CN109242785A - A method of lip gloss on the accurate portrait based on neural network and dithering - Google Patents
A method of lip gloss on the accurate portrait based on neural network and dithering Download PDFInfo
- Publication number
- CN109242785A CN109242785A CN201810909139.6A CN201810909139A CN109242785A CN 109242785 A CN109242785 A CN 109242785A CN 201810909139 A CN201810909139 A CN 201810909139A CN 109242785 A CN109242785 A CN 109242785A
- Authority
- CN
- China
- Prior art keywords
- lip
- neural network
- pixel
- dithering
- gloss
- 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.)
- Withdrawn
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 25
- 238000012545 processing Methods 0.000 claims abstract description 13
- 238000003062 neural network model Methods 0.000 claims abstract description 5
- 230000000694 effects Effects 0.000 claims description 7
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 3
- 239000003086 colorant Substances 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000000877 morphologic effect Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 239000002537 cosmetic Substances 0.000 description 1
- 238000004141 dimensional analysis Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000009466 transformation 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
- G06T5/77—Retouching; Inpainting; Scratch removal
-
- A—HUMAN NECESSITIES
- A45—HAND OR TRAVELLING ARTICLES
- A45D—HAIRDRESSING OR SHAVING EQUIPMENT; EQUIPMENT FOR COSMETICS OR COSMETIC TREATMENTS, e.g. FOR MANICURING OR PEDICURING
- A45D44/00—Other cosmetic or toiletry articles, e.g. for hairdressers' rooms
- A45D44/005—Other cosmetic or toiletry articles, e.g. for hairdressers' rooms for selecting or displaying personal cosmetic colours or hairstyle
-
- 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
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of to promote the precise degrees that lip gloss is handled based on the method for lip gloss on the accurate portrait of neural network and dithering, the case where capable of effectively judging lip position and whether having foreign matter or be blocked, the precision of lip gloss identification is improved, so that the virtual makeup processing of lip can be more life-like and natural.The present invention can be inferred that the approximate location of lip in face using neural network model, and then the pixel coverage for belonging to lip is further determined that using the scheme of dithering, can be with the accurate position coordinates of lip pixel in highly effective determining picture.When carrying out makeup to lip gloss can the pixel effectively to only lip handled rather than monolith lip position is handled.
Description
Technical field
The present invention relates to the processing of portrait image lip gloss, specifically a kind of accurate portrait based on neural network and dithering
The method of upper lip gloss.
Background technique
In the overall technology of digital picture makeup, the processing about lip gloss part is particularly important.One carrys out face lip
Portion's form is similar and different, and two carry out lip, and there are ever-changing cheilogrammas, reaches during cosmetic treatment
More true effect needs the more accurate lip gloss part identified and judgeed in face.
Chinese Patent Application No. 201210100239.7 proposes a kind of based on different lip morphological feature progress lip glosses
The method of makeup.The classification method classified in the method according to the morphological feature of lip to lip and by according to the classification
The classification map of the lip for the coordinate composition that method generates and the form for using the plane of delineation of lip and the analysis lip of solid, root
So that lip is reached suitable form balance according to plane and three-dimensional analysis result, and proposes the correction information of lip form.
The above method needs to carry out lip the calculation processing of complicated classification processing and balance, and in light complexity
In the case where be difficult to preferable expression effect.
Summary of the invention
In order to solve more life-like lip gloss treatment effect in portrait, the invention proposes one kind based on neural network and
On the accurate portrait of dithering the method for lip gloss come promoted lip gloss processing precise degrees, can effectively judge lip position
The case where whether having foreign matter or being blocked, improves the precision of lip gloss identification, so that the virtual makeup of lip is handled
It can be more life-like and natural.
