CN106611153A - Intelligent ball training action recognition system and method - Google Patents
Intelligent ball training action recognition system and method Download PDFInfo
- Publication number
- CN106611153A CN106611153A CN201610313334.3A CN201610313334A CN106611153A CN 106611153 A CN106611153 A CN 106611153A CN 201610313334 A CN201610313334 A CN 201610313334A CN 106611153 A CN106611153 A CN 106611153A
- Authority
- CN
- China
- Prior art keywords
- action
- intelligent sphere
- data
- identification module
- training
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention discloses an intelligent ball training action recognition system and method. The intelligent ball training action recognition system comprises an intelligent ball which is equipped with a micro control unit, a nine-shaft inertial motion sensing apparatus, a clock apparatus, a storage apparatus, a battery apparatus and Bluetooth equipment, and an intelligent ball training action recognition APP, wherein the intelligent ball training action recognition APP module is arranged in mobile equipment and matched with the intelligent ball in use; the intelligent ball training action recognition APP module specifically comprises a single action recognition module, a sequence action recognition module and a technology recognition module; the intelligent ball training action recognition system also comprises a cloud end server; and the cloud end server is connected with the mobile equipment through the internet and used for receiving combined actions and various grading results transmitted by the technology recognition module to be downloaded by other mobile terminals. According to the technical scheme, by combination of the intelligent ball for collecting motion related data and the mobile equipment, recognition and evaluation on training actions of ball fans can be realized.
Description
Technical field
The present invention relates to smart motion system regions, more particularly to a kind of intelligent sphere training action identifying system and method.
Background technology
In recent years, the ball game such as basketball, football, vollyball, rugby, handball, tennis increasingly liked by masses,
Associated training is also increasingly paid attention to.For the training of action, on the one hand by the training device of specialty, standard place
And man-to-man coach carries out, but this training expenses great number and not convenient.In the case of more, ball fan is regarded by viewing
Frequency is trained come the action for imitating standard operation or soccer star.Therefore, how the feedback of echopraxia is provided to ball fan,
Ball fan is allowed to recognize the gap of oneself action and standard operation or soccer star's action, and then by constantly correction improving special project
Ability is a problem demanding prompt solution.
The content of the invention
It is an object of the invention to overcome the deficiencies in the prior art, a kind of intelligent sphere training action identifying system and method are proposed, should
The intelligent sphere for gathering motion-dependent data is combined by system with mobile device, with sensor technology, filtering, bluetooth, mathematics
Model, SVM algorithm and technical regulation, realize the identification and assessment to ball fan's training action.
The technical solution adopted for the present invention to solve the technical problems is：
A kind of intelligent sphere training action identifying system, including being provided with micro-control unit, 9 axle inertia motion sensing devices, clock
The intelligent sphere of device, storage device, cell apparatus and bluetooth equipment and intelligent sphere training action identification APP modules, the intelligence
Ball training action identification APP modules are located at mobile device, coordinate intelligent sphere to use, and specifically, the intelligent sphere training action is known
Other APP modules are specifically included：
Individual part identification module, the individual part identification module are connected with the bluetooth equipment of mobile device, receive the intelligence
The 3-axis acceleration that with the addition of timestamp of ball transmission, three axis angular rates and three axle magnetometer data are simultaneously processed, and are stored
The data model of action carries out accurately mate or carries out fuzzy matching with storage action using SVM；By matching action and the intelligence
The sensing data that energy ball is uploaded is sent to sequence identification module；
Sequence identification module, the sequence identification module are connected with the individual part identification module, receive the list
The data that the matching action of individual action recognition module transmission and the intelligent sphere are uploaded, the individual part to receiving are combined simultaneously
Compare with the rule of combination for storing the combinative movement for judging that user makes, by combinative movement and the standard operation for storing
It is compared the completeness for calculating combinative movement；The biography that combinative movement, the completeness of combinative movement and the intelligent sphere are uploaded
Sense data is activation is to technology identification module；
Technology identification module, the technology identification module are connected with the sequence identification module, are received the sequence and are known
The sensing data that the other combinative movement of module transfer, the completeness of combinative movement and the intelligent sphere are uploaded, is carried out to combinative movement
Scoring, is compared with the technical regulation for storing and obtains the concrete scoring of each training, and obtain user after carrying out weighted average
Technical merit, will identification and appraisal result show on the mobile apparatus.
Preferably, the intelligent sphere realizes data communication with the mobile device by bluetooth equipment.
Preferably, the intelligent sphere training action identifying system also includes cloud server, and the cloud server passes through
Internet is connected with the mobile device, receives the combinative movement and every appraisal result of the technology identification module transmission,
Download for other mobile terminals.
Preferably, the 9 axle inertia motion sensing device includes 9 axle inertia motion sensors and pretreatment module, described 9
Axle inertia motion sensor is connected with pretreatment module, for detecting 3-axis acceleration, three axis angular rates and three axle magnetometer number
According to and sensing data to detecting be filtered, be fitted.
A kind of intelligent sphere training action recognition methodss, will be including individual part identification module, sequence identification module and technology
The intelligent sphere training action identification APP modules of identification module are located at mobile device, and intelligent sphere training action identification APP modules pass through
Bluetooth equipment realizes data communication with intelligent sphere, and concrete grammar includes：
Individual part identification module receives the 3-axis acceleration that with the addition of timestamp, three axis angular rates and the three axle magnetic that intelligent sphere sends
Power is counted and is processed, and is carried out accurately mate with the data model of storage action or is carried out with storage action using SVM
Fuzzy matching；The sensing data that matching action and the intelligent sphere are uploaded is sent to into sequence identification module；
Sequence identification module receives the biography that the matching action of the individual part identification module transmission and the intelligent sphere are uploaded
Sense data, the individual part to receiving are combined and are compared with the rule of combination for storing the group of judging that user makes
Conjunction action, combinative movement and the standard operation for storing are compared the completeness for calculating combinative movement；By combinative movement,
The sensing data that the completeness of combinative movement and the intelligent sphere are uploaded is sent to technology identification module；
Technology identification module receives the combinative movement of sequence identification module transmission, the completeness of combinative movement and described
The sensing data that intelligent sphere is uploaded, scores to combinative movement, compares with the technical regulation for storing and obtains each training
Concrete scoring, and after carrying out weighted average, obtain the technical merit of user, identification and appraisal result shown on the mobile apparatus.
Further, the accurately mate referred to by being modeled to sensing data, finds out the data model of every class action；
The eigenvalue sometime put is calculated according to the corresponding data model of action, and the threshold value for having been arranged with the action is compared,
If eigenvalue exceedes threshold value, it is inferred to corresponding actions and has occurred and that.
Further, the accurately mate also includes：If at the appointed time the outer eigenvalue of scope is pushed away again above threshold value
Break and corresponding action repetition generation.
Further, the fuzzy matching is specifically included：
Exceed the time point t that the resultant acceleration threshold value for having arranged derives data acquisition according to the resultant acceleration for calculating；
The sensing data in a certain section of continuous time section before and after collection t is used as training data；
Function f (x)=W with regard to x is setTX+b, finds a hyperplane using SVM and meets yi*f (x)>=1, and meet feature
It is worth the distance maximization for 1 and -1 two vector space；Wherein X represents training data, and yi is equal to 1 or -1；
After training data is input into, WTIn parameter constantly correct, data model is continued to optimize, when new training data is transfused to
To in model, if the f (x) for obtaining be on the occasion of, show action occur, if f (x) is negative value, show that action does not occur.
Further, the rule of combination includes：The quantitative range of action, the persistent period of action, the interval time of action and
The order of action.
Further, completeness S of the combinative movement is calculated with equation below：
Wherein, wiRepresent the weight of each completeness index, viRepresent the measured value of each index, viRepresent standard index
Measured value.
The beneficial effect brought of technical scheme that the present invention is provided is：
(1) 3-axis acceleration, three axis angular rates and the three axle magnetometer data that with the addition of timestamp are combined, using accurately mate
The identification to individualized training action is realized with fuzzy matching；
(2) rule of combination by being combined to individual part and with storing is compared the identification realized to combinative movement,
And completeness is estimated；
(3) combine the technical regulation that stored specifically to score each training action and comprehensive grading on the mobile apparatus
Show, so as to the feedback of echopraxia is provided for ball fan, allow ball fan recognize oneself action and standard operation or
The gap of soccer star's action, and then by constantly correction improving special ability.
The present invention is described in further detail below in conjunction with drawings and Examples, but a kind of intelligent sphere training action of the present invention is known
Other system and method are not limited to embodiment.
Description of the drawings
Composition figures of the Fig. 1 for present system；
Composition frame charts of the Fig. 2 for present system intelligent sphere；
Fig. 3 is that present system intelligent sphere training action recognizes APP module composition frame charts；
Fig. 4 is that present system intelligent sphere training action recognizes APP resume module logical schematics；
Intelligent sphere training action identification APP module workflow diagrams of the Fig. 5 for the inventive method.
Specific embodiment
Referring to Fig. 1, a kind of intelligent sphere training action identifying system of the present invention, including：Intelligent sphere 10, the first mobile device 11,
Cloud server 12 and the second mobile device 13, first mobile device 11 are provided with intelligent sphere training action identification APP
Module.
Specifically, referring to Fig. 2, the intelligent sphere 10 includes：Spheroid 20 and the micro-control unit 21 being installed in spheroid 20,
Inertial Sensor Unit, clock apparatus 24, storage device 25, wireless charging receiver 26, cell apparatus 27 and bluetooth set
Standby 28.
The Inertial Sensor Unit includes 9 axle motional inertia sensors 22 and pretreatment unit 23, and the 9 axle inertia motion is passed
Sensor 22 is connected with pretreatment unit 23, for detecting 3-axis acceleration, three axis angular rates and three axle magnetometer data right
The sensing data for detecting is filtered, is fitted, and the pretreatment unit 23 is connected with the micro-control unit 21.
The clock apparatus 24 are connected for providing clock signal with micro-control unit 21, so as to generation time stamp is record ball
The motion change of body 20 provides corresponding time relationship.The wireless charging receiver 26 is connected for right with cell apparatus 27
Cell apparatus 27 carry out wireless charging.The storage device 25 is connected with micro-control unit 21, for serving as the interim of data
Storage medium, when the network connection of bluetooth equipment is unstable or goes beyond the scope, interim storage motion-dependent data, to guarantee fortune
Dynamic data are not lost.The cell apparatus 27 are connected to provide power supply with above-mentioned each device.
The micro-control unit 21 be connected with bluetooth equipment 28 for will motion-dependent data and clock signal knot merga pass it is blue
Tooth equipment sends, and the micro-control unit 21 can adopt chip microcontroller.The bluetooth equipment 28 can be used to realize moving with first
The data communication of dynamic equipment 11, including send the motion-dependent data containing timestamp and receive control command, the control command
Title, setting data collection cycle, setting connection including setting ball etc..
In the present embodiment, after 9 axle motional inertia sensor, 22 data are collected, the filtering fitting of pretreatment unit 23 is (such as karr
Graceful filtering algorithm) the first mobile device 11 is sent to by the bluetooth equipment inside spheroid 20 afterwards.In first mobile device 11
Bluetooth equipment and intelligent sphere bluetooth equipment set up connect and be responsible for receive sensing data.Reception mode can be it is real-time,
Can also be batch (batch data receive need using flash make data keep in).Intelligent sphere in first mobile device 11
Training action identification APP modules are monitored and get sensing data and carry out data analysiss, and data flow can be by intelligent sphere training action
Data model in identification APP modules is matched, and draws matching result.The result of matching is by the first mobile device 11
The network equipment (such as 3G, 4G, wifi module) uploads to cloud server 12, and the data of cloud server 12 can be by
The crawl of two mobile devices, 13 client carries out data comparison.
By 9 axle inertia motion sensors, the data of 3-axis acceleration, angular acceleration and magnetometer can be collected.Intelligent sphere
10 groups of data are spread out of disorder of internal organs is become each time, respectively:Ts, ax, ay, az, gx, gy, gz, mx, my, mz, respectively
Represent：X-axis acceleration that timestamp, 9 axle inertia motion sensors spread out of, y-axis acceleration, z-axis acceleration, x-axis angle
Speed, y-axis angular velocity, z-axis angular velocity, x-axis magnetometer data, y-axis magnetometer data, z-axis magnetometer data, use tsn、
axn、ayn、azn、gxn、gyn、gzn、mxn、myn、mznRepresent the data that sensor n-th spreads out of.Such data meeting
The intelligent sphere training action that the first mobile device 11 is input to by bluetooth equipment is recognized in APP module oracle listeners.
Specifically, the intelligent sphere training action identification APP modules include that individual part identification module 31, sequence is recognized
Module 32 and technology identification module 33, Fig. 3 are that intelligent sphere training action recognizes APP resume module logical schematics.
Referring to Fig. 4, intelligent sphere training action identification APP module workflows of the present invention are as follows：
1st, open application program.Intelligent sphere training action identification APP modules carry out bluetooth connection with intelligent sphere is connected.
2nd, individual part cognitive phase.Individual part identification module 31 receives three axles that with the addition of timestamp of intelligent sphere transmission and adds
Speed, three axis angular rates and three axle magnetometer data are simultaneously processed, and are carried out with the data model for being stored in action in data base
Accurately mate carries out fuzzy matching with storage action using SVM；The sensing data that matching action and the intelligent sphere are uploaded
Store database concurrency and be sent to sequence identification module 32.
Individual part identification module 31 carries out accurately mate or fuzzy matching by existing action model, is implemented as follows：
(1) method of accurately mate finds out the data model of all kinds of actions generally by data modeling, these data models
If similar with the feature of input data, it can be inferred that corresponding actions have occurred and that.Such action include dribble on basketball,
Around ring, shooting, and dialling ball, juggle, shoot on football.
By taking a simple data model as an example, we can judge the fortune of ball by the change angle of resultant acceleration and direction of bowl
The generation of action F.
Wherein, atAnd btAll it is resultant acceleration, the resultant acceleration is that 3-axis acceleration quadratic sum opens radical sign, inside intelligent sphere ball
Multiple sensors can be embedded in, have plenty of high-precision, have plenty of high measurement scope, three axis angular rate therein is passed for high accuracy
Sensor collection；T represents some time point；FCtRepresent that the sensor obtained by some time point is directed to some feature
The eigenvalue of C (such as dribble and group ball of football of basketball).We can arrange a threshold values T (threshold) for C,
When the eigenvalue of a certain group of numerical value is F more than T valueCt>TCWhen, then it is assumed that this feature there occurs.
Further, another time range S is settIf, in the outer (t-t of the scopeC>St, tCRepresent the last feature C
The time of generation) above formula set up again, then and just repetition there occurs this feature C (kinestate).
(2) fuzzy matching is drawn by the sorting algorithm analysis of machine learning.It is following by example illustrating how into action
Make the identification of (ball kinestate).
So that the fixed point of basketball shoots basket as an example, to carry out whether judgement is knocked down to ball.Using SVM come according to 3-axis acceleration and
Three axis angular rate ax, ay, az, gx, gy, gz data carry out following judgement：
Data acquisition first is carrying out according to the resultant acceleration calculated after data filtering more than threshold values.Assume at some
Time point t0Resultant acceleration at0>aT, then think that basketball reaches the vicinity (beat on frame or network) of frame or ball exists
Land after selling.Pushed away from this time point forward, in interval (t-St,t0) in the range of, if also had so in time point t
Feature (at>aT), then it is assumed that t is the main time points of actual data acquisition, wherein aTAnd StIt is constant.
The sensor array selected in a certain section of section continuous time (t-m, t+n) before and after t is acquired, with a, gx,
Gy, gz are used as variable (data matrix that these variables are constituted is the training data of SVM), and record whether shooting every time enters
Net.
Training data is as follows：
1)xi, i=0 ..., 4* (m+n) (because having 4 accekerations every time taking advantage of 4)；
2)yi=1or-1；If networking is exactly 1, otherwise it is exactly -1.
Hypothesis is related to function f (the x)=W of xTThe target of X+b, SVM is to find a hyperplane, meets yi*f (x)>=1, and
And allow eigenvalue to be 1 and the maximization of the distance of -1 two vector space.After training data is input into, WTIn parameter can be continuous
It is corrected, data model can be continued to optimize.When one group of new X data are imported in model, if the f (x) for obtaining is
On the occasion of, then show that basketball networks, if negative value, then ball does not hit.
3rd, sequence cognitive phase.Sequence identification module 32 receive the individual part identification module 31 transmission
With the sensing data that action and the intelligent sphere are uploaded, the individual part to receiving be combined and with the rule of combination for storing
Compare the combinative movement for judging that user makes, combinative movement is compared to calculate with the standard operation for storing and is combined
The completeness of action；The sensing data that combinative movement, the completeness of combinative movement and the intelligent sphere are uploaded is sent to into technology knowledge
Other module.
32 concrete processing procedure of sequence identification module is as follows：
The result of accurately mate and fuzzy matching is to find action one by one.According to these actions, sequence identification module
32 can learn a series of combinative movements that user makes.Standard combination action and soccer star's combinative movement are defined within rule of combination
In, by taking superstar's combinative movement of basketball as an example, behind-the-back dribble, cross-leg dribble in storehouse, can be made, one is carried out after 5 to 8 times repeatedly
Secondary waist is finally laid up around ring.Rule of combination includes being defined as below：
1) quantitative range of action, such as 5 to 8 times dribbles；
2) persistent period of action, such as the regular hour limits around ring；
3) interval time of action, such as per double dribble, interval was less than 0.8 second；
4) order of action, such as dribble must be around rings and before laying up；
Sequence identification module 32 can do basic judgement according to these rules to the action completeness of user.Generally user is complete
Into data closer to standard operation or soccer star's action, then completeness is higher.Completeness can be calculated with the following methods：
wiRepresent the weight of each completeness index, viRepresent the measured value of each index, vsThe numerical value of expression standard, the two it
Between difference it is less, then the completeness of user is higher.
4th, technology cognitive phase.Technology identification module 33 receive the sequence identification module 32 transmission combinative movement,
The sensing data that the completeness of combinative movement and the intelligent sphere are uploaded, scores to combinative movement, with the technology rule for storing
Then compare and obtain the concrete scoring of each training, and after carrying out weighted average, obtain the technical merit of user, will recognize and comment
Result is divided to show on the first mobile device.
33 concrete processing procedure of technology identification module is as follows：
The action matching of user will be through the final process of technology identification module.The process logic of technology identification module 33 is also base
In certain sports rule/logic.By taking basketball as an example, the technical regulation of the impact user's final score that can be added includes：
1) dribble strength.In general, strength is bigger and will not lose ball and then proves that the ability of controlling the ball of user is stronger.
2) angle dribbled.This standard embodies cross edge wire spoke degree, and in general, amplitude is more big, and defender is taken advantage of
Deceive property higher.
3) dribble terminates or terminates to the time laid up around ring.This index embodies the stagnant empty ability of user.
4) rotation of ball.When laying up, selling for ball carry certain outward turning.
Corresponding score interval can be given to the technical regulation of each thin item, each skill of user can be obtained according to score interval
The concrete scoring of art, is exactly the technical merit of user after being weighted averagely.
After the assessment that technology identification module 33 completes last ins and outs, it is complete that last mobile device program provides user one
It is whole comprehensively to compare data, including the scoring of each action details, skill score, continuity scoring and comprehensive grading.User can
To see oneself strong point or weakness, it is trained and improves so as to more targeted.
The foregoing is only presently preferred embodiments of the present invention, not to limit the present invention, it is all the spirit and principles in the present invention it
Interior, any modification, equivalent substitution and improvements made etc. should be included within the scope of the present invention.
Claims (10)
1. a kind of intelligent sphere training action identifying system, including be provided with micro-control unit, 9 axle inertia motion sensing devices,
The intelligent sphere of clock apparatus, storage device, cell apparatus and bluetooth equipment and intelligent sphere training action identification APP modules, it is described
Intelligent sphere training action identification APP modules are located at mobile device, coordinate intelligent sphere to use, it is characterised in that the intelligent sphere instruction
Practice action recognition APP module to specifically include：
Individual part identification module, the individual part identification module are connected with the bluetooth equipment of mobile device, receive the intelligence
The 3-axis acceleration that with the addition of timestamp of ball transmission, three axis angular rates and three axle magnetometer data are simultaneously processed, and are stored
The data model of action carries out accurately mate or carries out fuzzy matching with storage action using SVM；By matching action and the intelligence
The sensing data that energy ball is uploaded is sent to sequence identification module；
Sequence identification module, the sequence identification module are connected with the individual part identification module, receive the list
The data that the matching action of individual action recognition module transmission and the intelligent sphere are uploaded, the individual part to receiving are combined simultaneously
Compare with the rule of combination for storing the combinative movement for judging that user makes, by combinative movement and the standard operation for storing
It is compared the completeness for calculating combinative movement；The biography that combinative movement, the completeness of combinative movement and the intelligent sphere are uploaded
Sense data is activation is to technology identification module；
Technology identification module, the technology identification module are connected with the sequence identification module, are received the sequence and are known
The sensing data that the other combinative movement of module transfer, the completeness of combinative movement and the intelligent sphere are uploaded, is carried out to combinative movement
Scoring, is compared with the technical regulation for storing and obtains the concrete scoring of each training, and obtain user after carrying out weighted average
Technical merit, will identification and appraisal result show on the mobile apparatus.
2. intelligent sphere training action identifying system according to claim 1, it is characterised in that：The intelligent sphere is by indigo plant
Tooth equipment realizes data communication with the mobile device.
3. intelligent sphere training action identifying system according to claim 1, it is characterised in that：The intelligent sphere training is dynamic
Making identifying system also includes cloud server, and the cloud server is connected with the mobile device by Internet, is received
The combinative movement and every appraisal result of the technology identification module transmission, downloads for other mobile terminals.
4. intelligent sphere training action identifying system according to claim 1, it is characterised in that：The 9 axle inertia motion
Sensing device includes 9 axle inertia motion sensors and pretreatment module, the 9 axle inertia motion sensor and pretreatment module phase
Connection, the sensing data for detecting 3-axis acceleration, three axis angular rates and three axle magnetometer data and to detecting is filtered,
Fitting.
5. a kind of intelligent sphere training action recognition methodss, it is characterised in that：Will be including individual part identification module, sequence
The intelligent sphere training action identification APP modules of identification module and technology identification module are located at mobile device, and intelligent sphere training action is known
Other APP modules realize data communication by bluetooth equipment and intelligent sphere, and concrete grammar includes：
Individual part identification module receives the 3-axis acceleration that with the addition of timestamp, three axis angular rates and the three axle magnetic that intelligent sphere sends
Power is counted and is processed, and is carried out accurately mate with the data model of storage action or is carried out with storage action using SVM
Fuzzy matching；The sensing data that matching action and the intelligent sphere are uploaded is sent to into sequence identification module；
Sequence identification module receives the biography that the matching action of the individual part identification module transmission and the intelligent sphere are uploaded
Sense data, the individual part to receiving are combined and are compared with the rule of combination for storing the group of judging that user makes
Conjunction action, combinative movement and the standard operation for storing are compared the completeness for calculating combinative movement；By combinative movement,
The sensing data that the completeness of combinative movement and the intelligent sphere are uploaded is sent to technology identification module；
Technology identification module receives the combinative movement of sequence identification module transmission, the completeness of combinative movement and described
The sensing data that intelligent sphere is uploaded, scores to combinative movement, compares with the technical regulation for storing and obtains each training
Concrete scoring, and after carrying out weighted average, obtain the technical merit of user, identification and appraisal result shown on the mobile apparatus.
6. intelligent sphere training action recognition methodss according to claim 5, it is characterised in that：The accurately mate is referred to
By being modeled to sensing data, the data model of every class action is found out；Certain is calculated according to the corresponding data model of action
The eigenvalue of one time point, and the threshold value for having been arranged with the action is compared, if eigenvalue exceedes threshold value, is inferred to phase
Action is answered to have occurred and that.
7. intelligent sphere training action recognition methodss according to claim 6, it is characterised in that the accurately mate is also wrapped
Include：If at the appointed time the outer eigenvalue of scope is inferred to corresponding action and repeats generation again above threshold value.
8. intelligent sphere training action recognition methodss according to claim 5, it is characterised in that the fuzzy matching is concrete
Including：
Exceed the time point t that the resultant acceleration threshold value for having arranged derives data acquisition according to the resultant acceleration for calculating；
The sensing data in a certain section of continuous time section before and after collection t is used as training data；
Function f (x)=W with regard to x is setTX+b, finds a hyperplane using SVM and meets yi*f (x)>=1, and meet feature
It is worth the distance maximization for 1 and -1 two vector space；Wherein X represents training data, and yi is equal to 1 or -1；
After training data is input into, WTIn parameter constantly correct, data model is continued to optimize, when new training data is transfused to
To in model, if the f (x) for obtaining be on the occasion of, show action occur, if f (x) is negative value, show that action does not occur.
9. intelligent sphere training action recognition methodss according to claim 5, it is characterised in that the rule of combination includes：
The order of the quantitative range of action, the persistent period of action, the interval time of action and action.
10. intelligent sphere training action recognition methodss according to claim 9, it is characterised in that：The combinative movement it is complete
Calculated into degree S with equation below：
Wherein, wiRepresent the weight of each completeness index, viRepresent the measured value of each index, viRepresent the survey of standard index
Value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610313334.3A CN106611153B (en) | 2016-05-12 | 2016-05-12 | Intelligent ball training action recognition system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610313334.3A CN106611153B (en) | 2016-05-12 | 2016-05-12 | Intelligent ball training action recognition system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106611153A true CN106611153A (en) | 2017-05-03 |
| CN106611153B CN106611153B (en) | 2020-01-14 |
Family
ID=58614790
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610313334.3A Active CN106611153B (en) | 2016-05-12 | 2016-05-12 | Intelligent ball training action recognition system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106611153B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107248124A (en) * | 2017-05-22 | 2017-10-13 | 浙江精益佰汇数字技术有限公司 | The mobile terminal and implementation method of aided education |
| CN107456748A (en) * | 2017-10-12 | 2017-12-12 | 周口师范学院 | A kind of digital volleyball training system |
| CN108721872A (en) * | 2018-07-30 | 2018-11-02 | 上海赢臻体育用品有限公司 | A kind of scoring system based on bluetooth |
| CN108921127A (en) * | 2018-07-19 | 2018-11-30 | 上海小蚁科技有限公司 | Method for testing motion and device, storage medium, terminal |
| CN109784133A (en) * | 2017-11-15 | 2019-05-21 | 财团法人资讯工业策进会 | Act evaluation model generating means and its movement evaluation model generating method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102221369A (en) * | 2011-04-29 | 2011-10-19 | 韩铮 | Gesture recognizing method and device of ball game and gesture auxiliary device |
| CN104524749A (en) * | 2015-01-09 | 2015-04-22 | 李漩 | Digital smart basketball |
| CN204680194U (en) * | 2015-06-25 | 2015-09-30 | 厦门市简极科技有限公司 | A kind of intelligent sphere and system thereof |
| CN105118253A (en) * | 2015-06-25 | 2015-12-02 | 厦门市简极科技有限公司 | Intelligent ball and system thereof |
| CN105107178A (en) * | 2015-08-03 | 2015-12-02 | 厦门市简极科技有限公司 | Shooting action training method |
| JP2016028712A (en) * | 2015-09-29 | 2016-03-03 | 株式会社三共 | Game machine |
-
2016
- 2016-05-12 CN CN201610313334.3A patent/CN106611153B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102221369A (en) * | 2011-04-29 | 2011-10-19 | 韩铮 | Gesture recognizing method and device of ball game and gesture auxiliary device |
| CN104524749A (en) * | 2015-01-09 | 2015-04-22 | 李漩 | Digital smart basketball |
| CN204680194U (en) * | 2015-06-25 | 2015-09-30 | 厦门市简极科技有限公司 | A kind of intelligent sphere and system thereof |
| CN105118253A (en) * | 2015-06-25 | 2015-12-02 | 厦门市简极科技有限公司 | Intelligent ball and system thereof |
| CN105107178A (en) * | 2015-08-03 | 2015-12-02 | 厦门市简极科技有限公司 | Shooting action training method |
| JP2016028712A (en) * | 2015-09-29 | 2016-03-03 | 株式会社三共 | Game machine |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107248124A (en) * | 2017-05-22 | 2017-10-13 | 浙江精益佰汇数字技术有限公司 | The mobile terminal and implementation method of aided education |
| CN107248124B (en) * | 2017-05-22 | 2021-10-01 | 郭山河 | Mobile terminal for assisting teaching and implementation method |
| CN107456748A (en) * | 2017-10-12 | 2017-12-12 | 周口师范学院 | A kind of digital volleyball training system |
| CN107456748B (en) * | 2017-10-12 | 2019-05-07 | 周口师范学院 | A kind of number volleyball training system |
| CN109784133A (en) * | 2017-11-15 | 2019-05-21 | 财团法人资讯工业策进会 | Act evaluation model generating means and its movement evaluation model generating method |
| CN109784133B (en) * | 2017-11-15 | 2021-11-02 | 财团法人资讯工业策进会 | Action evaluation model generation device and action evaluation model generation method thereof |
| CN108921127A (en) * | 2018-07-19 | 2018-11-30 | 上海小蚁科技有限公司 | Method for testing motion and device, storage medium, terminal |
| CN108721872A (en) * | 2018-07-30 | 2018-11-02 | 上海赢臻体育用品有限公司 | A kind of scoring system based on bluetooth |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106611153B (en) | 2020-01-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106611153A (en) | Intelligent ball training action recognition system and method | |
| US11219814B2 (en) | Autonomous personalized golf recommendation and analysis environment | |
| CN104007822B (en) | Motion recognition method and its device based on large database concept | |
| CN106464848B (en) | Motion event identification and video synchronizing system and method | |
| US20220362629A1 (en) | Method, apparatus, and computer program product for measuring and interpreting metrics of an athletic action and an object associated therewith | |
| US20100210975A1 (en) | Multi-state performance monitoring system | |
| CN108703760A (en) | Human motion gesture recognition system and method based on nine axle sensors | |
| CN105612475A (en) | Wrist-worn athletic device with gesture recognition and power management | |
| CN106874874A (en) | A kind of recognition methods of motion state and device | |
| CN103699795A (en) | Exercise behavior identification method and device and exercise intensity monitoring system | |
| CN104524760B (en) | A kind of method and system that basketball action is judged based on Intelligent bracelet | |
| CN110276384A (en) | The method, apparatus and system with annotation capture and movable group modeling for sensing data | |
| CN106067001A (en) | A kind of action identification method and system | |
| CN105651303A (en) | Pace counting system and pace counting method based on three-axis acceleration sensor | |
| CN108827290A (en) | A kind of human motion state inverting device and method | |
| CN105597296A (en) | Equipment and method for detecting movement | |
| US20220161121A1 (en) | Autonomous Tracking and Personalized Golf Recommendation and Analysis Environment | |
| CN109646902A (en) | A kind of body building metering method based on identification equipment | |
| CN106422211A (en) | Technical statistical method and device for ball training | |
| Bai et al. | Application and research of MEMS sensor in gait recognition algorithm | |
| CA3174383A1 (en) | Methods and system for training and improving machine learning models | |
| CN110013642B (en) | Method and system for determining motion state of intelligent stone lock and corresponding device | |
| CN107679623A (en) | A kind of intelligent sphere, system and method | |
| CN109758745A (en) | Artificial intelligence basketball training system based on Python/Java | |
| Chen | A Data Mining‐Based Model for Evaluating Tennis Players’ Training Movements |
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 |