CN110517250A - It is a kind of to fight the eye ground blood vessel segmentation system for generating network based on enhancing - Google Patents
It is a kind of to fight the eye ground blood vessel segmentation system for generating network based on enhancing Download PDFInfo
- Publication number
- CN110517250A CN110517250A CN201910797551.8A CN201910797551A CN110517250A CN 110517250 A CN110517250 A CN 110517250A CN 201910797551 A CN201910797551 A CN 201910797551A CN 110517250 A CN110517250 A CN 110517250A
- Authority
- CN
- China
- Prior art keywords
- blood vessel
- enhancing
- network
- network model
- fight
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 210000004204 blood vessel Anatomy 0.000 title claims abstract description 43
- 230000011218 segmentation Effects 0.000 title claims abstract description 39
- 230000002708 enhancing effect Effects 0.000 title claims abstract description 24
- 230000004256 retinal image Effects 0.000 claims abstract description 20
- 238000012545 processing Methods 0.000 claims abstract description 9
- 230000000694 effects Effects 0.000 claims abstract description 7
- 230000002207 retinal effect Effects 0.000 claims abstract description 6
- 230000003044 adaptive effect Effects 0.000 claims abstract description 4
- 230000004913 activation Effects 0.000 claims description 10
- 238000005070 sampling Methods 0.000 claims description 8
- 230000004069 differentiation Effects 0.000 claims description 6
- 239000000203 mixture Substances 0.000 claims description 2
- 238000003384 imaging method Methods 0.000 abstract description 2
- 238000000034 method Methods 0.000 description 11
- 239000008280 blood Substances 0.000 description 5
- 210000004369 blood Anatomy 0.000 description 5
- 210000001210 retinal vessel Anatomy 0.000 description 4
- 201000010099 disease Diseases 0.000 description 3
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 3
- 238000013480 data collection Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 210000001525 retina Anatomy 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 206010020772 Hypertension Diseases 0.000 description 1
- 230000003115 biocidal effect Effects 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 206010012601 diabetes mellitus Diseases 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000004233 retinal vasculature Effects 0.000 description 1
- 230000002792 vascular Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
-
- 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
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30041—Eye; Retina; Ophthalmic
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30101—Blood vessel; Artery; Vein; Vascular
Abstract
The present invention provides a kind of eye ground blood vessel segmentation system fought based on enhancing and generate network, belong to technical field of medical image processing, for solve the problems, such as existing eye ground blood vessel segmentation there are algorithm limitation, actual imaging contrast is lower, segmentation result occurs that rupture of blood vessel, there are redundancies for vessel branch details.Present system, including for carrying out enhancing processing to sample label and retinal images the comparison adaptive histogram equalization manually divided retinal images pretreatment unit, the data volume of training is enhanced again image the data enhancement unit of processing and treated that retinal data image is split to extract more minute blood vessels to data enhancement unit by increasing, improve the confrontation generation network model of segmentation effect accuracy.Effect is the great training for accelerating network, and tiny vessel branch details can also be split, and effectively eliminates the noise of retinal images, improves the accuracy rate of blood vessel segmentation.
Description
Technical field
The present invention relates to technical field of medical image processing, specifically, more particularly to a kind of based on enhancing confrontation generation
The eye ground blood vessel segmentation system of network.
Background technique
Retinal microvascular is uniquely can be with non-wound direct observation and micro- blood of deeper in Whole Body vascular system
Pipe, many diseases such as: hypertension, diabetes can all cause the change of retinal vessel state or structure.Pass through retinal blood
The information such as quantity, branch, the width of pipe can be used as the diagnosis basis of retinal vessel related disease.By to retinal blood
Pipe is analyzed, and diagnosis and the prevention some diseases are done with early stage are that have and its important clinical medicine meaning.Therefore it regards
Retinal vasculature cutting techniques are particularly important.
Since retinal images intensity profile is uneven, blood vessel structure is complicated, target blood and image background contrast compared with
Low and picture noise etc. influences, and retinal vessel segmentation is faced with huge challenge.Conventional segmentation methods have to be known based on mode
Method for distinguishing (supervised classification and unsupervised segmentation), based on matched filtering, based on mathematical morphology, based on tracking method
Deng.Retinal images divide branch as image segmentation, it is current there are many dividing method by numerous studies and report,
Existing method is based primarily upon above several aspects, or in terms of these on combination.
Above-mentioned method can extract most of retinal vessel, due to algorithm limitation or actual imaging contrast compared with
Low, the phenomenon that rupture of blood vessel and vessel branch details usually occurs in algorithm segmentation result, and there are redundancies.
Summary of the invention
According to technical problem set forth above, and provide a kind of eye ground blood vessel for fighting based on enhancing and generating network
Segmenting system.The present invention mainly generates network model using the confrontation of building, greatly accelerates the training of network, and makes tiny
Vessel branch details can also split, effectively eliminate the noise of retinal images, improve the standard of blood vessel segmentation
True rate.
The technological means that the present invention uses is as follows:
It is a kind of to fight the eye ground blood vessel segmentation system for generating network based on enhancing, comprising:
For being carried out at enhancing to sample label and retinal images the comparison adaptive histogram equalization manually divided
Reason improves the retinal images pretreatment unit of contrast;
By increasing the data volume of training to the retinal images pretreatment unit sample label that treated manually divides
It carries out enhancing again with retinal images to handle, to improve confrontation generation network model generalization ability and improve model robustness
Data enhancement unit;
For to data enhancement unit, treated that retinal data image is split, to extract more tiny blood
Pipe improves segmentation effect accuracy, and the confrontation for completing segmentation result generates network model.
Further, it includes generating network model and differentiation network model that the confrontation, which generates network model,;
The generation network model includes constricted path, extensions path, output layer;
The differentiation network model includes multiple convolution modules and pond layer and full articulamentum and sigmoid activation letter
Several layers.
Further, the constricted path is mainly made of multiple convolution blocks and down-sampling, the extensions path mainly by
Warp block and up-sampling composition, the output layer carries out blood vessel by sigmoid activation primitive and background pixel is classified, with reality
The fine segmentation of existing blood vessel.
Further, multiple convolution blocks in the constricted path are specially resnet convolution block.
Further, concatenate behaviour is also used in the extensions path before warp block carries out deconvolution
Make, by the image after corresponding constricted path progress convolution operation by channel link into extensions path.
Further, the confrontation generates the objective function of its confrontation generation network losses function of network model are as follows:
In formula, G (x, y) indicates that generator, D (x, y) indicate that arbiter, x are the blood vessel that people's work point is cut, and z makes a living into network
The blood vessel of generation.
Compared with the prior art, the invention has the following advantages that
1, provided by the invention that the eye ground blood vessel segmentation system for generating network is fought based on enhancing, it is suitable for view
The segmentation of film eye fundus image medium vessels.
2, provided by the invention that the eye ground blood vessel segmentation system for generating network is fought based on enhancing, confrontation generates
The convolution that network is generated in network uses the thought of resnet, greatly accelerates the training of network, and make tiny blood vessel
Branch's details can also be split.
3, provided by the invention that the eye ground blood vessel segmentation system for generating network is fought based on enhancing, by generating net
Flexible path, effectively eliminates the noise of retinal images, improves the accuracy rate of blood vessel segmentation in network.
The present invention can be widely popularized in fields such as Medical Image Processings based on the above reasons.
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 do simply to introduce, it should be apparent that, the accompanying drawings in the following description is this hair
Bright some embodiments for those of ordinary skill in the art without any creative labor, can be with
It obtains other drawings based on these drawings.
Fig. 1 is present system overall structure figure.
Fig. 2 is the generation network structure in present system.
Fig. 3 is the differentiation network configuration figure in present system.
Fig. 4 is the convolution block schematic diagram generated in network in constricted path and extensions path in present system.
Fig. 5 is segmentation result figure provided in an embodiment of the present invention.
Fig. 6 is that this system provided in an embodiment of the present invention and Wavelets, HED divide Detail contrast figure.
Specific embodiment
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the present invention can phase
Mutually combination.The present invention will be described in detail below with reference to the accompanying drawings and embodiments.
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is only
It is only a part of the embodiment of the present invention, instead of all the embodiments.It is real to the description of at least one exemplary embodiment below
It is merely illustrative on border, never as to the present invention and its application or any restrictions used.Based on the reality in the present invention
Example is applied, every other embodiment obtained by those of ordinary skill in the art without making creative efforts all belongs to
In the scope of protection of the invention.
It should be noted that term used herein above is merely to describe specific embodiment, and be not intended to restricted root
According to exemplary embodiments of the present invention.As used herein, unless the context clearly indicates otherwise, otherwise singular
Also it is intended to include plural form, additionally, it should be understood that, when in the present specification using term "comprising" and/or " packet
Include " when, indicate existing characteristics, step, operation, device, component and/or their combination.
Unless specifically stated otherwise, positioned opposite, the digital table of the component and step that otherwise illustrate in these embodiments
It is not limited the scope of the invention up to formula and numerical value.Simultaneously, it should be clear that for ease of description, each portion shown in attached drawing
The size divided not is to draw according to actual proportionate relationship.Technology known for person of ordinary skill in the relevant, side
Method and equipment may be not discussed in detail, but in the appropriate case, and the technology, method and apparatus should be considered as authorizing explanation
A part of book.In shown here and discussion all examples, appointing should be construed as merely illustratively to occurrence, and
Not by way of limitation.Therefore, the other examples of exemplary embodiment can have different values.It should also be noted that similar label
Similar terms are indicated in following attached drawing with letter, therefore, once it is defined in a certain Xiang Yi attached drawing, then subsequent attached
It does not need that it is further discussed in figure.
The present invention provides a kind of eye ground blood vessel segmentation systems that generation network is fought based on enhancing, comprising: view
Nethike embrane image pre-processing unit, data enhancement unit and confrontation generate network model;
As a preferred embodiment of the present invention, retinal images pretreatment unit is just for the sample label manually divided
With retinal images comparison adaptive histogram equalization carry out enhancing processing improve contrast, it is contemplated that by retinal images into
Row gray processing processing meeting lost part characteristic phenomenon, therefore RGB triple channel retinal images is used to carry out network training.
As a preferred embodiment of the present invention, data enhancement unit is by increasing the data volume of training to retinal images
The pretreatment unit sample label that treated manually divides and retinal images carry out enhancing again and handle, to improve to antibiosis
At network model generalization ability and improve model robustness;In the present embodiment, it is contemplated that retina data collection, DRIVE data
Collection totally 40, STARE data set 20 is opened, and number of training is few, and the present invention is by the eye fundus image of training sample and corresponding artificial
The image tag of segmentation carries out 3 degree of rotation respectively, so that a training sample becomes 120, to carry out to training sample
Data enhancing.
As a preferred embodiment of the present invention, confrontation generate network model be used for data enhancement unit treated view
Nethike embrane data image is split, and to extract more minute blood vessels, is improved segmentation effect accuracy, is completed segmentation result.This
Confrontation in embodiment generates network model, as shown in Figure 1, including generating network model and differentiation network model;
Generate network model:
In the present embodiment, generate in network model using the method for combining U-shaped network and resnet network.Such as Fig. 2
Shown, generating network model includes constricted path, extensions path and output layer.
Constricted path is mainly made of multiple resnet convolution blocks and down-sampling, and characteristic extraction part uses in constricted path
The thought of resnet network solves the problems, such as that gradient disappears or gradient is exploded, can protect the integrality of information, whole network
The differential section for only needing to learn to output and input, it is easier to optimize, convergence rate is faster.Its structure such as Fig. 3 (a) institute
Show, convolution transform optimizes adjustment using batch normalization BN, linearly using the combination of (3 × 3) BN+ReLu+Conv
Amending unit ReLu activation primitive is effectively reduced gradient in backpropagation and disappears, and ReLu activation primitive is as follows:
ReLU (x)=max (x, 0)
Down-sampling reduces parameter complexity using the maximum Chi Hualai compression and extraction feature of 2x2.
Extensions path is mainly made of warp block and up-sampling, is also used before warp block carries out deconvolution
Image after constricted path convolution on the left of U-shaped network is linked to by channel skip and is corresponded by concatenate operation
Extensions path in, solve the problems, such as in constricted path because down-sampling causes characteristic information to be lost.Its convolutional coding structure is such as
Shown in Fig. 3 (b), convolution transform expands receptive field using the combination of 2 Conv (3 × 3)+BN+ReLu.
Output layer carries out blood vessel by sigmoid activation primitive and background pixel is classified, to realize the fine segmentation of blood vessel.This
In embodiment, the convolution of 1x1 is used to realize information exchange and information integration across channel, and used sigmoid activation
Function carries out 2 classification and achievees the effect that blood vessel segmentation.
Differentiate network model:
Differentiate that network model includes multiple convolution modules and pond layer and full articulamentum and sigmoid activation primitive layer.
In the present embodiment, as shown in figure 4, differentiating that the entire conventional part of network model uses 2 convolution module Conv (3 × 3) at first 4 layers
Then+BN+ReLu combination and maximum pond are carried out complete in layer 5 using 2 Conv (3 × 3)+BN+ReLu and average pond
Attended operation finally carries out 2 classification using sigmoid activation primitive.
As a preferred embodiment of the present invention, in the present embodiment, its confrontation of confrontation generation network model generates network and exists
On segmentation problem, realized in such a way that optimization fights and generates network losses function, the objective function after optimization are as follows:
In formula, G (x, y) indicates that generator, D (x, y) indicate that arbiter, x are the blood vessel that people's work point is cut, and z makes a living into network
The blood vessel of generation.The present invention uses intersection entropy function, is updated and is identified by Adam gradient descent algorithm in backpropagation
Network and the parameter for generating network.
By above step, as shown in Fig. 5 (b), the finally obtained retina segmentation figure after iteration is complete.Such as Fig. 6 institute
Show, STARE data set number be im0255 blood-vessel image, present system compared with Wavelets algorithm, HED algorithm,
The present invention is obvious for the connection effect of rupture of blood vessel point, and when minor detail punishment is cut, blood vessel is more clear.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations;Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used
To modify to technical solution documented by previous embodiment, or some or all of the technical features are equal
Replacement;And these are modified or replaceed, the model for technical solution of the embodiment of the present invention that it does not separate the essence of the corresponding technical solution
It encloses.
Claims (6)
1. a kind of fight the eye ground blood vessel segmentation system for generating network based on enhancing characterized by comprising
It is mentioned for carrying out enhancing processing to sample label and retinal images the comparison adaptive histogram equalization manually divided
The retinal images pretreatment unit of high contrast;
By increasing the data volume of training to retinal images pretreatment unit sample label and view that treated manually divides
Nethike embrane image carries out enhancing again and handles, to improve the number for fighting and generating network model generalization ability and improving model robustness
According to enhancement unit;
For to data enhancement unit, treated that retinal data image is split, to extract more minute blood vessels, mention
High segmentation effect accuracy, the confrontation for completing segmentation result generate network model.
2. according to claim 1 fight the eye ground blood vessel segmentation system for generating network, feature based on enhancing
It is, it includes generating network model and differentiation network model that the confrontation, which generates network model,;
The generation network model includes constricted path, extensions path, output layer;
The differentiation network model includes multiple convolution modules and pond layer and full articulamentum and sigmoid activation primitive
Layer.
3. according to claim 1 or 2 fight the eye ground blood vessel segmentation system for generating network, spy based on enhancing
Sign is that the constricted path is mainly made of multiple convolution blocks and down-sampling, the extensions path mainly by warp block and
Up-sampling composition, the output layer carries out blood vessel by sigmoid activation primitive and background pixel is classified, to realize the essence of blood vessel
Subdivision is cut.
4. according to claim 3 fight the eye ground blood vessel segmentation system for generating network, feature based on enhancing
It is, multiple convolution blocks in the constricted path are specially resnet convolution block.
5. according to claim 3 fight the eye ground blood vessel segmentation system for generating network, feature based on enhancing
It is, also uses concatenate operation in the extensions path before warp block carries out deconvolution, it will in contrast
Image after the constricted path progress convolution operation answered is by channel link into extensions path.
6. according to claim 1 fight the eye ground blood vessel segmentation system for generating network, feature based on enhancing
It is, the confrontation generates the objective function of its confrontation generation network losses function of network model are as follows:
In formula, G (x, y) indicates that generator, D (x, y) indicate that arbiter, x are the blood vessel that people's work point is cut, and z makes a living into network and generates
Blood vessel.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910797551.8A CN110517250A (en) | 2019-08-27 | 2019-08-27 | It is a kind of to fight the eye ground blood vessel segmentation system for generating network based on enhancing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910797551.8A CN110517250A (en) | 2019-08-27 | 2019-08-27 | It is a kind of to fight the eye ground blood vessel segmentation system for generating network based on enhancing |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110517250A true CN110517250A (en) | 2019-11-29 |
Family
ID=68627304
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910797551.8A Pending CN110517250A (en) | 2019-08-27 | 2019-08-27 | It is a kind of to fight the eye ground blood vessel segmentation system for generating network based on enhancing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110517250A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111882566A (en) * | 2020-07-31 | 2020-11-03 | 华南理工大学 | Blood vessel segmentation method, device, equipment and storage medium of retina image |
CN112070767A (en) * | 2020-09-10 | 2020-12-11 | 哈尔滨理工大学 | Micro-vessel segmentation method in microscopic image based on generating type countermeasure network |
CN113112411A (en) * | 2020-01-13 | 2021-07-13 | 南京信息工程大学 | Human face image semantic restoration method based on multi-scale feature fusion |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107358612A (en) * | 2017-07-07 | 2017-11-17 | 东北大学 | A kind of retinal vessel segmenting system combined based on fractal dimension with gaussian filtering and method |
CN108154512A (en) * | 2017-11-08 | 2018-06-12 | 东北大学 | It is a kind of based on the multiple retinal images blood vessel segmentation system for going trend analysis |
CN108537801A (en) * | 2018-03-29 | 2018-09-14 | 山东大学 | Based on the retinal angiomatous image partition method for generating confrontation network |
CN109300107A (en) * | 2018-07-24 | 2019-02-01 | 深圳先进技术研究院 | Patch processing method, device and the calculating equipment of magnetic resonance vascular wall imaging |
CN109448006A (en) * | 2018-11-01 | 2019-03-08 | 江西理工大学 | A kind of U-shaped intensive connection Segmentation Method of Retinal Blood Vessels of attention mechanism |
CN109727259A (en) * | 2019-01-07 | 2019-05-07 | 哈尔滨理工大学 | A kind of retinal images partitioning algorithm based on residual error U-NET network |
-
2019
- 2019-08-27 CN CN201910797551.8A patent/CN110517250A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107358612A (en) * | 2017-07-07 | 2017-11-17 | 东北大学 | A kind of retinal vessel segmenting system combined based on fractal dimension with gaussian filtering and method |
CN108154512A (en) * | 2017-11-08 | 2018-06-12 | 东北大学 | It is a kind of based on the multiple retinal images blood vessel segmentation system for going trend analysis |
CN108537801A (en) * | 2018-03-29 | 2018-09-14 | 山东大学 | Based on the retinal angiomatous image partition method for generating confrontation network |
CN109300107A (en) * | 2018-07-24 | 2019-02-01 | 深圳先进技术研究院 | Patch processing method, device and the calculating equipment of magnetic resonance vascular wall imaging |
CN109448006A (en) * | 2018-11-01 | 2019-03-08 | 江西理工大学 | A kind of U-shaped intensive connection Segmentation Method of Retinal Blood Vessels of attention mechanism |
CN109727259A (en) * | 2019-01-07 | 2019-05-07 | 哈尔滨理工大学 | A kind of retinal images partitioning algorithm based on residual error U-NET network |
Non-Patent Citations (1)
Title |
---|
刘美丽: "《MATLAB语言与应用》", 国防工业出版社 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113112411A (en) * | 2020-01-13 | 2021-07-13 | 南京信息工程大学 | Human face image semantic restoration method based on multi-scale feature fusion |
CN113112411B (en) * | 2020-01-13 | 2023-11-24 | 南京信息工程大学 | Human face image semantic restoration method based on multi-scale feature fusion |
CN111882566A (en) * | 2020-07-31 | 2020-11-03 | 华南理工大学 | Blood vessel segmentation method, device, equipment and storage medium of retina image |
CN111882566B (en) * | 2020-07-31 | 2023-09-19 | 华南理工大学 | Blood vessel segmentation method, device, equipment and storage medium for retina image |
CN112070767A (en) * | 2020-09-10 | 2020-12-11 | 哈尔滨理工大学 | Micro-vessel segmentation method in microscopic image based on generating type countermeasure network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109615582B (en) | Face image super-resolution reconstruction method for generating countermeasure network based on attribute description | |
CN110517250A (en) | It is a kind of to fight the eye ground blood vessel segmentation system for generating network based on enhancing | |
US11308587B2 (en) | Learning method of generative adversarial network with multiple generators for image denoising | |
CN108198184B (en) | Method and system for vessel segmentation in contrast images | |
CN109685813A (en) | A kind of U-shaped Segmentation Method of Retinal Blood Vessels of adaptive scale information | |
IL308022A (en) | Neural network for eye image segmentation and image quality estimation | |
CN110097554A (en) | The Segmentation Method of Retinal Blood Vessels of convolution is separated based on intensive convolution sum depth | |
CN113066026B (en) | Endoscope image smoke purification method based on deep neural network | |
CN108764342B (en) | Semantic segmentation method for optic discs and optic cups in fundus image | |
CN112258488A (en) | Medical image focus segmentation method | |
CN111353390A (en) | Micro-expression recognition method based on deep learning | |
Guo et al. | Simultaneous classification and segmentation of intracranial hemorrhage using a fully convolutional neural network | |
Dhal et al. | Colour retinal images enhancement using modified histogram equalisation methods and firefly algorithm | |
CN111815562A (en) | Retinal vessel segmentation method combining U-Net and self-adaptive PCNN | |
CN112541924B (en) | Fundus image generation method, fundus image generation device, fundus image generation apparatus, and fundus image storage medium | |
Kumar et al. | Image processing in diabetic related causes | |
CN107610145B (en) | Automatic pancreas segmentation method based on adaptive threshold and template matching | |
Ghosh et al. | A novel approach of retinal image enhancement using PSO system and measure of fuzziness | |
Keerthiveena et al. | Computer‐aided diagnosis of retinal diseases using multidomain feature fusion | |
CN115018756A (en) | Method and device for classifying artery and vein of retina and storage medium | |
Zhang et al. | X-Net: A binocular summation network for foreground segmentation | |
Mehidi et al. | A high accuracy segmentation method for retinal blood vessel detection based on hybrid filters and an adaptive thresholding | |
Li et al. | Retinal Vessel Segmentation Network Based on Patch-GAN | |
hamzah Abed et al. | Diabetic retinopathy diagnosis based on convolutional neural network | |
Pai | Region Detection and Segmentation of Brain Hemorrhage using Algorithmic Approach of Image Processing |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191129 |
|
RJ01 | Rejection of invention patent application after publication |