CN107368820A - One kind becomes more meticulous gesture identification method, device and equipment - Google Patents
One kind becomes more meticulous gesture identification method, device and equipment Download PDFInfo
- Publication number
- CN107368820A CN107368820A CN201710656434.0A CN201710656434A CN107368820A CN 107368820 A CN107368820 A CN 107368820A CN 201710656434 A CN201710656434 A CN 201710656434A CN 107368820 A CN107368820 A CN 107368820A
- Authority
- CN
- China
- Prior art keywords
- feature
- displacement characteristic
- gesture
- cluster
- meticulous
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
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
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/28—Recognition of hand or arm movements, e.g. recognition of deaf sign language
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Social Psychology (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Psychiatry (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Image Analysis (AREA)
- User Interface Of Digital Computer (AREA)
Abstract
One kind gesture identification method that becomes more meticulous includes：The local feature of hand joint position is extracted, the local feature includes relative seat feature and/or sequential displacement characteristic；Cluster calculation is carried out according to the relative seat feature and/or sequential displacement characteristic, obtains the relative seat feature and/or the cluster feature corresponding to sequential displacement characteristic；According to the cluster feature and the other corresponding relation of gesture class, cluster feature model training is carried out, gesture identification is carried out according to the cluster feature model after training.It can realize that the dynamic gesture of Length discrepancy to the transformation of regular length feature, contributes to grader to carry out measuring similarity to gesture-type, and be advantageously implemented the judgement of finger refinement motion process and realize the detection of Large Amplitude Motion.
Description
Technical field
The invention belongs to gesture identification field, more particularly to one kind becomes more meticulous gesture identification method, device and equipment.
Background technology
In the man-machine interactive systems such as intelligent television, wearable mobile terminal, personal computer or virtual reality device, warp
Interactively entering to online gesture identification can often be used.
According to the difference of data acquisition modes, the method for gesture identification can be divided into based on Wearable both at home and abroad at present
With two kinds of view-based access control model.Wherein：
Gesture identification method based on Wearable mainly obtains gesture using sensors such as accelerometer, gyroscopes
In the motion track information of three dimensions, based on the advantages of Wearable progress gesture identification it is that multiple sensings can be equipped with
Device is to obtain the relative position information of accurate hand joint and space motion path, and the accuracy rate of identification is higher, but this method
Need to wear complicated device, such as data glove, position tracker, wearing is relatively complicated, to man-machine interactive system
Naturality brings certain influence.
The gesture identification method of view-based access control model can solve natural sex chromosome mosaicism during man-machine interaction well, be taken the photograph by visible ray
As the view data of head acquisition hand region, the segmentation, feature extraction, assorting process of hand target area are then carried out.But
The gesture identification method of existing view-based access control model can only handle static gesture (the gesture numeral knowledge of such as single image of single type
Not) or dynamic gesture (palm slides up and down page turning), in the gesture identification problem of processing Length discrepancy sequence, existing gesture is known
Other method carries out the similarity measurement of gesture motion track using dynamic time warping algorithm more, and the algorithm can solve significantly
The difference of the lower hand movement locus of motion, but computation complexity is high, can not realize become more meticulous, the knowledge of multifarious finger motion
Not.
The content of the invention
In view of this, become more meticulous gesture identification method, device and equipment the embodiments of the invention provide one kind, it is existing to solve
Have the gesture identification method in technology, because computation complexity is high, can not realize become more meticulous, multifarious finger motion identification
Problem.
The first aspect of the embodiment of the present invention provides one kind and become more meticulous gesture identification method, the gesture identification that becomes more meticulous
Method includes：
The local feature of hand joint position is extracted, the local feature includes relative seat feature and/or sequential displacement
Feature；
Cluster calculation is carried out according to the relative seat feature and/or sequential displacement characteristic, it is special to obtain the relative position
Cluster feature corresponding to sign and/or sequential displacement characteristic；
According to the cluster feature and the other corresponding relation of gesture class, cluster feature model training is carried out, after training
Cluster feature model carry out gesture identification.
With reference in a first aspect, in the first possible implementation of first aspect, the part of hand joint position is extracted
The step of relative seat feature in feature, includes：
T frame dynamic gesture images are obtained, it is determined that the position of the node per the hand in frame dynamic gesture image；
Position according to the root node that the node of hand includes with corresponding child node, be calculated child node relative to
The relative seat feature of root node.
With reference to the first possible implementation of first aspect, in second of possible implementation of first aspect, institute
The position according to the root node that the node of hand includes and corresponding child node is stated, child node is calculated relative to root node
Relative seat feature the step of include：
Obtain the relative seat feature of the t ' frames in T frame dynamic gesture imagesWherein：
The position of root node is represented,
The position of expression child node corresponding with the root node, u ∈ i | 1≤i≤N }, h=1 expression left hands, the h=2 expression right hands, 1
≤ t '≤T, N are the number of node.
With reference in a first aspect, in the third possible implementation of first aspect, the part of hand joint position is extracted
The step of sequential displacement characteristic in feature, includes：
T frame dynamic gesture images are obtained, it is determined that per the displacement characteristic reference point in frame dynamic gesture image；
The position of displacement characteristic reference point in the adjacent image of every two frame, determine that the dynamic gesture image institute is right
The sequential displacement characteristic answered.
With reference to the third possible implementation of first aspect, in the 4th kind of possible implementation of first aspect, institute
State the position of the displacement characteristic reference point in the adjacent image of every two frame, determine corresponding to the dynamic gesture image when
The step of sequence displacement characteristic, includes：
Obtain the sequential displacement characteristic of t " framesWherein, For the displacement characteristic reference point of t " frames,Referred to for the displacement characteristic of the frames of t " -1
Point, 1<T "≤T, v ∈ i | and 1≤i≤N }, M is displacement characteristic reference point number and 1≤v≤M, h=1 represent left hand, h=2 tables
Show the right hand, N is the number of node.
It is described special according to the relative position with reference in a first aspect, in the 5th kind of possible implementation of first aspect
Sign and/or sequential displacement characteristic carry out cluster calculation, obtain corresponding to the relative seat feature and/or sequential displacement characteristic
The step of cluster feature, includes：
The relative seat feature and/or sequential displacement characteristic are expressed as to the set of unified local feature；
The cluster that predetermined number is chosen in the set of the unified local feature clusters set；
Set is clustered to each cluster and carries out transformation calculations, obtains the relative seat feature and/or sequential displacement characteristic pair
The cluster feature answered.
The second aspect of the embodiment of the present invention provides one kind and become more meticulous gesture identifying device, the gesture identification that becomes more meticulous
Device includes：
Local shape factor unit, for extracting the local feature of hand joint position, the local feature includes relative
Position feature and/or sequential displacement characteristic；
Cluster feature computing unit, based on carrying out cluster according to the relative seat feature and/or sequential displacement characteristic
Calculate, obtain the relative seat feature and/or the cluster feature corresponding to sequential displacement characteristic；
Recognition unit is trained, for according to the cluster feature and the other corresponding relation of gesture class, carrying out cluster feature mould
Type training, gesture identification is carried out according to the cluster feature model after training.
With reference to second aspect, in the first possible implementation of second aspect, the local shape factor unit bag
Include：
First image obtains subelement, for obtaining T frame dynamic gesture images, it is determined that per the hand in frame dynamic gesture image
The position of the node in portion；
First computation subunit, for the position according to the root node that the node of hand includes with corresponding child node,
Relative seat feature of the child node relative to root node is calculated；
And/or
Second image obtains subelement, for obtaining T frame dynamic gesture images, it is determined that per the position in frame dynamic gesture image
Move feature reference point；
Second computation subunit, for the position of the displacement characteristic reference point in the adjacent image of every two frame, it is determined that
Sequential displacement characteristic corresponding to the dynamic gesture image.
The third aspect of the embodiment of the present invention provides one kind and become more meticulous gesture identification equipment, including：Memory, processor
And the computer program that can be run in the memory and on the processor is stored in, meter described in the computing device
The step of gesture identification method is become more meticulous as described in any one of first aspect is realized during calculation machine program.
The fourth aspect of the embodiment of the present invention provides a kind of computer-readable recording medium, the computer-readable storage
Media storage has computer program, is realized when the computer program is executed by processor fine as described in any one of first aspect
The step of changing gesture identification method.
Existing beneficial effect is the embodiment of the present invention compared with prior art：By extracting the local special of hand joint
Sign, the cluster feature for extracting local feature are trained, it is possible to achieve the dynamic gesture of Length discrepancy to regular length feature turns
Become, contribute to grader to gesture-type carry out measuring similarity, and using relative seat feature and/or sequential displacement characteristic come
Hand joint position is described, the judgement of finger refinement motion process is advantageously implemented and/or realizes the detection of Large Amplitude Motion.
Brief description of the drawings
Technical scheme in order to illustrate the embodiments of the present invention more clearly, below will be to embodiment or description of the prior art
In the required accompanying drawing used be briefly described, it should be apparent that, drawings in the following description be only the present invention some
Embodiment, for those of ordinary skill in the art, without having to pay creative labor, can also be according to these
Accompanying drawing obtains other accompanying drawings.
Fig. 1 is the implementation process schematic diagram of the gesture identification method provided in an embodiment of the present invention that becomes more meticulous；
Fig. 2 is dynamic gesture schematic diagram provided in an embodiment of the present invention；
Fig. 3 is hand joint position view provided in an embodiment of the present invention；
Fig. 4 is root node provided in an embodiment of the present invention and the corresponding relation schematic diagram of child node；
Fig. 5 is the displacement characteristic reference mode schematic diagram that the present invention implements to provide；
Fig. 6 is that offer progress cluster calculation of the embodiment of the present invention obtains the implementation process schematic diagram of cluster feature；
Fig. 7 is the structural representation for the dynamic gesture operation device that this hair inventive embodiments provide；
Fig. 8 is the schematic diagram of the gesture identification equipment provided in an embodiment of the present invention that becomes more meticulous.
Embodiment
In describing below, in order to illustrate rather than in order to limit, it is proposed that such as tool of particular system structure, technology etc
Body details, thoroughly to understand the embodiment of the present invention.However, it will be clear to one skilled in the art that there is no these specific
The present invention can also be realized in the other embodiments of details.In other situations, omit to well-known system, device, electricity
Road and the detailed description of method, in case unnecessary details hinders description of the invention.
In order to illustrate technical solutions according to the invention, illustrated below by specific embodiment.
As shown in figure 1, the gesture identification method that become more meticulous described in the embodiment of the present invention, including：
In step S101, the local feature of extraction hand joint position, the local feature includes relative seat feature
And/or sequential displacement characteristic.
Specifically, the extraction of the local feature, the multiple image for including user's hand based on acquisition, as shown in Figure 2
For the schematic diagram of the T two field pictures of " OK " gesture, from the 1st frame to T frames, including multiple images, and the change of frame over time,
The change of position can occur for the gesture in image, wherein, the change of position include the relative position of finger change and palm it is whole
The change of body position, can be described respectively by relative seat feature and timing position feature finger relative position change with
And the change of palm integral position.
In order to obtain relative seat feature and/or timing position feature, it is necessary to extract the node of hand in advance, pass through node
Position change, to react relative seat feature and the timing position feature.As optionally a kind of embodiment party of the application
Formula, as depicted in figs. 1 and 2, the node of the hand extracted include artis position, finger tips position and the palm of hand
Center, including 2 artis of thumb, 1 finger tips node, other four fingers include 3 artis
With 1 endpoint node, wrist portion includes an artis, and one node of palm center location, altogether including 22 features
Reference point, to distinguish left hand and the right hand, the right hand and left hand are represented respectively with R1, R2 ... R22, and L1, L2 ... L22 respectively
Node.
Wherein, the step of extracting the relative seat feature in the local feature of hand joint position includes：
T frame dynamic gesture images are obtained, it is determined that the position of the node per the hand in frame dynamic gesture image；
Position according to the root node that the node of hand includes with corresponding child node, be calculated child node relative to
The relative seat feature of root node.
Dynamic gesture G={ G for including T (T is for natural number and more than or equal to 2) two field picturet| 1≤t≤T }, such as Fig. 2
The shown right hand " OK " gesture, each frame gesture data contain the node of hand under world coordinate system estimated by depth camera
Position：
Here h=1 represents left hand, and h=2 represents right
Hand；I ∈ { 1,2,3 ..., N } represent i-th of node of corresponding hand, as shown in Fig. 2 each hand includes 22 nodes, then N
Value can be 22.
The position of the node of given the t ' framesHere 1≤t '≤T.As shown in figure 4, selection and Fig. 3
N-2 corresponding to the node (N is 22 in Fig. 3) group root nodeWith corresponding child nodePoint
Not Ji Suan child node relative to the root node relative position vectorI.e.
The then relative seat feature of the t ' framesIt can be expressed as：
Wherein, the selection of root node and child node, as shown in figure 3, being the position in the joint according to where node, phase is chosen
Every an artis two nodes respectively as root node and child node.
And include in this step, the step of sequential displacement characteristic in the local feature of extraction hand joint position：
T frame dynamic gesture images are obtained, it is determined that per the displacement characteristic reference point in frame dynamic gesture image；
The position of displacement characteristic reference point in the adjacent image of every two frame, determine that the dynamic gesture image institute is right
The sequential displacement characteristic answered.
The selection of the displacement characteristic reference point, can select the node in palm part, and so choosing more to be defined
The mass motion amount of true determination hand, avoid selecting the node of finger position from causing mistake to mass motion direction and size
Difference.As shown in figure 5, can choose the individual nodes of wherein M (in Fig. 2 M be 7) is used as displacement characteristic reference point, for give t " frame
Artis positionHere 1<T "≤T, N are the number of node, and M is the displacement characteristic reference point chosen
Number, selected displacement characteristic reference point are：Calculate present frameAnd former frameSequential
Motion vectorI.e.
The then sequential displacement characteristic of t " framesIt can be expressed as：
In step s 102, cluster calculation is carried out according to the relative seat feature and/or sequential displacement characteristic, obtains institute
State relative seat feature and/or the cluster feature corresponding to sequential displacement characteristic.
The process clustered to the local feature of extraction, can with as shown in fig. 6, including：
In step s 601, the relative seat feature and/or sequential displacement characteristic are expressed as unified local feature
Set.
For the relative seat feature of step S101 extractionsWith sequential displacement characteristicIt can be expressed as unified local
Feature representation form：Here can set：1≤s≤4,1<τ≤T.For only including relative position
Feature is put, or only includes sequential displacement characteristic to use corresponding unified local feature expression-form, such as
Or
In step S602, the cluster that predetermined number is chosen in the set of the unified local feature clusters set.
In the present invention, can be to s category feature setIt is individual using k-means++ algorithm picks K
Initialize cluster centre point { μs,k| 1≤k≤K }, and by the use of error sum of squares as clustering measure foundation, obtain K cluster cluster sets
Close Cs={ Cs,k| 1≤k≤K }, i.e.,：
WhereinFor, per cluster cluster centre point, wherein K is the natural number more than 2 after renewal.
In step S603, to each cluster cluster set carry out transformation calculations, obtain the relative seat feature and/or when
Cluster feature corresponding to sequence displacement characteristic.
C is clustered to every clusters,k, principal component analysis (Principal Component Analysis, PCA) operation is carried out,
And retain all the components, the cluster C ' after being converteds,k。
Calculate cluster C 's,kWith cluster centre point μ 's,kPoor accumulation and, obtain compact form cluster feature expression
νk：
The cluster feature of K cluster centre point is combined, forms the compact form expression of s category features：
Vs={ νs,k|1≤k≤K}。
Repeat step S602-S603, can be by local feature setBe expressed as and when
The character representation of regular length unrelated sequence τ, i.e.,：
V={ Vs|1≤s≤4}。
In step s 103, according to the cluster feature and the other corresponding relation of gesture class, cluster feature model instruction is carried out
Practice, gesture identification is carried out according to the cluster feature model after training.
After the character representation of the regular length unrelated with sequential τ is generated, SVMs or other instructions can be passed through
Practice model, carry out model training and test, after the model trained, gesture can be sentenced by the model trained
Disconnected and identification.
It should be understood that the size of the sequence number of each step is not meant to the priority of execution sequence, each process in above-described embodiment
Execution sequence should determine that the implementation process without tackling the embodiment of the present invention forms any limit with its function and internal logic
It is fixed.
Fig. 7 is a kind of structural representation for the gesture identifying device that becomes more meticulous provided in an embodiment of the present invention, as shown in fig. 7,
The gesture identifying device that becomes more meticulous includes：
Local shape factor unit 701, for extracting the local feature of hand joint position, the local feature includes phase
To position feature and/or sequential displacement characteristic；
Cluster feature computing unit 702, for being clustered according to the relative seat feature and/or sequential displacement characteristic
Calculate, obtain the relative seat feature and/or the cluster feature corresponding to sequential displacement characteristic；
Recognition unit 703 is trained, for according to the cluster feature and the other corresponding relation of gesture class, carrying out cluster feature
Model training, gesture identification is carried out according to the cluster feature model after training.
Preferably, the local shape factor unit includes：
First image obtains subelement, for obtaining T frame dynamic gesture images, it is determined that per the hand in frame dynamic gesture image
The position of the node in portion；
First computation subunit, for the position according to the root node that the node of hand includes with corresponding child node,
Relative seat feature of the child node relative to root node is calculated；
And/or
Second image obtains subelement, for obtaining T frame dynamic gesture images, it is determined that per the position in frame dynamic gesture image
Move feature reference point；
Second computation subunit, for the position of the displacement characteristic reference point in the adjacent image of every two frame, it is determined that
Sequential displacement characteristic corresponding to the dynamic gesture image.
Fig. 8 is the schematic diagram for the gesture identification equipment that becomes more meticulous that one embodiment of the invention provides.As shown in figure 8, the implementation
The gesture identification equipment 8 that becomes more meticulous of example includes：Processor 80, memory 81 and it is stored in the memory 81 and can be in institute
The computer program 82 run on processor 80 is stated, such as the gesture identification program that becomes more meticulous.The processor 80 performs the meter
The step in above-mentioned each gesture identification method embodiment that becomes more meticulous, such as the step 101 shown in Fig. 1 are realized during calculation machine program 82
To 103.Or the processor 80 realizes each module in above-mentioned each device embodiment/mono- when performing the computer program 82
The function of member, such as the function of module 701 to 703 shown in Fig. 7.
Exemplary, the computer program 82 can be divided into one or more module/units, it is one or
Multiple module/units are stored in the memory 81, and are performed by the processor 80, to complete the present invention.Described one
Individual or multiple module/units can be the series of computation machine programmed instruction section that can complete specific function, and the instruction segment is used for
Implementation procedure of the computer program 82 in the gesture identification equipment 8 that becomes more meticulous is described.For example, the computer program
82 can be divided into local shape factor unit, cluster feature computing unit and training recognition unit, each unit concrete function
It is as follows：
Local shape factor unit, for extracting the local feature of hand joint position, the local feature includes relative
Position feature and/or sequential displacement characteristic；
Cluster feature computing unit, based on carrying out cluster according to the relative seat feature and/or sequential displacement characteristic
Calculate, obtain the relative seat feature and/or the cluster feature corresponding to sequential displacement characteristic；
Recognition unit is trained, for according to the cluster feature and the other corresponding relation of gesture class, carrying out cluster feature mould
Type training, gesture identification is carried out according to the cluster feature model after training.
The gesture identification equipment 8 that becomes more meticulous can be desktop PC, notebook, palm PC and cloud server
Deng computing device.The gesture identification equipment that becomes more meticulous may include, but be not limited only to, processor 80, memory 81.This area skill
Art personnel are appreciated that Fig. 8 is only the example of gesture identification equipment 8 of becoming more meticulous, and do not form and the gesture identification that becomes more meticulous is set
It standby 8 restriction, can include than illustrating more or less parts, either combine some parts or different parts, such as
The gesture identification equipment that becomes more meticulous can also include input-output equipment, network access equipment, bus etc..
Alleged processor 80 can be CPU (Central Processing Unit, CPU), can also be
Other general processors, digital signal processor (Digital Signal Processor, DSP), application specific integrated circuit
(Application Specific Integrated Circuit, ASIC), ready-made programmable gate array (Field-
Programmable Gate Array, FPGA) either other PLDs, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor can also be any conventional processor
Deng.
The memory 81 can be the internal storage unit of the gesture identification equipment 8 that becomes more meticulous, such as the hand that becomes more meticulous
The hard disk or internal memory of gesture identification equipment 8.The memory 81 can also be the external storage of the gesture identification equipment 8 that becomes more meticulous
The plug-in type hard disk being equipped with equipment, such as the gesture identification equipment 8 that becomes more meticulous, intelligent memory card (Smart Media
Card, SMC), secure digital (Secure Digital, SD) card, flash card (Flash Card) etc..Further, it is described to deposit
Reservoir 81 can also both include the internal storage unit of the gesture identification equipment 8 that becomes more meticulous or including External memory equipment.Institute
Memory 81 is stated to be used to store the computer program and described other program sums to become more meticulous needed for gesture identification equipment
According to.The memory 81 can be also used for temporarily storing the data that has exported or will export.
It is apparent to those skilled in the art that for convenience of description and succinctly, only with above-mentioned each work(
Can unit, module division progress for example, in practical application, can be as needed and by above-mentioned function distribution by different
Functional unit, module are completed, i.e., the internal structure of described device are divided into different functional units or module, more than completion
The all or part of function of description.Each functional unit, module in embodiment can be integrated in a processing unit, also may be used
To be that unit is individually physically present, can also two or more units it is integrated in a unit, it is above-mentioned integrated
Unit can both be realized in the form of hardware, can also be realized in the form of SFU software functional unit.In addition, each function list
Member, the specific name of module are not limited to the protection domain of the application also only to facilitate mutually distinguish.Said system
The specific work process of middle unit, module, the corresponding process in preceding method embodiment is may be referred to, will not be repeated here.
In the above-described embodiments, the description to each embodiment all emphasizes particularly on different fields, and is not described in detail or remembers in some embodiment
The part of load, it may refer to the associated description of other embodiments.
Those of ordinary skill in the art are it is to be appreciated that the list of each example described with reference to the embodiments described herein
Member and algorithm steps, it can be realized with the combination of electronic hardware or computer software and electronic hardware.These functions are actually
Performed with hardware or software mode, application-specific and design constraint depending on technical scheme.Professional and technical personnel
Described function can be realized using distinct methods to each specific application, but this realization is it is not considered that exceed
The scope of the present invention.
In embodiment provided by the present invention, it should be understood that disclosed device/terminal device and method, can be with
Realize by another way.For example, device described above/terminal device embodiment is only schematical, for example, institute
The division of module or unit is stated, only a kind of division of logic function, there can be other dividing mode when actually realizing, such as
Multiple units or component can combine or be desirably integrated into another system, or some features can be ignored, or not perform.Separately
A bit, shown or discussed mutual coupling or direct-coupling or communication connection can be by some interfaces, device
Or INDIRECT COUPLING or the communication connection of unit, can be electrical, mechanical or other forms.
The unit illustrated as separating component can be or may not be physically separate, show as unit
The part shown can be or may not be physical location, you can with positioned at a place, or can also be distributed to multiple
On NE.Some or all of unit therein can be selected to realize the mesh of this embodiment scheme according to the actual needs
's.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit, can also
That unit is individually physically present, can also two or more units it is integrated in a unit.Above-mentioned integrated list
Member can both be realized in the form of hardware, can also be realized in the form of SFU software functional unit.
If the integrated module/unit realized in the form of SFU software functional unit and as independent production marketing or
In use, it can be stored in a computer read/write memory medium.Based on such understanding, the present invention realizes above-mentioned implementation
All or part of flow in example method, by computer program the hardware of correlation can also be instructed to complete, described meter
Calculation machine program can be stored in a computer-readable recording medium, and the computer program can be achieved when being executed by processor
The step of stating each embodiment of the method..Wherein, the computer program includes computer program code, the computer program
Code can be source code form, object identification code form, executable file or some intermediate forms etc..Computer-readable Jie
Matter can include：Can carry any entity or device of the computer program code, recording medium, USB flash disk, mobile hard disk,
Magnetic disc, CD, computer storage, read-only storage (ROM, Read-Only Memory), random access memory (RAM,
Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium etc..It is it should be noted that described
The content that computer-readable medium includes can carry out appropriate increasing according to legislation in jurisdiction and the requirement of patent practice
Subtract, such as in some jurisdictions, according to legislation and patent practice, computer-readable medium do not include be electric carrier signal and
Telecommunication signal.
Embodiment described above is merely illustrative of the technical solution of the present invention, rather than its limitations；Although with reference to foregoing reality
Example is applied the present invention is described in detail, it will be understood by those within the art that：It still can be to foregoing each
Technical scheme described in embodiment is modified, or carries out equivalent substitution to which part technical characteristic；And these are changed
Or replace, the essence of appropriate technical solution is departed from the spirit and scope of various embodiments of the present invention technical scheme, all should
Within protection scope of the present invention.
Claims (10)
- The gesture identification method 1. one kind becomes more meticulous, it is characterised in that the gesture identification method that becomes more meticulous includes：The local feature of hand joint position is extracted, the local feature includes relative seat feature and/or sequential displacement characteristic；Cluster calculation is carried out according to the relative seat feature and/or sequential displacement characteristic, obtain the relative seat feature and/ Or the cluster feature corresponding to sequential displacement characteristic；According to the cluster feature and the other corresponding relation of gesture class, cluster feature model training is carried out, according to poly- after training Category feature model carries out gesture identification.
- 2. the gesture identification method according to claim 1 that becomes more meticulous, it is characterised in that the part of extraction hand joint position The step of relative seat feature in feature, includes：T frame dynamic gesture images are obtained, it is determined that the position of the node per the hand in frame dynamic gesture image；Position according to the root node that the node of hand includes with corresponding child node, is calculated child node relative to root section The relative seat feature of point.
- 3. the gesture identification method according to claim 2 that becomes more meticulous, it is characterised in that wrapped in the node according to hand The position of the root node included and corresponding child node, be calculated child node relative to root node relative seat feature the step of Including：Obtain the relative seat feature of the t ' frames in T frame dynamic gesture imagesWherein： The position of root node is represented,Represent the position of child node corresponding with the root node, u ∈ i | and 1≤i≤N }, h=1 represents left hand, and h=2 represents the right hand, and 1≤t '≤T, N are the number of node.
- 4. the gesture identification method according to claim 1 that becomes more meticulous, it is characterised in that the part of extraction hand joint position The step of sequential displacement characteristic in feature, includes：T frame dynamic gesture images are obtained, it is determined that per the displacement characteristic reference point in frame dynamic gesture image；The position of displacement characteristic reference point in the adjacent image of every two frame, is determined corresponding to the dynamic gesture image Sequential displacement characteristic.
- 5. the gesture identification method according to claim 4 that becomes more meticulous, it is characterised in that the figure adjacent according to every two frame The position of displacement characteristic reference point as in, the step of determining the sequential displacement characteristic corresponding to the dynamic gesture image, wrap Include：Obtain the sequential displacement characteristic of t " framesWherein, For the displacement characteristic reference point of t " frames,For the displacement characteristic reference point of the frames of t " -1,1<T "≤T, v ∈ i | 1≤i ≤ N }, M is displacement characteristic reference point number and 1≤v≤M, h=1 represent left hand, and h=2 represents the right hand, and N is the number of node.
- 6. the gesture identification method according to claim 1 that becomes more meticulous, it is characterised in that described special according to the relative position Sign and/or sequential displacement characteristic carry out cluster calculation, obtain corresponding to the relative seat feature and/or sequential displacement characteristic The step of cluster feature, includes：The relative seat feature and/or sequential displacement characteristic are expressed as to the set of unified local feature；The cluster that predetermined number is chosen in the set of the unified local feature clusters set；Set is clustered to each cluster and carries out transformation calculations, is obtained corresponding to the relative seat feature and/or sequential displacement characteristic Cluster feature.
- The gesture identifying device 7. one kind becomes more meticulous, it is characterised in that the gesture identifying device that becomes more meticulous includes：Local shape factor unit, for extracting the local feature of hand joint position, the local feature includes relative position Feature and/or sequential displacement characteristic；Cluster feature computing unit, for carrying out cluster calculation according to the relative seat feature and/or sequential displacement characteristic, obtain To the cluster feature corresponding to the relative seat feature and/or sequential displacement characteristic；Recognition unit is trained, for according to the cluster feature and the other corresponding relation of gesture class, carrying out cluster feature model instruction Practice, gesture identification is carried out according to the cluster feature model after training.
- 8. the gesture identifying device according to claim 7 that becomes more meticulous, it is characterised in that the local shape factor unit bag Include：First image obtains subelement, for obtaining T frame dynamic gesture images, it is determined that per the hand in frame dynamic gesture image The position of node；First computation subunit, for the position according to the root node that the node of hand includes with corresponding child node, calculate Obtain relative seat feature of the child node relative to root node；And/orSecond image obtains subelement, for obtaining T frame dynamic gesture images, it is determined that special per the displacement in frame dynamic gesture image Levy reference point；Second computation subunit, for the position of the displacement characteristic reference point in the adjacent image of every two frame, it is determined that described Sequential displacement characteristic corresponding to dynamic gesture image.
- The gesture identification equipment 9. one kind becomes more meticulous, including memory, processor and be stored in the memory and can be in institute State the computer program run on processor, it is characterised in that realized described in the computing device during computer program as weighed Profit require to become more meticulous described in 1 to 6 any one gesture identification method the step of.
- 10. a kind of computer-readable recording medium, the computer-readable recording medium storage has computer program, and its feature exists In realizing and become more meticulous as described in any one of claim 1 to 6 gesture identification method when the computer program is executed by processor The step of.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710656434.0A CN107368820B (en) | 2017-08-03 | 2017-08-03 | Refined gesture recognition method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710656434.0A CN107368820B (en) | 2017-08-03 | 2017-08-03 | Refined gesture recognition method, device and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107368820A true CN107368820A (en) | 2017-11-21 |
| CN107368820B CN107368820B (en) | 2023-04-18 |
Family
ID=60309287
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710656434.0A Active CN107368820B (en) | 2017-08-03 | 2017-08-03 | Refined gesture recognition method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107368820B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108346168A (en) * | 2018-02-12 | 2018-07-31 | 腾讯科技（深圳）有限公司 | A kind of images of gestures generation method, device and storage medium |
| CN108921101A (en) * | 2018-07-04 | 2018-11-30 | 百度在线网络技术（北京）有限公司 | Processing method, equipment and readable storage medium storing program for executing based on gesture identification control instruction |
| CN109117766A (en) * | 2018-07-30 | 2019-01-01 | 上海斐讯数据通信技术有限公司 | A kind of dynamic gesture identification method and system |
| CN109117771A (en) * | 2018-08-01 | 2019-01-01 | 四川电科维云信息技术有限公司 | Incident of violence detection system and method in a kind of image based on anchor node |
| CN109992093A (en) * | 2017-12-29 | 2019-07-09 | 博世汽车部件(苏州)有限公司 | A kind of gesture comparative approach and gesture comparison system |
| CN110163130A (en) * | 2019-05-08 | 2019-08-23 | 清华大学 | A kind of random forest grader and classification method of the feature pre-align for gesture identification |
| CN111222486A (en) * | 2020-01-15 | 2020-06-02 | 腾讯科技（深圳）有限公司 | Training method, device and equipment for hand gesture recognition model and storage medium |
| US20210326657A1 (en) * | 2020-04-21 | 2021-10-21 | Pegatron Corporation | Image recognition method and device thereof and ai model training method and device thereof |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100071965A1 (en) * | 2008-09-23 | 2010-03-25 | Panasonic Corporation | System and method for grab and drop gesture recognition |
| US20100188519A1 (en) * | 2009-01-29 | 2010-07-29 | Keisuke Yamaoka | Information Processing Device and Method, Program, and Recording Medium |
| CN101976330A (en) * | 2010-09-26 | 2011-02-16 | 中国科学院深圳先进技术研究院 | Gesture recognition method and system |
| US20110158476A1 (en) * | 2009-12-24 | 2011-06-30 | National Taiwan University Of Science And Technology | Robot and method for recognizing human faces and gestures thereof |
| CN103246891A (en) * | 2013-05-28 | 2013-08-14 | 重庆邮电大学 | Chinese sign language recognition method based on kinect |
| US20140201126A1 (en) * | 2012-09-15 | 2014-07-17 | Lotfi A. Zadeh | Methods and Systems for Applications for Z-numbers |
| CN104598915A (en) * | 2014-01-24 | 2015-05-06 | 深圳奥比中光科技有限公司 | Gesture recognition method and gesture recognition device |
| CN106886751A (en) * | 2017-01-09 | 2017-06-23 | 深圳数字电视国家工程实验室股份有限公司 | A kind of gesture identification method and system |
| CN106937531A (en) * | 2014-06-14 | 2017-07-07 | 奇跃公司 | Method and system for producing virtual and augmented reality |
-
2017
- 2017-08-03 CN CN201710656434.0A patent/CN107368820B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100071965A1 (en) * | 2008-09-23 | 2010-03-25 | Panasonic Corporation | System and method for grab and drop gesture recognition |
| US20100188519A1 (en) * | 2009-01-29 | 2010-07-29 | Keisuke Yamaoka | Information Processing Device and Method, Program, and Recording Medium |
| US20110158476A1 (en) * | 2009-12-24 | 2011-06-30 | National Taiwan University Of Science And Technology | Robot and method for recognizing human faces and gestures thereof |
| CN101976330A (en) * | 2010-09-26 | 2011-02-16 | 中国科学院深圳先进技术研究院 | Gesture recognition method and system |
| US20140201126A1 (en) * | 2012-09-15 | 2014-07-17 | Lotfi A. Zadeh | Methods and Systems for Applications for Z-numbers |
| CN103246891A (en) * | 2013-05-28 | 2013-08-14 | 重庆邮电大学 | Chinese sign language recognition method based on kinect |
| CN104598915A (en) * | 2014-01-24 | 2015-05-06 | 深圳奥比中光科技有限公司 | Gesture recognition method and gesture recognition device |
| CN106937531A (en) * | 2014-06-14 | 2017-07-07 | 奇跃公司 | Method and system for producing virtual and augmented reality |
| CN106886751A (en) * | 2017-01-09 | 2017-06-23 | 深圳数字电视国家工程实验室股份有限公司 | A kind of gesture identification method and system |
Non-Patent Citations (2)
| Title |
|---|
| YUNGHER DON等: "Improving fine motor function after brain injury using gesture recognition biofeedback", 《DISABILITY AND REHABILITATION: ASSISTIVE TECHNOLOGY》 * |
| 曹洁等: "基于RGB-D信息的动态手势识别方法", 《计算机应用研究》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109992093A (en) * | 2017-12-29 | 2019-07-09 | 博世汽车部件(苏州)有限公司 | A kind of gesture comparative approach and gesture comparison system |
| CN109992093B (en) * | 2017-12-29 | 2024-05-03 | 博世汽车部件(苏州)有限公司 | Gesture comparison method and gesture comparison system |
| CN108346168A (en) * | 2018-02-12 | 2018-07-31 | 腾讯科技（深圳）有限公司 | A kind of images of gestures generation method, device and storage medium |
| CN108346168B (en) * | 2018-02-12 | 2019-08-13 | 腾讯科技（深圳）有限公司 | A kind of images of gestures generation method, device and storage medium |
| US11061479B2 (en) | 2018-07-04 | 2021-07-13 | Baidu Online Network Technology (Beijing) Co., Ltd. | Method, device and readable storage medium for processing control instruction based on gesture recognition |
| CN108921101A (en) * | 2018-07-04 | 2018-11-30 | 百度在线网络技术（北京）有限公司 | Processing method, equipment and readable storage medium storing program for executing based on gesture identification control instruction |
| CN109117766A (en) * | 2018-07-30 | 2019-01-01 | 上海斐讯数据通信技术有限公司 | A kind of dynamic gesture identification method and system |
| CN109117771A (en) * | 2018-08-01 | 2019-01-01 | 四川电科维云信息技术有限公司 | Incident of violence detection system and method in a kind of image based on anchor node |
| CN109117771B (en) * | 2018-08-01 | 2022-05-27 | 四川电科维云信息技术有限公司 | System and method for detecting violence events in image based on anchor nodes |
| CN110163130B (en) * | 2019-05-08 | 2021-05-28 | 清华大学 | Feature pre-alignment random forest classification system and method for gesture recognition |
| CN110163130A (en) * | 2019-05-08 | 2019-08-23 | 清华大学 | A kind of random forest grader and classification method of the feature pre-align for gesture identification |
| CN111222486A (en) * | 2020-01-15 | 2020-06-02 | 腾讯科技（深圳）有限公司 | Training method, device and equipment for hand gesture recognition model and storage medium |
| CN111222486B (en) * | 2020-01-15 | 2022-11-04 | 腾讯科技（深圳）有限公司 | Training method, device and equipment for hand gesture recognition model and storage medium |
| US20210326657A1 (en) * | 2020-04-21 | 2021-10-21 | Pegatron Corporation | Image recognition method and device thereof and ai model training method and device thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107368820B (en) | 2023-04-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107368820A (en) | One kind becomes more meticulous gesture identification method, device and equipment | |
| Cornea et al. | Computing hierarchical curve-skeletons of 3D objects | |
| Zhu et al. | A cuboid CNN model with an attention mechanism for skeleton-based action recognition | |
| CN107820593A (en) | A kind of virtual reality exchange method, apparatus and system | |
| WO2021120834A1 (en) | Biometrics-based gesture recognition method and apparatus, computer device, and medium | |
| Liang et al. | Model-based hand pose estimation via spatial-temporal hand parsing and 3D fingertip localization | |
| CN110363077A (en) | Sign Language Recognition Method, device, computer installation and storage medium | |
| CN103793683B (en) | Gesture recognition method and electronic device | |
| CN108229496A (en) | The detection method and device of dress ornament key point, electronic equipment, storage medium and program | |
| Liu et al. | Kinect-based hand gesture recognition using trajectory information, hand motion dynamics and neural networks | |
| CN107688824A (en) | Picture match method and terminal device | |
| Geng et al. | Gated path selection network for semantic segmentation | |
| CN107958230A (en) | Facial expression recognizing method and device | |
| CN108229559A (en) | Dress ornament detection method, device, electronic equipment, program and medium | |
| CN104685540A (en) | Image semantic segmentation method and apparatus | |
| De Smedt et al. | 3d hand gesture recognition by analysing set-of-joints trajectories | |
| CN110489424A (en) | A kind of method, apparatus, storage medium and the electronic equipment of tabular information extraction | |
| Hirzer et al. | Smart hypothesis generation for efficient and robust room layout estimation | |
| Zhang | Application of intelligent virtual reality technology in college art creation and design teaching | |
| Yoo et al. | Fast and accurate 3D hand pose estimation via recurrent neural network for capturing hand articulations | |
| CN104573737B (en) | The method and device of positioning feature point | |
| CN110046340A (en) | The training method and device of textual classification model | |
| Mousas et al. | Efficient hand-over motion reconstruction | |
| Zhang et al. | Accurate 3D hand pose estimation network utilizing joints information | |
| Li | Badminton motion capture with visual image detection of picking robotics |
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 |