CN108364286A - A kind of visual wrinkle of skin appraisal procedure and system - Google Patents
A kind of visual wrinkle of skin appraisal procedure and system Download PDFInfo
- Publication number
- CN108364286A CN108364286A CN201810085942.2A CN201810085942A CN108364286A CN 108364286 A CN108364286 A CN 108364286A CN 201810085942 A CN201810085942 A CN 201810085942A CN 108364286 A CN108364286 A CN 108364286A
- Authority
- CN
- China
- Prior art keywords
- skin
- wrinkle
- face
- pixel
- cloud server
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/44—Detecting, measuring or recording for evaluating the integumentary system, e.g. skin, hair or nails
- A61B5/441—Skin evaluation, e.g. for skin disorder diagnosis
- A61B5/442—Evaluating skin mechanical properties, e.g. elasticity, hardness, texture, wrinkle assessment
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30088—Skin; Dermal
Abstract
The invention discloses a kind of visual wrinkle of skin appraisal procedure and systems, belong to Skin Detection field;Method includes that the facial image of same face is associated with using image collecting device shooting, and facial image is sent in Cloud Server by skin detection mirror;Cloud Server handles to obtain the unit normal vector of each pixel by stereo photometry and processing obtains the case depth information of each pixel to form face's stereo-picture of face, then face's stereo-picture is judged using the wrinkle of skin assessment models that training is formed in advance, to obtain the wrinkle of skin assessment result of the different face locations in corresponding face stereo-picture, wrinkle of skin assessment result is associated with to the corresponding face location of indicating template and is exported with the wrinkle assessment result for forming image conversion.Advantage is:User can more directly check the wrinkle assessment result of different face locations in face's stereo-picture in the user terminal, and then take and targetedly maintain.
Description
Technical field
The present invention relates to Skin Detection field more particularly to a kind of visual wrinkle of skin appraisal procedure and it is
System.
Background technology
With the promotion of people's quality of the life, more and more people especially women begins to focus on the skin of itself,
More and more maintenance products for skin also occupy very important status on the market.Various beautifying apparatus simultaneously
Structure also continuously emerges in the visual field of people, and user can be by understanding the skin health state of oneself in relevant beauty mechanism
And skin care service etc. is obtained, when making assessment to the health status of user's skin in the beautician in beauty mechanism often
Rely on personal subjective experience, it can be seen that whether can make correctly assessment depending on the professional standards of beautician and a
People's experience, and the beauty mechanism of profession not only needs to pay higher beauty expense, while the solid door that requires a great deal of time
Shop obtains beauty services.
Therefore it is badly in need of a kind of equipment that can make skin detection at any time for the skin of user, the content of detection may include
Whether canthus has whether wrinkle, face have decree line etc..
It is current in the market, although there are some skin detection equipment, such as skin detection instrument etc., these skins inspection
The price of measurement equipment is more expensive, and operate it is extremely complex, be not suitable for user use at home.These skin detections are set simultaneously
Standby testing principle is generally only the data processing that plane is carried out by the collected sensing data of some sensors, be will not relate to
The problem of being rebuild to three dimensional face, therefore preferable detection result is had for skins such as Oily, skin water shortages, and
It is bad for the detection result of the skin of facial skin wrinkle, it cannot be satisfied the demand of user.
Invention content
According to the prior art protect present in the above problem, the skill of a kind of wrinkle of skin appraisal procedure and system is now provided
Art scheme, it is intended in conjunction with the function of skin detection and vanity mirror, the ability of wrinkle of skin three-dimensional reconstruction and three-dimensional assessment be provided, made
The wrinkle of skin situation that user is capable of accurate perception itself is obtained, user experience is promoted.
Above-mentioned technical proposal specifically includes:
A kind of visual wrinkle of skin appraisal procedure, wherein wrinkled to the skin of face using a skin detection mirror
Line is assessed, and the skin detection mirror includes an image collecting device, and the skin detection mirror remotely connects a Cloud Server, also
Including:
Step S1 is shot using described image harvester under different light source irradiations and is associated with same people respectively
The facial image is sent in the Cloud Server by the different facial images of face, the skin detection mirror;
Step S2, the Cloud Server handle to obtain the list of each pixel in the facial image by stereo photometry
Position normal vector;
Step S3, the case depth that the Cloud Server handles to obtain each pixel according to the unit normal vector are believed
Breath;
Step S4, the Cloud Server are located at according to the case depth information of each pixel and each pixel
Location information on the facial image forms face's stereo-picture of face;
Step S5 is directed to different face location training in the Cloud Server and is formed with corresponding wrinkle of skin and comments in advance
Model is estimated, according to the different face locations in the face stereo-picture using the corresponding wrinkle of skin assessment models to institute
It states face's stereo-picture to be assessed, to obtain the wrinkle of skin assessment result of the different face locations;
Step S6, the described Cloud Server will make the corresponding wrinkle of skin assessment models of the different face location
For indicating template, and the wrinkle of skin assessment result is associated with to the corresponding indicating template to form the wrinkle of image conversion
Assessment result;
Step S7, the described Cloud Server exports the wrinkle assessment result of described image to remotely connecting the cloud
The user terminal of server, so that user checks.
