CN109377454A - A kind of image processing method, device, equipment, storage medium and live broadcasting method - Google Patents
A kind of image processing method, device, equipment, storage medium and live broadcasting method Download PDFInfo
- Publication number
- CN109377454A CN109377454A CN201811117153.9A CN201811117153A CN109377454A CN 109377454 A CN109377454 A CN 109377454A CN 201811117153 A CN201811117153 A CN 201811117153A CN 109377454 A CN109377454 A CN 109377454A
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- processed
- color
- value
- 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
- 238000000034 method Methods 0.000 title claims abstract description 70
- 238000003672 processing method Methods 0.000 title claims abstract description 16
- 238000012545 processing Methods 0.000 claims abstract description 64
- 230000004927 fusion Effects 0.000 claims abstract description 32
- 230000002787 reinforcement Effects 0.000 claims abstract description 26
- 230000008569 process Effects 0.000 claims description 29
- 238000004590 computer program Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 12
- 238000004422 calculation algorithm Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 7
- 230000009885 systemic effect Effects 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 230000002087 whitening effect Effects 0.000 description 4
- 230000001815 facial effect Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 206010037660 Pyrexia Diseases 0.000 description 1
- 230000002146 bilateral effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 239000000686 essence Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000036632 reaction speed Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/73—Deblurring; Sharpening
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/77—Retouching; Inpainting; Scratch removal
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20172—Image enhancement details
- G06T2207/20192—Edge enhancement; Edge preservation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
This application discloses a kind of image processing method, device, equipment, storage medium and live broadcasting methods, and described method includes following steps: obtaining the weighted value of each pixel of image to be processed；Fuzzy Processing is carried out to the non-edge in image to be processed according to the weighted value of each pixel and obtains the first image；Reinforcement is carried out to the fringe region in the first image to handle to obtain the second image；Second image and the image to be processed are subjected to linear fusion and obtain third image.It aims to solve the problem that in the prior art, mill skin treated the details such as missing image dermatoglyph and hair, the blurred image technical problem of generation.
Description
Technical field
This application involves internet area more particularly to field of image processings.
Background technique
U.S. face is the very popular mode in image procossing, for portrait U.S. face, generally comprises mill skin, whitening etc.
Reason.Wherein, mill skin is to keep character facial finer and smoother by eliminating in image personage spot, flaw and variegated etc. on the face.Generally
Image will include marginal portion and non-edge part, and by taking face-image as an example, marginal portion includes the face contour of people, face wheel
Wide, hair and dermatoglyph etc., non-edge part can be the other parts in addition to edge contour.Treated for mill skin at present
The details such as missing image dermatoglyph and hair cause image fuzzy, and give a kind of false display effect of people.
Summary of the invention
The application provides a kind of image processing method, device, equipment, storage medium and live broadcasting method, it is intended to solve existing
In technology, mill skin treated the details such as missing image dermatoglyph and hair, the blurred image technical problem of generation.
The application's in a first aspect, providing a kind of image processing method, include the following steps:
Obtain the weighted value of each pixel of image to be processed；
Fuzzy Processing is carried out to the non-edge in image to be processed according to the weighted value of each pixel and obtains the
One image；
Reinforcement is carried out to the fringe region in the first image to handle to obtain the second image；
Second image and the image to be processed are subjected to linear fusion and obtain third image.
In some instances, the weighted value according to each pixel to the non-edge in image to be processed into
Row Fuzzy Processing obtains the first image, comprising:
According to the weighted value of each pixel, the color-weighted average value and face of each pixel of image to be processed are obtained
Color quadratic sum weighted average；
According to the color-weighted average value, color quadratic sum weighted average and adjusting parameter of the image to be processed,
Obtain Steerable filter treated the first image.
In some instances, the weighted value according to each pixel, obtains the face of each pixel of image to be processed
Color weighted average and color quadratic sum weighted average, comprising:
According to the weighted value of each pixel, each pixel of image to be processed in a first direction color-weighted is obtained
Average value and color quadratic sum weighted average；
According to the color-weighted average value and color quadratic sum weighted average of each pixel in a first direction,
And the weighted value of each pixel, obtain the color-weighted average value of each pixel of image to be processed in a second direction with
And color quadratic sum weighted average, wherein first direction is vertical with second direction.
In some instances, the weighted value according to each pixel, obtains the face of each pixel of image to be processed
Color weighted average and color quadratic sum weighted average, are obtained by following formula:
Wherein, the r is windows radius；
(i, j) characterizes the central pixel point of the window；
P (x, j), p (i, j) and p (i, y) are respectively the color of pixel (x, j), (i, j) and (i, y) in image to be processed
Value；
weightxFor the weighted value for the pixel that the distance of Range Profile vegetarian refreshments (i, j) in a first direction is x；
weightyFor the weighted value for the pixel that the distance of Range Profile vegetarian refreshments (i, j) in a second direction is y；
meanI1(i, j) and meanI1(i, y) is respectively pixel (i, j) and (i, y) in a first direction color-weighted
Average value；
meanI2(i, j) is the color-weighted average value of pixel (i, j) in a second direction；
meanII1(i, j) and meanII1(i, y) is respectively that the color of pixel (i, j) and (i, y) in a first direction is flat
Side and weighted average；
meanII2(i, j) is the color quadratic sum weighted average of pixel (i, j) in a second direction.
In some instances, the fringe region in the first image carries out reinforcement and handles to obtain the second image,
Include:
High contrast reservation process is carried out to the first image and obtains the second image.
In some instances, the high contrast reservation process includes improved high contrast reservation process；
The improved high contrast reservation process includes:
High contrast reservation process is carried out to the image to be processed and the first image and obtains the 4th image；
Result reinforcement is carried out to the 4th image to handle to obtain the 5th image；
High contrast reservation process is carried out to the 5th image and the image to be processed and obtains the second image.
In some instances, second image and the image to be processed are subjected to linear fusion and obtain third image,
Include:
It handles second image procossing progress Gaussian Blur to obtain the 6th image；
6th image and the image to be processed are subjected to linear fusion and obtain third image.
In some instances, the method also includes steps:
Image to be processed is carried out highlighting processing and obtains the 7th image；
7th image and the third linearity fusion treatment are obtained into the 8th image.
In some instances, the image to be processed includes each frame live video frame in live video stream, the method
Apply to main broadcaster's client of internet live streaming.
The second aspect of the application provides a kind of live broadcasting method, the method includes the steps:
Live video stream to be processed is obtained from main broadcaster's client, the live video stream includes several live video frames；
Obtain the weighted value of each pixel of live video frame；
The non-edge in the live video frame Fuzzy Processing is carried out according to the weighted value of each pixel to obtain
To the first image；
Reinforcement is carried out to the fringe region in the first image to handle to obtain the second image；
Second image and the image to be processed are subjected to linear fusion and obtain third image；
Institute's third image is sent to main broadcaster's client and specified spectator client is shown.
The third aspect of the application, provides a kind of image processing apparatus, and described device includes:
Processing module, for obtaining the weighted value of each pixel of image to be processed；
Fuzzy Processing is carried out to the non-edge in image to be processed according to the weighted value of each pixel and obtains the
One image；Reinforcement is carried out to the fringe region in the first image to handle to obtain the second image；
Fusion Module obtains third image for second image and the image to be processed to be carried out linear fusion.
The fourth aspect of the application, provides a kind of live broadcast device, and described device includes:
Module is obtained, for obtaining live video stream to be processed from main broadcaster's client, if the live video stream includes
Dry live video frame；
Processing module, for obtaining the weighted value of each pixel of live video frame according to one phase algorithm of Gauss；According to institute
Non-edge in the live video frame is carried out Fuzzy Processing and obtains the first image by the weighted value for stating each pixel；To institute
It states the fringe region in the first image and carries out reinforcement and handle to obtain the second image；By second image and the image to be processed
It carries out linear fusion and obtains third image；
Distribution module, for institute's third image being sent to main broadcaster's client and specified spectator client is shown
Show.
According to the 5th of the application the aspect, a kind of electronic equipment is provided, the equipment includes:
Memory, processor and storage are on a memory and the computer program that can run on a processor, wherein described
Processor realizes the operation such as above-mentioned first aspect and second aspect any one the method when executing described program.
According to the 7th of the application the aspect, a kind of storage medium is provided, is stored thereon with program, described program is by processor
The step of above-mentioned first aspect and second aspect any one the method are realized when execution.
The weighted value that the application passes through acquisition each pixel of image to be processed；It is treated according to the weighted value of each pixel
Non-edge in processing image carries out Fuzzy Processing and obtains the first image, so that the first image after fuzzy remains part
The information of fringe region；Non-edge progress Fuzzy Processing is obtained certainly for the image to be processed of some complexity first
The display effect of image is ideal not enough, then carries out reinforcement to the fringe region in the first image again and handles to obtain second
Image, the comparison of fringe region and non-edge is strengthened in second image, so that many detailed information are retained,
Second image and the image to be processed are subjected to linear fusion, the third image made not only remains basic thin
Information, such as skin quality and hair etc. are saved, and gives the effect of true nature.
Detailed description of the invention
Fig. 1 is a kind of flow chart of the image processing method schematically shown in the embodiment of the present application；
Fig. 2 is the specific flow chart of the S100 and S110 schematically shown in the embodiment of the present application a kind of；
What Fig. 3 was that the embodiment of the present application is schematically shown a kind of obtains showing for the weighted value of each pixel in image to be processed
It is intended to；
Fig. 4 is a kind of showing for the color-weighted average value for acquisition pixel (i, j) that the embodiment of the present application is schematically shown
It is intended to；
Fig. 5 is the flow chart that a kind of improved high contrast schematically shown in the embodiment of the present application retains；
Fig. 6 a is a kind of specific flow chart for S130 that the embodiment of the present application is schematically shown；
Fig. 6 b is the schematic diagram for another image procossing that the embodiment of the present application is schematically shown；
Fig. 7 a is the flow chart for another image procossing that the embodiment of the present application is schematically shown；
Fig. 7 b is the schematic diagram for another image procossing that the embodiment of the present application is schematically shown；
Fig. 8 is the schematic diagram for the live scene that the embodiment of the present application is schematically shown；
Fig. 9 is the flow chart of one of the embodiment of the present application live broadcasting method；
Figure 10 is the flow chart of another live broadcasting method in the embodiment of the present application；
Figure 11 is the schematic diagram of an electronic equipment in the embodiment of the present application；
Figure 12 is the schematic diagram of the image processing apparatus in the embodiment of the present application；
Figure 13 is the schematic diagram of the server apparatus in the embodiment of the present application；
Figure 14 is the schematic diagram of the live broadcast device in the embodiment of the present application.
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 embodiments consistent with the application.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 application.
It is only to be not intended to be limiting the application merely for for the purpose of describing particular embodiments in term used in this application.
It is also intended in the application and the "an" of singular used in the attached claims, " described " and "the" including majority
Form, unless the context clearly indicates other meaning.It is also understood that term "and/or" used herein refers to and wraps
It may be combined containing one or more associated any or all of project listed.
It will be appreciated that though various information, but this may be described using term first, second, third, etc. in the application
A little information should not necessarily be limited by these terms.These terms are only used to for same type of information being distinguished from each other out.For example, not departing from
In the case where the application range, the first information can also be referred to as the second information, and similarly, the second information can also be referred to as
One information.Depending on context, word as used in this " if " can be construed to " ... when " or " when ...
When " or " in response to determination ".
U.S. face is the very popular mode in image procossing, for portrait U.S. face, generally comprises mill skin, whitening etc.
Reason.Wherein, mill skin is by eliminating in image personage spot, flaw and variegated etc. on the face, so that character facial is finer and smoother.But
It is to grind skin treated details such as missing image dermatoglyph and hair at present, causes image fuzzy, and untrue to people's one kind
Display effect.
In order to solve the above-mentioned technical problem, the embodiment of the present application provides a kind of image processing method, device, equipment, storage
Medium and live broadcasting method.
A kind of flow chart of the image processing method schematically shown referring to Fig.1 for the embodiment of the present application, the method packet
Include step:
S100: the weighted value of each pixel of image to be processed is obtained.
S110: Fuzzy Processing is carried out to the non-edge in image to be processed according to the weighted value of each pixel and is obtained
To the first image.
S120: reinforcement is carried out to the fringe region in the first image and handles to obtain the second image.
S130: second image and the image to be processed are subjected to linear fusion and obtain third image.
General pattern will include fringe region and non-edge, by taking face-image as an example, what the embodiment of the present application proposed
Fringe region includes the face contour of people, face profile, hair and dermatoglyph etc., and non-edge can be except edge wheel
Other parts outside exterior feature.
Image processing method described in Fig. 1, by the weighted value for obtaining each pixel of image to be processed；According to each picture
The weighted value of vegetarian refreshments carries out Fuzzy Processing to the non-edge in image to be processed and obtains the first image, so that the after fuzzy
One image remains the information in part edge region；Mould is carried out to non-edge certainly for the image to be processed of some complexity
The display effect for the first image that paste is handled is ideal not enough, then carries out again to the fringe region in the first image
Reinforcement handles to obtain the second image, and the comparison of fringe region and non-edge is strengthened in second image, so that being permitted
More detailed information are retained, and second image and the image to be processed are carried out linear fusion, the third image made
Basic detailed information, such as skin quality and hair etc. are not only remained, and gives the effect of true nature.
In some instances, the S110 is right after specifically may is that the weighted value for obtaining each pixel of image to be processed
Image to be processed obtains the first image by protecting side Fuzzy Processing.Guarantor side Fuzzy Processing may include surface blur processing,
Bilateral filtering processing or Steerable filter processing etc..
In some instances, guarantor's side mode paste can be improved Steerable filter, referring to Fig. 2, the improved guiding
The process of filtering processing is as follows:
S210: the weighted value of each pixel of image to be processed is obtained using one phase algorithm of Gauss.
The improved Steerable filter is different from weighted value distribution mechanism in traditional Steerable filter processing, in the present embodiment
In, the weighted value of each pixel is obtained according to one order algorithm of Gauss.Specifically, the weighted value of each pixel can be according to following public affairs
Formula is calculated:
Wherein, x is pixel to be calculated in window at a distance from central pixel point；
weightxFor the weighted value for the pixel for being x apart from window center point distance.
For example, pixel can characterize its uniqueness with (i, j), it is in image that wherein i and j, which respectively represents the pixel,
The pixel of i-th row's jth column.By taking simple one-dimensional window as an example, Parameter Map 3, the windows radius of preset window is (for example, r),
In image 300 to be processed, the central pixel point 311 of window 310 is pixel (i, j), the pixel 312 in the window 310
(i, j0) is with the 311 distance x of central pixel point | j-j0 |, utilize pixel in distance x and formula (1) available window
The weighted value of 312 (i, j0) is put, the weighted value of each pixel available formula (1) is calculated in window.
Traditional Steerable filter processing, the weighted value of each pixel is obtained by way of mean value, even there is n in window
Pixel, then the weighted value of each pixel is 1/n, and the present embodiment obtains each pixel in window by one order algorithm of Gauss
Weighted value, the pixel weighted value for keeping distance center closer is bigger, and the remoter pixel weighted value of distance center is smaller, in utilization
The first image after stating the Steerable filter that weight distribution mechanism obtains can be realized preferably while " mill skin ", retain side
Edge region, such as hair and dermatoglyph.
S220: according to the weighted value of each pixel, the color-weighted average value of each pixel of image to be processed is obtained
And color quadratic sum weighted average.
In one example, obtained in window after the weighted value of each pixel by one order algorithm of Gauss or other modes, root
According to the weighted value of each pixel, the color-weighted average value and quadratic sum for calculating pixel all in the window add
Weight average value.It should be noted that the color value may include gray value, rgb value or YUV value；When color value is gray value
When, each window is only calculated once color-weighted average value and color quadratic sum weighted average；When color value be rgb value or
YUV value, since each pixel is characterized by three components, so each window calculation is directed to each component respectively and calculates color
Weighted average and color quadratic sum weighted average.
In some examples, S220 can be realized according to following formula, wherein using image to be processed as the guidance of Steerable filter
Figure:
Wherein, the r is windows radius；
(i, j) characterizes the central pixel point of the window；
P (x, j), p (i, j) and p (i, y) are respectively the color of pixel (x, j), (i, j) and (i, y) in image to be processed
Value；
weightxFor the weighted value for the pixel that the distance of Range Profile vegetarian refreshments (i, j) in a first direction is x；
weightyFor the weighted value for the pixel that the distance of Range Profile vegetarian refreshments (i, j) in a second direction is y；
meanI1(i, j) and meanI1(i, y) is respectively pixel (i, j) and (i, y) in a first direction color-weighted
Average value；
meanI2(i, j) is the color-weighted average value of pixel (i, j) in a second direction, i.e. pixel (i, j) is final
Color-weighted average value；
meanII1(i, j) and meanII1(i, y) is respectively that the color of pixel (i, j) and (i, y) in a first direction is flat
Side and weighted average；
meanII2(i, j) is the color quadratic sum weighted average of pixel (i, j) in a second direction, i.e. pixel
(i, j) final color quadratic sum weighted average.
In one example, color-weighted average to calculate pixel 411 (i, j) in image 400 to be processed referring to Fig. 4
For value, with reference first to the left figure of Fig. 4, one-dimensional window in a first direction (such as laterally) is obtained according to one order algorithm of Gauss
The weighted value of each pixel in 410 calculates the middle imago of the one-dimensional window 410 using formula (2) according to the weighted value
The mean of vegetarian refreshments 411I1(i, j), by taking r=1 as an example, the color-weighted average value of central pixel point 411 in a first direction is as follows:
meanI1(i, j)=weighti-1*p(i-1,j)+weighti*p(i,j)+weighti+1*p(i+1,j)；Then join
According to the right figure of Fig. 4, each pixel in the one-dimensional window 420 of second direction (such as vertical) is obtained according to one order algorithm of Gauss
Weighted value and above-mentioned meanI1(i, j) calculates the central pixel point 411 of the one-dimensional window 420 using formula (3)
meanI2(i, j), by taking r=1 as an example, the color-weighted average value of central pixel point 411 in a second direction is as follows:
meanI2(i, j)=weightj-1*meanI1(i,j-1)+weightj*meanI1(i,j)+weightj+1*meanI1
(i,j+1)；It should be understood that in some examples can it is first vertical again laterally, the specific first direction of the unlimited system of the application and
The specific direction of second direction is also possible to any other direction other than horizontal and vertical, in some examples, described first
Direction and second direction are mutually perpendicular to.
It, can (every shifting moves a step, and window is by the mobile default step of first direction by preset first step-length in some examples
It is long), whole picture image to be processed is traversed, the mean of each pixel in image to be processed is obtainedI1(i, j) and meanII1(i, j), then lead to
Preset second step-length (window is by the mobile preset step-length of second direction) is crossed, whole picture image to be processed is traversed, obtains figure to be processed
The mean of each pixel as inI2(i, j) and meanII2(i,j)。
Through the foregoing embodiment, can the biggish calculation amount for reducing weighted average and quadratic sum weighted average, with
For calculating color-weighted average value, different from traditional guarantor side Fuzzy Processing, to the rectangular window that a radius is r
(2r+1) * (2r+1) a pixel seeks color-weighted average value, and first radius is the one of r to the application in a first direction (such as laterally)
2r+1 pixel of rank window seeks color-weighted average value meanI1(i, j), then with meanI1Based on (i, j), with first
2r+1 pixel of the single order window that the vertical second direction in direction (such as longitudinal) radius is r seeks color-weighted average value
meanI2(i, j), so that calculation amount is reduced to 2* (2r+1) by (2r+1) * (2r+1) in traditional technology.
S230: it according to the weighted average, quadratic sum weighted average and adjusting parameter of the image to be processed, obtains
First image.
In some examples, this step can be realized by following formula, formula is as follows:
Var=meanII2(i, j)-meanI2(i, j) * meanI2(i, j) (6)
μA=var/ (var+ δ) (7)
μB=meanI2(i,j)-μA*meanI2(i,j) (8)
Q (i, j)=μA*p(i,j)+μB (9)
Wherein, δ is adjusting parameter；
P (i, j) is the color value of the pixel (i, j) in image to be processed；
meanI1(i, j) is the color-weighted average value of the pixel (i, j) in a first direction；
meanI2(i, j) is the color-weighted average value of the pixel (i, j) in a second direction, i.e. pixel (i, j)
Final color-weighted average value；
meanII1(i, j) is the color quadratic sum weighted average of the pixel (i, j) in a first direction；
meanII2(i, j) is the color quadratic sum weighted average of the pixel (i, j) in a second direction, i.e. pixel
The final color quadratic sum weighted average of point (i, j)；
Q (i, j) is the color value of pixel (i, j) described in the first image.
In some instances, the S120 specifically may is that the first image for obtaining S110 by high contrast reservation
Reason obtains the second image.
The high contrast reservation that the embodiment of the present application proposes can use following formula and obtain:
D (i, j)=p (i, j)-q (i, j)+α (10)
Wherein, d (i, j) is the color value of pixel (i, j) in the second image；
P (i, j) is the color value of pixel (i, j) in image to be processed；
Q (i, j) is the color value of pixel (i, j) in the first image；
α is the coefficient of systemic presupposition, described when step S120 is that GPU is executed in all formula of the embodiment of the present application
α can be 0.5；When step S120 is that GPU is executed, the α is 128.
In order to further retain fringe region, the application improves traditional high contrast reservation process, specifically,
It may also is that referring to Fig. 5, the S120
S510: high contrast reservation process is carried out to the image to be processed and the first image and obtains the 4th image.
In this step, the color value of each pixel can be obtained by following formula in the 4th image:
d(i,j)1=p (i, j)-q (i, j)+α (11)
Wherein, d (i, j)1For the color value of pixel (i, j) in the 4th image；
P (i, j) is the color value of pixel (i, j) in image to be processed；
Q (i, j) is the color value of pixel (i, j) in the first image；
α is the coefficient of systemic presupposition, described when step S120 is that GPU is executed in all formula of the embodiment of the present application
α can be 0.5；When step S120 is that GPU is executed, the α is 128.
S520: result reinforcement is carried out to the 4th image and handles to obtain the 5th image.
In this step, the color value of the 5th each pixel of image:
Wherein, the β is systemic presupposition constant, and when step S120 is that CPU is executed, the β is 255；As step S120
When executing for GPU, the β is 1；
α is the coefficient of systemic presupposition, described when step S120 is that GPU is executed in all formula of the embodiment of the present application
α can be 0.5；When step S120 is that GPU is executed, the α is 128；
d(i,j)1For the color value of pixel (i, j) in the 4th image；
d(i,j)2For the color value of pixel (i, j) in the 5th image.
S530: high contrast reservation process is carried out to the 5th image and the image to be processed, obtains the second image.
In this step, the color value of each pixel obtains according to the following formula in the second image:
D (i, j)=d (i, j)2-p(i,j)+α (13)
Wherein, p (i, j) is the color value of pixel (i, j) in image to be processed；
D (i, j) is the color value of pixel (i, j) in the second image.
Retain compared to traditional high contrast, the improved high contrast reservation process proposed by the embodiment of the present application makes
The fringe region for obtaining image is more clear, and preferably remains the detail edges such as texture and hair region.
In some instances, the algorithm of linear fusion described in the embodiment of the present application S130 can be is obtained by following formula
It arrives:
Dest (i, j)=p (i, j)+λ d (i, j)-η (14)
Wherein, p (i, j) is the color value of pixel (i, j) in image to be processed；
Dest (i, j) is the color value of pixel (i, j) in third image；
D (i, j) is the color value of pixel (i, j) in the second image；
λ and η is preset coefficient.
In practical application, the third image obtained by foregoing embodiments the method may still exist in some flaws
Defect, such as: by the third image that earlier figures 2 or/and Fig. 5 the embodiment described obtain although be effectively maintained dermatoglyph and
The details area at the edges such as hair, but a large amount of it was found that, there are there are some/a small amount of green in some third images
Color spot point；Again for example: although the third image obtained by Fig. 5 the embodiment described is also effectively maintained dermatoglyph and hair
The details area at equal edges, but may exist fringe region and non-edge contrast it is excessive caused by feeling of unreality etc.
Problem.To solve the above-mentioned problems, referring to Fig. 6 a, in some instances, the S130 be may also is that
S610: the 6th image is obtained after second image is carried out Gaussian Blur；
S620: the 6th image is merged to obtain third image with the linearity to be processed.
Specifically, referring to Fig. 6 b, executing improved Steerable filter in a specific example to image to be processed and handling
To the first image；Improved high contrast is executed to the first image to retain to obtain the second image；Gauss is executed to second image
Fuzzy Processing obtains the 6th image；6th image is merged to obtain third image with the linearity to be processed.
It is marginal zone in order to obtain to the purpose that the fringe region of image reinforce processing it is understood that in traditional technology
The image that domain is reinforced, such as high contrast reservation process, in order not to weaken the purpose of high contrast reservation process, ordinary circumstance
Under, Gaussian Blur, will not be after high contrast reservation before high contrast reservation process, and still, the present embodiment breaks tradition
Thinking carries out the second image that reinforcement is handled by the fringe region to image, then carries out a Gaussian Blur processing, makes
The part flaw of image must be not only eliminated, and increases the mildness of fringe region and non-edge transition in image,
So that the third image more true nature that processing obtains.
In some instances, referring to Fig. 7 a, the embodiment of the present application described image processing method further includes step step:
S710: carry out highlighting processing obtaining the 7th image to image to be processed；
S720: the 8th image will be obtained after treated the image and the 7th image co-registration processing.
Specifically, executing the above-mentioned steps S710 and step S720, institute after can be the instruction that user triggers " whitening "
The embodiment for the image co-registration processing that image co-registration processing is referred in above-mentioned steps S130 is stated, details are not described herein.
Specifically, referring to Fig. 7 b, the process of specific image processing method may is that detection in a specific example
After " mill skin " instruction triggered to user, improved Steerable filter is executed to image to be processed and handles to obtain the first image；To
One image executes improved high contrast and retains to obtain the second image；Gaussian Blur is executed to second image to handle to obtain the 6th
Image；6th image is merged to obtain third image with the linearity to be processed；Detect that user triggers " whitening "
Instruction after, carry out highlighting processing obtaining the 7th image to image to be processed；It will treated image and the 7th figure
As obtaining the 8th image after fusion treatment.
In some instances, the method can be used in live streaming field, referring to Fig. 8, not apply implementing to exemplify straight
It broadcasts schematic diagram of a scenario, the first spectator client and the second spectator client and main broadcaster's client is respectively installed in electronic equipment
810, on 820 and 830, main broadcaster's client can call camera recorded video, shooting photo or/and by screen capture etc.
Other modes make live video stream, are then sent to server 800 by network.Server 800 is straight for providing internet
The background service broadcast, such as the corresponding relationship of each main broadcaster's client and spectator client is saved, it is broadcast live the distribution of video flowing,
And distribution of interactive message etc., when the first spectator client and the second spectator client and main broadcaster's client are in same direct broadcasting room
Interior, the live video stream of desired displaying can be shared in the first spectator client of same direct broadcasting room and by main broadcaster's client
Two spectator clients, for the user of the first spectator client and the second spectator client viewing.The live video stream includes
Audio data and several frame live videos.
" main broadcaster's client " " spectator client " that the embodiment of the present application proposes can refer to installation on an electronic device soft
Part, in some cases, the live streaming client and spectator client are integrated on a software, when the identity of user is main broadcaster
When, which can be referred to as main broadcaster's client, and when the identity of user is spectators, which is referred to as spectators client
End.
If described image processing method is by main broadcaster's client executing, a kind of realization step reference Fig. 9 of live broadcasting method, part
Steps are as follows:
S900: the weighted value of main broadcaster's client acquisition each pixel of live video frame；
S910: the non-edge in the live video frame is carried out by fuzzy place according to the weighted value of each pixel
Reason obtains the first image；
S920: main broadcaster's client carries out reinforcement to the fringe region in the first image and handles to obtain the second image；
S930: second image and the image to be processed are carried out linear fusion and obtain third figure by main broadcaster's client
Picture；
S940: the third image is sent to server by main broadcaster's client, so that the server is by the third figure
As being distributed to corresponding spectator client.
It should be understood that main broadcaster's client passes through step S900, S910, S920 and S930 for live video to be processed
Each live video frame in stream is processed into third image, and the video flowing that the third image stabbed by several frame different times is formed
It is sent to server.
It should be noted that the specific implementation of each step is referred to previous embodiment in Fig. 9, details are not described herein again.
In practical applications, it since the equipment performance where certain main broadcaster's clients is poor, is set where main broadcaster's client
The method of standby upper execution described image processing may cause the place equipment reaction speed due to the excessive generation of occupancy to CPU compared with
The problems such as slow and fever, in some instances, executes described image processing by the server in live streaming to solve the above-mentioned problems
Method, referring to Fig.1 0, a kind of live streaming flow chart proposed for the embodiment of the present application:
S1000: server obtains live video stream to be processed from main broadcaster's client；
The live video stream includes several live video frames；
S1010: the weighted value of server acquisition each pixel of live video frame；
S1020: server according to the weighted value of each pixel by the non-edge in the live video frame into
Row Fuzzy Processing obtains the first image；
S1030: server carries out reinforcement to the fringe region in the first image and handles to obtain the second image；
S1040: second image and the image to be processed are carried out linear fusion and obtain third image by server；
S1050: the third image is sent to specified spectator client and main broadcaster's client by server；
S1060: main broadcaster's client will acquire the third image and show on the screen of the device.
In this step, the third image is coated on the live video frame in live video stream to be processed, with
Third image after making the user of main broadcaster's client can see that image procossing.
Pass through live broadcasting method described in Figure 10, it is possible to reduce the calculation amount of equipment where main broadcaster's client.
It should be noted that the specific implementation of each step is referred to previous embodiment in Figure 10, details are not described herein again.
In some instances, when determining the executing subject of image processing method of the embodiment of the present application proposition, by main broadcaster
The performance of equipment where client judges itself executes described image processing side by server if performance is more than preset condition
Method, concrete mode are referred to step described in Figure 10；If performance is less than preset condition, as described in main broadcaster's client executing
Image processing method, concrete mode are referred to step described in Fig. 9.
The application image processing method embodiment can also pass through hardware or software and hardware combining by software realization
Mode realize.It taking software implementation as an example, is the place by client device where it as the device on a logical meaning
Computer program instructions corresponding in nonvolatile memory are read into memory what operation was formed by reason device.From hardware view
Speech is as shown in figure 11 a kind of hardware structure diagram of electronic equipment where the application image processing apparatus, in addition to shown in Figure 11
The usual root of electronic equipment except processor, memory, network interface and nonvolatile memory, in embodiment where device
According to the actual functional capability of the equipment, it can also include other hardware, such as camera, microphone etc., this is repeated no more.The place
Reason device is used to carry out following operation:
Obtain the weighted value of each pixel of image to be processed；
Fuzzy Processing is carried out to the non-edge in image to be processed according to the weighted value of each pixel and obtains the
One image；
Reinforcement is carried out to the fringe region in the first image to handle to obtain the second image；
Second image and the image to be processed are subjected to linear fusion and obtain third image.
It should be noted that some images to be processed include each frame live video frame in live video stream, it is described
Electronic equipment is the electronic equipment where main broadcaster's client, and the electronic equipment can be smart phone, Intelligent flat, notebook
Computer, desktop computer and vehicle-mounted terminal equipment etc., the application do not limit the type of electronic equipment.
Referring to Figure 12, a kind of frame diagram for image processing apparatus 1200 that the embodiment of the present application is schematically shown, the figure
As processing unit 1200 includes:
Processing module 1210, for obtaining the weighted value of each pixel of image to be processed；According to the power of each pixel
Weight values carry out Fuzzy Processing to the non-edge in image to be processed and obtain the first image；To the edge in the first image
Region carries out reinforcement and handles to obtain the second image；
Fusion Module 1220 obtains third figure for second image and the image to be processed to be carried out linear fusion
Picture.
In some instances, processing module 1210 is specifically used for:
According to the weighted value of each pixel, the color-weighted average value and face of each pixel of image to be processed are obtained
Color quadratic sum weighted average；
According to the color-weighted average value, color quadratic sum weighted average and adjusting parameter of the image to be processed,
Obtain Steerable filter treated the first image.
In some instances, the weighted value according to each pixel, obtains the face of each pixel of image to be processed
When color weighted average and color quadratic sum weighted average, processing module 1210 is specifically used for:
According to the weighted value of each pixel, each pixel of image to be processed in a first direction color-weighted is obtained
Average value and color quadratic sum weighted average；
According to the color-weighted average value and color quadratic sum weighted average of each pixel in a first direction,
And the weighted value of each pixel, obtain the color-weighted average value of each pixel of image to be processed in a second direction with
And color quadratic sum weighted average, wherein first direction is vertical with second direction.
