CN110110807A - A kind of leucocyte extraction and classification method based on improvement K-means and convolutional neural networks - Google Patents
A kind of leucocyte extraction and classification method based on improvement K-means and convolutional neural networks Download PDFInfo
- Publication number
- CN110110807A CN110110807A CN201910404623.8A CN201910404623A CN110110807A CN 110110807 A CN110110807 A CN 110110807A CN 201910404623 A CN201910404623 A CN 201910404623A CN 110110807 A CN110110807 A CN 110110807A
- Authority
- CN
- China
- Prior art keywords
- leucocyte
- classification
- extraction
- algorithm
- neural networks
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 25
- 238000000605 extraction Methods 0.000 title claims abstract description 22
- 230000006872 improvement Effects 0.000 title description 2
- 210000004027 cell Anatomy 0.000 claims abstract description 20
- 230000011218 segmentation Effects 0.000 claims abstract description 17
- 230000001086 cytosolic effect Effects 0.000 claims abstract description 13
- 239000000284 extract Substances 0.000 claims abstract description 8
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 4
- 239000011159 matrix material Substances 0.000 claims description 12
- 230000006870 function Effects 0.000 claims description 11
- 238000005070 sampling Methods 0.000 claims description 5
- 102000002274 Matrix Metalloproteinases Human genes 0.000 claims description 3
- 108010000684 Matrix Metalloproteinases Proteins 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims description 3
- 238000004043 dyeing Methods 0.000 claims description 3
- 238000003064 k means clustering Methods 0.000 claims description 3
- 210000002569 neuron Anatomy 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 2
- 230000009467 reduction Effects 0.000 claims description 2
- 238000000926 separation method Methods 0.000 claims description 2
- 210000000265 leukocyte Anatomy 0.000 abstract description 7
- 210000000805 cytoplasm Anatomy 0.000 abstract description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000004040 coloring Methods 0.000 description 2
- 241000167880 Hirundinidae Species 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 238000009534 blood test Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 210000003743 erythrocyte Anatomy 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 210000000987 immune system Anatomy 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/60—Type of objects
- G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
- G06V20/695—Preprocessing, e.g. image segmentation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Probability & Statistics with Applications (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The present invention relates to a kind of based on the leucocyte extraction and classification method that improve K-means and convolutional neural networks.Firstly, selecting initial cluster center according to cell image intensity profile, initial clustering is carried out by nearby principle to image all pixels；Then, the Euclidean distance of FWSA-KM algorithm is improved；Before leucocyte extracts, color space decomposition is first carried out, nucleus and cytoplasmic extraction are carried out using the color component and improved K-means algorithm that are conducive to leucocyte segmentation；Then the part of complex overlapping is separated using watershed algorithm；Finally, being classified based on convolutional neural networks.The method of the present invention makes leukocyte cell core and cytoplasm segmentation precision be respectively 95.81% and 91.28%, improves a lot compared with conventional segmentation methods；Classification accuracy maximum can reach 98.96%, and classification average time is 0.39s, and relative to existing leukocyte differential count algorithm, CNN classification method not only has a clear superiority, while there are also very big rooms for promotion.
Description
Technical field
The present invention relates to medical image segmentation extractive technique fields, more particularly to one kind based on improvement K-means and volume
The leucocyte of product neural network extracts and classification method.
Background technique
Medically, leucocyte is the important component part of human immune system, is responsible for identifying and swallows abnormal cell.
The traditional classification of white blood corpuscle counts in blood routine examination and morphological analysis is dependent on the expert of artificial counting and blood test point
Analysis, low efficiency and have stronger subjectivity.Currently used is flow cytometer, can not realize Automated Classification of White Blood Cells, and
And there is limitation in clinical application.
In recent years, for preferably segmented image and leucocyte is identified, it is preferable that researcher proposes some effects in succession
Algorithm realize the Accurate Segmentation and sorting algorithm of leucocyte, but still there is also problems for leucocyte segmentation, these are asked
Topic is mainly derived from that image color brightness is different, and there are impurity in image, and leucocyte shape is varied, stained cells matter with
The color of red blood cell is close.Segmentation precision achieved by existing method can't reach clinical practice needs, therefore white
There are many more needs of work to carry out in cell segmentation field.
