CN111862128A - Image segmentation method and device - Google Patents
Image segmentation method and device Download PDFInfo
- Publication number
- CN111862128A CN111862128A CN202010537490.4A CN202010537490A CN111862128A CN 111862128 A CN111862128 A CN 111862128A CN 202010537490 A CN202010537490 A CN 202010537490A CN 111862128 A CN111862128 A CN 111862128A
- Authority
- CN
- China
- Prior art keywords
- image
- area
- segmented
- edge
- preprocessed
- 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
Links
- 238000003709 image segmentation Methods 0.000 title claims abstract description 49
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000003708 edge detection Methods 0.000 claims abstract description 24
- 238000007781 pre-processing Methods 0.000 claims abstract description 19
- 238000012545 processing Methods 0.000 claims description 45
- 230000009466 transformation Effects 0.000 claims description 18
- 238000006243 chemical reaction Methods 0.000 claims description 15
- 230000009467 reduction Effects 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 4
- 238000012549 training Methods 0.000 abstract description 7
- 230000008569 process Effects 0.000 description 10
- 238000005516 engineering process Methods 0.000 description 5
- 238000005457 optimization Methods 0.000 description 5
- 238000004801 process automation Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 239000000463 material Substances 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 238000012800 visualization Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 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
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G06T5/70—
-
- 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
- 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
- G06T7/90—Determination of colour characteristics
Abstract
In order to solve the problem that the existing image segmentation has high requirements on computer hardware, the disclosure provides an image segmentation method and an image segmentation device, so as to reduce the requirements of the image segmentation on the computer hardware. The image segmentation method comprises the following steps: preprocessing an image to be segmented to generate a preprocessed image; carrying out edge detection on the preprocessed image to obtain edge information of the preprocessed image; and segmenting the image to be segmented according to the edge information. The image segmentation method comprises the following steps: the device comprises a preprocessing module, an edge detection module and an image segmentation module. The image segmentation method and the image segmentation device can reduce the requirements of image segmentation on computer hardware, reduce the requirements on training data and improve the practicability of image segmentation.
Description
Technical Field
The present disclosure relates to the field of computer vision technologies, and in particular, to an image segmentation method and apparatus.
Background
In recent years, artificial intelligence technology is changing day by day, and various intelligent products are emerging, such as: industrial robots, food delivery robots, robot process automation systems, and the like have led to rapid changes in modern social development. The robot process automation system is popular among workers of enterprises and public institutions due to convenience and convenience of office processes. The image segmentation technology in the field of computer vision is a technology for dividing an image into different attribute regions, and plays a crucial role in robot automation.
However, most of the image segmentation technologies with better effect are realized by deep learning, a large amount of training data is needed for application of the deep learning to obtain a model with better generalization performance, a large amount of manpower, material resources and financial resources are needed for supporting the collected training data, and the model obtained after training occupies larger memory or needs GPU support when running, so that the requirement of the image segmentation on computer hardware is higher.
Disclosure of Invention
In order to solve one of the above technical problems, the present disclosure provides an image segmentation method and apparatus to reduce the requirement of image segmentation on computer hardware.
In a first aspect of the present disclosure, an image segmentation method includes:
preprocessing an image to be segmented to generate a preprocessed image;
performing edge detection on the preprocessed image to obtain edge information of the preprocessed image;
and segmenting the image to be segmented according to the edge information.
Further, the preprocessing the image to be segmented to generate a preprocessed image includes:
acquiring a gray level image of the image to be segmented;
performing conversion processing on the grayscale image to generate a first image;
performing pixel classification processing on the first image to generate a pixel classification image;
And generating a preprocessed image according to the pixel classification image.
Further, the acquiring the gray scale image of the image to be segmented comprises:
judging whether the image to be segmented is a color image or a gray image;
if the image to be segmented is a color image, carrying out gray processing on the image to be segmented;
and if the image to be segmented is a gray image, not performing gray processing on the image to be segmented.
Further, the performing conversion processing on the grayscale image includes:
and carrying out image dimension transformation and data type transformation on the gray level image.
Further, the performing pixel classification processing on the first image includes:
and carrying out pixel classification processing on the first image according to a clustering algorithm.
Further, the generating a preprocessed image from the pixel classified image includes:
converting the pixel classification image into a binary image according to a set threshold, wherein the set threshold is a median of central points obtained when the first image is subjected to pixel classification processing according to a clustering algorithm;
and performing noise reduction processing on the binary image to generate a preprocessed image.
Further, the performing edge detection on the preprocessed image to obtain edge information of the preprocessed image includes:
Identifying initial edge pixel points of the preprocessed image;
carrying out noise reduction processing on the initial edge pixel points to obtain low-noise edge pixel points;
and optimizing the low-noise edge pixel points to obtain edge information of the preprocessed image.
Further, the optimizing the low-noise edge pixel point to obtain the edge information of the preprocessed image includes:
if the area A is a subset of the area B, taking the low-noise edge pixel point of the area B as a target edge pixel point;
if the area A is not the subset of the area B and the intersection of the area A and the area B is not empty, taking the pixel point of the minimum matrix area containing the area A and the area B as a target edge pixel point;
if the area A is not the subset of the area B and the intersection of the area A and the area B is empty, taking the low-noise edge pixel points of the area A and the area B as target edge pixel points;
taking the information of each target edge pixel point as the edge information of the preprocessed image;
and the area A and the area B are areas surrounded by the low-noise edge pixels.
In a second aspect of the present disclosure, an image segmentation apparatus includes:
The preprocessing module is used for preprocessing the image to be segmented to generate a preprocessed image;
the edge detection module is used for carrying out edge detection on the preprocessed image so as to obtain edge information of the preprocessed image;
and the image segmentation module is used for segmenting the image to be segmented according to the edge information.
Further, the device also comprises an image acquisition module, wherein the image acquisition module is used for acquiring the image to be segmented locally or remotely.
The beneficial effects of this disclosure are: an image segmentation method and device are provided. The image to be segmented is preprocessed to reduce the complexity of the image, then the edge detection is carried out on the preprocessed image to be segmented, and finally the image is segmented according to the detection information of the edge detection, so that the visualization of useful information in the image content is enhanced, and the convenience of the user in operating the specific content in the image is improved; the need for training data is reduced; the requirements of image segmentation on computer hardware are reduced; the problem of dividing different small regions which are relatively close to each other into the same large region and the problem of dividing one large region into different small regions are avoided. In addition, the image pixels of the image are classified, so that the overall complexity of the image is reduced, and the readability of the image is improved; when the edge detection and optimization are carried out, firstly, the noise of the detected edge is reduced, the influence caused by some noises is eliminated, secondly, all the detected edges are optimized by adopting a special edge optimization method, the redundant information of the image edge is reduced, and the image segmentation effect is optimized.
Drawings
The accompanying drawings, which are included to provide a further understanding of the disclosure and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the disclosure and together with the description serve to explain the principles of the disclosure.
Fig. 1 is a flowchart of an image segmentation method provided in an embodiment of the present application;
FIG. 2 is a flow chart of generating a preprocessed image according to an embodiment of the present disclosure;
fig. 3 is a flowchart of a grayscale image determination provided in an embodiment of the present application;
fig. 4 is a flowchart of a process for converting a grayscale image according to an embodiment of the present application;
FIG. 5 is a flow chart of generating a pre-processed image from a pixel classification image according to an embodiment of the present application;
FIG. 6 is a flowchart for obtaining edge information of a preprocessed image according to an embodiment of the present disclosure;
fig. 7 is a schematic diagram of an image segmentation apparatus according to an embodiment of the present application.
Detailed Description
The present disclosure will be described in further detail with reference to the drawings and embodiments. It is to be understood that the specific embodiments described herein are for purposes of illustration only and are not to be construed as limitations of the present disclosure. It should be further noted that, for the convenience of description, only the portions relevant to the present disclosure are shown in the drawings.
It should be noted that the embodiments and features of the embodiments in the present disclosure may be combined with each other without conflict. The present disclosure will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.
Example 1:
referring to fig. 1, the present embodiment is a method, including:
s1: preprocessing an image to be segmented to generate a preprocessed image;
s2: carrying out edge detection on the preprocessed image to obtain edge information of the preprocessed image;
s3: the image to be segmented is segmented according to the segmentation.
According to the method, the preprocessing image of the image to be segmented is obtained after edge detection, and the complex image content to be segmented is automatically segmented; the visualization of useful information in the content of the image to be segmented is enhanced, and the convenience of the user in operating the specific content in the image is improved; the requirement on training data is reduced, and the requirement on computer hardware is lowered.
The image segmentation method disclosed by the invention can support online use and offline use, namely remote use and local use, and can be applied to more occasions compared with a deep learning model.
Compared with a deep learning model, the image segmentation method disclosed by the invention occupies a small memory, and has a high operation speed and high accuracy. The image segmentation method disclosed by the invention does not require the input of images with fixed sizes, and can segment images with any sizes.
In one embodiment, the pre-processing is used to reduce the complexity of the image to be segmented; specifically, referring to fig. 2, preprocessing an image to be segmented to generate a preprocessed image includes:
s11: acquiring a gray level image of the image to be segmented;
s12: performing conversion processing on the grayscale image to generate a first image;
s13: performing pixel classification processing on the first image to generate a pixel classification image;
s14: a pre-processed image is generated from the pixel classification image.
In the embodiment, the gray image of the image to be segmented is obtained, and the conversion processing and the pixel classification processing are performed on the gray image, so that the image complexity is reduced, and the edge detection is conveniently realized in the step S2; the pixel classification processing can reduce noise interference between different small regions which are relatively close to each other, and can avoid dividing different small regions into the same large region.
Specifically, the conversion processing in S12 refers to converting the dimensions and types of images so that the first image can be subjected to image classification processing;
for example, the conversion process involves image dimension transformation and data type transformation; if the image to be segmented is an M × N color image, the image to be segmented is represented in a computer as a three-dimensional array of M × N × 3, the three-dimensional array is converted into a gray image and then is converted into a two-dimensional array of M × N, and the image dimension conversion is that the two-dimensional array of M × N is converted into a one-dimensional array of MN. If the data type of the gray image is float32 (floating point type) and the gray image of the pixel classification processing is uint8 (unsigned integer type), the data type conversion converts float32 to uint 8.
Specifically, referring to fig. 3, acquiring the grayscale image of the image to be segmented includes:
s111: judging whether the image to be segmented is a color image or a gray image;
if the image to be segmented is a color image, S112: carrying out gray processing on an image to be segmented;
if the image to be segmented is a gray image, S113: carrying out graying treatment on the image to be segmented;
specifically, the converting process of the grayscale image includes: and carrying out image dimension transformation and data type transformation on the gray level image.
More specifically, referring to fig. 4, the performing of the conversion process on the grayscale image includes:
s121: carrying out image dimension transformation on the gray level image;
s122: carrying out data type transformation on the gray level image subjected to image dimension transformation;
the image dimension transformation may be from an image of dimension 3 to an image of dimension 2;
the data type transformation may be from the data type of agent 8 to the data type of float32, etc.; the gradation image is subjected to conversion processing to process the image more efficiently.
In one embodiment, the first image is subjected to a pixel classification process according to a clustering algorithm.
Referring to fig. 5, generating a preprocessed image from the pixel classified image includes:
S131: converting the pixel classification image into a binary image according to a set threshold, wherein the set threshold is a median of a central point obtained when the first image is subjected to pixel classification processing according to a clustering algorithm; the center point is the final clustering center point returned by the end of the clustering algorithm.
S132: and performing noise reduction processing on the binary image to generate a preprocessed image.
The method and the device for image segmentation and the system for image segmentation are characterized in that the clustering algorithm is used for carrying out pixel classification on the image to reduce the complexity of the segmented image, then the threshold value is calculated according to the clustering center point in the clustering algorithm, the image after the pixel classification is converted into a binary image to further reduce the complexity of the image, and finally the noise reduction is carried out on the binary image to reduce the influence of noise introduced during the pixel classification and the influence of wrongly classified pixel points on the image.
In the embodiment, noise interference between different small regions which are relatively close to each other is reduced through pixel classification processing, and the different small regions are prevented from being divided into the same large region; the pixel classification image is converted into the binary image, so that useful information which is easily filtered out in a large area during filtering can be enhanced, and the problem that one large area is divided into different small areas, and then different small areas which are relatively close to each other are divided into the same large area and one large area is divided into different small areas can be avoided. The image segmentation method disclosed by the invention can segment characters and objects in the image at the same time.
In this embodiment, the clustering algorithm needs to set the number of clusters first, set the conditions for stopping the algorithm second, and set the initial clustering center finally. The algorithm stopping conditions can be accuracy and iteration times respectively, and the accuracy and the iteration times can also be set simultaneously, and when the accuracy and the iteration times are set simultaneously, the algorithm stops when one of the conditions is met; the number of clusters and the initial cluster center can be set as desired.
In one embodiment, referring to fig. 6, performing edge detection on the preprocessed image to obtain edge information of the preprocessed image includes:
s21: identifying initial edge pixel points of the preprocessed image;
s22: carrying out noise reduction processing on the initial edge pixel points to obtain low-noise edge pixel points;
s23: and optimizing the low-noise edge pixel points to obtain edge information of the preprocessed image.
The method comprises the steps of firstly identifying initial edge pixel points of a preprocessed image with reduced complexity according to the existing edge detection algorithm, then reducing noise of the identified initial edge pixel points to eliminate noise caused by calculation and the like in the identification process, and finally optimizing the edge pixel points with the noise reduced by utilizing the optimization algorithm provided by the disclosure to generate edge information with higher confidence coefficient.
In particular, the method comprises the following steps of,
optimizing the low-noise edge pixel point to obtain edge information of the preprocessed image, including:
if the area A is a subset of the area B, taking the low-noise edge pixel point of the area B as a target edge pixel point;
if the area A is not the subset of the area B and the intersection of the area A and the area B is not empty, taking the pixel point of the minimum matrix area containing the area A and the area B as a target edge pixel point;
if the area A is not the subset of the area B and the intersection of the area A and the area B is empty, taking the low-noise edge pixel points of the area A and the area B as target edge pixel points;
taking the information of each target edge pixel point as the edge information of the preprocessed image;
and the area A and the area B are areas surrounded by the low-noise edge pixels.
By the optimization method, the edge pixel points after noise reduction are optimized to generate edge information with higher confidence coefficient, and the image segmentation precision is improved; meanwhile, redundant information of the image edge is reduced, and the image segmentation effect is optimized.
In one embodiment, segmenting the image to be segmented according to the edge information further comprises displaying the segmentation result.
In one embodiment, before step S1 of the image segmentation method, the method further includes acquiring an image to be segmented;
the method comprises the steps of obtaining an image to be segmented, judging the using mode of a user by a robot or selecting the using mode of the user by the user according to a flow configured by the user, obtaining the image to be segmented by the user in a remote control mode through the relevant atomic capability of the robot if the user uses a remote mode, and taking the image configured in the flow by the user as the image to be segmented if the user uses a local mode. Among them, atomic capability is a function of robot process automation (also called RPA), such as image segmentation, text detection, screen capture, mouse click, etc.
Example 2:
referring to fig. 7, the image segmentation apparatus of the present embodiment includes:
the preprocessing module 101 is configured to preprocess an image to be segmented to generate a preprocessed image;
an edge detection module 102, configured to perform edge detection on the preprocessed image to obtain edge information of the preprocessed image;
and the image segmentation module 103 is used for segmenting the image to be segmented according to the edge information.
Specifically, the image segmentation apparatus further includes an image obtaining module 100, configured to obtain an image to be segmented.
In this embodiment, the image obtaining module 100 obtains an image to be segmented, the preprocessing module 101 preprocesses the currently obtained image to be segmented to reduce complexity, the edge detecting module 102 performs edge detection on the preprocessed image, and the image segmenting module 103 segments the image according to the result of the edge detection.
The principle and effect of the image segmentation apparatus in the present embodiment are the same as those of the method in embodiment 1.
The image acquisition module 100 is used for remote data acquisition and local data acquisition; specifically, if the user uses the robot process automation in a remote manner, the user acquires an image in a remote control manner through the relevant atomic capability of the robot, and if the user uses the robot process automation in an off-line manner, the user directly designates image data to be segmented.
In one embodiment, the pre-processing is used to reduce the complexity of the image to be segmented;
specifically, the preprocessing the image to be segmented to generate a preprocessed image includes:
acquiring a gray level image of the image to be segmented;
performing conversion processing on the gray-scale image to generate a first image meeting the pixel classification requirement;
Performing pixel classification processing on the first image to generate a pixel classification image;
a pre-processed image is generated from the pixel classification image.
In the embodiment, the gray level image of the image to be segmented is obtained, and the conversion processing and the pixel classification processing are carried out on the gray level image, so that the image complexity is reduced, and the edge detection is convenient to realize; the pixel classification processing can reduce noise interference between different small regions which are relatively close to each other, and can avoid dividing different small regions into the same large region.
Specifically, the acquiring a gray scale image of the image to be segmented includes:
judging whether the image to be segmented is a color image or a gray image;
if the image to be segmented is a color image, converting the image to be segmented into a gray scale image;
if the image to be segmented is a gray scale image, keeping the image to be segmented unchanged.
Specifically, the converting process of the grayscale image includes: and carrying out image dimension transformation and data type transformation on the gray level image.
More specifically, the conversion processing of the gradation image includes:
carrying out image dimension transformation on the gray level image;
carrying out data type transformation on the gray level image subjected to image dimension transformation;
in one embodiment, the first image is subjected to a pixel classification process according to a clustering algorithm.
Generating a preprocessed image from the pixel-classified image, comprising:
converting the pixel classification image into a binary image according to a set threshold, wherein the set threshold is a median of a central point obtained when the first image is subjected to pixel classification processing according to a clustering algorithm;
and performing noise reduction processing on the binary image to generate a preprocessed image.
The method and the device for image segmentation and the system for image segmentation are characterized in that the clustering algorithm is used for carrying out pixel classification on the image to reduce the complexity of the segmented image, then the threshold value is calculated according to the clustering center point in the clustering algorithm, the image after the pixel classification is converted into a binary image to further reduce the complexity of the image, and finally the noise reduction is carried out on the binary image to reduce the influence of noise introduced during the pixel classification and the influence of wrongly classified pixel points on the image.
In the embodiment, noise interference between different small regions which are relatively close to each other is reduced through pixel classification processing, and the different small regions are prevented from being divided into the same large region; the useful information that filters when can strengthen being filtered easily among the big region with pixel classification image conversion binary image can avoid cutting apart into different little regions with a big region, and then has avoided cutting apart into same big region with the different little regions that are close apart to and cut apart into the problem in different little regions with a big region.
In one embodiment, edge detection is performed on the preprocessed image to obtain edge information of the preprocessed image, including:
identifying initial edge pixel points of the preprocessed image;
carrying out noise reduction processing on the initial edge pixel points to obtain low-noise edge pixel points;
and optimizing the low-noise edge pixel points to obtain edge information of the preprocessed image.
Specifically, the optimizing processing is performed on the low-noise edge pixel point to obtain the edge information of the preprocessed image, and the optimizing processing includes:
if the area A is a subset of the area B, taking the low-noise edge pixel point of the area B as a target edge pixel point;
if the area A is not the subset of the area B and the intersection of the area A and the area B is not empty, taking the pixel point of the minimum matrix area containing the area A and the area B as a target edge pixel point;
if the area A is not the subset of the area B and the intersection of the area A and the area B is empty, taking the low-noise edge pixel points of the area A and the area B as target edge pixel points;
taking the information of each target edge pixel point as the edge information of the preprocessed image;
and the area A and the area B are areas formed by surrounding low-noise edge pixels.
In one embodiment, the device further comprises an image acquisition module, wherein the image acquisition module is used for acquiring the image to be segmented locally or remotely;
the method comprises the steps of obtaining an image to be segmented, judging the using mode of a user by a robot or selecting the using mode of the user by the user according to a flow configured by the user, obtaining the image to be segmented by the user in a remote control mode through the relevant atomic capability of the robot if the user uses a remote mode, and taking the image configured in the flow by the user as the image to be segmented if the user uses a local mode.
In summary, the present disclosure provides an image segmentation method and an image segmentation apparatus, which reduce image complexity through image preprocessing and segment different regions in an image through a series of optimization algorithms. The problem that the robot flow automation needs a large amount of training data, occupies a large memory or needs GPU support, and the problem that the robot flow automation is difficult to use offline is solved, and the problem that different small areas are easily used as a large area when the different small areas are close to each other and the large area is divided into different small areas when the large area is large is solved.
In the description herein, reference to the description of the terms "one embodiment/mode," "some embodiments/modes," "example," "specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment/mode or example is included in at least one embodiment/mode or example of the application. In this specification, the schematic representations of the terms used above are not necessarily intended to be the same embodiment/mode or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments/modes or examples. Furthermore, the various embodiments/aspects or examples and features of the various embodiments/aspects or examples described in this specification can be combined and combined by one skilled in the art without conflicting therewith.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present application, "plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.
It will be understood by those skilled in the art that the foregoing embodiments are merely for clarity of illustration of the disclosure and are not intended to limit the scope of the disclosure. Other variations or modifications may occur to those skilled in the art, based on the foregoing disclosure, and are still within the scope of the present disclosure.
Claims (10)
1. An image segmentation method, comprising:
preprocessing an image to be segmented to generate a preprocessed image;
performing edge detection on the preprocessed image to obtain edge information of the preprocessed image;
and segmenting the image to be segmented according to the edge information.
2. The method of claim 1, wherein preprocessing the image to be segmented to generate a preprocessed image comprises:
Acquiring a gray level image of the image to be segmented;
performing conversion processing on the grayscale image to generate a first image;
performing pixel classification processing on the first image to generate a pixel classification image;
and generating a preprocessed image according to the pixel classification image.
3. The method according to claim 2, wherein the obtaining a grayscale image of the image to be segmented comprises:
judging whether the image to be segmented is a color image or a gray image;
if the image to be segmented is a color image, carrying out gray processing on the image to be segmented;
and if the image to be segmented is a gray image, not performing gray processing on the image to be segmented.
4. The method according to claim 2, wherein the converting the grayscale image comprises:
and carrying out image dimension transformation and data type transformation on the gray level image.
5. The method of claim 2, wherein the performing pixel classification processing on the first image comprises:
and carrying out pixel classification processing on the first image according to a clustering algorithm.
6. The method of claim 5, wherein generating a pre-processed image from the pixel classified image comprises:
Converting the pixel classification image into a binary image according to a set threshold, wherein the set threshold is a median of central points obtained when the first image is subjected to pixel classification processing according to a clustering algorithm;
and performing noise reduction processing on the binary image to generate a preprocessed image.
7. The method according to claim 1, wherein the performing edge detection on the preprocessed image to obtain edge information of the preprocessed image comprises:
identifying initial edge pixel points of the preprocessed image;
carrying out noise reduction processing on the initial edge pixel points to obtain low-noise edge pixel points;
and optimizing the low-noise edge pixel points to obtain edge information of the preprocessed image.
8. The method of claim 7, wherein the optimizing the low-noise edge pixel point to obtain the edge information of the preprocessed image comprises:
if the area A is a subset of the area B, taking the low-noise edge pixel point of the area B as a target edge pixel point;
if the area A is not the subset of the area B and the intersection of the area A and the area B is not empty, taking the pixel point of the minimum matrix area containing the area A and the area B as a target edge pixel point;
If the area A is not the subset of the area B and the intersection of the area A and the area B is empty, taking the low-noise edge pixel points of the area A and the area B as target edge pixel points;
taking the information of each target edge pixel point as the edge information of the preprocessed image;
and the area A and the area B are areas surrounded by the low-noise edge pixels.
9. An image segmentation apparatus, comprising:
the preprocessing module is used for preprocessing the image to be segmented to generate a preprocessed image;
the edge detection module is used for carrying out edge detection on the preprocessed image so as to obtain edge information of the preprocessed image;
and the image segmentation module is used for segmenting the image to be segmented according to the edge information.
10. The image segmentation apparatus according to claim 9, wherein the apparatus further comprises an image acquisition module configured to acquire the image to be segmented locally or remotely.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010537490.4A CN111862128B (en) | 2020-06-12 | 2020-06-12 | Image segmentation method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010537490.4A CN111862128B (en) | 2020-06-12 | 2020-06-12 | Image segmentation method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111862128A true CN111862128A (en) | 2020-10-30 |
| CN111862128B CN111862128B (en) | 2024-04-16 |
Family
ID=72987892
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010537490.4A Active CN111862128B (en) | 2020-06-12 | 2020-06-12 | Image segmentation method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111862128B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115145024A (en) * | 2022-07-06 | 2022-10-04 | 南开大学 | Adjusting and controlling method for shaping laser |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102324099A (en) * | 2011-09-05 | 2012-01-18 | 广东工业大学 | Step edge detection method oriented to humanoid robot |
| CN105261017A (en) * | 2015-10-14 | 2016-01-20 | 长春工业大学 | Method for extracting regions of interest of pedestrian by using image segmentation method on the basis of road restriction |
| CN106251332A (en) * | 2016-07-17 | 2016-12-21 | 西安电子科技大学 | SAR image airport target detection method based on edge feature |
| CN107644400A (en) * | 2016-07-21 | 2018-01-30 | 中国石油化工股份有限公司 | The Texture Segmentation Methods and device of seismic section image |
| CN109447036A (en) * | 2018-11-16 | 2019-03-08 | 齐鲁工业大学 | A kind of segmentation of image digitization and recognition methods and system |
| CN109492645A (en) * | 2018-11-01 | 2019-03-19 | 湖南文理学院 | A kind of registration number character dividing method and device |
| CN110163219A (en) * | 2019-04-17 | 2019-08-23 | 安阳师范学院 | Object detection method based on image border identification |
| CN110689553A (en) * | 2019-09-26 | 2020-01-14 | 湖州南太湖智能游艇研究院 | Automatic segmentation method of RGB-D image |
| CN111046862A (en) * | 2019-12-05 | 2020-04-21 | 北京嘉楠捷思信息技术有限公司 | Character segmentation method and device and computer readable storage medium |
| WO2020107716A1 (en) * | 2018-11-30 | 2020-06-04 | 长沙理工大学 | Target image segmentation method and apparatus, and device |
-
2020
- 2020-06-12 CN CN202010537490.4A patent/CN111862128B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102324099A (en) * | 2011-09-05 | 2012-01-18 | 广东工业大学 | Step edge detection method oriented to humanoid robot |
| CN105261017A (en) * | 2015-10-14 | 2016-01-20 | 长春工业大学 | Method for extracting regions of interest of pedestrian by using image segmentation method on the basis of road restriction |
| CN106251332A (en) * | 2016-07-17 | 2016-12-21 | 西安电子科技大学 | SAR image airport target detection method based on edge feature |
| CN107644400A (en) * | 2016-07-21 | 2018-01-30 | 中国石油化工股份有限公司 | The Texture Segmentation Methods and device of seismic section image |
| CN109492645A (en) * | 2018-11-01 | 2019-03-19 | 湖南文理学院 | A kind of registration number character dividing method and device |
| CN109447036A (en) * | 2018-11-16 | 2019-03-08 | 齐鲁工业大学 | A kind of segmentation of image digitization and recognition methods and system |
| WO2020107716A1 (en) * | 2018-11-30 | 2020-06-04 | 长沙理工大学 | Target image segmentation method and apparatus, and device |
| CN110163219A (en) * | 2019-04-17 | 2019-08-23 | 安阳师范学院 | Object detection method based on image border identification |
| CN110689553A (en) * | 2019-09-26 | 2020-01-14 | 湖州南太湖智能游艇研究院 | Automatic segmentation method of RGB-D image |
| CN111046862A (en) * | 2019-12-05 | 2020-04-21 | 北京嘉楠捷思信息技术有限公司 | Character segmentation method and device and computer readable storage medium |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115145024A (en) * | 2022-07-06 | 2022-10-04 | 南开大学 | Adjusting and controlling method for shaping laser |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111862128B (en) | 2024-04-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11361505B2 (en) | Model retrieval for objects in images using field descriptors | |
| CN115082683B (en) | Injection molding defect detection method based on image processing | |
| Ahmed et al. | Salient segmentation based object detection and recognition using hybrid genetic transform | |
| CN110148130B (en) | Method and device for detecting part defects | |
| CN112966684B (en) | Cooperative learning character recognition method under attention mechanism | |
| CN111275082A (en) | Indoor object target detection method based on improved end-to-end neural network | |
| US7653244B2 (en) | Intelligent importation of information from foreign applications user interface | |
| WO2022033095A1 (en) | Text region positioning method and apparatus | |
| US20190279368A1 (en) | Method and Apparatus for Multi-Model Primitive Fitting based on Deep Geometric Boundary and Instance Aware Segmentation | |
| Xiang et al. | Moving object detection and shadow removing under changing illumination condition | |
| CN108274476B (en) | Method for grabbing ball by humanoid robot | |
| CN111862128B (en) | Image segmentation method and device | |
| CN115331008A (en) | End-to-end target detection method based on target probability density graph | |
| CN111382741A (en) | Method, system and equipment for detecting text in natural scene picture | |
| Milyaev et al. | Improving the processing of machine vision images of robotic systems in the Arctic | |
| CN116071658B (en) | SAR image small target detection and recognition method and device based on deep learning | |
| US20230091374A1 (en) | Systems and Methods for Improved Computer Vision in On-Device Applications | |
| CN117710208A (en) | Self-training system of self-adaptive model and method for self-training self-adaptive model | |
| Aguiar et al. | Identification of External Defects on Fruits Using Deep Learning | |
| Pan et al. | Accuracy improvement of deep learning 3D point cloud instance segmentation | |
| Sharma et al. | Maximum entropy-based semi-supervised learning for automatic detection and recognition of objects using deep ConvNets | |
| Chen et al. | Design of gesture recognition system based on machine vision | |
| Vatchalaa et al. | Smart Household Object Detection Using CNN | |
| Liu et al. | Robust Real-Time Head Detection by Grayscale Template Matching Based on Depth Images | |
| CN117612020A (en) | SGAN-based detection method for resisting neural network remote sensing image element change |
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 |