CN113570628A - Leukocyte segmentation method based on active contour model - Google Patents

Leukocyte segmentation method based on active contour model Download PDF

Info

Publication number
CN113570628A
CN113570628A CN202110871648.6A CN202110871648A CN113570628A CN 113570628 A CN113570628 A CN 113570628A CN 202110871648 A CN202110871648 A CN 202110871648A CN 113570628 A CN113570628 A CN 113570628A
Authority
CN
China
Prior art keywords
image
cell nucleus
white
white cell
local
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.)
Granted
Application number
CN202110871648.6A
Other languages
Chinese (zh)
Other versions
CN113570628B (en
Inventor
陈伟
朱梦静
杨勇超
李利川
孟琦峰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xian University of Science and Technology
Original Assignee
Xian University of Science and Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Xian University of Science and Technology filed Critical Xian University of Science and Technology
Priority to CN202110871648.6A priority Critical patent/CN113570628B/en
Publication of CN113570628A publication Critical patent/CN113570628A/en
Application granted granted Critical
Publication of CN113570628B publication Critical patent/CN113570628B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • 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
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/155—Segmentation; Edge detection involving morphological operators
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • 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)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Artificial Intelligence (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a leukocyte segmentation method based on an active contour model, which comprises the following steps: step 1: obtaining the number of all pixel values in each small cube, assigning all pixel values in each small cube to be preset pixel values, and obtaining a preprocessed image, step 2: performing density clustering on the preprocessed image by using a density peak value clustering segmentation method to obtain an initial white blood cell image, then performing smooth processing on a white cell nucleus edge contour in the initial white blood cell image by using a morphological processing method to obtain a final white cell nucleus image, and step 3: processing the final white cell nucleus image by using a level set method to obtain a white cell nucleus contour curve, and taking the white cell nucleus contour curve as an initial contour curve of the active contour model to obtain a final white cell image; according to the local gray characteristic of the white blood cells, the white blood cells are segmented by evolving the cell nucleus contour line to the cytoplasm edge, so that a more accurate result is obtained.