Summary of the invention
It extracts and divides based on the leucocyte for improving K-means and convolutional neural networks the purpose of the present invention is to provide a kind of
Class method, this method can efficiently extract leucocyte and segmentation precision is higher, finally using convolutional neural networks (CNN) into
Row leukocyte differential count and identification.
To achieve the above object, the technical scheme is that it is a kind of based on improving K-means and convolutional neural networks
Leucocyte extracts and classification method, includes the following steps:
Step S1, before extracting leucocyte, the decomposition of color space is first carried out, using the face for being conducive to leucocyte segmentation
Colouring component；
Step S2, K-means clustering algorithm is improved according to cell image intensity profile, is selected in initial clustering
The heart makes in image all pixels carry out initial clustering by nearby principle, and improves to the Euclidean distance of FWSA-KM algorithm,
Improve the robustness of clustering algorithm；
Step S3, nucleus and cytoplasmic extraction are carried out using improved K-means algorithm；
Step S4, the leucocyte fraction of complex overlapping is separated using watershed algorithm；
Step S5, it is tested using convolutional neural networks to isolated leucocyte is extracted, realizes the knowledge of adhesion leucocyte
Not.
In an embodiment of the present invention, the step S1 is implemented as follows:
Step S11, establish color model: the processing of dialogue cell dyeing makes it in the chrominance component space (H) and saturation component
(S) all there is stronger contrasts with background image for the corresponding cytosolic domain in space and leucocyte region；
Step S12, in subsequent singulation, threshold value in setting saturation component space and chrominance component space, by leucocyte
Nuclear area and leucocyte region are roughly extracted from cell image.
In an embodiment of the present invention, the step S2 is implemented as follows:
Step S21, histogram distribution statistics is carried out to cell image gray scale to select initial cluster center, so that image institute
There is pixel to carry out initial clustering by nearby principle；
Step S22, it is based on non-Euclidean distance, feature weight is improved, K-means is improved, Improved-KM is obtained
Clustering algorithm, i.e., by the χ in the algorithm objective functionikAnd vjkEuclidean distance | χik-vjk| it is revised as non-Euclidean distanceThus the objective function of improved Improved-KM clustering algorithm is obtained are as follows:
In an embodiment of the present invention, in the step S22, steps are as follows for Improved-KM clustering algorithm:
Step S221, objective function:
Wherein U=(uij)n×cIt is subordinated-degree matrix；If i-th of data point xiBelong to j-th of class, then uij=1, otherwise uij=0, and
AndAnd V=[v1,v2,…,vc] it is the matrix that c cluster centre is constituted；Formula simultaneously
Meet:
Step S222, optimal subordinated-degree matrixWith cluster centre matrixIn element are as follows:With
Step S223, pass through three minimization problems of iterative solution；
Step S224: it enablesWith
Then akIt has measured and has clustered compactness in class total on kth dimensional feature, bkIt has measured and has divided between clustering class total on kth dimensional feature
From property measurement；
Step S225, with new objective function
Solve following feature weight matrix:
Wherein meet:
Step S226, it setsThe feature weight of iteration is walked for t, then following formula is represented by t+1 step
Feature weight:Wherein it is as follows to adjust residual quantity for feature weight:
In order to makeMeet constraint conditionStandardization processing is carried out to feature weight formula, obtains spy
Levy weight
In an embodiment of the present invention, the step S3 is implemented as follows:
Step S31, it the extraction of leucocyte: is selected by the different components of the different color model of the different cell images of observation
It selects, then carries out cluster segmentation；
Step S32, the extraction of nucleus: contain a large amount of noise and the universal area of noise using the bianry image after cluster
Small feature calculates the area of each connected region, rejects the connected region area for being less than threshold value, is filled in nuclear area
Medium and small hole；
Step S33, the leucocyte extracted cytoplasmic extraction: is subtracted into the method for leucocyte core to obtain cytoplasm portion
Point, then restored color images again.
In an embodiment of the present invention, the step S4 is implemented as follows:
Step S41, for there is the leucocyte being sticked together, denoising and holes filling processing are carried out；
Step S42, the leucocyte being sticked together is split processing with fractional spins.
In an embodiment of the present invention, in the step S5, the convolutional neural networks are by input layer, convolutional layer, sampling
Layer, articulamentum and output layer are constituted；The image that input layer input needs to classify, extracts corresponding feature by convolutional layer, to add
Fast pace of learning needs neuron number by the reduction of sample level down-sampling while retaining useful information, and articulamentum passes through activation letter
For number by classification results input to output layer, the dimension of output layer is the classification number of required classification.
Compared to the prior art, the invention has the following advantages: the method for the present invention makes leukocyte cell core and thin
Cytoplasm segmentation precision is respectively 95.81% and 91.28%, is improved a lot compared with conventional segmentation methods；Classification accuracy maximum energy
Reach 98.96%, classification average time is 0.39s, and relative to existing leukocyte differential count algorithm, CNN classification method not only has
Clear superiority, while there are also very big rooms for promotion
Detailed description of the invention
Fig. 1 is the method for the present invention flow chart.
Fig. 2 is the structure figures of convolutional neural networks of the present invention.
Specific embodiment
With reference to the accompanying drawing, technical solution of the present invention is specifically described.
As shown in Figure 1, extracted based on the leucocyte for improving K-means and convolutional neural networks the present invention provides a kind of and
Classification method includes the following steps:
Step S1, before extracting leucocyte, the decomposition of color space is first carried out, using the face for being conducive to leucocyte segmentation
Colouring component；It is specific as follows:
Step S11, establish color model: the processing of dialogue cell dyeing makes it in the chrominance component space (H) and saturation component
(S) all there is stronger contrasts with background image for the corresponding cytosolic domain in space and leucocyte region；
Step S12, in subsequent singulation, threshold value in setting saturation component space and chrominance component space, by leucocyte
Nuclear area and leucocyte region are roughly extracted from cell image；
Step S2, K-means clustering algorithm is improved according to cell image intensity profile, is selected in initial clustering
The heart makes in image all pixels carry out initial clustering by nearby principle, and improves to the Euclidean distance of FWSA-KM algorithm,
Improve the robustness of clustering algorithm；It is specific as follows:
Step S21, histogram distribution statistics is carried out to cell image gray scale to select initial cluster center, so that image institute
There is pixel to carry out initial clustering by nearby principle；
Step S22, it is based on non-Euclidean distance, feature weight is improved, K-means is improved, Improved-KM is obtained
Clustering algorithm, i.e., by the χ in the algorithm objective functionikAnd vjkEuclidean distance | χik-vjk| it is revised as non-Euclidean distanceThus the objective function of improved Improved-KM clustering algorithm is obtained are as follows:
In the step S22, steps are as follows for Improved-KM clustering algorithm:
Step S221, objective function:
Wherein U=(uij)n×cIt is subordinated-degree matrix；If i-th of data point xiBelong to j-th of class, then uij=1, otherwise uij=0, and
AndAnd V=[v1,v2,…,vc] it is the matrix that c cluster centre is constituted；Formula simultaneously
Meet:
Step S222, optimal subordinated-degree matrixWith cluster centre matrixIn element are as follows:With
Step S223, pass through three minimization problems of iterative solution；
Step S224: it enablesWithThen akDegree
It has measured and has clustered compactness in class total on kth dimensional feature, bkSeparation property degree between clustering class total on kth dimensional feature is measured
Amount；
Step S225, with new objective function
Solve following feature weight matrix:
Wherein meet:
Step S226, it setsThe feature weight of iteration is walked for t, then following formula is represented by t+1 step
Feature weight:Wherein it is as follows to adjust residual quantity for feature weight:
In order to makeMeet constraint conditionStandardization processing is carried out to feature weight formula, obtains spy
Levy weight
Step S3, nucleus and cytoplasmic extraction are carried out using improved K-means algorithm；It is specific as follows:
Step S31, it the extraction of leucocyte: is selected by the different components of the different color model of the different cell images of observation
It selects, then carries out cluster segmentation；
Step S32, the extraction of nucleus: contain a large amount of noise and the universal area of noise using the bianry image after cluster
Small feature calculates the area of each connected region, rejects the connected region area for being less than threshold value, is filled in nuclear area
Medium and small hole；
Step S33, cytoplasmic extraction: because of the cytoplasmic color often color phase with leucocyte or background image
Seemingly, so extremely difficult segmentation；Therefore cytoplasmic compartment is obtained using the leucocyte extracted to be subtracted to the method for leucocyte core, then
Again restored color images
Step S4, the leucocyte fraction of complex overlapping is separated using watershed algorithm；It is specific as follows:
Step S41, for there is the leucocyte being sticked together, denoising and holes filling processing are carried out；
Step S42, the leucocyte being sticked together is split processing with fractional spins.
Step S5, it is tested using convolutional neural networks to isolated leucocyte is extracted, realizes the knowledge of adhesion leucocyte
Not；As shown in Fig. 2, the convolutional neural networks are made of input layer, convolutional layer, sample level, articulamentum and output layer；Input layer
Input needs the image classified, and extracts corresponding feature by convolutional layer, to accelerate pace of learning to subtract by sample level down-sampling
Neuron number is needed less while retaining useful information, and articulamentum passes through activation primitive (sigmoid function etc.) for classification results
Input to output layer, the dimension of output layer are the classification number of required classification.
The above are preferred embodiments of the present invention, all any changes made according to the technical solution of the present invention, and generated function is made
When with range without departing from technical solution of the present invention, all belong to the scope of protection of the present invention.
Claims (7)
1. a kind of based on the leucocyte extraction and classification method that improve K-means and convolutional neural networks, which is characterized in that including
Following steps:
Step S1, before extracting leucocyte, the decomposition of color space is first carried out, using the color point for being conducive to leucocyte segmentation
Amount；
Step S2, K-means clustering algorithm is improved according to cell image intensity profile, selectes initial cluster center, makes
All pixels carry out initial clustering by nearby principle in image, and improve to the Euclidean distance of FWSA-KM algorithm, make to cluster
The robustness of algorithm improves；
Step S3, nucleus and cytoplasmic extraction are carried out using improved K-means algorithm；
Step S4, the leucocyte fraction of complex overlapping is separated using watershed algorithm；
Step S5, it is tested using convolutional neural networks to isolated leucocyte is extracted, realizes the identification of adhesion leucocyte.
2. according to claim 1 a kind of based on the leucocyte extraction and classification that improve K-means and convolutional neural networks
Method, which is characterized in that the step S1 is implemented as follows:
Step S11, establish color model: the processing of dialogue cell dyeing makes it in the chrominance component space (H) and saturation component (S)
All there is stronger contrasts with background image for the corresponding cytosolic domain in space and leucocyte region；
Step S12, in subsequent singulation, threshold value in setting saturation component space and chrominance component space, by the cell of leucocyte
Core region and leucocyte region are roughly extracted from cell image.
3. according to claim 1 a kind of based on the leucocyte extraction and classification that improve K-means and convolutional neural networks
Method, which is characterized in that the step S2 is implemented as follows:
Step S21, histogram distribution statistics is carried out to cell image gray scale to select initial cluster center, so that all pictures of image
Element carries out initial clustering by nearby principle；
Step S22, it is based on non-Euclidean distance, feature weight is improved, K-means is improved, obtains Improved-KM cluster
Algorithm, i.e., by the χ in the algorithm objective functionikAnd vjkEuclidean distance | χik-vjk| it is revised as non-Euclidean distanceThus the objective function of improved Improved-KM clustering algorithm is obtained are as follows:
4. according to claim 1 a kind of based on the leucocyte extraction and classification that improve K-means and convolutional neural networks
Method, which is characterized in that in the step S22, steps are as follows for Improved-KM clustering algorithm:
Step S221, objective function:Its
Middle U=(uij)n×cIt is subordinated-degree matrix；If i-th of data point xiBelong to j-th of class, then uij=1, otherwise uij=0, andAnd V=[v1,v2,…,vc] it is the matrix that c cluster centre is constituted；Formula is full simultaneously
Foot:
Step S222, optimal subordinated-degree matrixWith cluster centre matrixIn element are as follows:With
Step S223, pass through three minimization problems of iterative solution；
Step S224: it enablesWithThen ak
It has measured and has clustered compactness in class total on kth dimensional feature, bkSeparation property between clustering class total on kth dimensional feature is measured
Measurement；
Step S225, with new objective function
Solve following feature weight matrix:
Wherein meet:
Step S226, it setsThe feature weight of iteration is walked for t, then following formula is represented by the spy of t+1 step
Levy weight:Wherein it is as follows to adjust residual quantity for feature weight:
In order to makeMeet constraint conditionStandardization processing is carried out to feature weight formula, obtains feature power
Weight
5. according to claim 1 a kind of based on the leucocyte extraction and classification that improve K-means and convolutional neural networks
Method, which is characterized in that the step S3 is implemented as follows:
Step S31, the extraction of leucocyte: being selected by the different components of the different color model of the different cell images of observation,
Cluster segmentation is carried out again；
Step S32, the extraction of nucleus: contain a large amount of noise using the bianry image after cluster and the universal area of noise is small
Feature calculates the area of each connected region, rejects the connected region area for being less than threshold value, is filled among nuclear area
Small hole；
Step S33, cytoplasmic extraction: subtracting the method for leucocyte core for the leucocyte extracted to obtain cytoplasmic compartment,
Restored color images again again.
