CN109285178A - Image partition method, device and storage medium - Google Patents
Image partition method, device and storage medium Download PDFInfo
- Publication number
- CN109285178A CN109285178A CN201811252058.XA CN201811252058A CN109285178A CN 109285178 A CN109285178 A CN 109285178A CN 201811252058 A CN201811252058 A CN 201811252058A CN 109285178 A CN109285178 A CN 109285178A
- Authority
- CN
- China
- Prior art keywords
- pixel
- target image
- image
- identifier
- area
- 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.)
- Pending
Links
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
- G06T7/194—Segmentation; Edge detection involving foreground-background segmentation
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The disclosure is directed to a kind of image partition method, device and storage mediums, belong to field of image processing.Method includes: the pixel classifications model for obtaining and having trained, pixel classifications model is used to determine the class indication of each pixel in any image, class indication includes first identifier and second identifier, first identifier is used to indicate corresponding pixel and belongs to foreground area, and second identifier is used to indicate corresponding pixel and belongs to background area；Target image to be processed is input in pixel classifications model, pixel classifications model is based on, determines the class indication of each pixel in target image；According to the class indication of pixel each in target image, foreground area and the background area of target image are determined, so as to extract foreground area or the background area of target image.By way of training pixel classifications model, it can be directly based upon pixel classifications model, carry out image segmentation to target image reduces calculation amount, save the time without largely being calculated.
Description
Technical field
This disclosure relates to field of image processing more particularly to a kind of image partition method, device and storage medium.
Background technique
Piece image generally includes foreground area and background area, and image Segmentation Technology can determine the prospect in image
Region and background area, foreground area and background area are distinguished, independent to carry out to foreground area or background area
Ground processing, is widely used at present in the multiple fields such as target detection, object tracking, video monitoring, recognition of face.
The nontransparent degree of the foreground of the display color of any pixel and the pixel, background colour and foreground has in image
Following incidence relation: I=α F+ (1- α) B, wherein I indicates that the display color of pixel, F indicate the foreground of pixel, and B indicates picture
The background colour of element, α indicate the nontransparent degree of pixel foreground.When to be split to target image, calculated using matting
Method, can be according to the display color and above-mentioned incidence relation of pixels multiple in the target image, before calculating each pixel
The nontransparent degree of scenery generates prospect probability graph (mask) according to the nontransparent degree of the foreground of each pixel, the prospect probability
Figure is used to indicate the foreground area in the target image and the position where background area, according to the prospect probability in subsequent process
Figure handles the target image, and foreground area or background area can be extracted from target image.
Above scheme needs to carry out a large amount of nontransparent degree for calculating and capable of just calculating the foreground of each pixel, calculation amount
It is larger, it takes a long time.
Summary of the invention
Present disclose provides a kind of image partition method, device and storage mediums, can overcome present in the relevant technologies
Problem.
According to the first aspect of the embodiments of the present disclosure, a kind of image partition method is provided, which comprises
The pixel classifications model trained is obtained, the pixel classifications model is for determining each pixel in any image
Class indication, the class indication include first identifier and second identifier, and the first identifier is used to indicate corresponding pixel category
In foreground area, the second identifier is used to indicate corresponding pixel and belongs to background area；
Target image to be processed is input in the pixel classifications model, the pixel classifications model is based on, is determined
The class indication of each pixel in the target image；
According to the class indication of pixel each in the target image, the foreground area and background of the target image are determined
Region.
In a kind of mode in the cards, the method also includes:
Obtain the class indication of each pixel in multiple sample images and the multiple sample image；
Model is carried out according to the class indication of each pixel in the multiple sample image and the multiple sample image
Training, obtains the pixel classifications model.
In the mode of alternatively possible realization, the class indication according to pixel each in the target image, really
The foreground area of the fixed target image and background area, comprising:
According to the class indication of pixel each in the target image, prospect probability graph is generated, the prospect probability graph is used
Position where the foreground area and background area for indicating the target image.
In the mode of alternatively possible realization, the method also includes:
According to the prospect probability graph, enhancing processing is carried out to the foreground area of the target image；Alternatively,
According to the prospect probability graph, Fuzzy processing is carried out to the background area of the target image.
It is described according to the prospect probability graph in the mode of alternatively possible realization, to the back in the target image
Scene area carries out Fuzzy processing, comprising:
According to the prospect probability graph, the foreground area of the target image is extracted；
Fuzzy processing is carried out to the target image, obtains the first image, and according to the prospect probability graph, extracts institute
State the background area of the first image；
The background area of the foreground area of the target image and the first image is combined, the second figure is obtained
Picture.
In the mode of alternatively possible realization, in the prospect probability graph, the pixel value of pixel is 1 in foreground area,
The pixel value of pixel is 0 in background area, described according to the prospect probability graph, is carried out to the background area of the target image
Fuzzy processing, comprising:
The background area of the target image is carried out at blurring using following formula according to the prospect probability graph
Reason:
Target=Source*mask+Gaussian* (255-mask)；
Wherein, after Source indicates that the target image, Gaussian indicate that the target image carries out Fuzzy processing
Obtained image, mask indicate that the prospect probability graph, Target indicate that the background area of the target image is blurred
The image obtained after processing.
According to the second aspect of an embodiment of the present disclosure, a kind of image segmentation device is provided, described device includes:
Model obtains module, is configured as obtaining the pixel classifications model trained, the pixel classifications model is for true
Determine the class indication of each pixel in any image, the class indication includes first identifier and second identifier, first mark
Knowledge is used to indicate corresponding pixel and belongs to foreground area, and the second identifier is used to indicate corresponding pixel and belongs to background area；
Determining module is identified, is configured as target image to be processed being input in the pixel classifications model, be based on
The pixel classifications model, determines the class indication of each pixel in the target image；
Area determination module is configured as determining the mesh according to the class indication of pixel each in the target image
The foreground area of logo image and background area.
In a kind of mode in the cards, described device further include:
Sample acquisition module is configured as obtaining each pixel in multiple sample images and the multiple sample image
Class indication；
Training module is configured as according to each pixel in the multiple sample image and the multiple sample image
Class indication carries out model training, obtains the pixel classifications model.
In the mode of alternatively possible realization, the area determination module includes:
Generation unit is configured as the class indication according to pixel each in the target image, generates prospect probability graph,
The foreground area and the position where background area that the prospect probability graph is used to indicate the target image.
In the mode of alternatively possible realization, described device further include:
Processing module is configured as enhancing the foreground area of the target image according to the prospect probability graph
Processing；Alternatively, the processing module, is additionally configured to according to the prospect probability graph, to the background area of the target image
Carry out Fuzzy processing.
In the mode of alternatively possible realization, the processing module includes:
Extraction unit is configured as extracting the foreground area of the target image according to the prospect probability graph；
The extraction unit is additionally configured to carry out Fuzzy processing to the target image, obtains the first image, and root
According to the prospect probability graph, the background area of the first image is extracted；
Assembled unit is configured as carrying out the background area of the foreground area of the target image and the first image
Combination, obtains the second image.
