CN109753886A - A kind of evaluation method of facial image, device and equipment - Google Patents
A kind of evaluation method of facial image, device and equipment Download PDFInfo
- Publication number
- CN109753886A CN109753886A CN201811540914.1A CN201811540914A CN109753886A CN 109753886 A CN109753886 A CN 109753886A CN 201811540914 A CN201811540914 A CN 201811540914A CN 109753886 A CN109753886 A CN 109753886A
- Authority
- CN
- China
- Prior art keywords
- face
- characteristic point
- evaluation index
- region
- human 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.)
- Granted
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 105
- 230000001815 facial effect Effects 0.000 title claims abstract description 55
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 30
- 238000000034 method Methods 0.000 claims abstract description 29
- 238000004891 communication Methods 0.000 claims description 19
- 230000002596 correlated effect Effects 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 9
- 230000000875 corresponding effect Effects 0.000 claims description 7
- 238000013527 convolutional neural network Methods 0.000 claims description 6
- 238000010586 diagram Methods 0.000 description 15
- 238000001514 detection method Methods 0.000 description 11
- 238000012545 processing Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 238000007689 inspection Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000000802 evaporation-induced self-assembly Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
The embodiment of the invention provides a kind of evaluation method of facial image, device and equipment, wherein this method comprises: obtaining image to be detected；By Face datection algorithm, the Face datection region and human face characteristic point of image to be detected are determined；Determine the positional relationship between human face characteristic point and the proportionate relationship and human face characteristic point in Face datection region；According to the positional relationship between human face characteristic point and the proportionate relationship and human face characteristic point in Face datection region, face evaluation index is determined.Evaluation method, device and the equipment of the facial image provided through the embodiment of the present invention, can evaluate facial image, improve the accuracy of recognition of face.
Description
Technical field
The present invention relates to technical field of image processing, more particularly to a kind of evaluation method of facial image, device and set
It is standby.
Background technique
With the development of multimedia technology etc., facial image is more and more important.Such as field of video monitoring, pass through crawl face
Image carries out recognition of face, and then realizes and be monitored to the people of monitoring area；Intelligence shooting field, shoots facial image, leads to
Subsequent processing can be carried out to facial image by crossing recognition of face, such as is adjusted to the brightness of facial image, coloration.
As can be seen that facial image is the basis for carrying out recognition of face, the quality of facial image directly affects recognition of face
Accuracy, especially in the case where quality of human face image is poor, dramatically increase recognition of face erroneous judgement probability.
In this way, being evaluated the quality of facial image is just particularly important.
Summary of the invention
The evaluation method for being designed to provide a kind of facial image, device and the equipment of the embodiment of the present invention, to improve people
The accuracy of face identification.Specific technical solution is as follows:
In a first aspect, the embodiment of the invention provides a kind of evaluation methods of facial image, comprising:
Obtain image to be detected；
By Face datection algorithm, the Face datection region and human face characteristic point of described image to be detected are determined；
It determines between the human face characteristic point and the proportionate relationship and the human face characteristic point in the Face datection region
Positional relationship；
According between the human face characteristic point and the proportionate relationship and the human face characteristic point in the Face datection region
Positional relationship, determine face evaluation index.
Optionally, the human face characteristic point include: two, nose, the left corners of the mouth and the right corners of the mouth；Described two include left eye and
Right eye；
The proportionate relationship and the human face characteristic point in the determination human face characteristic point and the Face datection region
Between positional relationship, comprising:
Determine the first wide ratio in the distance of horizontal direction and the Face datection region between described two, described two
The distance of vertical direction and the second high ratio in the Face datection region and the nose and two companies between eye
The distance of the perpendicular bisector of line；
The proportionate relationship and the human face characteristic point according to the human face characteristic point and the Face datection region
Between positional relationship, determine face evaluation index, comprising:
According to the perpendicular bisector of first ratio, second ratio and the nose and two lines away from
From determining the face evaluation index.
Optionally, the face evaluation index and first ratio are positively correlated；The face evaluation index and the nose
It is sharp negatively correlated at a distance from the perpendicular bisector of two lines.
Optionally, the Face datection region and people of described image to be detected are determined by Face datection algorithm described
After face characteristic point, the method also includes:
Determine the size in the Face datection region and described image to be detected；
The proportionate relationship and the human face characteristic point according to the human face characteristic point and the Face datection region
Between positional relationship, determine face evaluation index, comprising:
According to the position between the proportionate relationship in the human face characteristic point and the Face datection region, the human face characteristic point
Relationship and the size are set, determines face evaluation index.
Optionally, described according to the proportionate relationship in the human face characteristic point and the Face datection region, the face
Positional relationship and the size between characteristic point, after determining face evaluation index, the method also includes:
Judge whether the face evaluation index is higher than preset threshold；
When the face evaluation index is higher than the preset threshold, determine the face evaluation index it is corresponding it is described to
Detection image is target image.
Optionally, the Face datection algorithm includes multitask concatenated convolutional neural network MTCNN algorithm, the face inspection
Surveying region includes rectangle frame region.
Second aspect, the embodiment of the invention provides a kind of evaluating apparatus of facial image, comprising:
Module is obtained, for obtaining image to be detected；
First determining module, for by Face datection algorithm, determine the Face datection region of described image to be detected with
And human face characteristic point；
Second determining module, for determining proportionate relationship and the institute of the human face characteristic point and the Face datection region
State the positional relationship between human face characteristic point；
Third determining module, for according to the human face characteristic point and the Face datection region proportionate relationship and institute
The positional relationship between human face characteristic point is stated, determines face evaluation index.
Optionally, the human face characteristic point include: two, nose, the left corners of the mouth and the right corners of the mouth；Described two include left eye and
Right eye；
Second determining module, distance and the Face datection specifically for horizontal direction between determination described two
The first wide ratio in region, it is two described between vertical direction distance and the Face datection region the second high ratio
And the nose is at a distance from the perpendicular bisector of two lines；
The third determining module, be specifically used for according to first ratio, second ratio and the nose with
The distance of the perpendicular bisector of two lines, determines the face evaluation index.
Optionally, the face evaluation index and first ratio are positively correlated；The face evaluation index and the nose
It is sharp negatively correlated at a distance from the perpendicular bisector of two lines.
Optionally, described device further include:
4th determining module, for determining the size in the Face datection region and described image to be detected；
The third determining module, for according to the proportionate relationship in the human face characteristic point and the Face datection region,
Positional relationship and the size between the human face characteristic point, determine face evaluation index.
Optionally, described device further include:
Judgment module, in the proportionate relationship according to the human face characteristic point with the Face datection region, institute
The positional relationship between human face characteristic point and the size are stated, after determining face evaluation index, judges the face
Whether evaluation index is higher than preset threshold；
5th determining module, for determining that the face is commented when the face evaluation index is higher than the preset threshold
The corresponding described image to be detected of valence index is target image.
Optionally, the Face datection algorithm includes multitask concatenated convolutional neural network MTCNN algorithm, the face inspection
Surveying region includes rectangle frame region.
The third aspect, the embodiment of the invention provides a kind of electronic equipment, including processor, communication interface, memory and
Communication bus, wherein the processor, the communication interface, the memory are completed each other by the communication bus
Communication；
The memory, for storing computer program；
The processor when for executing the program stored on the memory, realizes method described in first aspect
Step.
At the another aspect that the present invention is implemented, a kind of computer readable storage medium is additionally provided, it is described computer-readable
Instruction is stored in storage medium, when run on a computer, so that computer executes side described in above-mentioned first aspect
Method step.
At the another aspect that the present invention is implemented, a kind of computer program product comprising instruction is additionally provided, when it is being counted
When being run on calculation machine, so that computer executes method and step described in above-mentioned first aspect.
Evaluation method, device and the equipment of facial image provided in an embodiment of the present invention, available image to be detected；It is logical
Remarkable face detection algorithm determines the Face datection region and human face characteristic point of image to be detected；Determine human face characteristic point and people
Positional relationship between the proportionate relationship and human face characteristic point of face detection zone；According to human face characteristic point and Face datection region
Proportionate relationship and human face characteristic point between positional relationship, determine face evaluation index.In the embodiment of the present invention, Ke Yigen
It is closed according to the position between human face characteristic point in image to be detected and the proportionate relationship and human face characteristic point in Face datection region
System, determines the face evaluation index for evaluating face in image to be detected.It so, it is possible to evaluate facial image, mention
The accuracy of high recognition of face.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described.
Fig. 1 is a kind of flow chart of the evaluation method of facial image provided in an embodiment of the present invention；
Fig. 2 (a) is a kind of schematic diagram of face different conditions and human face characteristic point in the embodiment of the present invention；
Fig. 2 (b) is another schematic diagram of face different conditions and human face characteristic point in the embodiment of the present invention；
Fig. 2 (c) is another schematic diagram of face different conditions and human face characteristic point in the embodiment of the present invention；
Fig. 2 (d) is another schematic diagram of face different conditions and human face characteristic point in the embodiment of the present invention；
Fig. 2 (e) is another schematic diagram of face different conditions and human face characteristic point in the embodiment of the present invention；
Fig. 2 (f) is another schematic diagram of face different conditions and human face characteristic point in the embodiment of the present invention；
Fig. 3 (a) is that face different conditions human face monitors region in the embodiment of the present invention and one kind of human face characteristic point is shown
It is intended to；
Fig. 3 (b) is the another kind that face different conditions human face monitors region and human face characteristic point in the embodiment of the present invention
Schematic diagram；
Fig. 4 is another flow chart of the evaluation method of facial image provided in an embodiment of the present invention；
Fig. 5 (a) is one of embodiment of the present invention appraisal result schematic diagram；
Fig. 5 (b) is another appraisal result schematic diagram in the embodiment of the present invention；
Fig. 5 (c) is another appraisal result schematic diagram in the embodiment of the present invention；
Fig. 5 (d) is another appraisal result schematic diagram in the embodiment of the present invention；
Fig. 5 (e) is another appraisal result schematic diagram in the embodiment of the present invention；
Fig. 5 (f) is another appraisal result schematic diagram in the embodiment of the present invention；
Fig. 6 is another flow chart of the evaluation method of facial image provided in an embodiment of the present invention；
Fig. 7 is the structural schematic diagram of the evaluating apparatus of facial image provided in an embodiment of the present invention；
Fig. 8 is the structural schematic diagram of electronic equipment provided in an embodiment of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention is described.
Facial image is the basis for carrying out recognition of face, and the quality of facial image directly affects the accuracy of recognition of face,
Especially in the case where quality of human face image is poor, the probability of recognition of face erroneous judgement is dramatically increased.In this way, to people
The quality of face image, which is evaluated, to be just particularly important.
Wherein, evaluation, which can be understood as whether holding the face for including in facial image, to be carried out to the quality of facial image
Just, it clearly etc. is evaluated.
For example, usual video camera can not determine that current photo face is automatically in portrait shooting, especially intelligence shooting
It is no clear, proper, this meeting so that the photo finally shot be side face, it is too small etc., this to the work such as recognition of face can increase misjudgement,
The probability of erroneous judgement.Therefore, whether clear, proper mode is with regard to ten for the face in a kind of facial image of actively perceive current shooting
Divide important.
In the embodiment of the present invention, by the Face datection region of image to be detected and human face characteristic point, face spy is determined
Sign point and the positional relationship between the proportionate relationship and human face characteristic point in Face datection region, and according to human face characteristic point and people
Positional relationship between the proportionate relationship and human face characteristic point of face detection zone is determined for evaluating face in image to be detected
Proper degree evaluation index.Meanwhile it can determine the size of Face datection region and image to be detected；According to face
Positional relationship and the size between characteristic point and the proportionate relationship in Face datection region, human face characteristic point are determined and are used
The proper degree of face and the evaluation index of readability in evaluation image to be detected.In this way, face can be evaluated objectively
The proper degree of face in image is rectified and readability, and whether face is rectified, is clear in the facial image judged, into
The accuracy of one step raising recognition of face.
The evaluation method of facial image provided in an embodiment of the present invention is described in detail below.
The evaluation method of facial image provided in an embodiment of the present invention can be applied to electronic equipment, and specifically, electronics is set
It is standby to can be terminal, server, processor etc..
The embodiment of the invention provides a kind of evaluation methods of facial image, as shown in Figure 1, may include:
S101 obtains image to be detected.
Image to be detected can be the image that electronic equipment is got from other devices or electronic equipment itself includes figure
As acquisition module, image to be detected can be the image that electronic equipment is directly obtained by image capture module.
Image to be detected can be the image comprising face grabbed in intelligence shooting, can be field of video monitoring shooting
The image, etc. comprising face.
S102 determines the Face datection region and human face characteristic point of image to be detected by Face datection algorithm.
Face datection algorithm may include: Face datection algorithm based on histogram coarse segmentation and singular value features, be based on
The Face datection algorithm of dyadic wavelet transform, the Face datection algorithm based on AdaBoost algorithm are based on facial eyes structure spy
Face datection algorithm of sign, etc..In addition, determining Face datection region and determining that human face characteristic point can be calculated by Face datection
Method carries out simultaneously；It can also separate and carry out.
The embodiment of the present invention is not defined the mode of Face datection algorithm, any form that Face datection may be implemented
In the protection scope of the embodiment of the present invention.
In a kind of achievable mode, Face datection algorithm includes multitask concatenated convolutional neural network (Multi-task
Convolutional neural networks, MTCNN) algorithm, Face datection region may include rectangle frame region.MTCNN
Face datection region and human face characteristic point can be detected simultaneously, can be improved the efficiency of detection.
Human face characteristic point may include the face in face.Specifically, human face characteristic point may include: two, nose, a left side
The corners of the mouth and the right corners of the mouth, two include left eye and right eye.Such as Fig. 2 (a), Fig. 2 (b), Fig. 2 (c), Fig. 2 (d), Fig. 2 (e) and Fig. 2 (f)
Shown in medium and small black circle.
S103 determines that the position between human face characteristic point and the proportionate relationship and human face characteristic point in Face datection region is closed
System.
The proportionate relationship in human face characteristic point and Face datection region may include: the distance between two and Face datection area
The distance between ratio, nose and the mouth in domain and the ratio in Face datection region, etc..
Positional relationship between human face characteristic point may include: the distance between two, nose is relative to two lines
The distance of perpendicular bisector, etc. of the distance, mouth of perpendicular bisector relative to two lines.
When determining Face datection region and human face characteristic point, it can determine that position, size of the face detection zone etc. are believed
Breath and the location information of each human face characteristic point etc., such as coordinate of each human face characteristic point.In this way, can be according to face spy
The information such as the location information of point and position, the size in Face datection region are levied, determine human face characteristic point and Face datection region
Proportionate relationship and human face characteristic point between positional relationship.As the distance between two with the ratio in Face datection region,
The ratio of the distance between nose and mouth and Face datection region；The distance between two, nose is relative to two lines
The distance of perpendicular bisector, etc. of the distance, mouth of perpendicular bisector relative to two lines.
S104 is closed according to the position between human face characteristic point and the proportionate relationship and human face characteristic point in Face datection region
System, determines face evaluation index.
Face evaluation index herein is the evaluation index for evaluating the proper degree of face in image to be detected.
When face is in different conditions in facial image, the proportionate relationship and people in human face characteristic point and Face datection region
Positional relationship between face characteristic point is different.Therefore, it can be closed according to the ratio in human face characteristic point and Face datection region
Positional relationship between system and human face characteristic point, determination refer to for evaluating the evaluation of the proper degree of face in image to be detected
Mark.
In the embodiment of the present invention, it can be closed according to the ratio in human face characteristic point in image to be detected and Face datection region
Positional relationship between system and human face characteristic point, determines the face evaluation index for evaluating face in image to be detected.Such as
This, can evaluate facial image, improve the accuracy of recognition of face.
Analysis chart 2 (a), Fig. 2 (b), Fig. 2 (c), Fig. 2 (d), Fig. 2 (e) and Fig. 2 (f) as can be seen that Fig. 2 (a), Fig. 2 (b),
Face is positive face in facial image in Fig. 2 (c)；Fig. 2 (d), Fig. 2 (e), face is side face in facial image in Fig. 2 (f).Analysis
Fig. 2 (a), Fig. 2 (b), Fig. 2 (c) and Fig. 2 (d), Fig. 2 (e), Fig. 2 (f) are as can be seen that Fig. 2 (a), Fig. 2 (b), face in Fig. 2 (c)
Face is positive face in image；Fig. 2 (d), Fig. 2 (e), face is side face in facial image in Fig. 2 (f).In 5 human face characteristic points,
Positive face and the maximum differentiation of side face are the relative positions of nose, and when face is positive face, nose always hangs down in two lines
Near line, and when the face taken is side face, the position of nose always with the perpendicular bisector of two lines distance farther out, to scheme
For 2 (f), nose then may determine that the face is seen towards a left side on the perpendicular bisector left side of two lines.And its remoter, explanation of distance
More side.In this way, can using nose with respect to two lines perpendicular bisector relative position as estimate face departure degree according to
According to by the relative position of the perpendicular bisector of opposite two lines of nose, estimation face deviates positive degree, it is understood that is
Side degree of side face etc. determines the proper degree of face.
In a kind of optional embodiment of the present invention, step S103: the ratio of human face characteristic point and Face datection region is determined
Positional relationship between relationship and human face characteristic point may include:
Determine the first wide ratio in the distance of horizontal direction and Face datection region between two, Vertical Square between two
To distance at a distance from the perpendicular bisector of high second ratio and nose and two lines in Face datection region.
Step S104: according to the position between human face characteristic point and the proportionate relationship and human face characteristic point in Face datection region
Relationship is set, face evaluation index is determined, may include:
According to the first ratio, the second ratio and nose at a distance from the perpendicular bisector of two lines, determine that face evaluation refers to
Mark.
Specifically, face evaluation index and the first ratio are positively correlated.I.e. face evaluation index with the first ratio increase
And increase, reduce with the reduction of the first ratio.
Face evaluation index is negatively correlated at a distance from the perpendicular bisector of nose and two lines.That is face evaluation index is with nose
Increase of the point at a distance from the perpendicular bisector of two lines and reduce, with reduction of the nose at a distance from the perpendicular bisector of two lines
And increase.
In a kind of specific embodiment, as shown in Fig. 3 (a) and Fig. 3 (b), image to be detected 301, the face of image to be detected
Detection zone is as shown in rectangle frame 302, and the perpendicular bisector of two lines is as illustrated by the dotted lines 303.
Specifically, the width of image to be detected 301 is W, a height of H, the Face datection region of image to be detected 301, such as face
A height of h of detection block 302f, width wf；Left eye coordinates are (x in human face characteristic point1,y1), right eye coordinate is (x2,y2), nose is sat
It is designated as (x3,y3)。
The face of detection is more proper, and two spacing and the ratio of Face datection frame are bigger, and two in horizontal position, meaning
Taste face does not rotate, and can specifically be indicated by following formula (1-1).
Wherein,For the first ratio,For the second ratio.
Nose can specifically be indicated as close as possible to the perpendicular bisector of two lines by following formula (1-2).
Wherein,It is nose at a distance from the perpendicular bisector of two lines.
On the basis of the above embodiments, in a kind of optional embodiment of the present invention, in step S102: passing through Face datection
Algorithm, after the Face datection region and the human face characteristic point that determine image to be detected, as shown in figure 4, can also include:
S105 determines the size of Face datection region and image to be detected.
Under normal circumstances, the bigger expression facial image of the size in Face datection region is more clear.It specifically, can be by such as
Lower formula (1-3) indicates.
argmax(hf),argmax(wf) (1-3)
In order to avoid the difference of image to be detected size, cause solely to determine by the size in Face datection region itself
Error caused by Face datection area size can determine the size of Face datection region and image to be detected, with determination
The size of image to be detected shared by Face datection region.
Step S104: according to the position between human face characteristic point and the proportionate relationship and human face characteristic point in Face datection region
Relationship is set, face evaluation index is determined, may include:
S1041, according to the positional relationship between the proportionate relationship in human face characteristic point and Face datection region, human face characteristic point
And the size, determine face evaluation index.
Face evaluation index herein is for evaluating the proper degree of face and commenting for readability in image to be detected
Valence index.That is the face evaluation index proper degree and readability that can be used for evaluating face in image to be detected indicates
Whether face is proper and clear in image to be detected.
It, can be with when determining in the above-described embodiments for evaluating the evaluation index of the proper degree of face in image to be detected
The readability of face in image to be detected is considered simultaneously.It can be closed according to the ratio in human face characteristic point and Face datection region
Positional relationship between system, human face characteristic point, determines the evaluation index for evaluating the proper degree of face in image to be detected,
If considering simultaneously, the factor for influencing the readability of face then may be used such as the size in Face datection region and image to be detected
According to according to the positional relationship and face between the proportionate relationship in human face characteristic point and Face datection region, human face characteristic point
The size of detection zone and image to be detected determines proper degree and clear journey for evaluating face in image to be detected
The evaluation index of degree.
It specifically, can be according to the position between the proportionate relationship in human face characteristic point and Face datection region, human face characteristic point
Relationship and the size are set, by preset formula, determines face evaluation index.
Preset formula may include following scoring formula (1-4):
Wherein, the scoring that pface_prob is, as face evaluation index, a are the weight of face detection block size,
In a kind of achievable mode, 0.5 can be set by the weight a of Face datection frame size.
The readability of face and proper degree in facial image are comprehensively considered in formula (1-4), specifically, face inspection
Survey the size of frame region size；Relationship between Face datection region and human face characteristic point, as between two horizontal direction away from
From the first wide ratio with Face datection region, high second between two at a distance from vertical direction with Face datection region
Ratio；Relationship between human face characteristic point, if nose is at a distance from the perpendicular bisector of two lines.
This group of Fig. 5 (a), Fig. 5 (b), Fig. 5 (c), Fig. 5 (d), Fig. 5 (e) and Fig. 5 (f) picture are shown using formula (1-
4) it scores the facial image of face different angle, wherein the weight a of Face datection frame size is 0.5, is scored at every
The upper left corner of picture.It accordingly scores when being faced upward in Fig. 5 (a): 0.374；It accordingly scores when bowing in Fig. 5 (b): 0.483；Figure
It accordingly scores when face deflects to the left in 5 (c): 0.076；It accordingly scores when being deflected to the right in Fig. 5 (d): 0.008；Fig. 5
(e) it accordingly scores when face zooms out in: 0.518；Face is rectified and accordingly scores when furthering in Fig. 5 (f): 0.742.It can be with
Being interpreted as Fig. 5 (f) is that face is rectified and clearly optimal picture in this group of picture.
As can be seen that scoring be it is effective, either face upward still downward, towards left or towards the right side, or apart from drawing
Far, all without closer distance and proper face partition it is high.When only face is clear and proper, just there is higher scoring.
In this way, the readability and proper degree captured in shooting process can be chosen in face actively perceive task
Highest picture, as the capture result in face actively perceive task.
In the embodiment of the present invention, Face datection and a small number of facial key points, i.e. human face characteristic point are utilized, it is determined that Yi Zhongping
The proper degree of valence face and the scoring of readability, can evaluate facial image, and face knowledge can be improved
Not, the recognition capability of face snap etc., such as identification accuracy.And it can automatically determine for evaluating face in image to be detected
Proper degree evaluation index, for evaluating the proper degree of face and the evaluation index of readability in image to be detected,
It so, it is possible the proper degree for quickly and efficiently determining face and readability, recognition of face, face snap etc. can be improved
Efficiency.Meanwhile guidance can be provided for active capture apparatus, for example offers such as monitor camera, smart camera shooting refers to
It leads.
On the basis of the above embodiments, in a kind of optional embodiment of the present invention, in step S106: according to face characteristic
Positional relationship and size between the proportionate relationship in point and Face datection region, human face characteristic point, determine that face is evaluated
After index, as shown in fig. 6, can also include:
S107, judges whether face evaluation index is higher than preset threshold.
S108 determines that the corresponding image to be detected of face evaluation index is when face evaluation index is higher than preset threshold
Target image.
Preset threshold can be determined according to practical application.Specifically, journey can be rectified to face based on different applications
The difference that degree and readability require, determines corresponding preset threshold.In this way, meet demand can be selected according to the preset threshold
Image.
In order to select apparent, proper facial image, can choose face evaluation index higher than preset threshold to
Retrieval image is target image to be selected.
In this way, rectifying of meeting the requirements and clearly image can be selected according to different requirements,.
The embodiment of the invention also provides a kind of evaluating apparatus of facial image, as shown in fig. 7, comprises:
Module 701 is obtained, for obtaining image to be detected；
First determining module 702, for by Face datection algorithm, determine image to be detected Face datection region and
Human face characteristic point；
Second determining module 703, the proportionate relationship and face for determining human face characteristic point and Face datection region are special
Positional relationship between sign point；
Third determining module 704, for special according to the proportionate relationship and face of human face characteristic point and Face datection region
Positional relationship between sign point, determines face evaluation index.
In the embodiment of the present invention, it can be closed according to the ratio in human face characteristic point in image to be detected and Face datection region
Positional relationship between system and human face characteristic point, determines the face evaluation index for evaluating face in image to be detected.Such as
This, can evaluate facial image, improve the accuracy of recognition of face.
Optionally, human face characteristic point include: two, nose, the left corners of the mouth and the right corners of the mouth；Two include left eye and right eye；
Second determining module 703, specifically for determining the width of the distance of horizontal direction and Face datection region between two
The first ratio, between two the distance of vertical direction and Face datection region high the second ratio and nose and two companies
The distance of the perpendicular bisector of line；
Third determining module 704, specifically for hanging down according in the first ratio, the second ratio and nose and two lines
The distance of line determines face evaluation index.
Optionally, face evaluation index and the first ratio are positively correlated；In face evaluation index and nose and two lines
The distance of vertical line is negatively correlated.
Optionally, the device further include:
4th determining module, for determining the size in Face datection region and image to be detected；
Third determining module 704, specifically for special according to the proportionate relationship of human face characteristic point and Face datection region, face
Positional relationship and size between sign point, determine face evaluation index.
Optionally, the device further include:
Judgment module, for according between the proportionate relationship in human face characteristic point and Face datection region, human face characteristic point
Positional relationship and size, after determining face evaluation index, judge whether face evaluation index is higher than preset threshold；
5th determining module, for determining that face evaluation index is corresponding when face evaluation index is higher than preset threshold
Image to be detected is target image.
Optionally, Face datection algorithm includes multitask concatenated convolutional neural network MTCNN algorithm, Face datection region packet
Include rectangle frame region.
It should be noted that the evaluating apparatus of facial image provided in an embodiment of the present invention is using above-mentioned facial image
The device of evaluation method, then all embodiments of the evaluation method of above-mentioned facial image are suitable for the device, and can reach
The same or similar beneficial effect.
The embodiment of the invention also provides a kind of electronic equipment, as shown in figure 8, include processor 801, communication interface 802,
Memory 803 and communication bus 804, wherein processor 801, communication interface 802, memory 803 are complete by communication bus 804
At mutual communication.
Memory 803, for storing computer program；
Processor 801 when for executing the program stored on memory 803, realizes facial image in above-described embodiment
Evaluation method method and step.
In the embodiment of the present invention, it can be closed according to the ratio in human face characteristic point in image to be detected and Face datection region
Positional relationship between system and human face characteristic point, determines the face evaluation index for evaluating face in image to be detected.Such as
This, can evaluate facial image, improve the accuracy of recognition of face.
The communication bus that above-mentioned electronic equipment is mentioned can be Peripheral Component Interconnect standard (Peripheral Component
Interconnect, abbreviation PCI) bus or expanding the industrial standard structure (Extended Industry Standard
Architecture, abbreviation EISA) bus etc..The communication bus can be divided into address bus, data/address bus, control bus etc..
Only to be indicated with a thick line in figure, it is not intended that an only bus or a type of bus convenient for indicating.
Communication interface is for the communication between above-mentioned electronic equipment and other equipment.
Memory may include random access memory (Random Access Memory, abbreviation RAM), also may include
Nonvolatile memory (non-volatile memory), for example, at least a magnetic disk storage.Optionally, memory may be used also
To be storage device that at least one is located remotely from aforementioned processor.
Above-mentioned processor can be general processor, including central processing unit (Central Processing Unit,
Abbreviation CPU), network processing unit (Network Processor, abbreviation NP) etc.；It can also be digital signal processor
(Digital Signal Processing, abbreviation DSP), specific integrated circuit (Application Specific
Integrated Circuit, abbreviation ASIC), field programmable gate array (Field-Programmable Gate Array,
Abbreviation FPGA) either other programmable logic device, discrete gate or transistor logic, discrete hardware components.
In another embodiment provided by the invention, a kind of computer readable storage medium is additionally provided, which can
It reads to be stored with instruction in storage medium, when run on a computer, so that computer executes face figure in above-described embodiment
The method and step of the evaluation method of picture.
In the embodiment of the present invention, it can be closed according to the ratio in human face characteristic point in image to be detected and Face datection region
Positional relationship between system and human face characteristic point, determines the face evaluation index for evaluating face in image to be detected.Such as
This, can evaluate facial image, improve the accuracy of recognition of face.
In another embodiment provided by the invention, a kind of computer program product comprising instruction is additionally provided, when it
When running on computers, so that computer executes the method and step of the evaluation method of facial image in above-described embodiment.
In the embodiment of the present invention, it can be closed according to the ratio in human face characteristic point in image to be detected and Face datection region
Positional relationship between system and human face characteristic point, determines the face evaluation index for evaluating face in image to be detected.Such as
This, can evaluate facial image, improve the accuracy of recognition of face.
In the above-described embodiments, can come wholly or partly by software, hardware, firmware or any combination thereof real
It is existing.When implemented in software, it can entirely or partly realize in the form of a computer program product.The computer program
Product includes one or more computer instructions.When loading on computers and executing the computer program instructions, all or
It partly generates according to process or function described in the embodiment of the present invention.The computer can be general purpose computer, dedicated meter
Calculation machine, computer network or other programmable devices.The computer instruction can store in computer readable storage medium
In, or from a computer readable storage medium to the transmission of another computer readable storage medium, for example, the computer
Instruction can pass through wired (such as coaxial cable, optical fiber, number from a web-site, computer, server or data center
User's line (DSL)) or wireless (such as infrared, wireless, microwave etc.) mode to another web-site, computer, server or
Data center is transmitted.The computer readable storage medium can be any usable medium that computer can access or
It is comprising data storage devices such as one or more usable mediums integrated server, data centers.The usable medium can be with
It is magnetic medium, (for example, floppy disk, hard disk, tape), optical medium (for example, DVD) or semiconductor medium (such as solid state hard disk
Solid State Disk (SSD)) etc..
It should be noted that, in this document, relational terms such as first and second and the like are used merely to a reality
Body or operation are distinguished with another entity or operation, are deposited without necessarily requiring or implying between these entities or operation
In any actual relationship or order or sequence.Moreover, the terms "include", "comprise" or its any other variant are intended to
Non-exclusive inclusion, so that the process, method, article or equipment including a series of elements is not only wanted including those
Element, but also including other elements that are not explicitly listed, or further include for this process, method, article or equipment
Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that
There is also other identical elements in process, method, article or equipment including the element.
Each embodiment in this specification is all made of relevant mode and describes, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.Especially for device,
For equipment, computer readable storage medium and computer program product embodiments, since it is substantially similar to the method embodiment,
So being described relatively simple, the relevent part can refer to the partial explaination of embodiments of method.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the scope of the present invention.It is all
Any modification, equivalent replacement, improvement and so within the spirit and principles in the present invention, are all contained in protection scope of the present invention
It is interior.
Claims (13)
1. a kind of evaluation method of facial image characterized by comprising
Obtain image to be detected；
By Face datection algorithm, the Face datection region and human face characteristic point of described image to be detected are determined；
Determine the position between the human face characteristic point and the proportionate relationship and the human face characteristic point in the Face datection region
Set relationship；
According to the position between the human face characteristic point and the proportionate relationship and the human face characteristic point in the Face datection region
Relationship is set, determines face evaluation index.
2. the method according to claim 1, wherein the human face characteristic point includes: two, nose, the left corners of the mouth
With the right corners of the mouth；Described two include left eye and right eye；
Between the determination human face characteristic point and the proportionate relationship and the human face characteristic point in the Face datection region
Positional relationship, comprising:
Determine the first wide ratio of the distance of horizontal direction and the Face datection region between described two, it is described two it
Between the distance of vertical direction and the second high ratio in the Face datection region and the nose and two lines
The distance of perpendicular bisector；
It is described according between the human face characteristic point and the proportionate relationship and the human face characteristic point in the Face datection region
Positional relationship, determine face evaluation index, comprising:
According to first ratio, second ratio and the nose at a distance from the perpendicular bisector of two lines, really
The fixed face evaluation index.
3. according to the method described in claim 2, it is characterized in that, the face evaluation index and the first ratio positive
It closes；The face evaluation index is negatively correlated at a distance from the perpendicular bisector of the nose and two lines.
4. method according to any one of claims 1 to 3, which is characterized in that described by Face datection algorithm, determine
After the Face datection region of described image to be detected and human face characteristic point, the method also includes:
Determine the size in the Face datection region and described image to be detected；
It is described according between the human face characteristic point and the proportionate relationship and the human face characteristic point in the Face datection region
Positional relationship, determine face evaluation index, comprising:
It is closed according to the position between the proportionate relationship in the human face characteristic point and the Face datection region, the human face characteristic point
System and the size, determine face evaluation index.
5. according to the method described in claim 4, it is characterized in that, being examined described according to the human face characteristic point and the face
The positional relationship between the proportionate relationship in region, the human face characteristic point and the size are surveyed, determines that face evaluation refers to
After mark, the method also includes:
Judge whether the face evaluation index is higher than preset threshold；
When the face evaluation index is higher than the preset threshold, determine that the face evaluation index is corresponding described to be detected
Image is target image.
6. the method according to claim 1, wherein the Face datection algorithm includes multitask concatenated convolutional mind
Through network MTCNN algorithm, the Face datection region includes rectangle frame region.
7. a kind of evaluating apparatus of facial image characterized by comprising
Module is obtained, for obtaining image to be detected；
First determining module, for determining the Face datection region and people of described image to be detected by Face datection algorithm
Face characteristic point；
Second determining module, for determine the human face characteristic point and the Face datection region proportionate relationship and the people
Positional relationship between face characteristic point；
Third determining module, for according to the human face characteristic point and the Face datection region proportionate relationship and the people
Positional relationship between face characteristic point determines face evaluation index.
8. device according to claim 7, which is characterized in that the human face characteristic point includes: two, nose, the left corners of the mouth
With the right corners of the mouth；Described two include left eye and right eye；
Second determining module, distance and the Face datection region specifically for horizontal direction between determination described two
The first wide ratio, it is two described between the distance of vertical direction and high second ratio in the Face datection region and
The nose is at a distance from the perpendicular bisector of two lines；
The third determining module, be specifically used for according to first ratio, second ratio and the nose with it is described
The distance of the perpendicular bisector of two lines determines the face evaluation index.
9. device according to claim 8, which is characterized in that the face evaluation index and the first ratio positive
It closes；The face evaluation index is negatively correlated at a distance from the perpendicular bisector of the nose and two lines.
10. device according to any one of claims 7 to 9, which is characterized in that described device further include:
4th determining module, for determining the size in the Face datection region and described image to be detected；
The third determining module, specifically for according to the proportionate relationship in the human face characteristic point and the Face datection region,
Positional relationship and the size between the human face characteristic point, determine face evaluation index.
11. device according to claim 10, which is characterized in that described device further include:
Judgment module, in the proportionate relationship according to the human face characteristic point with the Face datection region, the people
Positional relationship and the size between face characteristic point after determining face evaluation index, judge the face evaluation
Whether index is higher than preset threshold；
5th determining module, for when the face evaluation index is higher than the preset threshold, determining that the face evaluation refers to
Marking corresponding described image to be detected is target image.
12. device according to claim 7, which is characterized in that the Face datection algorithm includes multitask concatenated convolutional
Neural network MTCNN algorithm, the Face datection region include rectangle frame region.
13. a kind of electronic equipment, which is characterized in that including processor, communication interface, memory and communication bus, wherein described
Processor, the communication interface, the memory complete mutual communication by the communication bus；
The memory, for storing computer program；
The processor when for executing the program stored on the memory, realizes any side claim 1-6
Method step.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811540914.1A CN109753886B (en) | 2018-12-17 | 2018-12-17 | Face image evaluation method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811540914.1A CN109753886B (en) | 2018-12-17 | 2018-12-17 | Face image evaluation method, device and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109753886A true CN109753886A (en) | 2019-05-14 |
| CN109753886B CN109753886B (en) | 2024-03-08 |
Family
ID=66403893
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811540914.1A Active CN109753886B (en) | 2018-12-17 | 2018-12-17 | Face image evaluation method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109753886B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110276308A (en) * | 2019-06-25 | 2019-09-24 | 上海商汤智能科技有限公司 | Image processing method and device |
| CN111081375A (en) * | 2019-12-27 | 2020-04-28 | 北京深测科技有限公司 | Early warning method and system for health monitoring |
| CN114155593A (en) * | 2022-02-09 | 2022-03-08 | 深圳市海清视讯科技有限公司 | Face recognition method, face recognition device, recognition terminal and storage medium |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101383001A (en) * | 2008-10-17 | 2009-03-11 | 中山大学 | Quick and precise front human face discriminating method |
| CN101615241A (en) * | 2008-06-24 | 2009-12-30 | 上海银晨智能识别科技有限公司 | A kind of screening technique of certificate photograph |
| US20130243268A1 (en) * | 2012-03-13 | 2013-09-19 | Honeywell International Inc. | Face image prioritization based on face quality analysis |
| CN104834898A (en) * | 2015-04-09 | 2015-08-12 | 华南理工大学 | Quality classification method for portrait photography image |
| CN105046246A (en) * | 2015-08-31 | 2015-11-11 | 广州市幸福网络技术有限公司 | Identification photo camera capable of performing human image posture photography prompting and human image posture detection method |
| CN105893946A (en) * | 2016-03-29 | 2016-08-24 | 中国科学院上海高等研究院 | Front face image detection method |
| CN107122054A (en) * | 2017-04-27 | 2017-09-01 | 青岛海信医疗设备股份有限公司 | A kind of detection method and device of face deflection angle and luffing angle |
| CN107590461A (en) * | 2017-09-12 | 2018-01-16 | 广东欧珀移动通信有限公司 | Face identification method and Related product |
| CN107958444A (en) * | 2017-12-28 | 2018-04-24 | 江西高创保安服务技术有限公司 | A kind of face super-resolution reconstruction method based on deep learning |
| CN108230293A (en) * | 2017-05-31 | 2018-06-29 | 深圳市商汤科技有限公司 | Determine method and apparatus, electronic equipment and the computer storage media of quality of human face image |
| CN108960156A (en) * | 2018-07-09 | 2018-12-07 | 苏州浪潮智能软件有限公司 | A kind of Face datection recognition methods and device |
-
2018
- 2018-12-17 CN CN201811540914.1A patent/CN109753886B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101615241A (en) * | 2008-06-24 | 2009-12-30 | 上海银晨智能识别科技有限公司 | A kind of screening technique of certificate photograph |
| CN101383001A (en) * | 2008-10-17 | 2009-03-11 | 中山大学 | Quick and precise front human face discriminating method |
| US20130243268A1 (en) * | 2012-03-13 | 2013-09-19 | Honeywell International Inc. | Face image prioritization based on face quality analysis |
| CN104834898A (en) * | 2015-04-09 | 2015-08-12 | 华南理工大学 | Quality classification method for portrait photography image |
| CN105046246A (en) * | 2015-08-31 | 2015-11-11 | 广州市幸福网络技术有限公司 | Identification photo camera capable of performing human image posture photography prompting and human image posture detection method |
| CN105893946A (en) * | 2016-03-29 | 2016-08-24 | 中国科学院上海高等研究院 | Front face image detection method |
| CN107122054A (en) * | 2017-04-27 | 2017-09-01 | 青岛海信医疗设备股份有限公司 | A kind of detection method and device of face deflection angle and luffing angle |
| CN108230293A (en) * | 2017-05-31 | 2018-06-29 | 深圳市商汤科技有限公司 | Determine method and apparatus, electronic equipment and the computer storage media of quality of human face image |
| CN107590461A (en) * | 2017-09-12 | 2018-01-16 | 广东欧珀移动通信有限公司 | Face identification method and Related product |
| CN107958444A (en) * | 2017-12-28 | 2018-04-24 | 江西高创保安服务技术有限公司 | A kind of face super-resolution reconstruction method based on deep learning |
| CN108960156A (en) * | 2018-07-09 | 2018-12-07 | 苏州浪潮智能软件有限公司 | A kind of Face datection recognition methods and device |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110276308A (en) * | 2019-06-25 | 2019-09-24 | 上海商汤智能科技有限公司 | Image processing method and device |
| CN111081375A (en) * | 2019-12-27 | 2020-04-28 | 北京深测科技有限公司 | Early warning method and system for health monitoring |
| CN111081375B (en) * | 2019-12-27 | 2023-04-18 | 北京深测科技有限公司 | Early warning method and system for health monitoring |
| CN114155593A (en) * | 2022-02-09 | 2022-03-08 | 深圳市海清视讯科技有限公司 | Face recognition method, face recognition device, recognition terminal and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109753886B (en) | 2024-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11062123B2 (en) | Method, terminal, and storage medium for tracking facial critical area | |
| WO2019033574A1 (en) | Electronic device, dynamic video face recognition method and system, and storage medium | |
| CN107871114A (en) | A kind of method, apparatus and system for pushing target person tracking information | |
| CN108269333A (en) | Face identification method, application server and computer readable storage medium | |
| CN109753886A (en) | A kind of evaluation method of facial image, device and equipment | |
| CN109241896A (en) | A kind of channel security detection method, device and electronic equipment | |
| CN109165589A (en) | Vehicle based on deep learning recognition methods and device again | |
| CN110390229B (en) | Face picture screening method and device, electronic equipment and storage medium | |
| CN105631418A (en) | People counting method and device | |
| CN109035330A (en) | Cabinet approximating method, equipment and computer readable storage medium | |
| CN111898486B (en) | Monitoring picture abnormality detection method, device and storage medium | |
| US11983951B2 (en) | Human detection device and human detection method | |
| WO2019033575A1 (en) | Electronic device, face tracking method and system, and storage medium | |
| CN108596135A (en) | Personal identification method and system | |
| CN107578021A (en) | Pedestrian detection method, apparatus and system based on deep learning network | |
| US11948363B2 (en) | Person detection device and person detection method | |
| CN113267828A (en) | Information association method and device, security check equipment and storage medium | |
| CN109389105A (en) | A kind of iris detection and viewpoint classification method based on multitask | |
| CN111079694A (en) | Counter assistant job function monitoring device and method | |
| CN104376323B (en) | A kind of method and device for determining target range | |
| JP2017174343A (en) | Customer attribute extraction device and customer attribute extraction program | |
| US10068335B2 (en) | Moving-object counter apparatus, moving-object counting method, and non-transitory computer readable medium | |
| US11770504B2 (en) | Person detection device and person detection method | |
| Huang et al. | Structural defect detection technology of transmission line damper based on UAV image | |
| CN108363944A (en) | Recognition of face terminal is double to take the photograph method for anti-counterfeit, apparatus and system |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |