CN106446207B - Makeups library banking process, personalized makeups householder method and its device - Google Patents
Makeups library banking process, personalized makeups householder method and its device Download PDFInfo
- Publication number
- CN106446207B CN106446207B CN201610873350.8A CN201610873350A CN106446207B CN 106446207 B CN106446207 B CN 106446207B CN 201610873350 A CN201610873350 A CN 201610873350A CN 106446207 B CN106446207 B CN 106446207B
- Authority
- CN
- China
- Prior art keywords
- makeups
- scheme
- keyword
- document
- makeup
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/93—Document management systems
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Business, Economics & Management (AREA)
- Processing Or Creating Images (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the present invention provides a kind of makeups library banking process, personalized makeups householder method and its device, and makeups library banking process includes: the extraction makeups keyword from the makeups scheme of acquisition, generates makeups according to the content structure of the makeups scheme and describes document；Different weights is assigned for the makeups keyword that different location occurs in the content structure that the makeups describe document, to realize the makeup skill for efficiently and effectively implementing to know in reality, personalized makeup scheme is completed and recommends.
Description
Technical field
The present embodiments relate to Internet technical field more particularly to a kind of makeups library banking process, personalized makeups
Householder method and its device.
Background technique
With cosmetics and tool, the step and skill in accordance with rule, control mirror and the manual behaviour that places one's entire reliance upon are taken
Work renders the face of human body, face and other positions, is drawn, being arranged, and enhancing three-dimensional print is as adjusting shape and color, covering up scarce
It falls into, shows expression, to achieve the purpose that beautify visual experience.
Makeup can be divided into foundation make up and emphasis makeup.Foundation make up refers to that color is applied on the basis of entire face, comprising: clear
Clean, moist, convergence, bottoming and face powder etc., the function with skin care.Emphasis makeup refers to the thin of the organs such as eye, eyelash, eyebrow, cheek, lip
Portion's makeup, comprising: add eye shadow, draw informer, brushing eyelash, apply nose shadow, wiping kermes and smear lipstick etc., can increase the beautiful of appearance and be in
Three-dimensional sense can change with different occasions.The method of makeup has daily general makeup method, adapts to the special of various occasions needs
Makeup method and simple and direct quick rapid-result makeup method etc..
And so far, except through learning outside makeup skill from TV, network and magazine, without other better canals
Road；Furthermore from the foregoing, it can be seen that makeup is related to numerous steps and disposal skill in itself, and for a user, even if user knows
Some makeup skills, but in practical application, since actual makeup environment such as cosmetics, cosmetic applicators are become
Change, leads to the makeup skill for being also difficult to efficiently and effectively implement to know in reality, and use really suitable makeup scheme.
Summary of the invention
In view of this, one of the technical issues of embodiment of the present invention is solved be to provide a kind of makeups library banking process,
Personalized makeups householder method and its device, to overcome above-mentioned technical problem in the prior art.
The embodiment of the present invention provides a kind of makeups library banking process comprising:
Makeups keyword is extracted from the makeups scheme of acquisition, and makeups are generated according to the content structure of the makeups scheme and are retouched
State document；
It is assigned for the makeups keyword that different location occurs in the content structure that the makeups describe document different
Weight.
Optionally, in one embodiment of this invention, makeups keyword is extracted in the makeups scheme of acquisition includes: from acquisition
To makeup knowledge in extract makeup knowledge keyword, the makeups scheme includes the makeup knowledge；
Generating makeups to describe document according to the content structure of the makeups scheme includes: the content according to the makeup knowledge
Structural generation makeup knowledge description document；
It is assigned for the makeups keyword that different location occurs in the content structure of the makeup knowledge description document
Different weights includes: the makeup knowledge for the different location appearance in the content structure of the makeup knowledge description document
Keyword assigns different weights.
Optionally, in one embodiment of this invention, makeups keyword is extracted from the makeups scheme of acquisition includes: from adopting
Modeling scheme keyword is extracted in the modeling scheme of collection, the makeups scheme includes the modeling scheme；
Generating makeups to describe document according to the content structure of the makeups scheme includes: the content according to the modeling scheme
Structural generation modeling scheme describes document；
It is assigned for the makeups keyword that different location occurs in the content structure of the makeup knowledge description document
Different weights includes: to describe the modeling scheme keyword that different location occurs in document in the modeling scheme and assigning
Different weights.
Optionally, in one embodiment of this invention, the makeup knowledge includes: knowledge entry and dressing data, described
Knowledge entry is used to characterize the static description of makeup, and the dressing data are used to characterize the Dynamic profiling of makeup.
Optionally, in one embodiment of this invention, the knowledge entry includes makeup skill, makeup region, makeup hand
One of method or a variety of combinations,；The dressing data include dressing description, makeup one of step and multimedia content or
A variety of combinations.
Optionally, in one embodiment of this invention, the method also includes: generate the management training coarse of the knowledge entry
And the management training coarse of the dressing data, and the management class of the management training coarse and the dressing data to the knowledge entry
Cheng Jinhang Data Integration.
Optionally, in one embodiment of this invention, the method also includes: according to tree hierarchy structure to the knowledge
The management training coarse of entry and the management training coarse of the dressing data carry out Content Management.
Optionally, in one embodiment of this invention, further includes: the makeups keyword of extraction is normalized.
Optionally, in one embodiment of this invention, further includes: to multiple beauty in the way of it can carry out concurrent type frog retrieval
Adornment describes document and is stored.
Optionally, in one embodiment of this invention, document is described to multiple makeups and carries out fragment storage, it is concurrent to carry out
Formula retrieval.
The embodiment of the present invention provides a kind of personalized makeups householder method comprising:
Matching is carried out to the real-time facial image for obtaining user in the five features classification model pre-established and obtains use
The personalized makeups label at family；
According to for carrying out makeups keyword in the personalized makeups the label in any embodiment database
Match；
According to the makeups keyword being matched in the different weights that corresponding makeups describe document different location and assign, to packet
The makeups scheme for including the makeups keyword being matched to carries out weight and counts to obtain the recommendation index of makeups scheme, according to the recommendation
Makeups scheme is presented to the user in index.
Optionally, in one embodiment of this invention, the method also includes: to sample facial image carry out feature mark
With extract and establish faceform, with to five features carry out classification generate feature classification model.
Optionally, in one embodiment of this invention, feature mark is carried out to sample facial image and face is established in extraction
Model, with to five features carry out classification generate feature classification model include:
Face shape modeling is carried out according to the face characteristic of extraction and the modeling of face texture respectively obtains face shape model
With face texture model；
Faceform is established according to the face shape model and face texture model, to carry out classification life to five features
At feature classification model.
Optionally, in one embodiment of this invention, the method also includes: according to the preference information of user, Yong Husuo
Makeups label is assisted in the Weather information on ground, the combination producing personalization that the birthday by information of user is any one or more；
Makeups keyword is carried out in the database described in any embodiment according to the personalized auxiliary makeups label
Matching.
Optionally, in one embodiment of this invention, the method also includes: to including the makeups keyword being matched to
Makeups scheme carries out level-one and secondary merger sequence respectively.
The embodiment of the present invention provides a kind of makeups library and builds library device comprising:
Document establishes unit, for extracting makeups keyword from the makeups scheme of acquisition, according to the makeups scheme
Content structure generates makeups and describes document；
Weight-assigning unit, for the beauty for the different location appearance in the content structure that the makeups describe document
Adornment keyword assigns different weights.
Optionally, in one embodiment of this invention, the document is established unit and is further used for from collected makeup
Makeup knowledge keyword is extracted in knowledge, the makeups scheme includes the makeup knowledge；According to the content of the makeup knowledge
Structural generation makeup knowledge description document；
The weight-assigning unit is further used for as the different positions in the content structure of the makeup knowledge description document
It includes: in the content structure of the makeup knowledge description document that the makeups keyword for setting appearance, which assigns different weights,
The makeup knowledge keyword that different location occurs assigns different weights.
Optionally, in one embodiment of this invention, the document establishes unit and is further used for moulding side from acquisition
Modeling scheme keyword is extracted in case, the makeups scheme includes the modeling scheme, according to the content knot of the modeling scheme
Structure generates modeling scheme and describes document；
The weight-assigning unit is further used for as the different positions in the content structure of the makeup knowledge description document
It includes: to describe different location in document in the modeling scheme and going out that the makeups keyword for setting appearance, which assigns different weights,
The existing modeling scheme keyword assigns different weights.
The embodiment of the present invention provides a kind of personalized makeups auxiliary device comprising:
Tag match unit, in the five features classification model pre-established to the real-time face figure for obtaining user
The personalized makeups label of user is obtained as carrying out matching；
Keywords matching unit, for according to the personalized makeups label in database described in any embodiment into
The matching of row makeups keyword；
Makeups recommendation unit is assigned for describing document different location in corresponding makeups according to the makeups keyword being matched to
The different weights given carry out weight to the makeups scheme for including the makeups keyword being matched to and count to obtain the recommendation of makeups scheme
Makeups scheme is presented to the user according to the recommendation index in index.
Optionally, in one embodiment of this invention, further includes: assisted tag unit, for being believed according to the preference of user
Breath, the Weather information in user location, the combination producing personalization that the birthday by information of user is any one or more assist makeups mark
Label；
The Keywords matching unit is further used for according to the personalized auxiliary makeups label in any embodiment institute
The matching of makeups keyword is carried out in the database stated.
It is crucial by extracting makeups from the makeups scheme of acquisition by above technical scheme as it can be seen that in the embodiment of the present invention
Word generates makeups according to the content structure of the makeups scheme and describes document；It is again to describe the content knot of document in the makeups
The makeups keyword that different location occurs in structure assigns different weights, so that completes makeups library builds library, when need to
When user recommends makeups scheme, the real-time facial image for obtaining user is carried out in the five features classification model pre-established
Matching obtains the personalized makeups label of user, describes document different location in corresponding makeups according to the makeups keyword being matched to
And the different weights assigned, weight is carried out to the makeups scheme for including the makeups keyword being matched to and counts to obtain makeups scheme
Recommend index, makeups scheme is presented to the user according to the recommendation index, is efficiently and effectively implemented in reality with realizing
The makeup skill known is completed personalized makeup scheme and is recommended.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
The some embodiments recorded in inventive embodiments can also obtain according to these attached drawings for those of ordinary skill in the art
Obtain other attached drawings.
Fig. 1 is one makeups library banking process flow diagram of the embodiment of the present invention；
Fig. 2 is two makeups library banking process flow diagram of the embodiment of the present invention；
Fig. 3 is three makeups library banking process flow diagram of the embodiment of the present invention；
Fig. 4 is the personalized makeups aided process flow sheet schematic diagram of the embodiment of the present invention four；
Fig. 5 is the personalized makeups aided process flow sheet schematic diagram of the embodiment of the present invention five；
Fig. 6 is that library apparatus structure schematic diagram is built in six makeups library of the embodiment of the present invention；
Fig. 7 is the personalized makeups assistant apparatus structure schematic diagram of the embodiment of the present invention seven.
