CN115984312B - Image segmentation method, electronic device and computer readable storage medium - Google Patents

Image segmentation method, electronic device and computer readable storage medium Download PDF

Info

Publication number
CN115984312B
CN115984312B CN202310247108.XA CN202310247108A CN115984312B CN 115984312 B CN115984312 B CN 115984312B CN 202310247108 A CN202310247108 A CN 202310247108A CN 115984312 B CN115984312 B CN 115984312B
Authority
CN
China
Prior art keywords
level set
iteration
image
set function
target 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.)
Active
Application number
CN202310247108.XA
Other languages
Chinese (zh)
Other versions
CN115984312A (en
Inventor
董斌
倪锦根
翁桂荣
卜倩倩
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Suzhou University
Original Assignee
Suzhou University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Suzhou University filed Critical Suzhou University
Priority to CN202310247108.XA priority Critical patent/CN115984312B/en
Publication of CN115984312A publication Critical patent/CN115984312A/en
Application granted granted Critical
Publication of CN115984312B publication Critical patent/CN115984312B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Landscapes

  • Image Analysis (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses an image segmentation method, electronic equipment and a computer readable storage medium. The method comprises the following steps: s1, setting an initial contour line in a target image, and representing the initial contour line by using an initial level set function; s2, updating and iterating the initial level set function; s3, taking the level set function obtained in the last iteration as a segmentation curve of the target image. The image segmentation method replaces a length term in the traditional active contour model by the mean filter and is used for smoothing the segmentation curve; the distance rule term in the traditional active contour model is replaced by an activation function, so that the level set function always keeps rules that the values on the contour line and the internal value of the contour line are negative and the external value of the contour line is positive in the iterative process. The image segmentation method has ideal segmentation effect on images with uneven gray scales, and has advantages in segmentation speed and segmentation accuracy.