In some instances, the weighted value according to each pixel, obtains the face of each pixel of image to be processed
Color weighted average and color quadratic sum weighted average, are obtained by following formula:
Wherein, the r is windows radius；
(i, j) characterizes the central pixel point of the window；
P (x, j), p (i, j) and p (i, y) are respectively the color of pixel (x, j), (i, j) and (i, y) in image to be processed
Value；
weightxFor the weighted value for the pixel that the distance of Range Profile vegetarian refreshments (i, j) in a first direction is x；
weightyFor the weighted value for the pixel that the distance of Range Profile vegetarian refreshments (i, j) in a second direction is y；
meanI1(i, j) and meanI1(i, y) is respectively pixel (i, j) and (i, y) in a first direction color-weighted
Average value；
meanI2(i, j) is the color-weighted average value of pixel (i, j) in a second direction；
meanII1(i, j) and meanII1(i, y) is respectively that the color of pixel (i, j) and (i, y) in a first direction is flat
Side and weighted average；
meanII2(i, j) is the color quadratic sum weighted average of pixel (i, j) in a second direction.
In some instances, the fringe region in the first image carries out reinforcement and handles to obtain the second image
When, processing module 1210 is specifically used for:
High contrast reservation process is carried out to the first image and obtains the second image.
In some instances, the high contrast reservation process includes improved high contrast reservation process；
The improved high contrast reservation process includes:
High contrast reservation process is carried out to the image to be processed and the first image and obtains the 4th image；
Result reinforcement is carried out to the 4th image to handle to obtain the 5th image；
High contrast reservation process is carried out to the 5th image and the image to be processed, obtains the second image.
In some instances, described that 4th image progress result reinforcement is handled to obtain the 5th image, it can pass through
Following formula obtains:
Wherein, the β is systemic presupposition constant；
The α is preset constant；
d(i,j)1For the color value of pixel (i, j) in the 4th image；
D (i, j) 2 is the color value of pixel (i, j) in the 5th image.
In some instances, second image and the image to be processed are subjected to linear fusion and obtain third image
When, Fusion Module 1220 is specifically used for:
It handles second image procossing progress Gaussian Blur to obtain the 6th image；
6th image and the image to be processed are subjected to linear fusion and obtain third image.
In some instances, described device further includes highlighting module, is used for:
Image to be processed is carried out highlighting processing and obtains the 7th image；
7th image and the third linearity fusion treatment are obtained into the 8th image.
In some instances, the image to be processed includes each frame live video frame in live video stream.
It as shown in figure 13, is a kind of hardware structure diagram of the application live broadcast device place server apparatus, in addition to Figure 13 institute
Processor, memory, network interface and the nonvolatile memory shown can also include other hardware.The processor by with
In performing the following operations:
Live video stream to be processed is obtained from main broadcaster's client, the live video stream includes several live video frames；
Obtain the weighted value of each pixel of live video frame；
The non-edge in the live video frame Fuzzy Processing is carried out according to the weighted value of each pixel to obtain
To the first image；
Reinforcement is carried out to the fringe region in the first image to handle to obtain the second image；
Second image and the image to be processed are subjected to linear fusion and obtain third image；
Institute's third image is sent to main broadcaster's client and specified spectator client is shown.
Referring to Figure 14, the frame diagram for a kind of live broadcast device 1400 that the embodiment of the present application is schematically shown, at described image
Managing device 1400 includes:
Module 1410 is obtained, for obtaining live video stream to be processed, the live video stream packet from main broadcaster's client
Include several live video frames；
Processing module 1420, for obtaining the weighted value of each pixel of live video frame；According to the power of each pixel
Non-edge in the live video frame is carried out Fuzzy Processing and obtains the first image by weight values；To in the first image
Fringe region carries out reinforcement and handles to obtain the second image；Second image is carried out linear fusion with the image to be processed to obtain
To third image；
Distribution module 1430, for by institute's third image be sent to main broadcaster's client and specified spectator client into
Row display.
In the embodiment of the present application, computer readable storage medium can be diversified forms, for example, in different examples
In, the machine readable storage medium may is that RAM (Radom Access Memory, random access memory), volatile deposit
Reservoir, nonvolatile memory, flash memory, memory driver (such as hard disk drive), solid state hard disk, any kind of storage dish
(such as CD, dvd) perhaps similar storage medium or their combination.Special, described computer-readable medium
Can also be paper or other be suitably capable of the medium of print routine.Using these media, these programs can be passed through
The mode of electricity gets (for example, optical scanner), can be compiled, be explained and processing in an appropriate manner, then can be by
It stores in computer media.
The function of each unit and the realization process of effect are specifically detailed in the above method and correspond to step in above-mentioned apparatus
Realization process, details are not described herein.
For device embodiment, since it corresponds essentially to embodiment of the method, so related place is referring to method reality
Apply the part explanation of example.The apparatus embodiments described above are merely exemplary, wherein described be used as separation unit
The unit of explanation may or may not be physically separated, and component shown as a unit can be or can also be with
It is not physical unit, it can it is in one place, or may be distributed over multiple network units.It can be according to actual
The purpose for needing to select some or all of the modules therein to realize application scheme.Those of ordinary skill in the art are not paying
Out in the case where creative work, it can understand and implement.
The foregoing is merely the preferred embodiments of the application, not to limit the application, all essences in the application
Within mind and principle, any modification, equivalent substitution, improvement and etc. done be should be included within the scope of the application protection.
Claims (14)
1. a kind of image processing method, which comprises the steps of:
Obtain the weighted value of each pixel of image to be processed；
Fuzzy Processing is carried out to the non-edge in image to be processed according to the weighted value of each pixel and obtains the first figure
Picture；
Reinforcement is carried out to the fringe region in the first image to handle to obtain the second image；
Second image and the image to be processed are subjected to linear fusion and obtain third image.
2. the method according to claim 1, wherein the weighted value according to each pixel is to be processed
Non-edge in image carries out Fuzzy Processing and obtains the first image, comprising:
According to the weighted value of each pixel, color-weighted average value and the color for obtaining each pixel of image to be processed are flat
Side and weighted average；
According to the color-weighted average value, color quadratic sum weighted average and adjusting parameter of the image to be processed, obtain
Steerable filter treated the first image.
3. according to the method described in claim 2, it is characterized in that, the weighted value according to each pixel, obtain to
Handle the color-weighted average value and color quadratic sum weighted average of each pixel of image, comprising:
According to the weighted value of each pixel, each pixel of image to be processed in a first direction color-weighted average is obtained
Value and color quadratic sum weighted average；
According to the color-weighted average value and color quadratic sum weighted average of each pixel in a first direction, and
The weighted value of each pixel obtains each pixel of image to be processed color-weighted average value in a second direction and face
Color quadratic sum weighted average, wherein first direction is vertical with second direction.
