CN109299714A - ROI template generation method, ROI extracting method and system, equipment, medium - Google Patents
ROI template generation method, ROI extracting method and system, equipment, medium Download PDFInfo
- Publication number
- CN109299714A CN109299714A CN201710609651.4A CN201710609651A CN109299714A CN 109299714 A CN109299714 A CN 109299714A CN 201710609651 A CN201710609651 A CN 201710609651A CN 109299714 A CN109299714 A CN 109299714A
- Authority
- CN
- China
- Prior art keywords
- region
- roi
- image
- roi region
- face
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/255—Detecting or recognising potential candidate objects based on visual cues, e.g. shapes
Abstract
The application provides a kind of ROI template generation method, ROI extracting method and system, equipment, medium.Wherein, the ROI region extracting method includes: the organic region in identification face-image obtained；Preset ROI region template is adjusted to the edge of the organic region identified, obtains the ROI region of the face-image.The application area data handled when solving the problems, such as in the prior art using the progress face-image processing of ROI region template is not accurate enough.
Description
Technical field
This application involves a kind of face-image processing techniques, extract more particularly to a kind of ROI template generation method, ROI
Method and system, equipment, medium.
Background technique
ROI region (area-of-interest) is machine vision, in image procossing, from processed image with box, circle, ellipse
The modes such as circle, irregular polygon sketch the contours of region to be treated.In field of image processing, area-of-interest (ROI) be from
The image-region selected in image, this region can be used as image analysis emphasis of interest, draw a circle to approve the region so as into
Row is further processed.The processing time can be reduced using ROI region, increases precision.
In face-image processing technique, technical staff carries out machine learning using multi-panel portion sample image to obtain face
ROI region, and obtained ROI region is utilized to carry out U.S. face, special efficacy, pathological analysis, face assessment etc..Due to utilization technology
Result obtained from processing can be intuitive obvious, and face-image processing technique has rapid development in medical and beauty treatment fields.
Currently, face detection instrument used in medical and beauty treatment fields utilizes the multiple face ROI areas determining through study
Domain template, and cover and be used in different user face-image, and then pathological analysis or face assessment are carried out to the face-image of user.So
And the face structure of each user is different, use counts the face-image region fallen into facial ROI region template
According to processing, so that ROI template and the disproportionate problem of practical ROI region must be will appear for user's monomer.Same problem
It is technical to also appear in some facial landscaping treatments.
Summary of the invention
In view of the foregoing deficiencies of prior art, the application is designed to provide a kind of ROI region template generation side
Method, ROI region extracting method, system and equipment are carried out at face-image using ROI region template in the prior art for solving
Handled area data not accurate enough problem when reason.
In order to achieve the above objects and other related objects, the first aspect of the application provides a kind of extraction side of ROI region
Method, comprising: identify the organic region in face-image obtained；By preset ROI region template to the organ area identified
The edge in domain adjusts, and obtains the ROI region of the face-image.
In the certain embodiments of the first aspect, organic region in identification face-image obtained
Mode includes: based on the organic region in face-image described in preset ROI region template extraction.
It is described based on face described in preset ROI region template extraction in the certain embodiments of the first aspect
The mode of organic region in image includes: that the face-image is carried out image segmentation based on ROI region template；Identification divides
The organic region in image-region after cutting.
In the certain embodiments of the first aspect, it is described by preset ROI region template to the organ identified
The mode of the edge adjustment in region includes: to extend the ROI region template to adjacent organic region, until meeting and corresponding device
The critical condition in official region.
It is described to extend ROI region template to adjacent organic region in the certain embodiments of the first aspect
Mode includes: that the characteristic point in the ROI region template is shifted to adjacent organic region, and based on the characteristic point position after movement
Sketch the contours ROI region.
In the certain embodiments of the first aspect, the organic region includes: ocular.
The application second aspect also provides a kind of ROI region template generation method, comprising: carries out feature using sample image
Training is to obtain average sample image；The characteristic point for sketching the contours the ROI region is determined based on the average sample image, and
ROI region template is sketched the contours based on selected characteristic point.
In the certain embodiments of the second aspect, the characteristic point includes: based in the average sample image
For sketching the contours the pixel of ROI region template contours.
The application third aspect provides a kind of face-image processing equipment, comprising: storage device, for storing ROI region
Template, face-image and the program for extracting ROI region in the face-image；Processing unit, for executing described program
To extract the ROI region in the face-image according to any the method as above.
In the certain embodiments of the third aspect, the face-image processing equipment further include: photographic device is used
In intake face-image and save in the storage device.
In the certain embodiments of the third aspect, the face-image processing equipment further include: shooting prompt dress
It sets, before being located at photographic device, for prompting tester to put on the head in photographic device intake direction.
The application fourth aspect provides a kind of computer equipment, comprising: storage device, for storing sample image and use
In the program for generating ROI region template；Processing unit, for executing described program according to such as above-mentioned any ROI region mould
Plate generation method generates ROI region template.
The 5th aspect of the application provides a kind of extraction system of ROI region, comprising: identification module is obtained for identification
Face-image in organic region；ROI extraction module, for by preset ROI region template to the organic region identified
Edge adjustment, obtain the ROI region of the face-image.
In the certain embodiments of the 5th aspect, the identification module is specifically based on preset ROI region template
Extract the organic region in the face-image.
In the certain embodiments of the 5th aspect, the identification module is used for will be described based on ROI region template
Face-image carries out image segmentation；And the organic region in the image-region after dividing for identification.
It is described 5th aspect certain embodiments in, the ROI extraction module be used for by the ROI region template to
Adjacent organic region extends, until meeting the critical condition with corresponding organ region.
In the certain embodiments of the 5th aspect, the ROI extraction module is specifically used for the ROI region mould
Characteristic point in plate shifts to adjacent organic region, and sketches the contours ROI region based on the characteristic point position after movement.
In the certain embodiments of the 5th aspect, the organic region includes: ocular.
The 6th aspect of the application provides a kind of ROI region template generating system, comprising: training module, for utilizing sample
Image carries out feature training to obtain average sample image；ROI region template generation module, for being based on the average sample figure
As determining the characteristic point for sketching the contours the ROI region, and ROI region template is sketched the contours based on selected characteristic point.
In the certain embodiments of the 6th aspect, the characteristic point includes: based in the average sample image
For sketching the contours the pixel of ROI region template contours.
The aspect of the application the 7th provides a kind of storage medium, which is characterized in that be stored with acquired face-image and
For carrying out the program of ROI region extraction；Wherein, described program is when being executed by processor, according to any extraction side as above
Step in method extracts the ROI region in the face-image.
As described above, the ROI template generation method of the application, ROI extracting method and system, equipment, medium, have following
The utility model has the advantages that by by ROI region template into face-image organic region it is mobile in the way of, realize and be directed to different face-images
ROI region adaptive adjustment, thus solve in the prior art using ROI region template carry out face-image processing when it is locating
The not accurate enough problem of the area data of reason.
Detailed description of the invention
Fig. 1 is the flow chart of the application ROI region template generation method in one embodiment.
Fig. 2 is the flow chart of the application ROI region extracting method in one embodiment.
Fig. 3-A is the ocular identified using the application ROI region extracting method and ROI region template in face-image
In schematic diagram.
Fig. 3-B is using in the application ROI region extracting method that ROI region template is mobile to obtain phase to ocular
Answer a kind of schematic diagram of embodiment of ROI region in face-image.
Fig. 3-C is using in the application ROI region extracting method that ROI region template is mobile to obtain phase to ocular
Answer the schematic diagram of another embodiment of ROI region in face-image.
Fig. 4 is the architecture diagram of the application ROI region template generating system method in one embodiment.
Fig. 5 is the architecture diagram of the application ROI region extraction system in one embodiment.
Fig. 6 is to be able to carry out the computer equipment of ROI region template generation method in a kind of embodiment using the application
In structural schematic diagram.
Fig. 7 is to be able to carry out the face-image processing equipment of ROI region extracting method in a kind of embodiment party using the application
Structural schematic diagram in formula.
Fig. 8 is that the face-image processing equipment of ROI region extracting method is able to carry out using the application in another implementation
Structural schematic diagram in mode.
Specific embodiment
Presently filed embodiment is illustrated by particular specific embodiment below, those skilled in the art can be by this explanation
Content disclosed by book understands other advantages and effect of the application easily.
It should be noted that this specification structure depicted in this specification institute accompanying drawings, ratio, size etc., only to cooperate
The bright revealed content of book is not limited to the enforceable limit of the application so that those skilled in the art understands and reads
Fixed condition, therefore do not have technical essential meaning, the modification of any structure, the change of proportionate relationship or the adjustment of size, not
It influences still fall in techniques disclosed in this application content under the effect of the application can be generated and the purpose that can reach and obtain
In the range of capable of covering.Meanwhile it is cited such as "upper", "lower", "left", "right", " centre " and " one " in this specification
Term is merely convenient to being illustrated for narration, rather than to limit the enforceable range of the application, the change of relativeness or tune
It is whole, under the content of no substantial changes in technology, when being also considered as the enforceable scope of the application.
It should be understood that singular " one ", "one" and "the" are intended to also include plural as used in herein
Form, unless having opposite instruction in context it will be further understood that term "comprising", " comprising " show that there are described
Feature, step, operation, element, component, project, type, and/or group, but it is not excluded for one or more other features, step, behaviour
Work, element, component, project, the presence of type, and/or group, appearance or addition term "or" used herein and "and/or" quilt
It is construed to inclusive, or means any one or any combination.Therefore, " A, B or C " or " A, B and/or C " mean " with
Descend any one: A；B；C；A and B；A and C；B and C；A, B and C ".Only when element, function, step or the combination of operation are in certain sides
When inherently mutually exclusive under formula, it just will appear the exception of this definition.
For medical and beauty treatment fields, the facial ROI region of user, which refers to, is concentrated with wrinkle, pore, any in spot
Kind or a variety of regions.Therefore, it in order to reduce computer equipment to the calculation amount of face-image, improves to the accurate of analysis result
Degree, computer equipment is by the image data in selective analysis face-image septum reset ROI region.In order to determine monomer user's face
The ROI region mould generated based on face-image sample (following abbreviation sample images) is provided previously in the ROI region of image, the application
Plate.The ROI region template be through several sample images training and obtain ROI region profile, face-image position and
The information such as the characteristic point for indicating ROI region.
Referring to Fig. 1, the application provides a kind of ROI region template generation side for medical and beauty treatment fields septum reset image
Method.The generation method is mainly executed by computer equipment.The computer equipment includes but is not limited to: PC, service
Device, industrial computer, embedded computer etc..The computer equipment may include memory, Memory Controller, one or
Multiple processing units (CPU), Peripheral Interface, RF circuit, voicefrequency circuit, loudspeaker, microphone, input/output (I/O) subsystem
System, touch screen, other outputs or control equipment and outside port.These components pass through one or more communication bus or letter
Number line is communicated.The component of the computer equipment can have more or fewer components than above-mentioned, or with different groups
Part configuration.
In step s 110, feature training is carried out using sample image.
Firstly, the user's face image collection that the sample image is shot based on equipment used in hospital, beauty clinic
Obtained by.The computer equipment carries out feature training using several collected sample images.Specifically, the computer is set
It is standby to utilize such as PCA (Principal Component Analysis principal component analysis) algorithm or LDA (Linear
The analysis of Discriminant Analysis linear discriminent) etc. machine learning algorithms feature is carried out to acquired sample image
Training is to obtain average sample image.
For example, sample image is converted into N-dimensional vector Ti, then the average image of all sample images is calculated, determine every width
The deviation matrix of all sample images is sought covariance to obtain feature sky by the deviation matrix of sample image and the average image
Between.
For example, step S111: obtaining set S, every sample image T comprising M sample imagesiPass through spatial alternation
Mode can be converted the vector S={ T an of N-dimensional1,T2,...,Tn, i=1,2 ..., n.
Step S112: after getting face vector set S, the average image is calculatedWherein, M is sample
Amount of images.
Step S113: the difference of every sample image and the average image is calculatedAnd it is inclined to obtain all sample images
Poor matrix φ={ φ1,φ2,...,φM}。
Step S114: by the deviation matrixΩ: Ω=φ of covariance matrix φ is obtained multiplied by its transposed matrixT；
The k nonzero eigenvalue of solution covariance matrix Ω and corresponding feature vector.
Step S115: eigenface is calculated.Specifically, it is transformation base with the feature vector that step S114 is solved, will owns
Sample image is respectively mapped in feature space and the feature vector merged after mapping obtains eigenface.The eigenface is as flat
Equal sample image.
Step S120 determines that the characteristic point for sketching the contours the ROI region is based on gained based on obtained average sample image
To average sample image determine sketch the contours the ROI region characteristic point be based on obtained average sample image determine sketch the contours institute
The characteristic point of ROI region is stated, and ROI region template is sketched the contours based on selected characteristic point.
Specifically, it is counted, will be obtained through statistics using principal component of the PCA algorithm to the preset quantity of the eigenface
Pixel of the characteristic value in the average sample image as the characteristic point for sketching the contours the ROI region.By each feature
Region obtained from point connection is as ROI region template.For example, using the approximate location of facial axial symmetry and face organ,
Each ROI region is sorted out to corresponding organ, with region shared by each organ of determination.Wherein, the characteristic point includes but is not limited to: right
It answers the pixel in region shared by two pixel of corner position on region contour shared by organ, connection organs and is manually set
Pixel etc..Obtained ROI region template includes but is not limited to: forehead region template, ocular template, cheekbone area
Template, corners of the mouth region template etc..
After obtaining ROI region template, each ROI region template is stored in facial figure by technical staff or computer equipment
As in processing equipment.The face-image processing equipment includes but is not limited to: face detection instrument, skinanalysis apparatus, mobile terminal
Deng.For example, the equipment is dedicated facial tester.For another example, the equipment is tablet computer etc..Referring to Fig. 2, the face
Image processing equipment is run according to the following steps, to determine the area ROI of each user's face on the basis of ROI region template
Domain.
In step S210, the organic region in face-image obtained is identified.
Here, the face-image can be received by network, or the photographic device by being arranged in equipment shoots and obtains.It is described
Face-image processing equipment identifies face based on the feature (such as shape feature, hair feature, color characteristic) of each organ of face
The organic region of image.For example, based on color characteristic and facial symmetry characteristic identification eyes and eyebrow region.