Preferably, visual wrinkle of skin appraisal procedure, wherein in the step S1, at least in 3 different light
Irradiation lower shooting in source obtains the facial image, and all light sources are not on same straight line.
Preferably, visual wrinkle of skin appraisal procedure, wherein the light source is 12.
Preferably, visual wrinkle of skin appraisal procedure, wherein the step S2 is specifically included:
Step S21, using a pixel as pending pixel;
Step S22, judge the pending pixel whether be bloom pixel:
If so, turning to step S23;
If it is not, then turning to step S24;
Step S23 is replaced, then using the pixel of the non-bloom of same position in the different facial images
Turn to step S24;
Step S24, processing obtain the surface normal of the pending pixel;
Step S25 handles to obtain the unit normal vector of the pixel according to the surface normal of the pixel.
Preferably, visual wrinkle of skin appraisal procedure, wherein the step S3 is specifically included:
For each pixel, preset constraint formulations are established according to the unit normal vector, and according to it is described about
Beam formula manipulation obtains the case depth information of the pixel.
Preferably, visual wrinkle of skin appraisal procedure, wherein the constraint formulations are specially:
Wherein,
V1And V2It is the tangential vector of the facial surface of face;
N is used to indicate the unit vector of the pixel;
X and y is used to indicate the location information of plane of the pixel on facial image;
Z is for indicating the case depth information.
Preferably, visual wrinkle of skin appraisal procedure, wherein in the step S5, the face location corresponds to
The wrinkle of skin assessment models include:The first assessment models for the crow's foot for assessing face;
First assessment models are gathered according to pre-prepd first training, learn to be formed using deep neural network;
The first training set includes multiple first training datas pair, and each first training data centering includes
The assessment score of the stereoscopic image data of face with crow's foot of different shapes and the corresponding stereoscopic image data.
Preferably, visual wrinkle of skin appraisal procedure, wherein in the step S5, the face location corresponds to
The wrinkle of skin assessment models include:The second assessment models for the decree line for assessing face;
Second assessment models are gathered according to pre-prepd second training, learn to be formed using deep neural network;
The second training set includes multiple second training datas pair, and each second training data centering includes
The assessment score of the stereoscopic image data of face with decree line of different shapes and the corresponding stereoscopic image data.
A kind of visual wrinkle of skin assessment system, wherein including:
An image collecting device is arranged in the skin detection mirror in skin detection mirror, under different light source irradiations,
Shoot the different facial images for being associated with same face respectively using described image harvester;
Cloud Server remotely connects the skin detection mirror, and the skin detection mirror is for sending the facial image
To the Cloud Server, the Cloud Server uses the appraisal procedure of above-mentioned wrinkle of skin, according to the facial image to people
The wrinkle of skin of face is assessed, and assessment result is exported to the user terminal remotely being connect with the Cloud Server, for
User checks.
The advantageous effect of above-mentioned technical proposal is:In conjunction with the function of skin detection and vanity mirror, wrinkle of skin three-dimensional is provided
Rebuild the ability with three-dimensional assessment so that user is capable of the wrinkle of skin situation of accurate perception itself, while user finally obtains
The assessment result of wrinkle of skin be to be shown the different assessment results of face location using face stereo-picture as indicating template
In the corresponding face location of indicating template, aspect user more intuitively checks that the assessment result of face's different location promotes user's body
It tests.
Description of the drawings
Fig. 1 is in the preferred embodiment of the present invention, and a kind of overall procedure of visual wrinkle of skin appraisal procedure shows
It is intended to;
Fig. 2 be the present invention preferred embodiment in, on the basis of Fig. 1, the idiographic flow schematic diagram of step S2;
Fig. 3 is the concrete structure schematic diagram of visual wrinkle of skin assessment system in the preferred embodiment of the present invention;
Fig. 4 is the schematic diagram of skin detection mirror in the preferred embodiment of the present invention.
Specific implementation mode
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 describes, 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, those of ordinary skill in the art obtained under the premise of not making creative work it is all its
His embodiment, shall fall within the protection scope of the present invention.
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the present invention can phase
Mutually combination.