4. according to the method described in claim 2, it is characterized in that, the weighted value according to each pixel, obtain to
The color-weighted average value and color quadratic sum weighted average for handling each pixel of image, are obtained by following formula:
Wherein, the r is windows radius；
(i, j) characterizes the central pixel point of the window；
P (x, j), p (i, j) and p (i, y) are respectively the color value of pixel (x, j), (i, j) and (i, y) in image to be processed；
weightxFor the weighted value for the pixel that the distance of Range Profile vegetarian refreshments (i, j) in a first direction is x；
weightyFor the weighted value for the pixel that the distance of Range Profile vegetarian refreshments (i, j) in a second direction is y；
meanI1(i, j) and meanI1(i, y) is respectively pixel (i, j) and (i, y) in a first direction color-weighted average
Value；
meanI2(i, j) is the color-weighted average value of pixel (i, j) in a second direction；
meanII1(i, j) and meanII1(i, y) is respectively the color quadratic sum of pixel (i, j) and (i, y) in a first direction
Weighted average；
meanII2(i, j) is the color quadratic sum weighted average of pixel (i, j) in a second direction.
5. the method according to claim 1, wherein the fringe region in the first image adds
It manages to obtain the second image in strength, comprising:
High contrast reservation process is carried out to the first image and obtains the second image.
6. according to the method described in claim 5, it is characterized in that, the high contrast reservation process includes improved high contrast
Reservation process；
The improved high contrast reservation process includes:
High contrast reservation process is carried out to the image to be processed and the first image and obtains the 4th image；
Result reinforcement is carried out to the 4th image to handle to obtain the 5th image；
High contrast reservation process is carried out to the 5th image and the image to be processed and obtains the second image.
7. the method according to claim 1, wherein second image and the image to be processed are carried out line
Property merges to obtain third image, comprising:
It handles second image procossing progress Gaussian Blur to obtain the 6th image；
6th image and the image to be processed are subjected to linear fusion and obtain third image.
8. the method according to claim 1, wherein the method also includes steps:
Image to be processed is carried out highlighting processing and obtains the 7th image；
7th image and the third linearity fusion treatment are obtained into the 8th image.
9. method according to any one of claims 1 to 8, which is characterized in that the image to be processed includes live video
Each frame live video frame in stream, the method apply to main broadcaster's client of internet live streaming.
10. a kind of live broadcasting method, which is characterized in that the method includes the steps:
Live video stream to be processed is obtained from main broadcaster's client, the live video stream includes several live video frames；
Obtain the weighted value of each pixel of live video frame；
The non-edge in the live video frame is subjected to Fuzzy Processing according to the weighted value of each pixel and obtains the
One image；
Reinforcement is carried out to the fringe region in the first image to handle to obtain the second image；
Second image and the image to be processed are subjected to linear fusion and obtain third image；
Institute's third image is sent to main broadcaster's client and specified spectator client is shown.
11. a kind of image processing apparatus, which is characterized in that described device includes:
Processing module, for obtaining the weighted value of each pixel of image to be processed；It is treated according to the weighted value of each pixel
Non-edge in processing image carries out Fuzzy Processing and obtains the first image；Fringe region in the first image is carried out
Reinforcement handles to obtain the second image；
Fusion Module obtains third image for second image and the image to be processed to be carried out linear fusion.
12. a kind of live broadcast device, which is characterized in that described device includes:
Module is obtained, for obtaining live video stream to be processed from main broadcaster's client, the live video stream includes several straight
Broadcast picture frame；
Processing module, for obtaining the weighted value of each pixel of live video frame；According to the weighted value of each pixel by institute
The non-edge progress Fuzzy Processing stated in live video frame obtains the first image；To the fringe region in the first image
Reinforcement is carried out to handle to obtain the second image；Second image and the image to be processed are subjected to linear fusion and obtain third figure
Picture；
Distribution module, for institute's third image being sent to main broadcaster's client and specified spectator client is shown.
13. a kind of electronic equipment characterized by comprising
Memory, processor and storage are on a memory and the computer program that can run on a processor, wherein the processing
Device realizes the operation as described in claims 1 to 10 any one method when executing described program.
14. a kind of storage medium, which is characterized in that be stored thereon with program, right is realized when described program is executed by processor
It is required that the step of 1 to 10 any one the method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811117153.9A CN109377454A (en) | 2018-09-25 | 2018-09-25 | A kind of image processing method, device, equipment, storage medium and live broadcasting method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811117153.9A CN109377454A (en) | 2018-09-25 | 2018-09-25 | A kind of image processing method, device, equipment, storage medium and live broadcasting method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109377454A true CN109377454A (en) | 2019-02-22 |
Family
ID=65402337
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811117153.9A Pending CN109377454A (en) | 2018-09-25 | 2018-09-25 | A kind of image processing method, device, equipment, storage medium and live broadcasting method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109377454A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110177287A (en) * | 2019-06-11 | 2019-08-27 | 广州虎牙科技有限公司 | A kind of image procossing and live broadcasting method, device, equipment and storage medium |
| CN110223246A (en) * | 2019-05-17 | 2019-09-10 | 杭州趣维科技有限公司 | A kind of windy lattice portrait U.S. face mill skin method and device |
| CN111028180A (en) * | 2019-12-23 | 2020-04-17 | 腾讯科技（深圳）有限公司 | Image processing method, video processing method and corresponding devices |
| CN112435173A (en) * | 2019-08-26 | 2021-03-02 | 广州虎牙科技有限公司 | Image processing and live broadcasting method, device, equipment and storage medium |
| WO2021253723A1 (en) * | 2020-06-16 | 2021-12-23 | 北京达佳互联信息技术有限公司 | Human body image processing method and apparatus, electronic device and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105430231A (en) * | 2015-11-24 | 2016-03-23 | 努比亚技术有限公司 | Image processing device and method |
| CN106169177A (en) * | 2016-06-27 | 2016-11-30 | 北京金山安全软件有限公司 | Image buffing method and device and electronic equipment |
| CN106228516A (en) * | 2016-07-14 | 2016-12-14 | 脸萌技术（深圳）有限公司 | The most U.S. face method, the device of a kind of high naturalness |
| CN106341696A (en) * | 2016-09-28 | 2017-01-18 | 北京奇虎科技有限公司 | Live video stream processing method and device |
| CN106447620A (en) * | 2016-08-26 | 2017-02-22 | 北京金山猎豹科技有限公司 | Face image polishing method and device, and terminal device |
| CN106600556A (en) * | 2016-12-16 | 2017-04-26 | 合网络技术（北京）有限公司 | Image processing method and apparatus |
| CN107798654A (en) * | 2017-11-13 | 2018-03-13 | 北京小米移动软件有限公司 | Image mill skin method and device, storage medium |
-
2018
- 2018-09-25 CN CN201811117153.9A patent/CN109377454A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105430231A (en) * | 2015-11-24 | 2016-03-23 | 努比亚技术有限公司 | Image processing device and method |
| CN106169177A (en) * | 2016-06-27 | 2016-11-30 | 北京金山安全软件有限公司 | Image buffing method and device and electronic equipment |
| CN106228516A (en) * | 2016-07-14 | 2016-12-14 | 脸萌技术（深圳）有限公司 | The most U.S. face method, the device of a kind of high naturalness |
| CN106447620A (en) * | 2016-08-26 | 2017-02-22 | 北京金山猎豹科技有限公司 | Face image polishing method and device, and terminal device |
| CN106341696A (en) * | 2016-09-28 | 2017-01-18 | 北京奇虎科技有限公司 | Live video stream processing method and device |
| CN106600556A (en) * | 2016-12-16 | 2017-04-26 | 合网络技术（北京）有限公司 | Image processing method and apparatus |
| CN107798654A (en) * | 2017-11-13 | 2018-03-13 | 北京小米移动软件有限公司 | Image mill skin method and device, storage medium |
Non-Patent Citations (8)
| Title |
|---|
| CAIN_HUANG: "Android OpenGLES 实时美颜(磨皮)的优化", 《HTTPS://WWW.JIANSHU.COM/P/A76A1201AE53》 * |
| KAIMING HE等: "Guided Image Filtering", 《IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE》 * |
| LAVIEWPBT: "简单探讨可牛影像软件中具有肤质保留功能的磨皮算法及其实现细节", 《HTTPS://WWW.CNBLOGS.COM/IMAGESHOP/P/4709710.HTML》 * |
| SANGU ARUNA KUMARI等: "Weighted Guided Image Filtering for Image Enhancement", 《PROCEEDINGS OF THE 2ND INTERNATIONAL CONFERENCE ON COMMUNICATION AND ELECTRONICS SYSTEMS (ICCES 2017)》 * |
| 吴昊雨: "Android平台Camera实时滤镜实现方法探讨(十一)--实时美颜滤镜", 《HTTPS://MY.OSCHINA.NET/WUHAOYU/BLOG/607815》 * |
| 拓幻科技: "知乎网站中的技术文章"妹纸们的最爱 - 美颜算法，美颜SDK", 《HTTPS://ZHUANLAN.ZHIHU.COM/P/32360617》 * |
| 爱鱼: "导向滤波算法的实现", 《HTTPS://WWW.CNBLOGS.COM/MIGHTYCODE/P/8018327.HTML》 * |
| 程序员扛把子: "Android平台Camera实时滤镜实现方法探讨(十一)--实时美颜滤镜", 《HTTPS://BLOG.CSDN.NET/OSHUNZ/ARTICLE/DETAILS/50536031》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110223246A (en) * | 2019-05-17 | 2019-09-10 | 杭州趣维科技有限公司 | A kind of windy lattice portrait U.S. face mill skin method and device |
| CN110223246B (en) * | 2019-05-17 | 2021-05-18 | 杭州趣维科技有限公司 | Multi-style portrait beautifying and skin-grinding method and device |
| CN110177287A (en) * | 2019-06-11 | 2019-08-27 | 广州虎牙科技有限公司 | A kind of image procossing and live broadcasting method, device, equipment and storage medium |
| CN112435173A (en) * | 2019-08-26 | 2021-03-02 | 广州虎牙科技有限公司 | Image processing and live broadcasting method, device, equipment and storage medium |
| CN111028180A (en) * | 2019-12-23 | 2020-04-17 | 腾讯科技（深圳）有限公司 | Image processing method, video processing method and corresponding devices |
| WO2021128835A1 (en) * | 2019-12-23 | 2021-07-01 | 腾讯科技（深圳）有限公司 | Image processing method and device, video processing method and device, electronic apparatus and storage medium |
| WO2021253723A1 (en) * | 2020-06-16 | 2021-12-23 | 北京达佳互联信息技术有限公司 | Human body image processing method and apparatus, electronic device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109377454A (en) | A kind of image processing method, device, equipment, storage medium and live broadcasting method | |
| Naruniec et al. | High‐resolution neural face swapping for visual effects | |
| CN109146823A (en) | Image processing method, device, equipment and the storage medium retained based on high contrast | |
| Tang et al. | Content-based photo quality assessment | |
| US9197789B2 (en) | Method and system for removal of fog, mist, or haze from images and videos | |
| US7982771B2 (en) | Method of emendation for attention trajectory in video content analysis | |
| AU2006252195B8 (en) | MPEG noise reduction | |
| US20180091768A1 (en) | Apparatus and methods for frame interpolation based on spatial considerations | |
| CN109767408B (en) | Image processing method, image processing device, storage medium and computer equipment | |
| CN107871303B (en) | Image processing method and device | |
| CN110620924B (en) | Method and device for processing coded data, computer equipment and storage medium | |
| CN109302628B (en) | Live broadcast-based face processing method, device, equipment and storage medium | |
| US11538138B2 (en) | Methods and apparatus for applying motion blur to overcaptured content | |
| WO2013135067A1 (en) | Method and apparatus for stabilizing target object image in video | |
| CN110766637A (en) | Video processing method, processing device, electronic equipment and storage medium | |
| CN112348736B (en) | Method, storage medium, device and system for removing black eye | |
| CN113253890B (en) | Video image matting method, system and medium | |
| Jiang et al. | Meflut: Unsupervised 1d lookup tables for multi-exposure image fusion | |
| US8629883B2 (en) | Method and system for generating online cartoon outputs | |
| US11544830B2 (en) | Enhancing image data with appearance controls | |
| CN114125471A (en) | Video coding pre-filtering method | |
| Guo et al. | No-reference omnidirectional video quality assessment based on generative adversarial networks | |
| Jiang et al. | Generating video sequence from photo image for mobile screens by content analysis | |
| WO2022110033A1 (en) | Video processing method and apparatus, and terminal device | |
| Tang et al. | Learning to Fuse Heterogeneous Features for Low-Light Image Enhancement |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20210115 Address after: 511442 3108, 79 Wanbo 2nd Road, Nancun Town, Panyu District, Guangzhou City, Guangdong Province Applicant after: GUANGZHOU CUBESILI INFORMATION TECHNOLOGY Co.,Ltd. Address before: 511442 24 floors, B-1 Building, Wanda Commercial Square North District, Wanbo Business District, 79 Wanbo Second Road, Nancun Town, Panyu District, Guangzhou City, Guangdong Province Applicant before: GUANGZHOU HUADUO NETWORK TECHNOLOGY Co.,Ltd. |
|
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20190222 Assignee: GUANGZHOU HUADUO NETWORK TECHNOLOGY Co.,Ltd. Assignor: GUANGZHOU CUBESILI INFORMATION TECHNOLOGY Co.,Ltd. Contract record no.: X2021440000052 Denomination of invention: The invention relates to an image processing method, a device, a storage medium and a live broadcast method License type: Common License Record date: 20210222 |