CN108376403B - Grid colony image segmentation method based on Hough circle transformation - Google Patents
Grid colony image segmentation method based on Hough circle transformation Download PDFInfo
- Publication number
- CN108376403B CN108376403B CN201810086989.0A CN201810086989A CN108376403B CN 108376403 B CN108376403 B CN 108376403B CN 201810086989 A CN201810086989 A CN 201810086989A CN 108376403 B CN108376403 B CN 108376403B
- Authority
- CN
- China
- Prior art keywords
- colony
- image
- grid
- target
- gray
- 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
- 238000000034 method Methods 0.000 title claims abstract description 29
- 230000009466 transformation Effects 0.000 title claims abstract description 26
- 238000003709 image segmentation Methods 0.000 title claims abstract description 19
- 230000011218 segmentation Effects 0.000 claims abstract description 17
- 239000003550 marker Substances 0.000 claims abstract description 16
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 230000001580 bacterial effect Effects 0.000 claims description 47
- 239000002609 medium Substances 0.000 claims description 26
- 238000001514 detection method Methods 0.000 claims description 11
- 101100269850 Caenorhabditis elegans mask-1 gene Proteins 0.000 claims description 9
- 238000001914 filtration Methods 0.000 claims description 5
- 239000001963 growth medium Substances 0.000 claims description 5
- 238000007619 statistical method Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 2
- 238000004088 simulation Methods 0.000 description 15
- 238000004458 analytical method Methods 0.000 description 4
- 241000894006 Bacteria Species 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 230000001186 cumulative effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
Images
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
- G06T7/11—Region-based segmentation
-
- 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
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/168—Segmentation; Edge detection involving transform domain methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20048—Transform domain processing
- G06T2207/20061—Hough transform
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
- G06T2207/20152—Watershed segmentation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
- Apparatus Associated With Microorganisms And Enzymes (AREA)
Abstract
The invention provides a grid colony image segmentation method based on Hough circle transformation, which is used for solving the technical problem that the grid colony image cannot be segmented in the prior art and comprises the following steps: preprocessing the grid colony image; acquiring a binaryzation boundary image of a colony target in a denoised grid colony gray level image; acquiring the average radius r and the average gray value g of a colony target in a denoised grid colony gray image by adopting Hough circle transformation; obtaining a candidate colony target marker image in the denoised grid colony gray level image by adopting Hough circle transformation; and acquiring a segmentation result of the colony image input into the grid. The invention can accurately segment the grid colony image and can be used for detecting, counting and classifying colony targets in the grid colony image.
Description
Technical Field
The invention belongs to the technical field of image processing, relates to a grid colony image segmentation method, and particularly relates to a grid colony image segmentation method based on Hough circle transformation, which can be used for detecting, counting and classifying colony targets in a grid colony image.
Background
The colony is a single bacterium group which grows on the surface of a solid culture medium and can be seen by naked eyes after the bacterium is inoculated on the surface of the solid culture medium and cultured. Counting the number of colonies generated after target sampling is a basic and important task for quality detection in agriculture, food, medicine and health analysis. The colony image is an image formed by shooting the colony with an industrial camera after the colony is cultured on a culture medium, so that the colony target information is counted conveniently. In order to improve the accuracy of analysis and counting of colony targets, analysts often culture bacteria in a grid background, thereby forming a grid colony image. However, the manual observation has the disadvantages of complicated procedures, long time consumption, low efficiency, subjectivity, large error, poor reproducibility and the like, so that the acquisition difficulty and the accuracy of the colony target information under the grid background are high. The image processing and analyzing method can relieve operators from the heavy work and greatly improve the counting and analyzing precision. Grid colony image segmentation is a technology for separating colony objects from an original image through analysis of the object image, so as to obtain information of the colony objects from the original image, and further analyze and process the colony objects.
