if the evolution difference of level set function φ is less than given threshold
Value, then termination of developing extract the zero level collection of function phi (x) as final segmentation result, and export and obtain segmented image；Instead
It, then return step 3 continues to develop to level set function φ.
It is specific as follows:
Utilize local entropyIt is interregional that one point x is described
Ωx, new image segmentation energy functional is redefined, is expressed as follows:
In formula, Ω1=inside (C), Ω2=outside (C), Er(x)=E (x, W (x, r)) is the local entropy of x ∈ Ω.
W (x, r) is rectangular window function, W (x, r)=y:| x-y |≤r }, r > 0.
And the energy theorem of LCV is then:
ELCV(c1,c2, φ) and=∫ | u0(x,y)-c1|2Hε(φ(x,y))dxdy+∫|u0(x,y)-c2|2(1-Hε(φ(x,
y)))dxdy
+∫|gk*u0(x,y)-u0(x,y)-d1|2)Hε(φ(x,y))dxdy+∫|gk*u0(x,y)-u0(x,y)-d2|2)(1-
Hε(φ(x,y)))dxdy
(1.11)
Wherein, gkIndicate the average convolution kernel for local message detection, k indicates window size, for controlling local entity
Sensibility control to noise；d1,d2Indicate original image and the gray average after convolved image operation inside curve C with outside.
The energy model that the present invention defines is as follows:
E(C,c1,c2,f1,f2)=(1- ω) ELCV(C,c1,c2)+ωENRSF(C,f1,f2) (1.12)
ω value range 0≤ω≤1, c1And c2Respectively develop the average gray value in the original image region of inside or outside of curve.
f1And f2It is match value of the image in point x.By using Heaviside function H (φ), formula (1.12) can become:
In formula, M1(φ)=H (φ), M2(φ)=1-H (φ), υ and μ are positive weighting constants.C in formula (1.13)1
(x), c2(x), d1, d2, f1(x) and f2(x) can by relational expression obtain:
The calculus of variations and gradient descent method are used to formula (1.13), obtain level set movements equation:
In formula:
Wherein, σ is the standard deviation of kernel function；R is the window size of entropy function；Δ t is sliced time；ω is weight, 0≤
ω≤1；δ is Dirac function；ε is the regularization parameter of δ, generally takes 1；μ=λ × 2552, λ ∈ [0,1]；The weighting that υ is positive is normal
Number；KσFor gaussian kernel function；ErFor local entropy；λ1And λ2It is the constant that value is positive, usually enables λ1=λ2=1；c1And c2Respectively
Develop the average gray value in the original image region of inside or outside of curve；I is image to be split.
Level set movements equation is solved using Hermite differential operator, as follows:
The occurrence of differential operator is calculated:
D=[0.083139007900672, -0.662858814766878,0.662858814766878, -
0.083139007900672](1.23)
The discrete differential (being represented with the direction x, the direction y form is identical) of image:
fx(x)=0.083139007900672 (f (x-2, y)-f (x+2, y)) -0.662858814766878 (f (x-1,
y)-f(x+1,y))(1.24)
Correspondingly, the Hermite difference of (i+1, j) point will be changed to along the forward difference of x-axis in image area, along y-axis
Forward difference is changed to the Hermite differential of (i, j+1) point；Backward difference along x-axis is changed to the Hermite difference of (i-1, j) point,
Forward difference along x-axis is changed to the Hermite difference of (i, j-1) point, obtains:
After Hermite difference operator, when calculating functional derivative, use the second order neighborhood of pixel poor
Value, therefore level set movements process greatly gets a promotion for the robustness of noise.
Embodiment
To verify effectiveness of the invention, the verification result of the present embodiment is as follows:
The non-uniform image segmentation result of gray scale, as shown in Figure 2: figure (a) is original image, and figure (b) is local entropy image,
Scheming (c) is to cut result；The segmentation result of figure (c) demonstrates this method with the ability to the uneven image segmentation of gray scale.
Raw video picture is added the Gaussian noise and salt-pepper noise of varying strength, with noisy image segmentation
As a result, as shown in Figure 3: where (a) is the segmentation result of untreated original image, and (b) and (c) is to figure respectively
The segmentation result after Gaussian noise as adding variance 0.01 and 0.1, (d) and (e) is to be respectively to image addition variance
Segmentation result after 0.005 and 0.01 salt-pepper noise.This method is obtained to the height by varying strength by the result divided
Picture after this noise and salt-pepper noise processing has good segmentation effect.
The present invention and tradition LBF model compare segmentation result, selection two it is different with noisy picture, use respectively
LBF model and this method are split, and segmentation result is as shown in Figure 4: segmentation result figure (c) is as can be seen that LBF model segmentation band
The effect of noisy picture is not that very well, and that this method can be divided is very accurate.
The present invention and tradition CV model compare segmentation result, are divided using two non-uniform images of different gray scales
It cuts, shown in segmentation result Fig. 5: it is non-uniform to gray scale can explicitly to obtain very much this method from the segmentation result figure of two kinds of models
Image segmentation is well many than traditional CV model is divided.
Divide true picture as a result, having chosen a true figure to verify the segmentation effect of this method more accurately
As being verified, as shown in Figure 6: the result of traditional LBF model segmentation is very poor, the segmentation effect of this method or more satisfactory
's.Analog image can not only be divided to demonstrate implementation method of the invention, simple true picture can also be divided.
The present invention first derives the grayscale image distribution of image by local entropy, redefines the energy of one He LBF model
The similar formula of formula, and with LCV models coupling.This new improved model not only can handle the non-uniform figure of gray scale
Picture also enhances model to the robustness of noise.Local entropy allows us to prejudge out target object institute as initial profile
Position, that is, object general profile, both avoided interference of the background information to target in this way, but also profile develop time
Number is greatly decreased, to improve the effect and efficiency of traditional activity skeleton pattern.
Disclosed above is only specific embodiments of the present invention, and still, the embodiment of the present invention is not limited to this, Ren Heben
What the technical staff in field can think variation should all fall into protection scope of the present invention.
Claims (3)
1. the level set image segmentation method based on local entropy, which comprises the following steps:
Step 1: choosing original image；
Step 2: calculating the local entropy of original image, obtain pretreatment image, thresholding is carried out to pretreatment image and handles to obtain
The coarse segmentation of the pretreatment image, resulting result is as initial level set profile；
Step 3: utilizing local entropy combination LBF Construction of A Model image segmentation energy functional, divide the image into energy functional and LCV mould
Linear combines to obtain the global and local movable contour model based on local entropy, and then obtains EVOLUTION EQUATION；
Step 4: solving EVOLUTION EQUATION using Hermite differential operator, fine segmentation is carried out to coarse segmentation image.
2. as described in claim 1 based on the level set image segmentation method of local entropy, which is characterized in that the coarse segmentation
Step are as follows:
Step 1: the neighborhood centered on each pixel i is 9 × 9, calculates the local entropy matrix E of original image；
Step 2: the cinetosis in the upper left corner and the upper right corner of detection local entropy matrix E, then removing the cinetosis makes E (i)=C gradually
Vignetting in dizzy region, wherein C is constant；
Step 3: local entropy matrix E is converted into gray level image, generate local entropy image Eim, intermediate value E (i) 0 and 255 it
Between carry out linear transformation；
Step 4: coarse segmentation result being obtained as Eim is split to local entropy diagram using OTSU threshold method.
3. as described in claim 1 based on the level set image segmentation method of local entropy, which is characterized in that the fine segmentation
The step of are as follows:
Step 1: obtaining initialization profile using coarse segmentation result, and using the initialization profile as level set function φ；
Step 2: the parameters in initialization level set movements equation, the parameter are σ, r, Δ t, ε, μ and υ；The level
Collect EVOLUTION EQUATION are as follows:
In formula:
Wherein, σ is the standard deviation of kernel function；R is the window size of entropy function；Δ t is sliced time；ω is weight, 0≤ω≤
1；δ is Dirac function；ε is the regularization parameter of δ；μ=λ × 2552, λ ∈ [0,1]；The weighting constant that υ is positive；KσFor Gauss
Kernel function；ErFor local entropy；λ1And λ2It is the constant that value is positive；c1And c2Respectively develop the original image region of inside or outside of curve
Average gray value；gkIndicate the average convolution kernel for local message detection, k indicates window size, for controlling local entity
Sensibility control to noise；I is image to be split；d1,d2The inside curve C is with outside after indicating original image and convolved image operation
The gray value in portion；
Step 3: on the basis of step 2, according to level set movements equation evolution level set function φ；
Step 4: the evolutionary process of analysis level set function φ, if the evolution difference of level set function φ is less than given threshold value,
Then develop termination, extracts the zero level collection of function phi (x) as final segmentation result, and export and obtain segmented image；Conversely,
Then return step 3 continues to develop to level set function φ.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910654340.9A CN110378924B (en) | 2019-07-19 | 2019-07-19 | Level set image segmentation method based on local entropy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910654340.9A CN110378924B (en) | 2019-07-19 | 2019-07-19 | Level set image segmentation method based on local entropy |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110378924A true CN110378924A (en) | 2019-10-25 |
| CN110378924B CN110378924B (en) | 2022-09-27 |
Family
ID=68254178
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910654340.9A Active CN110378924B (en) | 2019-07-19 | 2019-07-19 | Level set image segmentation method based on local entropy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110378924B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111105430A (en) * | 2019-11-28 | 2020-05-05 | 青岛大学 | Variation level set image segmentation method based on Landmark simplex constraint |
| CN111325727A (en) * | 2020-02-19 | 2020-06-23 | 重庆邮电大学 | Intracranial hemorrhage area three-dimensional segmentation method based on local entropy and level set algorithm |
| CN111709957A (en) * | 2020-06-22 | 2020-09-25 | 河南理工大学 | Medical image segmentation method based on two-dimensional maximum entropy threshold C-V model |
| CN113261973A (en) * | 2021-04-13 | 2021-08-17 | 山东师范大学 | Method and system for identifying atrial fibrillation electrocardiosignals through weighted multi-scale finite-crossing visual image |
| WO2021253732A1 (en) * | 2020-06-18 | 2021-12-23 | 飞依诺科技(苏州)有限公司 | Medical image processing method and apparatus, computer device, and storage medium |
| WO2024055446A1 (en) * | 2022-09-13 | 2024-03-21 | 深圳先进技术研究院 | Image segmentation method and apparatus, device, and readable storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080107351A1 (en) * | 2006-11-07 | 2008-05-08 | Siemens Corporate Research, Inc. | Efficient segmentation of piecewise smooth images |
| 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 |
| CN109472792A (en) * | 2018-10-29 | 2019-03-15 | 石家庄学院 | In conjunction with the local energy functional of local entropy and the image partition method of non-convex regular terms |
-
2019
- 2019-07-19 CN CN201910654340.9A patent/CN110378924B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080107351A1 (en) * | 2006-11-07 | 2008-05-08 | Siemens Corporate Research, Inc. | Efficient segmentation of piecewise smooth images |
| 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 |
| CN109472792A (en) * | 2018-10-29 | 2019-03-15 | 石家庄学院 | In conjunction with the local energy functional of local entropy and the image partition method of non-convex regular terms |
Non-Patent Citations (2)
| Title |
|---|
| 刘晨等: "一种新的活动轮廓模型图像分割方法", 《工程科学与技术》 * |
| 郝智慧等: "结合全局与局部信息活动轮廓的非同质图像分割", 《中国图象图形学报》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111105430A (en) * | 2019-11-28 | 2020-05-05 | 青岛大学 | Variation level set image segmentation method based on Landmark simplex constraint |
| CN111105430B (en) * | 2019-11-28 | 2023-08-08 | 青岛大学 | Variation level set image segmentation method based on Landmark simplex constraint |
| CN111325727A (en) * | 2020-02-19 | 2020-06-23 | 重庆邮电大学 | Intracranial hemorrhage area three-dimensional segmentation method based on local entropy and level set algorithm |
| CN111325727B (en) * | 2020-02-19 | 2023-06-16 | 重庆邮电大学 | Intracranial hemorrhage area three-dimensional segmentation method based on local entropy and level set algorithm |
| WO2021253732A1 (en) * | 2020-06-18 | 2021-12-23 | 飞依诺科技(苏州)有限公司 | Medical image processing method and apparatus, computer device, and storage medium |
| CN111709957A (en) * | 2020-06-22 | 2020-09-25 | 河南理工大学 | Medical image segmentation method based on two-dimensional maximum entropy threshold C-V model |
| CN113261973A (en) * | 2021-04-13 | 2021-08-17 | 山东师范大学 | Method and system for identifying atrial fibrillation electrocardiosignals through weighted multi-scale finite-crossing visual image |
| WO2024055446A1 (en) * | 2022-09-13 | 2024-03-21 | 深圳先进技术研究院 | Image segmentation method and apparatus, device, and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110378924B (en) | 2022-09-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110378924A (en) | Level set image segmentation method based on local entropy | |
| CN104574347B (en) | Satellite in orbit image geometry positioning accuracy evaluation method based on multi- source Remote Sensing Data data | |
| WO2018024030A1 (en) | Saliency-based method for extracting road target from night vision infrared image | |
| CN111242864B (en) | Finger vein image restoration method based on Gabor texture constraint | |
| CN110837768B (en) | Online detection and identification method for rare animal protection | |
| Zhang et al. | Exemplar-based image inpainting using color distribution analysis | |
| CN105068918B (en) | A kind of page method of testing and device | |
| CN103455991A (en) | Multi-focus image fusion method | |
| CN110807738B (en) | Blurred image non-blind restoration method based on edge image block sharpening | |
| JP2021168143A (en) | System and method for efficiently scoring probe in image by vision system | |
| CN110245600B (en) | Unmanned aerial vehicle road detection method for self-adaptive initial quick stroke width | |
| CN111161222A (en) | Printing roller defect detection method based on visual saliency | |
| CN110222661B (en) | Feature extraction method for moving target identification and tracking | |
| CN106952292B (en) | 3D moving object detection method based on 6-degree-of-freedom scene stream clustering | |
| CN103080979A (en) | System and method for synthesizing portrait sketch from photo | |
| Chen et al. | A color-guided, region-adaptive and depth-selective unified framework for Kinect depth recovery | |
| CN110766782A (en) | Large-scale construction scene real-time reconstruction method based on multi-unmanned aerial vehicle visual cooperation | |
| CN110751680A (en) | Image processing method with fast alignment algorithm | |
| CN110111368B (en) | Human body posture recognition-based similar moving target detection and tracking method | |
| CN102629368B (en) | Color image vignetting recovery method based on illumination surface modeling | |
| CN108717699B (en) | Ultrasonic image segmentation method based on continuous minimum segmentation | |
| CN116188826A (en) | Template matching method and device under complex illumination condition | |
| CN109035306A (en) | Moving-target automatic testing method and device | |
| CN115049552A (en) | Infrared small target detection method based on structure tensor weighted local contrast measurement | |
| CN114862883A (en) | Target edge extraction method, image segmentation method and system |
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 |