CN113393479B - Method for dividing test tube holes in cell plate image - Google Patents
Method for dividing test tube holes in cell plate image Download PDFInfo
- Publication number
- CN113393479B CN113393479B CN202110557045.9A CN202110557045A CN113393479B CN 113393479 B CN113393479 B CN 113393479B CN 202110557045 A CN202110557045 A CN 202110557045A CN 113393479 B CN113393479 B CN 113393479B
- Authority
- CN
- China
- Prior art keywords
- image
- test tube
- radius
- cell plate
- minimum bounding
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000012360 testing method Methods 0.000 title claims abstract description 64
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000003696 structure analysis method Methods 0.000 claims abstract description 5
- 238000012216 screening Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 claims description 2
- 230000002159 abnormal effect Effects 0.000 abstract description 9
- 238000004113 cell culture Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 230000001174 ascending effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000000969 carrier Substances 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000012258 culturing Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000007717 exclusion Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000008961 swelling Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration using local operators
- G06T5/30—Erosion or dilatation, e.g. thinning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- 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
- G06T2207/30024—Cell structures in vitro; Tissue sections in vitro
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention provides a method for dividing a test tube hole in a cell plate image. Firstly, a gray level image of a cell plate is read in, the gray level image is expanded according to two self-defined kernels to weaken image noise, the expanded image is subjected to inverse binary thresholding, then the two images subjected to binarization are subjected to intersection operation, then image white noise is removed by using open operation, and then contours in the image are extracted by using a topological structure analysis method which follows according to boundaries, so that test tube hole contours in the cell plate are output. However, partial noise exists in the acquired contour, minimum closed circle data in the contour is extracted on the basis, then a circle with a larger or smaller radius is removed on the basis of the circle radius, abnormal data of the circles with intersection of every two is removed, and finally the target cell bank data set is output. The method can divide the test tube holes rapidly and accurately, and is convenient for users to directly check the cell population culture condition in each test tube hole through image recognition.
Description
Technical Field
The invention belongs to the field of image segmentation, and particularly relates to a method for segmenting a test tube hole in a cell plate image.
Background
Digital images are one of the carriers of information transfer and play an important role in daily life or work. For example, in the biochemical field, in the process of culturing cloned cells using cell plates, it is necessary to check and confirm the culture conditions of cell populations in each test tube well from time to time, and record the growth states of cells in each test tube well too original and cumbersome. Therefore, there is a need for a method of dividing an image of a cell plate from an image of the cell plate, so that the cell plate image can be captured by a high-definition industrial camera, and the cell plate image can be divided, so that the image of each cell plate can be observed to screen the state of the cell population. However, there are many structural disturbances outside the test tube hole on the cell plate surface, and many surface structures of the cell plate have a staggered circle, polygon, etc. with the test tube hole in the image shot by the camera, so that it is difficult to divide the test tube hole in the cell plate image by adopting a conventional method.
Disclosure of Invention
The invention aims to solve the defects of too original and complicated growth state of cells in each test tube hole, and provides a method for dividing the test tube holes in a cell plate image
The cell plate image can be shot by a high-definition industrial camera, the cell plate image is divided into test tube holes, and then the image of each test tube hole is observed to screen the state of the cell group.
The technical scheme of the invention comprises the following steps:
a method for dividing a test tube hole in a cell plate image comprises the following steps:
s1, respectively performing image expansion on a cell plate gray level image to be segmented by using kernel matrixes with different sizes to obtain a first expanded image and a second expanded image;
s2, respectively carrying out binarization processing on the first expanded image and the second expanded image to obtain a first binarized image and a second binarized image;
s3, performing intersection operation on the first binarization image and the second binarization image to obtain an intersection operation result image;
s4, performing open operation on the intersection operation result image to remove white noise of the image, and obtaining a denoising image;
s5, extracting contours in the denoising image by using a topological structure analysis method followed by boundaries to obtain a contour set in the denoising image;
s6, aiming at each contour in the contour set, acquiring the circle center and the radius of the minimum bounding circle, so as to form the minimum bounding circle set;
s7, sorting all the minimum bounding circles in the minimum bounding circle set according to the radius, selecting a group of continuous minimum bounding circles with the radius deviation not exceeding a set threshold value from the sorting sequence as the representative of the outer contour of the test tube, and obtaining the average radius R of the group of continuous minimum bounding circles 1 The method comprises the steps of carrying out a first treatment on the surface of the At the average radius R 1 Setting a radius screening range which accords with the radius deviation amount of the test tube for the central value, and screening a minimum bounding circle to be selected, the radius of which is within the range, from the minimum bounding circle set;
s8, calculating the average radius R of all the minimum bounding circles to be selected 2 Then judging whether an intersection exists between the minimum bounding circles to be selected in pairs, if so, keeping the radius closest to the average radius R 2 Simultaneously deleting another circle, and outputting all the smallest circle to be selected, which does not have intersection, as the outer contour of the test tube;
s9, dividing the gray level image of the cell plate by utilizing all the output test tube outer outlines to obtain test tube hole areas in each test tube outer outline.
Preferably, each of S1 to S7 is implemented based on an open source software library Opencv.
Preferably, in the S1, in Opencv, image expansion is performed on the cell plate gray-scale image to be segmented by using the dialate () function with (3, 3) and (8, 8) as kernel matrices, so as to obtain a first expanded image and a second expanded image respectively.
Preferably, in S2, the inverse binary thresholding is performed on the first and second images after expansion using a threshold () function with 240 as a threshold.
Preferably, in the step S3, the first binarized image and the second binarized image are interleaved with a bitwise_and () function.
Preferably, in S4, the blending result image is subjected to an open operation by using a morphingyoex () function to remove white noise of the image.
Preferably, in S5, the following call method is used to extract the contour in the denoising image by using the topology analysis method that follows according to the boundary:
vector<vector<cv::Point>>contours;
findContours(srcImgGray,contours,RETR_LIST,
CHAIN_APPROX_SIMPLE);
where srcImgGray is the de-noised image, concurs is the set of contours in the de-noised image, RETR_LIST means that all contours are retrieved without establishing any hierarchical relationship, CHAIN_APPROX_SIMPLE means that the horizontal, vertical and diagonal segments are compressed, and only the endpoints thereof are preserved.
Preferably, in the step S6, the center and the radius of the minimum bounding circle are obtained for each contour in the contour set by using a minEnclosingCircle () function.
Preferably, in the step S7, when a group of consecutive minimum bounding circles is selected from the ordered sequence as a representative of the outer contour of the test tube, a set threshold value of 10% is adopted, that is, the deviation between the maximum radius and the minimum radius in the selected consecutive minimum bounding circles is not more than 10%; the radius selection range is [0.8R 1 ,1.2R 1 ]。
Preferably, in S8, it is determined whether there is an intersection between the two minimum bounding circles by calculating whether the center distances of the two minimum bounding circles are smaller than the sum of the radii.
Compared with the prior art, the invention has the following beneficial effects:
the invention can quickly and automatically divide the test tube hole area in the cell plate image, and can automatically identify the interference area which does not belong to the test tube hole in the dividing process, thereby ensuring the dividing accuracy. Compared with the traditional method for screening the cell plates, the method has the advantages that a person using the cell plates to culture cloned cells can directly check the culture condition of the cell population in each test tube hole through image recognition, and the image data of a single test tube hole selected in each time period in the cell culture process can be recorded so as to manually screen the cell population in the test tube hole by the staff, and the method has traceability. And the later stage can carry out software management to each test tube hole that cuts apart to cut apart again and detect the cell crowd, for the staff to carry out half intelligent even full intelligent screening to the cell crowd in the test tube hole.
Drawings
FIG. 1 is a flow chart of a method of segmenting a test tube aperture in a cell plate image.
Fig. 2 is a sample cell plate source image.
FIG. 3 is a sample image with abnormal data after primary segmentation of the cell plate source.
Fig. 4 is a flow chart of a radius exclusion anomaly data algorithm.
Fig. 5 is a sample image after segmentation of the cell plate.
Detailed Description
Specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.
In a preferred embodiment of the present invention, a method for dividing a tube hole in a cell plate image is provided, wherein the method is divided into two parts, firstly, the preliminary tube hole division detection from step 1 to step 7 is performed based on an open source software library Opencv, then, through an anomaly removal algorithm independently designed in step 8 and step 9, firstly, round anomaly data with larger or smaller radius is removed based on a round radius, then, intersection algorithm is utilized to remove round anomaly data with intersection, and finally, tube hole images in the cell plate are output, and the specific steps are described as shown in fig. 1. The cell plate image in the present invention is a top view image of a cell plate, and a cell plate image is obtained by photographing a cell plate in a horizontal state with a high-definition industrial camera. In such images, there are numerous test tube holes and other interfering structures, an example of a cell plate image is shown in FIG. 2.
Step 1, reading in cell plate gray level images, and respectively performing image expansion on the cell plate gray level images to be segmented by using kernel matrixes with different sizes to obtain a first expanded image and a second expanded image.
In this embodiment, the morphological operation tools in Opencv may be used, and the (3, 3) and (8, 8) are used as kernel matrices to obtain image custom kernels, which are kernel3 and kernel8 respectively, and the gray level map is expanded by using kernel3 and kernel8, which is as follows:
firstly, taking (3, 3) and (8, 8) as kernel matrixes to obtain image custom kernels, wherein the calling mode is as follows:
Mat kernel3=getStructuringElement(MORPH_RECT,Size(3,3));
Mat kernel6=getStructuringElement(MORPH_RECT,Size(8,8));
parameter 1 uses morph_rect, a rectangular structural element.
The parameter 2 is Size (3, 3), i.e. matrix Size.
Then, the gray level map is expanded by using a dialite () function in a manner of kernel3 and kernel8 as follows:
dilate(srcImgGray,out3,kernel3,Point(-1,-1),3);
dilate(srcImgGray,out6,kernel6,Point(-1,-1),3);
parameter 1 is srcImgGray, a gray scale image of the cell plate image.
The parameters 2 are out3 and out6, namely the image result after expansion.
And the parameters 3 are kernel3 and kernel6, namely the custom kernel output in the step 1.
The parameter 4 is Point (-1, -1), i.e. the anchor Point of the custom kernel, as a default value.
The parameter 5 is a value of 3, i.e. the number of times corrosion and swelling are applied.
And 2, performing inverse binary thresholding on the first and second inflated images to obtain first and second binarized images.
The threshold value used for inverse binary thresholding can be optimally adjusted according to the particular image. In this embodiment, the expanded image is subjected to inverse binary thresholding with a threshold 240 using a threshold () function, and the following manner is called:
threshold(out3,out3,240,255,THRESH_BINARY_INV);
threshold(out6,out6,240,255,THRESH_BINARY_INV);
and the parameters 1 are out3 and out6, namely the images after the expansion in the step 2.
The parameters 2 are out3 and out6, namely the output image after inverse binary thresholding.
Parameter 3 is 240, the binarization threshold.
Parameter 4 is 255, i.e., a value not greater than threshold 240 is set to a maximum value of 255.
The parameter 5 is thresh_binary_inv, i.e. inverse BINARY thresholding.
And step 3, performing intersection operation on the first binarized image and the second binarized image to obtain an intersection operation result image.
In this embodiment, the two images after binarization processing are interleaved by using a bitwise_and () function, and the invoking mode is as follows:
bitwise_and(out3,out6,srcImgGray);
the parameter 1 is out3, i.e. the binarized image of step 3.
And the parameter 2 is out6, namely the binarized image in the step 3.
And the parameter 3 is srcImgGray, namely an image after image intersection operation.
And 4, performing open operation on the intersection operation result image to remove white noise of the image, and obtaining a denoising image.
In this embodiment, the image white noise is removed by the on operation using the morphogex () function, and the calling method is as follows:
morphologyEx(srcImgGray,srcImgGray,MORPH_OPEN,kernel3,Point(-1,-1),2);
and the parameter 1 is srcImgGray, namely the image after the intersection operation in the step 4.
Parameter 2 is srcImgGray, i.e. the image after white noise removal.
Parameter 3 is MORPH_OPEN, the immediate-on operation.
Parameter 4 is kernel3, i.e. the custom kernel with (3, 3) as the kernel matrix.
The parameter 5 is Point (-1, -1), i.e. the anchor Point of the custom kernel, as a default value.
The parameter 6 is a value of 2, i.e. the number of iterations of the open operation.
And 5, extracting the contour in the denoising image by using a topological structure analysis method followed by the boundary to obtain a contour set in the denoising image.
In this embodiment, the contour in the denoising image is extracted by using the topology structure analysis method followed by the boundary, and the calling method is as follows:
vector<vector<cv::Point>>contours;
findContours(srcImgGray,contours,RETR_LIST,CHAIN_APPROX_SIMPLE);
and the parameter 1 is srcImgGray, namely the denoising image after white noise is removed in the step 4.
The parameter 2 is contours, i.e. a set of contours extracted from the de-noised image.
Parameter 3 is RETR_LIST, i.e. all contours are retrieved without establishing any hierarchical relationship.
Parameter 4 is CHAIN_APPROX_SIMPLE, i.e., compressing horizontal, vertical, and diagonal segments, leaving only their endpoints.
And 6, acquiring the circle center and the radius of the minimum bounding circle of each contour in the contour set, so as to form the minimum bounding circle set.
In this embodiment, the minEnclosingCircle () function is used to extract the minimum closed circle data in the contour, that is, the minimum bounding circle of the contour is extracted, and the following calling method is adopted:
minEnclosingCircle(contours[i],center,radius);
parameter 1 is contours [ i ], i.e., the ith contour in the set of contours.
The parameter 2 is center, i.e. the circular center of the smallest bounding circle.
The parameter 3 is radius, i.e. the radius of the circle of smallest bounding circle.
Through the above operation, the minimum closed circle of the outline, i.e., the minimum bounding circle, has been extracted from the image, but because the cell plate surface structure is complex, more abnormal data still occurs. As shown in fig. 3, a series of circles are extracted from the cell plate image in fig. 2, but part of the circles belongs to the circular structural members between the test tube holes, and a plurality of overlapped minimum bounding circles also occur due to a certain thickness of part of the test tube holes or the circular structural members or the quality problem of the image. The method is characterized in that the method belongs to abnormal data, and the abnormal data are required to be removed, wherein the rule of removing the abnormal data with larger or smaller radius is firstly based on the radius of a circle, and then the abnormal data with intersection are removed by utilizing an intersection algorithm, and the following detailed description is given below:
step 7, pressing all the minimum bounding circles in the minimum bounding circle setSorting according to radius, selecting a group of continuous minimum bounding circles with radius deviation not exceeding a set threshold value from the sorting sequence as representative of the outer contour of the test tube, and obtaining the average radius R of the group of continuous minimum bounding circles 1 The method comprises the steps of carrying out a first treatment on the surface of the At the average radius R 1 And setting a radius screening range which accords with the radius deviation amount of the test tube for the central value, and screening the minimum bounding circle to be selected with the radius within the range from the minimum bounding circle set.
In this embodiment, the circle anomaly data with a larger or smaller radius is removed based on the circle radius, and the specific algorithm for removing the circle anomaly data is shown in fig. 4, and specifically includes the following steps:
A. and (3) inputting the minimum closed circle data output in the step (6), namely the circle center and radius data of all the minimum closed circles in the minimum closed circle set.
B. The smallest bounding circles are ordered in ascending order according to the radius of the circle.
C. Selecting continuous x continuous minimum bounding circles from the ordered minimum bounding circle sequence as representative of the outer contour of the test tube, and obtaining the average radius R of the x circles 1 . Note that the x consecutive minimum bounding circles are not arbitrarily selected, but need to be ensured that their radius deviation does not exceed a set threshold. Since the dimensions of the remaining circular structures are generally much smaller than the dimensions of the tube wells, as is known from the characteristics of the cell plates, all tube wells are substantially constant in size, with at most variations in processing accuracy. Therefore, by setting a reasonable x value and a radius deviation threshold value, the extracted x continuous minimum bounding circles can be ensured to be the outer contour of the test tube. In this embodiment, x is set to 3, and the radius deviation threshold is set to 10%, i.e., the radius deviation of the first and last circles in the consecutive minimum bounding circle sequence must not exceed 10%.
D. At an average radius R 1 Setting a certain acceptable deviation for the center value, namely setting the minimum value and the maximum value of the target radius (the minimum radius is 0.8 average radius, the maximum value is 1.2 average radius), namely the radius screening range is [0.8R 1 ,1.2R 1 ]。
E. And screening the minimum bounding circles to be selected, of which the radius is within the range, from the minimum bounding circle set, regarding the minimum bounding circles as conventional circles, and outputting conventional circle data so as to facilitate subsequent intersection judgment.
Step 8, calculating the average radius R of all the minimum bounding circles to be selected 2 Then judging whether an intersection exists between the minimum bounding circles to be selected in pairs, if so, keeping the radius closest to the average radius R 2 And deleting another circle at the same time, and outputting all the minimum bounding circles to be selected which do not have intersection with each other as the outer contour of the test tube.
The purpose of this step is to remove the intersecting circular anomaly data which may be caused by repeated identification of the inner and outer boundaries of a test tube hole having a certain thickness, and the specific algorithm is as follows:
A. inputting the data of the minimum bounding circle to be selected in the normal radius range output in the step 7, and calculating the average radius R 2 。
B. Excluding certain circular data with the distance between the center points of the two circles being less than the sum of the radii (average radius R of all circles in step A) 2 For reference, a circle with the radius closest to the average radius is reserved in the intersecting circles, and the other circle is deleted).
C. After all abnormal data in the circles with intersections are removed, the rest is normal data, each circle represents one outline of a test tube hole in the image, the output is shown in fig. 5, and at the moment, the abnormal circle data are deleted.
And 9, dividing the gray level image of the cell plate by utilizing all the output test tube outer outlines to obtain test tube hole areas in each test tube outer outline, and accurately dividing each test tube hole (the thick circle is formed).
Through the method, a person using the cell plate to culture cloned cells can record cell plate images at different times through an industrial camera, then each test tube hole area is automatically segmented, and the cell group culture condition in each test tube hole can be checked directly through the test tube hole images without detecting the cell plate in real time. In addition, the image data of the single test tube hole selected in each time period in the cell culture process can be recorded by the method, so that workers can manually screen the cell population in the test tube hole, and the method has traceability. The cell population can be segmented and detected again by further matching with other software, so that a worker can perform semi-intelligent or even full-intelligent screening on the cell population in the test tube hole.
The above embodiment is only a preferred embodiment of the present invention, but it is not intended to limit the present invention. Various changes and modifications may be made by one of ordinary skill in the pertinent art without departing from the spirit and scope of the present invention. Therefore, all the technical schemes obtained by adopting the equivalent substitution or equivalent transformation are within the protection scope of the invention.
Claims (10)
1. A method for segmenting a test tube hole in a cell plate image, comprising the steps of:
s1, respectively performing image expansion on a cell plate gray level image to be segmented by using kernel matrixes with different sizes to obtain a first expanded image and a second expanded image;
s2, respectively carrying out binarization processing on the first expanded image and the second expanded image to obtain a first binarized image and a second binarized image;
s3, performing intersection operation on the first binarization image and the second binarization image to obtain an intersection operation result image;
s4, performing open operation on the intersection operation result image to remove white noise of the image, and obtaining a denoising image;
s5, extracting contours in the denoising image by using a topological structure analysis method followed by boundaries to obtain a contour set in the denoising image;
s6, aiming at each contour in the contour set, acquiring the circle center and the radius of the minimum bounding circle, so as to form the minimum bounding circle set;
s7, sorting all the minimum bounding circles in the minimum bounding circle set according to the radius, selecting a group of continuous minimum bounding circles with the radius deviation not exceeding a set threshold value from the sorting sequence as the representation of the outer contour of the test tube, and obtainingAverage radius R of the set of consecutive minimum bounding circles 1 The method comprises the steps of carrying out a first treatment on the surface of the At the average radius R 1 Setting a radius screening range which accords with the radius deviation amount of the test tube for the central value, and screening a minimum bounding circle to be selected, the radius of which is within the range, from the minimum bounding circle set;
s8, calculating the average radius R of all the minimum bounding circles to be selected 2 Then judging whether an intersection exists between the minimum bounding circles to be selected in pairs, if so, keeping the radius closest to the average radius R 2 Simultaneously deleting another circle, and outputting all the smallest circle to be selected, which does not have intersection, as the outer contour of the test tube;
s9, dividing the gray level image of the cell plate by utilizing all the output test tube outer outlines to obtain test tube hole areas in each test tube outer outline.
2. The method for segmenting a test tube hole in a cell plate image according to claim 1, wherein each of S1 to S7 is implemented based on an open source software library Opencv.
3. The method for segmenting a sample cell hole in a cell plate image according to claim 2, wherein in S1, image expansion is performed on a cell plate gray scale image to be segmented by using a dialite () function as kernel matrices of (3, 3) and (8, 8) respectively, to obtain a first expanded image and a second expanded image, respectively.
4. The method for segmenting the test tube hole in the cell plate image according to claim 2, wherein in S2, the first and second images after expansion are inverse binary thresholded with a threshold of 240 using a threshold () function.
5. The method of dividing a sample aperture in a cell plate image according to claim 2, wherein in S3, the first binarized image and the second binarized image are interleaved with a bitwise_and () function.
6. The method for segmenting the sample cell hole in the cell plate image according to claim 2, wherein in S4, the cross-operation result image is subjected to an open operation by using a morphogex () function to remove white noise of the image.
7. The method for segmenting a test tube hole in a cell plate image according to claim 2, wherein in S5, the following call method is used to extract the contour in the denoised image by using the topology analysis method followed by the boundary:
vector<vector<cv::Point>>contours;
findContours(srcImgGray,contours,RETR_LIST,CHAIN_APPROX_SIMPLE);
where srcImgGray is the de-noised image, concurs is the set of contours in the de-noised image, RETR_LIST means that all contours are retrieved without establishing any hierarchical relationship, CHAIN_APPROX_SIMPLE means that the horizontal, vertical and diagonal segments are compressed, and only the endpoints thereof are preserved.
8. The method of segmenting a test tube aperture in a cell plate image according to claim 2, wherein in S6, the center and radius of the smallest bounding circle is obtained for each contour in the set of contours with a minEnclosingCircle () function.
9. The method for segmenting a tube hole in a cell plate image according to claim 1, wherein in S7, when a group of consecutive minimum bounding circles is selected from the ordered sequence as a representative of the outer contour of the tube, a set threshold of 10% is used, that is, the deviation between the maximum radius and the minimum radius in the selected consecutive minimum bounding circles is not more than 10%; the radius selection range is [0.8R 1 ,1.2R 1 ]。
10. The method for segmenting a test tube hole in a cell plate image according to claim 1, wherein in S8, whether there is an intersection between the two minimum bounding circles is determined by calculating whether the center distances of the two minimum bounding circles to be selected are smaller than the sum of the radii thereof.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110557045.9A CN113393479B (en) | 2021-05-21 | 2021-05-21 | Method for dividing test tube holes in cell plate image |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110557045.9A CN113393479B (en) | 2021-05-21 | 2021-05-21 | Method for dividing test tube holes in cell plate image |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113393479A CN113393479A (en) | 2021-09-14 |
CN113393479B true CN113393479B (en) | 2024-02-02 |
Family
ID=77618938
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110557045.9A Active CN113393479B (en) | 2021-05-21 | 2021-05-21 | Method for dividing test tube holes in cell plate image |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113393479B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113869223A (en) * | 2021-09-28 | 2021-12-31 | 广东三维家信息科技有限公司 | Pattern recognition method, pattern recognition template determination method and device |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107481225A (en) * | 2017-07-26 | 2017-12-15 | 山东颐泽天泰医疗科技有限公司 | A kind of method of Automatic-searching optimized parameter segmentation adhesion cells |
CN109671052A (en) * | 2018-11-16 | 2019-04-23 | 华南理工大学 | A kind of mistake hole inspection method and hole inspection of the flexible IC package substrate extracted based on circular contour |
-
2021
- 2021-05-21 CN CN202110557045.9A patent/CN113393479B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107481225A (en) * | 2017-07-26 | 2017-12-15 | 山东颐泽天泰医疗科技有限公司 | A kind of method of Automatic-searching optimized parameter segmentation adhesion cells |
CN109671052A (en) * | 2018-11-16 | 2019-04-23 | 华南理工大学 | A kind of mistake hole inspection method and hole inspection of the flexible IC package substrate extracted based on circular contour |
Also Published As
Publication number | Publication date |
---|---|
CN113393479A (en) | 2021-09-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108596166B (en) | Container number identification method based on convolutional neural network classification | |
CN113450307B (en) | Product edge defect detection method | |
CN107967475A (en) | A kind of method for recognizing verification code based on window sliding and convolutional neural networks | |
CN108376403B (en) | Grid colony image segmentation method based on Hough circle transformation | |
CN106934795A (en) | The automatic testing method and Forecasting Methodology of a kind of glue into concrete beam cracks | |
CN113469951B (en) | Hub defect detection method based on cascade region convolutional neural network | |
CN114820625B (en) | Automobile top block defect detection method | |
CN113506286A (en) | Microwave chip defect detection method based on small sample data set of YOLOv5 algorithm | |
CN111754538B (en) | Threshold segmentation method for USB surface defect detection | |
JP4581090B2 (en) | Method for automatically drawing a portion of interest by image processing, apparatus therefor, and recording medium recording a program | |
CN109815870B (en) | High-throughput functional gene screening method and system for quantitative analysis of cell phenotype image | |
CN111369526B (en) | Multi-type old bridge crack identification method based on semi-supervised deep learning | |
CN110349125A (en) | A kind of LED chip open defect detection method and system based on machine vision | |
CN114782330B (en) | Grate abnormity detection method and system based on artificial intelligence | |
CN114495098B (en) | Diaxing algae cell statistical method and system based on microscope image | |
CN113012124B (en) | Shoe print hole and embedded object feature detection and description method | |
CN115797361B (en) | Aluminum template surface defect detection method | |
CN112288760B (en) | Adherent cell image screening method and system and cell image analysis method | |
CN117094975A (en) | Method and device for detecting surface defects of steel and electronic equipment | |
CN110458843A (en) | The dividing method and system of mask images | |
CN113393479B (en) | Method for dividing test tube holes in cell plate image | |
CN110349119B (en) | Pavement disease detection method and device based on edge detection neural network | |
CN115063603B (en) | Wood annual ring line accurate extraction and restoration method based on edge information | |
CN114384073B (en) | Subway tunnel crack detection method and system | |
CN112465817B (en) | Pavement crack detection method based on directional filter |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |