CN106650624A - Face tracking method and device - Google Patents
Face tracking method and device Download PDFInfo
- Publication number
- CN106650624A CN106650624A CN201611021502.8A CN201611021502A CN106650624A CN 106650624 A CN106650624 A CN 106650624A CN 201611021502 A CN201611021502 A CN 201611021502A CN 106650624 A CN106650624 A CN 106650624A
- Authority
- CN
- China
- Prior art keywords
- face
- frame
- present
- mark
- newly
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Studio Devices (AREA)
Abstract
The invention discloses a face tracking method and device and relates to the field of face recognition technology. Through the face tracking method and device, real-time tracking of a face can be realized. The method comprises the steps that when it is detected that a face frame exists in a current frame of a real-time image according to a face detection result of the real-time image, whether a face frame exists in a previous frame corresponding to the current frame or not is detected; if the face frame exists in the previous frame, face feature value distance is calculated for the face frame existing in the current frame and the face frame existing in the previous frame; the face frame in the current frame is subjected to face marking according to the calculation result, so that real-time face tracking is realized. The face tracking method and device are suitable for face tracking.
Description
Technical field
The present invention relates to a kind of technical field of face recognition, more particularly to a kind of face tracking methods and device.
Background technology
With the development of science and technology, face recognition technology is more and more ripe.Face recognition technology is based on the face of people
Characteristic information carries out a kind of biological identification technology of identification.The image containing face is gathered with video camera or camera or regard
Frequency flows, and detects in the picture automatically and follow the trail of face, and then the face to detecting carries out a series of correlation techniques of face.
At present, in recognition of face, it will usually in Face datection time-frame face, face mark is carried out after recognition.However,
When face is moved, because recognition of face speed is less than Face datection speed, the face after to movement is known
Before not, it is impossible to face mark is carried out to the face after movement, now only exist face frame but there is no face mark, Jin Erwu
Method is realized carrying out real-time tracing to face.
The content of the invention
In view of this, a kind of face tracking methods and device are embodiments provided, main purpose is can be real
Now real-time tracing is carried out to face.
According to the embodiment of the present invention on one side, there is provided a kind of face tracking methods, the method includes：
According to the Face datection result of realtime graphic, when there is face frame in the present frame for detecting the realtime graphic
When, detect and whether there is in previous frame corresponding with the present frame face frame；
If existing, face frame present in the present frame and face frame present in the previous frame are carried out into face
The calculating of characteristic value distance；
According to result of calculation, face mark is carried out to the face frame in the present frame, to realize face real-time tracing.
Further, it is described to carry out face frame present in the present frame and face frame present in the previous frame
Before the calculating of face characteristic value distance, methods described also includes：
According to face frame quantity present in the previous frame, detect in the present frame with the presence or absence of newly-increased face
Frame；
It is described by face frame present in the present frame and institute if there is no newly-increased face frame in the present frame
Stating face frame present in previous frame carries out the calculating of face characteristic value distance, specifically includes：
Choose face frame present in the present frame carries out respectively face characteristic with each face frame in the previous frame
The calculating of value distance；
It is described that face mark is carried out to the face frame in the present frame according to result of calculation, specifically include：
According to face mark corresponding with the face frame of the distance minimum of face characteristic value described in the previous frame, to choosing
The present frame in face frame carry out face mark.
Specifically, it is described by face present in the present frame if there is newly-increased face frame in the present frame
Frame carries out the calculating of face characteristic value distance with face frame present in the previous frame, specifically includes：
Choose face frame present in the previous frame carries out respectively face characteristic with each face frame in the present frame
The calculating of value distance；
It is described that face mark is carried out to the face frame in the present frame according to result of calculation, specifically include：
It is pair special with face described in the present frame according to face frame corresponding face mark in the previous frame chosen
The minimum face frame of value indicative distance carries out face mark.
Further, it is described according to face frame corresponding face mark in the previous frame chosen, it is pair current with described
The minimum face frame of the distance of face characteristic value described in frame is carried out after face mark, and methods described also includes：
The face frame of face mark will not be carried out in the present frame, be defined as the people increased newly present in the present frame
Face frame；
Identified according to preset newly-increased face, face mark is carried out to the newly-increased face frame.
Further, it is described according to face frame quantity present in the previous frame, detect whether deposited in the present frame
After newly-increased face frame, methods described also includes：
If there is newly-increased face frame in the present frame, by the present frame with each face in the previous frame
The face characteristic value distance that frame is calculated respectively is defined as newly-increased face frame both greater than or equal to the face frame of predetermined threshold；
Identified according to preset newly-increased face, face mark is carried out to the newly-increased face frame.
Further, methods described also includes：
Record carries out the corresponding face picture of face frame of face mark in the present frame, to carry out face tracking
Study.
Further, it is described the face frame in the present frame is carried out after face mark according to result of calculation, it is described
Method also includes：
According to the face recognition result to the present frame, detect in the present frame with the presence or absence of wrong face mark
Face frame；
If existing, the face frame that there is wrong face mark is corrected according to the face recognition result.
According to the embodiment of the present invention on the other hand, there is provided a kind of face tracking device, the device includes：
Detector unit, for according to the Face datection result of realtime graphic, when the present frame for detecting the realtime graphic
In when there is face frame, detect and whether there is in previous frame corresponding with the present frame face frame；
Computing unit, if detecting in previous frame corresponding with the present frame there is face for the detector unit
Frame, then carry out face characteristic value distance by face frame present in the present frame with face frame present in the previous frame
Calculate；
Indexing unit, for according to the result of calculation of the computing unit, to the face frame in the present frame pedestrian being entered
Face is marked, to realize face real-time tracing.
Further, the detector unit, is additionally operable to according to face frame quantity present in the previous frame, and detection is described
With the presence or absence of newly-increased face frame in present frame；
The computing unit, if detecting there is no newly-increased face frame in present frame specifically for the detector unit,
Choose face frame present in the present frame carries out respectively face characteristic value distance with each face frame in the previous frame
Calculate；
The indexing unit, specifically for the face frame that basis is minimum with the distance of face characteristic value described in the previous frame
Corresponding face mark, to the face frame in the present frame of selection face mark is carried out.
Specifically, the computing unit, increases newly if detecting to exist in the present frame specifically for the detector unit
Face frame, then choosing face frame present in the previous frame and each face frame in the present frame, to carry out face respectively special
The calculating of value indicative distance；
The indexing unit, specifically for according to face frame corresponding face mark in the previous frame chosen, pair with
The minimum face frame of the distance of face characteristic value described in the present frame carries out face mark.
Further, described device also includes：Determining unit；
The determining unit, for the face frame by face mark is not carried out in the present frame, is defined as described current
The face frame increased newly present in frame；
The indexing unit, is additionally operable to be identified according to preset newly-increased face, and to the newly-increased face frame face mark is carried out
Note.
The determining unit, if for there is newly-increased face frame in the present frame, by the present frame with institute
State the face characteristic value distance that each face frame is calculated respectively in previous frame to be both greater than or equal to the face frame of predetermined threshold, it is determined that
For the face frame for increasing newly；
The indexing unit, is additionally operable to be identified according to preset newly-increased face, and to the newly-increased face frame face mark is carried out
Note.
Further, described device also includes：Recording unit, for recording the present frame in carried out face mark
The corresponding face picture of face frame, to carry out face tracking study.
Further, described device also includes：Corrigendum unit；
The detector unit, is additionally operable to according to the face recognition result to the present frame, and detecting in the present frame is
The no face frame that there is wrong face mark；
, there is the face of wrong face mark in the present frame if detecting for the detector unit in the corrigendum unit
Frame, then correct according to the face recognition result to the face frame that there is wrong face mark.
By above-mentioned technical proposal, technical scheme provided in an embodiment of the present invention at least has following advantages：
A kind of face tracking methods provided in an embodiment of the present invention and device, tie first according to the Face datection of realtime graphic
Really, when there is face frame in the present frame for detecting the realtime graphic, in detecting previous frame corresponding with the present frame
With the presence or absence of face frame；If existing, by face frame present in the present frame and face frame present in the previous frame
Carry out the calculating of face characteristic value distance；Then according to result of calculation, face mark is carried out to the face frame in the present frame,
To realize face real-time tracing.With at present, compared with carrying out face after recognition of face again and marking, the embodiment of the present invention passes through
Real-time detection is carried out to two frames adjacent in realtime graphic, according to the face frame pair minimum with face characteristic value distance in previous frame
The face mark answered, carries out face mark, you can with before recognition of face is carried out to present frame to the face frame in present frame,
The immediate face frame of face characteristic carries out face mark between finding twice, by the method and device of the embodiment of the present invention, can
To carry out face mark in real time to the face frame in each two field picture in realtime graphic, and then can realize to markd
Face carries out real-time tracing, and can realize separately carrying out Face datection with the process of recognition of face, can improve and outline
The speed of face, so as to improve Consumer's Experience.
Described above is only the general introduction of technical solution of the present invention, in order to better understand the technological means of the present invention,
And can be practiced according to the content of specification, and in order to allow the above and other objects of the present invention, feature and advantage can
Become apparent, below especially exemplified by the specific embodiment of the present invention.
Description of the drawings
By the detailed description for reading hereafter preferred embodiment, various other advantages and benefit is common for this area
Technical staff will be clear from understanding.Accompanying drawing is only used for illustrating the purpose of preferred embodiment, and is not considered as to the present invention
Restriction.And in whole accompanying drawing, it is denoted by the same reference numerals identical part.In the accompanying drawings：
Fig. 1 shows a kind of face tracking methods' schematic flow sheet provided in an embodiment of the present invention；
Fig. 2 shows another kind of face tracking methods' schematic flow sheet provided in an embodiment of the present invention；
Fig. 3 shows a kind of face tracking apparatus structure schematic diagram provided in an embodiment of the present invention；
Fig. 4 shows another kind of face tracking apparatus structure schematic diagram provided in an embodiment of the present invention.
Specific embodiment
The exemplary embodiment of the disclosure is more fully described below with reference to accompanying drawings.Although showing the disclosure in accompanying drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure and should not be by embodiments set forth here
Limited.On the contrary, there is provided these embodiments are able to be best understood from the disclosure, and can be by the scope of the present disclosure
Complete conveys to those skilled in the art.
A kind of face tracking methods are embodiments provided, it is possible to achieve real-time tracing is carried out to face, such as Fig. 1 institutes
Show, methods described includes：
101st, according to the Face datection result of realtime graphic, when there is face frame in the present frame for detect realtime graphic
When, detect and whether there is in previous frame corresponding with present frame face frame.
Wherein, realtime graphic can be the image for carrying out Real-time Collection during recognition of face.
For the executive agent of the embodiment of the present invention can be to be configured in face recognition device to be used to control face tracking
Device.In embodiments of the present invention, the realtime graphic to collecting carries out first real-time face detection, i.e. detection collection successively
To realtime graphic each two field picture in whether there is face characteristic, if existing in detecting the current frame image of realtime graphic
Face characteristic, can carry out frame face and obtain face frame according to position of the face characteristic in current frame image, it should be noted that
There is the feature of multiple faces in current frame image if detecting, can respectively carry out frame face, obtain multiple face frames.
Carrying out after frame face obtains face frame, corresponding face frame coordinate being recorded to a two field picture, when detecting reality
When image current frame image in when there is face frame, can be by searching whether to have recorded corresponding face in previous frame image
Frame coordinate, judges with the presence or absence of face frame in previous frame image, if having recorded corresponding face frame coordinate in previous frame image,
Determine there is face frame in previous frame image；If not recording corresponding face frame coordinate in previous frame image, it is determined that upper one
There is no face frame in two field picture.
If the 102, detecting in previous frame corresponding with present frame there is face frame, by face frame present in present frame
With the calculating that face frame present in previous frame carries out face characteristic value distance.
For example, detect that horizontal direction has two face frames in current frame image, you can be divided into Zuoren face frame and the right side
Two face frames, respectively face frame A and face frame B are there is also in face frame, and previous frame image, Zuoren face frame correspondence is taken
Face characteristic value be (21,24,1,3,12 ... ...) and take the corresponding face characteristic values of face frame A be (10,21,41,13,
2 ... ...) distance of the two characteristic values, is calculated, i.e.,So
Calculate the distance of the corresponding face characteristic value of Zuoren face frame face characteristic value corresponding with face frame B according to same way afterwards；Separately
It is outer in order to carry out face mark to right face frame, in addition it is also necessary to calculate the distance of right face frame and the face characteristic value of face frame A, and
The distance of the face characteristic value of right face frame and face frame B.
103rd, according to result of calculation, face mark is carried out to the face frame in present frame.
Wherein, the content of face mark can be personage's pet name, code name, name, ID (Identity, identity number)
Number etc..
For example, detect that vertical direction has two face frames in current frame image, you can be divided into upper and lower two faces
Frame a and b, and two face frames are there is also in previous frame image, respectively face frame A and face frame B, through face characteristic value
The calculating of distance, find in previous frame image with the immediate face frame of the face characteristic of a be face frame A, according to face frame A
Face mark, a is marked, and found according to same way closest with the face characteristic of b with previous frame image
Face frame be face frame B, according to face frame B faces mark, b is marked.It is final that the face of a is marked and face frame
The face mark of A is identical, and the face of b mark is identical with the face mark of face frame B, and then realizes chasing after face in real time
Track, such that it is able to improve the experience of user.
In embodiments of the present invention, each face frame in current frame image carries out after face mark, to record
Corresponding face frame coordinate, to prepare for the face tracking of next two field picture.
A kind of face tracking methods provided in an embodiment of the present invention, and carry out again after recognition of face at present face mark
Compare, the embodiment of the present invention by carrying out real-time detection to two frames adjacent in realtime graphic, according to special with face in previous frame
The corresponding face mark of the minimum face frame of value indicative distance, carries out face mark, you can with right to the face frame in present frame
Present frame is carried out before recognition of face, and the immediate face frame of face characteristic carries out face mark between finding twice, by this
The method of bright embodiment, can in real time carry out face mark to the face frame in each two field picture in realtime graphic, and then can be with
Realization carries out real-time tracing to markd face, and can realize separating Face datection with the process of recognition of face
Carry out, the speed for outlining face can be improved, so as to improve Consumer's Experience.
In order to preferably understand the method shown in above-mentioned Fig. 1, as refinement and extension to above-mentioned embodiment,
Another kind of face tracking methods are embodiments provided, as shown in Fig. 2 methods described includes：
201st, according to the Face datection result of realtime graphic, when there is face frame in the present frame for detect realtime graphic
When, detect and whether there is in previous frame corresponding with present frame face frame.
In embodiments of the present invention, if there is no face frame in previous frame image, illustrate from the beginning of present frame, ability someone
Face occurs, and can carry out corresponding frame face operation, and can be tentatively the corresponding face mark of its distribution, to reduce face
Frame does not have the blank time that face is marked, and then improves Consumer's Experience, after recognition of face, is carried out again more according to recognition result
Just.
If the 202, there is face frame in previous frame, according to face frame quantity present in previous frame, in detection present frame
With the presence or absence of newly-increased face frame.
If the quantity of face frame is less than the quantity of face frame in current frame image in previous frame image, illustrate now there is new
The personnel of increasing are entered in effective acquisition range of realtime graphic, and then determine in current frame image there is newly-increased face frame, are connect
Execution step 203a of getting off to step 206a, to realize in addition to the face frame to original personnel carries out face and marks, may be used also
To carry out face mark to newly-increased personnel；If in previous frame image in the quantity of face frame and current frame image face frame number
Amount is identical, illustrates that the user that current frame image is present is identical with user present in previous frame image, is same a group of people, does not exist
Newly-increased personnel, further relate to without newly-increased face frame in current frame image, and following execution step 203b is to step 204b；Such as
When the quantity of face frame is more than the quantity of face frame in current frame image in fruit previous frame image, illustrate now there are groups of people to leave
Effective acquisition range of realtime graphic, but not newly-increased personnel enter, and further relate in current frame image without newly-increased
Face frame, next same execution step 203b is to step 204b.
If there is newly-increased face frame in 203a, present frame, present in selection previous frame in face frame and present frame
Each face frame carries out respectively the calculating of face characteristic value distance.
204a, according to face frame corresponding face mark in the previous frame chosen, pair with face characteristic value in present frame away from
Face mark is carried out from minimum face frame.
For example, there are 2 faces frame A1, A2 in previous frame image, there are 3 faces frame B1, B2 in current frame image,
B3, choose A1 carries out respectively the calculating of face characteristic value distance with B1, B2, B3, and selection A2 is carried out respectively with B1, B2, B3
The calculating of face characteristic value distance, according to result of calculation, for A1, can find the immediate B1 of face characteristic, by the people of A1
Face mark is allocated to B1；For A2, the immediate B2 of face characteristic can be found, the face mark of A2 is allocated to into B2.
It should be noted that the speed in order to accelerate face mark, by the way to a face frame of present frame
After carrying out face mark, face frame marked in the current frame can be carried out rejecting process, then under in previous frame
When one face frame carries out the calculating of face characteristic value distance, without the need for the face frame that weeded out in present frame entering pedestrian again
The calculating of face characteristic value distance, reduces calculating content, and then can improve calculating speed, so as to improve the effect of face mark
Rate.
205a, the face frame of face mark will not be carried out in present frame, be defined as the face increased newly present in present frame
Frame.
For example, there are 2 faces frame A1, A2 in previous frame image, there are 3 faces frame B1, B2 in current frame image,
B3, through the calculating of face characteristic value distance, for A1, can find the immediate B1 of face characteristic, and the face of A1 is marked
It is allocated to B1；For A2, the immediate B2 of face characteristic can be found, the face mark of A2 is allocated to into B2；So there was only B3
Face mark is not carried out, the face frame that B3 is defined as being increased newly present in present frame.
In embodiments of the present invention, if there is newly-increased face frame in present frame, in order to be marked in time, improve
The detection efficiency of newly-increased face frame, can also as follows determine newly-increased face frame, will in the present frame with institute
State the face characteristic value distance that each face frame is calculated respectively in previous frame to be both greater than or equal to the face frame of predetermined threshold, it is determined that
For the face frame for increasing newly.Wherein, predetermined threshold can rule of thumb constant be preset by technical staff.
For example, in current frame image there are 3 face frames in vertical direction, be divided into the face frame of upper, middle and lower three, and upper one
There are 2 face frames in two field picture, illustrate there is a newly-increased face frame in current frame image, through face characteristic value distance
Calculating, the face for only having in nethermost face frame and previous frame image 2 face frames to calculate respectively in current frame image is special
Value indicative distance both greater than predetermined threshold, accordingly, it is determined that nethermost face frame is newly-increased face frame.
Again for example, there are 3 faces frame B1, B2, B3 in current frame image, and there are 2 face frames in previous frame image
A1, A2, illustrate there is a newly-increased face frame in current frame image, through the calculating of face characteristic value distance, B1 and A1, A2
The face characteristic value distance both greater than predetermined threshold for calculating respectively, while the face characteristic value distance that B3 is calculated respectively with A1, A2
Also predetermined threshold is both greater than, now there is the object to be determined of two newly-increased face frames, can further detect the two faces
The closer predetermined threshold of face characteristic value distance which face frame is calculated respectively with A1, A2 in previous frame image in frame, if
The face characteristic value distance that B1 is calculated respectively with A1, A2 illustrates that the face characteristic value jump of B1 is less closer to predetermined threshold, and
The face characteristic value jump of B3 is larger, and B3 is likely to newly-increased face frame, and then B3 is defined as into newly-increased face frame.
It is determined that after the face frame for increasing newly, can be identified according to preset newly-increased face, the face frame to increasing newly enter pedestrian
Face is marked, and can record corresponding face picture to be tracked study to strange face, and current frame image is being entered
After row recognition of face, can again be corrected according to face recognition result.
206a, identified according to preset newly-increased face, the face frame to increasing newly carries out face mark.
Wherein, preset newly-increased face mark can be according to the actual requirements pre-configured with by technical staff.For example, inspection is worked as
When measuring newly-increased 1 face frame, can be labeled as " new head portrait "；When newly-increased 2 face frames are detected, can be respectively labeled as
" new head portrait A ", " new head portrait B ".
It should be noted that carrying out face mark by the face frame to increasing newly, it is possible to use family is carrying out recognition of face
During understand newly-increased personnel, the experience of user can be improved.
If there is no newly-increased face frame in step 203b, the present frame arranged side by side with step 203a, choose described current
Face frame present in frame carries out respectively the calculating of face characteristic value distance with each face frame in the previous frame.
For example, detect that horizontal direction has two face frames in current frame image, is divided into Zuoren face frame and right face frame,
And two face frames, respectively face frame A and face frame B are there is also in previous frame image, the corresponding face of Zuoren face frame is taken
Characteristic value be (10,20,13,17,15 ... ...) and take the corresponding face characteristic values of face frame A be (5,4,14,22,
30 ... ...) distance of the two characteristic values, is calculated, i.e.,
Then the distance of the corresponding face characteristic value of Zuoren face frame face characteristic value corresponding with face frame B is calculated according to same way.
204b, basis face mark corresponding with the face frame of face characteristic value distance minimum in previous frame, to what is chosen
Face frame in present frame carries out face mark.
For example, it is special that the face frame chosen in current frame image calculates respectively face with each face frame in previous frame
After value indicative distance, according to the corresponding face mark of the minimum face frame of the distance in previous frame, this face frame to choosing enters
Pedestrian's face is marked so that the two face marks are identical.
In embodiments of the present invention, in order to accelerate the speed that face is marked, by the way to a people of present frame
Face frame is carried out after face mark, can will carry out rejecting process with the immediate face frame of its face characteristic in previous frame, then
When next face frame in present frame carries out the calculating of face characteristic value distance, without the need for again with weeded out in previous frame
Face frame carry out the calculating of face characteristic value distance.
For example, detect in current frame image that horizontal direction has three face frames, be divided into ' left side ', ' in ', ' right side ' this three
Individual face frame, there is also three face frames, respectively face frame A, face frame B, face frame C, through Zuoren in previous frame image
Face frame respectively with face frame A, face frame B, the face characteristic value distance of face frame C calculating, it is determined that ' left side ' face frame and face
The face characteristic value distance of frame A is minimum, and ' left side ' face frame is carried out to mark with face frame A identicals face, then can be by people
Face frame A carries out rejecting process in previous frame image, then when in ' in ' face frame or ' right side ' face frame with previous frame image
When face frame carries out the distance calculating of face characteristic value, only face characteristic value distance need to be respectively calculated i.e. with face frame B, face frame C
Can, without the need for the calculating for again face characteristic value distance being carried out with face frame A, calculating content is reduced, and then calculating speed can be improved
Degree, so as to improve the efficiency of face mark.
For the embodiment of the present invention, each the face frame in current frame image is ensured has after face mark, Ke Yiting
Only face mark is carried out to the face frame in current frame image.Further, it is described in order to meet the demand of face self-aid learning
Method can also include：Record carries out the corresponding face picture of face frame of face mark in the present frame, to carry out
Face tracking learns.By method provided in an embodiment of the present invention, multiple can be distinguished in each two field picture of realtime graphic
Face, it is possible to record corresponding face picture, to carry out face tracking study, learns the variation characteristic of face, after being
Continuous face recognition process makees reference frame.
Further, in order to ensure the accuracy that face is marked, to carry out face tracking exactly, methods described may be used also
To include：According to the face recognition result to the present frame, detect in the present frame with the presence or absence of wrong face mark
Face frame；If existing, the face frame that there is wrong face mark is corrected according to the face recognition result.
For example, the method by calculating face characteristic value distance, has all carried out face mark, subsequently in current frame image
Recognition of face is carried out to this two field picture, exist in this two field picture wrong face mark if detected according to face recognition result
During the face frame of note, can accordingly be corrected, so that accurate face mark, Jin Erke can be carried out in follow-up two field picture
To realize face tracking exactly.
Another kind of face tracking methods provided in an embodiment of the present invention, and carry out again after recognition of face at present face mark
Note is compared, the embodiment of the present invention by carrying out real-time detection to two frames adjacent in realtime graphic, according to face in previous frame
The corresponding face mark of the minimum face frame of characteristic value distance, to the face frame in present frame face mark is carried out, you can with
Present frame is carried out before recognition of face, the immediate face frame of face characteristic carries out face mark between finding twice, by this
The method of inventive embodiments, can in real time carry out face mark, Jin Erke to the face frame in each two field picture in realtime graphic
To realize carrying out real-time tracing to markd face；And face mark can be carried out to the face frame for increasing newly, can be with
Make user that newly-increased personnel are understood during recognition of face is carried out, the experience of user can be improved；And according to present frame
Face recognition result, the face frame to there is wrong face mark corrected in time, can exactly carry out face tracking.
Further, implementing as Fig. 1 methods describeds, embodiments provides a kind of face tracking dress
Put, as shown in figure 3, described device includes：Detector unit 31, computing unit 32, indexing unit 33.
The detector unit 31, can be used for the Face datection result according to realtime graphic, when detecting the real-time figure
When there is face frame in the present frame of picture, detect and whether there is in previous frame corresponding with the present frame face frame.Detection is single
Unit 31 is the main functional modules that face frame detection is carried out in this device.Specifically can be sat according to the face frame recorded in previous frame
Mark, determines and whether there is in previous frame face frame.Computing unit 32 is triggered when there is face frame in detecting previous frame to be carried out
Work.
The computing unit 32, if can be used for the detector unit 31 detects previous frame corresponding with the present frame
In there is face frame, then face frame present in the present frame and face frame present in the previous frame are carried out into face special
The calculating of value indicative distance.The computing unit 32 is main apart from what is calculated to carry out face characteristic value between face frame in this device
Functional module.
The indexing unit 33, can be used for the result of calculation according to the computing unit 32, in the present frame
Face frame carries out face mark, and to the face frame in the present frame face mark is carried out.Further, in order to realizing face
Real-time tracing.The indexing unit 33 is the major function for carrying out face mark in this device to face frame, specifically can foundation
The result of calculation of computing unit 32 is that the face frame in present frame selects the immediate face frame of face characteristic to enter from previous frame
Pedestrian's face is marked.
It should be noted that other of each functional unit involved by a kind of face tracking device provided in an embodiment of the present invention
Corresponding description, may be referred to the correspondence description in Fig. 1, will not be described here.
A kind of face tracking device provided in an embodiment of the present invention, including：Detector unit, computing unit, indexing unit, with
Carry out face mark again after recognition of face at present to compare, the embodiment of the present invention is by entering two frames adjacent in realtime graphic
Row real-time detection, according to face mark corresponding with the face frame of face characteristic value distance minimum in previous frame, in present frame
Face frame carry out face mark, you can so that before recognition of face is carried out to present frame, face characteristic most connects between finding twice
Near face frame carries out face mark, by the method for the embodiment of the present invention, can be in each two field picture in realtime graphic
Face frame carries out in real time face mark, and then can realize carrying out real-time tracing to markd face, and can be real
Now Face datection is separately carried out with the process of recognition of face, the speed for outlining face can be improved, so as to improve user's body
Test.
Further, implementing as Fig. 2 methods describeds, embodiments provides another kind of face tracking dress
Put, as shown in figure 4, described device includes：Detector unit 41, computing unit 42, indexing unit 43.
The detector unit 41, can be used for the Face datection result according to realtime graphic, when detecting the real-time figure
When there is face frame in the present frame of picture, detect and whether there is in previous frame corresponding with the present frame face frame.Detection is single
Unit 41 is the main functional modules that face frame detection is carried out in this device.Specifically can be sat according to the face frame recorded in previous frame
Mark, determines and whether there is in previous frame face frame.Computing unit 42 is triggered when there is face frame in detecting previous frame to be carried out
Work.
The computing unit 42, if can be used for the detector unit 41 detects previous frame corresponding with the present frame
In there is face frame, then face frame present in the present frame and face frame present in the previous frame are carried out into face special
The calculating of value indicative distance.The computing unit 42 is main apart from what is calculated to carry out face characteristic value between face frame in this device
Functional module.
The indexing unit 43, can be used for the result of calculation according to the computing unit 42, in the present frame
Face frame carries out face mark.Further, in order to realizing face real-time tracing.The indexing unit 43 is right in this device
Face frame carries out the major function of face mark, can be specifically the face in present frame according to the result of calculation of computing unit 42
Frame selects the immediate face frame of face characteristic to carry out face mark from previous frame.
The detector unit 41, can be also used for according to face frame quantity present in the previous frame, and detection is described to work as
With the presence or absence of newly-increased face frame in previous frame.
The computing unit 42, if specifically can be used for the detector unit 41 detect do not exist in present frame it is newly-increased
Face frame, choose face frame present in the present frame carries out respectively face characteristic value with each face frame in the previous frame
The calculating of distance.
The indexing unit 43, specifically can be used for basis with the distance minimum of face characteristic value described in the previous frame
The corresponding face mark of face frame, to the face frame in the present frame of selection face mark is carried out.
The computing unit 42, increases newly if specifically can be used for the detector unit 41 and detect to exist in the present frame
Face frame, then choosing face frame present in the previous frame and each face frame in the present frame, to carry out face respectively special
The calculating of value indicative distance.
The indexing unit 43, specifically can be used for according to the corresponding face mark of face frame in the previous frame chosen
Note, pair face frame minimum with the distance of face characteristic value described in the present frame carries out face mark.
Further, described device also includes：Determining unit 44.
The determining unit 44, can be used for the face frame by face mark is not carried out in the present frame, be defined as institute
State the face frame increased newly present in present frame.
The determining unit 44, if can be also used for there is newly-increased face frame in the present frame, will be described current
The face characteristic value distance calculated respectively with each face frame in the previous frame in frame is both greater than or equal to the people of predetermined threshold
Face frame, is defined as newly-increased face frame.Newly-increased face frame can be in time detected, and then newly-increased face frame can be entered in time
Pedestrian's face is marked.
The indexing unit 43, can be used for being identified according to preset newly-increased face, and to the newly-increased face frame pedestrian is entered
Face is marked.It should be noted that carrying out face mark by the face frame to increasing newly, it is possible to use family is carrying out recognition of face
During understand newly-increased personnel, the experience of user can be improved.
Further, described device also includes：Recording unit 45.
The recording unit 45, can be used for recording the corresponding people of face frame for carrying out face mark in the present frame
Face picture, further, in order to carry out face tracking study, can meet the demand of face self-aid learning, learn face
Variation characteristic, can make reference frame for follow-up face recognition process.
Further, described device also includes：Corrigendum unit 46.
The detector unit 41, can be used for, according to the face recognition result to the present frame, detecting the present frame
In with the presence or absence of wrong face mark face frame.
The corrigendum unit 46, exist in the present frame wrong face mark if can be used for the detector unit 41 and detects
The face frame of note, then correct according to the face recognition result to the face frame that there is wrong face mark.So as to follow-up
Two field picture in can carry out accurate face mark, and then can exactly realize face tracking.
It should be noted that its of each functional unit involved by another kind of face tracking device provided in an embodiment of the present invention
He accordingly describes, and may be referred to the correspondence description in Fig. 2, will not be described here.
Another kind of face tracking device provided in an embodiment of the present invention, including：Detector unit, computing unit, indexing unit,
Processing unit, determining unit, recording unit, corrigendum unit etc., with present compared with carrying out face after recognition of face again and marking,
The embodiment of the present invention by carrying out real-time detection to two frames adjacent in realtime graphic, according to face characteristic value in previous frame away from
From the minimum corresponding face mark of face frame, face mark is carried out to the face frame in present frame, you can with to present frame
Before carrying out recognition of face, the immediate face frame of face characteristic carries out face mark between finding twice, by present invention enforcement
The method of example, can in real time carry out face mark to the face frame in each two field picture in realtime graphic, so can realize it is right
Markd face carries out real-time tracing；And face mark can be carried out to the face frame for increasing newly, it is possible to use family exists
Newly-increased personnel are understood during carrying out recognition of face, the experience of user can be improved；And known according to the face to present frame
Other result, the face frame to there is wrong face mark is corrected in time, can exactly carry out face tracking.
In the above-described embodiments, the description to each embodiment all emphasizes particularly on different fields, without the portion described in detail in certain embodiment
Point, may refer to the associated description of other embodiment.
It is understood that said method and the correlated characteristic in device can be referred to mutually.In addition, in above-described embodiment
" first ", " second " etc. be, for distinguishing each embodiment, and not represent the quality of each embodiment.
Those skilled in the art can be understood that, for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, may be referred to the corresponding process in preceding method embodiment, will not be described here.
Provided herein algorithm and display be not inherently related to any certain computer, virtual system or miscellaneous equipment.
Various general-purpose systems can also be used together based on teaching in this.As described above, construct required by this kind of system
Structure be obvious.Additionally, the present invention is also not for any certain programmed language.It is understood that, it is possible to use it is various
Programming language realizes the content of invention described herein, and the description done to language-specific above is to disclose this
Bright preferred forms.
In specification mentioned herein, a large amount of details are illustrated.It is to be appreciated, however, that the enforcement of the present invention
Example can be put into practice in the case of without these details.In some instances, known method, structure is not been shown in detail
And technology, so as not to obscure the understanding of this description.
Similarly, it will be appreciated that in order to simplify the disclosure and help understand one or more in each inventive aspect, exist
Above in the description of the exemplary embodiment of the present invention, each feature of the present invention is grouped together into single enforcement sometimes
In example, figure or descriptions thereof.However, the method for the disclosure should be construed to reflect following intention：I.e. required guarantor
The more features of feature that the application claims ratio of shield is expressly recited in each claim.More precisely, such as following
Claims reflect as, inventive aspect is all features less than single embodiment disclosed above.Therefore,
Thus the claims for following specific embodiment are expressly incorporated in the specific embodiment, wherein each claim itself
All as the separate embodiments of the present invention.
Those skilled in the art are appreciated that can be carried out adaptively to the module in the equipment in embodiment
Change and they are arranged in one or more equipment different from the embodiment.Can be the module or list in embodiment
Unit or component are combined into a module or unit or component, and can be divided in addition multiple submodule or subelement or
Sub-component.In addition at least some in such feature and/or process or unit is excluded each other, can adopt any
Combine to all features disclosed in this specification (including adjoint claim, summary and accompanying drawing) and so disclosed
Where all processes or unit of method or equipment are combined.Unless expressly stated otherwise, this specification is (including adjoint power
Profit is required, summary and accompanying drawing) disclosed in each feature can it is identical by offers, be equal to or the alternative features of similar purpose carry out generation
Replace.
Although additionally, it will be appreciated by those of skill in the art that some embodiments described herein include other embodiments
In included some features rather than further feature, but the combination of the feature of different embodiments means in of the invention
Within the scope of and form different embodiments.For example, in the following claims, embodiment required for protection appoint
One of meaning can in any combination mode using.
The present invention all parts embodiment can be realized with hardware, or with one or more processor operation
Software module realize, or with combinations thereof realization.It will be understood by those of skill in the art that can use in practice
Microprocessor or digital signal processor (DSP) are realizing a kind of face tracking methods according to embodiments of the present invention and device
In some or all parts some or all functions.The present invention is also implemented as described herein for performing
Some or all equipment of method or program of device (for example, computer program and computer program).So
Realization the present invention program can store on a computer-readable medium, or can have one or more signal shape
Formula.Such signal can be downloaded from internet website and obtained, or be provided on carrier signal, or with any other shape
Formula is provided.
It should be noted that above-described embodiment the present invention will be described rather than limits the invention, and ability
Field technique personnel can design without departing from the scope of the appended claims alternative embodiment.In the claims,
Any reference symbol between bracket should not be configured to limitations on claims.Word "comprising" is not excluded the presence of not
Element listed in the claims or step.Word "a" or "an" before element does not exclude the presence of multiple such
Element.The present invention can come real by means of the hardware for including some different elements and by means of properly programmed computer
It is existing.If in the unit claim for listing equipment for drying, several in these devices can be by same hardware branch
To embody.The use of word first, second, and third does not indicate that any order.These words can be explained and be run after fame
Claim.
Claims (10)
1. a kind of face tracking methods, it is characterised in that include：
According to the Face datection result of realtime graphic, when there is face frame in the present frame for detecting the realtime graphic, inspection
Survey in previous frame corresponding with the present frame and whether there is face frame；
If existing, face frame present in the present frame and face frame present in the previous frame are carried out into face characteristic
The calculating of value distance；
According to result of calculation, face mark is carried out to the face frame in the present frame, to realize face real-time tracing.
2. face tracking methods according to claim 1, it is characterised in that described by face present in the present frame
Frame is carried out before the calculating of face characteristic value distance with face frame present in the previous frame, and methods described also includes：
According to face frame quantity present in the previous frame, detect in the present frame with the presence or absence of newly-increased face frame；
If there is no newly-increased face frame in the present frame, it is described by face frame present in the present frame with it is described on
Face frame present in one frame carries out the calculating of face characteristic value distance, specifically includes：
Choose face frame present in the present frame and each face frame in the previous frame carry out respectively face characteristic value away from
From calculating；
It is described that face mark is carried out to the face frame in the present frame according to result of calculation, specifically include：
According to face mark corresponding with the face frame of the distance minimum of face characteristic value described in the previous frame, to the institute for choosing
The face frame stated in present frame carries out face mark.
3. face tracking methods according to claim 2, it is characterised in that if there is newly-increased face in the present frame
Frame, then it is described by face frame present in the present frame and face frame present in the previous frame carry out face characteristic value away from
From calculating, specifically include：
Choose face frame present in the previous frame and each face frame in the present frame carry out respectively face characteristic value away from
From calculating；
It is described that face mark is carried out to the face frame in the present frame according to result of calculation, specifically include：
According to face frame corresponding face mark in the previous frame chosen, pair with face characteristic value described in the present frame
The minimum face frame of distance carries out face mark.
4. face tracking methods according to claim 3, it is characterised in that described according to people in the previous frame chosen
The corresponding face mark of face frame, pair face frame minimum with the distance of face characteristic value described in the present frame carries out face mark
Afterwards, methods described also includes：
The face frame of face mark will not be carried out in the present frame, be defined as the face increased newly present in the present frame
Frame；
Identified according to preset newly-increased face, face mark is carried out to the newly-increased face frame.
5. face tracking methods according to claim 2, it is characterised in that described according to people present in the previous frame
Face frame quantity, detects in the present frame and whether there is after newly-increased face frame, methods described also includes：
If there is newly-increased face frame in the present frame, will divide with each face frame in the previous frame in the present frame
The face characteristic value distance not calculated is defined as newly-increased face frame both greater than or equal to the face frame of predetermined threshold；
Identified according to preset newly-increased face, face mark is carried out to the newly-increased face frame.
6. face tracking methods according to any one of claim 1 to 5, it is characterised in that methods described also includes：
Record carries out the corresponding face picture of face frame of face mark in the present frame, to carry out face tracking
Practise.
7. face tracking methods according to claim 1, it is characterised in that described according to result of calculation, to described current
Face frame in frame is carried out after face mark, and methods described also includes：
According to the face recognition result to the present frame, detect in the present frame with the presence or absence of the face of wrong face mark
Frame；
If existing, the face frame that there is wrong face mark is corrected according to the face recognition result.
8. a kind of face tracking device, it is characterised in that include：
Detector unit, for according to the Face datection result of realtime graphic, when depositing in the present frame for detecting the realtime graphic
In face frame, detect and whether there is in previous frame corresponding with the present frame face frame；
Computing unit, if detecting in previous frame corresponding with the present frame there is face frame for the detector unit,
Face frame present in the present frame and face frame present in the previous frame are carried out into the calculating of face characteristic value distance；
Indexing unit, for according to the result of calculation of the computing unit, to the face frame in the present frame face mark being carried out
Note, to realize face real-time tracing.
9. face tracking device according to claim 8, it is characterised in that
Whether the detector unit, is additionally operable to according to face frame quantity present in the previous frame, detect in the present frame
There is newly-increased face frame；
, there is no newly-increased face frame in the computing unit, choose if detecting specifically for the detector unit in present frame
Face frame present in the present frame carries out respectively the calculating of face characteristic value distance with each face frame in the previous frame；
The indexing unit, it is corresponding specifically for the basis face frame minimum with the distance of face characteristic value described in the previous frame
Face mark, to choose the present frame in face frame carry out face mark.
10. face tracking device according to claim 9, it is characterised in that
, there is newly-increased face frame in the present frame if detecting specifically for the detector unit in the computing unit,
Choose face frame present in the previous frame carries out respectively face characteristic value distance with each face frame in the present frame
Calculate；
The indexing unit, specifically for according to face frame corresponding face mark in the previous frame chosen, pair with it is described
The minimum face frame of the distance of face characteristic value described in present frame carries out face mark.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611021502.8A CN106650624A (en) | 2016-11-15 | 2016-11-15 | Face tracking method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611021502.8A CN106650624A (en) | 2016-11-15 | 2016-11-15 | Face tracking method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106650624A true CN106650624A (en) | 2017-05-10 |
Family
ID=58808196
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611021502.8A Pending CN106650624A (en) | 2016-11-15 | 2016-11-15 | Face tracking method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106650624A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106778482A (en) * | 2016-11-15 | 2017-05-31 | 东软集团股份有限公司 | Face tracking methods and device |
| CN108446585A (en) * | 2018-01-31 | 2018-08-24 | 深圳市阿西莫夫科技有限公司 | Method for tracking target, device, computer equipment and storage medium |
| CN109472278A (en) * | 2017-09-08 | 2019-03-15 | 上海银晨智能识别科技有限公司 | Acquisition method, device, computer-readable medium and the system of human face data |
| CN110363790A (en) * | 2018-04-11 | 2019-10-22 | 北京京东尚科信息技术有限公司 | Target tracking method, device and computer readable storage medium |
| WO2020052062A1 (en) * | 2018-09-14 | 2020-03-19 | 北京字节跳动网络技术有限公司 | Detection method and device |
| CN111613227A (en) * | 2020-03-31 | 2020-09-01 | 平安科技（深圳）有限公司 | Voiceprint data generation method and device, computer device and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1924894A (en) * | 2006-09-27 | 2007-03-07 | 北京中星微电子有限公司 | Multiple attitude human face detection and track system and method |
| CN101567043A (en) * | 2009-05-31 | 2009-10-28 | 中山大学 | Face tracking method based on classification and identification |
| CN105512627A (en) * | 2015-12-03 | 2016-04-20 | 腾讯科技（深圳）有限公司 | Key point positioning method and terminal |
-
2016
- 2016-11-15 CN CN201611021502.8A patent/CN106650624A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1924894A (en) * | 2006-09-27 | 2007-03-07 | 北京中星微电子有限公司 | Multiple attitude human face detection and track system and method |
| CN101567043A (en) * | 2009-05-31 | 2009-10-28 | 中山大学 | Face tracking method based on classification and identification |
| CN105512627A (en) * | 2015-12-03 | 2016-04-20 | 腾讯科技（深圳）有限公司 | Key point positioning method and terminal |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106778482A (en) * | 2016-11-15 | 2017-05-31 | 东软集团股份有限公司 | Face tracking methods and device |
| CN109472278A (en) * | 2017-09-08 | 2019-03-15 | 上海银晨智能识别科技有限公司 | Acquisition method, device, computer-readable medium and the system of human face data |
| CN108446585A (en) * | 2018-01-31 | 2018-08-24 | 深圳市阿西莫夫科技有限公司 | Method for tracking target, device, computer equipment and storage medium |
| CN108446585B (en) * | 2018-01-31 | 2020-10-30 | 深圳市阿西莫夫科技有限公司 | Target tracking method and device, computer equipment and storage medium |
| CN110363790A (en) * | 2018-04-11 | 2019-10-22 | 北京京东尚科信息技术有限公司 | Target tracking method, device and computer readable storage medium |
| CN110363790B (en) * | 2018-04-11 | 2024-06-14 | 北京京东尚科信息技术有限公司 | Target tracking method, apparatus and computer readable storage medium |
| WO2020052062A1 (en) * | 2018-09-14 | 2020-03-19 | 北京字节跳动网络技术有限公司 | Detection method and device |
| CN111613227A (en) * | 2020-03-31 | 2020-09-01 | 平安科技（深圳）有限公司 | Voiceprint data generation method and device, computer device and storage medium |
| WO2021196390A1 (en) * | 2020-03-31 | 2021-10-07 | 平安科技（深圳）有限公司 | Voiceprint data generation method and device, and computer device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106778482A (en) | Face tracking methods and device | |
| CN106650624A (en) | Face tracking method and device | |
| CN104615986B (en) | The method that pedestrian detection is carried out to the video image of scene changes using multi-detector | |
| CN111983935B (en) | Performance evaluation method and device | |
| CN105654507B (en) | A kind of vehicle overall dimension measurement method based on the tracking of image behavioral characteristics | |
| CN108388879A (en) | Mesh object detection method, device and storage medium | |
| CN103324937A (en) | Method and device for labeling targets | |
| CN105096307B (en) | The method of detection object in paired stereo-picture | |
| FI3452959T3 (en) | Model construction in a neural network for object detection | |
| CN104952073B (en) | Scene Incision method based on deep learning | |
| CN104346802B (en) | A kind of personnel leave the post monitoring method and equipment | |
| CN106897681A (en) | A kind of remote sensing images comparative analysis method and system | |
| KR101645959B1 (en) | The Apparatus and Method for Tracking Objects Based on Multiple Overhead Cameras and a Site Map | |
| CN110288032A (en) | A kind of vehicle driving trace type detection method and device | |
| CN106909886A (en) | A kind of high accuracy method for traffic sign detection and system based on deep learning | |
| CN107274679A (en) | Vehicle identification method, device, equipment and computer-readable recording medium | |
| CN108089695A (en) | A kind of method and apparatus for controlling movable equipment | |
| CN108399366A (en) | It is a kind of based on the remote sensing images scene classification extracting method classified pixel-by-pixel | |
| CN110502962A (en) | Mesh object detection method, device, equipment and medium in video flowing | |
| CN104751487B (en) | A kind of moving target detecting method based on the plane discoloration frames of colored RGB tri- difference | |
| CN110111370A (en) | A kind of vision object tracking methods based on TLD and the multiple dimensioned space-time characteristic of depth | |
| CN108234821A (en) | Detect the methods, devices and systems of the action in video | |
| CN108764264A (en) | Smog detection method, smoke detection system and computer installation | |
| CN110288629A (en) | Target detection automatic marking method and device based on moving Object Detection | |
| WO2020050973A3 (en) | Neural network-based systems and computer-implemented methods for identifying and/or evaluating one or more food items present in a visual input |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170510 |