Specific embodiment
Certainly, any technical solution for implementing the embodiment of the present invention must be not necessarily required to reach simultaneously above all excellent
Point.
In order to make those skilled in the art more fully understand the technical solution in the embodiment of the present invention, below in conjunction with the present invention
Attached drawing in embodiment, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described reality
Applying example only is a part of the embodiment of the embodiment of the present invention, instead of all the embodiments.Based on the implementation in the embodiment of the present invention
The range of protection of the embodiment of the present invention all should belong in example, those of ordinary skill in the art's every other embodiment obtained.
Below with reference to attached drawing of the embodiment of the present invention the embodiment of the present invention will be further explained specific implementation.
Fig. 1 is one makeups library banking process flow diagram of the embodiment of the present invention；As shown in Figure 1, comprising:
S101, makeups keyword is extracted from the makeups scheme of acquisition, generated according to the content structure of the makeups scheme
Makeups describe document；
In the present embodiment, specifically from collected makeup knowledge extract makeup knowledge keyword, such as by it is two-way most
Major term matching algorithm extracts.The makeups scheme includes the makeup knowledge；The makeup knowledge includes: knowledge entry and adornment
Hold data, the knowledge entry is used to characterize the static description of makeup, such as personalized makeup skill, makeup region, makeup
Gimmick etc..The dressing data are used to characterize the Dynamic profiling of makeup, such as dressing description, makeup step and multimedia content
Information such as (subsidiary voice, animations etc.), wherein dressing description may include to applicable scene, applicable crowd and dressing
The characteristics of introduce etc..
In the present embodiment, describing document according to the generation makeups of the content structure of the makeups scheme be can specifically include: root
Makeup knowledge description document is generated according to the content structure of the makeup knowledge；The storage form of various makeup knowledge in the database
It for file structure, is browsed convenient for subsequent user retrieval, it should be noted that the document can be common editable text text
Shelves, are also possible to multimedia document.
In the present embodiment, in order to reduce subsequent query, retrieval and the complexity of storage, the makeups keyword of extraction is carried out
Normalized.Further, storage makeups when describing document on the database, in the way of it can carry out concurrent type frog retrieval pair
Multiple makeups describe document and are stored.Specifically, multiple makeups can be described with document and carry out fragment storage, it is concurrent to carry out
Formula retrieval, to improve subsequent recall precision.
S102, it is assigned for the makeups keyword that different location occurs in the content structure that the makeups describe document
Different weights.
In the present embodiment, step S102 is specifically as follows the different positions in the content structure of the makeup knowledge description document
The makeup knowledge keyword for setting appearance assigns different weights.For describing document principal mode and be editable text,
If makeups keyword appears in the topic of document, show that the relationship of the document and the makeups keyword is more close, it can be with
Assign higher weight；And if there is lower weight in the body matter part of document, then can be assigned, and so on can
With the position occurred in description document by makeups keyword, the substantial connection of the document Yu makeups keyword, relationship are judged
Closer, corresponding weight is also larger, other more detailed weight distribution schemes are similar, repeats no more in detail.
Fig. 2 is two makeups library banking process flow diagram of the embodiment of the present invention；As shown in Fig. 2, comprising:
S201, modeling scheme keyword is extracted from the modeling scheme of acquisition, the makeups scheme includes the moulding side
Case；
It is unlike the embodiments above in the present embodiment, in the present embodiment, moulding is extracted from the modeling scheme of acquisition
Scheme keyword, modeling scheme can be a whole set of the complete content for recommending user, including rationale for the recommendation or user
Property neutralizing reading, knowledge entry, dressing data etc..
S202, document is described according to the content structure of modeling scheme generation modeling scheme；
Similar above-mentioned makeup knowledge description document in the present embodiment, also generates modeling scheme and describes document, the moulding side
Case, which describes document, can be editable text document, is also possible to multimedia document, can also be text document and multimedia
The combination of document.
S203, it is assigned not to describe the modeling scheme keyword that different location occurs in document in the modeling scheme
Same weight.
Similar above-described embodiment one, according to different location of the modeling scheme keyword in description document, to assign not
Same weight, substantial connection of the imparting principle of weight size similar to above-mentioned judgement the document and modeling scheme keyword, relationship
Closer, corresponding weight is also larger, on the contrary then smaller.
It should be noted that in an other embodiment, while including modeling scheme and change in above-described embodiment one and two
Adornment knowledge and corresponding modeling scheme keyword and makeup knowledge keyword, modeling scheme describes document and makeup knowledge is retouched
Document is stated, is repeated no more in detail.
Fig. 3 is three makeups library banking process flow diagram of the embodiment of the present invention；As shown in figure 3, comprising:
S301, makeups keyword is extracted from the makeups scheme of acquisition, generated according to the content structure of the makeups scheme
Makeups describe document；
In the present embodiment, makeups scheme may include one of modeling scheme and makeup knowledge kind in above-described embodiment one and two
Or both and corresponding modeling scheme keyword and makeup knowledge keyword, modeling scheme describe document and makeup knowledge
Document is described.
S302, it is assigned for the makeups keyword that different location occurs in the content structure that the makeups describe document
Different weights.
