CN103914841B - Based on the segmentation of the vaginal bacteria of super-pixel and deep learning and categorizing system - Google Patents
Based on the segmentation of the vaginal bacteria of super-pixel and deep learning and categorizing system Download PDFInfo
- Publication number
- CN103914841B CN103914841B CN201410133805.3A CN201410133805A CN103914841B CN 103914841 B CN103914841 B CN 103914841B CN 201410133805 A CN201410133805 A CN 201410133805A CN 103914841 B CN103914841 B CN 103914841B
- Authority
- CN
- China
- Prior art keywords
- super
- pixel
- deep learning
- segmentation
- module
- 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
Links
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
Present invention offer is a kind of to establish deep learning grader based on the segmentation of the vaginal bacteria of super-pixel and deep learning and categorizing system, including by training image collection, and test image and deep learning grader is contrasted, the result of output category;The contrast includes:The test image is done into medium filtering by each RGB channel, and removes noise;Calculate super-pixel;Calculate the color, shape, size characteristic in each super-pixel region;Each super-pixel region is tentatively filtered using priori, and split to determine candidate bacterial region;Feature extraction is carried out to candidate bacterial region;Criteria for classification in the feature and grader of extraction is contrasted, completes to classify according to similarity.The present invention is calculated by super-pixel to be split and is classified by deep learning, has the advantages of high identification, low cost, implement simply and easily promote.
Description
Technical field
The present invention relates to a kind of segmentation of image and sorting technique, more particularly to a kind of the moon based on super-pixel and deep learning
Road bacterium segmentation and categorizing system.
Background technology
By bacterial vaginal disease or infection (vaginal candida/trichomonad, gonococcus, Chlamydia) be among women most
Common disease, also cause the very high death rate.By bacteria types and the classification for carrying out vaginal disease is counted in clinical practice
In have very important effect.At present, widely using prevents that the diagnostic method of vaginal disease from being cytology examination, such a method
The experience of clinician is heavily dependent on, accurate diagnostic result is obtained with this.In order to solve this problem, design and
The automatic diagnostics of exploitation prevents vaginal disease, inflammation and cancer, has attracted the interest of many people, and turn into a focus.
Previously there is substantial amounts of research work to split nucleus, but only utilized the segmentation performance of nucleus information approach
It is still unsatisfactory.The dividing method of nucleus is broadly divided into multiple nucleus segmentation thresholds and chosen, Hough transform and
Watershed methods.However, these methods are the nucleus based on normal condition, rather than there is the nucleus of disease or exception.
Recently, segmentation also has pathology and normal cell, and the report of cell mixing segmentation.
Such as Yeoman, C.J., Thomas, S.M., Miller, M.E.B., Ulanov, A.V., Torralba, M.,
Lucas,S.,Gillis,M.,Cregger,M.,Gomez,A.,Ho,M.,Leigh,S.R.,Stumpf,R.,Creedon,
D.J.,Smith,M.A.,Weisbaum,J.S.,Nelson,K.E.,Wilson,B.A.,White,B.A.:AMulti-Omic
Systems-Based Approach Reveals Metabolic Markers of Bacterial Vaginosis and
Insight into the Disease.PLoS ONE 8(2),e56111(2013)[1]。
The work that cell is split from image, current many algorithms are all based on independent nucleus, and to a variety of
The segmentation research of cell is less.Parametric filtering, movable contour model can be divided into based on shape and the segmentation of marginal information nucleus
Maximized with difference.In addition, segmentation work has also obtained significant performance because merging previous knowledge and local feature.
Therefore, how to design it is a kind of based on the bacterium of super-pixel and deep learning segmentation with sorting technique and its application,
As new R&D direction.
The content of the invention
The present invention is directed to this problem, there is provided a kind of to be split and system of classifying based on the vaginal bacteria of super-pixel and deep learning
System.
It is provided by the invention based on the vaginal bacteria of super-pixel and deep learning segmentation and categorizing system, including:Training figure
Image set, for storing the sample image of vaginal bacteria;Deep learning grader, for storing based on obtained by the training image collection
The criteria for classification arrived;Pretreatment module, for test image to be pre-processed, calculate super-pixel and calculate each super picture
The color in plain region, shape, size characteristic, each super-pixel region is tentatively filtered using priori, and divided
Cut to determine candidate bacterial region, priori includes:The numerical value upper and lower limit of each RGB channel in bacterial region, bacterium length and width
Upper and lower limit, the upper and lower limit of bacterium color average, bacterium area, the upper and lower limit of girth;Characteristic extracting module, for described
Candidate bacterial region carries out feature extraction;Judge module, for by point in the feature of extraction and the deep learning grader
Class standard is contrasted, and completes to classify according to similarity;Wherein, the vaginal bacteria includes positive bacillus, negative bacillus, the positive
Coccus and negative cocci, the positive bacillus and positive cocci are in blueness, negative bacillus and the negative cocci pinkiness, institute
It is in shaft-like to state positive bacillus and negative bacillus, and the positive cocci and negative cocci are in spherical;The deep learning grader is also
Including:Initialization module, for initializing convolutional neural networks;Receiving module, for receiving from the training image collection
Image pattern;Adjusting module, for calculating reality output, the difference of calculating reality output and target output to transmission and utilizing pole
The backpropagation of smallization error approach adjusts weights;Counting module, for judging whether to reach predetermined frequency of training, and work as and reach
Afterwards, adjusting module output category standard is notified.
Preferably, the super-pixel that calculates is the algorithm using linear iteraction cluster.
The present invention based on the vaginal bacteria of super-pixel and deep learning segmentation and categorizing system, by super-pixel calculate into
Row is split and classified by deep learning, has the advantages of high identification, low cost, implement simply and easily promote.
Brief description of the drawings
Fig. 1 is the schematic flow sheet for the method that deep learning grader is established in the present invention.
Fig. 2 is schematic flow sheet of the present invention based on the segmentation of the bacterium of super-pixel and deep learning with sorting technique.
Fig. 3 is structural representation of the present invention based on the segmentation of the bacterium of super-pixel and deep learning with categorizing system.
Fig. 4 is the schematic diagram of medial vagina common bacteria of the present invention.
Fig. 5 is being calculated and the segmentation carried out and the effect diagram of feature extraction based on super-pixel in the present invention.
Fig. 6 is typical vagina image and the result to vaginal bacteria segmentation and classification in the present invention.
Fig. 7 is the exemplary plot of segmentation effect in the present invention.
Fig. 8 is the exemplary plot for splitting comparative result in the present invention.
Fig. 9 is the classification results comparative example figure of different classifications device in the present invention.
Embodiment
The invention provides a kind of based on the segmentation of the bacterium of super-pixel and deep learning and sorting technique, can be applied to all kinds of
The identification and classification of bacterium, have the advantages that quick, accurate, cost is cheap, idiographic flow refers to embodiment 1, additionally provides
Application of the above method in the diagnosis of vaginal bacteria, specific system architecture refer to embodiment 2.
Embodiment 1
It is shown incorporated by reference to Fig. 1 and Fig. 2, to be split and classification side based on the bacterium of super-pixel and deep learning in the present invention
Method.Mainly include the following steps that:
First, deep learning grader is established by training image collection;
2nd, test image and deep learning grader are contrasted, the result of output category.
Referring to Fig. 1, showing the step of establishing deep learning grader, specifically include:
In step S101, convolutional neural networks are initialized.
In step s 102, the image pattern from the training image collection is received.
In step s 103, positive transmit calculates reality output.
In step S104, reality output and the difference of target output are calculated.
In step S105, weights are adjusted using the backpropagation of minimization error approach.
In step s 106, judge whether to reach predetermined frequency of training.
If not up to, return to step S102 is after the above-mentioned image pattern of reception, and adjusts weights.
If reached, in step s 107, then output category standard and priori, and be stored in the deep learning point
In class device.
