CN109086688A - Face wrinkles' detection method, device, computer equipment and storage medium - Google Patents
Face wrinkles' detection method, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN109086688A CN109086688A CN201810768838.3A CN201810768838A CN109086688A CN 109086688 A CN109086688 A CN 109086688A CN 201810768838 A CN201810768838 A CN 201810768838A CN 109086688 A CN109086688 A CN 109086688A
- Authority
- CN
- China
- Prior art keywords
- wrinkle
- image
- face
- crumple zone
- pixel
- 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
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
- G06V40/165—Detection; Localisation; Normalisation using facial parts and geometric relationships
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06V10/30—Noise filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
- G06V40/171—Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships
Abstract
This application involves a kind of face wrinkles' detection method, device, computer equipment and storage mediums.The described method includes: obtaining face image, and mark the facial feature points of the face image；According to the facial feature points of the face image, the crumple zone of the face image is divided, and determines the wrinkle type of the crumple zone；According to the wrinkle type of the crumple zone, the area image in the crumple zone is extracted；Obtain the wrinkle characteristic component of the area image；Wrinkle detection is carried out according to the wrinkle characteristic component of the area image.According to the application, even if there are irregular face features there are more subtle wrinkle or some region of face in some region of face, accurate detection can also be carried out, improves the accuracy of wrinkle detection.
Description
Technical field
This application involves technical field of image processing, more particularly to a kind of face wrinkles' detection method, device, computer
Equipment and storage medium.
Background technique
Currently, face wrinkles' detection technique is increasingly used in multiple fields.For example, researching and developing field in cosmetics
In, it needs to design cosmetic product or the corresponding cosmetics of recommended user for different face wrinkles' features.In another example
Photo beautifies in special efficacy field, needs to carry out different degrees of beautification special efficacy for different face wrinkles.In another example in face
In identification, need according to face wrinkles as user characteristics to verify user identity.
Current relatively conventional face wrinkles' detection technique is three dimensions that user face is acquired by 3-D measuring apparatus
According to, modeled according to three-dimensional data, for facial model carry out wrinkle detection.
However, face wrinkles' detection method based on three-dimensional data, the problem of due to equipment precision, it is difficult to capture in face
More subtle wrinkle, and face feature similar with wrinkle irregular for face, then can be mistaken for wrinkle.
Therefore, current face wrinkles' detection method has wrinkle detection inaccuracy.
Summary of the invention
Based on this, it is necessary in view of the above technical problems, provide a kind of face's wrinkle for being able to ascend wrinkle detection accuracy
Marks detection method, device, computer equipment and storage medium.
A kind of face wrinkles' detection method, comprising:
Face image is obtained, and marks the facial feature points of the face image；
According to the facial feature points of the face image, the crumple zone of the face image is divided, and determines the wrinkle
The wrinkle type in line region；
According to the wrinkle type of the crumple zone, the area image in the crumple zone is extracted；
Obtain the wrinkle characteristic component of the area image；
Wrinkle detection is carried out according to the wrinkle characteristic component of the area image.
The wrinkle type according to the crumple zone in one of the embodiments, is extracted in the crumple zone
Area image, comprising:
Obtain the corresponding filtering parameter of wrinkle type of the crumple zone；
Using the filtering parameter, the image in the crumple zone is filtered, filtering image is obtained；
Processing is sharpened to the image in the crumple zone using the filtering image, obtains the area image.
The wrinkle characteristic component for obtaining the area image in one of the embodiments, comprising:
Obtain the pixel of the area image；
Calculate the image characteristic matrix of the pixel；
The maximum eigenvalue of described image eigenmatrix is calculated, and, calculate the minimal characteristic of described image eigenmatrix
Value；
According to the maximum eigenvalue and the minimal eigenvalue, the wrinkle similarity of the pixel is predicted；
When the pixel wrinkle similarity be greater than preset threshold, determine the pixel be wrinkle pixel；
According to the connected domain of the wrinkle pixel, connect the wrinkle pixel, using the wrinkle pixel of connection as
The wrinkle characteristic component.
The crumple zone has corresponding multiple filtering parameters, the filtering parameter tool in one of the embodiments,
There is corresponding wrinkle characteristic component, it is described that wrinkle detection is carried out according to the wrinkle characteristic component of the area image, comprising:
Merge the corresponding wrinkle characteristic component of the multiple filtering parameter, obtains initial wrinkle component set；
The initial wrinkle component set is denoised, targeted wrinkle component set is obtained；
According to the targeted wrinkle component set, wrinkle quantity, wrinkle length and wrinkle distribution characteristics are detected.
In one of the embodiments, further include:
Obtain sample face image and verification face image；
Using the sample face image and the verification face image, machine training is carried out to initial markers model, is obtained
To target label model；
The facial feature points of the label face image, comprising:
The face image is marked by the target label model, obtains the face feature of the face image
Point.
The filtering parameter is used described in one of the embodiments, the image in the crumple zone is carried out
Filtering processing, before obtaining filtering image, further includes:
To the image in the crumple zone, according to horizontally and vertically zooming in and out.
The filtering parameter includes that filtering direction, filtering sampling interval and filtering sampling are big in one of the embodiments,
It is small.
A kind of face wrinkles' detection device, comprising:
Mark module for obtaining face image, and marks the facial feature points of the face image；
Division module divides the crumple zone of the face image for the facial feature points according to the face image,
And determine the wrinkle type of the crumple zone；
Image zooming-out module extracts the region in the crumple zone for the wrinkle type according to the crumple zone
Image；
Component obtains module, for obtaining the wrinkle characteristic component of the area image；
Detection module, for carrying out wrinkle detection according to the wrinkle characteristic component of the area image.
A kind of computer equipment, including memory and processor, the memory are stored with computer program, the processing
Device performs the steps of when executing the computer program
Face image is obtained, and marks the facial feature points of the face image；
According to the facial feature points of the face image, the crumple zone of the face image is divided, and determines the wrinkle
The wrinkle type in line region；
According to the wrinkle type of the crumple zone, the area image in the crumple zone is extracted；
Obtain the wrinkle characteristic component of the area image；
Wrinkle detection is carried out according to the wrinkle characteristic component of the area image.
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
It is performed the steps of when row
Face image is obtained, and marks the facial feature points of the face image；
According to the facial feature points of the face image, the crumple zone of the face image is divided, and determines the wrinkle
The wrinkle type in line region；
According to the wrinkle type of the crumple zone, the area image in the crumple zone is extracted；
Obtain the wrinkle characteristic component of the area image；
Wrinkle detection is carried out according to the wrinkle characteristic component of the area image.
Above-mentioned face wrinkles' detection method, device, computer equipment and storage medium, by the face for marking face image
Characteristic point divides the crumple zone of face image according to the facial feature points of face image, and determines the wrinkle class of crumple zone
Type, it is thus possible to extract the area image in different crumple zones, and obtain based on area image according to different wrinkle types
Wrinkle characteristic component is taken, according to the wrinkle characteristic component of each region image, carries out wrinkle detection.To, though face some
There are more subtle wrinkle or some region of face, there are irregular face features in region, can also carry out standard
True detection improves the accuracy of wrinkle detection.
Detailed description of the invention
Fig. 1 is a kind of flow diagram of face wrinkles' detection method of the present embodiment；
Fig. 2 is a kind of schematic diagram in face image label facial feature points of the present embodiment；
Fig. 3 is the schematic diagram that a kind of crumple zone of the present embodiment divides；
Fig. 4 is a kind of applied environment figure of face wrinkles' detection method of the present embodiment；
Fig. 5 is a kind of flow chart of acquisition wrinkle component set of the present embodiment；
Fig. 6 is a kind of structural block diagram of face wrinkles' detection device of the present embodiment；
Fig. 7 is a kind of internal structure chart of computer equipment of the present embodiment.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only used to explain the application, not
For limiting the application.
In one embodiment, as shown in Figure 1, providing a kind of face wrinkles' detection method, this method may include with
Lower step:
Step S101 obtains face image, and marks the facial feature points of the face image.
Wherein, face image can be for for face acquired image.
Wherein, facial feature points can be the pixel of reflection human face five-sense-organ contour feature in face image.
In the specific implementation, image can be acquired for face, face image is obtained.Then, by markup model to face
The facial feature points of image are identified, and the facial feature points identified are marked.
Fig. 2 is a kind of schematic diagram in face image label facial feature points of the present embodiment.It can be seen that being directed to face
Portion as eyebrow, eyelid, nose, lip, shape of face outer profile, 94 characteristic points can be marked, as facial feature points.
Step S102 divides the crumple zone of the face image according to the facial feature points of the face image, and really
The wrinkle type of the fixed crumple zone.
Wherein, crumple zone can be in face image comprising regional area wrinkly.
Wherein, wrinkle type can be to the classification of the wrinkle with different shape feature in face's different zones.
In the specific implementation, area can be carried out to face image according to coordinate information of the facial feature points in face image
Domain divides, and obtains the crumple zone of several different wrinkle types.
For example, after determining 94 facial feature points, will wherein several facial feature points be formed by with reference to Fig. 2
Region, the crumple zone as some wrinkle type.
Specifically, according to totally 13 face features of facial feature points 62,66,60,64,58,65,61,67,63,7,1,0,6
Multiple facial feature points can be formed by region by point, the forehead line region as face image.
According to totally 7 facial feature points of facial feature points 0,1,2,3,64,58,65, the area Chuan Ziwen of face image is determined
Domain.
According to totally 6 facial feature points of facial feature points 2,3,18,19,30,31, the eye interband area of face image is determined
Domain.
According to totally 6 facial feature points of facial feature points 20,26,28,24,32,86, determine that the line now of face image is right
Region.
According to totally 6 facial feature points of facial feature points 21,27,29,25,33,87, determine that the line now of face image is left
Region.
According to totally 6 facial feature points of facial feature points 34,36,44,78,84,86, determine that the decree line of face image is right
Region.
According to totally 6 facial feature points of facial feature points 35,37,45,79,85,87, determine that the decree line of face image is left
Region.
According to totally 10 facial feature points of facial feature points 44,48,46,41,47,49,45,38,40,39, face is determined
Line region on the lip of image.
Fig. 3 is the schematic diagram that a kind of crumple zone of the present embodiment divides.It can be seen that for the multiple of face image
Face image is divided into forehead line region, the river region Zi Wen, eye interband area by 94 facial feature points by facial feature points
8 wrinkles such as line region on the right region of domain, now line, the now left region of line, the right region of decree line, the left region of decree line and lip
Region.
Certainly, above-mentioned division mode is given for example only explanation, and those skilled in the art can mark according to actual needs
Any number of facial feature points, and any number of crumple zone is divided according to facial feature points, the present embodiment is to face spy
The quantity of sign point and crumple zone is with no restriction.
Step S103 extracts the area image in the crumple zone according to the wrinkle type of the crumple zone.
Wherein, area image can be the topography in face image in crumple zone.
In the specific implementation, the corresponding wrinkle type of crumple zone can be directed to, image is extracted in crumple zone, as area
Area image.Specific extracting mode can obtain filtering parameter corresponding to wrinkle type first, corresponding according to wrinkle type
Filtering parameter is filtered the prime area image in crumple zone, obtains filtering image, then uses filtering image
Processing is sharpened to prime area image, obtains the area image by special effect processing.
Since the wrinkle of different wrinkle types is distinct in features such as shape, size, quantity, in order to accurately
Wrinkle feature is extracted, different filtering parameters can be used for different wrinkle types, the spies such as be filtered, sharpen to image
Effect processing, thus, in subsequent processing, wrinkle detection can be carried out based on the image that can accurately reflect wrinkle feature.
Step S104 obtains the wrinkle characteristic component of the area image.
Wherein, wrinkle characteristic component can be reflection wrinkle feature, multiple pixels interconnected in image.
In the specific implementation, wrinkle characteristic component can be got according to area image.For example, administrative division map can be extracted first
Multiple pixels as in, then calculate the image characteristic matrix of pixel, and calculate the maximum feature of the image characteristic matrix
Value and minimal eigenvalue can predict the wrinkle similarity of each pixel, will wrinkle according to maximum eigenvalue and minimal eigenvalue
Line similarity is greater than the pixel of preset threshold, is determined as wrinkle pixel, to obtain multiple wrinkle pixels, calculates multiple
The connected domain of wrinkle pixel, by the wrinkle pixel of connection, as wrinkle characteristic component.
Certainly, those skilled in the art can obtain the wrinkle characteristic component of each region image using other modes, this
Embodiment to specific wrinkle characteristic component acquisition modes with no restriction.
Step S105 carries out wrinkle detection according to the wrinkle characteristic component of the area image.
In the specific implementation, the wrinkle characteristic component of the area image of each crumple zone can be merged, wrinkled
Line component set can count the quantity of wrinkle, the length of wrinkle, distribution situation of wrinkle etc. according to the wrinkle component set.
It should be noted that face wrinkles' detection method provided by the present application, can be applied to application ring as shown in Figure 4
In border.Wherein, Image Acquisition terminal 402 is communicated with detection service device 404 by network by network.Wherein, image is adopted
Collection terminal 402 can be, but not limited to be various personal computers, laptop, smart phone, tablet computer and it is portable can
Wearable device, detection service device 404 can be with the server clusters of the either multiple server compositions of independent server come real
It is existing.Image Acquisition terminal 402 can acquire face image for face, and face image is sent to detection service device 404, by
Detection service device 404 executes above-mentioned face wrinkles' detection method.It certainly, in practical applications, can also be by Image Acquisition end
End 402 executes above-mentioned face wrinkles' detection method.
According to face wrinkles' detection method provided in this embodiment, by marking the facial feature points of face image, according to
The facial feature points of face image divide the crumple zone of face image, and determine the wrinkle type of crumple zone, thus, it can
To extract the area image in different crumple zones according to different wrinkle types, and wrinkle feature is obtained based on area image
Component carries out wrinkle detection according to the wrinkle characteristic component of each region image.To even if existing in some region of face
Having more subtle some region of wrinkle or face, there are irregular face features, can also carry out accurate detection, mention
The accuracy of wrinkle detection is risen.
One embodiment wherein, the step S103, comprising:
Obtain the corresponding filtering parameter of wrinkle type of the crumple zone；Using the filtering parameter, to the wrinkle
Image in region is filtered, and obtains filtering image；Using the filtering image to the image in the crumple zone
It is sharpened processing, obtains the area image.
Wherein, filtering parameter may include filtering direction, filtering sampling interval and filtering sampling size etc. for determining filter
The parameter of wave mode.Crumple zone has different wrinkle types, then different filtering parameters is corresponded to, for example, for eye interband
Region, now line region and crows feet area, filtering direction are horizontal direction, and filtering sampling size is 50*50 pixel, and filtering is adopted
10 pixels are divided between sample.
In the specific implementation, corresponding filtering direction, filtering sampling can be determined first according to the wrinkle type of crumple zone
Then the filtering parameters such as interval, filtering sampling size extract the image in the crumple zone, as prime area image, use
Filtering parameter is filtered the prime area image, obtains filtering image.It is then possible to using filtering image to initial
Area image is sharpened processing, the image with feature sharpening effect is obtained, as area image.Wherein it is possible to design filter
Wave device is rectangle frame filter, and filtering mode is mean filter.
For example, the crumple zone of forehead line type, due to the wrinkle direction of the wrinkle type, predominantly horizontal direction, because
This, the filtering sampling size for designing filter is 8*40 pixel, and filtering direction is vertical direction, is filtered by the filter
Processing.
In another example the crumple zone of river word line type, due to the wrinkle direction of the wrinkle type, predominantly vertical direction,
Therefore, the filtering sampling size for designing filter is 15*3 pixel, and filtering direction is horizontal direction, is filtered by the filter
Wave processing.
In another example the crumple zone of eye interband type, due to the wrinkle direction of the wrinkle type, predominantly horizontal direction,
Therefore, the filtering sampling size for designing filter is 6*30 pixel, and filtering direction is vertical direction, is filtered by the filter
Wave processing.
In another example the crumple zone of the right type of line now, due to the wrinkle direction of the wrinkle type, predominantly 120 ° of directions
With 150 ° of directions, therefore, the filtering sampling size for designing filter is 3*30 pixel, and filtering direction is 30 ° and 60 ° of directions, is led to
The filter is crossed to be filtered.
In another example the crumple zone of the left type of line now, due to the wrinkle direction of the wrinkle type, predominantly 30 ° of directions
With 60 ° of directions, therefore, the filtering sampling size for designing filter is 3*30 pixel, and filtering direction is 120 ° and 150 ° of directions, is led to
The filter is crossed to be filtered.
In another example the crumple zone of the left type of decree line, due to the wrinkle direction of the wrinkle type, predominantly 120 ° of sides
To therefore, the filtering sampling size for designing filter is 30*3 pixel, and filtering direction is 60 °, is filtered by the filter
Processing.
In another example the crumple zone of the right type of decree line, due to the wrinkle direction of the wrinkle type, predominantly 60 ° of directions,
Therefore, the filtering sampling size for designing filter is 30*3 pixel, and filtering direction is 120 °, is filtered place by the filter
Reason.
In another example on lip the right type of line crumple zone, due to the wrinkle direction of the wrinkle type, predominantly 70 ° of directions
With 110 ° of directions, therefore, the filtering sampling size for designing filter is 20*3 pixel, and filtering direction is 110 ° of directions and 70 ° of sides
To being filtered by the filter.
One embodiment wherein, the step S104, comprising:
Obtain the pixel of the area image；Calculate the image characteristic matrix of the pixel；It is special to calculate described image
The maximum eigenvalue of matrix is levied, and, calculate the minimal eigenvalue of described image eigenmatrix；According to the maximum eigenvalue and
The minimal eigenvalue predicts the wrinkle similarity of the pixel；When the wrinkle similarity of the pixel is greater than default threshold
Value determines that the pixel is wrinkle pixel；According to the connected domain of the wrinkle pixel, the wrinkle pixel is connected,
Using the wrinkle pixel of connection as the wrinkle characteristic component.
In the specific implementation, after obtaining area image image characteristic matrix can be carried out to each region image respectively
It calculates.Specifically, multiple pixels can be extracted in area image first, for each pixel, calculate its characteristics of image
Matrix, such as Hessian matrix (Hessian Matrix, Hessian matrix).
Wherein, Hessian matrix may include three submatrixs, the sub- square for respectively asking second order to lead horizontal direction
Battle array Hxx, the submatrix H that asks second order to lead vertical directionyy, and, first to horizontal direction derivation, again to vertical direction derivation
Obtained submatrix Hxy, and, the submatrix H first obtained to vertical direction derivation, again to horizontal direction derivationyx=Hxy.By
This, it is as follows to obtain Hessian matrix:
Then, the maximum eigenvalue and minimal eigenvalue of calculating matrix H, respectively λ1、λ2, according to λ1、λ2Calculate pixel
Wrinkle similarity value, calculation formula is as follows:
Wherein, constant C is used to control the size of wrinkle similarity, and constant beta is used to control the shape of wrinkle.
By the wrinkle similarity f of pixel and preset threshold f0It is compared, when f value is greater than or equal to threshold value f0When, then recognize
It is wrinkle pixel for the pixel；Less than threshold value f0When, then it is assumed that the pixel is non-creped pixel.In practical application,
The pixel value of wrinkle pixel can be set to 1, the pixel value of non-creped pixel is set as 0.According to the pixel of pixel
Value, can distinguish wrinkle pixel and non-creped pixel.
Finally, the connected domain of detection wrinkle pixel, determining has multiple wrinkle pixels of syntople, and it is more to connect this
A wrinkle pixel, multiple wrinkle pixels of connection then form the connected domain of multiple wrinkle pixels, as wrinkle feature point
Amount.
One embodiment wherein, the crumple zone have corresponding multiple filtering parameters, and the filtering parameter has
Corresponding wrinkle characteristic component, the step S105, comprising:
Merge the corresponding wrinkle characteristic component of the multiple filtering parameter, obtains initial wrinkle component set；To described first
Beginning wrinkle component set is denoised, and targeted wrinkle component set is obtained；According to the targeted wrinkle component set, wrinkle is detected
Quantity, wrinkle length and wrinkle distribution characteristics.
In the specific implementation, thus generating different filters due to the filtering that can carry out multiple directions for a crumple zone
The area image in wave direction, it is available to arrive corresponding wrinkle characteristic component for each area image.It therefore, can will be same
Multiple wrinkle characteristic components of one area image merge, and obtain initial wrinkle component set W0。
Due to initial wrinkle component set W0There may be many noises, and hence it is also possible to initial wrinkle component set W0
Denoising is carried out, targeted wrinkle component set W is obtained, and be based on targeted wrinkle component set W, carries out wrinkle quantity, wrinkle
The detection processing of length, wrinkle distribution characteristics.
Fig. 5 is a kind of flow chart of acquisition wrinkle component set of the present embodiment.It wrinkles in figure as it can be seen that being divided to face image
After line region, for the prime area image in a crumple zone, filtering, the Edge contrast of multiple directions are carried out respectively,
Obtain the area image D of corresponding different directionsn, image characteristic matrix H is calculated for the pixel of each area imagen, and according to
Image characteristic matrix determines wrinkle characteristic component, is finally synthesizing the wrinkle characteristic component W of the area image of all directionsn, go forward side by side
Row denoising obtains targeted wrinkle component set as illustrated in the drawing.
One embodiment wherein, further includes:
Obtain sample face image and verification face image；Schemed using the sample face image and the verification face
Picture carries out machine training to initial markers model, obtains target label model；
The facial feature points of the label face image, comprising:
The face image is marked by the target label model, obtains the face feature of the face image
Point.
In the specific implementation, several frontal one images can be obtained first before carrying out face wrinkles' detection, middle part
Divide for the machine training to model, is partially used for the verification of model.Face image is marked into face feature by marker software
Result is stored in .xml file by point.It can be using face landmark detection algorithm (face in dlib algorithms library
Detect labeling algorithm in portion), dlib is the library open source c++ comprising many machine learning algorithms, wherein face landmark
Detection algorithm is a kind of machine learning training algorithm based on face mark point.
After the completion of markup model training, the markup model of facial feature points is obtained.Using the markup model to face image
It is marked, the accuracy rate of label can be promoted, in subsequent processing, to divide wrinkle area based on accurate facial feature points
Domain, to further promote the accuracy of wrinkle detection.
One embodiment wherein uses the filtering parameter described, filters to the image in the crumple zone
Wave processing, before obtaining filtering image, further includes:
To the image in the crumple zone, according to horizontally and vertically zooming in and out.
In the specific implementation, the scaling of multiple directions can be carried out to the prime area image in crumple zone.For example, according to
It horizontally and vertically zooms in and out, obtains the image of one half-size scale of prime area image, further to promote wrinkle inspection
The accuracy of survey.
In practical applications, for the river region Zi Wen and forehead region, be easy among wrinkle to disconnect, and part wrinkle
The case where width is larger, be easy to cause wrinkle missing inspection.Therefore, can to the prime area image in the crumple zone of the type,
A certain proportion of scaling is carried out, to avoid wrinkle missing inspection.
It should be understood that although each step in the flow chart of Fig. 1 is successively shown according to the instruction of arrow, this
A little steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly state otherwise herein, these steps
It executes there is no the limitation of stringent sequence, these steps can execute in other order.Moreover, at least part in Fig. 1
Step may include that perhaps these sub-steps of multiple stages or stage are executed in synchronization to multiple sub-steps
It completes, but can execute at different times, the execution sequence in these sub-steps or stage, which is also not necessarily, successively to be carried out,
But it can be executed in turn or alternately at least part of the sub-step or stage of other steps or other steps.
In one embodiment, as shown in fig. 6, providing a kind of face wrinkles' detection device, comprising: mark module 601,
Division module 602, image zooming-out module 603, component obtain module 604 and detection module 605, in which:
Mark module 601 for obtaining face image, and marks the facial feature points of the face image；
Division module 602 divides the wrinkle area of the face image for the facial feature points according to the face image
Domain, and determine the wrinkle type of the crumple zone；
Image zooming-out module 603 extracts the area in the crumple zone for the wrinkle type according to the crumple zone
Area image；
Component obtains module 604, for obtaining the wrinkle characteristic component of the area image；
Detection module 605, for carrying out wrinkle detection according to the wrinkle characteristic component of the area image.
One embodiment wherein, described image extraction module 603, comprising:
Filtering parameter acquisition submodule, for obtaining the corresponding filtering parameter of wrinkle type of the crumple zone；
Submodule is filtered, for using the filtering parameter, the image in the crumple zone is filtered, is obtained
To filtering image；
Submodule is sharpened, for being sharpened processing to the image in the crumple zone using the filtering image, is obtained
To the area image.
One embodiment wherein, the component obtain module 604, comprising:
Pixel acquisition submodule, for obtaining the pixel of the area image；
Matrix computational submodule, for calculating the image characteristic matrix of the pixel；
Characteristic value computational submodule, for calculating the maximum eigenvalue of described image eigenmatrix, and, calculate the figure
As the minimal eigenvalue of eigenmatrix；
Similarity prediction module, for predicting the pixel according to the maximum eigenvalue and the minimal eigenvalue
Wrinkle similarity；
Wrinkle pixel determines submodule, is greater than preset threshold for the wrinkle similarity when the pixel, determines institute
Stating pixel is wrinkle pixel；
Component determines submodule, for the connected domain according to the wrinkle pixel, connects the wrinkle pixel, will even
The wrinkle pixel connect is as the wrinkle characteristic component.
One embodiment wherein, the crumple zone have corresponding multiple filtering parameters, and the filtering parameter has
Corresponding wrinkle characteristic component, the detection module 605, comprising:
Merge submodule, for merging the corresponding wrinkle characteristic component of the multiple filtering parameter, obtains initial wrinkle point
Duration set；
It denoises submodule and obtains targeted wrinkle component set for denoising to the initial wrinkle component set；
Wrinkle detection sub-module, for detecting wrinkle quantity, wrinkle length and wrinkle according to the targeted wrinkle component set
Line distribution characteristics.
One embodiment wherein, further includes:
Sample acquisition module, for obtaining sample face image and verification face image；
Training module, for use the sample face image and the verification face image, to initial markers model into
The training of row machine, obtains target label model；
The mark module 601, comprising:
Signature submodule obtains institute for the face image to be marked by the target label model
State the facial feature points of face image.
One embodiment wherein, further includes:
Zoom module, for the image in the crumple zone, according to horizontally and vertically zooming in and out.
One embodiment wherein, the filtering parameter include filtering direction, filtering sampling interval and filtering sampling size.
Specific about face wrinkles' detection device limits the limit that may refer to above for face wrinkles' detection method
Fixed, details are not described herein.Modules in above-mentioned face wrinkles' detection device can fully or partially through software, hardware and its
Combination is to realize.Above-mentioned each module can be embedded in the form of hardware or independently of in the processor in computer equipment, can also be with
It is stored in the memory in computer equipment in a software form, in order to which processor calls the above modules of execution corresponding
Operation.
In one embodiment, a kind of computer equipment is provided, which can be server, internal junction
Composition can be as shown in Figure 7.The computer equipment include by system bus connect processor, memory, network interface and
Database.Wherein, the processor of the computer equipment is for providing calculating and control ability.The memory packet of the computer equipment
Include non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system, computer program and data
Library.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The database of machine equipment is used to store the related data of wrinkle detection.The network interface of the computer equipment is used for and external end
End passes through network connection communication.To realize a kind of face wrinkles' detection method when the computer program is executed by processor.
It will be understood by those skilled in the art that structure shown in Fig. 7, only part relevant to application scheme is tied
The block diagram of structure does not constitute the restriction for the computer equipment being applied thereon to application scheme, specific computer equipment
It may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
In one embodiment, a kind of computer equipment, including memory and processor are provided, is stored in memory
Computer program, the processor perform the steps of when executing computer program
Face image is obtained, and marks the facial feature points of the face image；
According to the facial feature points of the face image, the crumple zone of the face image is divided, and determines the wrinkle
The wrinkle type in line region；
According to the wrinkle type of the crumple zone, the area image in the crumple zone is extracted；
Obtain the wrinkle characteristic component of the area image；
Wrinkle detection is carried out according to the wrinkle characteristic component of the area image.
In one embodiment, it is also performed the steps of when processor executes computer program
Obtain the corresponding filtering parameter of wrinkle type of the crumple zone；
Using the filtering parameter, the image in the crumple zone is filtered, filtering image is obtained；
Processing is sharpened to the image in the crumple zone using the filtering image, obtains the area image.
In one embodiment, it is also performed the steps of when processor executes computer program
Obtain the pixel of the area image；
Calculate the image characteristic matrix of the pixel；
The maximum eigenvalue of described image eigenmatrix is calculated, and, calculate the minimal characteristic of described image eigenmatrix
Value；
According to the maximum eigenvalue and the minimal eigenvalue, the wrinkle similarity of the pixel is predicted；
When the pixel wrinkle similarity be greater than preset threshold, determine the pixel be wrinkle pixel；
According to the connected domain of the wrinkle pixel, connect the wrinkle pixel, using the wrinkle pixel of connection as
The wrinkle characteristic component.
In one embodiment, it is also performed the steps of when processor executes computer program
Merge the corresponding wrinkle characteristic component of the multiple filtering parameter, obtains initial wrinkle component set；
The initial wrinkle component set is denoised, targeted wrinkle component set is obtained；
According to the targeted wrinkle component set, wrinkle quantity, wrinkle length and wrinkle distribution characteristics are detected.
In one embodiment, it is also performed the steps of when processor executes computer program
Obtain sample face image and verification face image；
Using the sample face image and the verification face image, machine training is carried out to initial markers model, is obtained
To target label model；
The facial feature points of the label face image, comprising:
The face image is marked by the target label model, obtains the face feature of the face image
Point.
In one embodiment, it is also performed the steps of when processor executes computer program
To the image in the crumple zone, according to horizontally and vertically zooming in and out.
In one embodiment, the filtering parameter includes filtering direction, filtering sampling interval and filtering sampling size.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program performs the steps of when being executed by processor
Face image is obtained, and marks the facial feature points of the face image；
According to the facial feature points of the face image, the crumple zone of the face image is divided, and determines the wrinkle
The wrinkle type in line region；
According to the wrinkle type of the crumple zone, the area image in the crumple zone is extracted；
Obtain the wrinkle characteristic component of the area image；
Wrinkle detection is carried out according to the wrinkle characteristic component of the area image.
In one embodiment, it is also performed the steps of when computer program is executed by processor
Obtain the corresponding filtering parameter of wrinkle type of the crumple zone；
Using the filtering parameter, the image in the crumple zone is filtered, filtering image is obtained；
Processing is sharpened to the image in the crumple zone using the filtering image, obtains the area image.
In one embodiment, it is also performed the steps of when computer program is executed by processor
Obtain the pixel of the area image；
Calculate the image characteristic matrix of the pixel；
The maximum eigenvalue of described image eigenmatrix is calculated, and, calculate the minimal characteristic of described image eigenmatrix
Value；
According to the maximum eigenvalue and the minimal eigenvalue, the wrinkle similarity of the pixel is predicted；
When the pixel wrinkle similarity be greater than preset threshold, determine the pixel be wrinkle pixel；
According to the connected domain of the wrinkle pixel, connect the wrinkle pixel, using the wrinkle pixel of connection as
The wrinkle characteristic component.
In one embodiment, it is also performed the steps of when computer program is executed by processor
Merge the corresponding wrinkle characteristic component of the multiple filtering parameter, obtains initial wrinkle component set；
The initial wrinkle component set is denoised, targeted wrinkle component set is obtained；
According to the targeted wrinkle component set, wrinkle quantity, wrinkle length and wrinkle distribution characteristics are detected.
In one embodiment, it is also performed the steps of when computer program is executed by processor
Obtain sample face image and verification face image；
Using the sample face image and the verification face image, machine training is carried out to initial markers model, is obtained
To target label model；
The facial feature points of the label face image, comprising:
The face image is marked by the target label model, obtains the face feature of the face image
Point.
In one embodiment, it is also performed the steps of when computer program is executed by processor
To the image in the crumple zone, according to horizontally and vertically zooming in and out.
In one embodiment, the filtering parameter includes filtering direction, filtering sampling interval and filtering sampling size.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each technical characteristic of above embodiments can be combined arbitrarily, for simplicity of description, not to above-described embodiment
In each technical characteristic it is all possible combination be all described, as long as however, the combination of these technical characteristics be not present lance
Shield all should be considered as described in this specification.
The several embodiments of the application above described embodiment only expresses, the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the concept of this application, various modifications and improvements can be made, these belong to the protection of the application
Range.Therefore, the scope of protection shall be subject to the appended claims for the application patent.
Claims (10)
1. a kind of face wrinkles' detection method characterized by comprising
Face image is obtained, and marks the facial feature points of the face image；
According to the facial feature points of the face image, the crumple zone of the face image is divided, and determines the wrinkle area
The wrinkle type in domain；
According to the wrinkle type of the crumple zone, the area image in the crumple zone is extracted；
Obtain the wrinkle characteristic component of the area image；
Wrinkle detection is carried out according to the wrinkle characteristic component of the area image.
2. the method according to claim 1, wherein the wrinkle type according to the crumple zone, is extracted
Area image in the crumple zone, comprising:
Obtain the corresponding filtering parameter of wrinkle type of the crumple zone；
Using the filtering parameter, the image in the crumple zone is filtered, filtering image is obtained；
Processing is sharpened to the image in the crumple zone using the filtering image, obtains the area image.
3. the method according to claim 1, wherein the wrinkle characteristic component for obtaining the area image,
Include:
Obtain the pixel of the area image；
Calculate the image characteristic matrix of the pixel；
The maximum eigenvalue of described image eigenmatrix is calculated, and, calculate the minimal eigenvalue of described image eigenmatrix；
According to the maximum eigenvalue and the minimal eigenvalue, the wrinkle similarity of the pixel is predicted；
When the pixel wrinkle similarity be greater than preset threshold, determine the pixel be wrinkle pixel；
According to the connected domain of the wrinkle pixel, the wrinkle pixel is connected, using the wrinkle pixel of connection as described in
Wrinkle characteristic component.
4. the method according to claim 1, wherein the crumple zone have corresponding multiple filtering parameters,
The filtering parameter has corresponding wrinkle characteristic component, described to carry out wrinkle according to the wrinkle characteristic component of the area image
Detection, comprising:
Merge the corresponding wrinkle characteristic component of the multiple filtering parameter, obtains initial wrinkle component set；
The initial wrinkle component set is denoised, targeted wrinkle component set is obtained；
According to the targeted wrinkle component set, wrinkle quantity, wrinkle length and wrinkle distribution characteristics are detected.
5. the method according to claim 1, wherein further include:
Obtain sample face image and verification face image；
Using the sample face image and the verification face image, machine training is carried out to initial markers model, obtains mesh
Mark markup model；
The facial feature points of the label face image, comprising:
The face image is marked by the target label model, obtains the facial feature points of the face image.
6. according to the method described in claim 2, it is characterized in that, the filtering parameter is used described, to the wrinkle area
Image in domain is filtered, before obtaining filtering image, further includes:
To the image in the crumple zone, according to horizontally and vertically zooming in and out.
7. according to the method described in claim 2, it is characterized in that, the filtering parameter includes between filtering direction, filtering sampling
Every with filtering sampling size.
8. a kind of face wrinkles' detection device characterized by comprising
Mark module for obtaining face image, and marks the facial feature points of the face image；
Division module divides the crumple zone of the face image for the facial feature points according to the face image, and really
The wrinkle type of the fixed crumple zone；
Image zooming-out module extracts the area image in the crumple zone for the wrinkle type according to the crumple zone；
Component obtains module, for obtaining the wrinkle characteristic component of the area image；
Detection module, for carrying out wrinkle detection according to the wrinkle characteristic component of the area image.
9. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the step of processor realizes any one of claims 1 to 7 the method when executing the computer program.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of method described in any one of claims 1 to 7 is realized when being executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810768838.3A CN109086688A (en) | 2018-07-13 | 2018-07-13 | Face wrinkles' detection method, device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810768838.3A CN109086688A (en) | 2018-07-13 | 2018-07-13 | Face wrinkles' detection method, device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109086688A true CN109086688A (en) | 2018-12-25 |
Family
ID=64837812
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810768838.3A Pending CN109086688A (en) | 2018-07-13 | 2018-07-13 | Face wrinkles' detection method, device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109086688A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110110637A (en) * | 2019-04-25 | 2019-08-09 | 深圳市华嘉生物智能科技有限公司 | A kind of method of face wrinkle of skin automatic identification and wrinkle severity automatic classification |
| CN110443765A (en) * | 2019-08-02 | 2019-11-12 | 厦门美图之家科技有限公司 | Image processing method, device and electronic equipment |
| CN111814520A (en) * | 2019-04-12 | 2020-10-23 | 虹软科技股份有限公司 | Skin type detection method, skin type grade classification method, and skin type detection device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| CN104021380A (en) * | 2014-05-02 | 2014-09-03 | 香港应用科技研究院有限公司 | Method and device performing facial recognition through calculating device |
| CN105310652A (en) * | 2014-06-09 | 2016-02-10 | 松下知识产权经营株式会社 | Wrinkle detection apparatus and wrinkle detection method |
| CN105310690A (en) * | 2014-06-09 | 2016-02-10 | 松下知识产权经营株式会社 | Wrinkle detection apparatus and wrinkle detection method |
| 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 |
| US20170079599A1 (en) * | 2014-07-18 | 2017-03-23 | Fujifilm Corporation | Moisture feeling evaluation device, moisture feeling evaluation method, and moisture feeling evaluation program |
-
2018
- 2018-07-13 CN CN201810768838.3A patent/CN109086688A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| CN104021380A (en) * | 2014-05-02 | 2014-09-03 | 香港应用科技研究院有限公司 | Method and device performing facial recognition through calculating device |
| CN105310652A (en) * | 2014-06-09 | 2016-02-10 | 松下知识产权经营株式会社 | Wrinkle detection apparatus and wrinkle detection method |
| CN105310690A (en) * | 2014-06-09 | 2016-02-10 | 松下知识产权经营株式会社 | Wrinkle detection apparatus and wrinkle detection method |
| US20170079599A1 (en) * | 2014-07-18 | 2017-03-23 | Fujifilm Corporation | Moisture feeling evaluation device, moisture feeling evaluation method, and moisture feeling evaluation program |
| 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 |
Non-Patent Citations (4)
| Title |
|---|
| CHOON-CHING NG ET.AL: ""Automatic Wrinkle Detection using Hybrid Hessian Filter"", 《COMPUTER VISION--ACCV 2014》 * |
| CHOON-CHING NG ET.AL: ""Will Wrinkle Estimate the Face Age?"", 《2015 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS,MAN,AND CYBERNETICS》 * |
| CHOON-CHING NG ET.AL: ""Wrinkle Detection Using Hessian Line Tracking"", 《IEEE ACCESS》 * |
| ERJIN ZHOU ET.AL: ""Extensive Facial Landmark Localization with Coarse-to-fine Convolutional Network Cascade"", 《2013 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION WORKSHOPS》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111814520A (en) * | 2019-04-12 | 2020-10-23 | 虹软科技股份有限公司 | Skin type detection method, skin type grade classification method, and skin type detection device |
| CN110110637A (en) * | 2019-04-25 | 2019-08-09 | 深圳市华嘉生物智能科技有限公司 | A kind of method of face wrinkle of skin automatic identification and wrinkle severity automatic classification |
| CN110443765A (en) * | 2019-08-02 | 2019-11-12 | 厦门美图之家科技有限公司 | Image processing method, device and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10872420B2 (en) | Electronic device and method for automatic human segmentation in image | |
| CN110136243B (en) | Three-dimensional face reconstruction method, system, device and storage medium thereof | |
| CN109657583B (en) | Face key point detection method and device, computer equipment and storage medium | |
| US9305240B2 (en) | Motion aligned distance calculations for image comparisons | |
| CN106156692B (en) | method and device for positioning human face edge feature points | |
| CN109086688A (en) | Face wrinkles' detection method, device, computer equipment and storage medium | |
| CN106447592B (en) | Online personalization service per feature descriptor | |
| CN108805203A (en) | Image procossing and object recognition methods, device, equipment and storage medium again | |
| CN111797650A (en) | Obstacle identification method and device, computer equipment and storage medium | |
| CN111461170A (en) | Vehicle image detection method and device, computer equipment and storage medium | |
| Bethmann et al. | Object-based multi-image semi-global matching–concept and first results | |
| CN112101195A (en) | Crowd density estimation method and device, computer equipment and storage medium | |
| Wu et al. | Single-shot face anti-spoofing for dual pixel camera | |
| Tsoli et al. | Patch-based reconstruction of a textureless deformable 3d surface from a single rgb image | |
| CN109685029B (en) | Flexible object identification method, device, equipment and medium with complex space structure | |
| CN111144404A (en) | Legacy object detection method, device, system, computer device, and storage medium | |
| CN109063601B (en) | Lip print detection method and device, computer equipment and storage medium | |
| CN111291716B (en) | Sperm cell identification method, sperm cell identification device, computer equipment and storage medium | |
| CN110781739B (en) | Method, device, computer equipment and storage medium for extracting pedestrian characteristics | |
| KR101904480B1 (en) | Object recognition system and method considering camera distortion | |
| CN108876713B (en) | Mapping method and device of two-dimensional template image, terminal equipment and storage medium | |
| CN110490020A (en) | Bar code recognition, device, computer equipment and storage medium | |
| WO2022206679A1 (en) | Image processing method and apparatus, computer device and storage medium | |
| CN114757822A (en) | Binocular-based human body three-dimensional key point detection method and system | |
| Basavaraj et al. | Real Time Object Distance and Dimension Measurement using Deep Learning and OpenCV |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20201119 Address after: Room 201, No. 9, Tengfei 2nd Street, Huangpu District, Guangzhou City, Guangdong Province Applicant after: GUANGZHOU LAIDE PU DETECTION TECHNOLOGY Co.,Ltd. Address before: 102200 room 1, Building 29, life park, Changping District science and Technology Park, Beijing, B316-82 Applicant before: BEIJING KELAI PUYUN TECHNOLOGY Co.,Ltd. |
|
| TA01 | Transfer of patent application right |