In some embodiments, this step can be based on the device in face-image described in preset ROI region template extraction
Official region.Specifically, this step includes step S211 and S212.
In step S211, the face-image is carried out according to organ site by image segmentation based on ROI region template.Tool
Body, the face-image processing equipment identifies the profile of face-image, further according to described in preset ROI region template
The ROI region template contours are placed on acquired face by position of the ROI region template contours in average sample image
In image.According still further to ROI region template by face-image piecemeal.
It should be noted that described image partitioned mode can divide face-image merely with ROI region template contours
Block label, and not necessarily carry out image block FIG pull handle.
In step S212, the organic region in the image-region after dividing is identified.Specifically, at the face-image
Facial area represented by equipment utilization ROI region template is managed, organic region is identified in face-image.For example, using corresponding to
The ROI region template of cheekbone area, in the image block adjacent with the region template identify monomer user ocular and/
Or nasal region.
In step S220, preset ROI region template is adjusted to the edge of the organic region identified, is obtained described
The ROI region of face-image.
Specifically, the face-image processing equipment will be outside the profile of ROI region template according to identified organic region
It connects or on the adjacent organs region, to obtain the ROI region in the face-image of corresponding monomer user.Wherein, described
The profile of ROI region template can extend to a kind of organic region, or extend to adjacent a variety of organic regions.Here, the ROI
The extension of region template includes that the outer of ROI region template contours extends to contraction.More specifically, the face-image processing equipment will
ROI region template extends to adjacent organic region, until meeting the critical condition with corresponding organ region.Wherein, described critical
Condition includes but is not limited to: two regions share external contour line, two regions share external profile point, two regions it is most short
Distance is less than preset value etc..
In certain embodiments, the face-image processing equipment determines organ in ROI region template contours and adjacent
The spacing of region contour is less than the partial contour of pre-determined distance threshold value, and the partial contour replacement in ROI region template is grown up to be a useful person
The corresponding part profile in official region, to obtain the ROI region of corresponding monomer user.As shown in Fig. 3-A and 3-B, examine to traversal formula
The spacing between cheekbone ROI region template contours 11 and adjacent ocular profile 12 is surveyed, to determine cheekbone ROI region template
Spacing between profile 11 and ocular profile 12, by least part profile in cheekbone ROI region template 11 to ocular
12 corresponding part contour extension (as shown in dotted line in Fig. 3-B), to obtain the cheekbone ROI region of corresponding monomer user.Wherein,
The partial contour extended is based on spacing determination less than pre-determined threshold.
For another example, as shown in Fig. 3-A and 3-C, the characteristic point in ROI region template is moved to adjacent organic region profile
It is dynamic, until the characteristic point moved is less than preset value with the spacing of organ contours or connects.Here, the moving direction of characteristic point and
Moving distance can be by artificially disposing.Alternatively, the characteristic point is moved to along place normal direction to organic region profile.It is described
Face-image processing equipment is using the characteristic point area defined after movement as the ROI region of corresponding monomer user.
It should be noted that the moving direction of characteristic point described in this application is only for example, rather than limit.The feature
Point can also be moved according to the shortest distance with adjacent organs region contour along shortest distance direction.
Obtained ROI region profile can also be carried out round and smooth processing by the face-image processing equipment.The face
Image processing equipment can emphasis detect determined by wrinkle condition, skin wrinkle, pore situation etc. in ROI region, and then pass through
Matching detection result determines the current situation of skin, such as determines the skin age, determines skin wrinkle lesion type.Alternatively, according to
Testing result such as beautifies at the face-image.
It should be noted that it should be appreciated by those skilled in the art that the above-mentioned cheekbone ROI region for monomer user really
Surely be only for example rather than the limitation to the application range.In fact, those skilled in the art are in combination with above-mentioned example and want true
The characteristics of fixed ROI region, carries out R & D design, and the technological means used based on technological frame described herein should all
The specific example being considered as within the scope of the application.
Referring to Fig. 4, the application provides a kind of ROI region template generation system for medical and beauty treatment fields septum reset image
System.The generation system mainly includes the software and hardware in computer equipment.The computer equipment includes but is not limited to: a
People's computer, server, industrial computer, embedded computer etc..The computer equipment may include memory, memory control
Device processed, one or more processing units (CPU), Peripheral Interface, RF circuit, voicefrequency circuit, loudspeaker, microphone, input/output
(I/O) subsystem, touch screen, other outputs or control equipment and outside port.These components pass through one or more communication
Bus or signal wire are communicated.The component of the computer equipment can have more or fewer components than above-mentioned, or have
Different component Configurations.
The generation system 4 includes training module 41 and ROI region template generation module 42.
The training module 41 is used to carry out feature training using sample image.
Firstly, the user's face image collection that the sample image is shot based on equipment used in hospital, beauty clinic
Obtained by.The training module 41 carries out feature training using several collected sample images.Specifically, the training module
41 using such as PCA (Principal Component Analysis principal component analysis) algorithm or LDA (Linear
The analysis of Discriminant Analysis linear discriminent) etc. machine learning algorithms feature is carried out to acquired sample image
It trains, can reflect the face organs such as eye, nose, mouth integrated distribution in a width average sample image in obtained feature space
Region.
For example, sample image is converted into N-dimensional vector Ti, then the average image of all sample images is calculated, determine every width
The deviation matrix of all sample images is sought covariance to obtain feature sky by the deviation matrix of sample image and the average image
Between.
For example, the training module 41 first obtains set S, every sample image T comprising M sample imagesiPass through sky
Between the mode that converts can be converted the vector S={ T an of N-dimensional1,T2,...,Tn, i=1,2 ..., n.
Then, the average image is calculated after getting face vector set S in the training module 41Its
In, M is sample image quantity.
The training module 41 calculates the difference of every sample image and the average image againAnd obtain all samples
This image deviations matrix φ={ φ1,φ2,...,φM}。
The training module 41 is by the deviation matrixMultiplied by its transposed matrix obtain covariance matrix Ω: Ω=
φ·φT；The k nonzero eigenvalue of solution covariance matrix Ω and corresponding feature vector.
The training module 41 calculates eigenface.It specifically, is transformation base, the training module 41 with described eigenvector
All sample images are respectively mapped in feature space and the feature vector merged after mapping obtains eigenface.The eigenface
As average sample image.
The ROI region template generation module 42 is used to sketch the contours the ROI based on the determination of obtained average sample image
The characteristic point in region determines that the characteristic point for sketching the contours the ROI region is based on obtained put down based on obtained average sample image
Equal sample image determines the characteristic point for sketching the contours the ROI region, and sketches the contours ROI region template based on selected characteristic point.
Specifically, the ROI region template generation module 42 is using PCA algorithm to the master of the preset quantity of the eigenface
Ingredient is counted, and it is described as being used to sketch the contours will to be counted pixel of the obtained characteristic value in the average sample image
The characteristic point of ROI region.Region obtained from the ROI region template generation module 42 connects each characteristic point is as the area ROI
Domain template.For example, approximate location of the ROI region template generation module 42 using facial axial symmetry and face organ, it will
Each ROI region is sorted out to corresponding organ, with region shared by each organ of determination.Wherein, the characteristic point includes but is not limited to: corresponding
The pixel of corner position on region contour shared by organ, region shared by two organs of connection pixel and be manually set
Pixel etc..Obtained ROI region template includes but is not limited to: forehead region template, ocular template, cheekbone area mould
Plate, corners of the mouth region template etc..
After obtaining ROI region template, each ROI region template is stored in facial figure by technical staff or computer equipment
As in processing equipment.The face-image processing equipment includes but is not limited to: face detection instrument, skinanalysis apparatus, mobile terminal
Deng.For example, the equipment is dedicated facial tester.For another example, the equipment is tablet computer etc..
Referring to Fig. 5, the application also provides a kind of ROI region extraction system.The extraction system is described comprising being mounted on
Software and hardware in face-image processing equipment, the extraction system 5 is by executing identification module 51 and ROI extraction module
52, realize the ROI region that single face-image is determined on the basis of ROI region template.
The identification module 51 organic region in face-image obtained for identification.
Here, the face-image can be received by network, or the photographic device by being arranged in equipment shoots and obtains.It is described
Identification module 51 is based on feature (such as shape feature, hair feature, color characteristic) the identification face-image of each organ of face
Organic region.For example, based on color characteristic and facial symmetry characteristic identification eyes and eyebrow region.
In some embodiments, the identification module 51 can be based on the figure of face described in preset ROI region template extraction
Organic region as in.Specifically, the identification module 51 executes following steps S211 and S212.
In step S211, the identification module 51 is based on ROI region template for the face-image according to organ site
Carry out image segmentation.Specifically, the identification module 51 identifies the profile of face-image, further according to preset ROI region mould
The ROI region template contours are placed on institute by position of the ROI region template contours in average sample image described in plate
In the face-image of acquisition.According still further to ROI region template by face-image piecemeal.
It should be noted that described image partitioned mode can divide face-image merely with ROI region template contours
Block label, and not necessarily carry out image block FIG pull handle.
In step S212, the identification module 51 identifies the organic region in the image-region after dividing.Specifically,
The identification module 51 identifies organic region using facial area represented by ROI region template in face-image.For example,
Using the ROI region template of corresponding cheekbone area, the eye of monomer user is identified in the image block adjacent with the region template
Portion region and/or nasal region.
The ROI extraction module 52 is used to adjust preset ROI region template to the edge of the organic region identified,
Obtain the ROI region of the face-image.
Specifically, the ROI extraction module 52 it is according to identified organic region that the profile of ROI region template is external,
Or on adjacent organs region, to obtain the ROI region in the face-image of corresponding monomer user.Wherein, the area ROI
The profile of domain template can extend to a kind of organic region, or extend to adjacent a variety of organic regions.Here, the ROI region
The extension of template includes that the outer of ROI region template contours extends to contraction.More specifically, the ROI extraction module 52 is by ROI region
Template extends to adjacent organic region, until meeting the critical condition with corresponding organ region.Wherein, the critical condition includes
But be not limited to: two regions share external contour line, two regions share external profile point, the shortest distance in two regions is less than
Preset value etc..
In certain embodiments, the ROI extraction module 52 determine in ROI region template contours with adjacent organic region
The spacing of profile is less than the partial contour of pre-determined distance threshold value, and the partial contour in ROI region template is substituted for organ area
The corresponding part profile in domain, to obtain the ROI region of corresponding monomer user.As shown in Fig. 3-A and 3-B, traversal formula cheekbone is detected
Spacing between bone ROI region template contours 11 and adjacent ocular profile 12, to determine cheekbone ROI region template contours
Spacing between 11 and ocular profile 12, by least part profile in cheekbone ROI region template 11 to ocular 12
Corresponding part contour extension (as shown in dotted line in Fig. 3-B), to obtain the cheekbone ROI region of corresponding monomer user.Wherein, prolonged
The partial contour stretched is based on spacing determination less than pre-determined threshold.
For another example, as shown in Fig. 3-A and 3-C, the characteristic point in ROI region template is moved to adjacent organic region profile
It is dynamic, until the characteristic point moved is less than preset value with the spacing of organ contours or connects.Here, the moving direction of characteristic point and
Moving distance can be by artificially disposing.Alternatively, the characteristic point is moved to along place normal direction to organic region profile.It is described
ROI extraction module 52 is using the characteristic point area defined after movement as the ROI region of corresponding monomer user.
It should be noted that the moving direction of characteristic point described in this application is only for example, rather than limit.The feature
Point can also be moved according to the shortest distance with adjacent organs region contour along shortest distance direction.
Obtained ROI region profile can also be carried out round and smooth processing by the ROI extraction module 52.The ROI is extracted
Module 52 can emphasis detect determined by wrinkle condition, skin wrinkle, pore situation etc. in ROI region, and then pass through matching inspection
It surveys result and determines the current situation of skin, such as determine the skin age, determine skin wrinkle lesion type.Alternatively, being tied according to detection
Fruit such as beautifies at the face-image.
It should be noted that it should be appreciated by those skilled in the art that the above-mentioned cheekbone ROI region for monomer user really
Surely be only for example rather than the limitation to the application range.In fact, those skilled in the art are in combination with above-mentioned example and want true
The characteristics of fixed ROI region, carries out R & D design, and the technological means used based on technological frame described herein should all
The specific example being considered as within the scope of the application.
Referring to Fig. 6, the application also provides a kind of computer equipment, it is used to generate facial ROI region template.The meter
Calculating machine equipment 2 includes: storage device 21 and processing unit 22.
The storage device 21 is for storing sample image and the program for generating ROI region template.The storage
Device 21 may include high-speed random access memory, and may also include nonvolatile memory, such as one or more disks
Store equipment, flash memory device or other non-volatile solid-state memory devices.In certain embodiments, storage device 21 can also wrap
The memory far from one or more processors is included, such as via the network attached storage of communication network access, wherein described
Communication network can be internet, one or more intranets, local area network (LAN), wide area network (WLAN), storage area network
(SAN) etc. or its is appropriately combined.It further include Memory Controller in storage device 21, the Memory Controller can control all
Such as access of the other assemblies to memory of CPU and Peripheral Interface etc.The component software being stored in storage device 21 includes
Operating system, communication module (or instruction set), contact/motion module (or instruction set), figure module (or instruction set), tactile are anti-
Present module (or instruction set), text input module (or instruction set) and program (or instruction set).
The processing unit 22 is used to execute ROI of the described program to generate face according to ROI region template generation method
Region template.
Here, the processing unit 22 includes processor, the processor operationally with memory and/or non-volatile
Memory coupling.More specifically, the instruction stored in memory and/or nonvolatile memory can be performed in terms of in processor
It calculates in equipment and executes operation, such as generate image data and/or image data is transferred to display circuit.In this way, processor can
It is patrolled including one or more general purpose microprocessors, one or more application specific processors (ASIC), one or more field-programmables
Collect array (FPGA) or any combination of them.
Processing unit 22 is also operationally coupled with network interface, will be calculated equipment and is communicatively coupled to either network.
For example, network interface, which can will calculate equipment, is connected to personal area network (PAN) (such as blueteeth network), local area network (LAN) (such as
802.11x Wi-Fi network), and/or wide area network (WAN).In addition, processing unit 22 is operatively coupled to power supply, the power supply
Electric power can be provided to the various parts calculated in equipment such as electronic console.In this way, power supply may include any suitable energy,
Such as rechargeable lighium polymer (Li-poly) battery and/or alternating current (AC) power adapter.
The processing unit 22 is also operatively coupled to the port I/O and input structure, which may make calculating to set
It is standby to be interacted with various other electronic equipments (such as dedicated for the instrument of skin detection or mobile terminal), the input knot
Structure aloows user to interact with equipment is calculated.Therefore, input structure may include button, keyboard, mouse, Trackpad
Deng.
The processing unit 22 is also operationally coupled with network interface, will be calculated equipment and is communicatively coupled to either net
Network.For example, network interface, which can will calculate equipment, is connected to personal area network (PAN) (such as blueteeth network), local area network (LAN)
(such as 802.11x Wi-Fi network), and/or wide area network (WAN) (injection 4G or LTE cellular network).
In this application, the processing unit 22 can be called based on the enabled instruction that input structure is inputted for generating
The program of ROI region template, and then when executing described program using the sample image in the storage device 21 as machine learning
Basis, to generate facial ROI region template.The process that the processing unit 22 executes described program can be as shown in Figure 1.
In step s 110, feature training is carried out using sample image.
Firstly, the user's face image collection that the sample image is shot based on equipment used in hospital, beauty clinic
Obtained by.The processing unit 22 carries out feature training using several collected sample images.Specifically, the processing unit
22 using such as PCA (Principal Component Analysis principal component analysis) algorithm or LDA (Linear
The analysis of Discriminant Analysis linear discriminent) etc. machine learning algorithms feature is carried out to acquired sample image
It trains, can reflect the face organs such as eye, nose, mouth integrated distribution in a width average sample image in obtained feature space
Region.
For example, sample image is converted into N-dimensional vector Ti, then the average image of all sample images is calculated, determine every width
The deviation matrix of all sample images is sought covariance to obtain feature sky by the deviation matrix of sample image and the average image
Between.
For example, step S111: obtaining set S, every sample image T comprising M sample imagesiPass through spatial alternation
Mode can be converted the vector S={ T an of N-dimensional1,T2,...,Tn, i=1,2 ..., n.
Step S112: after getting face vector set S, the average image is calculatedWherein, M is sample
Amount of images.
Step S113: the difference of every sample image and the average image is calculatedAnd it is inclined to obtain all sample images
Poor matrix φ={ φ1,φ2,...,φM}。
Step S114: by the deviation matrixΩ: Ω=φ of covariance matrix φ is obtained multiplied by its transposed matrixT；
The k nonzero eigenvalue of solution covariance matrix Ω and corresponding feature vector.
Step S115: eigenface is calculated.Specifically, it is transformation base with the feature vector that step S114 is solved, will owns
Sample image is respectively mapped in feature space and the feature vector merged after mapping obtains eigenface.The eigenface is as flat
Equal sample image.
Step S120 determines that the characteristic point for sketching the contours the ROI region is based on gained based on obtained average sample image
To average sample image determine sketch the contours the ROI region characteristic point be based on obtained average sample image determine sketch the contours institute
The characteristic point of ROI region is stated, and ROI region template is sketched the contours based on selected characteristic point.
Specifically, it is counted, will be obtained through statistics using principal component of the PCA algorithm to the preset quantity of the eigenface
Pixel of the characteristic value in the average sample image as the characteristic point for sketching the contours the ROI region.By each feature
Region obtained from point connection is as ROI region template.For example, using the approximate location of facial axial symmetry and face organ,
Each ROI region is sorted out to corresponding organ, with region shared by each organ of determination.Wherein, the characteristic point includes but is not limited to: right
It answers the pixel in region shared by two pixel of corner position on region contour shared by organ, connection organs and is manually set
Pixel etc..Obtained ROI region template includes but is not limited to: forehead region template, ocular template, cheekbone area
Template, corners of the mouth region template etc..
In addition to this, the computer equipment 2 further include: I/O subsystem, display circuit, telecommunication circuit etc..Wherein, described
I/O subsystem provide equipment input/output peripheral hardware and Peripheral Interface between interface, input/output peripheral such as display screen and
Other input/control devicess.The I/O subsystem includes displaying screen controller and one of equipment is exported or controlled for other
Or multiple input controllers.Other inputs or control equipment come from/are gone in one or more of input controller reception/transmissions
Electric signal.Other described input/control devicess may include keyboard, mouse etc..
The display circuit includes but is not limited to liquid crystal (LCD) display, Organic Light Emitting Diode (OLED) display etc..
However, the operation between different types of electronic console may be different.For example, LCD display can be by being based on liquid crystal
Orientation the brightness (for example, brightness value and/or gray value) of LCD display pixel is controlled to show picture frame.More
Body, in LCD display pixel, the voltage based on received image data can be applied to pixel electrode, thus raw
The electric field that pairs of liquid crystal is orientated.In contrast, OLED display can be by based on the luminous component for flowing through display picture element
The magnitude of the source current of (for example, OLED) carries out the brightness (for example, brightness value and/or gray value) of OLED display pixel
Control is to show picture frame.More specifically, OLED display pixel can be applied to the voltage based on the image data received
In switchgear (such as thin film transistor (TFT)) grid, with control flow to its luminous component source current magnitude.
The telecommunication circuit facilitate through one or more outside ports and with other equipment (such as server, from equipment) into
Row communication, and it further includes the various component softwares for handling RF circuit and/or the received data of outside port.Outer end
Mouthful (such as universal serial bus (USB), FIREWIRE etc.) is suitable for directly or through network (such as internet, Wireless LAN
Etc.) it is indirectly coupled to other equipment.
The processing unit 22 can utilize coupled telecommunication circuit or I/O subsystem, or will be obtained by technical staff
ROI region template is supplied to technical staff and/or facial treatment device.
It is corresponding, referring to Fig. 7, the application also provides a kind of face-image processing equipment.The face-image processing is set
It is standby to include but is not limited to: face detection instrument, skinanalysis apparatus, mobile terminal etc..For example, the face-image processing equipment is special
Facial test equipment.For another example, the face-image processing equipment is tablet computer etc..The face-image processing equipment 3
It include: storage device 31, processing unit 32.
The storage device 31 is for storing ROI region template, face-image and for extracting ROI in the face-image
The program in region.The storage device 31 may include high-speed random access memory, and may also include nonvolatile memory,
Such as one or more disk storage equipments, flash memory device or other non-volatile solid-state memory devices.In certain embodiments,
Storage device 31 can also include the memory far from one or more processors, such as attached via the network of communication network access
Add memory, wherein the communication network can be internet, one or more intranets, local area network (LAN), wide area network
(WLAN), storage area network (SAN) etc. or its is appropriately combined.It further include Memory Controller in storage device 31, the storage
Device controller can control access of the other assemblies of such as CPU and Peripheral Interface etc to memory.It is stored in storage device 31
In component software include operating system, communication module (or instruction set), contact/motion module (or instruction set), figure module
(or instruction set), haptic feedback module (or instruction set), text input module (or instruction set) and program (or instruction set).
The processing unit 32 is for executing described program to extract the face-image according to the extracting method of ROI region
In ROI region.
Here, the processing unit 32 includes processor, the processor operationally with memory and/or non-volatile
Memory coupling.More specifically, the instruction stored in memory and/or nonvolatile memory can be performed in terms of in processor
It calculates in equipment and executes operation, such as generate image data and/or image data is transferred to display circuit.In this way, processor can
It is patrolled including one or more general purpose microprocessors, one or more application specific processors (ASIC), one or more field-programmables
Collect array (FPGA) or any combination of them.
Processing unit 32 is also operationally coupled with network interface, will be calculated equipment and is communicatively coupled to either network.
For example, network interface, which can will calculate equipment, is connected to personal area network (PAN) (such as blueteeth network), local area network (LAN) (such as
802.11x Wi-Fi network), and/or wide area network (WAN).In addition, processing unit 32 is operatively coupled to power supply, the power supply
Electric power can be provided to the various parts calculated in equipment such as electronic console.In this way, power supply may include any suitable energy,
Such as rechargeable lighium polymer (Li-poly) battery and/or alternating current (AC) power adapter.
The processing unit 32 is also operatively coupled to the port I/O and input structure, which may make calculating to set
It is standby to be interacted with various other electronic equipments (such as dedicated for the instrument of skin detection or mobile terminal), the input knot
Structure aloows user to interact with equipment is calculated.Therefore, input structure may include button, keyboard, mouse, Trackpad
Deng.
The processing unit 32 is also operationally coupled with network interface, will be calculated equipment and is communicatively coupled to either net
Network.For example, network interface, which can will calculate equipment, is connected to personal area network (PAN) (such as blueteeth network), local area network (LAN)
(such as 802.11x Wi-Fi network), and/or wide area network (WAN) (injection 4G or LTE cellular network).
In this application, the processing unit 32 can be called based on the enabled instruction that input structure is inputted for generating
The program of ROI region template, and then when executing described program using the sample image in the storage device 31 as machine learning
Basis, to generate facial ROI region template.The process that the processing unit 32 executes described program can be as shown in Figure 2.
In step S210, the organic region in face-image obtained is identified.
Here, the face-image can be received by network, or the photographic device 33 by being arranged in equipment shoots and obtains.Institute
It states processing unit 32 and face-image is identified based on the feature (such as shape feature, hair feature, color characteristic) of each organ of face
Organic region.For example, based on color characteristic and facial symmetry characteristic identification eyes and eyebrow region.
In some embodiments, this step can be based on the device in face-image described in preset ROI region template extraction
Official region.Specifically, this step includes step S211 and S212.
In step S211, the face-image is carried out according to organ site by image segmentation based on ROI region template.Tool
Body, the processing unit 32 identifies the profile of face-image, further according to the area ROI described in preset ROI region template
The ROI region template contours are placed on acquired face-image by position of the domain template contours in average sample image
In.According still further to ROI region template by face-image piecemeal.
It should be noted that described image partitioned mode can divide face-image merely with ROI region template contours
Block label, and not necessarily carry out image block FIG pull handle.
In step S212, the organic region in the image-region after dividing is identified.Specifically, the processing unit 32
Using facial area represented by ROI region template, organic region is identified in face-image.For example, utilizing corresponding cheekbone area
The ROI region template in domain identifies the ocular and/or nose of monomer user in the image block adjacent with the region template
Region.
In step S220, preset ROI region template is adjusted to the edge of the organic region identified, is obtained described
The ROI region of face-image.
Specifically, the processing unit 32 according to identified organic region by the profile of ROI region template it is external or
On adjacent organs region, to obtain the ROI region in the face-image of corresponding monomer user.Wherein, the ROI region mould
The profile of plate can extend to a kind of organic region, or extend to adjacent a variety of organic regions.Here, the ROI region template
Extension include the outer of ROI region template contours extend to contraction.More specifically, the processing unit 32 by ROI region template to
Adjacent organic region extends, until meeting the critical condition with corresponding organ region.Wherein, the critical condition includes but unlimited
In: two regions share external contour line, two regions share external profile point, the shortest distance in two regions is less than preset value
Deng.
In certain embodiments, the processing unit 32 determine in ROI region template contours with adjacent organic region wheel
Wide spacing is less than the partial contour of pre-determined distance threshold value, and the partial contour in ROI region template is substituted for organic region
Corresponding part profile, to obtain the ROI region of corresponding monomer user.As shown in Fig. 3-A and 3-B, traversal formula cheekbone is detected
Spacing between ROI region template contours 11 and adjacent ocular profile 12, to determine cheekbone ROI region template contours 11
The spacing between ocular profile 12, by least part profile in cheekbone ROI region template 11 to pair of ocular 12
Partial contour is answered to extend (as shown in dotted line in Fig. 3-B), to obtain the cheekbone ROI region of corresponding monomer user.Wherein, extended
Partial contour be that pre-determined threshold and determination are less than based on spacing.
For another example, as shown in Fig. 3-A and 3-C, the characteristic point in ROI region template is moved to adjacent organic region profile
It is dynamic, until the characteristic point moved is less than preset value with the spacing of organ contours or connects.Here, the moving direction of characteristic point and
Moving distance can be by artificially disposing.Alternatively, the characteristic point is moved to along place normal direction to organic region profile.It is described
Face-image processing equipment is using the characteristic point area defined after movement as the ROI region of corresponding monomer user.
It should be noted that the moving direction of characteristic point described in this application is only for example, rather than limit.The feature
Point can also be moved according to the shortest distance with adjacent organs region contour along shortest distance direction.
Obtained ROI region profile can also be carried out round and smooth processing by the processing unit 32.The processing unit 32
Can emphasis detect determined by wrinkle condition, skin wrinkle, pore situation etc. in ROI region, and then pass through matching detection result
It determines the current situation of skin, such as determines the skin age, determines skin wrinkle lesion type.Alternatively, opposite according to testing result
Portion's image beautifies etc..
It should be noted that it should be appreciated by those skilled in the art that the above-mentioned cheekbone ROI region for monomer user really
Surely be only for example rather than the limitation to the application range.In fact, those skilled in the art are in combination with above-mentioned example and want true
The characteristics of fixed ROI region, carries out R & D design, and the technological means used based on technological frame described herein should all
The specific example being considered as within the scope of the application.
In some embodiments, the face-image processing equipment further include: photographic device 33.As shown in Figure 8.It is described
Photographic device 33 connects storage device 31 by processing unit 32.
The photographic device 33 is for absorbing face-image and being stored in the storage device 31.The photographic device 33
It can be a part being built in face-image processing equipment, such as photographic device 33 built-in in mobile terminal.Or it is described
Photographic device 33 is individual digital camera, and is connected with processing unit 32 by I/O subsystem.Wherein, the I/O subsystem can
It is packaged together with processing unit 32 comprising but be not limited to: the serial line interfaces such as USB.The photographic device 33 include lens group,
Imaging sensor, picture processing chip etc..Wherein, lens group is made of muti-piece eyeglass, will be taken the photograph using eyeglass to optical path change
The entity scene imaging taken is on an imaging sensor.Optical imagery is converted into electronic signal by the imaging sensor.With product
Class discrimination, imaging sensor product are broadly divided into three kinds of CCD, CMOS and CIS sensor.The imaging sensor is by gained
Image transfers to picture processing chip (ISP, Image Signal Processing) to carry out image rectification, noise removes, bad point is repaired
The image procossings such as benefit, color interpolation, white balance correction, exposure correction.
The face-image absorbed is supplied to processing unit 32 by the photographic device 33, is mentioned with being executed by it ROI region
Take method.
The face-image absorbed in order to ensure photographic device 33 phase in dimension scale with the ROI region template prestored
Match, the face-image processing equipment further include: shooting suggestion device 34, as shown in Figure 6.
Before the shooting suggestion device 34 is located at photographic device 33, for prompting tester to absorb in the photographic device 33
It puts on the head in direction.Here, the shooting suggestion device 34 can be a specific shooting prompt pattern, point, prompt are such as prompted
Line etc..In some embodiments, it includes the first support member for being used to support user's lower jaw and fixed photographic devices 33
Second support member, spacing between two support members is depending on ratio of the face-image in entire image.First
The height of support part part is related to the complete face-image that photographic device 33 can be shot.First support member is adjustable.Example
Such as, first support member includes an elevating lever and locking part, and a jaw support is equipped on the elevating lever.User can use
The height of the first support member of preceding adjustment, so that the photographic device 33 takes complete face-image.
It should be noted that herein described scheme can also be in use to beauty direction of taking pictures.For example, according to the side of the application
Case extracts the facial ROI region in image of taking pictures, and emphasis carries out beauty operation to extracted ROI region.
It should be noted that a part that the respective embodiments described above are also used as in U.S. face processing software be stored in as
In the storage medium of server-side.Meanwhile by the description of above each embodiment, those skilled in the art can be clearly
Solving some or all of the application can realize by software and in conjunction with required general hardware platform.Based on such reason
Solution, substantially the part that contributes to existing technology can body in the form of software products in other words for the technical solution of the application
Reveal and, which may include machine readable Jie of one or more for being stored thereon with machine-executable instruction
Matter, these instructions can make when being executed by one or more machines such as computer, computer network or other electronic equipments
It obtains the one or more machine and executes operation according to an embodiment of the present application.Machine readable media may include, but be not limited to, soft
Disk, CD, CD-ROM (compact-disc-read-only memory), magneto-optic disk, ROM (read-only memory), RAM (random access memory),
EPROM (Erasable Programmable Read Only Memory EPROM), EEPROM (electrically erasable programmable read-only memory), magnetic or optical card, sudden strain of a muscle
Deposit or suitable for store machine-executable instruction other kinds of medium/machine readable media.Wherein, the storage medium can
It may be alternatively located in third-party server positioned at terminal device (such as face detection equipment or intelligent terminal), be such as located at and certain is provided
Using in the server in store.With no restrictions to concrete application store at this, store, apple are applied using store, Huawei such as millet
Fruit is using store etc..
The application can be used in numerous general or special purpose computing system environments or configuration.Such as: personal computer, service
Device computer, handheld device or portable device, laptop device, multicomputer system, microprocessor-based system, top set
Box, programmable consumer-elcetronics devices, network PC, minicomputer, mainframe computer, including any of the above system or equipment
Distributed computing environment etc..
The application can describe in the general context of computer-executable instructions executed by a computer, such as program
Module.Generally, program module includes routines performing specific tasks or implementing specific abstract data types, programs, objects, group
Part, data structure etc..The application can also be practiced in a distributed computing environment, in these distributed computing environments, by
Task is executed by the connected remote processing devices of communication network.In a distributed computing environment, program module can be with
In the local and remote computer storage media including storage equipment.
The principles and effects of the application are only illustrated in above-described embodiment, not for limitation the application.It is any ripe
Know the personage of this technology all can without prejudice to spirit herein and under the scope of, carry out modifications and changes to above-described embodiment.Cause
This, those of ordinary skill in the art is complete without departing from spirit disclosed herein and institute under technical idea such as
At all equivalent modifications or change, should be covered by claims hereof.
Claims (21)
1. a kind of extracting method of ROI region characterized by comprising
Identify the organic region in face-image obtained；
Preset ROI region template is adjusted to the edge of the organic region identified, obtains the area ROI of the face-image
Domain.
2. the extracting method of ROI region according to claim 1, which is characterized in that the identification face figure obtained
The mode of organic region as in includes: based on the organic region in face-image described in preset ROI region template extraction.
3. the extracting method of ROI region according to claim 2, which is characterized in that described to be based on preset ROI region mould
The mode that plate extracts the organic region in the face-image includes:
The face-image is subjected to image segmentation based on ROI region template；
Identify the organic region in the image-region after dividing.
4. the extracting method of ROI region according to claim 1, which is characterized in that described by preset ROI region template
The mode adjusted to the edge of the organic region identified includes: to extend the ROI region template to adjacent organic region, directly
To the critical condition met with corresponding organ region.
5. the extracting method of ROI region according to claim 4, which is characterized in that it is described by ROI region template to adjacent
The mode that organic region extends includes: the characteristic point in the ROI region template to be shifted to adjacent organic region, and be based on movement
Characteristic point position afterwards sketches the contours ROI region.
6. the extracting method of ROI region according to claim 1, which is characterized in that the organic region includes: eye area
Domain.
7. a kind of ROI region template generation method characterized by comprising
Feature training is carried out using sample image to obtain average sample image；
The characteristic point for sketching the contours the ROI region is determined based on the average sample image, and based on selected characteristic point
Sketch the contours ROI region template.
8. ROI region template generation method according to claim 7, which is characterized in that the characteristic point includes: based on institute
It states in average sample image for sketching the contours the pixel of ROI region template contours.
9. a kind of extraction system of ROI region characterized by comprising
Identification module, for identification organic region in face-image obtained；
ROI extraction module obtains described for adjusting preset ROI region template to the edge of the organic region identified
The ROI region of face-image.
10. the extraction system of ROI region according to claim 9, which is characterized in that the identification module is specifically based on pre-
If ROI region template extraction described in organic region in face-image.
11. the extraction system of ROI region according to claim 10, which is characterized in that the identification module is for being based on
The face-image is carried out image segmentation by ROI region template；And the organ in the image-region after dividing for identification
Region.
12. the extraction system of ROI region according to claim 9, which is characterized in that the ROI extraction module is used for will
The ROI region template extends to adjacent organic region, until meeting the critical condition with corresponding organ region.
13. the extraction system of ROI region according to claim 12, which is characterized in that the ROI extraction module is specifically used
In the characteristic point in the ROI region template is shifted to adjacent organic region, and ROI is sketched the contours based on the characteristic point position after movement
Region.
14. the extraction system of ROI region according to claim 9, which is characterized in that the organic region includes: eye
Region.
15. a kind of ROI region template generating system characterized by comprising
Training module, for carrying out feature training using sample image to obtain average sample image；
ROI region template generation module, for determining the feature for sketching the contours the ROI region based on the average sample image
Point, and ROI region template is sketched the contours based on selected characteristic point.
16. ROI region template generating system according to claim 15, which is characterized in that the characteristic point includes: to be based on
For sketching the contours the pixel of ROI region template contours in the average sample image.
17. a kind of face-image processing equipment characterized by comprising
Storage device, for storing ROI region template, face-image and journey for extracting ROI region in the face-image
Sequence；
Processing unit, for executing described program to scheme according to the face as described in being extracted the method any in claim 1-6
ROI region as in.
18. face-image processing equipment according to claim 17, which is characterized in that further include: photographic device, for taking the photograph
It takes face-image and saves in the storage device.
19. face-image processing equipment according to claim 18, which is characterized in that further include: shooting suggestion device, position
Before photographic device, for prompting tester to put on the head in photographic device intake direction.
20. a kind of computer equipment characterized by comprising
Storage device, the program for storing sample image and for generating ROI region template；
Processing unit, for executing described program to generate ROI region template according to the method as described in claim 7 or 8.
21. a kind of storage medium, which is characterized in that be stored with acquired face-image and for carrying out ROI region extraction
Program；Wherein, described program is when being executed by processor, according to the step in any extracting method in such as claim 1-6
Suddenly the ROI region in the face-image is extracted.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710609651.4A CN109299714A (en) | 2017-07-25 | 2017-07-25 | ROI template generation method, ROI extracting method and system, equipment, medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710609651.4A CN109299714A (en) | 2017-07-25 | 2017-07-25 | ROI template generation method, ROI extracting method and system, equipment, medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109299714A true CN109299714A (en) | 2019-02-01 |
Family
ID=65167711
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710609651.4A Pending CN109299714A (en) | 2017-07-25 | 2017-07-25 | ROI template generation method, ROI extracting method and system, equipment, medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109299714A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110188711A (en) * | 2019-06-03 | 2019-08-30 | 北京字节跳动网络技术有限公司 | Method and apparatus for output information |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070183661A1 (en) * | 2006-02-07 | 2007-08-09 | El-Maleh Khaled H | Multi-mode region-of-interest video object segmentation |
| CN101593425A (en) * | 2009-05-06 | 2009-12-02 | 深圳市汉华安道科技有限责任公司 | A kind of fatigue driving monitoring method and system based on machine vision |
| CN102073776A (en) * | 2011-01-20 | 2011-05-25 | 西北大学 | Zonal statistic model based facial reconstruction method |
| US8908904B2 (en) * | 2011-12-28 | 2014-12-09 | Samsung Electrônica da Amazônia Ltda. | Method and system for make-up simulation on portable devices having digital cameras |
| CN104951743A (en) * | 2015-03-04 | 2015-09-30 | 苏州大学 | Active-shape-model-algorithm-based method for analyzing face expression |
| CN105205827A (en) * | 2015-10-16 | 2015-12-30 | 中科院成都信息技术股份有限公司 | Auxiliary feature point labeling method for statistical shape model |
| CN105825519A (en) * | 2016-02-05 | 2016-08-03 | 北京雅森科技发展有限公司 | Method and apparatus for processing medical image |
-
2017
- 2017-07-25 CN CN201710609651.4A patent/CN109299714A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070183661A1 (en) * | 2006-02-07 | 2007-08-09 | El-Maleh Khaled H | Multi-mode region-of-interest video object segmentation |
| CN101593425A (en) * | 2009-05-06 | 2009-12-02 | 深圳市汉华安道科技有限责任公司 | A kind of fatigue driving monitoring method and system based on machine vision |
| CN102073776A (en) * | 2011-01-20 | 2011-05-25 | 西北大学 | Zonal statistic model based facial reconstruction method |
| US8908904B2 (en) * | 2011-12-28 | 2014-12-09 | Samsung Electrônica da Amazônia Ltda. | Method and system for make-up simulation on portable devices having digital cameras |
| CN104951743A (en) * | 2015-03-04 | 2015-09-30 | 苏州大学 | Active-shape-model-algorithm-based method for analyzing face expression |
| CN105205827A (en) * | 2015-10-16 | 2015-12-30 | 中科院成都信息技术股份有限公司 | Auxiliary feature point labeling method for statistical shape model |
| CN105825519A (en) * | 2016-02-05 | 2016-08-03 | 北京雅森科技发展有限公司 | Method and apparatus for processing medical image |
Non-Patent Citations (4)
| Title |
|---|
| WOO H J等: ""Eye and Mouth Images Based Facial Expressions Recognition Using PCA and Template Matching"", 《THE JOURNAL OF THE KOREA CONTENTS ASSOCIATION》 * |
| 王元全等: ""基于先验知识改进Snake模型的脸部特征分割"", 《计算机辅助设计与图形学学报》 * |
| 范小九: ""基于人脸特征定位和建模理论的视频编码关键技术研究"", 《中国博士学位论文全文数据库·信息科技辑》 * |
| 赵瑞: ""基于肤色和模板的人脸检测算法研究"", 《中国优秀硕士学位论文全文数据库·信息科技辑》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110188711A (en) * | 2019-06-03 | 2019-08-30 | 北京字节跳动网络技术有限公司 | Method and apparatus for output information |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106469302B (en) | A kind of face skin quality detection method based on artificial neural network | |
| Jain et al. | Three-dimensional CNN-inspired deep learning architecture for Yoga pose recognition in the real-world environment | |
| CN103443804B (en) | Method of facial landmark detection | |
| CN106796449A (en) | Eye-controlling focus method and device | |
| CN106056064B (en) | A kind of face identification method and face identification device | |
| CN103810490B (en) | A kind of method and apparatus for the attribute for determining facial image | |
| Alnajar et al. | Calibration-free gaze estimation using human gaze patterns | |
| WO2020103417A1 (en) | Bmi evaluation method and device, and computer readable storage medium | |
| CN108875485A (en) | A kind of base map input method, apparatus and system | |
| CN108875633A (en) | Expression detection and expression driving method, device and system and storage medium | |
| CN109242789A (en) | Image processing method, image processing apparatus and storage medium | |
| Pal et al. | Severity grading of psoriatic plaques using deep CNN based multi-task learning | |
| CN104143086A (en) | Application technology of portrait comparison to mobile terminal operating system | |
| WO2017161734A1 (en) | Correction of human body movements via television and motion-sensing accessory and system | |
| CN107292299B (en) | Side face recognition methods based on kernel specification correlation analysis | |
| CN109299632A (en) | Skin detecting method, system, equipment and storage medium | |
| US20160203358A1 (en) | Systems and methods for using curvatures to analyze facial and body features | |
| JP2014194617A (en) | Visual line direction estimating device, visual line direction estimating method, and visual line direction estimating program | |
| CN109977764A (en) | Vivo identification method, device, terminal and storage medium based on plane monitoring-network | |
| CN114078278A (en) | Method and device for positioning fixation point, electronic equipment and storage medium | |
| Wu et al. | Appearance-based gaze block estimation via CNN classification | |
| CN109145716A (en) | Boarding gate verifying bench based on face recognition | |
| Lanz et al. | Automated classification of therapeutic face exercises using the Kinect | |
| WO2022088290A1 (en) | Motion assessment method, apparatus and system, and storage medium | |
| CN109299714A (en) | ROI template generation method, ROI extracting method and system, equipment, medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190201 |