CN104850860A - Cell image recognition method and cell image recognition device - Google Patents
Cell image recognition method and cell image recognition device Download PDFInfo
- Publication number
- CN104850860A CN104850860A CN201510269977.8A CN201510269977A CN104850860A CN 104850860 A CN104850860 A CN 104850860A CN 201510269977 A CN201510269977 A CN 201510269977A CN 104850860 A CN104850860 A CN 104850860A
- Authority
- CN
- China
- Prior art keywords
- cell image
- cell
- image recognition
- matrix
- recognition
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2136—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on sparsity criteria, e.g. with an overcomplete basis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/04—Recognition of patterns in DNA microarrays
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention provides a cell image recognition method and a cell image recognition device. In the cell image recognition method, cells are divided into to-be-tested cells and various kinds of training cells. The cell image recognition process comprises the following steps: S1, preprocessing a cell image, specifically including cell image gray processing, image denoising processing, image gray histogram equalization processing, and image segmentation processing; S2, using a compressed sensing technology to establish sparse coefficient representation on the preprocessed cell image; S3, recognizing cells, calculating difference value of linear weighting of the to-be-tested cells and each kind of training sample cells ri(y), and selecting the classification the training sample with the lowest difference value belongs to as the classification of a test sample. Through using the compressed sensing technology in cell recognition, operand is reduced, and the method and the device has self-adaptive capability, so influence of external factors is reduced, and cell image recognition precision and speed are improved.
Description
Technical field
The invention belongs to medical cell image processing field, be specifically related to a kind of cell image recognition method and cell image recognition device.
Background technology
Cell image process is an important branch of Medical Imaging, clinical cytology diagnosis is by sampling human lesion position, and make cell smear, doctor examines under a microscope form, the Texture eigenvalue of cell, and gain knowledge in conjunction with cell pathology, early stage diagnosis and early warning are made to disease.Current cyto-diagnosis depends on the artificial diagosis of doctor, therefore, the accuracy of diagnostic result and reliability are subject to the subjective factor impact of the clinical experience of doctor, professional ability and doctor, the methodological standardization popularization degree of artificial diagosis is low, identification judging efficiency is low, and when the workload of diagosis increases, the error of diagnosis also can increase thereupon.In order to improve the diagnostic techniques of clinical cytology, need Import computer automatic aided diagnosis technique, realize the intellectuality of cell recognition technology, real-time, effectively avoid the impact of manually-operated subjective factor, the diagnosis for pre-cancerous provides convenient and swift, time saving and energy saving robust techniques means.
Current cell recognition technology is roughly divided into several step: first feature extraction: the cell image feature of abstraction reaction cell essential characteristic from unicellular image; Then model is set up: gained knowledge by specific cells field cell pathology, adopts and specifies sorter to the training of designated cell sample, set up cell recognition disaggregated model; Carry out pattern-recognition again: by the cell tests sample collected, by the recognition classifier set up, generic is made to test sample book and judges.Such method makes the data calculated amount when model is set up large, and the data compared when pattern-recognition are large, and the interference by extraneous factor is many, and cannot carry out reaching adaptive effect, therefore produces adverse influence to recognition result large.
Summary of the invention
It is large and cannot adaptive problem that the present invention is intended to solve in prior art the data calculated amount existed.For this reason, the invention provides a kind of cell image recognition method and cell image recognition device, by adopting the application in cell recognition of compressed sensing principle art, decrease operand, there is adaptive ability simultaneously, reduce the impact of extraneous factor, improve identification accuracy and the speed of cell image, can intelligent, identify cell image classification fast and accurately, consistance is good, and standardized and popularized degree is high.
A kind of cell image recognition method provided according to a first aspect of the present invention, wherein said cell is divided into cell to be tested and multiclass training cell, cell image recognition process comprises the steps: S1: to described cell image pre-service, specifically comprises the process of cell image gray processing, image denoising sonication, image greyscale histogram equalizing process, Iamge Segmentation processing procedure; S2: adopt compressed sensing technology to set up sparse coefficient to pretreated described cell image and represent, this step comprises following process:
S21: first by row, formation cell image sample column vector is extracted to cell image;
S22: the perception matrix A ∈ R that the training sample cell image in database is formed
m × nas the calculation matrix of compressed sensing, wherein m is sample characteristics dimension, and n is sample size;
S23: according to compressive sensing theory, to the cell image sample y ∈ R that Real-time Collection arrives
m, by solving optimum l
1norm constructs sparse coefficient x, and the mode of solving is
s.t.Ax=y;
S3: cell recognition, calculates the difference r of the linear weighted function of cell to be identified and each class training sample cell
i(y), select the classification of generic as test sample book of the minimum class training sample of difference, specific formula for calculation is;
In formula,
represent the coefficient that the cell sparse coefficient x to be identified of extraction is corresponding with the i-th class cell image.
The present invention, by overcoming the impact of extraneous factor on cell image quality to the pre-service of cell image, improves identification accuracy and the speed of cell image, by solving optimum l by employing compressed sensing technology
1norm constructs the computation complexity that sparse coefficient x can effectively reduce cell recognition algorithm, not only increase computing velocity and can also realize self-adaptation, also reduce extraneous factor to the impact of cell image recognition, this method integrating cell image acquisition, process and decision-making integration, judgment standard is consistent, standardization level is high, may be used for artificial diagosis identification and combines, and reduces diagnostic error rate further.
Further, optimum l is solved
1the employing orthogonal matching pursuit algorithm (OMP) of norm realizes, and comprises following detailed process
S231: definition length is null vector S and the residual vector r of N
1, initial residual vector r
1equal measured value vector y;
S232: with each column element in perception matrix A and residual vector r
1do inner product operation, find out maximum inner product absolute value and the position number λ of respective column thereof
1, find out its column element corresponding in perception matrix A simultaneously and be designated as A
max1, this train value in perception matrix A is set to 0, then utilizes matrix V
1preserve this column element value V
1=[A
max1];
S233: definition matrix W
1=(V
1 hv
1)
-1v
1 hy;
S234: calculate residual vector r
2, i.e. residual vector r
2=y-V
1w
1.
S235: with the new residual vector r obtained
2do inner product with the column element in perception matrix A, be designated as A by the row that calculate in perception matrix A corresponding to maximum inner product
max2, upgrade matrix V
1, obtain matrix V
2, i.e. V
2=[V
1, A
max2];
S236: repetitive process S233, S234, S235, after completing m iteration, finally obtain matrix W
m;
S237: by matrix W
min element according to the sequence number λ of respective column
m, put into vector x, namely obtain the sparse coefficient x of cell image feature
.
Compressed sensing (compressive Sensing, CS) theory is proposed by people such as D Donoho, E Candes and scientist T Tao of Chinese origin for 2006, its core concept utilizes transformation space to describe signal, by directly collecting the Systems with Linear Observation data that minority " is chosen carefully ", being the sampling of information by the sample transition of signal, from the data of compression observation, recovering original signal by solving an optimization problem.Under this theory, the sampling desired data amount of signal is far fewer than the required data volume of traditional sampling method, high-resolution sampling rate is made no longer to depend on Shannon's sampling theorem like this, make to recover high-resolution signal from low resolution observation and become possibility, therefore the present invention can adopt compressive sensing theory to build a kind of compressed sensing sorter, by the rarefaction representation to cell image, the identification of cell is realized from sparse coefficient, both decrease computational complexity, improved arithmetic speed, and improve the accuracy that cell image distinguishes.
The present invention, also provide a kind of cell image recognition device on the other hand, described cell image recognition device comprises cell image input receiver module, cell image processing module, cell image recognition module, wherein said cell image input receiver module receives the input of cell image and exports described cell image processing module to, described cell image processing module adopts as the step S1 in claim 1, S2 processes to cell image the sparse coefficient x obtaining cytological map, obtained cell image sparse coefficient x is also delivered to described cell image recognition module by described cell image processing module and described cell image recognition model calling, described cell image recognition module carries out classification identification according to the described cervical cell image of the step S3 in claim 1 to input.
Further, cell image recognition device also comprises with cell image recognition model calling for pointing out the reminding module that predicts the outcome of cell image Forecasting recognition result, described in the reminding module that predicts the outcome comprise voice or/and picture cues.
This cell image recognition device improves degree of accuracy and the recognition speed of cell image recognition, intellectualized operation and judgement can be realized, the class state of cell can be identified fast, can help doctor rapidly diagnosis cell whether be in health status, prevention of disease and treatment have good using value.
Accompanying drawing explanation
Above-mentioned and/or additional aspect of the present invention and advantage will become obvious and easy understand from accompanying drawing below combining to the description of embodiment, wherein:
Fig. 1 is cell recognition method principle schematic of the present invention;
Fig. 2 is the Image semantic classification principle schematic in cell recognition method of the present invention;
Fig. 3 is the compressed sensing sorter schematic diagram in cell recognition method of the present invention;
Fig. 4 is OMP algorithm flow schematic diagram in cell recognition method of the present invention;
Fig. 5 is that the present invention tests six kinds of HEp-2 cell images used;
Fig. 6 is HEp-2 cell recognition residual plot of the present invention;
Fig. 7 is that cell recognition method of the present invention is to HEp-2 cell test result figure;
Fig. 8 is that cell recognition device of the present invention forms schematic diagram.
Embodiment
In order to more clearly understand above-mentioned purpose of the present invention, feature and advantage, below in conjunction with the drawings and specific embodiments, the present invention is further described in detail.It should be noted that, when not conflicting, the feature in the embodiment of the application and embodiment can combine mutually.
Set forth a lot of detail in the following description so that fully understand the present invention; but; the present invention can also adopt other to be different from mode described here to implement, and therefore, protection scope of the present invention is not by the restriction of following public specific embodiment.
The cell image recognition method proposed to the first aspect of the present invention referring to Fig. 1-7 is further described.
Shown in Figure 1, cell image recognition Method And Principle schematic diagram of the present invention, wherein said cell image is divided into cell image to be tested and multiclass training cell image, these cell images are sent to compressed sensing sorter identifying processing after pre-service, and recognition result is exported, see Fig. 2 and Fig. 3, cell image recognition process comprises the steps:
S1: to cell image pre-service, specifically comprises the process of cell image gray processing, image denoising sonication, image greyscale histogram equalizing process, Iamge Segmentation processing procedure;
S2: adopt compressed sensing technology to set up sparse coefficient to pretreated described cell image and represent, this step comprises following process:
S21: first by row, formation cell image sample column vector y is extracted to test cell image;
S22: the perception matrix A ∈ R that the training sample cell image in database is formed
m × nas the calculation matrix of compressed sensing after rectangular array normalized, wherein m is sample characteristics dimension, and n is sample size;
S23: according to compressive sensing theory, to the test cell image pattern y ∈ R that Real-time Collection arrives
m, by solving optimum l
1norm constructs sparse coefficient x, and the mode of solving is
s.t.Ax=y;
S3: cell recognition, calculates the residual values r of the linear weighted function of test cell and each class training sample cell
i(y), select the classification of generic as test sample book of the minimum class training sample of residual values, specific formula for calculation is;
In formula,
represent the coefficient that the cell sparse coefficient x to be identified of extraction is corresponding with the i-th class cell image.
Solve optimum l
1the employing orthogonal matching pursuit algorithm (OMP) of norm realizes, and comprises following detailed process see Fig. 4, OMP
S231: definition length is null vector S and the residual vector r of N
1, iterations k=1, initial residual vector r
1equal measured value vector y;
S232: with each column element in perception matrix A and residual vector r
1do inner product operation, find out maximum inner product absolute value and the position number λ of respective column thereof
1, find out its column element corresponding in perception matrix A simultaneously and be designated as A
max1, this train value in perception matrix A is set to 0, then utilizes matrix V
1preserve this column element value V
1=[A
max1];
S233: definition matrix W
1=(V
1 hv
1)
-1v
1 hy;
S234: calculate residual vector r
2, i.e. residual vector r
2=y-V
1w
1;
S235: with the new residual vector r obtained
2do inner product with the column element in perception matrix A, be designated as A by the row that calculate in perception matrix A corresponding to maximum inner product
max2, upgrade matrix V
1, obtain matrix V
2, i.e. V
2=[V
1, A
max2];
S236: repetitive process S233, S234, S235, after completing m iteration, finally obtain matrix W
m;
S237: by matrix W
min element according to the sequence number λ of respective column
m, put into vector x, namely obtain the sparse coefficient x of cell image feature.
The embodiment of the present invention adopts HEp-2 cell data set to test, HEp-2 cell data set derives from International Model identification conference (International Conference On Pattern Recognition, ICPR) the HEp-2 cell data set (http://mivia.unisa.it/hep2contest/index.shtml) that " contest of the HEp-2 cell classification " official held for 2012 provides, HEp-2 cell image is obtained by fluorescent microscope, this microscope magnification is 40 times, additional 50W mercury vapor lamp and digital camera, the resolution of digital camera is 6.45 μm, the HEp-2 cell image size that this microscope obtains is 1388 × 1038, and color depth is 24, data centralization contains the indirect immunofluorescence image of 28 HEp-2 cells, and is partitioned into 1455 unicellular images of HEp-2 (721 training sample image, 734 test sample image), and concrete all kinds of HEp-2 cell quantity is as shown in table 1.What data centralization HEp-2 cell image all obtained immunology expert checks confirmation, and this data centralization comprises 6 kinds of HEp-2 cells, is respectively:
1. homogeneous pattern---cell caryoplasm even dyeing is consistent, and this type is relevant with histonic antibody and anti-DNA antibody;
2. spotted type---cell feulgen's stain is mottled, there is nuclear membrane, and this type is how relevant with solubility nuclear antigen (ENA) antibody;
3. nucleolar pattern---only entoblast has fluorescent effect, relevant to 4-6SRNA antibody, and this type is more common in chorionitis;
4. nuclear membrane type---fluorescence is centered around around nuclear membrane, and this type is how relevant with anti-dsDNA antibody;
5. kinetochore type---this type cell presents the toroidal of kinetochore hash particular point composition, main relevant with Raynaud's phenomenon.
Spotted type is divided into again mat patch point-type and thin spotted type two class by HEp-2 cell recognition contest official.
HEp-2 cell six type image as shown in Figure 5.Fig. 6 is the residual plot of HEp-2 cell recognition, the left figure of Fig. 6 represent the 4th class HEp-2 cell and kinetochore type HEp-2 cell residual error minimum, the HEp-2 cell type that can obtain thus testing is kinetochore type, the right figure of Fig. 6 represent the 1st class HEp-2 cell and homogeneous pattern HEp-2 cell residual error minimum, can obtain thus test HEp-2 cell type be homogeneous pattern.Show through test, the cell recognition rate adopting the present invention to obtain is 69.82%, discrimination optimum compared with other unartificial recognition methodss existing, also closest to expert to artificial cognition rate 74%, the HEp-2 cell recognition rate comparison diagram of cell see such as Fig. 7.
The present invention, by overcoming the impact of extraneous factor on cell image quality to the pre-service of cell image, improves identification accuracy and the speed of cell image, by solving optimum l by employing compressed sensing technology
1norm constructs the computation complexity that sparse coefficient x can effectively reduce cell recognition algorithm, not only increase computing velocity and can also realize self-adaptation, also reduce extraneous factor to the impact of cell image recognition, this method integrating cell image acquisition, process and decision-making integration, judgment standard is consistent, standardization level is high, may be used for artificial diagosis identification and combines, and reduces diagnostic error rate further.
The present invention also provides a kind of cell image recognition device on the other hand, shown in Figure 8, cell image recognition device comprises cell image input receiver module 10, cell image processing module 20, cell image recognition module 30, wherein said cell image input receiver module 10 receives the input of cell image and exports described cell image processing module 20 to, described cell image processing module 20 adopts the step S1 in foregoing cell image recognition method, S2 processes to cell image the sparse coefficient x obtaining cytological map, cell image processing module 20 is connected with described cell image recognition module 30 and obtained cell image sparse coefficient x is delivered to described cell image recognition module 30, cell image recognition module 30 carries out classification identification according to the described cervical cell image of the step S3 in foregoing cell image recognition method to input.
In addition, see Fig. 8, cervical cell pattern recognition device also comprises the reminding module 40 that predicts the outcome be connected with cell image recognition module 30 for pointing out cell image Forecasting recognition result, described in the reminding module 40 that predicts the outcome comprise voice or/and picture cues.
Above embodiment is only the preferred embodiments of the present invention, is not limited to the present invention, and for a person skilled in the art, the present invention can have various modifications and variations.All within creative spirit of the present invention and principle, any amendment done, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.
Table 1
| HEp-2 cell type | Training sample quantity | Test sample book quantity |
| Homogeneous pattern | 150 | 180 |
| Mat patch point-type | 109 | 101 |
| Nucleolar pattern | 102 | 139 |
| Kinetochore type | 208 | 149 |
| Thin spotted type | 94 | 114 |
| Nuclear membrane type | 58 | 51 |
Claims (4)
1. a cell image recognition method, wherein said cell is divided into cell to be tested and multiclass training cell, and cell image recognition process comprises the steps:
S1: to described cell image pre-service, specifically comprises the process of cell image gray processing, image denoising sonication, image greyscale histogram equalizing process, Iamge Segmentation processing procedure;
S2: adopt compressed sensing technology to set up sparse coefficient to pretreated described cell image and represent, this step comprises following process:
S21: first by row, formation cell image sample column vector is extracted to cell image;
S22: the perception matrix A ∈ R that the training sample cell image in database is formed
m × nas the calculation matrix of compressed sensing, wherein m is sample characteristics dimension, and n is sample size;
S23: according to compressive sensing theory, to the cell image sample y ∈ R that Real-time Collection arrives
m, by solving optimum l
1norm constructs sparse coefficient x, and the mode of solving is
S3: cell recognition, calculates the difference r of the linear weighted function of cell to be identified and each class training sample cell
i(y), select the classification of generic as test sample book of the minimum class training sample of difference, specific formula for calculation is;
In formula,
represent the coefficient that the cell sparse coefficient x to be identified of extraction is corresponding with the i-th class cell image.
2. cell image recognition method as claimed in claim 1, is characterized in that in step S23, solves optimum l
1the employing orthogonal matching pursuit algorithm (OMP) of norm realizes, and comprises following detailed process:
S231: definition length is null vector S and the residual vector r of N
1, initial residual vector r
1equal measured value vector y;
S232: with each column element in perception matrix A and residual vector r
1do inner product operation, find out maximum inner product absolute value and the position number λ of respective column thereof
1, find out its column element corresponding in perception matrix A simultaneously and be designated as A
max1, this train value in perception matrix A is set to 0, then utilizes matrix V
1preserve this column element value V
1=[A
max1];
S233: definition matrix W
1=(V
1 hv
1)
-1v
1 hy;
S234: calculate residual vector r
2, i.e. residual vector r
2=y-V
1w
1.
S235: with the new residual vector r obtained
2do inner product with the column element in perception matrix A, be designated as A by the row that calculate in perception matrix A corresponding to maximum inner product
max2, upgrade matrix V
1, obtain matrix V
2, i.e. V
2=[V
1, A
max2];
S236: repetitive process S233, S234, S235, after completing m iteration, finally obtain matrix W
m;
S237: by matrix W
min element according to the sequence number λ of respective column
m, put into vector x, namely obtain the sparse coefficient x of cell image feature.
3. a cell image recognition device, comprise cell image input receiver module, cell image processing module, cell image recognition module, wherein said cell image input receiver module receives the input of cell image and exports described cell image processing module to, described cell image processing module adopts as the step S1 in claim 1, S2 processes to cell image the sparse coefficient x obtaining cytological map, obtained cell image sparse coefficient x is also delivered to described cell image recognition module by described cell image processing module and described cell image recognition model calling, described cell image recognition module carries out classification identification according to the described cervical cell image of the step S3 in claim 1 to input.
4. cell image recognition device as claimed in claim 3, characterized by further comprising with cell image recognition model calling for pointing out the reminding module that predicts the outcome of cell image Forecasting recognition result, described in the reminding module that predicts the outcome comprise voice or/and picture cues.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510269977.8A CN104850860A (en) | 2015-05-25 | 2015-05-25 | Cell image recognition method and cell image recognition device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510269977.8A CN104850860A (en) | 2015-05-25 | 2015-05-25 | Cell image recognition method and cell image recognition device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104850860A true CN104850860A (en) | 2015-08-19 |
Family
ID=53850495
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510269977.8A Pending CN104850860A (en) | 2015-05-25 | 2015-05-25 | Cell image recognition method and cell image recognition device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104850860A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106407777A (en) * | 2016-08-30 | 2017-02-15 | 孟玲 | Intelligent mobile terminal based on biological identification |
| CN106600577A (en) * | 2016-11-10 | 2017-04-26 | 华南理工大学 | Cell counting method based on depth deconvolution neural network |
| CN106991673A (en) * | 2017-05-18 | 2017-07-28 | 深思考人工智能机器人科技（北京）有限公司 | A kind of cervical cell image rapid classification recognition methods of interpretation and system |
| CN108776823A (en) * | 2018-07-06 | 2018-11-09 | 武汉兰丁医学高科技有限公司 | Cervical carcinoma lesion analysis method based on cell image recognition |
| CN109191434A (en) * | 2018-08-13 | 2019-01-11 | 阜阳师范学院 | Image detecting system and detection method in a kind of cell differentiation |
| CN109670472A (en) * | 2018-12-28 | 2019-04-23 | 广东省心血管病研究所 | The image processing system and method for umbilical cord mesenchymal stem cells in vitro culture and amplification |
| WO2020024227A1 (en) * | 2018-08-02 | 2020-02-06 | 深圳迈瑞生物医疗电子股份有限公司 | Cell analysis method, cell analysis device, and storage medium |
| CN110765855A (en) * | 2019-09-12 | 2020-02-07 | 杭州迪英加科技有限公司 | Pathological image processing method and system |
| CN110796668A (en) * | 2019-10-28 | 2020-02-14 | 闽江学院 | White cell nucleus segmentation method based on sparse limitation |
| CN111261187A (en) * | 2020-02-04 | 2020-06-09 | 清华珠三角研究院 | Method, system, device and storage medium for converting voice into lip shape |
| CN112183210A (en) * | 2020-08-28 | 2021-01-05 | 江苏康尚生物医疗科技有限公司 | Sparse matrix-based particle automatic focusing method and system and storage medium |
| WO2022042509A1 (en) * | 2020-08-26 | 2022-03-03 | 深圳太力生物技术有限责任公司 | Cell screening method and apparatus based on expression level prediction model |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103295198A (en) * | 2013-05-13 | 2013-09-11 | 西安电子科技大学 | Non-convex compressed sensing image reconstruction method based on redundant dictionary and structure sparsity |
| CN104268841A (en) * | 2014-10-09 | 2015-01-07 | 哈尔滨工程大学 | Infrared image preprocessing method based on compressed sensing algorithm |
-
2015
- 2015-05-25 CN CN201510269977.8A patent/CN104850860A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103295198A (en) * | 2013-05-13 | 2013-09-11 | 西安电子科技大学 | Non-convex compressed sensing image reconstruction method based on redundant dictionary and structure sparsity |
| CN104268841A (en) * | 2014-10-09 | 2015-01-07 | 哈尔滨工程大学 | Infrared image preprocessing method based on compressed sensing algorithm |
Non-Patent Citations (6)
| Title |
|---|
| NAZIM BURAK KARAHANO˘GLU等: "A* Orthogonal Matching Pursuit: Best-First Search for Compressed Sensing Signal Recovery", 《PREPRINT SUBMITTED TO DIGITAL SIGNAL PROCESSING》 * |
| YONG FANG等: "GPU Implementation of Orthogonal Matching Pursuit for Compressive Sensing", 《2011 IEEE 17TH INTERNATIONAL CONFERENCE ON PARALLEL AND DISTRIBUTED SYSTEMS》 * |
| 王娟: "量子免疫克隆算法研究及在压缩感知重构中的应用", 《中国优秀硕士学位论文全文数据库 基础科学辑》 * |
| 谢丽娟: "基于压缩感知和神经网络的肿瘤细胞图像识别研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 邹伟等: "基于压缩感知的人脸识别方法", 《计算机工程》 * |
| 魏冬梅等: "采用压缩感知的人脸识别算法", 《计算机工程》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106407777A (en) * | 2016-08-30 | 2017-02-15 | 孟玲 | Intelligent mobile terminal based on biological identification |
| CN106600577A (en) * | 2016-11-10 | 2017-04-26 | 华南理工大学 | Cell counting method based on depth deconvolution neural network |
| CN106600577B (en) * | 2016-11-10 | 2019-10-18 | 华南理工大学 | A kind of method for cell count based on depth deconvolution neural network |
| CN106991673B (en) * | 2017-05-18 | 2019-10-22 | 深思考人工智能机器人科技（北京）有限公司 | A kind of the cervical cell image rapid classification recognition methods and system of interpretation |
| CN106991673A (en) * | 2017-05-18 | 2017-07-28 | 深思考人工智能机器人科技（北京）有限公司 | A kind of cervical cell image rapid classification recognition methods of interpretation and system |
| CN108776823A (en) * | 2018-07-06 | 2018-11-09 | 武汉兰丁医学高科技有限公司 | Cervical carcinoma lesion analysis method based on cell image recognition |
| WO2020024227A1 (en) * | 2018-08-02 | 2020-02-06 | 深圳迈瑞生物医疗电子股份有限公司 | Cell analysis method, cell analysis device, and storage medium |
| CN109191434A (en) * | 2018-08-13 | 2019-01-11 | 阜阳师范学院 | Image detecting system and detection method in a kind of cell differentiation |
| CN109670472A (en) * | 2018-12-28 | 2019-04-23 | 广东省心血管病研究所 | The image processing system and method for umbilical cord mesenchymal stem cells in vitro culture and amplification |
| CN110765855A (en) * | 2019-09-12 | 2020-02-07 | 杭州迪英加科技有限公司 | Pathological image processing method and system |
| CN110796668A (en) * | 2019-10-28 | 2020-02-14 | 闽江学院 | White cell nucleus segmentation method based on sparse limitation |
| CN110796668B (en) * | 2019-10-28 | 2022-04-01 | 闽江学院 | White cell nucleus segmentation method based on sparse limitation |
| CN111261187A (en) * | 2020-02-04 | 2020-06-09 | 清华珠三角研究院 | Method, system, device and storage medium for converting voice into lip shape |
| CN111261187B (en) * | 2020-02-04 | 2023-02-14 | 清华珠三角研究院 | Method, system, device and storage medium for converting voice into lip shape |
| WO2022042509A1 (en) * | 2020-08-26 | 2022-03-03 | 深圳太力生物技术有限责任公司 | Cell screening method and apparatus based on expression level prediction model |
| CN112183210A (en) * | 2020-08-28 | 2021-01-05 | 江苏康尚生物医疗科技有限公司 | Sparse matrix-based particle automatic focusing method and system and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104850860A (en) | Cell image recognition method and cell image recognition device | |
| CN108182441B (en) | Parallel multichannel convolutional neural network, construction method and image feature extraction method | |
| CN110443143B (en) | Multi-branch convolutional neural network fused remote sensing image scene classification method | |
| CN110598029B (en) | Fine-grained image classification method based on attention transfer mechanism | |
| CN107316013A (en) | Hyperspectral image classification method with DCNN is converted based on NSCT | |
| CN110728224A (en) | Remote sensing image classification method based on attention mechanism depth Contourlet network | |
| CN110045015B (en) | Concrete structure internal defect detection method based on deep learning | |
| CN111582225B (en) | Remote sensing image scene classification method and device | |
| CN107451565B (en) | Semi-supervised small sample deep learning image mode classification and identification method | |
| CN105760821A (en) | Classification and aggregation sparse representation face identification method based on nuclear space | |
| CN110619352A (en) | Typical infrared target classification method based on deep convolutional neural network | |
| CN105654141A (en) | Isomap and SVM algorithm-based overlooked herded pig individual recognition method | |
| CN109635726B (en) | Landslide identification method based on combination of symmetric deep network and multi-scale pooling | |
| CN110543916B (en) | Method and system for classifying missing multi-view data | |
| CN113052185A (en) | Small sample target detection method based on fast R-CNN | |
| CN104699781B (en) | SAR image search method based on double-deck anchor figure hash | |
| CN103268494B (en) | Parasite egg recognition methods based on rarefaction representation | |
| CN114299567B (en) | Model training method, living body detection method, electronic device, and storage medium | |
| CN111680575A (en) | Human epithelial cell staining and classifying device, equipment and storage medium | |
| CN116310466A (en) | Small sample image classification method based on local irrelevant area screening graph neural network | |
| CN115131503A (en) | Health monitoring method and system for iris three-dimensional recognition | |
| CN101630369B (en) | Pedestrian detection method based on wavelet fractal characteristic | |
| CN109558803A (en) | SAR target discrimination method based on convolutional neural networks Yu NP criterion | |
| CN113344046A (en) | Method for improving SAR image ship classification precision | |
| CN109241315A (en) | A kind of fast face search method based on deep learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150819 |
|
| RJ01 | Rejection of invention patent application after publication |