Modeling scheme keyword assigns different weights when modeling scheme describes different location in document；Knowledge of making up is crucial
Word assigns different weights when different location in knowledge description document of making up, and the specific distribution principle of weight is similar to above-described embodiment
One, two, it repeats no more in detail.
The management training coarse of S303, the management training coarse for generating knowledge entry and the dressing data, and to the knowledge item
Purpose management training coarse and the management training coarse of the dressing data carry out Data Integration.
It, specifically, can be according to tree hierarchy structure to the management training coarse of the knowledge entry and institute in the present embodiment
The management training coarse for stating dressing data carries out Content Management.Such as by the management of dressing data, by dressing data organization at an adornment
Hold-" makeup region-" and dressing step a tree hierarchy structure, this level of dressing specifically includes that title, description, secondary mark
The information such as topic, crowd, difficulty on probation.One dressing can be corresponding with multiple makeup regions, generally include bottom adornment, eye shadow, informer,
Eyelashes, blush, lip, repair 8 makeup regions such as appearance at eyebrow.Each makeup region depending on dressing theme need situation include it is several not
Deng region step, by taking eye shadow as an example, eye shadow region may include upper eye shadow, superposition eye shadow, several steps such as lower eye shadow, dressing
There are also the data such as subsidiary makeup region, animation, track, official documents and correspondence, voice in step.
Management training coarse and the dressing number in the present embodiment, for the recommendation of modeling scheme, to the knowledge entry
According to management training coarse carry out Data Integration.
Fig. 4 is the personalized makeups aided process flow sheet schematic diagram of the embodiment of the present invention four；As shown in figure 4, comprising:
S401, the real-time facial image for obtaining user match in the five features classification model pre-established and is obtained
Take the personalized makeups label at family；
In the present embodiment, by carrying out feature mark and extraction to sample facial image and establishing faceform, to sample
Five features in this facial image carries out classification and generates feature classification model.
In the present embodiment, feature mark is carried out to sample facial image and faceform is established in extraction, to sample face
Five features in image, which carries out classification generation feature classification model, to be specifically included: carrying out people according to the face characteristic of extraction
Face shape modeling and the modeling of face texture respectively obtain face shape model and face texture model；According to the face shape mould
Type and face texture model establish faceform, generate feature classification mould to carry out classification to five features in sample facial image
Type.
In the present embodiment, the generation of feature classification model can be completed by aam model, detailed process is as follows:
(1) select facial image sample as learning sample；
(2) manual characteristic point label is carried out to the learning sample of selection, so that marking the set of v good characteristic point position
It can constitute shape S, S=(x1, y1, x2, y2 ... xv, yv)；
(3) shape is normalized, is referred to all by normalizing for the face shape removal rotation of study, scaling
With translation etc. global changes；
(4) principal component analysis (Principal Component Analysis, abbreviation PCA) is carried out to normalized shape
Transformation obtains the average shape S0 and the corresponding shape eigenvectors Si of preceding n characteristic value of corresponding training set.
(5) any face shape S can be expressed with linear equation:
This completes the modelings to face shape.
B) texture models:
(1) by the face shape in S0 and training set, difference Delaunay trigonometric ratio；
(2) texture information in sample set face shape is mapped to by average shape S0 by the affine method of piecewise linearity
In, it realizes and texture is normalized；
(3) PCA transformation is carried out to the texture information after normalization, obtains average texture A0 and preceding m characteristic value is corresponding
Texture feature vector Ai (x).
(4) texture and shape are closely similar, and the texture information of any face can also be indicated with linear representation:
A (x) indicates texture example, and A0 indicates average line
Reason, γ i indicate that parametric texture, Ai (x) indicate i-th of texture feature vector.
The modeling to texture is also just completed in this way.
C) AAM faceform example generates
The generation step of AAM faceform's example is as follows, after first obtaining any one group of form parameter p, with shape into
Row linear expression, it will be able to obtain a corresponding shape S, after then obtaining one group of parametric texture γ i, be carried out with texture model
Linear expression obtains a corresponding texture example A (x).Finally the texture example A (x) in average shape S0 is mapped to currently
Shape S in, thus generate the model instance of an AAM.In addition to AAM model, it can also be known using other models
Others' face and face such as Hmax+ neural network classification.
By AAM model be obtained 68 region points, including ocular or so be 6 each, each 5 of supercilium region or so,
Lip-region 20, nasal region 9, face area 17.After identifying facial feature localization, need to carry out human face five-sense-organ feature point
Class, using svm disaggregated model.
By taking eye-shaped disaggregated model training as an example,
(1) first normalize facial feature points: two-dimensional coordinate is translated in three dimensions, rotates, is scaled, so that face is special
Sign normalizing is positive forward-facing；
(2) characteristic point for influencing eye-shaped is chosen；
(3) feature extraction is carried out to characteristic point, such as: eye height, eye width, canthus and eye head relative position.
(4) using the feature of all labeled data as the input of svm, model is generated.Svm principle of classification is to pass through mark
The classification that 1 and -1 is carried out to sample, seeks a linear algorithm, and two class vector maximum degree are separated and clearance space is bigger,
Basis vector calculation formula is as follows, and in following formula, w is that represent be the plane that can separate, and s.t represents restrictive condition and belt restraining
Quadratic programming, b represent intercept, and y represents classification:.Basis vector calculation formula is as follows:
Each face characteristic has respective classification foundation feature, and each face characteristic will carry out svm classification, these are special
Sign mainly includes image texture characteristic, face ratio characteristic, face's Aspect Ratio, eyes eye head and eye tail ratio, eyeball high point
It is long same with nose to chin ratio, eyebrow peak and eyebrow tail and brows gradient, eyebrow with eye tail terminal and eye head starting slope, nose length
Eyebrow gradient etc..Final output is accurate and recall rate meets feature classification model.
S402, makeups keyword is carried out according in the personalized makeups the label in any embodiment database
Matching；
In the present embodiment, the five features of personalized makeups tagging user, for example, such as shape of face: oval face, rectangular
Face, diamond shape face, triangle face, heart-shaped face, round face and elongated face；Such as chin: square jaw, pointed chin and circle chin；Than
Such as eye-shaped: almond-eyed, birdeye, circle eye, slot mesh, telecentricity eye, loser and vertical eye；Such as camber: arched eyebrows, slanted eyebrows, one
Word eyebrow, on choose eyebrow and standard eyebrow etc..
According to these above-mentioned illustrative personalized makeups labels, matched in the data established according to above-mentioned Fig. 1-Fig. 3
Makeups keyword, to retrieve the description document for including these makeups keywords, the description text including modeling scheme and makeup knowledge
Shelves, such as round face are suitble to the makeups scheme of that type, and arched eyebrows is suitble to which type of makeups scheme, and so on.
S403, it is weighed according to the makeups keyword being matched in the difference that corresponding makeups describe document different location and assign
Weight carries out weight to the makeups scheme for including the makeups keyword being matched to and counts to obtain the recommendation index of makeups scheme, according to
Makeups scheme is presented to the user in the recommendation index.
In the present embodiment, if makeups keyword only includes makeup knowledge keyword, since data include multiple makeups
Knowledge document, including it is described makeup knowledge keyword description document may have it is multiple, therefore, in order to accurately suitable to user
Makeups scheme, according to the keyword being matched to description document different location there is the weight assigned, count corresponding makeups
The weight of scheme, for example, the weight for the same makeup knowledge keyword that can occur to different location in description document carries out letter
Single adduction operation obtains the comprehensive weight of corresponding makeups scheme, obtains the recommendation index of makeups scheme according to the comprehensive weight,
Comprehensive weight is bigger, recommends index higher, higher with the personalized makeups tag match degree of user.
In other embodiments, the situation that modeling scheme keyword is only included for makeups keyword is only included similar to above-mentioned
The situation of makeup knowledge keyword, repeats no more in detail.
And in other embodiments, if makeups keyword includes that makeup knowledge keyword and modeling scheme are crucial simultaneously
Word, corresponding makeups scheme includes modeling scheme and makeup knowledge, then when generating the weight of makeups scheme, to modeling scheme and
The comprehensive weight of makeup knowledge is fitted, for example the comprehensive weight of the two modeling schemes and the comprehensive weight of makeup knowledge are asked
With the comprehensive weight of acquisition makeups scheme；It is also possible to be carried out respectively according to the weight of modeling scheme and the weight of makeup knowledge
The recommendation of modeling scheme and makeup knowledge.
Fig. 5 is the personalized makeups aided process flow sheet schematic diagram of the embodiment of the present invention five；As shown in figure 5, comprising:
S501, the real-time facial image for obtaining user match in the five features classification model pre-established and is obtained
Take the personalized makeups label at family；
In the present embodiment, step S501 is recorded similar to above-described embodiment correlation, and details are not described herein.
It is S502, any or more according to the birthday by information of the preference information of user, the Weather information in user location, user
The combination producing personalization of kind assists makeups label；
It is unlike the embodiments above, in the present embodiment, in addition to there is the personalization of five features, also add user's
Preference information, the Weather information in user location, the personalization that the birthday by information of user characterizes assist makeups label.
In the present embodiment, different users has different styles of wearing the clothes, and collects this kind of preference information of user and returns
One change processing, such as style of wearing the clothes are worn the clothes style classification are as follows: extremely letter is neutral, pure and fresh literature and art, vigor maiden, former place unusual charm, mixed
It is gently ripe to take changeable, intellectual, charming imperial elder sister, makings.
The weather in user location is obtained according to the geographical location of the user reflected on intelligent terminal and is normalized
Processing obtains: fine day, cloudy day, rainy day etc..
Red-letter day situation, while the lucky color of dynamic acquisition user's constellation are judged according to the birthday by information of user, are sorted out, than
Such as are as follows: brown, red, yellow, green, navy blue, white, blue, gold, black, pink, blue-green, grey, dark brown
Color, purple, orange etc..
S503, makeups keyword is carried out according in the personalized makeups the label in any embodiment database
Matching；
Related personalization makeups label is referring to above-mentioned related embodiment, and details are not described herein.
S504, makeups key is carried out according in the personalized database for assisting makeups label described in any embodiment
The matching of word；
In the present embodiment, the matching of similar above-mentioned personalized makeups label, retrieval includes in the database and personalization is beautiful
The description document of the makeups keyword of adornment tag match, for example be suitble to the makeups scheme at cloudy day, be suitble to lucky color for red makeups
Scheme etc..
KMP algorithm can be used to realize in matching process in above-mentioned steps S503 and 504, repeats no more in detail.
S505, it is weighed according to the makeups keyword being matched in the difference that corresponding makeups describe document different location and assign
Weight carries out weight to the makeups scheme for including the makeups keyword being matched to and counts to obtain the recommendation index of makeups scheme；
In the present embodiment, the weight of makeups scheme counts similar above-mentioned related embodiment and records, and details are not described herein.When same
When determining makeups scheme according to personalization auxiliary makeups label and personalized makeups label, can be according in same makeups scheme
The weight of the personalized auxiliary makeups label and personalized makeups label that include simultaneously generates the comprehensive weight of the makeups scheme.