The invention will be further described in the following with reference to the drawings and specific embodiments, but not as limiting to the invention.
According to the above-mentioned problems in the prior art, a kind of visual wrinkle of skin appraisal procedure, the party are now provided
In method, wrinkle assessment is carried out to the skin of face using a skin detection mirror, skin detection mirror includes an image collecting device,
Skin detection mirror remotely connects a Cloud Server, further includes:
Step S1 is shot using image collecting device under different light source irradiations and is associated with same face respectively
Facial image is sent in Cloud Server by different facial images, skin detection mirror;
Step S2, Cloud Server handle to obtain the unit normal direction of each pixel in facial image by stereo photometry
Amount;
Step S3, Cloud Server handle to obtain the case depth information of each pixel according to unit normal vector;
Step S4, Cloud Server are located at facial image according to the case depth information of each pixel and each pixel
On location information formed face face's stereo-picture;
Step S5 is directed to different face location training in the Cloud Server and is formed with corresponding wrinkle of skin and comments in advance
Model is estimated, according to the different face locations in the face stereo-picture using the corresponding wrinkle of skin assessment models to institute
It states face's stereo-picture to be assessed, to obtain the wrinkle of skin assessment result of the different face locations;
Step S6, the described Cloud Server will make the corresponding wrinkle of skin assessment models of the different face location
For indicating template, and the wrinkle of skin assessment result is associated with to the corresponding indicating template to form the wrinkle of image conversion
Assessment result;
Step S7, the described Cloud Server exports the wrinkle assessment result of described image to remotely connecting the cloud
The user terminal of server, so that user checks.
Specifically, in the present embodiment, above-mentioned skin detection mirror can be carried out using common vanity mirror as the ontology of minute surface
Transformation, i.e., above-mentioned skin detection mirror is made one can be daily in offer people in this way to the vanity mirror of people's normal use
The function of wrinkle of skin detect and assess can be integrated into while use in vanity mirror.
In the present embodiment, an image collecting device is set on above-mentioned skin detection mirror, which can be
Camera on skin detection mirror is set, the surface of skin detection mirror is particularly provided at, facilitates shooting user people
The general image of face.
In the present embodiment, the image collecting device on skin detection mirror shoots the same face respectively under different light sources
Facial image, above-mentioned different light source be different directions point light source, be placed on the different positions around skin detection mirror
It sets, to form the different facial images of same face under different point light source irradiations.It is same in these facial images
Pixel is identical relative to the position of facial image, differs only in the brightness of pixel, can be hereinafter described in detail.
In the present embodiment, skin detection mirror remotely connects a Cloud Server, and above-mentioned shooting is associated with same
The different facial images of face are uploaded in Cloud Server.Specifically, a wireless telecommunications can be set in skin detection mirror
Module such as WiFi module is remotely connected to Cloud Server by indoor routing.
In the present embodiment, after the facial image that cloud server is uploaded to skin detection mirror, it will be clapped under different point light sources
The different facial images taken the photograph carry out integrated treatment.Specifically, it is according to each pixel on facial image in Cloud Server
What point was handled, and the light source direction that differs only in point light source of the same pixel in different facial images is different
And the brightness of the pixel thereby resulted in is different, therefore the attribute of different facial images can be integrated into the same pixel
In point, such as the attribute of a pixel includes the letters such as its light source direction and brightness value in different facial images
Breath.
In the present embodiment, Cloud Server is handled each pixel by stereo photometry, to obtain different pixels
The unit normal vector of point, then handles to obtain the case depth information of each pixel according to unit normal vector.There is surface deep
After spending information, Cloud Server can establish face's stereo-picture of face according to pixel, i.e., carry out 3-D view to face
It rebuilds.
In the present embodiment, after the face's stereo-picture for obtaining face, the wrinkle of skin formed using advance training is assessed
Model assesses face's stereo-picture, to form the wrinkle of skin assessment result of face's stereo-picture different location;
Then using face's stereo-picture as indicating template, while the wrinkle of skin assessment result that assessment is obtained is about aobvious
Show the corresponding face location of template, and then form the wrinkle of skin assessment result of image conversion, by Cloud Server by image conversion
Wrinkle of skin assessment result be remotely issued to corresponding user terminal, for user terminal with after assessment result, user can be with
The wrinkle assessment result of different face locations in face's stereo-picture is more directly checked in the user terminal, and then is taken and be directed to
The maintenance of property.
Specifically, in above-mentioned wrinkle of skin assessment models, input data is the stereo-picture of face, and output data is vertical for this
The assessment result in corresponding wrinkle of skin region in body image.Further, which can be assessment score.Above-mentioned skin
The concrete operating principle of skin wrinkle assessment models can hereinafter be described in detail.
To sum up, in technical solution of the present invention, using the image collecting device on skin detection mirror, in different light source knit stitch
Different facial images shot respectively to same face, Cloud Server according to these facial images, using stereo photometry into
Row processing ultimately forms the stereo-picture of face, and stereo-picture is sent into the assessment models of image recognition and wrinkle assessment
In, the corresponding assessment result of final output.Compared with prior art, in technical solution of the present invention, using stereo photometry and
The three-dimensional image that the modes such as case depth information establish face is solved, and wrinkle assessment, Neng Gouzhen are carried out to the image
The situation of reaction user wrinkle of skin on the face on the spot.And the function of skin detection is integrated into vanity mirror, user day is facilitated
It is often used.
It should be noted that shooting integrality based on the considerations of facial image, the skin detection mirror in the present invention is should be
Vanity mirror with certain minute surface area, such as it is placed in the vanity mirror made up on table, or it is directly installed on bedroom metope
On vanity mirror etc., avoid using the technology that the present invention can be realized with the hand held middle vanity mirror used more small and exquisite on the market
Scheme.
In the preferred embodiment of the present invention, in step S1, at least shooting obtains face under 3 different light source irradiations
Image, all light sources are not on same straight line.
Further, in preferred embodiment of the invention, above-mentioned light source is specially 12, i.e., around skin detection mirror
12 light sources are set, and the light source direction of these light sources is different, and all light sources are not on same straight line.
In the preferred embodiment of the present invention, as shown in Figure 2, above-mentioned steps S2 is specifically included:
Step S21, using a pixel as pending pixel;
Step S22, judge pending pixel whether be bloom pixel:
If so, turning to step S23;
If it is not, then turning to step S24;
Step S23 is replaced using the pixel of the non-bloom of same position in different facial images, is subsequently diverted to
Step S24;
Step S24, processing obtain the surface normal of pending pixel;
Step S25 handles to obtain the unit normal vector of pixel according to the surface normal of pixel.
Specifically, in the present embodiment, since the pixel of bloom is unable to judge accurately its wrinkle of skin image, into
Row face image needs to reject the pixel of bloom first when rebuilding.The mode of rejecting is:Pixel is judged by threshold method first
Point whether be bloom pixel (its brightness value is bloom pixel higher than preset threshold value).Subsequently, for being judged as
The pixel of bloom is substituted using the pixel of the non-bloom of same position in other facial images, specifically, such as above
Described, the attribute of different facial images is participated in calculating by the attribute as each pixel, such as pixel
Include brightness value of the pixel in different facial images in attribute value, then for a pixel, rejects high
The principle of light image vegetarian refreshments is exactly to reject the brightness value for being higher than predetermined threshold value in a pixel, leaves behind the brightness of non-bloom
Value.
In the present embodiment, after rejecting bloom pixel, the pixel of all non-blooms is handled successively, with
The surface normal of each pixel is obtained, and obtains the unit normal vector of each pixel in turn.
The process for solving surface normal specifically includes:
For coloured image, the brightness of each pixel is indicated by the value of tri- Color Channels of R, G, B.This reality
It applies in example and the processing procedure of surface normal is illustrated by taking R values as an example.
Assuming that the face surface in the present invention meets ideal Lambertian scattering models, then the luminance equation of pixel
It should be:
IR=ρRL·nR; (1)
Wherein, I is used to indicate that the brightness of pixel, L to be used to indicate the direction vector of light source, and ρ is for indicating pixel pair
The texture reflectivity for the surface region answered, n are used to indicate the unit normal vector of the corresponding surface region of pixel.
The direction vector of above-mentioned each light source knows in advance, may be used shot under different light source directions it is one group high
Light black ball image is realized, can obtain the side of illumination by searching for the position of highlight in each bloom black ball image
To the light source direction as point light source.
In above-mentioned steps, the brightness of the pixel of non-bloom under the irradiation of different light sources can be represented as:
IR=(I1R,I2R,...,IqR)T; (2)
Wherein, q is used to indicate the quantity of light source, wherein eliminating the light source corresponding to bloom pixel, T is turning for matrix
Set calculating.
Correspondingly, the unit normal vector of each pixel can be represented as:
nR=(n1R,n2R,...,nqR)T; (3)
By calculating, the direction vector of q light source is should be:
It is multiplied by L simultaneously at the both ends of above-mentioned formula (1)-1It can obtain:
L-1IR=ρR·nR; (5)
Then the equation left side vector field homoemorphism of above-mentioned formula (5) is the value of texture reflectivity, and the direction of left side vector is
The direction of surface normal.
