CN106815849A - A kind of method for recognizing biopsy tissues - Google Patents
A kind of method for recognizing biopsy tissues Download PDFInfo
- Publication number
- CN106815849A CN106815849A CN201710040116.1A CN201710040116A CN106815849A CN 106815849 A CN106815849 A CN 106815849A CN 201710040116 A CN201710040116 A CN 201710040116A CN 106815849 A CN106815849 A CN 106815849A
- Authority
- CN
- China
- Prior art keywords
- connected domain
- biopsy tissues
- slice information
- area
- spot
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Geometry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The present invention relates to pathological section field, more particularly to a kind of method for recognizing biopsy tissues.The section of one blank and histotomy to be identified are provided, including：Scan the Background of blank section and the preview graph of histotomy to be identified；The gray value of preview graph is subtracted the gray value of Background, a section hum pattern is obtained；The connected domain of traversal processing slice information figure, rejects the spot in slice information figure；Attribute according to connected domain judges whether each connected domain in slice information figure is biopsy tissues respectively, and reject be judged as be not biopsy tissues connected domain；Retain the slice information figure after rejecting the connected domain for not being biopsy tissues and exported as biopsy tissues picture.Differentiate in the attributive character such as color shape by tissue and impurity, tissue part is effectively identified as far as possible, while most impurity and spot are identified and are rejected, so as to more accurately isolate the tissue part in section preview graph.
Description
Technical field
The present invention relates to pathological section field, more particularly to a kind of method for recognizing biopsy tissues.
Background technology
In recent years, the development with computer and network technology and constantly improve, digital pathology system are arisen at the historic moment.With biography
System pathological section is compared, digital pathological section section browse, transmit, preserve and impart knowledge to students and hold a consultation etc. many-side all there occurs it is huge
Great change so that digital pathology system turns into a kind of indispensable important tool in terms of medical pathological diagnosis from now on.
It is accurate to identify tissue, it is necessary to the preview graph to cutting into slices carries out preliminary treatment before scanning of cutting into slices, and carry
Relevant information is taken to scan.Traditional recognition methods is to use image processing techniques, by preview graph gray processing, further according to brightness
Difference distinguishes the dyeing stain of tissue, background, impurity or the section of section, so as to recognize and extract tissue part.
Conventional method is identified generally be directed to a certain or several specific biopsy tissues classifications, is not suitable for mostly
Number section, and during using traditional brightness recognition methods, often occur omitting tissue or miss impurity and spot
Be identified as the situation of tissue, cause be difficult accurately and effectively recognized, so as to reduce identification biopsy tissues accuracy rate and
Efficiency.
The content of the invention
For that can not recognize most of biopsy tissues and the accurate low situation of identification at present, the present invention provides a kind of identification and cuts
The method of piece tissue.
The technical proposal for solving the technical problem of the invention is：
A kind of method for recognizing biopsy tissues a, there is provided blank is cut into slices and histotomy to be identified, methods described bag
Include：
Step 1, scans the Background of the blank section and the preview graph of histotomy to be identified；
Step 2, the gray value of the preview graph is subtracted the gray value of the Background, obtains a section hum pattern；
Step 3, the connected domain of slice information figure described in traversal processing rejects the spot in the slice information figure；
Step 4, the attribute according to the connected domain judge respectively each connected domain in the slice information figure whether be
Biopsy tissues, and reject be judged as be not the biopsy tissues the connected domain；
Step 5, retains the slice information figure after rejecting the connected domain for not being the biopsy tissues and conduct is cut
Piece tissue picture is exported.
Preferably, the step 4 includes：
By finding out the maximum connection of area in the remaining connected domain after the spot in rejecting the slice information figure
Domain, and the area of other remaining connected domains is calculated one compared with the area of the maximum connected domain of the area respectively
Area difference, biopsy tissues are judged as by connected domain of the area difference in the range of a default area difference.
Preferably, the step 4 includes：
Calculate equal by the color of each passages of RGB of the remaining connected domain after the spot in rejecting the slice information figure
It is worth, and the color value of each passages of RGB respectively according to remaining each connected domain is obtained with the color average value processing of each passages of RGB
Corresponding RGB differences, the RGB differences of the same connected domain is corresponding in a default RGB difference ranges
The connected domain be judged as biopsy tissues.
Preferably, in the step 4, the side edge with the slice information figure in the slice information figure is touched
Connected domain be judged as be not the biopsy tissues the connected domain.
Preferably, in the step 4, by by the remaining connected domain after the spot in rejecting the slice information figure
Connected domain of the brightness value in a default brightness range is judged as YES the connected domain of the biopsy tissues.
Preferably, in the step 3, the method for rejecting the spot in the slice information figure includes：
Reject connected domain of the area less than a default area；And/or
Reject connected domain of the Aspect Ratio less than a default length and width coefficient；And/or
Reject connected domain of the area accounting coefficient less than a default area accounting coefficient.
Preferably, before the step 2 is performed, following step is first carried out：
Step 1A, wipes out the marginal portion of the Background of the blank section and the preview graph of the histotomy, with
Step 2 described in rear steering.
Preferably, before the step 3 is performed, following step is first carried out：
Step 2A, image binaryzation operation is carried out to the slice information figure, is subsequently diverted to the step 3.
Preferably, before the step 3 is performed, following step is first carried out：
Step 2A, image binaryzation operation is carried out to the slice information figure；
Step 2B, opening operation is carried out to the slice information figure after image binaryzation is processed, and is subsequently diverted to the step
3。
Beneficial effects of the present invention：The present invention uses image processing techniques, by tissue and impurity in color characteristic and
Differentiated in geometric characteristic, tissue part is effectively identified as far as possible, while entering to most impurity and spot
Row identification is simultaneously rejected, so as to more accurately isolate the tissue part in section preview graph.
Brief description of the drawings
Fig. 1 is a kind of flow chart of embodiment of identification biopsy tissues of the invention；
Fig. 2 is the Background of blank section；
Fig. 3 is the preview graph of the biopsy tissues for shooting；
Fig. 4 is the figure after the edge for wiping out Fig. 3；
Fig. 5 is the slice information figure of biopsy tissues；
Fig. 6 is by the binary picture after image binaryzation treatment by Fig. 5；
Fig. 7 is by the figure after opening operation by Fig. 6；
Fig. 8 is Fig. 7 by the figure after geometry differentiation；
Fig. 9 is Fig. 8 by the figure after color differentiation；
Figure 10 is the figure by being generated after differentiation.
Specific embodiment
The invention will be further described with specific embodiment below in conjunction with the accompanying drawings, but not as limiting to the invention.
As Figure 1-10 shows, a kind of method for recognizing biopsy tissues a, there is provided blank is cut into slices and tissue to be identified is cut
Piece, method includes：
Step 1, scans the Background of blank section and the preview graph of histotomy to be identified；
Step 2, the gray value of preview graph is subtracted the gray value of Background, obtains a section hum pattern；
Step 3, the connected domain of traversal processing slice information figure rejects the spot in slice information figure；
Step 4, the attribute according to connected domain judges whether each connected domain in slice information figure is biopsy tissues respectively,
And reject be judged as be not biopsy tissues connected domain；
Step 5, retains the slice information figure and defeated as biopsy tissues picture after rejecting the connected domain for not being biopsy tissues
Go out.
Above method step is as shown in Figure 1.The present invention uses image processing techniques, according to tissue and miscellaneous qualitative attribution, example
Such as color characteristic and geometric characteristic, biopsy tissues and non-sliced tissue (such as impurity, spot) are differentiated, as far as possible
Effectively tissue part is identified, and reject non-sliced tissue part, so that cutting in more accurately isolating section preview graph
Piece tissue.
The preferred embodiment of the invention, step 4 includes：
In the connected domain by finding out area maximum in the remaining connected domain after the spot in rejecting slice information figure, and
The area of other remaining connected domains is calculated a difference in areas value compared with the area of the maximum connected domain of area respectively,
Connected domain of the area difference in the range of a default area difference is judged as biopsy tissues.
On the one hand, the Area comparison of general impurity is small, and the Area comparison organized is big, thus can by geometry come
Whether determine is biopsy tissues.On the other hand, it is necessary to give face on tissue sections using dyestuff before tissue is scanned
Color, makes it be had an effect with tissue or certain intracellular composition, by spectral absorption and refraction after transparent, makes its various
Fine structure can manifest different colours, and histiocytic various composition can be so shown under the microscope.And it is most of miscellaneous
Matter color is partially yellow, therefore can determine whether biopsy tissues using impurity and histiocytic color distortion.For example, picking
The connected domain maximum except area is found out in the connected domain after spot, by another connected domain and the area of the maximum connected domain of the area
It is compared, obtains a difference in areas value, if the area difference is in the range of default area difference, illustrates another connected domain
Area is also than larger, it can be determined that be connected domain.
The preferred embodiment of the invention, step 4 includes：
Calculate equal by the color of each passages of RGB of the remaining connected domain after the spot in rejecting the slice information figure
It is worth, and the color value of each passages of RGB respectively according to remaining each connected domain is obtained with the color average value processing of each passages of RGB
Corresponding RGB differences, the RGB differences of the same connected domain is corresponding in a default RGB difference ranges
The connected domain be judged as biopsy tissues.
In color method of discrimination, total colour of R passages, total colour of G passages and the B of all connected domains that add up first are logical
Total colour in road, then tries to achieve the average of the average, the average of G passages and channel B of the R passages of all connected domains, when being sentenced
The average of the R colours of disconnected remaining connected domain, G colours, B colours and the average, the average of G passages and channel B of the R passages for calculating
Difference than it is larger when, such as the RGB differences of same connected domain in a default RGB difference ranges, then the remaining company
Logical domain is judged as stained tissue.R colours, G colours, B colours when the remaining connected domain to be judged are led to the corresponding R for calculating
When the average difference of the average, the average of G passages and channel B in road is smaller, now the RGB differences of same connected domain are not pre-
If RGB difference ranges in, then the remaining connected domain is judged as impurity.
The preferred embodiment of the present invention, in the step 4, will in the slice information figure with the slice information figure
The tactile connected domain of side edge be judged as be not the biopsy tissues the connected domain.
The embodiment can be referred to as touch side method of discrimination, its judge connected domain whether the side edge with slice information figure
Touch, the connected domain tactile with the side edge of slice information figure is judged as impurity.
The preferred embodiment of the present invention, in the step 4, by by after the spot in rejecting the slice information figure
Connected domain of the brightness value of remaining connected domain in a default brightness range is judged as YES the company of the biopsy tissues
Logical domain.
Impurity is different with the brightness value of tissue, therefore impurity and tissue can be determined according to brightness.
4 kinds of embodiments for being judged as biopsy tissues in step 4 are enumerated above, in the present invention, wherein one can be used
Plant or numerous embodiments.To can at least twice be judged as that the connected domain of biopsy tissues is considered as connected domain in 4 methods, finally
The section organization chart piece of generation one.In order to improve precision, the connected domain for being all judged as biopsy tissues for 4 times can be preferably selected,
Exported so as to generate biopsy tissues picture.
In the preferred embodiment of the present invention, step 3, the method for rejecting the spot in slice information figure includes：
Reject connected domain of the area less than a default area；And/or
Reject connected domain of the Aspect Ratio less than a default length and width coefficient；And/or
Reject connected domain of the area accounting coefficient less than a default area accounting coefficient.
In this embodiment it is possible to area, length and width and area accounting coefficient according to connected domain these three aspects shapes respectively
Into three steps of the spot for rejecting slice information figure.It should be noted that these three steps are not influenceed by sequencing, can
To perform above three step simultaneously, it is also possible to reject the spot in slice information figure according to any order that can be performed,
Those skilled in the art will envision that the execution sequence of above three step does not influence the effect of the spot for rejecting slice information figure.
Step 3 is spot method of discrimination, rejects all of spot in slice information figure.It is pre- less than one using area is rejected
If area connected domain the step for because the area of the point-like spot on dust in impurity and section is all at hundreds of
Pixel scale, much smaller than the area of tissue, therefore can delete area less than the connected domain of a default area, and this is default
Area could be arranged to, such as 1200 pixels close with the size of general impurity.For being less than one using rejecting Aspect Ratio
Default length and width coefficient, for example, reject Aspect Ratio less than 1/5 or the step for connected domain more than 5, and it is to delete length and width
The special connected domain of ratio.This is that the dyeing spot at cover glass edge is in often water due to section statining or other operation reasons
Flat strip or vertically long bar shaped, its length and width either very big or very little, therefore, it can reject this part Aspect Ratio
Special connected domain.In the step for area accounting coefficient is less than the connected domain of a default area accounting coefficient is rejected, will
The area of connected domain obtains area accounting coefficient divided by the length and width product of the surface area, when area accounting coefficient is pre- less than one
If area accounting coefficient, such as when 1/9, illustrate that the profile of this connected domain is very big, but slice information seldom because this be by
The region of multiple strip spots connection composition at cover glass edge, therefore should give rejecting.
The preferred embodiment of the present invention, before the step 2 is performed, is first carried out following step：Step 1A, wipes out
The marginal portion of the Background of the blank section and the preview graph of the histotomy, is subsequently diverted to the step 2.
Tissue is placed on slide, and organizationally covers a cover glass, on the one hand can prevent dust from falling tissue
On, on the other hand can fix tissue.The marginal portion of Background and preview graph is the marginal portion of slide and cover glass,
It is exclusive PCR as far as possible herein without organizational information, edge is cut and is rejected, to reject distracter.
The preferred embodiment of the present invention, before the step 3 is performed, is first carried out following step：Step 2A, to institute
Stating slice information figure carries out image binaryzation operation, is subsequently diverted to the step 3.
The step is that background is rejected.The preview graph of blank section is impinged upon for recording light source and cut as Background
Optical field distribution on piece.The Background that blank is cut into slices is subtracted each other with the gray value of the correspondence position of the preview graph of histotomy, is obtained
To a slice information figure.This slice information figure is carried out into image binaryzation treatment, can preliminarily dividing tissue.In order to more comprehensively
Identification tissue, the threshold value in binaryzation can be set more relatively low, will tissue and most of impurities identification out.
The preferred embodiment of the present invention, before the step 3 is performed, is first carried out following step：
Step 2A, image binaryzation operation is carried out to the slice information figure；
Step 2B, opening operation is carried out to the slice information figure after image binaryzation is processed, and is subsequently diverted to the step
3。
An opening operation is carried out after image binaryzation treatment, mainly for separating tiny connected domain and larger connection
Edge between domain, smoothed image profile.
The method of the present invention is further illustrated, as shown in figs. 2-10.Fig. 2 show the Background of blank section, and Fig. 3 is
The organized preview graph for shooting.The edge of Background and preview graph is easy to photograph instrument portion, therefore to exclude preview graph
And/or the interference at Background edge, the edge in Fig. 2 and/or Fig. 3 can be wiped out.Fig. 4 shows the reality at the edge for wiping out Fig. 3
Apply example.Then, the gray value of the Background after wiping out subtracts the gray value of the preview graph after wiping out, and obtains one as shown in Figure 5
Slice information figure.On the basis of Fig. 5, image binaryzation treatment is carried out, be used to further discriminate between tissue and impurity, obtain figure
6.Then, opening operation is carried out to the picture shown in Fig. 6, removes minute impurities, obtain Fig. 7.Further, can be by geometric form
Shape method of discrimination obtains the histotomy figure shown in Fig. 8；The histotomy shown in Fig. 9 can also be obtained by color method of discrimination
Figure；Based on Fig. 8 and Fig. 9, comprehensive descision goes out whether connected domain is to be considered as tissue, such as Figure 10 of the result after identification.
Preferred embodiments of the present invention are these are only, embodiments of the present invention and protection domain is not thereby limited, it is right
For those skilled in the art, should can appreciate that all utilization description of the invention and diagramatic content made equivalent replaces
Change and obviously change resulting scheme, should be included in protection scope of the present invention.
Claims (9)
1. it is a kind of recognize biopsy tissues method, it is characterised in that provide a blank section and histotomy to be identified, institute
The method of stating includes：
Step 1, scans the Background of the blank section and the preview graph of histotomy to be identified；
Step 2, the gray value of the preview graph is subtracted the gray value of the Background, obtains a section hum pattern；
Step 3, the connected domain of slice information figure described in traversal processing rejects the spot in the slice information figure；
Step 4, the attribute according to the connected domain judges whether each connected domain in the slice information figure is section respectively
Tissue, and reject be judged as be not the biopsy tissues the connected domain；
Step 5, retains the slice information figure after rejecting the connected domain for not being the biopsy tissues and as section group
Knit picture output.
2. it is according to claim 1 identification biopsy tissues method, it is characterised in that the step 4 includes：
In the connected domain by finding out area maximum in the remaining connected domain after the spot in rejecting the slice information figure, and
The area of other remaining connected domains is calculated an area compared with the area of the maximum connected domain of the area respectively
Difference, biopsy tissues are judged as by connected domain of the area difference in the range of a default area difference.
3. it is according to claim 1 identification biopsy tissues method, it is characterised in that the step 4 includes：
The color average of each passages of RGB by the remaining connected domain after the spot in rejecting the slice information figure is calculated,
And the color value of each passages of RGB respectively according to remaining each connected domain obtains right with the color average value processing of each passages of RGB
The RGB differences answered, the RGB differences of the same connected domain is corresponding in a default RGB difference ranges
The connected domain is judged as biopsy tissues.
4. it is according to claim 1 identification biopsy tissues method, it is characterised in that in the step 4, will be cut described
In piece hum pattern the connected domain tactile with the side edge of the slice information figure be judged as be not the biopsy tissues the company
Logical domain.
5. it is according to claim 1 identification biopsy tissues method, it is characterised in that in the step 4, will by rejecting
Connected domain of the brightness value of the remaining connected domain after spot in the slice information figure in a default brightness range is sentenced
Break to be the connected domain of the biopsy tissues.
6. it is according to claim 1 identification biopsy tissues method, it is characterised in that in the step 3, reject described in cut
The method of the spot in piece hum pattern includes：
Reject connected domain of the area less than a default area；And/or
Reject connected domain of the Aspect Ratio less than a default length and width coefficient；And/or
Reject connected domain of the area accounting coefficient less than a default area accounting coefficient.
7. the method for identification biopsy tissues according to claim 1, it is characterised in that first before the step 2 is performed
First carry out following step：
Step 1A, wipes out the marginal portion of the Background of the blank section and the preview graph of the histotomy, then turns
To the step 2.
8. the method for identification biopsy tissues according to claim 1, it is characterised in that first before the step 3 is performed
First carry out following step：
Step 2A, image binaryzation operation is carried out to the slice information figure, is subsequently diverted to the step 3.
9. the method for identification biopsy tissues according to claim 1, it is characterised in that first before the step 3 is performed
First carry out following step：
Step 2A, image binaryzation operation is carried out to the slice information figure；
Step 2B, opening operation is carried out to the slice information figure after image binaryzation is processed, and is subsequently diverted to the step 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710040116.1A CN106815849A (en) | 2017-01-18 | 2017-01-18 | A kind of method for recognizing biopsy tissues |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710040116.1A CN106815849A (en) | 2017-01-18 | 2017-01-18 | A kind of method for recognizing biopsy tissues |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106815849A true CN106815849A (en) | 2017-06-09 |
Family
ID=59112416
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710040116.1A Pending CN106815849A (en) | 2017-01-18 | 2017-01-18 | A kind of method for recognizing biopsy tissues |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106815849A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107493403A (en) * | 2017-08-11 | 2017-12-19 | 宁波江丰生物信息技术有限公司 | A kind of digital pathological section scanning system |
| CN108766534A (en) * | 2018-04-09 | 2018-11-06 | 新乡医学院 | A kind of digital pathological section scanning system |
| CN112580663A (en) * | 2020-12-09 | 2021-03-30 | 山东志盈医学科技有限公司 | Method and apparatus for coverslip boundary removal in digital slide scanner |
| CN112633197A (en) * | 2020-12-28 | 2021-04-09 | 宁波江丰生物信息技术有限公司 | Method and system for tissue region identification of fluorescence section |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101877074A (en) * | 2009-11-23 | 2010-11-03 | 常州达奇信息科技有限公司 | Tubercle bacillus target recognizing and counting algorithm based on diverse characteristics |
| CN102298700A (en) * | 2011-06-09 | 2011-12-28 | 华东师范大学 | Method for recognizing and positioning cells in bone marrow pathology image |
| CN104899584A (en) * | 2015-06-29 | 2015-09-09 | 宁波江丰生物信息技术有限公司 | Stained section identification method based on fuzzy thought |
| CN105139377A (en) * | 2015-07-24 | 2015-12-09 | 中南大学 | Rapid robustness auto-partitioning method for abdomen computed tomography (CT) sequence image of liver |
| CN105513028A (en) * | 2016-01-18 | 2016-04-20 | 宁波江丰生物信息技术有限公司 | Slice preview recognition method for mycobacteria tuberculosis |
| CN105741293A (en) * | 2016-01-30 | 2016-07-06 | 上海联影医疗科技有限公司 | Method for positioning organs in medical image |
-
2017
- 2017-01-18 CN CN201710040116.1A patent/CN106815849A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101877074A (en) * | 2009-11-23 | 2010-11-03 | 常州达奇信息科技有限公司 | Tubercle bacillus target recognizing and counting algorithm based on diverse characteristics |
| CN102298700A (en) * | 2011-06-09 | 2011-12-28 | 华东师范大学 | Method for recognizing and positioning cells in bone marrow pathology image |
| CN104899584A (en) * | 2015-06-29 | 2015-09-09 | 宁波江丰生物信息技术有限公司 | Stained section identification method based on fuzzy thought |
| CN105139377A (en) * | 2015-07-24 | 2015-12-09 | 中南大学 | Rapid robustness auto-partitioning method for abdomen computed tomography (CT) sequence image of liver |
| CN105513028A (en) * | 2016-01-18 | 2016-04-20 | 宁波江丰生物信息技术有限公司 | Slice preview recognition method for mycobacteria tuberculosis |
| CN105741293A (en) * | 2016-01-30 | 2016-07-06 | 上海联影医疗科技有限公司 | Method for positioning organs in medical image |
Non-Patent Citations (1)
| Title |
|---|
| 王奇文 等: "基于形态学的小鼠舌头切片图像分割与实现", 《计算机工程》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107493403A (en) * | 2017-08-11 | 2017-12-19 | 宁波江丰生物信息技术有限公司 | A kind of digital pathological section scanning system |
| CN107493403B (en) * | 2017-08-11 | 2019-09-24 | 宁波江丰生物信息技术有限公司 | A kind of digital pathological section scanning system |
| CN108766534A (en) * | 2018-04-09 | 2018-11-06 | 新乡医学院 | A kind of digital pathological section scanning system |
| CN112580663A (en) * | 2020-12-09 | 2021-03-30 | 山东志盈医学科技有限公司 | Method and apparatus for coverslip boundary removal in digital slide scanner |
| CN112633197A (en) * | 2020-12-28 | 2021-04-09 | 宁波江丰生物信息技术有限公司 | Method and system for tissue region identification of fluorescence section |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106815849A (en) | A kind of method for recognizing biopsy tissues | |
| CN114820627B (en) | Textile quality detection method based on optical means | |
| CN102426649B (en) | Simple steel seal digital automatic identification method with high accuracy rate | |
| TWI467515B (en) | Multi-color dropout for scanned document | |
| US7068840B2 (en) | Determination of an illuminant of digital color image by segmentation and filtering | |
| CN108181316B (en) | Bamboo strip defect detection method based on machine vision | |
| Mehrjou et al. | Automatic malaria diagnosis system | |
| KR20050090945A (en) | Character recognition method | |
| CN109668909A (en) | A kind of glass defect detection method | |
| WO2004044845A2 (en) | Image analysis | |
| CN109523524B (en) | Eye fundus image hard exudation detection method based on ensemble learning | |
| CN109241973B (en) | Full-automatic soft segmentation method for characters under texture background | |
| CN104834890B (en) | A kind of extracting method to word expression information in calligraphy work | |
| CN111986157B (en) | Digital pathological image quality evaluation system | |
| CN110687122A (en) | Method and system for detecting surface cracks of ceramic tile | |
| CN110648312A (en) | Method for identifying wool and cashmere fibers based on scale morphological characteristic analysis | |
| CN106960424B (en) | Tubercle bacillus image segmentation and identification method and device based on optimized watershed algorithm | |
| CN115620075A (en) | Method, system and equipment for generating data set for leukocyte classification model | |
| KR20100128501A (en) | Crack and center segregation estimation method | |
| CN107145888A (en) | Video caption real time translating method | |
| CN110458042B (en) | Method for detecting number of probes in fluorescent CTC | |
| JP2015194700A (en) | Focusing determination device, focusing determination method, and imaging device | |
| CN113657401B (en) | Probability map visualization method and device for defect detection | |
| CN112270683B (en) | IHC digital preview image identification and organization foreground segmentation method and system | |
| JP3955467B2 (en) | Image processing program and image processing apparatus |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170609 |
|
| RJ01 | Rejection of invention patent application after publication |