A method of lip gloss on the accurate portrait based on neural network and dithering, comprising the following steps:
1), determine the point of the key position of lip as characteristic point respectively;
2) picture of portrait, is collected, and according to the lip characteristic point defined in step 1) in the picture being collected into
Portrait lip be marked, the sample as neural network;
3) neural network, is constructed, lip feature point value is returned using the homing method in neural network
Method training, obtains the neural network model of a regression forecasting;
4), the digital picture inputted, obtains the prediction of lip position by step 3), obtains the key point about lip;
5), the range of key point is expanded, prevents neural network from predicting lip position not accurate enough, and then be lost one
Partial lip pixel;
6) pixel in key point after, expanding range in step 5 switchs to YCrCb color space from RGB color;
7) two values space R, is definedcbAnd Rcr, the numberical range in the channel Cb and the channel Cr that respectively represent;
8), selecting step 6) in color space the value in the channel Cb and the channel Cr in Cb ∈ Rcb、Cr∈RcrPixel, symbol
The pixel for closing above-mentioned condition is the pixel set of lip, records the coordinate value of its pixel;
9) pixel color space, is converted into RGB color from YCrCb;
10) primary colors processing, is carried out using the method that lip gloss is handled to the pixel set got in step 8), reaches and behaves excellently
The effect of transformation;
11), output treated result images.
The present invention can be inferred that the approximate location of lip in face using neural network model, and then utilize dithering
Scheme further determine that the pixel coverage for belonging to lip, can be with the exact position of lip pixel in highly effective determining picture
Coordinate.To lip gloss carry out makeup when can the pixel effectively to only lip handled rather than monolith lip
Position is handled.
Detailed description of the invention
Attached drawing is used to provide further understanding of the present invention, and constitutes part of specification, with reality of the invention
It applies example to be used to explain the present invention together, not be construed as limiting the invention.In the accompanying drawings:
Fig. 1 is flow diagram of the embodiment of the present invention;
Fig. 2 is the point schematic diagram of the key position of lip of the embodiment of the present invention;
Fig. 3 is that the range of the key point of lip of the embodiment of the present invention expands schematic diagram.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
In the virtual makeup of image lip gloss, the processing of monolith is often carried out just for the position of lip, is generally occurred within
The problem of be lip position that excessive dependence neural network is inferred, and non-lip gloss position is also handled.Such as work as lip
There are certain will to have carried out the shelter same lip gloss processing when blocking, cause effect very awkward.Pass through
The solution of the present invention can effectively avoid the occurrence of this.
As shown in Figure 1, a kind of accurate portrait lip gloss processing method based on neural network and dithering, essential core step
It is rapid as follows:
1, determine respectively the key position of lip point (position in such as labial angle position, people, upper lip camber line arc top position and/
Or lower lip camber line arc top position), totally 21, wherein the characteristic point of upper lip is 10, and the characteristic point of lower lip is 11, such as Fig. 2 institute
Show.
2, the picture of portrait is collected, and according to 21 characteristic points of the lip defined in step 1 to the picture being collected into
In portrait lip (horizontal axis coordinate and ordinate of orthogonal axes position), the sample as neural network is marked.
3, a neural network is constructed, using the homing method in neural network to the total 21x2=42 number of 21 points of lip
Value carries out homing method training, obtains the neural network model of 21 points of a regression forecasting.
Due to the diversity of the movement posture of the diversity and lip of human face posture, the neural network mould that is obtained in step 3
Type tends not to access very accurately portrait lip prediction result, especially in some extreme situations, as illumination is strong
Either in dim image or digital picture.
4, the prediction for the lip position that the digital picture inputted is obtained by step 3, available 21 about lip
Key point.
5, the range of 21 key points is expanded into (as shown in Fig. 3).It is widened the reason is that neural network pair in order to prevent
Lip position prediction is not accurate enough, and then is lost the lip pixel of a part.
6, the pixel in 21 points (upperlip) in step 5 is switched into YCrCb color space from RGB color, turned
The formula changed is as follows:
Y=0.257*R+0.564*G+0.098*B+16
Cb=-0.148*R-0.291*G+0.439*B+128
Cr=0.439*R-0.368*G-0.071*B+128
7, two values space, R are definedcbAnd Rcr, what is respectively represented is the numberical range in the channel Cb and the channel Cr.
8, in selecting step 6 in color space the value in the channel Cb and the channel Cr in Cb ∈ RcbCr∈RcrPixel, meet
The pixel for stating condition is the pixel set of lip, records the coordinate value of its pixel.
9, pixel color space is converted into RGB color, the formula of conversion from YCrCb are as follows:
R=1.164* (Y-16)+1.596* (Cr-128)
G=1.164* (Y-16) -0.392* (Cb-128) -0.813* (Cr-128)
B=1.164* (Y-16)+2.017* (Cb-128)
10, the method that lip gloss processing is utilized to the pixel set got in step 8, as the methods of color mixing carries out original
Color processing, achievees the effect that color conversion out.
11, output treated result images.
Method in the present invention is only the value range for judging pixel, and calculation amount is relatively low, can satisfy various performances
Real-time requirement of the lower processor in the processing to lip gloss all has extensively in hand-held terminal device and embedded device
Application value.
It although an embodiment of the present invention has been shown and described, for the ordinary skill in the art, can be with
A variety of variations, modification, replacement can be carried out to these embodiments without departing from the principles and spirit of the present invention by understanding
And modification, the scope of the present invention is defined by the appended.
Claims (4)
1. a kind of method of lip gloss on accurate portrait based on neural network and dithering, it is characterised in that including following step
It is rapid:
1), determine the point of the key position of lip as characteristic point respectively;
2) picture of portrait, is collected, and according to the lip characteristic point defined in step 1) to the people in the picture being collected into
As lip is marked, the sample as neural network;
3) neural network, is constructed, homing method is carried out to lip feature point value using the homing method in neural network
Training, obtains the neural network model of a regression forecasting;
4), the digital picture inputted, obtains the prediction of lip position by step 3), obtains the key point about lip;
5), the range of key point is expanded, prevents neural network from predicting lip position not accurate enough, and then be lost a part
Lip pixel;
6) pixel in key point after, expanding range in step 5 switchs to YCrCb color space from RGB color;
7) two values space R, is definedcbAnd Rcr, the numberical range in the channel Cb and the channel Cr that respectively represent;
8), selecting step 6) in color space the value in the channel Cb and the channel Cr in Cb ∈ Rcb、Cr∈RcrPixel, meet above-mentioned
The pixel of condition is the pixel set of lip, records the coordinate value of its pixel;
9) pixel color space, is converted into RGB color from YCrCb;
10) primary colors processing, is carried out using the method that lip gloss is handled to the pixel set got in step 8), reaches color conversion out
Effect;
11), output treated result images.
2. the method for lip gloss, special on a kind of accurate portrait based on neural network and dithering as described in claim 1
Sign is:
In the step 1), the key position of lip include labial angle position, position in people, upper lip camber line arc top position and/or under
Lip camber line arc top position.
3. the method for lip gloss, special on a kind of accurate portrait based on neural network and dithering as described in claim 1
Sign is in the step 6) that conversion formula is as follows:
Y=0.257*R+0.564*G+0.098*B+16
Cb=-0.148*R-0.291*G+0.439*B+128
Cr=0.439*R-0.368*G-0.071*B+128.
4. the method for lip gloss, special on a kind of accurate portrait based on neural network and dithering as described in claim 1
Sign is in the step 8) that conversion formula is as follows:
R=1.164* (Y-16)+1.596* (Cr-128)
G=1.164* (Y-16) -0.392* (Cb-128) -0.813* (Cr-128)
B=1.164* (Y-16)+2.017* (Cb-128).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810909139.6A CN109242785A (en) | 2018-08-10 | 2018-08-10 | A method of lip gloss on the accurate portrait based on neural network and dithering |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810909139.6A CN109242785A (en) | 2018-08-10 | 2018-08-10 | A method of lip gloss on the accurate portrait based on neural network and dithering |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109242785A true CN109242785A (en) | 2019-01-18 |
Family
ID=65070931
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810909139.6A Withdrawn CN109242785A (en) | 2018-08-10 | 2018-08-10 | A method of lip gloss on the accurate portrait based on neural network and dithering |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109242785A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20190065831A1 (en) * | 2017-08-24 | 2019-02-28 | Cal-Comp Big Data, Inc. | Body information analysis apparatus and lip-makeup analysis method thereof |
CN110827195A (en) * | 2019-10-31 | 2020-02-21 | 北京达佳互联信息技术有限公司 | Virtual article adding method and device, electronic equipment and storage medium |
TWI752473B (en) * | 2019-11-22 | 2022-01-11 | 大陸商北京市商湯科技開發有限公司 | Image processing method and apparatus, electronic device and computer-readable storage medium |
US11403788B2 (en) | 2019-11-22 | 2022-08-02 | Beijing Sensetime Technology Development Co., Ltd. | Image processing method and apparatus, electronic device, and storage medium |
-
2018
- 2018-08-10 CN CN201810909139.6A patent/CN109242785A/en not_active Withdrawn
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20190065831A1 (en) * | 2017-08-24 | 2019-02-28 | Cal-Comp Big Data, Inc. | Body information analysis apparatus and lip-makeup analysis method thereof |
US10515260B2 (en) * | 2017-08-24 | 2019-12-24 | Cal-Comp Big Data, Inc. | Body information analysis apparatus and lip-makeup analysis method thereof |
CN110827195A (en) * | 2019-10-31 | 2020-02-21 | 北京达佳互联信息技术有限公司 | Virtual article adding method and device, electronic equipment and storage medium |
CN110827195B (en) * | 2019-10-31 | 2023-09-22 | 北京达佳互联信息技术有限公司 | Virtual article adding method and device, electronic equipment and storage medium |
TWI752473B (en) * | 2019-11-22 | 2022-01-11 | 大陸商北京市商湯科技開發有限公司 | Image processing method and apparatus, electronic device and computer-readable storage medium |
US11403788B2 (en) | 2019-11-22 | 2022-08-02 | Beijing Sensetime Technology Development Co., Ltd. | Image processing method and apparatus, electronic device, and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109242785A (en) | A method of lip gloss on the accurate portrait based on neural network and dithering | |
CN104282002B (en) | A kind of quick beauty method of digital picture | |
CN110084161B (en) | Method and system for rapidly detecting key points of human skeleton | |
CN109559310B (en) | Power transmission and transformation inspection image quality evaluation method and system based on significance detection | |
WO2022179215A1 (en) | Image processing method and apparatus, electronic device, and storage medium | |
CN107204034B (en) | A kind of image processing method and terminal | |
CN107993216A (en) | A kind of image interfusion method and its equipment, storage medium, terminal | |
CN113344836B (en) | Face image processing method and device, computer readable storage medium and terminal | |
CN106504264A (en) | Video foreground image extraction method and device | |
CN106530265A (en) | Adaptive image fusion method based on chromaticity coordinates | |
CN109657612A (en) | A kind of quality-ordered system and its application method based on facial image feature | |
CN108024105A (en) | Image color adjusting method, device, electronic equipment and storage medium | |
CN106340025A (en) | Background replacement visual communication method based on chromatic adaptation transformation | |
CN112308944A (en) | Augmented reality display method of simulated lip makeup | |
CN110533732A (en) | The recognition methods of the colour of skin, device, electronic equipment and storage medium in image | |
CN109919030A (en) | Black eye kind identification method, device, computer equipment and storage medium | |
CN105069745A (en) | face-changing system based on common image sensor and enhanced augmented reality technology and method | |
CN109064431B (en) | Picture brightness adjusting method, equipment and storage medium thereof | |
CN107154046A (en) | A kind of method of video background processing and secret protection | |
KR20210032489A (en) | Simulation method of rendering of makeup products on the body area | |
KR0134701B1 (en) | Image generating method and device | |
CN113052783A (en) | Face image fusion method based on face key points | |
CN114155569B (en) | Cosmetic progress detection method, device, equipment and storage medium | |
JP2006031688A (en) | Image processing device, method, and program | |
CN106360941A (en) | Method for locating fingernails and manicure equipment |
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 | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190118 |
|
WW01 | Invention patent application withdrawn after publication |