Description

Image segmentation method, electronic device and computer readable storage medium
Technical Field
The present invention relates to the field of image processing technologies, and in particular, to an image segmentation method, an electronic device, and a computer readable storage medium.
Background
Image segmentation is an important research topic in the field of image processing, and it divides an image into a plurality of mutually non-overlapping areas, and extracts interested targets (such as foreground and background) from the areas, thus laying a foundation for subsequent image target identification and feature analysis.
The active contour model is the most representative image segmentation method in recent decades, and is also a research hot spot of the current image segmentation method. The basic idea of the active contour model is to set an initial contour line on an image, and drive the contour line to evolve by using an energy function based on a level set method to approach a target boundary so as to realize target segmentation. The algorithm can obtain the accuracy of the sub-pixel level of the target boundary, and provides a smooth closed contour as a segmentation result.
The main active contour model of the current main stream mainly has the following technical problems: the partial region gray level is not good in uniform dividing effect, and false targets in the image cannot be distinguished. Therefore, a new image segmentation method is needed to solve the above problems.
Disclosure of Invention
Therefore, the technical problem to be solved by the invention is to provide an image segmentation method which can effectively segment an image with uneven gray scale and has high segmentation accuracy.
In order to solve the above technical problems, the present invention provides an image segmentation method, which includes the following steps:
s1, setting an initial contour line in a target image, and representing the initial contour line by using an initial level set function;
s2, updating and iterating the initial level set function; the level set function at the nth iteration is:
Figure GDA0004196427350000021
wherein,,
Figure GDA0004196427350000022
is an average filter with a window size of k×k; softsign is a softsign activation function; η is a constant; phi (x, n) is the level set function at the nth iteration; x is a vector pixel point in the target image; phi (x, n-1) is the level set function at the n-1 th iteration; u (x, n-1)) is the gradient descent equation at the n-1 th iteration; Δt is the time step;
s3, stopping iteration when the I phi (x, n) -phi (x, n-1) I < epsilon, and taking a level set function obtained in the last iteration as a segmentation curve of the target image; where ε is a constant.
In one embodiment of the invention, the gradient descent equation at the n-1 th iteration is:
U(φ(x，n-1))＝-αS′(φ(x，n-1))softsign((d1(x，n-1)-d 2 (x，n-1))/β)，
wherein α is a constant; beta is the standard deviation of the gray scale of the target image; the function S' (z) is a function
Figure GDA0004196427350000023
Is of (1)Function (F)>
Figure GDA0004196427350000024
Figure GDA0004196427350000025
For the data driving term, y is the vector pixel point in the target image, theta y For a square neighborhood with y as the center point and a side length of 4σ+1,
Figure GDA0004196427350000026
as Gaussian kernel function, σ is standard deviation of Gaussian kernel function, a is such that +.>
Figure GDA0004196427350000027
Is a normalized constant of (2); i (x) is the true gray value of the target image; />
Figure GDA0004196427350000028
To calculate the shadow values of the image to be fitted,
Figure GDA0004196427350000029
the method is characterized in that the method is an average filter with a window size of 2w+1, w is a constant, and the convolution operation is performed;
Figure GDA0004196427350000031
the information value is the edge reflection structure information value at the n-1 th iteration; when i=1, _a->
Figure GDA0004196427350000032
Is the area theta y ∩Ω 1 Information value of edge reflection structure in +.>
Figure GDA0004196427350000033
When i=2, _a->
Figure GDA0004196427350000034
Is the area theta y ∩Ω 2 The edge reflection structure information value in the inner layer,
Figure GDA0004196427350000035
Ω 1 is the outer region of the contour line Ω 2 Is a contour line and an inner region of the contour line.
In one embodiment of the present invention, the standard deviation of the target image gray scale is:
Figure GDA0004196427350000036
wherein Ω is an image domain of the target image; q is the total number of pixel points in the target image;
Figure GDA0004196427350000037
is the average gray value of the target image.
In one embodiment of the invention, the initial level set function is:
Figure GDA0004196427350000038
in one embodiment of the invention, ε may be valued in the range of [0.001,0.0001].
In one embodiment of the invention, Δt has a value in the range of [0.5,2].
The invention also provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of any of the methods described above when the program is executed.
The invention also provides a computer readable storage medium having stored thereon a computer program which when executed by a processor performs the steps of any of the methods described above.
The invention also provides an image segmentation system, which comprises the following modules:
an initial level set module for setting an initial contour line in a target image and representing the initial contour line by an initial level set function;
the iteration module is used for carrying out updating iteration on the initial level set function; the level set function at the nth iteration is:
Figure GDA0004196427350000041
wherein,,
Figure GDA0004196427350000042
is an average filter with a window size of k×k; softsign is a softsign activation function; η is a constant; phi (x, n) is the level set function at the nth iteration; x is a vector pixel point in the target image; phi (x, n-1) is the level set function at the n-1 th iteration; u (x, n-1)) is the gradient descent equation at the n-1 th iteration; Δt is the time step;
the segmentation curve determining module is used for taking the level set function obtained in the last iteration as a segmentation curve of the target image; when |φ (x, n) - φ (x, n-1) | < ε, the iteration is stopped, where ε is a constant.
Compared with the prior art, the technical scheme of the invention has the following advantages:
the image segmentation method of the invention updates the discrete level set function by setting the initial level set function and adopting the optimized level set function during each iteration. The length term in the traditional active contour model is replaced by the mean filter, so that the length term is used for smoothing the segmentation curve; the distance rule term in the traditional active contour model is replaced by an activation function, so that the level set function always keeps rules that the values on the contour line and the internal value of the contour line are negative and the external value of the contour line is positive in the iterative process. Thereby realizing an ideal division effect of an image having uneven gradation, and having advantages in both division speed and division accuracy.
The foregoing description is only an overview of the present invention, and is intended to be implemented in accordance with the teachings of the present invention, as well as the preferred embodiments thereof, together with the following detailed description of the invention, given by way of illustration only, together with the accompanying drawings.
Drawings
In order that the invention may be more readily understood, a more particular description of the invention will be rendered by reference to specific embodiments thereof that are illustrated in the appended drawings, in which
FIG. 1 is a flow chart of an image segmentation method in an embodiment of the invention;
FIG. 2 is a schematic diagram of an image segmentation method in an embodiment of the present invention;
FIG. 3 is a schematic diagram of the segmentation of a gray-scale non-uniform image using an image segmentation method in an embodiment of the present invention;
FIG. 4 is a schematic diagram of robustness to noise using an image segmentation method in an embodiment of the present invention;
FIG. 5 is a schematic diagram showing the segmentation of an image using different models in accordance with an embodiment of the present invention.
Detailed Description
The present invention will be further described with reference to the accompanying drawings and specific examples, which are not intended to be limiting, so that those skilled in the art will better understand the invention and practice it.
Example 1
Referring to fig. 1, the present embodiment discloses an image segmentation method, which includes the steps of:
s1, setting an initial contour line in a target image, and representing the initial contour line by using an initial level set function;
specifically, referring to fig. 2, for the target image I, whose continuous image domain is Ω, x and y are vector pixels within the image domain Ω, an initial contour line C is set 0 Dividing the image domain Ω into two sub-regions Ω 1 And omega 2 ，Ω 1 An outer region, Ω, representing the initial contour line 2 Representing the initial contour and the interior region of the initial contour. For any point y in the image domain omega, there is a square neighborhood theta with y as the center point and the side length of 4sigma+1 y Wherein, the method comprises the steps of, wherein,
Figure GDA0004196427350000051
as Gaussian kernel function, σ is standard deviation of Gaussian kernel function, a is such that +.>
Figure GDA0004196427350000061
Is a constant of normalization of (c). When the point y is close to C 0 At the time of theta y And subregion omega 1 And omega 2 All have intersection, namely theta y ∩Ω 1 And theta y ∩Ω 2 。
Further, the initial contour line C is represented by an initial level set function phi (x, 0) based on a level set method 0 And is defined at an initial contour line C 0 Upper and initial contour line C 0 In (C), the values of the initial level set function phi (x, 0) are all-1, and in the initial contour line C 0 Outside of (1), the initial level set function phi (x, 0) has a value of 1. The curve C is represented by a level set function phi (x, n) obtained by the nth iteration (n is greater than or equal to 1) n As a final segmentation result.
Specifically, the initial level set function is expressed as:
Figure GDA0004196427350000062
s2, updating and iterating the initial level set function; the level set function at the nth iteration is:
Figure GDA0004196427350000063
wherein,,
Figure GDA0004196427350000064
for an average filter with a window size of k×k, passing the average filter +.>
Figure GDA0004196427350000065
The method replaces length terms in the traditional active contour model and is used for smoothing the segmentation curve; softsign is softsign activating functionA number; η is a constant; phi (x, n) is the level set function at the nth iteration; x is a vector pixel point in the target image; phi (x, n-1) is the level set function at the n-1 th iteration; u (x, n-1)) is the gradient descent equation at the n-1 th iteration; Δt is the time step; optionally, Δt has a value in the range of [0.5,2]In the present embodiment, Δt=1.
Specifically, the definition of the activation function softsign is:
Figure GDA0004196427350000066
the distance rule term in the traditional active contour model is replaced by an activation function softsign, which is used to keep the level set function always in the iterative process with negative values on the contour and inside the contour and positive values outside the contour.
Specifically, the gradient descent equation at the n-1 th iteration is:
U(φ(x，n-1))＝-αS′(φ(x，n-1))softsign((d 1 (x，n-1)-d 2 (x，n-1))/β)，
wherein α is a constant; beta is the standard deviation of the gray scale of the target image; the function S' (z) is a function
Figure GDA0004196427350000071
Is a derivative of->
Figure GDA0004196427350000072
Figure GDA0004196427350000073
The data driving item shows the difference between the target and the background. I (x) is the true gray value of the target image;
Figure GDA0004196427350000074
for the image shading value fitted by calculation, +.>
Figure GDA0004196427350000075
Is an average filter with a window size of 2w+1, w is a constant, and is convolutionCalculating;
Figure GDA0004196427350000076
the information value is the edge reflection structure information value at the n-1 th iteration; when i=1, _a->
Figure GDA0004196427350000077
Is the area theta y ∩Ω 1 The edge reflection structure information value in the inner layer,
Figure GDA0004196427350000078
when i=2, _a->
Figure GDA0004196427350000079
Is the area theta y ∩Ω 2 Information value of edge reflection structure in +.>
Figure GDA00041964273500000710
Ω 1 Is the outer region of the contour line Ω 2 Is the initial contour and the inner region of the contour.
Further, the standard deviation of the target image gray scale is:
Figure GDA0004196427350000081
wherein Ω is an image domain of the target image; q is the total number of pixel points in the target image;
Figure GDA0004196427350000085
is the average gray value of the target image.
And step S3, taking the level set function obtained in the last iteration as a segmentation curve of the target image.
Specifically, when |phi (x, n) -phi (x, n-1) | < epsilon, stopping iteration, and taking a level set function obtained by the last iteration as a segmentation curve of the target image; where ε is a constant, optionally, ε is in the range of [0.001,0.0001], and ε is 0.001 in this embodiment.
The image shading value proposed in the present embodiment
Figure GDA0004196427350000082
Is a comprehensive measure of the overall image intensity, the calculation result of which is disposable and does not need to be updated through iteration. Edge reflection Structure information value +.>
Figure GDA0004196427350000083
The region with obviously changed brightness in the image is reflected, so that iteration of the level set function is guided, and the contour line evolves towards the target boundary.
In the optimized length term calculation method in the present embodiment,
Figure GDA0004196427350000084
in fact, the level set function after each iteration is subjected to filtering treatment, and compared with the traditional method, the method has the advantage that the length term is required to be updated through iteration, and the method has higher calculation speed. Meanwhile, an activation function softsign is used for realizing a distance rule term, only one constant eta is provided in the formula except a level set function phi (x, n-1), no parameter to be adjusted is provided, and the robustness of the method is improved.
In order to prove the effectiveness of the invention, the performance of the proposed method is verified by a computer experimental method, and a comparison experiment is carried out with two multiplicative bias field models.
A. Experimental conditions:
all experiments were performed in MATLAB2015b on a 2.6-GHz Intel Kuri 5 personal computer. The images used for the experiments were all from a standard gallery. The color image is converted into a gray scale image before segmentation. In all experiments, the rectangular solid line represents the initial contour line and the dashed line represents the final segmentation curve.
B. The experimental steps are as follows:
the setting parameters are as follows: w=35, k=11, σ=5, α=3, η=10, ε=0.001, Δt=1, and the maximum number of iterations N and the initial contour line are set.
C. The experimental results are as follows:
(1) Segmentation experiment
5 images with uneven gray scale are selected from the BSDS gallery, and the five images are segmented by the image segmentation method. As shown in fig. 3, the first column is the original image and the initial contour, the second column is the shadow of the image, and the third column is d at the nth iteration 1 (x)-d 2 (x) Schematic diagram, fourth column is final segmentation result. Table 1 shows the results of 10 independent experiments averaged for the number of iterations and segmentation.
TABLE 1 iteration times and segmentation time of the image segmentation method of the present invention for five images
Figure GDA0004196427350000091
(2) Noise robustness experiment
In order to evaluate the robustness of the image segmentation method to noise, 5 gray images are selected from a BSDS image library, gaussian noise with the mean value of 0 and the variance of 0.03 is artificially added, and the original image and the image added with the noise are respectively segmented by the image segmentation method. As shown in fig. 4, the first column is the original image and the initial contour, the second column is the segmentation result of the image segmentation method of the present invention on the original image, and the third column is the segmentation result of the image segmentation method of the present invention on the image to which noise is added. The segmentation result of the noiseless and noisy image proximity shows that the image segmentation method has robustness to noise.
(3) Comparative experiments
7 complex color images were selected from the BSDS gallery and the 7 images were segmented using PBCACM (pre-fitting bias correction active contour model, pre-fitting off-field correction active contour model), JDACM (Jeffreys divergence active contour model ) and the image segmentation method of the present invention, respectively. As shown in fig. 5, the first column is the original image and the initial contour, the second column is the segmentation result of the PBC model, the third column is the segmentation result of the JDACM model, and the fourth column is the segmentation result of the image segmentation method of the present invention.
TABLE 2 time (seconds) for dividing three models
Figure GDA0004196427350000101
TABLE 3 DSC and IOU (DSC/IOU) of three models
Figure GDA0004196427350000102
Table 2 shows how the ERSI model is superior in speed to the other models as seen from the data in the table when the three models are segmented. The segmentation accuracy of each model was quantitatively compared using two reference standards, DSC (Dice Similariy Coefficient, dice similarity coefficient) and IOU (Intersection of Union, degree of overlap). DSC definition is
Figure GDA0004196427350000111
IOU is defined as->
Figure GDA0004196427350000112
Wherein S is 1 Is the target area obtained by experiment, S 2 Is the standard target area provided in the BSDS gallery. The values of DSC and IOU are closer to 1, indicating that the segmentation effect is better. Table 3 shows the DSC and IOU values of the three models. As can be seen from the data in the table, the image segmentation method of the present invention also has an advantage in segmentation accuracy.
Example two
The present embodiment discloses an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, which processor implements the steps of the method described in embodiment one when executing the program.
Example III
The present embodiment discloses a computer readable storage medium having stored thereon a computer program which when executed by a processor realizes the steps of the method described in the first embodiment.
Example IV
The embodiment discloses an image segmentation system, which comprises the following modules:
an initial level set module for setting an initial contour line in a target image and representing the initial contour line by an initial level set function;
the iteration module is used for carrying out updating iteration on the initial level set function; the level set function at the nth iteration is:
Figure GDA0004196427350000114
wherein (1)>
Figure GDA0004196427350000113
Is an average filter with a window size of k×k; softsign is a softsign activation function; η is a constant; phi (x, n) is the level set function at the nth iteration; x is a vector pixel point in the target image; phi (x, n-1) is the level set function at the n-1 th iteration; u (x, n-1)) is the gradient descent equation at the n-1 th iteration; Δt=1;
and the segmentation curve determining module is used for taking the level set function obtained in the last iteration as a segmentation curve of the target image.
The image segmentation system in the embodiment of the present invention is used for implementing the foregoing image segmentation method, so that the detailed implementation of the system can be seen from the foregoing example part of the image segmentation method, and therefore, the detailed implementation of the system can refer to the description of the corresponding examples of the respective parts, which are not further described herein.
In addition, since the image segmentation system of the present embodiment is used to implement the foregoing image segmentation method, the effects thereof correspond to those of the foregoing method, and will not be described herein.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It is apparent that the above examples are given by way of illustration only and are not limiting of the embodiments. Other variations and modifications of the present invention will be apparent to those of ordinary skill in the art in light of the foregoing description. It is not necessary here nor is it exhaustive of all embodiments. And obvious variations or modifications thereof are contemplated as falling within the scope of the present invention.

Claims (9)

1. An image segmentation method, characterized by comprising the steps of:
s1, setting an initial contour line in a target image, and representing the initial contour line by using an initial level set function;
s2, updating and iterating the initial level set function; the level set function at the nth iteration is:
Figure FDA0004209911650000011
wherein,,
Figure FDA0004209911650000012
is an average filter with a window size of k×k; softsign is a softsign activation function; η is a constant; phi (x, n) is the level set function at the nth iteration; x is a vector pixel point in the target image; phi (x, n-1) is the level set function at the n-1 th iteration; u (x, n-1)) is the gradient descent equation at the n-1 th iteration; Δt is the time step;
s3, stopping iteration when the I phi (x, n) -phi (x, n-1) I < epsilon, and taking a level set function obtained in the last iteration as a segmentation curve of the target image; where ε is a constant.
2. The image segmentation method according to claim 1, wherein the gradient descent equation at the n-1 th iteration is:
U(φ(x，n-1))＝-αS′(φ(x，n-1))softsign((d 1 (x，n-1)-d 2 (x, n-1))/β, wherein α is a constant; beta is the standard deviation of the gray scale of the target image; the function S' (z) is a function
Figure FDA0004209911650000013
Is a derivative of->
Figure FDA0004209911650000014
Figure FDA0004209911650000015
For the data driving term, y is the vector pixel point in the target image, theta y Is a square neighborhood with y as a center point and a side length of 4sigma+1, and is +.>
Figure FDA0004209911650000021
As Gaussian kernel function, σ is standard deviation of Gaussian kernel function, a is such that +.>
Figure FDA0004209911650000022
Is a normalized constant of (2); i (x) is the true gray value of the target image;
Figure FDA0004209911650000023
for the image shading value fitted by calculation, +.>
Figure FDA0004209911650000024
The method is characterized in that the method is an average filter with a window size of 2w+1, w is a constant, and the convolution operation is performed;
Figure FDA0004209911650000025
the information value is the edge reflection structure information value at the n-1 th iteration; when i=1, _a->
Figure FDA0004209911650000026
Is the area theta y ∩Ω 1 The edge reflection structure information value in the inner layer,
Figure FDA0004209911650000027
when i=2, _a->
Figure FDA0004209911650000028
Is the region omega y ∩Ω 2 Information value of edge reflection structure in +.>
Figure FDA0004209911650000029
Ω 1 Is the outer region of the contour line Ω 2 Is a contour line and an inner region of the contour line.
3. The image segmentation method according to claim 2, wherein the standard deviation of the gray scale of the target image is:
Figure FDA00042099116500000210
wherein Ω is an image domain of the target image; q is the total number of pixel points in the target image;
Figure FDA00042099116500000211
is the average gray value of the target image.
4. The image segmentation method as set forth in claim 1, wherein the initial level set function is:
Figure FDA0004209911650000031
5. the image segmentation method according to claim 1, wherein epsilon has a value in the range of [0.001,0.0001].
6. The image segmentation method according to claim 1, wherein Δt has a value in the range of [0.5,2].
7. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method of any of claims 1 to 6 when the program is executed.
8. A computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps of the method according to any one of claims 1 to 6.
9. An image segmentation system, comprising the following modules:
an initial level set module for setting an initial contour line in a target image and representing the initial contour line by an initial level set function;
the iteration module is used for carrying out updating iteration on the initial level set function; the level set function at the nth iteration is:
Figure FDA0004209911650000032
wherein,,
Figure FDA0004209911650000033
is an average filter with a window size of k×k; softsign is a softsign activation function; η is a constant; phi (x, n) is the level set function at the nth iteration; x is a vector pixel point in the target image; phi (x, n-1) is the level set function at the n-1 th iteration; u (x, n-1)) is the gradient descent equation at the n-1 th iteration; Δt is the time step;
the segmentation curve determining module is used for taking the level set function obtained in the last iteration as a segmentation curve of the target image; when |φ (x, n) - φ (x, n-1) | < ε, the iteration is stopped, where ε is a constant.
CN202310247108.XA 2023-03-15 2023-03-15 Image segmentation method, electronic device and computer readable storage medium Active CN115984312B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310247108.XA CN115984312B (en) 2023-03-15 2023-03-15 Image segmentation method, electronic device and computer readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310247108.XA CN115984312B (en) 2023-03-15 2023-03-15 Image segmentation method, electronic device and computer readable storage medium

Publications (2)

Publication Number Publication Date
CN115984312A CN115984312A (en) 2023-04-18
CN115984312B true CN115984312B (en) 2023-07-07

Family

ID=85968353

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310247108.XA Active CN115984312B (en) 2023-03-15 2023-03-15 Image segmentation method, electronic device and computer readable storage medium

Country Status (1)

Country Link
CN (1) CN115984312B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117593323B (en) * 2024-01-19 2024-03-29 苏州大学 Image segmentation method, system, medium and device based on non-local features

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3327963A1 (en) * 2016-11-29 2018-05-30 Alcatel Lucent Optical multiplexing device and associated equipment and method
CN114584870A (en) * 2022-02-28 2022-06-03 武汉光迅科技股份有限公司 Device and method for compensating originating signal frequency spectrum imbalance in optical module

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111462145B (en) * 2020-04-01 2023-04-07 重庆大学 Active contour image segmentation method based on double-weight symbol pressure function
CN112424835B (en) * 2020-05-18 2023-11-24 上海联影医疗科技股份有限公司 System and method for image reconstruction

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3327963A1 (en) * 2016-11-29 2018-05-30 Alcatel Lucent Optical multiplexing device and associated equipment and method
CN114584870A (en) * 2022-02-28 2022-06-03 武汉光迅科技股份有限公司 Device and method for compensating originating signal frequency spectrum imbalance in optical module

Also Published As

Publication number Publication date
CN115984312A (en) 2023-04-18

Similar Documents

Publication Publication Date Title
Li et al. Deep retinex network for single image dehazing
US8437570B2 (en) Geodesic image and video processing
CN115984312B (en) Image segmentation method, electronic device and computer readable storage medium
Yoshizawa et al. Fast gauss bilateral filtering
Yang et al. Research and analysis of Image edge detection algorithm Based on the MATLAB
Jebadass et al. Low contrast enhancement technique for color images using interval-valued intuitionistic fuzzy sets with contrast limited adaptive histogram equalization
CN109472792A (en) In conjunction with the local energy functional of local entropy and the image partition method of non-convex regular terms
Kapoor et al. Fog removal in images using improved dark channel prior and contrast limited adaptive histogram equalization
US20130064468A1 (en) Methods and Apparatus for Image Analysis and Modification Using Fast Sliding Parabola Erosian
Jalab et al. A new medical image enhancement algorithm based on fractional calculus
CN113643302B (en) Active contour model-based unsupervised medical image segmentation method and system
CN117173201B (en) Second order differential image segmentation method, system, medium and device
CN111898408B (en) Quick face recognition method and device
US20230103929A1 (en) Onion convolution-based inpainting of images
CN107578393B (en) Aerial image brightness adjusting method based on manual interaction
Dong et al. An active contour model based on shadow image and reflection edge for image segmentation
CN114596320A (en) Image segmentation method and device based on ALSLCV model
Long et al. Spectrum decomposition in Gaussian scale space for uneven illumination image binarization
Lei et al. A robust hybrid active contour model based on pre-fitting bias field correction for fast image segmentation
Yadav et al. Enhancement of region of interest from a single backlit image with multiple features
Han et al. An efficient estimation method for intensity factor of illumination changes
CN117994277B (en) Target segmentation method, device, medium and system for non-uniform illumination image
Hu et al. Dehazing for images with sun in the sky
CN114926476B (en) Accurate segmentation method for variational level set model image based on local correction function
Li et al. A structure and texture revealing retinex model for low-light image enhancement

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