CN105488794B - A kind of action prediction method and system based on space orientation and cluster - Google Patents
A kind of action prediction method and system based on space orientation and cluster Download PDFInfo
- Publication number
- CN105488794B CN105488794B CN201510837116.5A CN201510837116A CN105488794B CN 105488794 B CN105488794 B CN 105488794B CN 201510837116 A CN201510837116 A CN 201510837116A CN 105488794 B CN105488794 B CN 105488794B
- Authority
- CN
- China
- Prior art keywords
- information
- target point
- cluster
- frame
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000009471 action Effects 0.000 title claims abstract description 30
- 238000000034 method Methods 0.000 title claims abstract description 24
- 230000033001 locomotion Effects 0.000 claims abstract description 33
- 238000010276 construction Methods 0.000 claims description 8
- 230000005540 biological transmission Effects 0.000 abstract description 4
- 230000001133 acceleration Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 230000002035 prolonged effect Effects 0.000 description 2
- 208000012886 Vertigo Diseases 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000002604 ultrasonography Methods 0.000 description 1
- 231100000889 vertigo Toxicity 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The action prediction method and system based on space orientation and cluster that the invention discloses a kind of, wherein the method includes：Space orientation is carried out to target point, obtains the location information and movement state information of each frame of the target point；According at least two frames, continuously the location information of the target point and motion state letter carry out information coding, establish the model of at least two frame continuous informations；The model of at least two frame continuous informations is clustered, cluster result is obtained；Training dataset is established according to the cluster result, and operating mode is obtained according to the training dataset；Action is predicted according to the operation mode.Implement the embodiment of the present invention, space orientation is carried out to target point and based under externally input huge data, effective prediction is carried out to reduce the time delay because being generated in signals transmission by clustering technique, to effectively reduce subjective delay, improves the usage experience sense of user.
Description
Technical field
The present invention relates to action prediction technical field more particularly to a kind of action prediction sides based on space orientation and cluster
Method and system.
Background technology
Space orientation technique is the key technology in virtual reality system.In the application of virtual reality, need all the time
User is wanted to be interacted with virtual environment, if continuing to use traditional interaction technique, the situation that user experience can be caused bad.With certainly
It is a kind of better interactive mode that right body action, which interacts, this just needs to position the body of user.
Common indoor wireless location technology also has：Wi-Fi, bluetooth, infrared ray, ultra wide band, RFID, ZigBee and ultrasound
Wave.Current space positioning system can reach the sample effect of 1000 frame per second using high-speed camera, and user carries out picture
Switching is than smoother and fine and smooth.However the universal disadvantage of current space positioning system be delayed it is higher.Delay height leads to picture
Transformation it is not prompt enough, cause the vertigo symptoms of user to happen occasionally.
Therefore, it is also the key component in this system to reduce the delay in space positioning system.
Invention content
It is an object of the invention to overcome the deficiencies in the prior art, and the present invention provides one kind being based on space orientation and cluster
Action prediction method and system, space orientation is carried out to target point and based under externally input huge data, passes through cluster
Technology carries out effective prediction to reduce the time delay because being generated in signals transmission, to effectively reduce subjective delay, improves
The usage experience sense of user.
In order to solve the above technical problem, the present invention provides a kind of action prediction sides based on space orientation and cluster
Method, the method includes：
Space orientation is carried out to target point, obtains the location information and movement state information of each frame of the target point；
According at least two frames, continuously the location information of the target point and motion state letter carry out information coding, establish extremely
The model of few two frame continuous informations；
The model of at least two frame continuous informations is clustered, cluster result is obtained；
Training dataset is established according to the cluster result, and operating mode is obtained according to the training dataset；
Action is predicted according to the operation mode.
Preferably, the movement state information includes instantaneous velocity information and acceleration information.
Preferably, the basis at least two frames continuously believe into row information by the location information of the target point and motion state
Coding establishes the model of at least two frame continuous informations, including：
The location information and movement state information of the target point are recorded in each frame；
The location information and motion state letter of the target point of continuous at least two frames of record are gone forward side by side row information coding, acquisition
Information coding result；
According to the information vector of described information coding result structure at least bidimensional；
According to the information vector of at least bidimensional, the model of at least two frame continuous informations is obtained.
