CN106372673A - Apparatus motion identification method - Google Patents
Apparatus motion identification method Download PDFInfo
- Publication number
- CN106372673A CN106372673A CN201610809925.XA CN201610809925A CN106372673A CN 106372673 A CN106372673 A CN 106372673A CN 201610809925 A CN201610809925 A CN 201610809925A CN 106372673 A CN106372673 A CN 106372673A
- Authority
- CN
- China
- Prior art keywords
- information
- motion
- movement
- action recognition
- recognition methodss
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/03—Arrangements for converting the position or the displacement of a member into a coded form
- G06F3/033—Pointing devices displaced or positioned by the user, e.g. mice, trackballs, pens or joysticks; Accessories therefor
- G06F3/0346—Pointing devices displaced or positioned by the user, e.g. mice, trackballs, pens or joysticks; Accessories therefor with detection of the device orientation or free movement in a 3D space, e.g. 3D mice, 6-DOF [six degrees of freedom] pointers using gyroscopes, accelerometers or tilt-sensors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/10—Image acquisition
- G06V10/12—Details of acquisition arrangements; Constructional details thereof
- G06V10/14—Optical characteristics of the device performing the acquisition or on the illumination arrangements
- G06V10/147—Details of sensors, e.g. sensor lenses
Abstract
The invention discloses an apparatus motion identification method. The method comprises the following steps: S1, setting a movement sensor and a timer for an apparatus, and obtaining corresponding movement parameters, wherein the movement sensor provides movement sensing of six axes and comprises three rotation movement sensors and three accelerometers, and the movement parameters comprise triaxial angular speeds, triaxial accelerated speeds and time parameters provided by the timer; S2, performing clustering processing on the movement parameters to obtain information to be compared; and S3, performing comparison identification on the information to be compared with a template database and outputting corresponding motion information. According to the invention, movement data of the apparatus is measured through the movement sensor, the time parameters corresponding to the movement data are measured through the timer, through combination of the two kinds of parameters and comparison with the training template database of corresponding motion determination, motion of the apparatus is obtained, the motion information of the apparatus can be provided, corresponding control can be performed according to motion, and thus the intelligent processing capability of the apparatus is improved.
Description
Technical field
The present invention relates to a kind of device action recognition methodss, belong to action recognition field.
Background technology
With the development in industrial 4.0 epoch, the oriented miniaturization of development of equipment, intelligentized trend, in order to meet to dress
That puts is precisely controlled, and the state of correct identifying device just seems very important.
The motion sensor such as the accelerometer of inertia sensing type or gyroscope can use in the electronic device.Plus
Velometer can be used for measuring linear acceleration, and gyroscope can be used for measuring the angular velocity of mobile device.Motion sensor
Market include mobile phone, PlayStation 3 videogame console/PS3, pda, mobile internet device (mid), personal navigation equipment (pnd),
Digital camera, digital camera etc..For example, cell phone can carry out testing equipment inclining in space using accelerometer
Tiltedly, this permission shows video image on direction corresponding with this inclination.Video game console controller can be using acceleration
Degree meter the motion being used for providing the manual controller of input to game is detected.
But which kind of motion the judgement equipment that existing technology is mainly concentrated has been by rather than has been carried out which kind of action,
Therefore face a problem that can carry out action identifying.
Content of the invention
In order to solve the above problems, the present invention passes through to provide kind of device action recognition methodss.
The technical solution used in the present invention one side is a kind of device action recognition methodss, comprises the following steps:
Step s1: motion sensor and timer are set for device, motion sensor provides the motion sensing of six axles, including three
Individual rotational motion sensor and three accelerometers, obtain corresponding kinematic parameter, described kinematic parameter include three axis angular rates, three
Axle acceleration and the time parameter of described timer offer;
Step s2: clustering processing is carried out to described kinematic parameter and obtains treating comparative information;
Step s3: will treat that comparative information carries out contrast with template base and identifies and export corresponding action message.
Preferably, described step s2 includes carrying out windowing process, smoothing processing and quantification treatment to kinematic parameter.
Preferably, described step s1 also includes, and calculates Eulerian angles based on pid algorithm.
Preferably, described step s1 also includes being converted to the angle of motion sensor and earth axes based on Eulerian angles
Information.
Preferably, described step s3 realizes the contrast identification of comparative information based on dtw algorithm.
Preferably, described step s3 also includes identifying treating that comparative information and template base carry out contrasting, if contrast identification
Successful then output information.
Preferably, described information includes acoustic information and light information.
Beneficial effects of the present invention are by setting up a kind of device action recognition methodss, by motion sensor measurement apparatus
Motion data, by the corresponding time parameter of timer measuring motion data, in conjunction with two parameter and contrast corresponding action and sentence
Disconnected training template base, obtains the action of this device it is provided that the action message of device, is conducive to making correspondence according to action
Control, improve device intelligent processing capacity.
Brief description
Fig. 1 show a kind of schematic diagram of the device action recognition methodss based on the embodiment of the present invention.
Specific embodiment
The present invention will be described with reference to embodiments.
Based on inventive embodiment, a kind of device action recognition methodss, comprise the following steps: step s1: be as shown in Figure 1
Device setting motion sensor and timer, motion sensor provides the motion sensing of six axles, passes including three rotary motions
Sensor and three accelerometers, obtain corresponding kinematic parameter, and described kinematic parameter includes three axis angular rates, 3-axis acceleration and institute
The time parameter of timer offer is provided;Step s2: clustering processing is carried out to described kinematic parameter and obtains treating comparative information;Step
S3: will treat that comparative information carries out contrast with template base and identifies and export corresponding action message.
First, for needing to carry out on the device of action judgement, motion sensor (six common axle sensors) and meter are disposed
When device, motion sensor provides the motion sensing of six axles, including three rotational motion sensor and three accelerometers, wherein
Rotational motion sensor is used for measurement sensing roll(and tilts), pitch(tilts forward and back), yaw(vacillates now to the left, now to the right) full side
Position multidate information, accelerometer is used for sensing three directions of xyz(solid space, all around up and down) acceleration in axial direction,
The parameter of the above-mentioned measuring instrument of labelling is kinematic parameter;Clustering processing is carried out to described kinematic parameter, clustering processing is included to fortune
Dynamic parameter carries out windowing process, smoothing processing and quantification treatment, the first step, kinematic parameter is carried out with windowing process (based on fractionation
Action, action that will be complicated, progressively cut into the combination of individual part, for example, jump, in the stage of squatting down, downward acceleration
For some value vanishing, the acceleration very little of left and right directions, can be considered do not have relative movement simultaneously, squats down and stops arriving upwards
Stage, downwards and the speed of left and right directions is all zero, ascent stage then upwards acceleration be a changing value, left and right is zero, on
Rise stop phase, acceleration and the relation of time can be passed through, calculating the relatively original state of object is to move up, then root
May determine that according to the situation of left and right acceleration be liftoff upper jump rather than jump to the left and right, the effect of adding window predominantly carries out action
Decompose), smoothing processing is that unexpected impact is eliminated, and predominantly arranges a value of the confidence, when motion change by parameter
Difference is more than when the value of the confidence then it is assumed that this value is exceptional value, is processed (if the object of regular motion, then different
Constant value can seldom, but actually used in, the factor such as example external shock, cause object to produce abnormal, now, by smooth
Process, be able to maintain that original judgement, even if object is knocked, but the action of its original is jump this point still without changing
Become, declaration of exception can be sent to support trouble shooting according to the situation of exceptional value simultaneously), quantification treatment is continuously to be counted
Turn to discrete data and processed according to becoming, when so can reduce identification, the difficulty that compares, kinematic parameter is carried out above-mentioned
After process, according to the similarity principle of clustering algorithm, the parameter sets that can be classified as the behavior of a class (are likely at one
It is similar action and intersection, and the result of this intersection is it is simply that by a series of continuously similar action, gradually enter
One complete action of row), obtain treating comparative information;Template base pre-sets, through some fortune tested in advance and obtain
The data base obtaining that logins of dynamic parameter and corresponding action, contrast is mainly the contrast according to various parameters, and that is, data is one
Then belong to certain correspondence within the scope of fixed and represent numeral, represent numeral correspondence and represent some action.
Described step s2 includes carrying out windowing process, smoothing processing and quantification treatment to kinematic parameter.
Described step s1 also includes, and calculates Eulerian angles based on pid algorithm.
Pid algorithm is in process control, the pid control being (i) controlled with differential (d) in the ratio (p) of deviation, integration
Device (also known as pid actuator) processed is a kind of automatic controller being most widely used.It is simple that it has principle, it is easy to accomplish, fit
With wide, control parameter is separate, parameter selected fairly simple the advantages of;And may certify that in theory, for mistake
The typical subject of process control -- " first-order lag+purely retarded " and the control object of " second-order lag+purely retarded ", pid controller
It is a kind of optimum control.Pid Rule adjusting is a kind of effective ways of continuous system dynamic quality correction, its parameter tuning side
Formula is easy, and structural change is flexibly (pi, pd ...).
By pid algorithm, kinematic parameter the parameter of motion sensor (mainly) can be processed to obtain accurately
Change procedure (change that object is fastened in own coordinate, i.e. Eulerian angles).
Described step s1 also includes being converted to the angle information of motion sensor and earth axes based on Eulerian angles.
Based on the conversion between existing quaternary number and Eulerian angles, obtain the angle information between each system of earth axes
(above-mentioned treating that comparative information combines with the type of abundant action for coordinating).
Described step s3 realizes the contrast identification of comparative information based on dtw algorithm.
Dtw(dynamic time warping, dynamic time consolidation) algorithm, the think of Ji Yu dynamic programming (dp) for this algorithm
Think, solve the problems, such as Parameters variation template matching different in size, time serieses are a kind of common representations of data.For
For time Series Processing, a universal task is exactly to compare the similarity of two sequences, and same action is although name one
Sample, but its persistent period is all different with corresponding momentum change, therefore passes through dtw algorithm, will one of (or two
Individual) sequence warping distortion under time shafts, to reach more preferable alignment.Obtain Time alignment function w with meeting certain condition
N () describes the corresponding time relationship of test template and reference template, corresponding to cumulative distance minimum when solving two template matching
Warping function.The data of same change type is classified as a class, in such manner, it is possible to be not provided with the template base (comparison database) of substantial amounts
And mainly store the judge templet of key, and by after the parameter (after comparative information) that judges is processed, carry out with judge templet
It is equipped with realization contrast identification.
Described step s3 also includes identifying treating that comparative information and template base carry out contrasting, if contrast identifies successfully, defeated
Go out information.
Described information includes acoustic information and light information.
One prompting mechanism of setting, successfully identifies, output acoustic information (for example carry out any action now, or
Person's output " action recognition success " etc.) and light information (for example display lamp is lighted).
The above, simply presently preferred embodiments of the present invention, the invention is not limited in above-mentioned embodiment, as long as
It reaches the technique effect of the present invention with identical means, all should belong to protection scope of the present invention.Protection model in the present invention
In enclosing, its technical scheme and/or embodiment can have various different modifications and variations.
Claims (7)
1. a kind of device action recognition methodss are it is characterised in that comprise the following steps:
Step s1: motion sensor and timer are set for device, motion sensor provides the motion sensing of six axles, including three
Individual rotational motion sensor and three accelerometers, obtain corresponding kinematic parameter, described kinematic parameter include three axis angular rates, three
Axle acceleration and the time parameter of described timer offer;
Step s2: clustering processing is carried out to described kinematic parameter and obtains treating comparative information;
Step s3: will treat that comparative information carries out contrast with template base and identifies and export corresponding action message.
2. a kind of device action recognition methodss according to claim 1 are it is characterised in that described step s2 is included to motion
Parameter carries out windowing process, smoothing processing and quantification treatment.
3. a kind of device action recognition methodss according to claim 1 are it is characterised in that described step s1 also includes, base
Calculate Eulerian angles in pid algorithm.
4. a kind of device action recognition methodss according to claim 3 are it is characterised in that described step s1 also includes being based on
Eulerian angles are converted to the angle information of motion sensor and earth axes.
5. a kind of device action recognition methodss according to claim 1 are it is characterised in that described step s3 is based on dtw calculation
Method realizes the contrast identification of comparative information.
6. a kind of device action recognition methodss according to claim 1 are it is characterised in that described step s3 also includes treating
Comparative information and template base carry out contrast identification, if contrast identifies successfully, export information.
7. a kind of device action recognition methodss according to claim 6 are it is characterised in that described information includes sound
Information and light information.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610809925.XA CN106372673A (en) | 2016-09-06 | 2016-09-06 | Apparatus motion identification method |
PCT/CN2017/099685 WO2018045902A1 (en) | 2016-09-06 | 2017-08-30 | Apparatus action recognition method, computer device, and computer readable storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610809925.XA CN106372673A (en) | 2016-09-06 | 2016-09-06 | Apparatus motion identification method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106372673A true CN106372673A (en) | 2017-02-01 |
Family
ID=57899066
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610809925.XA Pending CN106372673A (en) | 2016-09-06 | 2016-09-06 | Apparatus motion identification method |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN106372673A (en) |
WO (1) | WO2018045902A1 (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2018045902A1 (en) * | 2016-09-06 | 2018-03-15 | 深圳市民展科技开发有限公司 | Apparatus action recognition method, computer device, and computer readable storage medium |
CN108703444A (en) * | 2018-06-15 | 2018-10-26 | 吉林大学 | The Intelligent bracelet and method of occupant and driver for identification |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103892840A (en) * | 2014-03-06 | 2014-07-02 | 深圳市德凯瑞科技有限公司 | Intelligent wearing device and method for extracting human body motion features |
CN104021295A (en) * | 2014-06-12 | 2014-09-03 | 广州三星通信技术研究有限公司 | Clustering feature fusion method and device for motion recognition |
CN104407617A (en) * | 2014-12-22 | 2015-03-11 | 大连理工大学 | Programmable aircraft attitude control IP core |
CN104731307A (en) * | 2013-12-20 | 2015-06-24 | 孙伯元 | Somatic action identifying method and man-machine interaction device |
CN105184325A (en) * | 2015-09-23 | 2015-12-23 | 歌尔声学股份有限公司 | Human body action recognition method and mobile intelligent terminal |
CN105549746A (en) * | 2016-01-28 | 2016-05-04 | 广州成潮智能科技有限公司 | Action identification method based on acceleration sensing chip |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103984416B (en) * | 2014-06-10 | 2017-02-08 | 北京邮电大学 | Gesture recognition method based on acceleration sensor |
CN104866099B (en) * | 2015-05-27 | 2017-12-26 | 东南大学 | The error compensating method of smart machine gesture identification precision is improved based on motion sensor |
CN105824420B (en) * | 2016-03-21 | 2018-09-14 | 李骁 | A kind of gesture identification method based on acceleration transducer |
CN106372673A (en) * | 2016-09-06 | 2017-02-01 | 深圳市民展科技开发有限公司 | Apparatus motion identification method |
-
2016
- 2016-09-06 CN CN201610809925.XA patent/CN106372673A/en active Pending
-
2017
- 2017-08-30 WO PCT/CN2017/099685 patent/WO2018045902A1/en active Application Filing
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104731307A (en) * | 2013-12-20 | 2015-06-24 | 孙伯元 | Somatic action identifying method and man-machine interaction device |
CN103892840A (en) * | 2014-03-06 | 2014-07-02 | 深圳市德凯瑞科技有限公司 | Intelligent wearing device and method for extracting human body motion features |
CN104021295A (en) * | 2014-06-12 | 2014-09-03 | 广州三星通信技术研究有限公司 | Clustering feature fusion method and device for motion recognition |
CN104407617A (en) * | 2014-12-22 | 2015-03-11 | 大连理工大学 | Programmable aircraft attitude control IP core |
CN105184325A (en) * | 2015-09-23 | 2015-12-23 | 歌尔声学股份有限公司 | Human body action recognition method and mobile intelligent terminal |
CN105549746A (en) * | 2016-01-28 | 2016-05-04 | 广州成潮智能科技有限公司 | Action identification method based on acceleration sensing chip |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2018045902A1 (en) * | 2016-09-06 | 2018-03-15 | 深圳市民展科技开发有限公司 | Apparatus action recognition method, computer device, and computer readable storage medium |
CN108703444A (en) * | 2018-06-15 | 2018-10-26 | 吉林大学 | The Intelligent bracelet and method of occupant and driver for identification |
Also Published As
Publication number | Publication date |
---|---|
WO2018045902A1 (en) | 2018-03-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10466775B2 (en) | Method and apparatus for changing a field of view without synchronization with movement of a head-mounted display | |
KR102347067B1 (en) | Portable device for controlling external apparatus via gesture and operating method for same | |
CN108132053B (en) | Pedestrian track construction method and system and inertia measurement device | |
CN106525071B (en) | A kind of motion state monitoring method and device | |
CN104536558A (en) | Intelligent ring and method for controlling intelligent equipment | |
CN105807798A (en) | Method and device for controlling jittering phenomenon of head-mounted intelligent glasses | |
CN103940442A (en) | Location method and device adopting accelerating convergence algorithm | |
CN104063074B (en) | Information Display Device And Method, | |
CN106813679A (en) | The method and device of the Attitude estimation of moving object | |
Kim et al. | Golf swing analysis system with a dual band and motion analysis algorithm | |
CN106372673A (en) | Apparatus motion identification method | |
Vogt et al. | Acceleration sensors of smartphones | |
Li et al. | Real-time hand gesture tracking for human–computer interface based on multi-sensor data fusion | |
TW201118662A (en) | Trace-generating systems and methods thereof | |
CN103186255B (en) | Based on gyroscope to light target moving processing method and system, user terminal | |
CN112363196B (en) | Vehicle attribute determining method, device, storage medium and electronic equipment | |
CN103984407B (en) | The method and device of movement identification is carried out using motion sensor fusion | |
CN104765476B (en) | Handwriting tracks generation method and device | |
CN114247081B (en) | Method and system for determining motion energy consumption of intelligent stone lock | |
CN103822645B (en) | A kind of angle fusion proof of algorithm bearing calibration | |
Raharja | Development of circular motion experiment tool using sensor smartphone for high school students | |
CN106062680A (en) | Air pointer with improved user experience | |
CN108762528A (en) | Attitude algorithm method suitable for aerial flying squirrel | |
EP3669759A1 (en) | System and method for dynamic posturography | |
Bi et al. | Ramt: Real-time attitude and motion tracking for mobile devices in moving vehicle |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | 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: 20170201 |