In the mode of alternatively possible realization, in the prospect probability graph, the pixel value of pixel is 1 in foreground area,
The pixel value of pixel is 0 in background area, and the processing module is additionally configured to according to the prospect probability graph, and use is following
Formula carries out Fuzzy processing to the background area of the target image:
Target=Source*mask+Gaussian* (255-mask)；
Wherein, after Source indicates that the target image, Gaussian indicate that the target image carries out Fuzzy processing
Obtained image, mask indicate that the prospect probability graph, Target indicate that the background area of the target image is blurred
The image obtained after processing.
According to the third aspect of an embodiment of the present disclosure, a kind of image segmentation device is provided, described device includes:
Processor；
Memory for storage processor executable command；
Wherein, the processor is configured to:
The pixel classifications model trained is obtained, the pixel classifications model is for determining each pixel in any image
Class indication, the class indication include first identifier and second identifier, and the first identifier is used to indicate corresponding pixel category
In foreground area, the second identifier is used to indicate corresponding pixel and belongs to background area；
Target image to be processed is input in the pixel classifications model, the pixel classifications model is based on, is determined
The class indication of each pixel in the target image；
According to the class indication of pixel each in the target image, the foreground area and background of the target image are determined
Region.
According to a fourth aspect of embodiments of the present disclosure, a kind of non-transitorycomputer readable storage medium is provided, when described
When instruction in storage medium is executed by the processor of processing unit, so that processing unit is able to carry out a kind of image segmentation side
Method, which comprises
The pixel classifications model trained is obtained, the pixel classifications model is for determining each pixel in any image
Class indication, the class indication include first identifier and second identifier, and the first identifier is used to indicate corresponding pixel category
In foreground area, the second identifier is used to indicate corresponding pixel and belongs to background area；
Target image to be processed is input in the pixel classifications model, the pixel classifications model is based on, is determined
The class indication of each pixel in the target image；
According to the class indication of pixel each in the target image, the foreground area and background of the target image are determined
Region.
According to a fifth aspect of the embodiments of the present disclosure, a kind of computer program product is provided, when the computer program produces
When instruction in product is executed by the processor of processing unit, so that processing unit is able to carry out a kind of image partition method, it is described
Method includes:
The pixel classifications model trained is obtained, the pixel classifications model is for determining each pixel in any image
Class indication, the class indication include first identifier and second identifier, and the first identifier is used to indicate corresponding pixel category
In foreground area, the second identifier is used to indicate corresponding pixel and belongs to background area；
Target image to be processed is input in the pixel classifications model, the pixel classifications model is based on, is determined
The class indication of each pixel in the target image；
According to the class indication of pixel each in the target image, the foreground area and background of the target image are determined
Region.
The technical scheme provided by this disclosed embodiment can include the following benefits:
The pixel classifications model trained is obtained, pixel classifications model is used to determine the classification of each pixel in any image
Mark, class indication includes first identifier and second identifier, and first identifier is used to indicate corresponding pixel and belongs to foreground area, the
Two marks are used to indicate corresponding pixel and belong to background area, and target image to be processed is input in pixel classifications model,
Based on pixel classifications model, the class indication of each pixel in target image is determined, according to point of pixel each in target image
Class mark, determines foreground area and the background area of target image, so as to extract the foreground area or background of target image
Region.By way of training pixel classifications model, it can be directly based upon pixel classifications model, image point is carried out to target image
It cuts, without largely being calculated, reduces calculation amount, save the time.
It should be understood that above general description and following detailed description be only it is exemplary and explanatory, not
The disclosure can be limited.
Detailed description of the invention
The drawings herein are incorporated into the specification and forms part of this specification, and shows the implementation for meeting the disclosure
Example, and together with specification for explaining the principles of this disclosure.
Fig. 1 is a kind of flow chart of image partition method shown according to an exemplary embodiment.
Fig. 2 is a kind of flow chart of image partition method shown according to an exemplary embodiment.
Fig. 3 is a kind of block diagram of processing unit shown according to an exemplary embodiment.
Fig. 4 is a kind of block diagram of terminal for image segmentation shown according to an exemplary embodiment.
Fig. 5 is a kind of structural schematic diagram of server shown according to an exemplary embodiment.
Specific embodiment
Example embodiments are described in detail here, and the example is illustrated in the accompanying drawings.Following description is related to
When attached drawing, unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements.Following exemplary embodiment
Described in embodiment do not represent all implementations consistent with this disclosure.On the contrary, they be only with it is such as appended
The example of the consistent device and method of some aspects be described in detail in claims, the disclosure.
Fig. 1 is a kind of flow chart of image partition method shown according to an exemplary embodiment, as shown in Figure 1, the figure
As dividing method is in processing unit, comprising the following steps:
In a step 101, the pixel classifications model trained is obtained, pixel classifications model is every in any image for determining
The class indication of a pixel, class indication include first identifier and second identifier, and first identifier is used to indicate corresponding pixel category
In foreground area, second identifier is used to indicate corresponding pixel and belongs to background area.
In a step 102, target image to be processed is input in pixel classifications model, is based on pixel classifications model,
Determine the class indication of each pixel in target image.
In step 103, according to the class indication of pixel each in target image, determine target image foreground area and
Background area.
The method that the embodiment of the present disclosure provides obtains the pixel classifications model trained, and pixel classifications model is for determining
The class indication of each pixel in any image, class indication include first identifier and second identifier, and first identifier is used to indicate
Corresponding pixel belongs to foreground area, and second identifier is used to indicate corresponding pixel and belongs to background area, by target to be processed
Image is input in pixel classifications model, is based on pixel classifications model, is determined the class indication of each pixel in target image, root
According to the class indication of pixel each in target image, foreground area and the background area of target image are determined, so as to extract
The foreground area of target image or background area.By way of training pixel classifications model, pixel classifications can be directly based upon
Model, carry out image segmentation to target image reduces calculation amount, saves the time without largely being calculated.
In a kind of mode in the cards, method further include:
Obtain the class indication of each pixel in multiple sample images and multiple sample images；
Model training is carried out according to the class indication of each pixel in multiple sample images and multiple sample images, is obtained
Pixel classifications model.
In the mode of alternatively possible realization, according to the class indication of pixel each in target image, target figure is determined
The foreground area of picture and background area, comprising:
According to the class indication of pixel each in target image, prospect probability graph is generated, prospect probability graph is used to indicate mesh
The foreground area of logo image and the position where background area.
In the mode of alternatively possible realization, method further include:
According to prospect probability graph, enhancing processing is carried out to the foreground area of target image；Alternatively,
According to prospect probability graph, Fuzzy processing is carried out to the background area of target image.
In the mode of alternatively possible realization, according to prospect probability graph, mould is carried out to the background area in target image
Gelatinization processing, comprising:
According to prospect probability graph, the foreground area of target image is extracted；
Fuzzy processing is carried out to target image, obtains the first image, and according to prospect probability graph, extracts the first image
Background area；
The background area of the foreground area of target image and the first image is combined, the second image is obtained.
In the mode of alternatively possible realization, in prospect probability graph, the pixel value of pixel is 1 in foreground area, background
The pixel value of pixel is 0 in region, according to prospect probability graph, carries out Fuzzy processing to the background area of target image, comprising:
According to prospect probability graph, using following formula, Fuzzy processing is carried out to the background area of target image:
Target=Source*mask+Gaussian* (255-mask)；
Wherein, Source indicates that target image, Gaussian indicate that target image carries out the figure obtained after Fuzzy processing
Picture, mask indicate that prospect probability graph, Target indicate that the background area of target image carries out the image obtained after Fuzzy processing.
Fig. 2 is a kind of flow chart of image partition method shown according to an exemplary embodiment, as shown in Fig. 2, the figure
As dividing method is in processing unit, processing unit can be the tool such as mobile phone, computer, server, camera, monitoring device
There is the device of image processing function, method includes the following steps:
In step 201, processing unit obtains target image to be processed, and obtains the pixel classifications model trained.
Wherein, which can be shot to obtain by processing unit, or be mentioned from the video that processing unit takes
It obtains, is perhaps downloaded from internet by processing unit and obtain or be sent to processing unit by other equipment.Alternatively,
During processing unit carries out net cast, every image in available video flowing, using every picture as mesh
Logo image, to handle every image in video flowing.
The target image can be a plurality of types of images, such as character image, scenery picture, cuisines picture, the target
Image includes foreground area and background area, and shape of color, object that foreground area is shown with background area etc. can phase
Together, or different, difference is only that: the area where object when foreground area refers to the shooting target image close to camera
Domain is equivalent to " center of interest " of the target image, and background area is in addition to foreground area, for serving as a contrast or foil foreground area
Region.
It further include the environment of personage behind for example, both having included personage in the portrait photographs of shooting, personage region at this time
It may be considered foreground area, behind environment region may be considered background area, alternatively, both including table in cuisines picture
Son further includes the cuisines on desk, and desk region may be considered background area at this time, and cuisines region can consider
It is foreground area.
It, can be by the foreground area of the target image and background area area based on pixel classifications model in the embodiment of the present disclosure
It branches away.Wherein, pixel classifications model is used to determine that the class indication of each pixel in any image, class indication to include first
Mark and second identifier, first identifier are used to indicate corresponding pixel and belong to foreground area, and second identifier is used to indicate corresponding
Pixel belongs to background area, that is to say, when the class indication of any pixel in image is first identifier, indicates the pixel category
In foreground area, when the class indication of any pixel in image is second identifier, indicate that the pixel belongs to background area.
First identifier is two different marks from second identifier, for example, first identifier is 1, and second identifier is 0, or
Person, first identifier 0, and second identifier is 1.
The pixel classifications model can be obtained by processing unit training, and be stored by the processing unit, alternatively, the pixel point
Class model is sent to processing unit after being trained by other equipment, and is stored by the processing unit.
In training pixel classifications model, each pixel in multiple sample images and multiple sample images can be first obtained
Class indication, according in multiple sample images and multiple sample images each pixel class indication carry out model training,
Obtain pixel classifications model.
In a kind of mode in the cards, in training pixel classifications model, initial pixel classifications mould can be constructed
Type, obtains training dataset and test data set, training dataset and test data concentrate include multiple sample images and
The class indication of each pixel in each sample image.
In the training process, multiple sample images training data concentrated are as the input of pixel classifications model, by this
Output of the class indication of each pixel as pixel classifications model in multiple sample images, instructs pixel classifications model
Practice, learn pixel classifications model to the difference between foreground area and background area, has pixel classifications model and draw
Divide the ability of foreground area and background area.Later, each sample image that test data is concentrated is input to pixel classifications mould
In type, determine that the testing classification of each pixel in each sample image identifies respectively based on pixel classifications model, by testing classification
It identifies and is compared with the actual classification of mark mark, pixel classifications model is modified according to comparing result.
, can be using default training algorithm in training pixel classifications model in a kind of possible implementation, this is default
Training algorithm can be deep learning network algorithm, decision Tree algorithms, artificial neural network algorithm etc..
In subsequent process, the class indication of each pixel in new sample image and the sample image can also be obtained,
Continue to train the pixel classifications model, so as to improve the classification accuracy of the pixel classifications model.
In step 202, target image is input in pixel classifications model by processing unit, is based on pixel classifications model,
Determine the class indication of each pixel in target image.
Processing unit is handled the target image by the way that target image to be input in pixel classifications model, can be with
Classified according to the class indication of each pixel in target image to the pixel in target image.
When the class indication of pixel is first identifier, indicate that the pixel belongs to foreground area, when the class indication of pixel
When for second identifier, indicate that the pixel belongs to background area, the class indication of each pixel and each in combining target image
The position of pixel in the target image can determine foreground area and the background area of target image, therefore the pixel classifications mould
Type is actually a kind of semantic segmentation model, target image can be divided into foreground area and background area.
In step 203, processing unit generates prospect probability graph according to the class indication of pixel each in target image.
Prospect probability graph is used to indicate the position in target image where foreground area and background area, general according to the prospect
Rate figure can determine that the pixel in target image on each position belongs to foreground area and still falls within background area, so as to incite somebody to action
Target image is divided into foreground area and background area, realizes to the foreground area of target image or taking for background area.
In a kind of possible implementation, using the class indication of pixel each in target image as right in prospect probability graph
The pixel value of pixel on position is answered, to generate prospect probability graph.Then in prospect probability graph, the pixel value of any pixel is the
One mark indicates that the pixel corresponding pixel in the target image belongs to foreground area, and the pixel value of any pixel is the second mark
Knowing indicates that the pixel corresponding pixel in the target image belongs to background area.
For example, first identifier is 1, second identifier 0, the white area in prospect probability graph indicates phase in the target image
The position answered is foreground area, and the black region in prospect probability graph indicates that corresponding position is background area in the target image
Domain.
In step 204, according to prospect probability graph, foreground area or background area to target image carry out processing unit
Processing.
In order to improve the clarity of foreground area, both enhancing processing can be carried out to foreground area, it can also be to background area
Domain carries out virtualization processing.Correspondingly, step 204 includes the following steps 2041 and 2042 at least one:
In step 2041, according to prospect probability graph, enhancing processing is carried out to the foreground area of target image.
Processing unit determines that pixel value is the pixel group of first identifier according to the pixel value of pixel each in prospect probability graph
At image-region, by the image-region, corresponding image-region is extracted in the target image, before target image
Scene area, and determine the image-region that the pixel that pixel value is second identifier forms, the image-region is right in the target image
The image-region answered extracts, the background area as target image.
Later, using algorithm for image enhancement, enhancing processing is carried out to foreground area, the foreground area that obtains that treated will
Foreground area that treated and background area are combined, the image that obtains that treated, to improve the quality of foreground area, before making
Scene area is more clear, thus more prominent foreground area.
For example, algorithm for image enhancement can for algorithm of histogram equalization, contrast enhancement algorithms or other can increase
The algorithm of strong image, this is no longer going to repeat them.
In step 2042, according to prospect probability graph, Fuzzy processing is carried out to the background area of target image.
In a kind of possible implementation, processing unit carries out Fuzzy processing to target image, obtains the first image, and
According to the pixel value of pixel each in prospect probability graph, the image-region that the pixel that pixel value is second identifier forms is determined, it will
The image-region corresponding image-region in the first image extracts, which is that the background area of target image passes through
Region after crossing Fuzzy processing, the as background area after Fuzzy processing, and determine that pixel value is the pixel of first identifier
The image-region of composition, by the image-region, corresponding image-region is extracted in the target image, as target image
Background area after foreground area and Fuzzy processing is combined by foreground area later, the image that obtains that treated, real
The Fuzzy processing to the background area of target image is showed, thus more prominent foreground area.
In alternatively possible implementation, in prospect probability graph, the pixel value of pixel is 1 in foreground area, background
The pixel value of pixel is 0 in region, according to prospect probability graph, using following formula, carries out mould to the background area of target image
Gelatinization processing:
Target=Source*mask+Gaussian* (255-mask).
Wherein, Source indicates that target image, Gaussian indicate that target image carries out the figure obtained after Fuzzy processing
Picture, mask indicate that prospect probability graph, Target indicate that the background area of target image carries out the image obtained after Fuzzy processing.
Source*mask indicates the foreground area of target image, and Gaussian* (255-mask) indicates target image by blurring
The background area of the image obtained after reason obtains the foreground area of target image and target image after Fuzzy processing
It combines the background area of image, so that it may which the background area for obtaining target image carries out the image obtained after Fuzzy processing.
It should be noted that when processing unit handles target image in the embodiment of the present disclosure, it can be first to target
Image carries out enhancing processing, then carries out Fuzzy processing to target image.Alternatively, processing unit handles target image
When, enhancing processing only can be carried out to target image, alternatively, only carrying out Fuzzy processing to target image.
The method that the embodiment of the present disclosure provides obtains the pixel classifications model trained, and pixel classifications model is for determining
The class indication of each pixel in any image, class indication include first identifier and second identifier, and first identifier is used to indicate
Corresponding pixel belongs to foreground area, and second identifier is used to indicate corresponding pixel and belongs to background area, by target to be processed
Image is input in pixel classifications model, is based on pixel classifications model, is determined the class indication of each pixel in target image, root
According to the class indication of pixel each in target image, foreground area and the background area of target image are determined, so as to extract
The foreground area of target image or background area.By way of training pixel classifications model, pixel classifications can be directly based upon
Model, carry out image segmentation to target image reduces calculation amount, saves the time, improve without largely being calculated
Background blurring process speed.
Also, by the way that foreground area is carried out enhancing processing, foreground area can be made to be more clear, by background area
Fuzzy processing is carried out, background area can be weakened, enhance the difference between foreground area and background area, so as to
More prominent foreground area, keeps foreground area more obvious.
The embodiment of the present disclosure can be applied to recognition of face, under the scenes such as video monitoring, net cast and beautification picture,
For example, human face region can be extracted from target image under recognition of face scene, ignore the area in addition to human face region
Domain, to carry out recognition of face to human face region.Under video monitoring scene, it can be extracted from every image in video flowing
The region in addition to target object region is ignored in target object region, to obtain the dynamic change of target object in video flowing
Information realizes the tracking to target object.It, can be real in every image in the video flowing of live streaming under net cast scene
When distinguish foreground area and background area, and virtualization processing is carried out to background area, so that prominent foreground area, makes foreground area
It is more clear, enhancing live streaming effect.
Fig. 3 is a kind of block diagram of processing unit shown according to an exemplary embodiment.Referring to Fig. 3, which includes mould
Type obtains module 301, mark determining module 302 and area determination module 303.
Model obtains module 301, is configured as obtaining the pixel classifications model trained, pixel classifications model is for determining
The class indication of each pixel in any image, class indication include first identifier and second identifier, and first identifier is used to indicate
Corresponding pixel belongs to foreground area, and second identifier is used to indicate corresponding pixel and belongs to background area；
Determining module 302 is identified, is configured as target image to be processed being input in pixel classifications model, is based on picture
Plain disaggregated model determines the class indication of each pixel in target image；
Area determination module 303 is configured as determining target image according to the class indication of pixel each in target image
Foreground area and background area.
In a kind of mode in the cards, device further include:
Sample acquisition module is configured as obtaining the classification of each pixel in multiple sample images and multiple sample images
Mark；
Training module is configured as the class indication according to each pixel in multiple sample images and multiple sample images
Model training is carried out, pixel classifications model is obtained.
In the mode of alternatively possible realization, area determination module 303 includes:
Generation unit is configured as the class indication according to pixel each in target image, generates prospect probability graph, prospect
The foreground area and the position where background area that probability graph is used to indicate target image.
In the mode of alternatively possible realization, device further include:
Processing module, is configured as according to prospect probability graph, carries out enhancing processing to the foreground area of target image；Or
Person, processing module are additionally configured to according to prospect probability graph, carry out Fuzzy processing to the background area of target image.
In the mode of alternatively possible realization, processing module includes:
Extraction unit is configured as extracting the foreground area of target image according to prospect probability graph；
Extraction unit is additionally configured to carry out Fuzzy processing to target image, obtains the first image, and general according to prospect
Rate figure extracts the background area of the first image；
Assembled unit is configured as being combined the background area of the foreground area of target image and the first image, obtain
To the second image.
In the mode of alternatively possible realization, in prospect probability graph, the pixel value of pixel is 1 in foreground area, background
The pixel value of pixel is 0 in region, and processing module is additionally configured to according to prospect probability graph, using following formula, to target figure
The background area of picture carries out Fuzzy processing:
Target=Source*mask+Gaussian* (255-mask)；
Wherein, Source indicates that target image, Gaussian indicate that target image carries out the figure obtained after Fuzzy processing
Picture, mask indicate that prospect probability graph, Target indicate that the background area of target image carries out the image obtained after Fuzzy processing.
About the device in above-described embodiment, wherein modules execute the concrete mode of operation in related this method
Embodiment in be described in detail, no detailed explanation will be given here.
Fig. 4 is a kind of block diagram of terminal 400 for image segmentation shown according to an exemplary embodiment.The terminal
400 are used to execute step performed by processing unit in above-mentioned image partition method, can be portable mobile termianl, such as:
Smart phone, tablet computer, MP3 player (Moving Picture Experts Group Audio Layer III, dynamic
Image expert's compression standard audio level 3), MP4 (Moving Picture Experts Group Audio Layer IV, move
State image expert's compression standard audio level 4) player, laptop or desktop computer.Terminal 400 is also possible to referred to as use
Other titles such as family equipment, portable terminal, laptop terminal, terminal console.
In general, terminal 400 includes: processor 401 and memory 402.
Processor 401 may include one or more processing cores, such as 4 core processors, 8 core processors etc..Place
Reason device 401 can use DSP (Digital Signal Processing, Digital Signal Processing), FPGA (Field-
Programmable Gate Array, field programmable gate array), PLA (Programmable Logic Array, may be programmed
Logic array) at least one of example, in hardware realize.Processor 401 also may include primary processor and coprocessor, master
Processor is the processor for being handled data in the awake state, also referred to as CPU (Central Processing
Unit, central processing unit)；Coprocessor is the low power processor for being handled data in the standby state.?
In some embodiments, processor 401 can be integrated with GPU (Graphics Processing Unit, image processor),
GPU is used to be responsible for the rendering and drafting of content to be shown needed for display screen.In some embodiments, processor 401 can also be wrapped
AI (Artificial Intelligence, artificial intelligence) processor is included, the AI processor is for handling related machine learning
Calculating operation.
Memory 402 may include one or more computer readable storage mediums, which can
To be non-transient.Memory 402 may also include high-speed random access memory and nonvolatile memory, such as one
Or multiple disk storage equipments, flash memory device.In some embodiments, the non-transient computer in memory 402 can
Storage medium is read for storing at least one instruction, at least one instruction by processor 401 for being had to realize this Shen
Please in embodiment of the method provide image partition method.
In some embodiments, terminal 400 is also optional includes: peripheral device interface 403 and at least one peripheral equipment.
It can be connected by bus or signal wire between processor 401, memory 402 and peripheral device interface 403.Each peripheral equipment
It can be connected by bus, signal wire or circuit board with peripheral device interface 403.Specifically, peripheral equipment includes: radio circuit
404, at least one of touch display screen 405, camera 406, voicefrequency circuit 407, positioning component 408 and power supply 409.
Peripheral device interface 403 can be used for I/O (Input/Output, input/output) is relevant outside at least one
Peripheral equipment is connected to processor 401 and memory 402.In some embodiments, processor 401, memory 402 and peripheral equipment
Interface 403 is integrated on same chip or circuit board；In some other embodiments, processor 401, memory 402 and outer
Any one or two in peripheral equipment interface 403 can realize on individual chip or circuit board, the present embodiment to this not
It is limited.
Radio circuit 404 is for receiving and emitting RF (Radio Frequency, radio frequency) signal, also referred to as electromagnetic signal.It penetrates
Frequency circuit 404 is communicated by electromagnetic signal with communication network and other communication equipments.Radio circuit 404 turns electric signal
It is changed to electromagnetic signal to be sent, alternatively, the electromagnetic signal received is converted to electric signal.Optionally, radio circuit 404 wraps
It includes: antenna system, RF transceiver, one or more amplifiers, tuner, oscillator, digital signal processor, codec chip
Group, user identity module card etc..Radio circuit 404 can be carried out by least one wireless communication protocol with other terminals
Communication.The wireless communication protocol includes but is not limited to: Metropolitan Area Network (MAN), each third generation mobile communication network (2G, 3G, 4G and 13G), wireless office
Domain net and/or WiFi (Wireless Fidelity, Wireless Fidelity) network.In some embodiments, radio circuit 404 may be used also
To include the related circuit of NFC (Near Field Communication, wireless near field communication), the application is not subject to this
It limits.
Display screen 405 is for showing UI (User Interface, user interface).The UI may include figure, text, figure
Mark, video and its their any combination.When display screen 405 is touch display screen, display screen 405 also there is acquisition to show
The ability of the touch signal on the surface or surface of screen 405.The touch signal can be used as control signal and be input to processor
401 are handled.At this point, display screen 405 can be also used for providing virtual push button and/or dummy keyboard, also referred to as soft button and/or
Soft keyboard.In some embodiments, display screen 405 can be one, and the front panel of terminal 400 is arranged；In other embodiments
In, display screen 405 can be at least two, be separately positioned on the different surfaces of terminal 400 or in foldover design；In still other reality
It applies in example, display screen 405 can be flexible display screen, be arranged on the curved surface of terminal 400 or on fold plane.Even, it shows
Display screen 405 can also be arranged to non-rectangle irregular figure, namely abnormity screen.Display screen 405 can use LCD (Liquid
Crystal Display, liquid crystal display), OLED (Organic Light-EmittingDiode, Organic Light Emitting Diode)
Etc. materials preparation.
CCD camera assembly 406 is for acquiring image or video.Optionally, CCD camera assembly 406 include front camera and
Rear camera.In general, the front panel of terminal is arranged in front camera, the back side of terminal is arranged in rear camera.One
In a little embodiments, rear camera at least two is main camera, depth of field camera, wide-angle camera, focal length camera shooting respectively
Any one in head, to realize that main camera and the fusion of depth of field camera realize background blurring function, main camera and wide-angle
Camera fusion realizes that pan-shot and VR (Virtual Reality, virtual reality) shooting function or other fusions are clapped
Camera shooting function.In some embodiments, CCD camera assembly 406 can also include flash lamp.Flash lamp can be monochromatic warm flash lamp,
It is also possible to double-colored temperature flash lamp.Double-colored temperature flash lamp refers to the combination of warm light flash lamp and cold light flash lamp, can be used for not
With the light compensation under colour temperature.
Voicefrequency circuit 407 may include microphone and loudspeaker.Microphone is used to acquire the sound wave of user and environment, and will
Sound wave, which is converted to electric signal and is input to processor 401, to be handled, or is input to radio circuit 404 to realize voice communication.
For stereo acquisition or the purpose of noise reduction, microphone can be separately positioned on the different parts of terminal 400 to be multiple.Mike
Wind can also be array microphone or omnidirectional's acquisition type microphone.Loudspeaker is then used to that processor 401 or radio circuit will to be come from
404 electric signal is converted to sound wave.Loudspeaker can be traditional wafer speaker, be also possible to piezoelectric ceramic loudspeaker.When
When loudspeaker is piezoelectric ceramic loudspeaker, the audible sound wave of the mankind can be not only converted electrical signals to, it can also be by telecommunications
Number the sound wave that the mankind do not hear is converted to carry out the purposes such as ranging.In some embodiments, voicefrequency circuit 407 can also include
Earphone jack.
Positioning component 408 is used for the current geographic position of positioning terminal 400, to realize navigation or LBS (Location
Based Service, location based service).Positioning component 408 can be the GPS (Global based on the U.S.
Positioning System, global positioning system), the dipper system of China, Russia Gray receive this system or European Union
The positioning component of Galileo system.
Power supply 409 is used to be powered for the various components in terminal 400.Power supply 409 can be alternating current, direct current,
Disposable battery or rechargeable battery.When power supply 409 includes rechargeable battery, which can support wired charging
Or wireless charging.The rechargeable battery can be also used for supporting fast charge technology.
In some embodiments, terminal 400 further includes having one or more sensors 410.The one or more sensors
410 include but is not limited to: acceleration transducer 411, gyro sensor 412, pressure sensor 413, fingerprint sensor 414,
Optical sensor 415 and proximity sensor 416.
The acceleration that acceleration transducer 411 can detecte in three reference axis of the coordinate system established with terminal 400 is big
It is small.For example, acceleration transducer 411 can be used for detecting component of the acceleration of gravity in three reference axis.Processor 401 can
With the acceleration of gravity signal acquired according to acceleration transducer 411, touch display screen 405 is controlled with transverse views or longitudinal view
Figure carries out the display of user interface.Acceleration transducer 411 can be also used for the acquisition of game or the exercise data of user.
Gyro sensor 412 can detecte body direction and the rotational angle of terminal 400, and gyro sensor 412 can
To cooperate with acquisition user to act the 3D of terminal 400 with acceleration transducer 411.Processor 401 is according to gyro sensor 412
Following function may be implemented in the data of acquisition: when action induction (for example changing UI according to the tilt operation of user), shooting
Image stabilization, game control and inertial navigation.
The lower layer of side frame and/or touch display screen 405 in terminal 400 can be set in pressure sensor 413.Work as pressure
When the side frame of terminal 400 is arranged in sensor 413, user can detecte to the gripping signal of terminal 400, by processor 401
Right-hand man's identification or prompt operation are carried out according to the gripping signal that pressure sensor 413 acquires.When the setting of pressure sensor 413 exists
When the lower layer of touch display screen 405, the pressure operation of touch display screen 405 is realized to UI circle according to user by processor 401
Operability control on face is controlled.Operability control includes button control, scroll bar control, icon control, menu
At least one of control.
Fingerprint sensor 414 is used to acquire the fingerprint of user, collected according to fingerprint sensor 414 by processor 401
The identity of fingerprint recognition user, alternatively, by fingerprint sensor 414 according to the identity of collected fingerprint recognition user.It is identifying
When the identity of user is trusted identity out, authorize the user that there is relevant sensitive operation, the sensitive operation packet by processor 401
Include solution lock screen, check encryption information, downloading software, payment and change setting etc..Terminal can be set in fingerprint sensor 414
400 front, the back side or side.When being provided with physical button or manufacturer Logo in terminal 400, fingerprint sensor 414 can be with
It is integrated with physical button or manufacturer's mark.
Optical sensor 415 is for acquiring ambient light intensity.In one embodiment, processor 401 can be according to optics
The ambient light intensity that sensor 415 acquires controls the display brightness of touch display screen 405.Specifically, when ambient light intensity is higher
When, the display brightness of touch display screen 405 is turned up；When ambient light intensity is lower, the display for turning down touch display screen 405 is bright
Degree.In another embodiment, the ambient light intensity that processor 401 can also be acquired according to optical sensor 415, dynamic adjust
The acquisition parameters of CCD camera assembly 406.
Proximity sensor 416, also referred to as range sensor are generally arranged at the front panel of terminal 400.Proximity sensor 416
For acquiring the distance between the front of user Yu terminal 400.In one embodiment, when proximity sensor 416 detects use
When family and the distance between the front of terminal 400 gradually become smaller, touch display screen 405 is controlled from bright screen state by processor 401
It is switched to breath screen state；When proximity sensor 416 detects user and the distance between the front of terminal 400 becomes larger,
Touch display screen 405 is controlled by processor 401 and is switched to bright screen state from breath screen state.
It will be understood by those skilled in the art that the restriction of the not structure paired terminal 400 of structure shown in Fig. 4, can wrap
It includes than illustrating more or fewer components, perhaps combine certain components or is arranged using different components.
Fig. 5 is a kind of structural schematic diagram of server shown according to an exemplary embodiment, which can be because matching
It sets or performance is different and generate bigger difference, may include one or more processors (central
Processing units, CPU) 501 and one or more memory 502, wherein it is stored in the memory 502
There is at least one instruction, at least one instruction is loaded by the processor 501 and executed to realize that above-mentioned each method is real
The method that example offer is provided.Certainly, which can also have wired or wireless network interface, keyboard and input/output interface
Equal components, to carry out input and output, which can also include other for realizing the component of functions of the equipments, not do herein
It repeats.
Server 500 can be used for executing step performed by processing unit in above-mentioned image partition method.
In the exemplary embodiment, a kind of non-transitorycomputer readable storage medium is additionally provided, when in storage medium
Instruction by processing unit processor execute when so that processing unit is able to carry out a kind of image partition method, method includes:
The pixel classifications model trained is obtained, pixel classifications model is used to determine the classification of each pixel in any image
Mark, class indication includes first identifier and second identifier, and first identifier is used to indicate corresponding pixel and belongs to foreground area, the
Two marks are used to indicate corresponding pixel and belong to background area；
Target image to be processed is input in pixel classifications model, pixel classifications model is based on, determines target image
In each pixel class indication；
According to the class indication of pixel each in target image, foreground area and the background area of target image are determined.
In the exemplary embodiment, a kind of computer program product is additionally provided, when the instruction in computer program product
When being executed by the processor of processing unit, so that processing unit is able to carry out a kind of image partition method, method includes:
The pixel classifications model trained is obtained, pixel classifications model is used to determine the classification of each pixel in any image
Mark, class indication includes first identifier and second identifier, and first identifier is used to indicate corresponding pixel and belongs to foreground area, the
Two marks are used to indicate corresponding pixel and belong to background area；
Target image to be processed is input in pixel classifications model, pixel classifications model is based on, determines target image
In each pixel class indication；
According to the class indication of pixel each in target image, foreground area and the background area of target image are determined.
Those skilled in the art will readily occur to other realities of the disclosure after considering specification and practicing disclosure herein
Apply scheme.This application is intended to cover any variations, uses, or adaptations of the disclosure, these modifications, purposes or suitable
The variation of answering property follows the general principles of this disclosure and including the undocumented common knowledge in the art of the disclosure or used
Use technological means.The description and examples are only to be considered as illustrative, and the true scope and spirit of the disclosure are by following right
It is required that pointing out.
It should be understood that the present disclosure is not limited to the precise structures that have been described above and shown in the drawings, and
And various modifications and changes may be made without departing from the scope thereof.The scope of the present disclosure is only limited by the accompanying claims.
Claims (10)
1. a kind of image partition method, which is characterized in that the described method includes:
The pixel classifications model trained is obtained, the pixel classifications model is used to determine the classification of each pixel in any image
Mark, the class indication include first identifier and second identifier, and the first identifier is used to indicate before corresponding pixel belongs to
Scene area, the second identifier are used to indicate corresponding pixel and belong to background area；
Target image to be processed is input in the pixel classifications model, be based on the pixel classifications model, determine described in
The class indication of each pixel in target image；
According to the class indication of pixel each in the target image, the foreground area and background area of the target image are determined
Domain.
2. the method according to claim 1, wherein the method also includes:
Obtain the class indication of each pixel in multiple sample images and the multiple sample image；
Model training is carried out according to the class indication of each pixel in the multiple sample image and the multiple sample image,
Obtain the pixel classifications model.