Description

Leukocyte segmentation method based on active contour model
Technical Field
The invention belongs to the field of medical image processing, and particularly relates to a leukocyte segmentation method based on an active contour model.
Background
Medical image segmentation is one of the main research directions in the image processing field at present due to its great contribution to the fields of assistant medicine, clinical medicine, and the like. For certain medical images, such as white blood cell images in peripheral blood cell images. The change in the number of leukocytes can be used as a judgment of whether the human body is infected. However, generally, due to the characteristics of white blood cells or the influence of equipment, the nucleus and cytoplasm of the white blood cells are inevitably affected by gray scale unevenness and noise in the dyeing and image acquisition stages, and although segmentation algorithms are continuously proposed, the accuracy and speed of a single traditional segmentation algorithm are still not ideal.
The peripheral blood image contains a large amount of red blood cells, platelets, and the like in addition to various types of white blood cells. Although white blood cells can be well marked after a standard staining method is adopted for the white blood cells, the nucleus of the white blood cells is deeply stained with dark purple and the cytoplasm of the white blood cells is excessively purple. Meanwhile, due to the problems of interference of red blood cells in the background, staining leakage and the like, most of segmentation algorithms have excellent segmentation performance on white cell nuclei, but have unsatisfactory segmentation effect on white cytoplasm.
Generally, the system for identifying and classifying leukocytes comprises the following aspects: preprocessing of images, segmentation of white blood cells, feature extraction and classification. The segmentation is used as a link before a classification task and is of great importance to the accuracy of classification. The main segmentation methods at present have two aspects: one is a segmentation method based on the graph theory, mainly comprising a threshold value method, a clustering method and the like; the other is a segmentation method based on a variation theory, which mainly comprises a moving contour method. The segmentation method based on the atlas theory does not have a method which can effectively overcome the problem of uneven gray level of a white blood cell image under various conditions, but a segmentation method based on the variation theory is simple, because of the huge gray level difference between the cell nucleus and the cell cytoplasm of the white blood cell, the evolution curve of the active contour is difficult to fit with the cell nucleus edge or can be converged at the cell nucleus edge, and the white blood cell is difficult to segment.
Disclosure of Invention
The invention aims to provide a leukocyte segmentation method based on an active contour model, which can accurately segment the nucleus and cytoplasm of a leukocyte.
The invention adopts the following technical scheme: a leukocyte segmentation method based on an active contour model comprises the following steps:
step 1: dividing an RGB pixel space of an original stained leukocyte image into small cubes with equal length, calculating pixel values positioned at a central point in all the small cubes, namely predetermined pixel values, obtaining the number of all the pixel values in each small cube, assigning all the pixel values in each small cube to be the predetermined pixel values, and obtaining a preprocessed image,
step 2: performing density clustering on the preprocessed image by using a density peak value clustering segmentation method to obtain an initial white blood cell image, then performing smooth processing on the white cell nucleus edge contour in the initial white blood cell image by using a morphological processing method to obtain a final white cell nucleus image,
and step 3: and processing the final white cell nucleus image by using a level set method to obtain a white cell nucleus contour curve, taking the white cell nucleus contour curve as an initial contour curve of the active contour model, constructing a local gray information expression by using local JS divergence as an edge stop function, and gradually calculating the initial contour curve to obtain a white cell edge so as to obtain a final white cell image.
Further, the edge stop function in step 3 is:
Figure BDA0003189416260000021
wherein I (x) denotes a local image region，C1And C2The gray scale values are respectively the average gray scale value of the inner part and the outer part divided by the contour curve in the local area, p and q respectively represent the gray scale distribution of the inner part and the outer part divided by the contour curve, JS (p, q) represents JS divergence information between the gray scale distribution of the inner part and the outer part, alpha is a local threshold value, 0.02 is taken to measure the uniformity of the local gray scale, alpha-JS (p, q) > 0 represents that the gray scale is uniform, otherwise, the gray scale is non-uniform, and a local window is selected to be 5.
Further, step 2 consists of the following steps:
step 21: firstly, calculating the local density rho and the relative distance delta of a preprocessed image pixel by pixel, and determining the truncation distance dc；
Wherein, the local density rho of the density cluster is:
Figure BDA0003189416260000031
in the above formula, di，jRepresenting the distance between pixel i and pixel j, calculated using the Euclidean distance, dcutoffRepresents the truncation distance, i.e. the effective density radius, which is the only parameter in the clustering of density peaks, taking 0.5% of the maximum relative distance, then ρiThe distribution condition of the pixel points i around the pixel points, namely the local density, can be represented;
wherein, the relative distance δ of the density cluster is:
Figure BDA0003189416260000032
step 22: arranging products of local density rho and relative distance delta of all pixel points in a descending order, taking the first N points as clustering center points, taking the N value of the clustering center as 2, and representing that clustering structures are divided into two types, one type is a background part, and the other type is a foreground white cell nucleus;
step 23: classifying and dividing the rest points, namely pixel points except the N clustering center points according to the clustering centers to obtain clustered images, namely white cell nucleus images, binarizing the clustered white cell nucleus images, taking the white cell nuclei as a foreground, taking the rest parts as a background, and eliminating small areas by using a morphological processing method and utilizing morphological expansion and corrosion operation; and processing the foreground part to obtain a white cell nucleus edge contour, and then smoothing the white cell nucleus edge contour, namely smoothing the edge of the image by Gaussian filtering, and taking 1.8 from Sigma to obtain a final white cell nucleus image.
The invention has the beneficial effects that: according to the method, the white blood cell image is segmented by combining a clustering method of a map theory and an active contour method of a difference theory, the coarse contour of a white blood cell nucleus is obtained through density clustering, the self characteristics of the white blood cell are considered, and the white blood cell is segmented by evolving a cell nucleus contour line to a cytoplasm edge according to the local gray characteristic of the white blood cell, so that a more accurate result is obtained.
Drawings
FIG. 1 shows the results of the processing of stained leukocyte images in step 1 in example 1 of the present invention; FIG. 1a is a pre-processed image, FIG. 1b is a post-processed image, and FIG. 1c is a post-processed image;
fig. 2 shows results before and after density peak clustering segmentation of white cell nuclei for a white cell image according to embodiment 1 of the present invention, where fig. 2a shows a white cell nucleus image obtained after density clustering, and fig. 2b shows a white cell nucleus image after binarization, morphological processing, and smoothing;
fig. 3 is a result of fine segmentation of a white blood cell image by using a moving contour method according to embodiment 1 of the present invention, where fig. 3a is a result of initialization of a cell nucleus contour to an initial contour line image, fig. 3b is a result of curve evolution, and fig. 3c is a result of final segmentation.
Detailed Description
The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
The invention discloses a leukocyte segmentation method based on an active contour model, which comprises the following steps:
step 1: dividing an RGB pixel space of an original stained leukocyte image into small cubes with equal length, calculating pixel values positioned at a central point in all the small cubes, namely, preset pixel values, obtaining the number of all the pixel values in each small cube, assigning all the pixel values in each small cube to be the preset pixel values, and obtaining a preprocessed image.
Since the problem of huge data processing is faced when the density clustering processing is carried out on the original leukocyte image, the leukocyte image is preprocessed by adopting a mini-RGBCUBE mode, so that the number of pixels is reduced.
Step 2: performing density clustering on the preprocessed image by using a density peak value clustering segmentation method to obtain an initial white blood cell image, and then performing smooth processing on the white cell nucleus edge contour in the initial white blood cell image by using a morphological processing method to obtain a final white cell nucleus image.
Besides the density peak clustering segmentation method in step 2, other segmentation methods such as K-Means (K-Means) and C-Means (C-Means) can be adopted, but the color consistency of the segmented small regions and the time complexity of the algorithm are considered comprehensively, so that the performance of the density peak clustering algorithm is relatively optimal.
Performing density clustering on the preprocessed color image in the step 1, and binarizing a clustering result by adopting Euclidean distance measurement, wherein white cell nuclei are used as a foreground, and the rest parts are used as a background; and then, smoothing the white cell nucleus edge contour in the initial white cell image by using a morphological processing method to obtain a final white cell nucleus image.
Wherein, the step 2 comprises the following steps:
step 21: firstly, calculating the local density rho and the relative distance delta of a preprocessed image pixel by pixel, and determining the truncation distance dc。
Wherein, the local density rho of the density cluster is:
Figure BDA0003189416260000051
in the above formula, di,jAnd representing the distance between the pixel point i and the pixel point j, and calculating by using the Euclidean distance. dcutoffRepresenting the cutoff distance, i.e. effective density radiusIt is the only parameter in the density peak clustering, taking 0.5% of the maximum relative distance. Then ρiThe distribution condition of the pixel points i around the pixel points, namely the local density, can be represented.
Wherein, the relative distance δ of the density cluster is:
Figure BDA0003189416260000061
step 22: and arranging products of the local density rho and the relative distance delta of all the pixel points in a descending order, taking the first N points as clustering center points, taking the N value of the clustering center as 2, and representing that the clustering structure is divided into two types, one type is a background part, and the other type is a foreground white cell nucleus.
Step 23: classifying and dividing the rest points, namely pixel points except the N clustering center points according to the clustering centers to obtain clustered images, namely white cell nucleus images, binarizing the clustered white cell nucleus images, taking the white cell nuclei as a foreground, taking the rest parts as a background, and eliminating small areas by using a morphological processing method and utilizing morphological expansion and corrosion operation; and processing the foreground part to obtain a white cell nucleus edge contour, and then smoothing the white cell nucleus edge contour, namely smoothing the edge of the image by Gaussian filtering, and taking 1.8 from Sigma to obtain a final white cell nucleus image.
And step 3: and processing the final white cell nucleus image by using a level set method to obtain a cell nucleus contour curve, taking the cell nucleus contour curve as an initial contour curve of the active contour model, constructing a local gray information expression by using local JS divergence as an edge stop function, and gradually calculating the initial contour curve to obtain a white cell edge so as to obtain a final white cell image.
Wherein, the step 3 comprises the following steps:
step 31: and processing the final white cell nucleus image by using a level set method to obtain a cell nucleus contour curve, taking the cell nucleus contour curve as an initial contour line, and processing the white cell nucleus edge contour obtained in the step 24 by using the level set method, wherein the white cell nucleus part is set to be 1, and the rest part is set to be 0.
Step 32: constructing an edge stop function of local JS divergence drive;
the edge stop function is:
Figure BDA0003189416260000071
wherein I (x) denotes a local image region, C1 and C2The gray values are respectively the average gray values of the local area divided into an inner part and an outer part by the contour curve. JS (p | | q) represents JS divergence information between the gray distribution of the inner part and the outer part, alpha is a local threshold value, and 0.02 is taken to measure the uniformity of local gray.
Step 33: and (5) evolving an active contour curve according to the local gray level change condition of the white blood cells by utilizing an edge stop function to obtain a final white blood cell image.
The method utilizes the evolution of the activity contour curve to finely divide the white blood cells, takes the cell nucleus contour curve as the initial contour curve of the activity contour model, constructs a local gray information expression by using the local JS divergence as an edge stop function, and guides the evolution of the initial contour curve to the white blood cell edge, thereby obtaining the accurate white blood cell division image.
The method comprises the steps of preprocessing an original white blood cell image, roughly segmenting cell nucleuses, and accurately segmenting an active contour model of curve evolution by using roughly segmented contours. On one hand, the white blood cells are segmented by a two-step method according to the problem of uneven staining of cytoplasm and cell nucleus of the white blood cell image; on the other hand, a variation active contour model segmentation method with a solid mathematical theory basis is adopted, the mathematical theory is effectively combined with the actual environment, an edge stop function of the active contour model is obtained, white blood cells with uneven gray levels are more accurately segmented, the dyeing characteristics of the white blood cell nuclei in the peripheral blood image are considered, white blood cell nucleus parts are obtained in a clustering mode, the edge stop function is adopted to evolve contour curves aiming at the characteristic of uneven gray levels of white blood cell cytoplasm parts, and the two aspects are combined, so that the accurate white blood cell segmentation method is realized.
Example 1
Step 1: the input original stained leukocyte color image is preprocessed by using an RGB-CUBE method as shown in fig. 1a, the number of local pixels is reduced, an RGB pixel space (256 × 256) is divided into small CUBEs with each side length of 16, and a length of 16 is defined.
Calculating the pixel value of the central point of all the small cubes, namely the preset pixel value, obtaining the number npixel of all the pixel values of each small cube, assigning all the pixel values of each small cube to be the preset pixel value, and obtaining the preprocessed image so as to reduce the number of the pixels of the original image.
Selecting a cube length of 2nIn 16(n is 4), the actual n can be (0-9), but as the value of n is not increased, after the pixel points in the small cubes are classified to the center of each small cube, the whole pixels of the image are reduced sharply, so that the original image is not distinguishable, when n is 0, the length of the small cube is 1, and at this time, the central pixel is itself, which is equivalent to that the image is not changed; when n is 9, the cube length is 256, and the center pixel is (127, 127, 127), which corresponds to changing the whole image to a gray value, as shown in fig. 1 b. Therefore, through experiments, when n is 4, both efficiency and accuracy can be considered, and meanwhile, the calculation method of the pixel at the central point of each microcube is as follows:
Figure BDA0003189416260000081
wherein,
Figure BDA0003189416260000082
indicating a rounded-down symbol and pixels representing pixel values of the original image.
The RGBCUBE method is used to reduce the number of original pixels, which facilitates the subsequent clustering process, thereby obtaining a preprocessed image, as shown in fig. 1 c.
Step 2: and (3) carrying out density clustering on the preprocessed images in the step (1), and binarizing a clustering result by adopting Euclidean distance measurement, wherein a white cell nucleus is used as a foreground (white), and the rest part is used as a background (black). And then processing the foreground part by using a morphological processing method, and smoothing the white cell nucleus edge contour in the initial white cell image by adopting a connected domain marking mode to obtain a final white cell nucleus image.
Obtaining the number of pixels npixel in each small cube and the pixel value centerpoints of the center point of the small cube aiming at the preprocessed image in the step 1, and then calculating the local density ρ of each central pixel value centerpoints, wherein the local density ρ of the density cluster is as follows:
Figure BDA0003189416260000091
in the above formula, di，jAnd representing the distance between the pixel point i and the pixel point j, and calculating by using the Euclidean distance. dcutoffRepresents the truncation distance, i.e. the effective density radius, which is the only parameter in the clustering of density peaks, taking 0.5% of the maximum relative distance. Experimental tests prove that: dcA better clustering result can be obtained by taking 0.5% of the maximum relative distance; then ρiThe distribution condition of the pixel points i around the pixel points, namely the local density, can be represented.
Wherein, the relative distance δ of the density cluster is:
Figure BDA0003189416260000092
and after the local density and the relative distance of each pixel point centerpoints are calculated, the products of the two are arranged in a descending order, the point with the first N being 2 is taken as a clustering center, and the number N of the clustering centers is selected so as to obtain a white cell nucleus part, so that the number of the selected clusters is 2.
After the clustering center is found, the rest of the pixel points are subjected to extended clustering, and are distributed to the cluster of the data point which is closest to the pixel points and has a local density larger than that of the pixel points, and finally, a density peak value clustering result is obtained, as shown in fig. 2 a. And eliminating the tiny regions of the clustered results by morphological expansion and corrosion operation, wherein the core size is 5X 5. The image after the morphological processing is subjected to gaussian smoothing filtering, edge smoothing, and smoothing operations may also adopt median filtering, mean filtering, etc., but the gaussian filtering is adopted to smooth the image, so that the overall gray distribution characteristics of the image can be more retained, and a final white nucleus image is obtained, as shown in fig. 2 b.
And step 3: precise segmentation of white blood cells by active contour algorithm
Taking the white nucleus contour curve as the initial level set, as shown in fig. 3a, the final white nucleus image obtained in step 2 is set to 1 in the nucleus part and 0 in the rest part, and then is used as the function of the initial level set.
Calculating the gray average C divided into an inner part and an outer part by the contour curve in the local area1And C2：
Figure BDA0003189416260000101
Figure BDA0003189416260000102
Wherein, I is a local image, H (Φ) is a smooth form of a unit step function, and is used for calculating the inner and outer parts of the local region, specifically:
Figure BDA0003189416260000103
where phi is the profile curve, eps is a minimal constant, and eps is 1 e-3.
And (3) constructing an edge stop function driven by local Jensen-Shannon divergence by using the local gray level image mean value:
Figure BDA0003189416260000104
wherein I (x) represents a local image region, C1And C2The gray scale values are respectively the average gray scale value of the inner part and the outer part divided by the contour curve in the local area, p and q respectively represent the gray scale distribution of the inner part and the outer part divided by the contour curve, JS (p, q) represents JS divergence information between the gray scale distribution of the inner part and the outer part, alpha is a local threshold value, 0.02 is taken to measure the uniformity of the local gray scale, alpha-JS (p, q) > 0 represents that the gray scale is uniform, otherwise, the gray scale is non-uniform, and a local window is selected to be 5.
And guiding the evolution of the contour curve by using the constructed edge stopping function, and assuming that the gray density of the target region in the image is greater than that of the background region (vice versa), when JS (p, q) > alpha, namely alpha-JS (p, q) < 0, gjspf< 0, meaning that the profile curve moves towards the region with large gray scale change, and is finally located at the edge of the image; when JS (p, q) < alpha, i.e. alpha-JS (p, q) > 0, gjspfThe fact that the evolution curve is in a gray level flat area in the image is meant to be greater than 0, and the gray value of each pixel is not greatly different. Continuously detecting g during the evolution of the contour curvejspfThe contour curve is continuously approaching the target edge and finally stops at the target boundary.
Through gjspfThe contour curve is finally converged at the cell edge due to the continuous change of the signs, at this time, the signs inside the contour curve are negative, and the signs outside the contour curve are positive, so as to achieve the purpose of accurately dividing the white blood cells, as shown in fig. 3b and 3 c.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.

Claims (3)

1. A leukocyte segmentation method based on an active contour model is characterized by comprising the following steps:
step 1: dividing an RGB pixel space of an original stained leukocyte image into small cubes with equal length, calculating pixel values positioned at a central point in all the small cubes, namely predetermined pixel values, obtaining the number of all the pixel values in each small cube, assigning all the pixel values in each small cube to be the predetermined pixel values, and obtaining a preprocessed image,
step 2: performing density clustering on the preprocessed image by using a density peak value clustering segmentation method to obtain an initial white blood cell image, then performing smooth processing on the white cell nucleus edge contour in the initial white blood cell image by using a morphological processing method to obtain a final white cell nucleus image,
and step 3: and processing the final white cell nucleus image by using a level set method to obtain a white cell nucleus contour curve, taking the white cell nucleus contour curve as an initial contour curve of the active contour model, constructing a local gray information expression by using local JS divergence as an edge stop function, and gradually calculating the initial contour curve to obtain a white cell edge so as to obtain a final white cell image.
2. The method for segmenting white blood cells based on the active contour model according to claim 1, wherein the edge stop function in step 3 is:
Figure FDA0003189416250000011
wherein I (x) denotes a local image region, C1And C2The gray scale values are respectively the average gray scale value of the inner part and the outer part divided by the contour curve in the local area, p and q respectively represent the gray scale distribution of the inner part and the outer part divided by the contour curve, JS (p, q) represents JS divergence information between the gray scale distribution of the inner part and the outer part, alpha is a local threshold value, 0.02 is taken to measure the uniformity of the local gray scale, alpha-JS (p, q) > 0 represents that the gray scale is uniform, otherwise, the gray scale is non-uniform, and a local window is selected to be 5.
3. The method for segmenting white blood cells based on the active contour model according to claim 1, wherein the step 2 comprises the following steps:
step 21: firstly, calculating the local density rho and the relative distance delta pixel by pixel of a preprocessed image, and determiningFixed cut-off distance dc；
Wherein, the local density rho of the density cluster is:
Figure FDA0003189416250000021
in the above formula, di，jRepresenting the distance between pixel i and pixel j, calculated using the Euclidean distance, dcutoffRepresents the truncation distance, i.e. the effective density radius, which is the only parameter in the clustering of density peaks, taking 0.5% of the maximum relative distance, then ρiThe distribution condition of the pixel points i around the pixel points, namely the local density, can be represented;
wherein, the relative distance δ of the density cluster is:
Figure FDA0003189416250000022
step 22: arranging products of local density rho and relative distance delta of all pixel points in a descending order, taking the first N points as clustering center points, taking the N value of the clustering center as 2, and representing that clustering structures are divided into two types, one type is a background part, and the other type is a foreground white cell nucleus;
step 23: classifying and dividing the rest points, namely pixel points except the N clustering center points according to the clustering centers to obtain clustered images, namely white cell nucleus images, binarizing the clustered white cell nucleus images, taking the white cell nuclei as a foreground, taking the rest parts as a background, and eliminating small areas by using a morphological processing method and utilizing morphological expansion and corrosion operation; and processing the foreground part to obtain a white cell nucleus edge contour, and then smoothing the white cell nucleus edge contour, namely smoothing the edge of the image by Gaussian filtering, and taking 1.8 from Sigma to obtain a final white cell nucleus image.
CN202110871648.6A 2021-07-30 2021-07-30 White blood cell segmentation method based on movable contour model Active CN113570628B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110871648.6A CN113570628B (en) 2021-07-30 2021-07-30 White blood cell segmentation method based on movable contour model

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110871648.6A CN113570628B (en) 2021-07-30 2021-07-30 White blood cell segmentation method based on movable contour model

Publications (2)

Publication Number Publication Date
CN113570628A true CN113570628A (en) 2021-10-29
CN113570628B CN113570628B (en) 2024-04-02

Family

ID=78169445

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110871648.6A Active CN113570628B (en) 2021-07-30 2021-07-30 White blood cell segmentation method based on movable contour model

Country Status (1)

Country Link
CN (1) CN113570628B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115620075A (en) * 2022-12-16 2023-01-17 南昌大学 Method, system and equipment for generating data set for leukocyte classification model
CN116433701A (en) * 2023-06-15 2023-07-14 武汉中观自动化科技有限公司 Workpiece hole profile extraction method, device, equipment and storage medium
CN117237646A (en) * 2023-11-15 2023-12-15 深圳市润海电子有限公司 PET high-temperature flame-retardant adhesive tape flaw extraction method and system based on image segmentation

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103473739A (en) * 2013-08-15 2013-12-25 华中科技大学 White blood cell image accurate segmentation method and system based on support vector machine
WO2016091016A1 (en) * 2014-12-12 2016-06-16 山东大学 Nucleus marker watershed transformation-based method for splitting adhered white blood cells
CN110288582A (en) * 2019-06-26 2019-09-27 电子科技大学 A kind of accurate dividing method of nucleus
WO2019223706A1 (en) * 2018-05-22 2019-11-28 杭州智微信息科技有限公司 Saturation clustering-based method for positioning bone marrow white blood cells

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103473739A (en) * 2013-08-15 2013-12-25 华中科技大学 White blood cell image accurate segmentation method and system based on support vector machine
WO2016091016A1 (en) * 2014-12-12 2016-06-16 山东大学 Nucleus marker watershed transformation-based method for splitting adhered white blood cells
WO2019223706A1 (en) * 2018-05-22 2019-11-28 杭州智微信息科技有限公司 Saturation clustering-based method for positioning bone marrow white blood cells
CN110288582A (en) * 2019-06-26 2019-09-27 电子科技大学 A kind of accurate dividing method of nucleus

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
潘晨, 闫相国, 郑崇勋, 梁成文: "利用单类支持向量机分割血细胞图像", 西安交通大学学报, no. 02 *
蔡隽;鲍旭东;吴磊;罗立民;: "基于活动轮廓模型的彩色白细胞图像自动分割方法研究", 生物医学工程研究, no. 04 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115620075A (en) * 2022-12-16 2023-01-17 南昌大学 Method, system and equipment for generating data set for leukocyte classification model
CN116433701A (en) * 2023-06-15 2023-07-14 武汉中观自动化科技有限公司 Workpiece hole profile extraction method, device, equipment and storage medium
CN116433701B (en) * 2023-06-15 2023-10-10 武汉中观自动化科技有限公司 Workpiece hole profile extraction method, device, equipment and storage medium
CN117237646A (en) * 2023-11-15 2023-12-15 深圳市润海电子有限公司 PET high-temperature flame-retardant adhesive tape flaw extraction method and system based on image segmentation
CN117237646B (en) * 2023-11-15 2024-01-30 深圳市润海电子有限公司 PET high-temperature flame-retardant adhesive tape flaw extraction method and system based on image segmentation

Also Published As

Publication number Publication date
CN113570628B (en) 2024-04-02

Similar Documents

Publication Publication Date Title
CN113570628B (en) White blood cell segmentation method based on movable contour model
CN109522908B (en) Image significance detection method based on region label fusion
Zhou et al. Automatic detection of exudates in digital color fundus images using superpixel multi-feature classification
CN106340016B (en) A kind of DNA quantitative analysis method based on microcytoscope image
CN106296653A (en) Brain CT image hemorrhagic areas dividing method based on semi-supervised learning and system
CN108664976B (en) Super-pixel-based fuzzy spectral clustering brain tumor image automatic segmentation method
CN114067114B (en) Area nuclear segmentation counting method based on algae cell morphology
CN103761726B (en) Block adaptive image partition method based on FCM
CN104616308A (en) Multiscale level set image segmenting method based on kernel fuzzy clustering
CN104217213B (en) A kind of medical image multistage sorting technique based on symmetric theory
CN115661467A (en) Cerebrovascular image segmentation method, device, electronic equipment and storage medium
CN116912255B (en) Follicular region segmentation method for ovarian tissue analysis
El-Khamy et al. An efficient brain mass detection with adaptive clustered based fuzzy C-mean and thresholding
CN106327490A (en) Nucleus segmentation method based on white blood cell detection
CN111429461B (en) Novel segmentation method for overlapped and exfoliated epithelial cells
CN109285176B (en) Brain tissue segmentation method based on regularization graph segmentation
CN113850792A (en) Cell classification counting method and system based on computer vision
Zhang et al. A Morphological Image Segmentation Algorithm for Circular Overlapping Cells.
CN111091578B (en) Segmentation method for vascular medical image
Gunawan et al. Fuzzy Region Merging Using Fuzzy Similarity Measurement on Image Segmentation
CN114764810A (en) Medical image segmentation method
Karthik et al. An enhanced approach for spot segmentation of microarray images
CN108182684B (en) Image segmentation method and device based on weighted kernel function fuzzy clustering
CN110136100B (en) Automatic classification method and device for CT slice images
CN103208116A (en) Neighborhood-information-based method for segmentation of images with non-uniform gray levels for blurred active contour model

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