Hough circle transform belongs to a special case of Hough transform and is mainly used for detecting a circular target. The Hough circle transformation finds the coordinates of the circle center by using partial boundary points of the circular target, so that the whole circular boundary is restored. The basic idea of Hough circle transformation is that each non-zero pixel point on an image is considered to be a potential point on a circle, a cumulative coordinate plane is generated through voting, and a cumulative weight is set to position the circle. The standard Hough circle transformation combines a Cartesian coordinate system and a three-dimensional coordinate system, and determines whether circles in the two-dimensional coordinate system corresponding to the three-dimensional coordinate system are reserved or not by judging whether the number of intersections of each point in the three-dimensional coordinate system is greater than a certain threshold value or not based on the principle that all circles passing through a certain point in the Cartesian coordinate system are mapped to a three-dimensional curve in the three-dimensional coordinate system, and the three-dimensional curve is used as a final circle fitting result. To improve the computational efficiency, the improved hough circle transformation is directly processed under a two-dimensional coordinate system, for example: and taking all boundary points as circle centers, drawing circles by taking the minimum radius and the maximum radius in the input parameters as radii, recording the number of intersection points at corresponding pixel points in the image, recording pixel points with the number larger than the minimum required point in the input parameters as the central point of a fitting circle, and taking the radius of the same large and small circle with the largest number of intersection points as the radius of the fitting circle, thereby fitting the circle target. Since the colony target is generally a circular target, a part of the colony target can be found by adopting Hough circle transformation, so that the related characteristics of the colony target can be obtained.
The related colony image segmentation technology is to segment non-grid colony images, and can not process grid colony images, so that no document and technology for specially segmenting the grid colony images are found at present.
Disclosure of Invention
The invention aims to provide a grid colony image segmentation method based on Hough circle transformation aiming at a more special grid colony image, and the method is used for solving the technical problem that the grid colony image cannot be segmented in the prior art.
In order to achieve the purpose, the technical scheme adopted by the invention comprises the following steps:
(1) preprocessing the grid colony image:
(1.1) converting the input grid colony color image with the size of m multiplied by n pixels into a grid colony gray image I1M is more than or equal to 300 and n is more than or equal to 300;
(1.2) recalculating grid colony gray level image I by using gray level histogram statistical method1Background value m ofbAnd using the calculated background value to obtain a grid colony gray image I1The background is set to obtain a grid bacterial colony gray image I of the reset background2；
(1.3) Gray image of grid colony I2Carrying out median filtering to obtain a denoised grid bacterial colony gray image I3；
(2) Obtaining a denoised grid bacterial colony gray image I3Binary boundary image I of medium colony target5：
(2.1) grid colony gray level image I after denoising3And (3) carrying out boundary detection:
grid colony gray level image I after denoising by adopting boundary detection algorithm3Carrying out boundary detection to obtain a grid bacterial colony binaryzation boundary image I4；
(2.2) removing the binary boundary of the grid bacterial colonyImage I4The grid boundary line of (a):
calculating grid bacterial colony binaryzation boundary image I4The length and width of each boundary and the length and width of the minimum bounding rectangle, and excluding I4The middle length is greater than a preset parameter a1Or the aspect ratio of the minimum bounding rectangle is larger than the preset parameter a2Obtaining a denoised grid colony gray image I3Binary boundary image I of medium colony target5；
(3) Obtaining a denoised grid colony gray level image I by adopting Hough circle transformation3Mean radius r and mean gray value g of middle colony target:
(3.1) Gray-level image of grid colony I3Binary boundary image I of medium colony target5Carrying out circle fitting:
setting a circle fitting judgment parameter b1And a binary boundary image I of the colony target is transformed by adopting Hough circle5Performing circle fitting to obtain a binary boundary image I containing part of colony targets6；
(3.2) Using the binarized boundary image I containing part of the colony target6Calculating the denoised grid colony gray level image I3The average radius r and the average gray value g of the medium colony target;
(4) obtaining a denoised grid colony gray level image I by adopting Hough circle transformation3Candidate colony target marker image mask in (1):
(4.1) binarized boundary image I of colony target5Carrying out Hough circle transformation to obtain I3The first type candidate target mark image mask 1;
(4.2) use of I3Average gray value g of middle colony target is obtained3The second type candidate target mark image mask 2;
(4.3) adding the first type of candidate target marker image mask1 and the second type of candidate target marker image mask2 to obtain a de-noised grid bacterial colony gray level image I3The candidate colony target marker image mask in (1);
(5) acquiring a segmentation result of an input grid colony image:
(5.1) marking the colony target image mask and grid colony gray level image I1Background value m ofbPerforming certain mathematical operation to obtain a colony target gray level image I with the grid line background removed12；
(5.2) acquiring a denoised grid bacterial colony gray image I3Initial binary image I of medium colony target13：
To the colony target gray image I excluding the grid line background12Thresholding is carried out to obtain a denoised grid colony gray image I3Initial binary image I of medium colony target13；
(5.3) on the initial binary image I13The adhesion target in (1) is divided:
adopting an adhesion segmentation algorithm to perform initial binarization on the image I13Processing the adhered target to obtain a denoised grid bacterial colony gray image I3Binary image I of medium colony target14And is combined with14As a result of segmentation of the input grid colony image.
Compared with the prior art, the invention has the following advantages:
the method adopts Hough circle transformation to obtain a denoised grid colony gray level image I3And constructing a new gray map by using the obtained average radius and average gray value of the medium colony target and the candidate colony target marker image through certain mathematical operation, and segmenting the newly constructed gray map by adopting a threshold segmentation algorithm, thereby realizing the segmentation of the grid colony image.
Drawings
FIG. 1 is a flow chart of an implementation of the present invention;
FIG. 2 is a simulation of grid colony image segmentation with a large number of colony targets according to the present invention;
FIG. 3 is a simulation of grid colony image segmentation with fuzzy boundary colony targets according to the present invention.
Detailed description of the invention
The invention is described in further detail below with reference to the figures and specific examples.
Referring to fig. 1, a mesh colony image segmentation method based on hough circle transformation includes the following steps:
step 1) preprocessing a grid colony image:
step 1.1) converting the input grid colony color image with the size of m multiplied by n pixels into a grid colony gray image I1M is more than or equal to 300 and n is more than or equal to 300; in this embodiment, m is 1280, n is 960;
step 1.2) recalculating grid colony gray level image I by using gray level histogram statistical method1Background value m ofbAnd using the calculated background value to obtain a grid colony gray image I1The background is set to obtain a grid bacterial colony gray image I of the reset background2The method comprises the following implementation steps:
since the culture dish is generally circular or rectangular, the colony target is generally concentrated in a circular or rectangular area in the middle of the image, however, the color change of the edge of the culture dish and the external background thereof is random, and the influence on the colony target segmentation operation is great. To eliminate interference of culture dish edge, grid colony gray image I1In the method, an effective operation area with any shape is defined at any position without the culture dish boundary, the defined shape is usually a circle or a rectangle, the defined shape is a circle, the effective operation area refers to an area containing a colony target, then, a gray histogram of the effective operation area is counted, and a gray value corresponding to a peak value in the gray histogram is used as a grid colony gray image I1Background value m ofbThen resetting the grid colony gray level image I1The gray value of all pixels outside the middle operation region is the background value mbObtaining a grid bacterial colony gray image I with the background reset2；
Step 1.3) Gray level image I of grid bacterial colony2Carrying out median filtering to obtain a denoised grid bacterial colony gray image I3：
In order to eliminate the interference of the fine Gaussian noise, an 8-neighborhood median filtering method is adopted for filtering.
Step 2) obtaining a denoised grid bacterial colony gray image I3Binary boundary image I of medium colony target5：
Step 2.1) of denoising grid bacterial colony gray image I3And (3) carrying out boundary detection:
this embodiment employs CannyGrid colony gray level image I by boundary detection algorithm3Carrying out boundary detection to obtain a grid bacterial colony binaryzation boundary image I4Including colony targets and boundaries of grid lines;
step 2.2) removing the binaryzation boundary image I of the grid bacterial colony4The grid boundary line of (a):
grid bacterial colony binaryzation boundary image I4The result in (1) comprises a grid boundary and a colony target boundary, and in order to improve the accuracy of acquiring the characteristics of the colony target, the interference of the grid line is eliminated by utilizing the characteristics that the grid line is long and narrow and the colony target boundary is similar to a circle. For this purpose calculate I4The length and width of each boundary and the length and width of the minimum bounding rectangle, and excluding I4The middle length is greater than a preset parameter a1Or the aspect ratio of the minimum bounding rectangle is larger than the preset parameter a2To obtain grid colony gray image I3Binary boundary image I of medium colony target5. Wherein, a1Depending on the width w of a single grid in the image, a is typically taken1In the actual grid colony image, the width of a single grid is in direct proportion to the size of the grid colony image, and the grid width w is assumed to be large in the invention
And is provided with
a2In this example, a is 31＝55。
Step 3) obtaining a denoised grid bacterial colony gray image I by adopting Hough circle transformation3Mean radius r and mean gray value g of middle colony target:
using colony targetsCircle-like feature, binary boundary image of colony target5Circle fitting operation is carried out by adopting Hough circle transformation, higher circle fitting conditions are set, namely circle fitting is carried out when more boundary points belong to the same circle, so that the positions of part of colony targets are accurately determined, and various characteristics of the colony targets are obtained. The colony target characteristics comprise morphology, size, gray value, texture and the like, and proper characteristic information is extracted aiming at different colony targets. The method comprises the following specific steps:
step 3.1) Gray image I of grid bacterial colony3Binary boundary image I of medium colony target5Carrying out circle fitting:
in order to improve the fitting accuracy, a set circle fitting judgment parameter b1It should be larger, b in this embodiment1The simulation result proves that the Hough circle transformation under the parameter can accurately obtain part of colony targets, and the Hough circle transformation is adopted to carry out binarization boundary image I on the colony targets5Performing circle fitting to obtain a binary boundary image I containing part of colony targets6；
Step 3.2) utilizing a binary boundary image I containing part of colony targets6Calculating the denoised grid colony gray level image I3The average radius r and the average gray value g of the medium colony target comprise the following specific steps:
step 3.2.1) of binarizing the boundary image I containing part of the colony targets6Filling the binary boundary to obtain a binary image I containing part of colony target7；
Step 3.2.2) extraction of I7Coordinates of each non-zero connected region in the image and counting I3Calculating to obtain the average value of the size and the gray value of the colony target at the corresponding coordinate, and obtaining a de-noised grid colony gray image I3Average radius r and average gray value g of the middle colony target.
Step 4) obtaining a denoised grid bacterial colony gray image I by adopting Hough circle transformation3Candidate colony target marker in (1)Image mask:
step 4.1) of binarizing boundary image I of colony target5Carrying out Hough circle transformation to obtain I3The first type candidate target mark image mask1 comprises the following specific steps:
step 4.1.1) to prevent missing detection of the circular colony target, a lower circle fitting judgment parameter b is set2In this embodiment, b2A large number of simulation experiments prove that most of similar circular colony targets can be obtained by Hough circle transformation under the parameters, fewer missing colony targets are ensured, and then a binary boundary image I of the colony targets is subjected to Hough circle transformation5Performing circle fitting to obtain a circle fitting result image I8；
Step 4.1.2) fitting result image I to circle8Filling a binaryzation boundary, namely representing a candidate target by a connected region formed by a circular boundary and an internal region thereof to obtain a binaryzation circular colony target image I9；
Step 4.1.3) obtaining a binaryzation circular colony target image I9And calculating the coordinate of each circular colony target, and calculating the denoised gray level image I3Taking the average gray value of the circular target at the corresponding coordinate as the characteristic value of the circular target;
step 4.1.4) removing the binaryzation circular colony target image I9The difference between the characteristic value and the average gray value g of the colony target is more than a preset parameter a3A round target, this example takes a3Simulation experiments show that the parameters enable the obtained first-class candidate target to be accurate, and a de-noised grid bacterial colony gray image I is obtained3The first type candidate object mark image mask1, wherein 1 represents a candidate object and 0 represents a background;
step 4.2) Using I3Average gray value g of middle colony target is obtained3The second type candidate target mark image mask2 comprises the following specific steps:
step 4.2.1) traversing denoised grid bacterial colony gray image I3The difference between the gray value and the average gray value g is smaller than a preset parameter a3The pixel of (2) is used as a candidate target pixel to obtain a binaryzation colony target image I10；
Step 4.2.2) of binarizing the target image I of the bacterial colony10Filling a binarization target to obtain a binarization colony target image I11；
Step 4.2.3) calculating a binaryzation colony target image I11The radius and roundness of each colony target in the culture medium are eliminated by I11The difference between the medium radius and the average radius is larger than a preset parameter a4In order to further eliminate the interference of partial impurities, the roundness C of the colony target is calculated, and the roundness C smaller than a is eliminated5The colony object of (1), this example takes a4＝5，a5Experiments show that the two parameters can effectively eliminate the candidate colony target which is detected wrongly and obtain a de-noised grid colony gray image I3The second type of candidate object marker image mask2, where 1 represents a candidate object, 0 represents a background, and circularity is defined as:
C＝4πS2/L2s represents the area of a colony target, L represents the length of a boundary line of the colony target, the roundness reflects the approaching degree of the colony target and a circle, and the closer to 1 represents the closer to the circle;
step 4.3) adding the first type of candidate target marker image mask1 and the second type of candidate target marker image mask2 to obtain a de-noised grid colony gray level image I3The candidate colony target marker image mask in (1).
Step 5) obtaining a segmentation result of the colony image of the input grid:
step 5.1) marking the colony target image mask and the grid colony gray level image I1Background value m ofbUsing the formula: i is12＝I3×mask+mbCalculating by x (1-mask) to obtain a colony target gray level image I with the grid line background excluded12；
Step 5.2) obtaining a denoised grid bacterial colony gray image I3Initial binary image I of medium colony target13：
For bacterial colony target with the background of grid lines eliminatedGrayscale image I12Ostu thresholding segmentation is carried out to obtain a denoised grid colony gray image I3Initial binary image I of medium colony target13；
Step 5.3) for the initial binary image I13The adhesion target in (1) is divided:
adopting a watershed segmentation algorithm based on seed points to carry out initial binarization on the image I13Processing the adhered target to obtain a denoised grid bacterial colony gray image I3Binary image I of medium colony target14And is combined with14As a result of segmentation of the input grid colony image.
The technical effects of the present invention will be further explained below by combining with simulation experiments.
1. Simulation content and conditions:
to verify the effectiveness and correctness of the invention, two types of grid colony images are used for simulation experiments. Simulation I, namely an experiment for segmenting a grid bacterial colony image containing a large number of bacterial colony targets by using the method; simulation two is an experiment for segmenting a grid colony image containing a fuzzy edge colony target by using the method. All simulation experiments are realized by adopting VS2015 software + OpenCv library under a Windows 7 operating system.
2. And (3) simulation result analysis:
simulation one results as shown in fig. 2 (e).
With reference to figure 2 of the drawings,
FIG. 2(a) is a grayscale image of an input grid colony image;
fig. 2(b) is a grid colony gray image after preprocessing the input grid colony image, and it can be seen from fig. 2(b) that the interference of uneven gray values at the boundary of the culture dish is eliminated after the input grid colony gray image is reset with the background, and simultaneously the gray values of the pixels except the grid lines and the colony targets tend to be consistent;
FIG. 2(c) is a binarized boundary image of a colony target in a grid colony gray scale image after preprocessing, and it can be seen from FIG. 2(c) that grid lines have been substantially excluded;
FIG. 2(d) is a diagram showing candidate targets obtained by Hough circle transformation on a preprocessed grid colony gray image, wherein white connected regions represent candidate colony targets, and it can be seen from FIG. 2(d) that the candidate colony targets basically cover all colony targets, but are not all true colony targets;
fig. 2(e) is the final segmentation result of the present invention, wherein the white connected regions represent colony targets, and it can be seen from fig. 2(e) that the false targets in the candidate targets have been eliminated, and the obtained grid colony image segmentation result is more accurate.
The results from simulation two are shown in fig. 3 (e).
With reference to figure 3 of the drawings,
FIG. 3(a) is a grayscale image of an input grid colony image;
fig. 3(b) is a grid colony gray image after preprocessing the input grid colony image, and it can be seen from fig. 3(b) that the interference of uneven gray values at the boundary of the culture dish is eliminated after the input grid colony gray image is reset with the background, and simultaneously the gray values of the pixels except the grid lines and the colony targets tend to be consistent;
fig. 3(c) is a binarized boundary image of a colony target in the preprocessed grid colony grayscale image, and it can be seen from fig. 3(c) that grid lines have been substantially excluded, and although the edge of the colony target is blurred, partial boundaries of most colony targets are still detected;
FIG. 3(d) is a diagram showing candidate targets obtained by Hough circle transformation on a preprocessed grid colony gray image, wherein white connected regions represent candidate colony targets, and it can be seen from FIG. 3(d) that the candidate colony targets substantially cover all colony targets, but are not all true colony targets;
fig. 3(e) is the final segmentation result of the present invention, wherein the white connected regions represent colony targets, and it can be seen from fig. 3(e) that the false targets in the candidate targets have been eliminated, and the obtained grid colony image segmentation result is more accurate.
As can be seen from the simulation experiments, the method can obtain accurate segmentation results no matter whether the grid bacterial colony image containing a large number of bacterial colony targets is segmented or the grid bacterial colony image with fuzzy target boundaries is segmented.
Claims (6)
1. A grid colony image segmentation method based on Hough circle transformation is characterized by comprising the following steps:
(1) preprocessing the grid colony image:
(1.1) converting the input grid colony color image with the size of m multiplied by n pixels into a grid colony gray image I1M is more than or equal to 300 and n is more than or equal to 300;
(1.2) recalculating grid colony gray level image I by using gray level histogram statistical method1Background value m ofbAnd using the calculated background value to obtain a grid colony gray image I1The background is set to obtain a grid bacterial colony gray image I of the reset background2；
(1.3) Gray image of grid colony I2Carrying out median filtering to obtain a denoised grid bacterial colony gray image I3；
(2) Obtaining a denoised grid bacterial colony gray image I3Binary boundary image I of medium colony target5：
(2.1) grid colony gray level image I after denoising3And (3) carrying out boundary detection:
grid colony gray level image I after denoising by adopting boundary detection algorithm3Carrying out boundary detection to obtain a grid bacterial colony binaryzation boundary image I4；
(2.2) removing grid bacterial colony binaryzation boundary image I4The grid boundary line of (a):
calculating grid bacterial colony binaryzation boundary image I4The length and width of each boundary and the length and width of the minimum bounding rectangle, and excluding I4The middle length is greater than a preset parameter a1Or the aspect ratio of the minimum bounding rectangle is larger than the preset parameter a2Obtaining a denoised grid colony gray image I3Binary boundary image I of medium colony target5；
(3) MiningObtaining a denoised grid colony gray image I by Hough circle transformation3Mean radius r and mean gray value g of middle colony target:
(3.1) Gray-level image of grid colony I3Binary boundary image I of medium colony target5Carrying out circle fitting:
setting a circle fitting judgment parameter b1And a binary boundary image I of the colony target is transformed by adopting Hough circle5Performing circle fitting to obtain a binary boundary image I containing part of colony targets6；
(3.2) Using the binarized boundary image I containing part of the colony target6Calculating the denoised grid colony gray level image I3The average radius r and the average gray value g of the medium colony target;
(4) obtaining a denoised grid colony gray level image I by adopting Hough circle transformation3Candidate colony target marker image mask in (1):
(4.1) binarized boundary image I of colony target5Carrying out Hough circle transformation to obtain I3The first type candidate target mark image mask 1;
(4.2) use of I3Average gray value g of middle colony target is obtained3The second type candidate target mark image mask 2;
(4.3) adding the first type of candidate target marker image mask1 and the second type of candidate target marker image mask2 to obtain a de-noised grid bacterial colony gray level image I3The candidate colony target marker image mask in (1);
(5) acquiring a segmentation result of an input grid colony image:
(5.1) marking the colony target image mask and grid colony gray level image I1Background value m ofbPerforming certain mathematical operation to obtain a colony target gray level image I with the grid line background removed12；
(5.2) acquiring a denoised grid bacterial colony gray image I3Initial binary image I of medium colony target13：
To the colony target gray image I excluding the grid line background12Performing thresholdingObtaining a denoised grid bacterial colony gray image I3Initial binary image I of medium colony target13；
(5.3) on the initial binary image I13The adhesion target in (1) is divided:
adopting an adhesion segmentation algorithm to perform initial binarization on the image I13Processing the adhered target to obtain a denoised grid bacterial colony gray image I3Binary image I of medium colony target14And is combined with14As a result of segmentation of the input grid colony image.
2. The method for grid colony image segmentation based on Hough circle transformation as claimed in claim 1, wherein the step (1.2) of recalculating the grid colony gray image I by using a gray histogram statistical method1Background value m ofbAnd using the calculated background value to obtain a grid colony gray image I1The background is set to obtain a grid bacterial colony gray image I of the reset background2The method comprises the following implementation steps:
gray scale image of colonies on grid I1In the method, an effective operation area in any shape is defined at any position without the edge of the culture dish, a gray level histogram of the effective operation area is counted, and a gray level value corresponding to a peak value in the gray level histogram is used as a grid bacterial colony gray level image I1Background value m ofbThen resetting the grid colony gray level image I1The gray value of all pixels outside the middle operation region is the background value mbObtaining a grid bacterial colony gray image I with the background reset2。
3. The method for mesh colony image segmentation based on Hough circle transform as claimed in claim 1, wherein the step (3.2) is implemented by using a binarized boundary image I containing part of colony targets6Calculating the denoised grid colony gray level image I3The average radius r and the average gray value g of the medium colony target comprise the following specific steps:
(3.2.1) binarizing edge containing part of colony targetBoundary image I6Filling the binary boundary to obtain a binary image I containing part of colony target7；
(3.2.2) extraction of I7Coordinates of each non-zero connected region in the image and counting I3The radius and the gray value of the colony target at the corresponding coordinate are calculated to obtain the average value of the radius and the gray value of the colony target, and a grid colony gray image I after denoising is obtained3Average radius r and average gray value g of the middle colony target.
4. The method for mesh colony image segmentation based on Hough circle transform as claimed in claim 1, wherein the binarized boundary image I of colony target in step (4.1)5Carrying out Hough circle transformation to obtain I3The first type candidate target mark image mask1 comprises the following specific steps:
(4.1.1) setting the circle fitting judgment parameter as b2Using Hough circle transform to binarize boundary image I of colony target5Performing circle fitting to obtain a circle fitting result image I8；
(4.1.2) fitting result to circle image I8Filling a binaryzation boundary to obtain a binaryzation circular colony target image I9；
(4.1.3) obtaining a binaryzation circular colony target image I9And calculating the coordinate of each circular colony target, and calculating the denoised gray level image I3Taking the average gray value of the circular target at the corresponding coordinate as the characteristic value of the circular target;
(4.1.4) removing the binarized circular colony target image I9The difference between the characteristic value and the average gray value g of the colony target is more than a preset parameter a3Obtaining a denoised grid colony gray image I3The first type candidate target mark image mask 1.
5. The method for mesh colony image segmentation based on Hough circle transform as claimed in claim 1, wherein step (4.2) utilizes I3Medium colony targetAverage gray value g obtaining I3The second type candidate target mark image mask2 comprises the following specific steps:
(4.2.1) traversing the denoised grid bacterial colony gray image I3The difference between the gray value and the average gray value g is smaller than a preset parameter a4The pixel of (2) is used as a candidate target pixel to obtain a binaryzation colony target image I10；
(4.2.2) on the binarized colony target image I10Filling a binarization target to obtain a binarization colony target image I11；
(4.2.3) calculating a binarized colony target image I11The radius and roundness of each colony target in the culture medium are eliminated by I11The difference between the medium radius and the average radius is larger than a preset parameter a5And removing the binarized colony target image I11The middle roundness C is less than the preset parameter a6Obtaining a denoised grid bacterial colony gray image I3The second type of candidate target marker image mask2, where circularity is defined as:
C＝4πS2/L2s represents the area of the colony target, and L represents the boundary length of the colony target.
6. The method for mesh colony image segmentation based on Hough circle transform as claimed in claim 1, wherein the step (5.1) comprises marking the colony target image mask with the mesh colony gray level image I1Background value m ofbPerforming certain mathematical operation to obtain a colony target gray level image I with the grid line background excluded12The method comprises the following specific operations:
using the formula: i is12＝I3×mask+mbX (1-mask), obtaining a colony target gray level image I with the grid line background excluded12。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810086989.0A CN108376403B (en) | 2018-01-30 | 2018-01-30 | Grid colony image segmentation method based on Hough circle transformation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810086989.0A CN108376403B (en) | 2018-01-30 | 2018-01-30 | Grid colony image segmentation method based on Hough circle transformation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108376403A CN108376403A (en) | 2018-08-07 |
| CN108376403B true CN108376403B (en) | 2020-02-07 |
Family
ID=63016979
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810086989.0A Expired - Fee Related CN108376403B (en) | 2018-01-30 | 2018-01-30 | Grid colony image segmentation method based on Hough circle transformation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108376403B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110334601A (en) * | 2019-06-04 | 2019-10-15 | 武汉极目智能技术有限公司 | A kind of speed(-)limit sign board recognition methods of combination machine learning and computer vision |
| CN111368643B (en) * | 2020-02-12 | 2023-04-14 | 杭州电子科技大学 | Escherichia coli dynamic growth monitoring method |
| CN111429361A (en) * | 2020-02-17 | 2020-07-17 | 马健锦 | Bacterial colony counting method based on machine vision |
| CN113110265A (en) * | 2021-05-26 | 2021-07-13 | 焦作大学 | Fermentation environment adjusting method for biochemical industry by adopting Internet of things |
| CN113817585A (en) * | 2021-09-10 | 2021-12-21 | 武汉理工大学 | Bacterial colony selection comprehensive workstation and selection method |
| CN114723804A (en) * | 2022-04-13 | 2022-07-08 | 盐城工学院 | Colony counting method based on image classification |
| CN116309193B (en) * | 2023-05-23 | 2023-08-25 | 武汉大学人民医院（湖北省人民医院） | Image processing method and device for immune sample |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8417011B2 (en) * | 2008-09-18 | 2013-04-09 | Molecular Devices (New Milton) Ltd. | Colony detection |
| CN103164858B (en) * | 2013-03-20 | 2015-09-09 | 浙江大学 | Adhesion crowd based on super-pixel and graph model is split and tracking |
| CN105574528B (en) * | 2015-12-15 | 2019-01-22 | 安徽工业大学 | It is a kind of that cell image segmentation method is adhered based on multiphase mutual exclusion level set |
| CN106447020B (en) * | 2016-09-13 | 2018-11-23 | 上海海事大学 | A kind of intelligence method for counting colonies |
-
2018
- 2018-01-30 CN CN201810086989.0A patent/CN108376403B/en not_active Expired - Fee Related
Also Published As
| Publication number | Publication date |
|---|---|
| CN108376403A (en) | 2018-08-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108376403B (en) | Grid colony image segmentation method based on Hough circle transformation | |
| CN107230202B (en) | Automatic identification method and system for road surface disease image | |
| CN107292310B (en) | Visual positioning and automatic reading method for circular pointer type dial plate | |
| CN109325935B (en) | Power transmission line detection method based on unmanned aerial vehicle image | |
| CN109636824B (en) | Multi-target counting method based on image recognition technology | |
| CN110210477B (en) | Digital instrument reading identification method | |
| CN112614062B (en) | Colony counting method, colony counting device and computer storage medium | |
| CN110415208A (en) | A kind of adaptive targets detection method and its device, equipment, storage medium | |
| CN112233116B (en) | Concave-convex mark visual detection method based on neighborhood decision and gray level co-occurrence matrix description | |
| CN110827235B (en) | Steel plate surface defect detection method | |
| WO2021109697A1 (en) | Character segmentation method and apparatus, and computer-readable storage medium | |
| CN109658391B (en) | Circle radius measuring method based on contour merging and convex hull fitting | |
| CN111382658B (en) | Road traffic sign detection method in natural environment based on image gray gradient consistency | |
| CN113902713B (en) | Image segmentation algorithm-based multi-colony counting method | |
| CN114596551A (en) | Vehicle-mounted forward-looking image crack detection method | |
| CN115063430A (en) | Electric pipeline crack detection method based on image processing | |
| CN107516315B (en) | Tunneling machine slag tapping monitoring method based on machine vision | |
| CN115587966A (en) | Method and system for detecting whether parts are missing or not under condition of uneven illumination | |
| CN110288619B (en) | Vision-based sunflower module surface screw hole position detection method | |
| CN113781413A (en) | Electrolytic capacitor positioning method based on Hough gradient method | |
| CN111292346B (en) | Method for detecting contour of casting box body in noise environment | |
| CN116524269A (en) | Visual recognition detection system | |
| CN113643290B (en) | Straw counting method and device based on image processing and storage medium | |
| CN112396638A (en) | Image processing method, terminal and computer readable storage medium | |
| Fu et al. | A circle detection algorithm based on mathematical morphology and chain code |
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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20200901 Address after: 400030, stone road 70, Shapingba District, Chongqing, attached to 4-15-3 Patentee after: CHONGQING JIANGXUE TECHNOLOGY Co.,Ltd. Address before: 710071, No. 2 Taibai South Road, Yanta District, Shaanxi, Xi'an Patentee before: XIDIAN University |
|
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200207 Termination date: 20210130 |