After obtaining the surface normal in the channels R of pixel, the unit normal vector in the channels R is calculated, i.e.,:
nR=(nRa,nRb,nRc)T; (6)
Wherein, (a, b, c) is vector direction coordinate of the pixel in normal direction quantity space.
The unit normal vector of the pixel channel B and the channels G can be respectively obtained according to above-mentioned formula (1)-(6):
nB=(nBa,nBb,nBc)T; (7)
nG=(nGa,nGb,nGc)T; (8)
The unit normal vector of a final pixel can take the average value of the unit normal vector of above-mentioned RGB channel, and table
It is shown as:
N=(na,nb,nc)T; (9)
The gradient (- n of each pixel can also be calculated according to above-mentioned formula (9)a/nc,-nb/nc), and can be with
Establish the normal direction spirogram of facial skin wrinkle.
In the preferred embodiment of the present invention, above-mentioned steps S3 is specifically included:
For each pixel, preset constraint formulations are established according to unit normal vector, and handle according to constraint formulations
To the case depth information of pixel.
Further, be based on tangent plane principle, the normal direction of body surface every bit should with it is tangential vertical, then can establish
Following constraint formulations:
Wherein, V1And V2It is the tangential vector on face surface, (x, y, z) is spatial coordinate of the pixel on face surface,
Wherein (x, y) is the plane coordinates of pixel, i.e. location information (the institute in step S4 of plane of the pixel on facial image
The location information stated), z is for indicating depth value of the pixel on facial image.
There is the facial image of m pixel for a width, so that it may to obtain 2m constraint equation, the per unit system of each pixel
Vector n is known, and the location information x and y of the plane of each pixel are also known, and depth value z is a scalar, because
The constraint equation of all pixels point can be formed an equation matrix by this, and each picture is obtained with by solving equation matrix
The case depth information of vegetarian refreshments.
After obtaining the case depth information of each pixel, just there is spatial coordinate (x, y, z), can thus build people
Face's stereo-picture of face and carry out 3D display.
In the preferred embodiment of the present invention, in above-mentioned steps S5, the corresponding wrinkle of skin of face location assesses mould
Type includes:The first assessment models for the crow's foot for assessing face;
First assessment models are gathered according to pre-prepd first training, learn to be formed using deep neural network;
First training set includes multiple first training datas pair, and each first training data centering includes having difference
The assessment score of the stereoscopic image data of the face of the crow's foot of shape and corresponding stereoscopic image data.
Specifically, in the present embodiment, multiple first training datas pair are prepared in advance, each first training data centering includes
The stereo-picture of one width crow's foot, and the assessment score to being obtained after stereo-picture progress manually evaluation.First training
The quantity of data pair is more, and the assessment result for the first assessment models that training is formed is more accurate.
In the present embodiment, above-mentioned first assessment models dedicated for assess face crow's foot.In the mistake of practical application
Cheng Zhong, above-mentioned first assessment models find out the stereo-picture of crow's foot from whole face's stereo-picture first, then right
The stereo-picture of crow's foot is assessed, one assessment result of final output, which may be used assessment score
Mode provides.
In another preferred embodiment of the present invention, in above-mentioned steps S5, the corresponding wrinkle of skin assessment of face location
Model includes:The second assessment models for the decree line for assessing face;
Second assessment models are gathered according to pre-prepd second training, learn to be formed using deep neural network;
Second training set includes multiple second training datas pair, and each second training data centering includes having difference
The assessment score of the stereoscopic image data of the face of the decree line of shape and corresponding stereoscopic image data.
Specifically, in the present embodiment, multiple second training datas pair are prepared in advance, each second training data centering includes
The stereo-picture of one width decree line, and the assessment score to being obtained after stereo-picture progress manually evaluation.Second training number
According to quantity it is more, training formed the second assessment models assessment result it is more accurate.
In the present embodiment, above-mentioned second assessment models dedicated for assess face decree line.In the process of practical application
In, above-mentioned second assessment models find out the stereo-picture of decree line from whole face's stereo-picture first, then to decree
The stereo-picture of line is assessed, one assessment result of final output, which equally may be used the side of assessment score
Formula provides.
In another preferred embodiment of the present invention, above-mentioned first can be used simultaneously in wrinkle of skin assessment models
Assessment models and the second assessment models, so as to comprehensively assess user crow's foot and decree line situation.
In the preferred embodiment of the present invention, based on the above visual wrinkle of skin appraisal procedure, now carry
For a kind of visual wrinkle of skin assessment system A, the system is specifically as shown in Figure 3, including:
An image collecting device A11 is set in skin detection mirror A1, skin detection mirror A1, for being irradiated in different light sources
Under, shoot the different facial images for being associated with same face respectively using image collecting device A11;
Cloud Server A2, long-range to connect skin detection mirror A1, skin detection mirror A1 is used to facial image being sent to cloud clothes
Be engaged in device A2, and Cloud Server A2 uses the appraisal procedure of the above wrinkle of skin, according to facial image to the skin of face
Wrinkle is assessed, and assessment result is exported to the user terminal B remotely being connect with Cloud Server A2, so that user checks.
Further, the specular appearance schematic diagram of above-mentioned skin detection mirror A1 is as shown in Figure 4, skin detection mirror A1's
Point light sources of the plurality of LEDs lamp bead A13 as different directions can also be set around minute surface ontology A12, carry out assistant images acquisition
Device A11 shoots face.
The foregoing is merely preferred embodiments of the present invention, are not intended to limit embodiments of the present invention and protection model
It encloses, to those skilled in the art, should can appreciate that all with made by description of the invention and diagramatic content
Equivalent replacement and obviously change obtained scheme, should all be included within the scope of the present invention.
Claims (9)
1. a kind of visual wrinkle of skin appraisal procedure, which is characterized in that using a skin detection mirror to the skin of face into
Row wrinkle is assessed, and the skin detection mirror includes an image collecting device, the skin detection mirror remotely one cloud service of connection
Device further includes:
Step S1 is shot using described image harvester under different light source irradiations and is associated with same face respectively
The facial image is sent in the Cloud Server by different facial images, the skin detection mirror;
Step S2, the Cloud Server handle to obtain the per unit system of each pixel in the facial image by stereo photometry
Vector;
Step S3, the Cloud Server handle to obtain the case depth information of each pixel according to the unit normal vector;
Step S4, the Cloud Server are located at described according to the case depth information of each pixel and each pixel
Location information on facial image forms face's stereo-picture of face;
Step S5 is directed to different face location training in the Cloud Server and is formed with corresponding wrinkle of skin assessment mould in advance
Type, according to the different face locations in the face stereo-picture using the corresponding wrinkle of skin assessment models to the face
Portion's stereo-picture is assessed, to obtain the wrinkle of skin assessment result of the different face locations;
Step S6, the described Cloud Server will be using the corresponding wrinkle of skin assessment models of the different face location as aobvious
Show template, and the wrinkle of skin assessment result is associated with to the corresponding indicating template and is assessed with the wrinkle for forming image conversion
As a result;
Step S7, the described Cloud Server exports the wrinkle assessment result of described image to remotely connecting the cloud service
The user terminal of device, so that user checks.
2. visual wrinkle of skin appraisal procedure as described in claim 1, which is characterized in that in the step S1, at least
Shooting obtains the facial image under 3 different light source irradiations, and all light sources are not on same straight line.
3. visual wrinkle of skin appraisal procedure as claimed in claim 2, which is characterized in that the light source is 12.
4. visual wrinkle of skin appraisal procedure as described in claim 1, which is characterized in that the step S2 is specifically wrapped
It includes:
Step S21, using a pixel as pending pixel;
Step S22, judge the pending pixel whether be bloom pixel:
If so, turning to step S23;
If it is not, then turning to step S24;
Step S23 is replaced using the pixel of the non-bloom of same position in the different facial images, is subsequently diverted to
Step S24;
Step S24, processing obtain the surface normal of the pending pixel;
Step S25 handles to obtain the unit normal vector of the pixel according to the surface normal of the pixel.
5. visual wrinkle of skin appraisal procedure as described in claim 1, which is characterized in that the step S3 is specifically wrapped
It includes:
For each pixel, preset constraint formulations are established according to the unit normal vector, and according to constraint public affairs
Formula handles to obtain the case depth information of the pixel.
6. visual wrinkle of skin appraisal procedure as claimed in claim 5, which is characterized in that the constraint formulations are specific
For:
Wherein,
V1And V2It is the tangential vector of the facial surface of face;
N is used to indicate the unit vector of the pixel;
X and y is used to indicate the location information of plane of the pixel on facial image;
Z is for indicating the case depth information.
7. visual wrinkle of skin appraisal procedure as described in claim 1, which is characterized in that described in the step S5
The corresponding wrinkle of skin assessment models of face location include:The first assessment models for the crow's foot for assessing face;
First assessment models are gathered according to pre-prepd first training, learn to be formed using deep neural network;
The first training set includes multiple first training datas pair, and each first training data centering includes having
The assessment score of the stereoscopic image data of the face of crow's foot of different shapes and the corresponding stereoscopic image data.
8. visual wrinkle of skin appraisal procedure as described in claim 1, which is characterized in that described in the step S5
The corresponding wrinkle of skin assessment models of face location include:The second assessment models for the decree line for assessing face;
Second assessment models are gathered according to pre-prepd second training, learn to be formed using deep neural network;
The second training set includes multiple second training datas pair, and each second training data centering includes having
The assessment score of the stereoscopic image data of the face of decree line of different shapes and the corresponding stereoscopic image data.
9. a kind of visual wrinkle of skin assessment system, which is characterized in that including:
An image collecting device is arranged in the skin detection mirror in skin detection mirror, under different light source irradiations, using
Described image harvester shoots the different facial images for being associated with same face respectively;
Cloud Server, remotely connects the skin detection mirror, and the skin detection mirror is used to the facial image being sent to institute
Cloud Server is stated, the Cloud Server is using the appraisal procedure of the wrinkle of skin as described in any one of claim 1-8, root
The wrinkle of skin of face is assessed according to the facial image, and assessment result is exported to the Cloud Server and is remotely connected
The user terminal connect, so that user checks.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810085942.2A CN108364286A (en) | 2018-01-29 | 2018-01-29 | A kind of visual wrinkle of skin appraisal procedure and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810085942.2A CN108364286A (en) | 2018-01-29 | 2018-01-29 | A kind of visual wrinkle of skin appraisal procedure and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108364286A true CN108364286A (en) | 2018-08-03 |
Family
ID=63007451
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810085942.2A Pending CN108364286A (en) | 2018-01-29 | 2018-01-29 | A kind of visual wrinkle of skin appraisal procedure and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108364286A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110301894A (en) * | 2019-06-26 | 2019-10-08 | 杜剑波 | Detection device before skin surface liquid layered water is removed |
CN110956623A (en) * | 2019-11-29 | 2020-04-03 | 深圳和而泰家居在线网络科技有限公司 | Wrinkle detection method, apparatus, device, and computer-readable storage medium |
CN111028214A (en) * | 2019-12-05 | 2020-04-17 | 北京美医医学技术研究院有限公司 | Skin detection device based on cloud platform |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101299267A (en) * | 2008-07-02 | 2008-11-05 | 北京中星微电子有限公司 | Method and device for processing human face image |
US20090195545A1 (en) * | 2008-01-31 | 2009-08-06 | University Fo Southern California | Facial Performance Synthesis Using Deformation Driven Polynomial Displacement Maps |
CN102819864A (en) * | 2012-08-14 | 2012-12-12 | 北京航空航天大学 | Luminosity diagram-based geometry and surface material reconstructing method |
CN103827916A (en) * | 2011-09-22 | 2014-05-28 | 富士胶片株式会社 | Wrinkle detection method, wrinkle detection device and wrinkle detection program, as well as wrinkle evaluation method, wrinkle evaluation device and wrinkle evaluation program |
CN105787974A (en) * | 2014-12-24 | 2016-07-20 | 中国科学院苏州纳米技术与纳米仿生研究所 | Establishment method for establishing bionic human facial aging model |
CN106127250A (en) * | 2016-06-24 | 2016-11-16 | 深圳市唯特视科技有限公司 | A kind of face method for evaluating quality based on three dimensional point cloud |
CN106469302A (en) * | 2016-09-07 | 2017-03-01 | 成都知识视觉科技有限公司 | A kind of face skin quality detection method based on artificial neural network |
CN107392858A (en) * | 2017-06-16 | 2017-11-24 | 广东欧珀移动通信有限公司 | Image highlight area processing method, device and terminal device |
-
2018
- 2018-01-29 CN CN201810085942.2A patent/CN108364286A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090195545A1 (en) * | 2008-01-31 | 2009-08-06 | University Fo Southern California | Facial Performance Synthesis Using Deformation Driven Polynomial Displacement Maps |
CN101299267A (en) * | 2008-07-02 | 2008-11-05 | 北京中星微电子有限公司 | Method and device for processing human face image |
CN103827916A (en) * | 2011-09-22 | 2014-05-28 | 富士胶片株式会社 | Wrinkle detection method, wrinkle detection device and wrinkle detection program, as well as wrinkle evaluation method, wrinkle evaluation device and wrinkle evaluation program |
CN102819864A (en) * | 2012-08-14 | 2012-12-12 | 北京航空航天大学 | Luminosity diagram-based geometry and surface material reconstructing method |
CN105787974A (en) * | 2014-12-24 | 2016-07-20 | 中国科学院苏州纳米技术与纳米仿生研究所 | Establishment method for establishing bionic human facial aging model |
CN106127250A (en) * | 2016-06-24 | 2016-11-16 | 深圳市唯特视科技有限公司 | A kind of face method for evaluating quality based on three dimensional point cloud |
CN106469302A (en) * | 2016-09-07 | 2017-03-01 | 成都知识视觉科技有限公司 | A kind of face skin quality detection method based on artificial neural network |
CN107392858A (en) * | 2017-06-16 | 2017-11-24 | 广东欧珀移动通信有限公司 | Image highlight area processing method, device and terminal device |
Non-Patent Citations (7)
Title |
---|
ALI SOHAIB: ""Use of Photometric Stereo for the Accurate Modelling of Three-Dimensional Skin Microrelief"", 《A THESIS OF THE UNIVERSITY OF THE WEST OF ENGLAND》 * |
ALI SOHAIB等: ""3D In-Vivo Measurement of Skin Topography Using Photometric Stereo"", 《3RD INTERNATIONAL CONFERENCE ON 3D BODY SCANNING TECHNOLOGIES》 * |
ALI SOHAIB等: ""In vivo measurement of skin microrelief using photometric stereo in the presence of interreflections"", 《JOURNAL OF THE OPTICAL SOCIETY OF AMERICA》 * |
E.JULIASTUTI等: ""Prototype of Photometric Stereo System using 6 and 9 LED light source to reconstruct human skin texture"", 《APPLIED MECHANICS AND MATERIALS》 * |
SHAHZAD ANWAR等: ""3D Skin Texture Analysis: A Neural Network and Photometric Stereo Perspective"", 《3RD INTERNATIONAL CONFERENCE ON 3D BODY SCANNING TECHNOLOGIES》 * |
张家林: ""基于立体视觉的皮肤表面检测技术的研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
许鹏翔: ""皮肤评价系统的研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110301894A (en) * | 2019-06-26 | 2019-10-08 | 杜剑波 | Detection device before skin surface liquid layered water is removed |
CN110956623A (en) * | 2019-11-29 | 2020-04-03 | 深圳和而泰家居在线网络科技有限公司 | Wrinkle detection method, apparatus, device, and computer-readable storage medium |
CN110956623B (en) * | 2019-11-29 | 2023-11-07 | 深圳数联天下智能科技有限公司 | Wrinkle detection method, wrinkle detection device, wrinkle detection equipment and computer-readable storage medium |
CN111028214A (en) * | 2019-12-05 | 2020-04-17 | 北京美医医学技术研究院有限公司 | Skin detection device based on cloud platform |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108324247A (en) | A kind of designated position wrinkle of skin appraisal procedure and system | |
CN108363964A (en) | A kind of pretreated wrinkle of skin appraisal procedure and system | |
Yuan et al. | A mixed reality virtual clothes try-on system | |
CN110609617B (en) | Apparatus, system and method for virtual mirror | |
CN105701447B (en) | Guest-meeting robot | |
US9269157B2 (en) | Methods for extracting objects from digital images and for performing color change on the object | |
US20140225977A1 (en) | Devices, systems and methods of virtualizing a mirror | |
CN110335343A (en) | Based on RGBD single-view image human body three-dimensional method for reconstructing and device | |
CN108364286A (en) | A kind of visual wrinkle of skin appraisal procedure and system | |
Zhao et al. | 3-D reconstruction of human body shape from a single commodity depth camera | |
CN107211165A (en) | Devices, systems, and methods for automatically delaying video display | |
CN110074788B (en) | Body data acquisition method and device based on machine learning | |
KR101792541B1 (en) | Smart Device Skin care service system using Skin Diagnostic Information | |
CN103871099A (en) | 3D simulation matching processing method and system based on mobile terminal | |
CN107976157A (en) | A kind of wireless hand-held three-dimensional scanning device in acquisition object surface three-dimensional morphology | |
CN106251376B (en) | One kind is towards colored structures pumped FIR laser and edge extracting method | |
US11875468B2 (en) | Three-dimensional (3D) image modeling systems and methods for determining respective mid-section dimensions of individuals | |
CN108334836A (en) | A kind of wrinkle of skin appraisal procedure and system | |
CN107533580A (en) | The multiple parameters for the biological information that part as live plant is shown over the display | |
CN108154142A (en) | A kind of wrinkle of skin appraisal procedure and system based on speech recognition | |
CN105184856A (en) | Two-phase human skin three-dimensional reconstruction method based on density matching | |
CN102419861A (en) | Color image correcting method based on topology subdivision of uniform color space | |
Pan et al. | Accuracy of RGB-D camera-based and stereophotogrammetric facial scanners: a comparative study | |
CN107343151B (en) | Image processing method, device and terminal | |
Harari et al. | A computer-based method for the assessment of body-image distortions in anorexia-nervosa patients |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
AD01 | Patent right deemed abandoned | ||
AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20210528 |