Referring to Fig. 2, it show based on the segmentation of the bacterium of super-pixel and deep learning and the flow chart of sorting technique, the figure
What is more stressed is the contrast step of test image, including:
In step s 201, the test image is done into medium filtering by each GRB passages, and removes noise.
In step S202, super-pixel is calculated.
Due to super-pixel algorithm closely with it is flexible the characteristics of, and can by with perceive meaning atomic region pixel
Flock together, therefore, applied to cost and labour can be reduced in computer aided calculation.
In step S203, the color, shape, size characteristic in each super-pixel region are calculated.
In step S204, each super-pixel region is tentatively filtered using priori, and is split with true
Determine candidate bacterial region.
Wherein, priori, including:The numerical value upper and lower limit of each RGB channel in bacterial region, the upper and lower limit of bacterium length and width,
The upper and lower limit of bacterial region color average, bacterium area, the upper and lower limit of girth, that is, by deep learning classification chart right
The rough estimates value that enough samples are done.
In step S205, feature extraction is carried out to the candidate bacterial region.
In step S206, the feature of extraction and the criteria for classification in the grader are contrasted.
In step S207, complete to classify according to similarity.
Similarity, that is, the categorized device of feature of candidate region criteria for classification contrasted after score, test institute
Fraction it is more close with the numerical value of the label of affiliated bacteria types, it is higher to be considered as similarity, and just the test zone is classified as
Change the bacterium of type.
In other embodiments, when similarity is less than certain predetermined threshold value, according to the state of an illness or the needs of pathology, to similar
Low candidate region is spent to be abandoned or paid close attention to.
The present invention based on the bacterium of super-pixel and deep learning segmentation and sorting technique, by super-pixel calculate divided
Cut and classified by deep learning, there is the advantages of high identification, low cost, implement simply and easily promote.
Embodiment 2
Referring to Fig. 3, it show a kind of based on the segmentation of the vaginal bacteria of super-pixel and deep learning and classification in the present invention
System.The system include training image collection 10, deep learning grader 20, pretreatment module 30, characteristic extracting module 40 with
And judge module 50.
Training image collection 10, for storing the sample image of vaginal bacteria.
Deep learning grader 20, for storing based on the criteria for classification obtained by the training image collection.
In the present embodiment, the deep learning grader 20, in addition to:Initialization module 21, rolled up for initializing
Product neutral net;Receiving module 22, for receiving the image pattern from the training image collection 10;Adjusting module 23, is used for
Forward direction, which is transmitted, to be calculated reality output, the difference of calculating reality output and target output and utilizes minimization error approach backpropagation
Adjust weights;Counting module 24, for judging whether to reach predetermined frequency of training, and after reaching, notify adjusting module 23
Output category standard.
Referring to Fig. 4, it show the description of the criteria for classification of vaginal bacteria.Learnt from practice, vaginal bacteria is divided into four kinds
Type:Including positive bacillus, negative bacillus, positive cocci and negative cocci.
Distinguished from color:Positive is general in blueness, negative general pinkiness;
Distinguished from shape:Bacillus typically in shaft-like, coccus typically in spherical, but it is overlapping in the case of have different shapes
State;
Distinguished from area:Positive bacillus is both greater than 250 substantially, and negative and positive bacillus is both greater than 150 but is slightly smaller than the positive substantially,
The area of positive cocci and negative cocci is generally less than 200.
In addition, deep learning grader 20, being additionally operable to be in course of adjustment calculates and stores priori, including:Bacterium
The numerical value upper and lower limit of each RGB channel in region, the upper and lower limit of bacterium length and width, the upper and lower limit of bacterial region color average, bacterium
The upper and lower limit of area, girth.
Pretreatment module 30, for test image to be pre-processed, calculate super-pixel and calculate each super-pixel area
The color in domain, shape, size characteristic, each super-pixel region is tentatively filtered using priori, and split with
Determine candidate bacterial region.
Wherein, super-pixel is calculated, the algorithm of the super-pixel of simple linear iteraction cluster (SLIC) can be utilized, be bacterium
Good border can be provided.SLIC algorithms have the characteristics of memory simply, quickly and efficiently, can pass through a pretreatment
Step reduces complexity and improves speed.And there is cluster advantage and superperformance, it is especially suitable for splitting bacterium.Feature extraction mould
Block 40, for carrying out feature extraction to the candidate bacterial region.
Referring to Fig. 5, show segmentation that pretreatment module and characteristic extracting module calculated and carried out based on super-pixel and
The design sketch of feature extraction.Characteristic value resolving power in Fig. 5 is higher, can quick separating go out coccus and bacillus.Each super-pixel
Border there is real image boundary to be closely related.Also, because the border of image is retained, the method for super-pixel is to segmentation
It is highly effective, and ensure that the accuracy of its follow-up extraction.
Judge module 50, for the feature of extraction and the criteria for classification in the deep learning grader to be contrasted,
Complete to classify according to similarity.
Similarity, that is, the categorized device of feature of candidate region criteria for classification contrasted after score, test institute
Fraction it is more close with the numerical value of the label of affiliated bacteria types, it is higher to be considered as similarity, and just the test zone is classified as
Change the bacterium of type.
In other embodiments, when similarity is less than certain predetermined threshold value, according to the state of an illness or the needs of pathology, to similar
Low candidate region is spent to be abandoned or paid close attention to.
Experimental result
First, material and data
Collection for vaginal cell data comes from Shenzhen South Mountain hospital, to the women of 18 to 50 years old, have collected 105 altogether
Individual slide.Then Olympus BX43 microscopes are used, (oil mirror, its numerical aperture are under conditions of being 100 times in objective lens magnification
1.25) 105 relatively good visuals field, are at least gathered to each slide as processing data, obtained picture is 1360*1024's
Jpg rgb images.After obtaining initial data, under the guidance of doctor, using the picture that doctor frequently encounters at work as mark
Standard, test pictures of 319 pictures as algorithm are have chosen altogether.
2nd, segmentation result
The popular approach of assessment to systematic function such as accuracy rate, recall rate, accuracy, sensitivity, specificity, F1 estimate
All it is used with Zijdenbos index of similarity (ZSI) etc..Based in 6 kinds of typical cell images, the hand of different bacterium is identified
Dynamic and automatic segmentation result comparison is as shown in Figure 6.From fig. 6, it can be seen that the manual operations based on doctors experience, automatic segmentation
Effect it is also very good.Quantitative analysis for segmentation result is shown in Fig. 7.Segmentation performance is very good as can be seen from Figure 7.
Fig. 8 lists the comparison on the distinct methods in segmentation result.It will be seen that the segmentation result proposed is better than background skill
Partitioning algorithm used in art [1].
3rd, classification results
Utilize different graders such as deep learning (CNN), reverse transmittance nerve network (BNN), probabilistic neural network
(PNN), SVMs (SVM) and learning vector quantizations (LVQ) are classified to vaginal cell, and carry out performance comparision.BNN
It is the sorting technique based on supervision, this method adjusts the weight of each layer by using gradient minimisation object function.
Fig. 9 summarizes the performance based on the different grader of representational feature.The result quantitative from Fig. 9 can be seen that
CNN sorting technique is better than other sorting techniques, while the method for indicating deep learning is very effective vaginal bacteria point
The method of class.However, because its limited Internet, BNN result are quite poor.It will be apparent that the classification to vaginal bacteria
Can be sufficiently high.From the results, it was seen that the classification of vaginal cell meets the requirement used in real life.
4th, conclusion
Apply segmentation and classification with vaginal cell to be studied super-pixel and deep learning art herein, using this two
Vaginal bacteria is divided into four classes by kind technology.Test result indicates that segmentation of the algorithm of super-pixel segmentation to vaginal bacteria is closed very much
It is suitable, achieve the effect of highly significant.Super-pixel label is trained based on deep learning, and vaginal bacteria is identified, is also obtained
Obtain extraordinary Classification and Identification result.The algorithm proposed obtains the effect significantly split and classified and shown, vaginal bacteria is certainly
Dynamic classification and identification contribute to the diagnosis of doctor, and accuracy rate is high.
Described above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
Member, under the premise without departing from the principles of the invention, can also make some improvements and modifications, these improvements and modifications also should be regarded as
Protection scope of the present invention.
Claims (3)
- It is 1. a kind of based on the segmentation of the vaginal bacteria of super-pixel and deep learning and categorizing system, it is characterised in that including:Training image collection, for storing the sample image of vaginal bacteria;Deep learning grader, for storing based on the criteria for classification obtained by the training image collection;Pretreatment module, for test image to be pre-processed, calculate super-pixel and calculate the face in each super-pixel region Color, shape, size characteristic, each super-pixel region is tentatively filtered using priori, and split to determine to wait Bacterial region is selected, the priori includes:The numerical value upper and lower limit of each RGB channel in bacterial region, bacterium length and width it is upper and lower Limit, the upper and lower limit of bacterium color average, bacterium area, the upper and lower limit of girth;Characteristic extracting module, for carrying out feature extraction to the candidate bacterial region;Judge module, for the feature of extraction and the criteria for classification in the deep learning grader to be contrasted, according to phase Classification is completed like degree;Wherein, the vaginal bacteria includes positive bacillus, negative bacillus, positive cocci and negative cocci, the positive bacillus and For positive cocci in blueness, negative bacillus and the negative cocci pinkiness, the positive bacillus and negative bacillus are in shaft-like, institute Positive cocci and negative cocci are stated in spherical;The deep learning grader also includes:Initialization module, for initializing convolutional neural networks;Receiving module, for receiving the image pattern from the training image collection;Adjusting module, for the positive difference transmitted calculating reality output, calculate reality output and target output and utilize minimization Error approach backpropagation adjusts weights;Counting module, for judging whether to reach predetermined frequency of training, and after reaching, notify adjusting module output category mark It is accurate.
- 2. vaginal bacteria segmentation as claimed in claim 1 and categorizing system, it is characterised in that the calculating super-pixel is to use The algorithm of linear iteraction cluster.
- 3. vaginal bacteria segmentation as claimed in claim 1 and categorizing system, it is characterised in that when the similarity is less than default During threshold value, its corresponding candidate bacterial region is abandoned or paid close attention to.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410133805.3A CN103914841B (en) | 2014-04-03 | 2014-04-03 | Based on the segmentation of the vaginal bacteria of super-pixel and deep learning and categorizing system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410133805.3A CN103914841B (en) | 2014-04-03 | 2014-04-03 | Based on the segmentation of the vaginal bacteria of super-pixel and deep learning and categorizing system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103914841A CN103914841A (en) | 2014-07-09 |
CN103914841B true CN103914841B (en) | 2018-03-09 |
Family
ID=51040497
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410133805.3A Active CN103914841B (en) | 2014-04-03 | 2014-04-03 | Based on the segmentation of the vaginal bacteria of super-pixel and deep learning and categorizing system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103914841B (en) |
Families Citing this family (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104866727A (en) * | 2015-06-02 | 2015-08-26 | 陈宽 | Deep learning-based method for analyzing medical data and intelligent analyzer thereof |
CN105117611B (en) * | 2015-09-23 | 2018-06-12 | 北京科技大学 | Based on the determining method and system of the TCM tongue diagnosis model of convolutional Neural metanetwork |
CN106651886A (en) * | 2017-01-03 | 2017-05-10 | 北京工业大学 | Cloud image segmentation method based on superpixel clustering optimization CNN |
WO2019073312A1 (en) * | 2017-10-13 | 2019-04-18 | Sigtuple Technologies Private Limited | Method and device for integrating image channels in a deep learning model for classification |
CN107945144B (en) * | 2017-11-20 | 2022-02-25 | 宁波江丰生物信息技术有限公司 | Mycobacterium tuberculosis fluorescence sputum smear identification method based on deep neural network |
CN108231190B (en) * | 2017-12-12 | 2020-10-30 | 北京市商汤科技开发有限公司 | Method of processing image, neural network system, device, and medium |
CN108447047A (en) * | 2018-02-11 | 2018-08-24 | 深圳市恒扬数据股份有限公司 | Acid-fast bacilli detection method and device |
CN108629768B (en) * | 2018-04-29 | 2022-01-21 | 山东省计算中心(国家超级计算济南中心) | Method for segmenting epithelial tissue in esophageal pathology image |
CN108961301B (en) * | 2018-07-12 | 2021-11-02 | 中国海洋大学 | Chaetoceros image segmentation method based on unsupervised pixel-by-pixel classification |
JP7134421B2 (en) * | 2019-01-28 | 2022-09-12 | 日本たばこ産業株式会社 | Sample quality determination method, program, and device using fluorescence image |
CN112907456B (en) * | 2019-12-04 | 2022-06-10 | 四川大学 | Deep neural network image denoising method based on global smooth constraint prior model |
CN111062346A (en) * | 2019-12-21 | 2020-04-24 | 电子科技大学 | Automatic leukocyte positioning detection and classification recognition system and method |
CN111429361A (en) * | 2020-02-17 | 2020-07-17 | 马健锦 | Bacterial colony counting method based on machine vision |
CN112949378A (en) * | 2020-12-30 | 2021-06-11 | 至微生物智能科技(厦门)有限公司 | Bacterial microscopic image segmentation method based on deep learning network |
CN112966645B (en) * | 2021-03-24 | 2022-04-08 | 山东仕达思生物产业有限公司 | Intelligent detection and classification counting method for multiple types of bacilli in gynecological microecology |
CN116602712A (en) * | 2023-05-25 | 2023-08-18 | 广州斐博生物科技有限公司 | Female vaginal secretion sample based evaluation system |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101900737A (en) * | 2010-06-10 | 2010-12-01 | 上海理工大学 | Automatic identification system for urinary sediment visible components based on support vector machine |
CN102750544A (en) * | 2012-06-01 | 2012-10-24 | 浙江捷尚视觉科技有限公司 | Detection system and detection method of rule-breaking driving that safety belt is not fastened and based on plate number recognition |
WO2013037357A1 (en) * | 2011-09-14 | 2013-03-21 | Eads Deutschland Gmbh | Automatic learning method for the automatic learning of forms of appearance of objects in images |
CN103093470A (en) * | 2013-01-23 | 2013-05-08 | 天津大学 | Rapid multi-modal image synergy segmentation method with unrelated scale feature |
CN103164858A (en) * | 2013-03-20 | 2013-06-19 | 浙江大学 | Adhered crowd segmenting and tracking methods based on superpixel and graph model |
US8526723B2 (en) * | 2009-06-23 | 2013-09-03 | Los Alamos National Security, Llc | System and method for the detection of anomalies in an image |
CN103353938A (en) * | 2013-06-14 | 2013-10-16 | 山东大学 | Cell membrane segmentation method based on hierarchy-level characteristic |
CN103400368A (en) * | 2013-06-25 | 2013-11-20 | 西安电子科技大学 | Parallel rapid SAR image segmentation method based on graph theory and superpixel |
CN103544697A (en) * | 2013-09-30 | 2014-01-29 | 南京信息工程大学 | Hypergraph spectrum analysis based image segmentation method |
CN103679719A (en) * | 2013-12-06 | 2014-03-26 | 河海大学 | Image segmentation method |
-
2014
- 2014-04-03 CN CN201410133805.3A patent/CN103914841B/en active Active
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8526723B2 (en) * | 2009-06-23 | 2013-09-03 | Los Alamos National Security, Llc | System and method for the detection of anomalies in an image |
CN101900737A (en) * | 2010-06-10 | 2010-12-01 | 上海理工大学 | Automatic identification system for urinary sediment visible components based on support vector machine |
WO2013037357A1 (en) * | 2011-09-14 | 2013-03-21 | Eads Deutschland Gmbh | Automatic learning method for the automatic learning of forms of appearance of objects in images |
CN102750544A (en) * | 2012-06-01 | 2012-10-24 | 浙江捷尚视觉科技有限公司 | Detection system and detection method of rule-breaking driving that safety belt is not fastened and based on plate number recognition |
CN103093470A (en) * | 2013-01-23 | 2013-05-08 | 天津大学 | Rapid multi-modal image synergy segmentation method with unrelated scale feature |
CN103164858A (en) * | 2013-03-20 | 2013-06-19 | 浙江大学 | Adhered crowd segmenting and tracking methods based on superpixel and graph model |
CN103353938A (en) * | 2013-06-14 | 2013-10-16 | 山东大学 | Cell membrane segmentation method based on hierarchy-level characteristic |
CN103400368A (en) * | 2013-06-25 | 2013-11-20 | 西安电子科技大学 | Parallel rapid SAR image segmentation method based on graph theory and superpixel |
CN103544697A (en) * | 2013-09-30 | 2014-01-29 | 南京信息工程大学 | Hypergraph spectrum analysis based image segmentation method |
CN103679719A (en) * | 2013-12-06 | 2014-03-26 | 河海大学 | Image segmentation method |
Non-Patent Citations (2)
Title |
---|
基于卷积神经网络的模式分类器;李葆青;《大连大学学报》;20030430;第24卷(第2期);全文 * |
细胞病理图像的特征分析与分类识别;赵明珠;《中国优秀硕士学位论文全文数据库 信息科技辑》;20130715(第7期);全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN103914841A (en) | 2014-07-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103914841B (en) | Based on the segmentation of the vaginal bacteria of super-pixel and deep learning and categorizing system | |
CN110428432B (en) | Deep neural network algorithm for automatically segmenting colon gland image | |
Chagas et al. | Classification of glomerular hypercellularity using convolutional features and support vector machine | |
Ta et al. | Graph-based tools for microscopic cellular image segmentation | |
Salem et al. | Segmentation of retinal blood vessels using a novel clustering algorithm (RACAL) with a partial supervision strategy | |
Tashk et al. | An automatic mitosis detection method for breast cancer histopathology slide images based on objective and pixel-wise textural features classification | |
CN105122308A (en) | Systems and methods for multiplexed biomarker quantitation using single cell segmentation on sequentially stained tissue | |
Stoklasa et al. | Efficient k-NN based HEp-2 cells classifier | |
WO2021227708A1 (en) | Method and device for identifying abnormal cell in sample to be tested, and storage medium | |
CN101900737A (en) | Automatic identification system for urinary sediment visible components based on support vector machine | |
CN114445670B (en) | Training method, device and equipment of image processing model and storage medium | |
Mao et al. | Iteratively training classifiers for circulating tumor cell detection | |
Shanthi et al. | Deep convolution neural network for malignancy detection and classification in microscopic uterine cervix cell images | |
Zhang et al. | Segmentation of overlapping cells in cervical smears based on spatial relationship and overlapping translucency light transmission model | |
Mao et al. | A deep convolutional neural network trained on representative samples for circulating tumor cell detection | |
CN104156945A (en) | Method for segmenting gray scale image based on multi-objective particle swarm optimization algorithm | |
Sornapudi et al. | Comparing deep learning models for multi-cell classification in liquid-based cervical cytology image | |
CN112784767A (en) | Cell example segmentation algorithm based on leukocyte microscopic image | |
Rulaningtyas et al. | Multi patch approach in K-means clustering method for color image segmentation in pulmonary tuberculosis identification | |
Sheela Shiney et al. | Deep auto encoder based extreme learning system for automatic segmentation of cervical cells | |
CN115206495A (en) | Renal cancer pathological image analysis method and system based on CoAtNet deep learning and intelligent microscopic device | |
Bokhorst et al. | Automatic detection of tumor budding in colorectal carcinoma with deep learning | |
Arya et al. | Three segmentation techniques to predict the dysplasia in cervical cells in the presence of debris | |
Sulaiman et al. | Semi-automated pseudo colour features extraction technique for cervical cancer's pap smear images | |
Vale et al. | Automatic segmentation and classification of blood components in microscopic images using a fuzzy approach |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |