CN100495034C - Blood cell analysis method based on machine vision - Google Patents
Blood cell analysis method based on machine vision Download PDFInfo
- Publication number
- CN100495034C CN100495034C CNB2006100315045A CN200610031504A CN100495034C CN 100495034 C CN100495034 C CN 100495034C CN B2006100315045 A CNB2006100315045 A CN B2006100315045A CN 200610031504 A CN200610031504 A CN 200610031504A CN 100495034 C CN100495034 C CN 100495034C
- Authority
- CN
- China
- Prior art keywords
- image
- area
- blood cell
- red blood
- segment
- 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.)
- Expired - Fee Related
Links
Images
Classifications
-
- C—CHEMISTRY; METALLURGY
- C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
- C12M—APPARATUS FOR ENZYMOLOGY OR MICROBIOLOGY; APPARATUS FOR CULTURING MICROORGANISMS FOR PRODUCING BIOMASS, FOR GROWING CELLS OR FOR OBTAINING FERMENTATION OR METABOLIC PRODUCTS, i.e. BIOREACTORS OR FERMENTERS
- C12M41/00—Means for regulation, monitoring, measurement or control, e.g. flow regulation
- C12M41/30—Means for regulation, monitoring, measurement or control, e.g. flow regulation of concentration
- C12M41/36—Means for regulation, monitoring, measurement or control, e.g. flow regulation of concentration of biomass, e.g. colony counters or by turbidity measurements
Landscapes
- Investigating Or Analysing Biological Materials (AREA)
Abstract
This invention relates to a blood cell analysis method based on machine vision. Inputting the blood sample into a computer by a camera to convert the color image into the black-and-white and then the binary one and scale the binary image; calculating the area and taking count of the scaled image to obtain the number of RBC, WBC and platelet. This invention has high efficiency and accuracy with low cost.
Description
Technical field
The present invention relates to a kind of blood cell analysis method based on machine vision.
Background technology
At present, hemocytometer counting method soil will have microscope manual observation counting method, resistance-type hemocytometer counting method, laser radiation counting method.Microscope manual observation counting method is with the blood sample on the microscopic examination blood cell counting plate, more various cells is manually counted, thereby draws each parameter, and the result is more accurate, but inefficiency, speed is very slow.Resistance-type hemocytometer counting method is proposed in nineteen fifty-three by U.S. Ku Erte, manages the cell of counting in the blood according to the bad electric conductivity of haemocyte and the antigen that has a resistance, and is engaged in graininess and measures, thereby calculates the data of blood routine examination.Various big hospital all is extensive use of this cellanalyzer at present, this automation equipment degree height, but price is expensive, and the maintenance cost height needs calibration after use a period of time, and accuracy is sometimes not as microscope manual observation counting method.The laser radiation counting method is utilized the laser beam irradiation haemocyte of shaping, compiles and convert to corresponding data by the laser beam after refraction, transmission, the reflection.Or utilize the high frequency waves reflection, collect conversion of signals and become data.The shortcoming that this method of counting exists is: costliness, and inaccurate, job insecurity.
Summary of the invention
For overcoming the shortcoming of above-mentioned existing hemocytometer counting method, the invention provides that a kind of cost is lower, efficient is high, counting is accurately based on the blood cell analysis method of machine vision.
For realizing above-mentioned purpose, the blood cell analysis method based on machine vision of the present invention may further comprise the steps:
Leukocytic counting step is as follows:
A. blood sample is melted blood and dyeing processing, dissolve red blood cell, put into microscopically and amplify, again by camera head input computing machine;
B. the coloured image with input converts black white image to;
C. calculate the brightness value of black white image, calculate mean flow rate;
D. image is converted into binary image, pixel is set to 1 greater than the picture point of mean flow rate in the image, otherwise is 0;
E. to being that 1 segment is demarcated in the image;
F. the segment of being demarcated is carried out the cavity and judge, if having, filling cavity then;
G. to the figure block count, and calculate the area of each segment, the area multiplication by constants is volume, so just obtain number of WBC and volume separately;
Red blood cell and thrombocyte counting step are as follows:
1), will put into microscopically and amplify with after blood sample dilution of once extracting and the dyeing, again by camera head input computing machine;
2), the coloured image with input converts black white image to;
3), calculate the brightness value of black white image, calculate mean flow rate;
4), image is converted into binary image, pixel is set to 1 greater than the picture point of mean flow rate in the image, otherwise is 0;
5), to being that 1 segment is demarcated in the image;
6), have the cavity to judge to each segment of demarcating, if having, filling cavity then;
7), to each segment and carry out circularity coupling, near circular segment is a part of single red blood cell and blood platelet, calculate the area of all segments, be divided into two classes by size, a big class is the red blood cell of single red blood cell or adhesion, and a little class is a blood platelet, calculates these near the circular erythrocytic average area of a class, as single erythrocytic standard area, in like manner can obtain single hematoblastic standard area;
8), the area of the class segment that all are big and standard red blood cell area are relatively, 1.5 times of persons of overgauge area are considered as two above red cell adherences, how many red cell adherences have been calculated by the area multiple, and calculating number, be near the mark the red blood cell area for single red blood cell, also count together and add up;
9), all little class segments are blood platelet, to platelet count, obtain red blood cell and hematoblastic number and area thereof like this, and the area multiplication by constants is volume; The red blood cell number is deducted with the leukocyte count in the volume, be RBC number actual in institute's test sample product.
Beneficial effect of the present invention: the present invention will dye with microscope and camera or undyed blood sample picked-up picture is sent into computing machine, by computing machine samples pictures is carried out the figure identification, identification and number go out various haemocyte from digital photo, and draw the data of blood routine examination thus, the inventive method has adopted automation mechanized operation, has that cost is lower, high efficiency and an advantage of high accuracy.Overcome the shortcoming that resistance-type hemocytometer counting method price is expensive, maintenance cost is high, accuracy rate is lower.
Description of drawings
Fig. 1 is the situation after the micro-image of haemocyte among the present invention is converted into the bi-values image.
Fig. 2 is the filling synoptic diagram of cell figure among the present invention.
Embodiment
Leukocytic counting step is as follows among the present invention:
The blood sample that extracts is melted blood and dyeing processing, dissolve red blood cell, only can see leucocyte at microscopically, reduce enlargement factor or a plurality of images are spliced, make leukocyte count among the figure between 100-200, transfer image to 8 black white images, if the image that collects is a black white image, can ignore this step, as be coloured image, the wherein one deck that can extract HSL or RGB element transfers black white image to;
Calculate the brightness value of full figure, calculate mean flow rate, be about to the value addition of all pixels, again divided by the pixel number;
Image is converted into the BINARY image, the image of 0 and 1 two value is promptly only arranged, be as the criterion with mean flow rate, the picture point greater than this brightness in the image is set to 1, otherwise is 0;
Carry out the MORPHOLOGY algorithm operating, each segment on the uncalibrated image;
The segment of being demarcated is carried out the cavity judge, if having, filling cavity then, as shown in Figure 2;
To the figure block count, and calculate the area of each segment, the area multiplication by constants is volume, so just obtain number of WBC and volume separately;
Hiving off by leukocytic volume, is lymphocyte less than 90FL person, is the maxicell district greater than 160FL person, and in the middle of being positioned at is monocyte;
Red blood cell and enumeration of thrombocytes step are as follows:
Will be after dilution and dyeing with the blood sample that once extracts, place microscopically, gather picture input computing machine by digital camera, it is good that extension rate should see that red blood cell can not be bonded as in a large number with microscopically, and the camera lens enlargement factor should make the red blood cell in the visual field appear at more than 200;
Convert the coloured image of input to black white image;
Calculate the brightness value of full figure, calculate mean flow rate;
Image is converted into BINARY (binaryzation) image, the image of 0 and 1 two value is promptly only arranged, be as the criterion with mean flow rate, the picture point greater than this brightness in the image is set to 1, otherwise is 0;
Carry out the segment identification operation in the MORPHOLOGY algorithm, can recognize each thumbnail on the image, be about to each value and be 1 graph block and find out and demarcate; There are many algorithms can finish this operation, wherein a kind of is the method for exhaustion, light from first pixel, if value is 1, then demarcating is a segment, and check the value up and down that this point is adjacent, if be 1 then also be included in this segment, find up to the pixel of all these segments, from entire image, reject picture point under this segment, these segments may be independent red blood cell and blood platelets, also may be to be linked to be one cell body, may be the noise spot of image also, and a class segment of removing the area minimum is (such as less than 9 pigment points, or select the general minimum following numerical value of the due size of blood platelet), can get rid of noise spot;
To each segment that finds, whether there is the cavity to judge, if having, filling cavity then;
To each segment and carry out circularity coupling, near circular segment is part red blood cell and blood platelet, to these segment reference areas, be divided into two classes by size, a big class is the red blood cell of single red blood cell or adhesion, and a little class is a blood platelet, calculates these erythrocytic average areas, be considered as erythrocytic standard area, in like manner calculate hematoblastic standard area;
All segment areas and standard red blood cell area are compared, and 1.5 times of persons of overgauge area are considered as two above red cell adherences, have calculated how many red cell adherences by the area multiple, and calculating number;
To all non-adhesion segments (promptly near or less than the segment of standard red blood cell area) reference area, magnitude classification in proportion, segment can be classified as two significantly districts, and a big district is non-adhesion red blood cell and minute quantity leucocyte, a little district is a blood platelet, respectively counting;
So far drawn red blood cell and blood platelet and area thereof, the area multiplication by constants has been volume; Promptly draw red blood cell and hematoblastic number and volume thereof, the more by volume big or small leucocyte number that cuts the equal volume size of the red blood cell number that obtains is promptly obtained red blood cell number actual in the sample;
Press red blood cell, leucocyte, hematoblastic volume and number can obtain red blood cell, leucocyte, hematoblastic histogram.Measure haemoglobin with the spectrophotometer colorimetric principle, by hemoglobin concentration, red blood cell count(RBC), white blood cell count(WBC) and classification, platelet count can obtain the detected parameters of tripartite group blood cell detector.
Claims (1)
1, a kind of blood cell analysis method based on machine vision may further comprise the steps:
The leukocytic counting and the step of hiving off are as follows:
A. blood sample is melted blood and dyeing processing, dissolve red blood cell, put into microscopically and amplify, again by camera head input computing machine;
B. the coloured image with input converts black white image to;
C. calculate the brightness value of black white image, calculate mean flow rate;
D. image is converted into binary image, pixel is set to 1 greater than the picture point of mean flow rate in the image, otherwise is 0;
E. to being that 1 segment is demarcated in the image;
F. the segment of being demarcated is carried out the cavity and judge, if having, filling cavity then;
G. to the figure block count, and calculate the area of each segment, the area multiplication by constants is volume, so just obtain number of WBC and volume separately;
H. hiving off by leukocytic volume, is lymphocyte less than 90FL person, is the maxicell district greater than 160FL person, and in the middle of being positioned at is monocyte;
Red blood cell and platelet count step are as follows:
1), will put into microscopically and amplify with after blood sample dilution of once extracting and the dyeing, again by camera head input computing machine;
2), the coloured image with input converts black white image to;
3), calculate the brightness value of black white image, calculate mean flow rate;
4), image is converted into binary image, pixel is set to 1 greater than the picture point of mean flow rate in the image, otherwise is 0;
5), to being that 1 segment is demarcated in the image;
6), have the cavity to judge to each segment of demarcating, if having, filling cavity then;
7), to each segment and carry out circularity coupling, near circular segment is a part of single red blood cell and blood platelet, calculate the area of all segments, be divided into two classes by size, a big class is the red blood cell of single red blood cell or adhesion, and a little class is a blood platelet, calculates these near the circular erythrocytic average area of a class, as single erythrocytic standard area, in like manner can obtain single hematoblastic standard area;
8), the area of the class segment that all are big and standard red blood cell area are relatively, 1.5 times of persons of overgauge area are considered as two above red cell adherences, how many red cell adherences have been calculated by the area multiple, and calculating number, be near the mark the red blood cell area for single red blood cell, also count together and add up;
9), all little class segments are blood platelet, to platelet count, obtain red blood cell and hematoblastic number and area thereof like this, and the area multiplication by constants is volume; The red blood cell number is deducted with the leukocyte count in the volume, be RBC number actual in institute's test sample product.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB2006100315045A CN100495034C (en) | 2006-04-18 | 2006-04-18 | Blood cell analysis method based on machine vision |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNB2006100315045A CN100495034C (en) | 2006-04-18 | 2006-04-18 | Blood cell analysis method based on machine vision |
Publications (2)
Publication Number | Publication Date |
---|---|
CN1837819A CN1837819A (en) | 2006-09-27 |
CN100495034C true CN100495034C (en) | 2009-06-03 |
Family
ID=37015278
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNB2006100315045A Expired - Fee Related CN100495034C (en) | 2006-04-18 | 2006-04-18 | Blood cell analysis method based on machine vision |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN100495034C (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104655546A (en) * | 2015-03-12 | 2015-05-27 | 重庆大学 | Counting method and counting system for milk body cells based on mobile equipment |
Families Citing this family (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102519861B (en) * | 2011-12-09 | 2013-10-30 | 南昌百特生物高新技术股份有限公司 | Blood cell form multi-mode image screening method |
JP5447546B2 (en) * | 2012-01-31 | 2014-03-19 | 東洋製罐グループホールディングス株式会社 | Cell counting method, cell counting apparatus, and cell counting program |
CN102819765B (en) * | 2012-02-28 | 2016-01-20 | 浙江工业大学 | A kind of milk somatic cell method of counting based on computer vision |
JP5413501B1 (en) * | 2012-12-07 | 2014-02-12 | 富士ゼロックス株式会社 | Image processing apparatus, image processing system, and program |
CN109655383A (en) * | 2017-10-11 | 2019-04-19 | 南京大学 | A kind of detection device and its method based on blood platelet projection imaging |
CN108627637A (en) * | 2018-06-04 | 2018-10-09 | 江苏柯伦迪医疗技术有限公司 | A kind of platelet aggregation detecting system and method |
CN109117937B (en) * | 2018-08-16 | 2022-03-22 | 杭州电子科技大学信息工程学院 | Leukocyte image processing method and system based on connected area subtraction |
CN109580625A (en) * | 2018-12-17 | 2019-04-05 | 李伟 | Integrated blood inspection platform and its application method |
CN111445448A (en) * | 2020-03-19 | 2020-07-24 | 中国医学科学院北京协和医院 | Single-cell hemoglobin determination method and device based on image processing |
CN113759129A (en) * | 2021-09-26 | 2021-12-07 | 北京倍肯恒业科技发展股份有限公司 | Rapid detection card for leucocyte-bound C-reactive protein and preparation method thereof |
CN114047151A (en) * | 2021-11-15 | 2022-02-15 | 四川丹诺迪科技有限公司 | Instrument and detection method for simultaneously carrying out sample analysis and immunity measurement |
CN114813522B (en) * | 2022-06-29 | 2022-10-21 | 深圳安侣医学科技有限公司 | Blood cell analysis method and system based on microscopic amplification digital image |
-
2006
- 2006-04-18 CN CNB2006100315045A patent/CN100495034C/en not_active Expired - Fee Related
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104655546A (en) * | 2015-03-12 | 2015-05-27 | 重庆大学 | Counting method and counting system for milk body cells based on mobile equipment |
CN104655546B (en) * | 2015-03-12 | 2017-07-07 | 重庆大学 | A kind of milk somatic cell method of counting and system based on mobile device |
Also Published As
Publication number | Publication date |
---|---|
CN1837819A (en) | 2006-09-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN100495034C (en) | Blood cell analysis method based on machine vision | |
CN103076288B (en) | A kind of flesh of fish self-grading device based on computer vision and method | |
CN105136795B (en) | Blood sample detection device, method and system | |
EP3279657B1 (en) | Urinalysis system and urinalysis method | |
CN105445170B (en) | Identify the method for biological sample target components and the device of analyzing biologic fluids sample | |
Prasad et al. | Image analysis approach for development of a decision support system for detection of malaria parasites in thin blood smear images | |
CN101782524B (en) | Cell image processing apparatus, cell image processing method and computer system | |
CN109612911B (en) | Full-automatic sperm cell detector | |
EP2083268A1 (en) | Analysis method for 5-differential complete blood cell based on visual image | |
CN102662048B (en) | Method and device for analyzing urine | |
CN102147523B (en) | Biological digital microscope with double ccd (charge coupled device) light sensitive elements and photographic image processing method thereof | |
CN105143850A (en) | Autofocus systems and methods for particle analysis in blood samples | |
CN102967586B (en) | Device for online automatically classifying pearl color and gloss based on monocular multi-view machine vision | |
CN104897556A (en) | Milk somatic cell counting device and method based on intelligent terminal and micro-fluidic chip | |
CN107064079A (en) | For recognizing the hematoblastic method and apparatus in whole blood sample | |
CN111504885B (en) | Analysis method of full-automatic blood smear morphological analysis device based on machine vision | |
CN103984979A (en) | Lens-diffraction-imaging-free automatic algae detection and counting device and method | |
CN110186836A (en) | The optofluidic flow cytometer of circulating tumor cell separation analysis and Classification Count | |
CN109967143A (en) | A kind of cell size detection method based on micro-fluidic microscopic system | |
Meimban et al. | Blood cells counting using python opencv | |
Schonbrun et al. | Dye exclusion microfluidic microscopy | |
CN117576103B (en) | Urinary sediment microscopic examination analysis system integrating electric control microscope and deep learning algorithm | |
WO2020100667A1 (en) | Information processing device, information processing method, and computer program | |
CN114441452A (en) | Optical fiber pigtail detection method | |
CN102297833A (en) | Method for determination of content of various cells in human blood or bone marrow |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
C17 | Cessation of patent right | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090603 Termination date: 20110418 |