6. according to claim 1 a kind of based on the leucocyte extraction and classification that improve K-means and convolutional neural networks
Method, which is characterized in that the step S4 is implemented as follows:
Step S41, for there is the leucocyte being sticked together, denoising and holes filling processing are carried out；
Step S42, the leucocyte being sticked together is split processing with fractional spins.
7. according to claim 1 a kind of based on the leucocyte extraction and classification that improve K-means and convolutional neural networks
Method, which is characterized in that in the step S5, the convolutional neural networks by input layer, convolutional layer, sample level, articulamentum and
Output layer is constituted；The image that input layer input needs to classify, extracts corresponding feature by convolutional layer, to accelerate pace of learning logical
The reduction of over-sampling layer down-sampling needs neuron number while retaining useful information, and articulamentum passes through activation primitive for classification results
Input to output layer, the dimension of output layer are the classification number of required classification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910404623.8A CN110110807B (en) | 2019-05-16 | 2019-05-16 | Leukocyte extraction and classification method based on improved K-means and convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910404623.8A CN110110807B (en) | 2019-05-16 | 2019-05-16 | Leukocyte extraction and classification method based on improved K-means and convolutional neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110110807A true CN110110807A (en) | 2019-08-09 |
| CN110110807B CN110110807B (en) | 2023-01-03 |
Family
ID=67490357
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910404623.8A Expired - Fee Related CN110110807B (en) | 2019-05-16 | 2019-05-16 | Leukocyte extraction and classification method based on improved K-means and convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110110807B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115019305A (en) * | 2022-08-08 | 2022-09-06 | 成都西交智汇大数据科技有限公司 | Method, device and equipment for identifying root tip cells and readable storage medium |
| CN115661667A (en) * | 2022-12-13 | 2023-01-31 | 济宁市土哥农业服务有限公司 | Method for identifying impurities of descurainia sophia seeds based on computer vision |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100035940A1 (en) * | 2005-10-25 | 2010-02-11 | University Of Flordia Research Foundation | Cyclin Dependent Kinase Inhibitors |
| CN103473739A (en) * | 2013-08-15 | 2013-12-25 | 华中科技大学 | White blood cell image accurate segmentation method and system based on support vector machine |
| CN104392460A (en) * | 2014-12-12 | 2015-03-04 | 山东大学 | Adherent white blood cell segmentation method based on nucleus-marked watershed transformation |
| CN104751462A (en) * | 2015-03-29 | 2015-07-01 | 嘉善加斯戴克医疗器械有限公司 | White cell segmentation method based on multi-feature nonlinear combination |
| CN109034045A (en) * | 2018-07-20 | 2018-12-18 | 中南大学 | A kind of leucocyte automatic identifying method based on convolutional neural networks |
-
2019
- 2019-05-16 CN CN201910404623.8A patent/CN110110807B/en not_active Expired - Fee Related
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100035940A1 (en) * | 2005-10-25 | 2010-02-11 | University Of Flordia Research Foundation | Cyclin Dependent Kinase Inhibitors |
| CN103473739A (en) * | 2013-08-15 | 2013-12-25 | 华中科技大学 | White blood cell image accurate segmentation method and system based on support vector machine |
| CN104392460A (en) * | 2014-12-12 | 2015-03-04 | 山东大学 | Adherent white blood cell segmentation method based on nucleus-marked watershed transformation |
| CN104751462A (en) * | 2015-03-29 | 2015-07-01 | 嘉善加斯戴克医疗器械有限公司 | White cell segmentation method based on multi-feature nonlinear combination |
| CN109034045A (en) * | 2018-07-20 | 2018-12-18 | 中南大学 | A kind of leucocyte automatic identifying method based on convolutional neural networks |
Non-Patent Citations (3)
| Title |
|---|
| LI-QUN LIN ET AL.: "Automatic Extraction of Fuzzy and Touching Leukocyte Using Improved FWSA K-means in Peripheral Blood and Bone Marrow Cell Images", 《JOURNAL OF COMPUTERS》 * |
| MAN YAN ET AL.: "K-means cluster algorithm based on color image enhancement for cell segmentation", 《2012 5TH INTERNATIONAL CONFERENCE ON BIOMEDICAL ENGINEERING AND INFORMATICS》 * |
| 林丽群 等: "改进的分数阶微分及图论的粘连血细胞图像分割", 《福州大学学报(自然科学版)》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115019305A (en) * | 2022-08-08 | 2022-09-06 | 成都西交智汇大数据科技有限公司 | Method, device and equipment for identifying root tip cells and readable storage medium |
| CN115019305B (en) * | 2022-08-08 | 2022-11-11 | 成都西交智汇大数据科技有限公司 | Method, device and equipment for identifying root tip cells and readable storage medium |
| CN115661667A (en) * | 2022-12-13 | 2023-01-31 | 济宁市土哥农业服务有限公司 | Method for identifying impurities of descurainia sophia seeds based on computer vision |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110110807B (en) | 2023-01-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112070772B (en) | Blood leukocyte image segmentation method based on UNet++ and ResNet | |
| US11900598B2 (en) | System and method of classification of biological particles | |
| Sinha et al. | Automation of differential blood count | |
| CN108376400B (en) | Automatic bone marrow cell classification method | |
| Mehanian et al. | Computer-automated malaria diagnosis and quantitation using convolutional neural networks | |
| CN103745210B (en) | A kind of leucocyte classification method and device | |
| Kothari et al. | Automatic batch-invariant color segmentation of histological cancer images | |
| Mathur et al. | Scalable system for classification of white blood cells from Leishman stained blood stain images | |
| Dhieb et al. | An automated blood cells counting and classification framework using mask R-CNN deep learning model | |
| Shkolyar et al. | Automatic detection of cell divisions (mitosis) in live-imaging microscopy images using convolutional neural networks | |
| CN108074243A (en) | A kind of cellular localization method and cell segmentation method | |
| CN105404887B (en) | A kind of five classification method of leucocyte based on random forest | |
| CN110120056A (en) | Blood leucocyte dividing method based on self-adapting histogram threshold value and contour detecting | |
| Pandit et al. | Survey on automatic rbc detection and counting | |
| Mohamed et al. | Automated classification of bacterial images extracted from digital microscope via bag of words model | |
| Meimban et al. | Blood cells counting using python opencv | |
| Gautam et al. | White blood nucleus extraction using K-Mean clustering and mathematical morphing | |
| CN110110807A (en) | A kind of leucocyte extraction and classification method based on improvement K-means and convolutional neural networks | |
| Revanda et al. | Classification of Acute Lymphoblastic Leukemia on White Blood Cell Microscopy Images Based on Instance Segmentation Using Mask R-CNN. | |
| CN104751461B (en) | A kind of leukocyte cell core dividing method based on histogram thresholding and low-rank representation | |
| CN109191467A (en) | A kind of prediction technique and device of cell autophagy phenotype | |
| CN104573668A (en) | Automatic face color adaptability identification method based on spectral reflectivity | |
| Jamil et al. | A rule-based segmentation method for fruit images under natural illumination | |
| Zhou et al. | Color space volume and superpixel based leukocyte image segmentation | |
| Chen et al. | Robust segmentation for low quality cell images from blood and bone marrow |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20230103 |