CN113723214B - Face key point labeling method, system, electronic equipment and storage medium - Google Patents
Face key point labeling method, system, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN113723214B CN113723214B CN202110904274.3A CN202110904274A CN113723214B CN 113723214 B CN113723214 B CN 113723214B CN 202110904274 A CN202110904274 A CN 202110904274A CN 113723214 B CN113723214 B CN 113723214B
- Authority
- CN
- China
- Prior art keywords
- face
- key points
- coordinates
- point
- face key
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Image Analysis (AREA)
Abstract
The invention provides a method, a system, electronic equipment and a storage medium for labeling key points of a human face, wherein the method comprises the following steps: marking a target frame of the outline of the face in the face picture, and calculating the position coordinates of each key point of the face according to the relative relation between the positions of each key point of the face and the outline of the face; or detecting the key points of the face in the face picture according to the existing face key point detection algorithm of the point type; and acquiring the coordinates of the key points of the second number of faces on the face picture based on the detected coordinates of the key points of the first number of faces. The method for improving the marking efficiency of the key points based on the human face can greatly reduce the problem that the marking of the key points of the human face is not correct, does not need to mark one point by one point, realizes automatic marking by a coordinate calculation method, and finally finely adjusts a few inaccurate positions, thereby greatly improving the working efficiency of marking personnel and saving the labor time cost.
Description
Technical Field
The invention relates to the field of key point labeling, in particular to a method, a system, electronic equipment and a storage medium for labeling key points of a human face.
Background
With the advent of the artificial intelligence era, the training data for AI training is increasingly demanded, and higher demands are also put on the data volume and quality of the labeling data, for example: target detection, face recognition, semantic segmentation, point cloud, face key points and the like, and if the labeling efficiency can be improved, a great deal of labor and financial cost can be saved for enterprises in the face of labeling quantity above hundreds of thousands of levels.
Taking a face key point as an example, a user has the requirement of marking the face key points such as 20 points, 40 points, 60 points, 80 points and the like in practical application, an open-source face key point detection algorithm can only aim at one point type, such as 68 points, can not realize simultaneous compatibility of a plurality of different point types, and a method is needed to realize coverage of the plurality of different point types by using the existing face key point detection algorithm or realize rapid marking of the face key points by using a simple technology under the condition of not using the existing AI algorithm.
Disclosure of Invention
Aiming at the technical problems in the prior art, the invention provides a method, a system, electronic equipment and a storage medium for labeling key points of a human face.
According to a first aspect of the present invention, there is provided a face key point labeling method, including: marking a target frame of the outline of the face in the face picture, and calculating the position coordinates of each key point of the face according to the relative relation between the positions of each key point of the face and the outline of the face; or detecting the face key points in the face picture according to the face key point detection algorithm of the existing point type, and obtaining the coordinates of the first number of face key points in the face picture; and acquiring the coordinates of the key points of the second number of faces on the face picture based on the detected coordinates of the key points of the first number of faces, wherein the second number is unequal to the first number.
On the basis of the technical scheme, the invention can also make the following improvements.
Optionally, the marking a target frame of the outline of the face in the face picture, calculating the position coordinates of each key point of the face according to the relative relation between the positions of each key point of the face and the outline of the face, including: and recording the upper left corner coordinate point and the lower right corner coordinate point of the outline target frame of the human face, calculating the coordinates of each key point of the human face according to the relative positions of each key point of the standard human face on the human face, and marking the positions of each key point of the human face.
Optionally, the labeling the key points of the face further includes: and fine-tuning the positions of the key points of the human face with deviation of the labeling positions.
Optionally, the fine tuning the position of the face key point with deviation from the labeling position includes: dragging the positions of the key points of the human faces with deviation to the target positions on the marking tool; or, the face key points with the deviation are framed, the accurate positions of the face key points with the deviation are calculated through frame coordinates, and the positions of the face key points with the deviation are adjusted according to the accurate positions of the face key points.
Optionally, the acquiring, based on the detected first number of face key points, a second number of face key points on the face picture, where the second number is unequal to the first number, includes: if the second number is greater than the first number, determining the first number of face key point position coordinates in the second number of face key points according to the first number of face key point position coordinates; and calculating the position coordinates of other face key points according to the relative relation between the other face key points and the first number of face key points.
Optionally, if the second number is greater than the first number, determining the first number of face key point coordinates in the second number of face key points according to the first number of face key point coordinates, including: dividing the first number of key point position coordinates of the face to obtain third number of key point position coordinates of different parts of the face; and calculating the position coordinates of each other key point of any part of the face based on the position relative relation between each other key point and the third number of key points, and obtaining the fourth number of key point position coordinates of any part.
Optionally, the acquiring, based on the detected first number of face key points, a second number of face key points on the face picture, where the second number is unequal to the first number, includes: and if the second number is smaller than the first number, omitting redundant face key points in the first number of face key points, and acquiring coordinates of the second number of face key points.
According to a second aspect of the present invention, there is provided a face key point labeling system, comprising: the computing module is used for marking out a target frame of the outline of the face in the face picture, and computing the position coordinates of each key point of the face according to the relative relation between the positions of each key point of the face and the outline of the face; or the first acquisition module is used for detecting the face key points in the face picture according to the face key point detection algorithm of the existing point type to acquire the coordinates of the first number of face key points in the face picture; the second acquisition module is used for acquiring the coordinates of the key points of the face of the second number on the face picture based on the detected coordinates of the key points of the face of the first number, wherein the second number is unequal to the first number.
According to a third aspect of the present invention, there is provided an electronic device including a memory, and a processor, where the processor is configured to implement the steps of the face key point labeling method when executing a computer management program stored in the memory.
According to a fourth aspect of the present invention, there is provided a computer-readable storage medium having stored thereon a computer management class program which, when executed by a processor, implements the steps of a face key point labeling method.
According to the method, the system, the electronic equipment and the storage medium for marking the key points of the human face, the problem that the marking of the key points of the human face is incorrect can be greatly reduced, the key points do not need to be marked point by point, automatic marking is realized through a coordinate calculation method, and finally, a small number of inaccurate positions are finely adjusted, so that the working efficiency of marking personnel is greatly improved, and the labor time cost is saved.
Drawings
FIG. 1 is a flow chart of a face key point labeling method provided by the invention;
FIG. 2 is a schematic flow chart of key points of a face calculated according to the outline of the face;
FIG. 3 is a schematic diagram of a frame for each location on a face;
FIG. 4-1 is a schematic diagram of a face 68 point keypoint detection algorithm using an open source to identify 5 nose keypoint coordinates;
FIG. 4-2 is a schematic illustration of the calculation of 9 keypoint coordinates from 5 keypoint coordinates of the nose;
FIG. 5-1 is a schematic diagram of a face 68 point keypoint detection algorithm using an open source to identify 4 eye keypoint coordinates;
FIG. 5-2 is a schematic illustration of calculating 8 keypoint coordinates from 4 keypoint coordinates of an eye;
FIG. 6 is a schematic diagram of key points of a face finally marked;
FIG. 7 is a schematic diagram of a face key point labeling system according to the present invention;
fig. 8 is a schematic hardware structure of a possible electronic device according to the present invention;
fig. 9 is a schematic hardware structure of a possible computer readable storage medium according to the present invention.
Detailed Description
The following describes in further detail the embodiments of the present invention with reference to the drawings and examples. The following examples are illustrative of the invention and are not intended to limit the scope of the invention.
Fig. 1 is a flowchart of a face key point labeling method provided by the present invention, as shown in fig. 1, the method includes: s1, marking a target frame of an outline of a face in a face picture, and calculating position coordinates of key points of the face according to the relative relation between the positions of the key points of the face and the outline of the face; or S1', detecting the key points of the face in the face picture according to the existing point type face key point detection algorithm, and obtaining the coordinates of the key points of the face in the first number in the face picture; s2', acquiring the coordinates of the key points of the second number of faces on the face picture based on the coordinates of the key points of the first number of faces, wherein the second number is unequal to the first number.
It can be understood that, aiming at the defects in the background technology, the embodiment of the invention provides a method capable of improving the labeling efficiency of key points of a human face, which mainly comprises two modes, wherein the first mode is to automatically label a plurality of key points of the human face without depending on the existing on-source key point detection algorithm of the human face. The second mode is that the existing open source face key point detection algorithm is utilized to detect face key points with one dot type number, then the face key points with other dot types are calculated according to the detected face key points with the known dot type number.
The method is specifically realized in that for a face picture to be marked, a face outline target frame in the face picture is marked, coordinates of the face outline target frame are obtained, and position coordinates of key points of the face are calculated according to the relative relation between the positions of the key points of the face and the face outline.
The second mode is specifically implemented by detecting face key points in a face picture according to an existing point type face key point detection algorithm of an open source, and acquiring coordinates of a first number of face key points in the face picture. And calculating the coordinates of the key points of the second number of faces on the face picture according to the position relation between the key points of other faces and the key points of the detected faces based on the coordinates of the key points of the first number of faces, wherein the second number is unequal to the first number.
The method for improving the marking efficiency of the key points based on the human face can greatly reduce the problem that the marking of the key points of the human face is not correct, does not need to mark one point by one point, realizes automatic marking by a coordinate calculation method, and finally finely adjusts a few inaccurate positions, thereby greatly improving the working efficiency of marking personnel and saving the labor time cost.
In a possible embodiment, marking a target frame of an outline of a face in a face picture, calculating position coordinates of each key point of the face according to a relative relation between the positions of each key point of the face and the outline of the face, including: and recording the upper left corner coordinate point and the lower right corner coordinate point of the outline target frame of the human face, calculating the coordinates of each key point of the human face according to the relative positions of each key point of the standard human face on the human face, and marking the positions of each key point of the human face.
It can be understood that, under the condition of not depending on the existing open source face key point detection algorithm, for a face picture to be marked, first, the left upper corner coordinate and the right lower corner coordinate of a face outline target frame on the face picture are obtained, then the position coordinates of each key point are calculated based on the relative positions of each key point of the face key points of the standard face on the face, and the calculated face key points are marked.
Specifically, referring to fig. 2, for marking the outline of the face, the upper left corner point1 (x 1 ,y 1 ) And lower right corner coordinate point2 (x 2 ，y 2 ). Coordinate calculation of coordinates pointn of each point of the key points of the face is precisely calculated according to the relative positions of each key point of the standard face on the face, for example: position coordinates point_chip of chin (x) i ,y i ) Then x i ＝(x 1 +x 2 )/2，y i ＝y 2 Other positions are similar, and all the coordinates of the key points of the human face are calculated based on the initial coordinates of the upper left corner and the lower right corner of the outer wheel library, that is, only the two points need to be marked, and all the coordinates of the key points of the human face can be automatically generated.
In a possible embodiment, the labeling of each key point of the face further includes: and fine-tuning the positions of the key points of the human face with deviation of the labeling positions.
The specific mode of fine tuning the positions of the face key points with deviation is that fine tuning the positions of the face key points with deviation in labeling positions comprises the following steps: dragging the positions of the key points of the human faces with deviation to the target positions on the marking tool; or, the face key points with the deviation are framed, the accurate positions of the face key points with the deviation are calculated through frame coordinates, and the positions of the face key points with the deviation are adjusted according to the accurate positions of the face key points.
It can be understood that the position coordinates of each key point of different faces are slightly different, if the calculated key point position coordinates are accurate, the key points with deviation of the position coordinates can be finely adjusted without adjustment, and specifically, two fine adjustment methods exist, namely, the first method: the key mark points are directly dragged to the target positions by directly clicking a mouse on the marking tool, and the positions of the key points of different faces are different, but the approximate positions are opposite. The second method is as follows: the method is characterized in that the positions of the face are respectively framed, such as eyes, a nose, a mouth, a chin and the like, the accurate calculation of key points of the different positions is realized through frame coordinates, the position deviation caused by the difference of the different positions of the face can be accurately solved, the framed is still a quick method due to the fact that the number of key points of the eyes, the mouth and the nose is large, the method II can be used as a method for fine adjustment under the condition that the coordinates of the key points are automatically generated in the step S1, and the method can also be independently used independently of the step S1, and the specific operation is that: (1) Taking a face picture, respectively creating a frame for eyebrow, eye, nose, mouth and chin, for example, fig. 3 is a schematic diagram of the frame for each part on the face, and respectively obtaining coordinates p1 (x 1, y 1) and p2 (x 2, y 2) of the upper left corner and the lower right corner of the frame, then the coordinates of the key points of eyes of the face are represented as Pa (xa, ya), pb (xb, yb), pc (xc, yc), pd (xd, yd), pe (xe, ye). (2) the calculation method of the coordinates of each key point on the eyes is as follows: xa=x1, ya= (y1+y2)/2; xb=x2, yb= (y1+y2)/2; xc= (x1+x2)/2, yc=y1; xd= (x1+x2)/2, yd=y2; xe= (x1+x2)/2, ye= (y1+y2)/2, different face eyes are different in size and shape, but different in frame size and shape of eyes, so that accurate positioning of 5 point coordinates of the eyes is achieved through the upper left corner and the lower right corner coordinates, errors are small, negligible basically, and fine adjustment of key point coordinates is not needed. For nose and mouth. Although the nose and mouth shapes of different faces are different, the frame coordinates and the key point coordinates have a certain relationship, or the rest key points have a position relationship with the on-line points and the middle points on the frame line or in the middle, and when the number of key points changes, the position coordinates of each key point of the face can be calculated by adopting the mode of S1 'and S2' described above.
In a possible embodiment, based on the detected first number of face keypoints, a second number of face keypoints on the face picture is obtained, where the second number is unequal to the first number, and the method includes: if the second number is greater than the first number, determining the first number of face key point position coordinates in the second number of face key points according to the first number of face key point position coordinates; and calculating the position coordinates of other face key points according to the relative relation between the other face key points and the first number of face key points.
It can be understood that if the existing face key point detection algorithm of the open source is utilized to detect the face key points of one dot type number in the face image to be marked, the face key points of the other outer dot type number are calculated based on the face key points of the one dot type number.
The method comprises the steps of detecting a first number of key points in a face picture to be marked according to an existing face detection algorithm of an on-off source, and obtaining the detected first number of face key points. And determining a second number of face key points in the second number of face key points according to the detected first number of face key points, and calculating other face key points according to the relative relation between other face key points and the first number of face key points, so as to obtain all the second number of face key points.
If the second number is greater than the first number, determining the first number of face key point coordinates in the second number of face key points according to the first number of face key point coordinates, including: dividing the first number of key point position coordinates of the face to obtain third number of key point position coordinates of different parts of the face; and calculating the position coordinates of each other key point of any part of the face based on the position relative relation between each other key point and the third number of key points, and obtaining the fourth number of key point position coordinates of any part.
And if the second number is smaller than the first number, omitting redundant face key points in the first number of face key points, and acquiring coordinates of the second number of face key points.
Specifically, if the open-source face key point detection algorithm is used, for example, the trained model is used for 68-point coordinate detection, then under the condition of labeling 40-point and 80-point face key points, calculation is only needed according to the positions of fewer points or more points and 68 points of a person, for example, a face nose with 68 points has 5 points, a face nose with 40 points has 3 points, and a face nose with 80 points has 9 points, and then the coordinate positions of the 3 points and the 9 points are calculated according to the coordinates of the existing 5 points respectively. For example, using the open source face 68-point keypoint detection algorithm to identify the 5 nose keypoint coordinates of fig. 4-1, how is the face 80-point keypoint coordinate labeling? It is known that 5-point coordinates of a nose of a face are detected by using an open-source face key point detection algorithm, as shown in fig. 4-1, p1 (x 1, y 1), p2 (x 2, y 2), p3 (x 3, y 3), p4 (x 4, y 4), p5 (x 5, y 5), and nose 9-point coordinates p6 to p14 of 80 points are calculated, and since some point positions are the same, coordinates of p1 and p6, p2 and p8, p4 and p11, p5 and p12, p3 and p10 are the same, and coordinates pb of an upper left corner pa and a lower right corner of fig. 4-2, and coordinates of p6 to p14 are also known in order: p6 (x 6, y 6) to p14 (x 14, y 14), for example: x6= (xa+xb)/2, y6=ya+0.15 x (yb-ya), x7= (xa+xb)/2, y7=ya+0.2 x (yb-ya), x8= (xa+xb)/2, y8=ya+0.25 x (yb-ya), where the p6 and p8 coordinates have been derived from p1 and p2, i.e. p6 (x 1, y 1), p8 (x 2, y 2), then the unknown p7 point coordinates, x7= (x6+x8)/2, y7= (y6+y8)/2, i.e. x7= (x1+x2)/2, y7= (y1+y2)/2, the remaining p9 is intermediate p8 and p10, p13 is intermediate p10 and p11, and p14 is intermediate p10 and p 12.
4 key point coordinates of the eyes are identified by using a face 68-point key point detection algorithm of an on-source, as shown in fig. 5-1, then how to automatically calculate 8 key point coordinates of the eyes is similar to nose calculation, 4 point coordinates are identical, the rest 4 coordinates are p6 located between p5 and p7, p6 and p12 are symmetrical about p5, p8 is located between p7 and p9, p8 and p10 are symmetrical about p9, thus the rest 4 point coordinates can be easily calculated, and the 8 key point coordinates of the eyes finally obtained are shown in fig. 5-2. The method uses the open-source algorithm that the few points calculate the majority points, the majority points calculate the minority points in reverse, redundant point coordinates are omitted, the method does not need to draw an original face outer frame when marking each time, a coordinate formula is well defined, some points are defined by using the existing open-source face key point detection algorithm, the points are accurate, and the rest of the coordinate relation formula is reused to obtain the face key point detection method. The points thus obtained are substantially accurate, if the individual points are not accurate, it is sufficient to drag with a mouse on the marking tool.
The key points of the face are automatically calculated by adopting the two modes, a few inaccurate key points are finely adjusted, and the last labeled key point of the face after fine adjustment can be shown in fig. 6.
By the method, the face key points with various point types can be automatically marked, compared with the traditional point-by-point marking, the marking efficiency is high, the face key points with one point type number can be detected according to the existing face key point detection algorithm of an open source, the face key points with other point types are calculated according to the detected face key points with one point type number, and then the quick marking of the face key points with various different point types is realized.
Fig. 7 is a block diagram of a face key point labeling system according to an embodiment of the present invention, as shown in fig. 7, and the face key point labeling system includes a calculating module 701, a first obtaining module 702, and a second obtaining module 703, where:
the calculating module 701 is configured to label a target frame of an outline of a face in the face picture, and calculate position coordinates of each key point of the face according to a relative relationship between the position of each key point of the face and the outline of the face; or alternatively, the process may be performed,
the first obtaining module 702 is configured to detect, according to a face key point detection algorithm of an existing point type, face key points in a face picture, and obtain coordinates of a first number of face key points in the face picture;
the second obtaining module 703 is configured to obtain coordinates of a second number of face key points on the face picture based on the detected coordinates of the first number of face key points, where the second number is unequal to the first number.
It can be understood that the face key point labeling system provided by the invention corresponds to the face key point labeling method provided by the foregoing embodiments, and the relevant technical features of the face key point labeling system may refer to the relevant technical features of the face key point labeling method, which are not described herein.
Referring to fig. 8, fig. 8 is a schematic diagram of an embodiment of an electronic device according to an embodiment of the invention. As shown in fig. 8, an embodiment of the present invention provides an electronic device 800, including a memory 810, a processor 820, and a computer program 811 stored on the memory 810 and executable on the processor 820, the processor 820 implementing the following steps when executing the computer program 811: marking a target frame of the outline of the face in the face picture, and calculating the position coordinates of each key point of the face according to the relative relation between the positions of each key point of the face and the outline of the face; or detecting the face key points in the face picture according to the face key point detection algorithm of the existing point type, and obtaining the coordinates of the first number of face key points in the face picture; and acquiring the coordinates of the key points of the second number of faces on the face picture based on the detected coordinates of the key points of the first number of faces, wherein the second number is unequal to the first number.
Referring to fig. 9, fig. 9 is a schematic diagram of an embodiment of a computer readable storage medium according to the present invention. As shown in fig. 9, the present embodiment provides a computer-readable storage medium 900 having stored thereon a computer program 911, which when executed by a processor, implements the steps of: marking a target frame of the outline of the face in the face picture, and calculating the position coordinates of each key point of the face according to the relative relation between the positions of each key point of the face and the outline of the face; or detecting the face key points in the face picture according to the face key point detection algorithm of the existing point type, and obtaining the coordinates of the first number of face key points in the face picture; and acquiring the coordinates of the key points of the second number of faces on the face picture based on the detected coordinates of the key points of the first number of faces, wherein the second number is unequal to the first number.
According to the method, the system and the storage medium for marking the key points of the human face, the problem that the marking of the key points of the human face is incorrect can be greatly reduced, the mark is not needed point by point, automatic marking is realized through a coordinate calculation method, and finally, a small number of inaccurate positions are finely adjusted, so that the working efficiency of marking personnel is greatly improved, and the labor time cost is saved.
The method can detect the face key points of one dot type number according to the existing face key point detection algorithm of the on-off source, calculate the face key points of other dot type numbers according to the detected face key points of one dot type number, and further realize the rapid labeling of the face key points of various different dot types.
In the foregoing embodiments, the descriptions of the embodiments are focused on, and for those portions of one embodiment that are not described in detail, reference may be made to the related descriptions of other embodiments.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
Claims (8)
1. The method for labeling the key points of the human face is characterized by comprising the following steps of:
detecting face key points in a face picture according to a face key point detection algorithm of the existing point type, and obtaining coordinates of a first number of face key points in the face picture;
acquiring the coordinates of the key points of the second number of faces on the face picture based on the detected coordinates of the key points of the first number of faces, wherein the second number is unequal to the first number;
the method comprises the steps of acquiring a second number of face key points on a face picture based on a first number of detected face key points, marking the face picture based on coordinates of the second number of face key points, wherein the second number is unequal to the first number, and the method comprises the following steps:
if the second number is greater than the first number, determining the first number of face key point position coordinates in the second number of face key points according to the first number of face key point position coordinates;
and calculating the position coordinates of other face key points according to the relative relation between the other face key points and the first number of face key points through a preset coordinate relation formula.
2. The method for labeling face keypoints according to claim 1, wherein the labeling on the face picture based on the second number of face keypoint coordinates further comprises:
and fine-tuning the positions of the key points of the human face with deviation of the labeling positions.
3. The method for labeling face key points according to claim 2, wherein the fine-tuning the positions of the face key points with deviations in labeling positions comprises:
dragging the positions of the key points of the human faces with deviation to the target positions on the marking tool;
or alternatively, the process may be performed,
and (3) carrying out picture frame on the face key points with the deviation, calculating the accurate positions of the face key points with the deviation through frame coordinates, and adjusting the positions of the face key points with the deviation according to the accurate positions of the face key points.
4. The method of claim 1, wherein if the second number is greater than the first number, determining the coordinates of the first number of face keypoints among the second number of face keypoints according to the coordinates of the first number of face keypoints, comprises:
dividing the first number of key point position coordinates of the face to obtain third number of key point position coordinates of different parts of the face;
and calculating the position coordinates of each other key point of any part of the face based on the position relative relation between each other key point and the third number of key points through a preset coordinate relation formula, and obtaining the fourth number of key point position coordinates of any part.
5. The method for labeling face keypoints according to claim 1, wherein the obtaining a second number of face keypoints on the face picture based on the detected first number of face keypoints, wherein the second number is unequal to the first number, comprises:
and if the second number is smaller than the first number, omitting redundant face key points in the first number of face key points, and acquiring coordinates of the second number of face key points.
6. A face key point labeling system, comprising:
the first acquisition module is used for detecting the face key points in the face picture according to the face key point detection algorithm of the existing point type to acquire the coordinates of the first number of face key points in the face picture;
the second acquisition module is used for acquiring the coordinates of the key points of the second number of faces on the face picture based on the detected coordinates of the key points of the first number of faces, and marking the face picture based on the coordinates of the key points of the second number of faces, wherein the second number is unequal to the first number;
the second obtaining module is specifically configured to:
if the second number is greater than the first number, determining the first number of face key point position coordinates in the second number of face key points according to the first number of face key point position coordinates;
and calculating the position coordinates of other face key points according to the relative relation between the other face key points and the first number of face key points through a preset coordinate relation formula.
7. An electronic device, comprising a memory and a processor, wherein the processor is configured to implement the steps of the face key point labeling method according to any one of claims 1-5 when executing a computer management class program stored in the memory.
8. A computer-readable storage medium, having stored thereon a computer-management-class program which, when executed by a processor, implements the steps of the face-key-point labeling method of any of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110904274.3A CN113723214B (en) | 2021-08-06 | 2021-08-06 | Face key point labeling method, system, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110904274.3A CN113723214B (en) | 2021-08-06 | 2021-08-06 | Face key point labeling method, system, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113723214A CN113723214A (en) | 2021-11-30 |
| CN113723214B true CN113723214B (en) | 2023-10-13 |
Family
ID=78675070
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110904274.3A Active CN113723214B (en) | 2021-08-06 | 2021-08-06 | Face key point labeling method, system, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113723214B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104866807A (en) * | 2014-02-24 | 2015-08-26 | 腾讯科技（深圳）有限公司 | Face positioning method and system |
| CN105184283A (en) * | 2015-10-16 | 2015-12-23 | 天津中科智能识别产业技术研究院有限公司 | Method and system for marking key points in human face images |
| WO2018001092A1 (en) * | 2016-06-29 | 2018-01-04 | 中兴通讯股份有限公司 | Face recognition method and apparatus |
| WO2018153294A1 (en) * | 2017-02-27 | 2018-08-30 | 腾讯科技（深圳）有限公司 | Face tracking method, storage medium, and terminal device |
| CN109034095A (en) * | 2018-08-10 | 2018-12-18 | 杭州登虹科技有限公司 | A kind of face alignment detection method, apparatus and storage medium |
| CN109034131A (en) * | 2018-09-03 | 2018-12-18 | 福州海景科技开发有限公司 | A kind of semi-automatic face key point mask method and storage medium |
| CN109543495A (en) * | 2017-09-22 | 2019-03-29 | 中国移动通信有限公司研究院 | A kind of face key point mask method, device, electronic equipment and storage medium |
| CN109558837A (en) * | 2018-11-28 | 2019-04-02 | 北京达佳互联信息技术有限公司 | Face critical point detection method, apparatus and storage medium |
| CN110956071A (en) * | 2019-06-21 | 2020-04-03 | 初速度（苏州）科技有限公司 | Eye key point labeling and detection model training method and device |
| CN112434562A (en) * | 2020-11-03 | 2021-03-02 | 浙江大华技术股份有限公司 | Method and device for detecting wearing state of mask, electronic device and storage medium |
| CN112560639A (en) * | 2020-12-11 | 2021-03-26 | 上海明略人工智能(集团)有限公司 | Face key point number conversion method, system, electronic equipment and storage medium |
| CN112613448A (en) * | 2020-12-28 | 2021-04-06 | 北京的卢深视科技有限公司 | Face data labeling method and system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109086719A (en) * | 2018-08-03 | 2018-12-25 | 北京字节跳动网络技术有限公司 | Method and apparatus for output data |
-
2021
- 2021-08-06 CN CN202110904274.3A patent/CN113723214B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104866807A (en) * | 2014-02-24 | 2015-08-26 | 腾讯科技（深圳）有限公司 | Face positioning method and system |
| CN105184283A (en) * | 2015-10-16 | 2015-12-23 | 天津中科智能识别产业技术研究院有限公司 | Method and system for marking key points in human face images |
| WO2018001092A1 (en) * | 2016-06-29 | 2018-01-04 | 中兴通讯股份有限公司 | Face recognition method and apparatus |
| WO2018153294A1 (en) * | 2017-02-27 | 2018-08-30 | 腾讯科技（深圳）有限公司 | Face tracking method, storage medium, and terminal device |
| CN109543495A (en) * | 2017-09-22 | 2019-03-29 | 中国移动通信有限公司研究院 | A kind of face key point mask method, device, electronic equipment and storage medium |
| CN109034095A (en) * | 2018-08-10 | 2018-12-18 | 杭州登虹科技有限公司 | A kind of face alignment detection method, apparatus and storage medium |
| CN109034131A (en) * | 2018-09-03 | 2018-12-18 | 福州海景科技开发有限公司 | A kind of semi-automatic face key point mask method and storage medium |
| CN109558837A (en) * | 2018-11-28 | 2019-04-02 | 北京达佳互联信息技术有限公司 | Face critical point detection method, apparatus and storage medium |
| CN110956071A (en) * | 2019-06-21 | 2020-04-03 | 初速度（苏州）科技有限公司 | Eye key point labeling and detection model training method and device |
| CN112434562A (en) * | 2020-11-03 | 2021-03-02 | 浙江大华技术股份有限公司 | Method and device for detecting wearing state of mask, electronic device and storage medium |
| CN112560639A (en) * | 2020-12-11 | 2021-03-26 | 上海明略人工智能(集团)有限公司 | Face key point number conversion method, system, electronic equipment and storage medium |
| CN112613448A (en) * | 2020-12-28 | 2021-04-06 | 北京的卢深视科技有限公司 | Face data labeling method and system |
Non-Patent Citations (2)
| Title |
|---|
| Look at Boundary: A Boundary-Aware Face Alignment Algorithm;Wayne Wu et al.;arXiv:1805.10483v1;全文 * |
| 基于深度学习的人脸关键点识别定位方法仿真;许佳;田维坚;樊养余;;计算机仿真(第06期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113723214A (en) | 2021-11-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200210702A1 (en) | Apparatus and method for image processing to calculate likelihood of image of target object detected from input image | |
| CN112330526B (en) | Training method of face conversion model, storage medium and terminal equipment | |
| US10964057B2 (en) | Information processing apparatus, method for controlling information processing apparatus, and storage medium | |
| US9164583B2 (en) | Method and apparatus for gaze point mapping | |
| CN106845508B (en) | The methods, devices and systems of release in a kind of detection image | |
| US20180075291A1 (en) | Biometrics authentication based on a normalized image of an object | |
| JPWO2011148596A1 (en) | Facial feature point position correction apparatus, face feature point position correction method, and face feature point position correction program | |
| US10977767B2 (en) | Propagation of spot healing edits from one image to multiple images | |
| KR102344373B1 (en) | Apparatus and method for generating feature maps | |
| US20190066311A1 (en) | Object tracking | |
| CN107844803B (en) | Picture comparison method and device | |
| CN114283079A (en) | Method and equipment for shooting correction based on graphic card | |
| CN115937003A (en) | Image processing method, image processing device, terminal equipment and readable storage medium | |
| CN112488067A (en) | Face pose estimation method and device, electronic equipment and storage medium | |
| CN115984796A (en) | Image annotation method and system | |
| JP2018036770A (en) | Position attitude estimation device, position attitude estimation method, and position attitude estimation program | |
| CN111798422A (en) | Checkerboard angular point identification method, device, equipment and storage medium | |
| CN113723214B (en) | Face key point labeling method, system, electronic equipment and storage medium | |
| US20210248357A1 (en) | Image processing method and image processing apparatus | |
| CN113723187A (en) | Semi-automatic labeling method and system for gesture key points | |
| CN112434581A (en) | Outdoor target color identification method and system, electronic device and storage medium | |
| US20220392107A1 (en) | Image processing apparatus, image processing method, image capturing apparatus, and non-transitory computer-readable storage medium | |
| CN111626244B (en) | Image recognition method, device, electronic equipment and medium | |
| CN114627542A (en) | Eye movement position determination method and device, storage medium and electronic equipment | |
| US20220130138A1 (en) | Training data generation apparatus, method and program |
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 |