3. the method according to claim 1, wherein the classification according to pixel each in the target image
Mark, determines foreground area and the background area of the target image, comprising:
According to the class indication of pixel each in the target image, prospect probability graph is generated, the prospect probability graph is for referring to
The foreground area for showing the target image and the position where background area.
4. according to the method described in claim 3, it is characterized in that, the method also includes:
According to the prospect probability graph, enhancing processing is carried out to the foreground area of the target image；Alternatively,
According to the prospect probability graph, Fuzzy processing is carried out to the background area of the target image.
5. according to the method described in claim 4, it is characterized in that, described according to the prospect probability graph, to the target figure
Background area as in carries out Fuzzy processing, comprising:
According to the prospect probability graph, the foreground area of the target image is extracted；
Fuzzy processing is carried out to the target image, obtains the first image, and according to the prospect probability graph, extracts described the
The background area of one image；
The background area of the foreground area of the target image and the first image is combined, the second image is obtained.
6. according to the method described in claim 4, it is characterized in that, in the prospect probability graph, the picture of pixel in foreground area
Element value is 1, and the pixel value of pixel is 0 in background area, described according to the prospect probability graph, to the background of the target image
Region carries out Fuzzy processing, comprising:
According to the prospect probability graph, using following formula, Fuzzy processing is carried out to the background area of the target image:
Target=Source*mask+Gaussian* (255-mask)；
Wherein, Source indicates that the target image, Gaussian indicate to obtain after the target image carries out Fuzzy processing
Image, mask indicates that the prospect probability graph, Target indicate that the background area of the target image carries out Fuzzy processing
The image obtained afterwards.
7. a kind of image segmentation device, which is characterized in that described device includes:
Model obtains module, is configured as obtaining the pixel classifications model trained, the pixel classifications model is appointed for determining
The class indication of each pixel in one image, the class indication include first identifier and second identifier, and the first identifier is used
Belong to foreground area in the corresponding pixel of instruction, the second identifier is used to indicate corresponding pixel and belongs to background area；
Determining module is identified, is configured as target image to be processed being input in the pixel classifications model, based on described
Pixel classifications model determines the class indication of each pixel in the target image；
Area determination module is configured as the class indication according to pixel each in the target image, determines the target figure
The foreground area of picture and background area.
8. device according to claim 7, which is characterized in that described device further include:
Sample acquisition module is configured as obtaining the classification of each pixel in multiple sample images and the multiple sample image
Mark；
Training module is configured as the classification according to each pixel in the multiple sample image and the multiple sample image
Mark carries out model training, obtains the pixel classifications model.
9. a kind of image segmentation device, which is characterized in that described device includes:
Processor；
Memory for storage processor executable command；
Wherein, the processor is configured to:
The pixel classifications model trained is obtained, the pixel classifications model is used to determine the classification of each pixel in any image
Mark, the class indication include first identifier and second identifier, and the first identifier is used to indicate before corresponding pixel belongs to
Scene area, the second identifier are used to indicate corresponding pixel and belong to background area；
Target image to be processed is input in the pixel classifications model, be based on the pixel classifications model, determine described in
The class indication of each pixel in target image；
According to the class indication of pixel each in the target image, the foreground area and background area of the target image are determined
Domain.
10. a kind of non-transitorycomputer readable storage medium, when the instruction in the storage medium is by the processing of processing unit
When device executes, so that processing unit is able to carry out a kind of image partition method, which comprises
The pixel classifications model trained is obtained, the pixel classifications model is used to determine the classification of each pixel in any image
Mark, the class indication include first identifier and second identifier, and the first identifier is used to indicate before corresponding pixel belongs to
Scene area, the second identifier are used to indicate corresponding pixel and belong to background area；
Target image to be processed is input in the pixel classifications model, be based on the pixel classifications model, determine described in
The class indication of each pixel in target image；
According to the class indication of pixel each in the target image, the foreground area and background area of the target image are determined
Domain.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811252058.XA CN109285178A (en) | 2018-10-25 | 2018-10-25 | Image partition method, device and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811252058.XA CN109285178A (en) | 2018-10-25 | 2018-10-25 | Image partition method, device and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109285178A true CN109285178A (en) | 2019-01-29 |
Family
ID=65178440
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811252058.XA Pending CN109285178A (en) | 2018-10-25 | 2018-10-25 | Image partition method, device and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109285178A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109887045A (en) * | 2019-03-29 | 2019-06-14 | 南京维沃软件技术有限公司 | A kind of image display method and terminal device |
| CN110321865A (en) * | 2019-07-09 | 2019-10-11 | 北京字节跳动网络技术有限公司 | Head effect processing method and device, storage medium |
| CN110728632A (en) * | 2019-09-04 | 2020-01-24 | 北京奇艺世纪科技有限公司 | Image blurring processing method, image blurring processing device, computer device and storage medium |
| CN111539960A (en) * | 2019-03-25 | 2020-08-14 | 华为技术有限公司 | Image processing method and related device |
| CN111758118A (en) * | 2020-05-26 | 2020-10-09 | 蜂图科技有限公司 | Visual positioning method, device and equipment and readable storage medium |
| CN112270686A (en) * | 2020-12-24 | 2021-01-26 | 北京达佳互联信息技术有限公司 | Image segmentation model training method, image segmentation device and electronic equipment |
| CN112489053A (en) * | 2020-11-26 | 2021-03-12 | 深圳市艾合芯科智慧医疗有限公司 | Tongue image segmentation method and device and storage medium |
| CN113240611A (en) * | 2021-05-28 | 2021-08-10 | 中建材信息技术股份有限公司 | Foreign matter detection method based on picture sequence |
| CN113784207A (en) * | 2021-07-30 | 2021-12-10 | 北京达佳互联信息技术有限公司 | Video picture display method and device, electronic equipment and storage medium |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7783075B2 (en) * | 2006-06-07 | 2010-08-24 | Microsoft Corp. | Background blurring for video conferencing |
| CN102436637A (en) * | 2010-09-29 | 2012-05-02 | 中国科学院计算技术研究所 | Method and system for automatically segmenting hairs in head images |
| CN103262119A (en) * | 2010-12-08 | 2013-08-21 | 阿尔卡特朗讯公司 | Method and system for segmenting an image |
| CN104063865A (en) * | 2014-06-27 | 2014-09-24 | 小米科技有限责任公司 | Classification model creation method, image segmentation method and related device |
| CN105513105A (en) * | 2015-12-07 | 2016-04-20 | 天津大学 | Image background blurring method based on saliency map |
| CN106683147A (en) * | 2017-01-23 | 2017-05-17 | 浙江大学 | Method of image background blur |
| CN106920243A (en) * | 2017-03-09 | 2017-07-04 | 桂林电子科技大学 | The ceramic material part method for sequence image segmentation of improved full convolutional neural networks |
| CN107341805A (en) * | 2016-08-19 | 2017-11-10 | 北京市商汤科技开发有限公司 | Background segment and network model training, image processing method and device before image |
| CN107347166A (en) * | 2016-08-19 | 2017-11-14 | 北京市商汤科技开发有限公司 | Processing method, device and the terminal device of video image |
| CN107944499A (en) * | 2017-12-10 | 2018-04-20 | 上海童慧科技股份有限公司 | A kind of background detection method modeled at the same time for prospect background |
-
2018
- 2018-10-25 CN CN201811252058.XA patent/CN109285178A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7783075B2 (en) * | 2006-06-07 | 2010-08-24 | Microsoft Corp. | Background blurring for video conferencing |
| CN102436637A (en) * | 2010-09-29 | 2012-05-02 | 中国科学院计算技术研究所 | Method and system for automatically segmenting hairs in head images |
| CN103262119A (en) * | 2010-12-08 | 2013-08-21 | 阿尔卡特朗讯公司 | Method and system for segmenting an image |
| CN104063865A (en) * | 2014-06-27 | 2014-09-24 | 小米科技有限责任公司 | Classification model creation method, image segmentation method and related device |
| CN105513105A (en) * | 2015-12-07 | 2016-04-20 | 天津大学 | Image background blurring method based on saliency map |
| CN107341805A (en) * | 2016-08-19 | 2017-11-10 | 北京市商汤科技开发有限公司 | Background segment and network model training, image processing method and device before image |
| CN107347166A (en) * | 2016-08-19 | 2017-11-14 | 北京市商汤科技开发有限公司 | Processing method, device and the terminal device of video image |
| CN106683147A (en) * | 2017-01-23 | 2017-05-17 | 浙江大学 | Method of image background blur |
| CN106920243A (en) * | 2017-03-09 | 2017-07-04 | 桂林电子科技大学 | The ceramic material part method for sequence image segmentation of improved full convolutional neural networks |
| CN107944499A (en) * | 2017-12-10 | 2018-04-20 | 上海童慧科技股份有限公司 | A kind of background detection method modeled at the same time for prospect background |
Non-Patent Citations (1)
| Title |
|---|
| 周修庄 等，: "《视觉跟踪中的马氏链蒙特卡洛方法》", 30 June 2018, 北京邮电大学出版社， * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111539960B (en) * | 2019-03-25 | 2023-10-24 | 华为技术有限公司 | Image processing method and related device |
| CN111539960A (en) * | 2019-03-25 | 2020-08-14 | 华为技术有限公司 | Image processing method and related device |
| WO2020192692A1 (en) * | 2019-03-25 | 2020-10-01 | 华为技术有限公司 | Image processing method and related apparatus |
| CN109887045A (en) * | 2019-03-29 | 2019-06-14 | 南京维沃软件技术有限公司 | A kind of image display method and terminal device |
| CN110321865A (en) * | 2019-07-09 | 2019-10-11 | 北京字节跳动网络技术有限公司 | Head effect processing method and device, storage medium |
| CN110728632B (en) * | 2019-09-04 | 2022-07-12 | 北京奇艺世纪科技有限公司 | Image blurring processing method, image blurring processing device, computer device and storage medium |
| CN110728632A (en) * | 2019-09-04 | 2020-01-24 | 北京奇艺世纪科技有限公司 | Image blurring processing method, image blurring processing device, computer device and storage medium |
| CN111758118A (en) * | 2020-05-26 | 2020-10-09 | 蜂图科技有限公司 | Visual positioning method, device and equipment and readable storage medium |
| CN111758118B (en) * | 2020-05-26 | 2024-04-16 | 蜂图志科技控股有限公司 | Visual positioning method, device, equipment and readable storage medium |
| CN112489053A (en) * | 2020-11-26 | 2021-03-12 | 深圳市艾合芯科智慧医疗有限公司 | Tongue image segmentation method and device and storage medium |
| CN112489053B (en) * | 2020-11-26 | 2021-07-13 | 深圳市艾合芯科智慧医疗有限公司 | Tongue image segmentation method and device and storage medium |
| CN112270686A (en) * | 2020-12-24 | 2021-01-26 | 北京达佳互联信息技术有限公司 | Image segmentation model training method, image segmentation device and electronic equipment |
| CN112270686B (en) * | 2020-12-24 | 2021-03-16 | 北京达佳互联信息技术有限公司 | Image segmentation model training method, image segmentation device and electronic equipment |
| CN113240611A (en) * | 2021-05-28 | 2021-08-10 | 中建材信息技术股份有限公司 | Foreign matter detection method based on picture sequence |
| CN113240611B (en) * | 2021-05-28 | 2024-05-07 | 中建材信息技术股份有限公司 | Foreign matter detection method based on picture sequence |
| CN113784207A (en) * | 2021-07-30 | 2021-12-10 | 北京达佳互联信息技术有限公司 | Video picture display method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110189340B (en) | Image segmentation method and device, electronic equipment and storage medium | |
| CN109285178A (en) | Image partition method, device and storage medium | |
| CN108594997B (en) | Gesture skeleton construction method, device, equipment and storage medium | |
| EP3828769B1 (en) | Image processing method and apparatus, terminal and computer-readable storage medium | |
| CN110992493B (en) | Image processing method, device, electronic equipment and storage medium | |
| CN109829456A (en) | Image-recognizing method, device and terminal | |
| CN110139142A (en) | Virtual objects display methods, device, terminal and storage medium | |
| WO2019024717A1 (en) | Anti-counterfeiting processing method and related product | |
| CN110650379B (en) | Video abstract generation method and device, electronic equipment and storage medium | |
| CN108401124A (en) | The method and apparatus of video record | |
| CN109360222A (en) | Image partition method, device and storage medium | |
| CN112907725B (en) | Image generation, training of image processing model and image processing method and device | |
| CN110290421A (en) | Frame per second method of adjustment, device, computer equipment and storage medium | |
| CN110222789A (en) | Image-recognizing method and storage medium | |
| CN112287852B (en) | Face image processing method, face image display method, face image processing device and face image display equipment | |
| CN113763228B (en) | Image processing method, device, electronic equipment and storage medium | |
| CN110827195B (en) | Virtual article adding method and device, electronic equipment and storage medium | |
| CN111144365A (en) | Living body detection method, living body detection device, computer equipment and storage medium | |
| CN111723803B (en) | Image processing method, device, equipment and storage medium | |
| CN109558837A (en) | Face critical point detection method, apparatus and storage medium | |
| CN109978996B (en) | Method, device, terminal and storage medium for generating expression three-dimensional model | |
| CN110149517A (en) | Method, apparatus, electronic equipment and the computer storage medium of video processing | |
| CN110290426A (en) | Method, apparatus, equipment and the storage medium of showing resource | |
| CN110139143A (en) | Virtual objects display methods, device, computer equipment and storage medium | |
| CN110991457A (en) | Two-dimensional code processing method and device, electronic equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190129 |
|
| RJ01 | Rejection of invention patent application after publication |