Preferably, the model at least two frame continuous informations clusters, and obtains cluster result, including：
The information vector of each that structure is completed at least bidimensional carries out category division；
The identical information vector of classification is clustered, cluster result is obtained.
Preferably, the operating mode includes normal mode of operation and prediction work pattern.
The action prediction system based on space orientation and cluster that the present invention also provides a kind of, the system comprises：
Locating module：For carrying out space orientation to target point, the location information and fortune of each frame of the target point are obtained
Dynamic status information；
Model construction module：For according at least two frames continuously the location information of the target point and motion state believe into
Row information encodes, and establishes the model of at least two frame continuous informations；
Cluster module：It is clustered for the model at least two frame continuous informations, obtains cluster result；
Pattern acquiring module：For establishing training dataset according to the cluster result, and according to the training dataset
Obtain operating mode；
Prediction module：For predicting according to the operation mode action.
Preferably, the movement state information includes instantaneous velocity information and acceleration information.
Preferably, the model construction module includes：
Information recording unit：Location information for recording the target point in each frame and movement state information；
Information coding unit：The location information and motion state letter of the target point for recording continuous at least two frames are simultaneously
Information coding is carried out, information coding result is obtained；
Information vector construction unit：For the information vector according to described information coding result structure at least bidimensional；
Model acquiring unit：For the information vector according at least bidimensional, the mould of acquisition at least two frame continuous informations
Type.
Preferably, the cluster module includes：
Category division unit：Information vector for that will build each at least bidimensional completed carries out category division；
Cluster cell：For clustering the identical information vector of classification, cluster result is obtained.
Preferably, the operating mode includes normal mode of operation and prediction work pattern.
In embodiments of the present invention, by carrying out space orientation to target point and being based under externally input huge data,
Effective prediction is carried out to reduce the time delay because being generated in signals transmission by clustering technique, is prolonged to effectively reduce subjectivity
When, improve the usage experience sense of user.
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is the flow diagram of the action prediction method based on space orientation and cluster in the embodiment of the present invention；
Fig. 2 is the flow diagram for establishing at least model of two frame continuous informations in the embodiment of the present invention；
Fig. 3 is the structure composition signal of the action prediction system based on space orientation and cluster in the embodiment of the present invention
Figure.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other without creative efforts
Embodiment shall fall within the protection scope of the present invention.
Fig. 1 is the flow diagram of the action prediction method based on space orientation and cluster in the embodiment of the present invention, such as
Shown in Fig. 1, this method includes：
S11：Space orientation is carried out to target point, obtains the location information and movement state information of each frame of target point；
S12：Information coding is carried out according to the location information of the continuous target point of at least two frames and motion state letter, is established extremely
The model of few two frame continuous informations；
S13：The model of at least two frame continuous informations is clustered, cluster result is obtained；
S14：Training dataset is established according to cluster result, and operating mode is obtained according to training dataset；
S15：Action is predicted according to operating mode.
S11 is described further：
Space orientation is carried out to target point by way of camera or gravity sensor in the present embodiment, obtains the mesh
The spatial positional information of punctuate；The fortune of target point is obtained by instruments such as gyroscope, speed sensor, accelerator and magnetometers
Dynamic status information, wherein movement state information include being not limited to instantaneous velocity information and acceleration information.
In the present embodiment, by the location information of the target point in space and status information using mathematical linguistics mode come table
Show, i.e., the location information of each target point uses position vectorThe instantaneous velocity information of each target point
All use instantaneous velocity vectorThe acceleration information of each target point uses vector acceleration
S12 is described further：
S12 is illustrated in conjunction with Fig. 2, Fig. 2 is the model of the foundation at least two frame continuous informations in the embodiment of the present invention
Flow diagram, as shown in Fig. 2, the flow includes the following steps：
S121：The location information and movement state information of target point are recorded in each frame；
S122：The record continuous at least location information of the target point of two frames and motion state letter row information of going forward side by side encode, and obtain
It wins the confidence breath coding result；
S123：According to the information vector of information coding result structure at least bidimensional；
S124：According to the information vector of at least bidimensional, the model of at least two frame continuous informations is obtained.
S121 is described further：
In embodiments of the present invention, the location information and movement state information of each frame target point are recorded.
S122 is described further：
In the embodiment of the present invention, carried out to recording at least location information of the continuous target point of N frames and movement state information
Rank-numeral encodes, wherein N >=2, by being encoded, gets the location information and movement state information after encoding.
S123 is described further：
To after coding location information and movement state information carry out vectorization processing, be converted into the information of vectorization.
S124 is described further：
It is as follows according to the information architecture one of these vectorizations at least two frame continuous information matrix models：
Wherein, N >=2 and N >=L, the preceding L vector for defining a series of movements is key value, and subsequent N-L vector is prediction
Value.
S13 is described further：
Information vector by that will build each at least bidimensional completed carries out category division, by the identical information of classification
Vector is regrouped divides a new classification into together, to obtain cluster result.
S14 is described further
Cluster result is trained, is new data by the aiming spot movement (action element) that N number of element forms
, the pattern of work is got according to these data item, wherein operating mode has normal mode of operation and prediction work pattern,
Middle N >=2.
S15 is described further：
First, work as N=L, enter normal mode of operation, which terminates；When N >=L enters prediction mode, output is
Prediction data carries out action reduction according to these prediction data, obtains prediction action；And by the data of prediction and the data that transmit
It compares, verifies the accuracy of prediction data, subsequently to provide more accurate prediction result.
Fig. 3 is the structure composition signal of the action prediction system based on space orientation and cluster in the embodiment of the present invention
Figure, as shown in figure 3, the system includes：
Locating module：For carrying out space orientation to target point, the location information and movement shape of each frame of target point are obtained
State information；
Model construction module：For carrying out letter according to the location information and motion state letter of the continuous target point of at least two frames
Breath coding, establishes the model of at least two frame continuous informations；
Cluster module：It is clustered for the model at least two frame continuous informations, obtains cluster result；
Pattern acquiring module：For establishing training dataset according to cluster result, and work is obtained according to training dataset
Pattern；
Prediction module：For being predicted action according to operating mode.
Preferably, movement state information includes instantaneous velocity information and acceleration information.
Preferably, model construction module includes：
Information recording unit：Location information for recording target point in each frame and movement state information；
Information coding unit：Believe and carry out for recording the continuous at least location information of the target point of two frames and motion state
Information coding obtains information coding result；
Information vector construction unit：For the information vector according to information coding result structure at least bidimensional；
Model acquiring unit：For the information vector according at least bidimensional, the model of at least two frame continuous informations is obtained.
Preferably, cluster module includes：
Category division unit：Information vector for that will build each at least bidimensional completed carries out category division；
Cluster cell：For clustering the identical information vector of classification, cluster result is obtained.
Preferably, operating mode includes normal mode of operation and prediction work pattern.
Specifically, the operation principle of the system related functions module of the embodiment of the present invention can be found in the correlation of embodiment of the method
Description, which is not described herein again.
In embodiments of the present invention, by carrying out space orientation to target point and being based under externally input huge data,
Effective prediction is carried out to reduce the time delay because being generated in signals transmission by clustering technique, is prolonged to effectively reduce subjectivity
When, improve the usage experience sense of user.
One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment is can
It is completed with instructing relevant hardware by program, which can be stored in a computer readable storage medium, storage
Medium may include：Read-only memory (ROM, Read Only Memory), random access memory (RAM, Random
Access Memory), disk or CD etc..
In addition, be provided for the embodiments of the invention above a kind of action prediction method based on space orientation and cluster and
System is described in detail, and principle and implementation of the present invention are described for specific case used herein, with
The explanation of upper embodiment is merely used to help understand the method and its core concept of the present invention；Meanwhile for the general of this field
Technical staff, according to the thought of the present invention, there will be changes in the specific implementation manner and application range, in conclusion
The content of the present specification should not be construed as limiting the invention.
Claims (6)
1. a kind of action prediction method based on space orientation and cluster, which is characterized in that the method includes：
Space orientation is carried out to target point, obtains the location information and movement state information of each frame of the target point；
The location information and movement state information of the target point are recorded in each frame；
The location information and movement state information of the target point of continuous at least two frames of record are gone forward side by side row information coding, and letter is obtained
Cease coding result；
According to the information vector of described information coding result structure at least bidimensional；
According to the information vector of at least bidimensional, the model of at least two frame continuous informations is obtained；
The model of at least two frame continuous informations is clustered, cluster result is obtained；
Training dataset is established according to the cluster result, and operating mode is obtained according to the training dataset；
Action is predicted according to the operation mode.
2. action prediction method according to claim 1, which is characterized in that described at least two frame continuous informations
Model is clustered, and cluster result is obtained, including：
The information vector of each that structure is completed at least bidimensional carries out category division；
The identical information vector of classification is clustered, cluster result is obtained.
3. action prediction method according to claim 1, which is characterized in that the operating mode includes normal mode of operation
With prediction work pattern.
4. a kind of action prediction system based on space orientation and cluster, which is characterized in that the system comprises：
Locating module：For carrying out space orientation to target point, the location information and movement shape of each frame of the target point are obtained
State information；
Model construction module, including：Information recording unit：Location information for recording the target point in each frame and fortune
Dynamic status information；
Information coding unit：The location information and motion state of the target point for recording continuous at least two frames are believed and are carried out
Information coding obtains information coding result；
Information vector construction unit：For the information vector according to described information coding result structure at least bidimensional；
Model acquiring unit：For the information vector according at least bidimensional, the model of acquisition at least two frame continuous informations；
Cluster module：It is clustered for the model at least two frame continuous informations, obtains cluster result；
Pattern acquiring module：For establishing training dataset according to the cluster result, and obtained according to the training dataset
Operating mode；
Prediction module：For predicting according to the operation mode action.
5. action prediction system according to claim 4, which is characterized in that the cluster module includes：
Category division unit：Information vector for that will build each at least bidimensional completed carries out category division；
Cluster cell：For clustering the identical information vector of classification, cluster result is obtained.
6. action prediction system according to claim 4, which is characterized in that the operating mode includes normal mode of operation
With prediction work pattern.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510837116.5A CN105488794B (en) | 2015-11-26 | 2015-11-26 | A kind of action prediction method and system based on space orientation and cluster |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510837116.5A CN105488794B (en) | 2015-11-26 | 2015-11-26 | A kind of action prediction method and system based on space orientation and cluster |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105488794A CN105488794A (en) | 2016-04-13 |
| CN105488794B true CN105488794B (en) | 2018-08-24 |
Family
ID=55675759
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510837116.5A Active CN105488794B (en) | 2015-11-26 | 2015-11-26 | A kind of action prediction method and system based on space orientation and cluster |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105488794B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106504266B (en) * | 2016-09-29 | 2019-06-14 | 北京市商汤科技开发有限公司 | The prediction technique and device of walking behavior, data processing equipment and electronic equipment |
| CN106644397B (en) * | 2016-12-16 | 2019-06-21 | 捷开通讯(深圳)有限公司 | The detection device and detection method of the delay time of VR device |
| CN106644396B (en) * | 2016-12-16 | 2019-06-25 | 捷开通讯(深圳)有限公司 | The detection device and detection method of the delay time of VR glasses |
| CN109740418B (en) * | 2018-11-21 | 2022-10-14 | 中山大学 | Yoga action identification method based on multiple acceleration sensors |
| CN114943936B (en) * | 2022-06-17 | 2023-06-20 | 北京百度网讯科技有限公司 | Target behavior recognition method and device, electronic equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1212526A (en) * | 1997-08-07 | 1999-03-31 | 松下电器产业株式会社 | Device and method for detecting action vector |
| CN102427528A (en) * | 2011-09-30 | 2012-04-25 | 北京航空航天大学 | Video motion estimating method based on clustering statistics |
| CN103661165A (en) * | 2012-09-19 | 2014-03-26 | 阿尔卑斯电气株式会社 | Movement prediction device and input apparatus using the same |
-
2015
- 2015-11-26 CN CN201510837116.5A patent/CN105488794B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1212526A (en) * | 1997-08-07 | 1999-03-31 | 松下电器产业株式会社 | Device and method for detecting action vector |
| CN102427528A (en) * | 2011-09-30 | 2012-04-25 | 北京航空航天大学 | Video motion estimating method based on clustering statistics |
| CN103661165A (en) * | 2012-09-19 | 2014-03-26 | 阿尔卑斯电气株式会社 | Movement prediction device and input apparatus using the same |
Non-Patent Citations (2)
| Title |
|---|
| 《Discriminative Density Propagation for 3D Human Motion Estimation》;Cristian Sminchisescu et al.;《Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition》;20051231;第1-8页 * |
| 《基于聚类的相位相关运动估计方法》;马红星等;《光学与光电技术》;20120430;第2卷(第10期);第93-96、100页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105488794A (en) | 2016-04-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105488794B (en) | A kind of action prediction method and system based on space orientation and cluster | |
| JP6716650B2 (en) | System and method for deep learning based hand gesture recognition in first person view | |
| CN103404169B (en) | Focus on the method and system of microphone array | |
| WO2021120157A1 (en) | Light weight multi-branch and multi-scale person re-identification | |
| JP2018198081A (en) | Activity recognition robot apparatus | |
| CN109996039A (en) | A kind of target tracking method and device based on edge calculations | |
| CN111185008B (en) | Method and apparatus for controlling virtual character in game | |
| US11537857B2 (en) | Pooling processing method and system applied to convolutional neural network | |
| US11351458B2 (en) | Method for controlling target object, apparatus, device, and storage medium | |
| CN109074497A (en) | Use the activity in depth information identification sequence of video images | |
| US20220270314A1 (en) | Robust facial animation from video using neural networks | |
| Zhao et al. | Intelligent wireless sensing driven metaverse: A survey | |
| WO2019144346A1 (en) | Object processing method in virtual scene, device and storage medium | |
| CN115222847A (en) | Animation data generation method and device based on neural network and related products | |
| Dey et al. | Umpire’s Signal Recognition in Cricket Using an Attention based DC-GRU Network | |
| CN111078008B (en) | Control method of early education robot | |
| KR102590916B1 (en) | Method and device for inferring dynamic relationship between objects in video | |
| US20230005225A1 (en) | Cloud-based Production of High-Quality Virtual And Augmented Reality Video Of User Activities | |
| US20190236364A1 (en) | Systems and methods for providing video presentation and video analytics for live sporting events | |
| Wang | [Retracted] Long Jump Action Recognition Based on Deep Convolutional Neural Network | |
| KR20230096826A (en) | Control method for server providing block puzzle with customized quiz difficulty for user level | |
| Zhang et al. | Skeleton‐Guided Action Recognition with Multistream 3D Convolutional Neural Network for Elderly‐Care Robot | |
| KR101577624B1 (en) | Multi display system | |
| Tang et al. | The application prospects of robot pose estimation technology: exploring new directions based on YOLOv8-ApexNet | |
| Wang | Scaling Wearable Cognitive Assistance |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information | ||
| CB03 | Change of inventor or designer information |
Inventor after: Zhou Fan Inventor after: Chen Xiaoyan Inventor after: Zheng Guifeng Inventor before: Zhou Fan Inventor before: Li Wanhua Inventor before: Huang Zhenye Inventor before: Chen Xiaoyan Inventor before: Cao Siqi Inventor before: Chen Tuli |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20160413 Assignee: LONGSE TECHNOLOGY Co.,Ltd. Assignor: SUN YAT-SEN University Contract record no.: X2020440000043 Denomination of invention: Spatial positioning and clustering based action prediction method and system Granted publication date: 20180824 License type: Common License Record date: 20200320 |