S506, level-one and secondary merger sequence are carried out respectively to the makeups scheme for including the makeups keyword being matched to, with
Makeups scheme is presented to the user according to the recommendation index.
In the present embodiment, a retrieval trunk module can be set, second level merger sequence is carried out to makeups scheme, removal repeats
Document improves recall precision.
Fig. 6 is that library apparatus structure schematic diagram is built in six makeups library of the embodiment of the present invention；As shown in fig. 6, comprising:
Document establishes unit 601, for extracting makeups keyword from the makeups scheme of acquisition, according to the makeups scheme
Content structure generate makeups document is described；
Weight-assigning unit 602, for the institute for the different location appearance in the content structure that the makeups describe document
It states makeups keyword and assigns different weights.
Optionally, in one embodiment of this invention, the document is established unit 601 and is further used for from collectedization
Makeup knowledge keyword is extracted in adornment knowledge, the makeups scheme includes the makeup knowledge；According in the makeup knowledge
Hold structural generation makeup knowledge description document；
The weight-assigning unit 602 is further used for being different in the content structure of the makeup knowledge description document
It includes: in the content structure of the makeup knowledge description document that the makeups keyword that position occurs, which assigns different weights,
The makeup knowledge keyword that middle different location occurs assigns different weights.
Optionally, in one embodiment of this invention, the document establishes unit 601 and is further used for moulding from acquisition
Modeling scheme keyword is extracted in scheme, the makeups scheme includes the modeling scheme, according to the content of the modeling scheme
Structural generation modeling scheme describes document；
The weight-assigning unit 602 is further used for being different in the content structure of the makeup knowledge description document
It includes: to describe different location in document in the modeling scheme that the makeups keyword that position occurs, which assigns different weights,
The modeling scheme keyword occurred assigns different weights.
Fig. 7 is the personalized makeups assistant apparatus structure schematic diagram of the embodiment of the present invention seven；As shown in fig. 7, comprising:
Tag match unit 701, in the five features classification model pre-established to the real-time people for obtaining user
Face image carries out the personalized makeups label that matching obtains user；
Keywords matching unit 702, for the database according to the personalized makeups label described in any embodiment
The middle matching for carrying out makeups keyword；
Makeups recommendation unit 703, for describing document different location in corresponding makeups according to the makeups keyword being matched to
And the different weights assigned, weight is carried out to the makeups scheme for including the makeups keyword being matched to and counts to obtain makeups scheme
Recommend index, makeups scheme is presented to the user according to the recommendation index.
Optionally, in one embodiment of this invention, further includes: assisted tag unit 704, for the preference according to user
Information, the Weather information in user location, the combination producing personalization that the birthday by information of user is any one or more assist makeups
Label；
The Keywords matching unit 702 is further used for according to the personalized auxiliary makeups label in any embodiment
The matching of makeups keyword is carried out in the database.
It will be understood by those skilled in the art that the embodiment of the embodiment of the present invention can provide as method, apparatus (equipment) or
Computer program product.Therefore, the embodiment of the present invention can be used complete hardware embodiment, complete software embodiment or combine soft
The form of the embodiment of part and hardware aspect.Moreover, it wherein includes to calculate that the embodiment of the present invention, which can be used in one or more,
Computer-usable storage medium (including but not limited to magnetic disk storage, CD-ROM, the optical memory of machine usable program code
Deng) on the form of computer program product implemented.
The embodiment of the present invention referring to according to the method for the embodiment of the present invention, device (equipment) and computer program product
Flowchart and/or the block diagram describes.It should be understood that can be realized by computer program instructions every in flowchart and/or the block diagram
The combination of process and/or box in one process and/or box and flowchart and/or the block diagram.It can provide these computers
Processor of the program instruction to general purpose computer, special purpose computer, Embedded Processor or other programmable data processing devices
To generate a machine, so that generating use by the instruction that computer or the processor of other programmable data processing devices execute
In the dress for realizing the function of specifying in one or more flows of the flowchart and/or one or more blocks of the block diagram
It sets.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
Although the preferred embodiment of the embodiment of the present invention has been described, once a person skilled in the art knows bases
This creative concept, then additional changes and modifications may be made to these embodiments.So the following claims are intended to be interpreted as
Including preferred embodiment and fall into all change and modification of range of embodiment of the invention.Obviously, those skilled in the art
Various modification and variations can be carried out to the embodiment of the present invention without departing from the spirit and scope of the embodiment of the present invention.If in this way,
If these modifications and variations of the embodiment of the present invention belong within the scope of claim of the embodiment of the present invention and its equivalent technologies,
Then the embodiment of the present invention is also intended to include these modifications and variations.
Claims (18)
1. a kind of personalization makeups householder method characterized by comprising
Matching is carried out to the real-time facial image for obtaining user in the five features classification model pre-established and obtains user's
Personalized makeups label；
Makeups keyword is extracted from the makeups scheme of acquisition, and makeups description text is generated according to the content structure of the makeups scheme
Shelves, to establish makeups library；
Different power is assigned for the makeups keyword that different location occurs in the content structure that the makeups describe document
Weight；
The matching of makeups keyword is carried out in makeups library according to the personalized makeups label；
According to the makeups keyword being matched in the different weights that corresponding makeups describe document different location and assign, to including
The makeups scheme for the makeups keyword being fitted on carries out weight and counts to obtain the recommendation index of makeups scheme, according to the recommendation index
Makeups scheme is presented to the user；
Wherein, the five features classification model generating process includes:
(1) select facial image sample as learning sample；
(2) manual characteristic point label is carried out to the learning sample of selection, enables the set of the v characteristic point position marked
Constitute shape S, S=(x1, y1；x2,y2；...xv,yv);
(3) the shape S is normalized;
(4) principal component analysis is carried out to normalized shape S, obtains the average shape S0 and preceding n spy of corresponding training set
The corresponding shape eigenvectors Si of value indicative；
(5) arbitrary shape S is expressed with linear equation:
。
2. the method according to claim 1, wherein extracting makeups keyword packet from the makeups scheme of acquisition
It includes:
Makeup knowledge keyword is extracted from collected makeup knowledge, the makeups scheme includes the makeup knowledge；
Generating makeups to describe document according to the content structure of the makeups scheme includes: the content structure according to the makeup knowledge
Generate makeup knowledge description document；
It is assigned for the makeups keyword that different location occurs in the content structure of the makeup knowledge description document different
Weight include: for different location occurs in the content structure of the makeup knowledge description document the makeup knowledge it is crucial
Word assigns different weights.
3. the method according to claim 1, wherein extracting makeups keyword packet from the makeups scheme of acquisition
It includes:
Modeling scheme keyword is extracted from the modeling scheme of acquisition, the makeups scheme includes the modeling scheme；
Generating makeups to describe document according to the content structure of the makeups scheme includes: the content structure according to the modeling scheme
It generates modeling scheme and describes document；
It is assigned for the makeups keyword that different location occurs in the content structure that the modeling scheme describes document different
Weight include: for the modeling scheme describe the modeling scheme keyword that different location in document occurs assign it is different
Weight.
4. according to the method described in claim 2, it is characterized in that, the makeup knowledge includes: knowledge entry and dressing data,
The knowledge entry is used to characterize the static description of makeup, and the dressing data are used to characterize the Dynamic profiling of makeup.
5. according to the method described in claim 4, it is characterized in that, the knowledge entry includes personalized makeup skill, changes
One of adornment region, makeup gimmick or a variety of combinations；The dressing data include that dressing description, makeup step and correspondence are more
One of media content or a variety of combinations.
6. according to the method described in claim 5, it is characterized by further comprising: generate the management training coarse of the knowledge entry with
And the management training coarse of the dressing data, and the management training coarse of the management training coarse and the dressing data to the knowledge entry
Carry out Data Integration.
7. according to the method described in claim 6, it is characterized by further comprising: according to tree hierarchy structure to the knowledge item
Purpose management training coarse and the management training coarse of the dressing data carry out Content Management.
8. method according to claim 1-7, which is characterized in that further include: to the makeups keyword of extraction into
Row normalized.
9. according to the method described in claim 8, it is characterized by further comprising: to more in the way of carrying out concurrent type frog retrieval
A makeups describe document and are stored.
10. according to the method described in claim 9, it is characterized in that, to multiple makeups describe document carry out fragment storage, with into
The retrieval of row concurrent type frog.
11. the method according to claim 1, wherein further include: to sample facial image carry out feature mark and
Faceform is extracted and established, generates feature classification model to carry out classification to five features.
12. according to the method for claim 11, which is characterized in that carry out feature mark to sample facial image and extraction is built
Faceform is stood, includes: to carry out classification generation feature classification model to five features
Face shape modeling is carried out according to the face characteristic of extraction and the modeling of face texture respectively obtains face shape model and people
Face texture model；
Faceform is established according to the face shape model and face texture model, generates spy to carry out classification to five features
Levy classification model.
13. the method according to claim 1, wherein further include: according to where the preference information of user, user
The Weather information on ground, the combination producing personalization that the birthday by information of user is any one or more assist makeups label；
The matching of makeups keyword is carried out in the makeups library according to the personalized auxiliary makeups label.
14. the method according to claim 1, wherein further include: to the beauty for including the makeups keyword being matched to
Adornment scheme carries out level-one and secondary merger sequence respectively.
15. a kind of personalization makeups auxiliary device characterized by comprising
Tag match unit, in the five features classification model pre-established to the real-time facial image for obtaining user into
Row matching obtains the personalized makeups label of user；
Document establishes unit, for extracting makeups keyword from the makeups scheme of acquisition, according to the content of the makeups scheme
Structural generation makeups describe document, to establish makeups library；
Weight-assigning unit, for being closed for the makeups that different location occurs in the content structure that the makeups describe document
Keyword assigns different weights；
Keywords matching unit, for carrying out of makeups keyword in the makeups library according to the personalized makeups label
Match；
Makeups recommendation unit is assigned for describing document different location in corresponding makeups according to the makeups keyword being matched to
Different weights, the recommendation for counting to obtain makeups scheme to the makeups scheme progress weight for including the makeups keyword being matched to refer to
Makeups scheme is presented to the user according to the recommendation index in number；
The process of the five features classification model of the foundation includes:
(1) select facial image sample as learning sample；
(2) manual characteristic point label is carried out to the learning sample of selection, enables the set of the v characteristic point position marked
Constitute shape S, S=(x1, y1；x2,y2；...xv,yv);
(3) the shape S is normalized;
(4) principal component analysis is carried out to normalized shape S, obtains the average shape S0 and preceding n spy of corresponding training set
The corresponding shape eigenvectors Si of value indicative；
(5) arbitrary shape S is expressed with linear equation:
。
16. device according to claim 15, which is characterized in that the document is established unit and is further used for from collecting
Makeup knowledge in extract makeup knowledge keyword, the makeups scheme includes the makeup knowledge；According to the makeup knowledge
Content structure generate makeup knowledge description document；
The weight-assigning unit be further used for for it is described makeup knowledge description document content structure in different location go out
It is different in the content structure of the makeup knowledge description document that the existing makeups keyword, which assigns different weights to include:,
The makeup knowledge keyword that position occurs assigns different weights.
17. device according to claim 15 or 16, which is characterized in that the document is established unit and is further used for from adopting
Modeling scheme keyword is extracted in the modeling scheme of collection, the makeups scheme includes the modeling scheme, according to the moulding side
The content structure of case generates modeling scheme and describes document；
The weight-assigning unit be further used for in the content structure that the modeling scheme describes document different location go out
It includes: to describe different location in document in the modeling scheme and occurring that the existing makeups keyword, which assigns different weights,
The modeling scheme keyword assigns different weights.
18. device according to claim 15, which is characterized in that further include: assisted tag unit, for according to user's
Preference information, the Weather information in user location, the combination producing personalization that the birthday by information of user is any one or more assist
Makeups label；
The Keywords matching unit is further used for being carried out in the makeups library according to the personalized auxiliary makeups label
The matching of makeups keyword.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610873350.8A CN106446207B (en) | 2016-09-30 | 2016-09-30 | Makeups library banking process, personalized makeups householder method and its device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610873350.8A CN106446207B (en) | 2016-09-30 | 2016-09-30 | Makeups library banking process, personalized makeups householder method and its device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106446207A CN106446207A (en) | 2017-02-22 |
| CN106446207B true CN106446207B (en) | 2019-11-12 |
Family
ID=58172717
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610873350.8A Active CN106446207B (en) | 2016-09-30 | 2016-09-30 | Makeups library banking process, personalized makeups householder method and its device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106446207B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108121957B (en) * | 2017-12-19 | 2021-09-03 | 麒麟合盛网络技术股份有限公司 | Method and device for pushing beauty material |
| CN108596094B (en) * | 2018-04-24 | 2021-02-05 | 杭州数为科技有限公司 | Character style detection system, method, terminal and medium |
| CN110020187A (en) * | 2018-05-22 | 2019-07-16 | 京东方科技集团股份有限公司 | A kind of makeup proposal recommending method, device and relevant device |
| CN112819718A (en) * | 2021-02-01 | 2021-05-18 | 深圳市商汤科技有限公司 | Image processing method and device, electronic device and storage medium |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101599886B (en) * | 2008-06-05 | 2013-01-02 | 华为技术有限公司 | Query method, system and device in distributed structured network |
| CN102693288A (en) * | 2012-04-27 | 2012-09-26 | 上海申视汽车新技术有限公司 | Automatic recommendation method for makeup scheme |
| CN104750801A (en) * | 2015-03-24 | 2015-07-01 | 华迪计算机集团有限公司 | Generation method and system of structured document |
| CN105138648A (en) * | 2015-08-26 | 2015-12-09 | 宇龙计算机通信科技(深圳)有限公司 | Information recommendation method and user terminal |
-
2016
- 2016-09-30 CN CN201610873350.8A patent/CN106446207B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN106446207A (en) | 2017-02-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101324961B (en) | Human face portion three-dimensional picture pasting method in computer virtual world | |
| Liu et al. | Wow! you are so beautiful today! | |
| CN106446207B (en) | Makeups library banking process, personalized makeups householder method and its device | |
| CN110443189B (en) | Face attribute identification method based on multitask multi-label learning convolutional neural network | |
| CN104205168B (en) | Makeup application assistance device, makeup application assistance method, and makeup application assistance program | |
| CN105513125B (en) | Composograph generating means and method, the recording medium for executing this method | |
| Zhang et al. | Computer models for facial beauty analysis | |
| CN105426850A (en) | Human face identification based related information pushing device and method | |
| CN107153805A (en) | Customize makeups servicing unit and method | |
| CN107506559B (en) | Star face shaping makeup recommendation method and device based on face similarity analysis | |
| CN108537628A (en) | Method and system for creating customed product | |
| CN104331564B (en) | Adorn guidance method and terminal device based on terminal device | |
| CN109310196A (en) | Makeup auxiliary device and cosmetic auxiliary method | |
| US11562536B2 (en) | Methods and systems for personalized 3D head model deformation | |
| JP2004094917A (en) | Virtual makeup device and method therefor | |
| CN104463938A (en) | Three-dimensional virtual make-up trial method and device | |
| CN108537126A (en) | A kind of face image processing system and method | |
| CN108846792A (en) | Image processing method, device, electronic equipment and computer-readable medium | |
| US11587288B2 (en) | Methods and systems for constructing facial position map | |
| US11417053B1 (en) | Methods and systems for forming personalized 3D head and facial models | |
| CN112819718A (en) | Image processing method and device, electronic device and storage medium | |
| CN114283052A (en) | Method and device for cosmetic transfer and training of cosmetic transfer network | |
| Yi et al. | Quality metric guided portrait line drawing generation from unpaired training data | |
| WO2022197429A1 (en) | Methods and systems for extracting color from facial image | |
| CN104933742A (en) | Automatic cartoon image generation method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |