CN107273857A - The recognition methods of athletic performance and device, electronic equipment - Google Patents
The recognition methods of athletic performance and device, electronic equipment Download PDFInfo
- Publication number
- CN107273857A CN107273857A CN201710463245.1A CN201710463245A CN107273857A CN 107273857 A CN107273857 A CN 107273857A CN 201710463245 A CN201710463245 A CN 201710463245A CN 107273857 A CN107273857 A CN 107273857A
- Authority
- CN
- China
- Prior art keywords
- action
- type
- signal
- probability
- athletic performance
- 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
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Abstract
The disclosure discloses recognition methods and device, electronic equipment, the computer-readable recording medium of a kind of athletic performance, and this method includes:The motion process that moving object tracking is performed, obtains original motion signal;Original motion signal is converted by default change scaling method, some features of original motion signal are extracted from the signal after conversion;Obtain and be pre-stored the probability that each feature occurs in various type of action;There is the maximum type of action of some feature probabilities simultaneously in screening from various type of action;Type of action according to filtering out obtains the recognition result of athletic performance.The technical scheme that the disclosure is provided, the recognition accuracy of athletic performance is high, and identification amount of calculation is small, and recognition efficiency is high.
Description
Technical field
This disclosure relates to technical field of data processing, the recognition methods of more particularly to a kind of athletic performance and device, electronics
Equipment, computer-readable recording medium.
Background technology
With comprehensive development of nationwide fitness programs, increasing people are participated among sport and body-building.With intelligence
Body-building sports apparatus also becomes very universal.Sport and body-building action is lack of standardization to cause a variety of harmful effects, such as injury gained in sports
Risk increase, muscle coordination reduction, the reduction of body-building efficiency etc..Therefore, for the body building of more efficient more standard, need pair
Human action is identified, for example, the action for waving racket is identified, cut is identified.
The identification method of existing sports equipment control action includes the motion trace data for gathering all kinds of actions, sets up sample
Database, the motion trace data for then gathering action to be measured identifies that human body manipulates sports equipment by track similarity mode
Action.This mode only can recognize that whether the movement locus of two classes action is similar, but for the similar action in track simultaneously
It can not make a distinction, thus the accuracy of action recognition is not high.
Also a kind of mode is the kinematic parameter that human action is gathered by sensor, then by extracting kinematic parameter
Characteristic value, determines the action of human body by way of characteristic value is matched.Which can be realized dynamic to human body to a certain extent
The identification of work, but be due to many actions characteristic value approach, by way of characteristic value is matched, accuracy is not also high.
To sum up, prior art is not high to the identification accuracy of sports equipment control action.
The content of the invention
The problem of in order to solve not high to the identification accuracy of athletic performance present in correlation technique, present disclose provides
The recognition methods of another athletic performance, to improve the accuracy of identification.
On the one hand, present disclose provides a kind of recognition methods of athletic performance, this method includes:
The motion process that moving object tracking is performed, obtains original motion signal;
The original motion signal is converted by default change scaling method, extracts described from the signal after conversion
Some features of original motion signal;
Obtain and be pre-stored the probability that each feature occurs in various type of action;
There is the maximum type of action of some feature probabilities simultaneously in screening from various type of action;
The type of action according to filtering out obtains the recognition result of athletic performance.
In one embodiment, there are some feature probability maximums simultaneously in the screening from various type of action
Type of action, including:
The probability occurred according to each feature in various type of action, if calculating each type of action simultaneously in the presence of described
The probability of dry feature;
There is the probability of some features simultaneously according to each type of action, filter out the maximum action class of probability
Type.
In one embodiment, the type of action that the basis is filtered out obtains the recognition result of athletic performance, bag
Include:
Judge whether the type of action filtered out is more than one kind, if it is not, the type of action conduct then filtered out
The recognition result of the athletic performance;
If so, original motion signal sample motor message corresponding with each type of action filtered out is carried out into waveform
Matching, searches the sample motor message with the original motion Signal Matching, and the sample motor message of the matching is corresponding
Type of action as the athletic performance recognition result.
In one embodiment, the motion process that the moving object tracking is performed, obtains original motion signal, bag
Include:
Pass through the 3-axis acceleration component of Moving Objects described in the sensor continuous collecting on the Moving Objects and angle speed
Degree;
Posture and the position of the Moving Objects are obtained according to the 3-axis acceleration component and angular speed of collection;
Acceleration is obtained according to the 3-axis acceleration component of the Moving Objects, angular speed, posture and the change of position to become
Change waveform signal, angular speed change waveform signal, attitudes vibration waveform signal and change in location waveform signal.
In one embodiment, the original motion signal is converted by default change scaling method, after conversion
Signal in extract some features of the original motion signal, including:
The acceleration change waveform signal, angular speed change waveform signal, attitudes vibration waveform signal and position are become
Change waveform signal and carry out waveform conversion, some features of waveform signal after being converted by a variety of operation rules.
On the other hand, the disclosure has additionally provided a kind of identifying device of athletic performance, and the device includes:
Signal acquisition module, the motion process being performed for moving object tracking obtains original motion signal;
Characteristic extracting module, for the original motion signal to be converted by default change scaling method, from conversion
Some features of the original motion signal are extracted in signal afterwards;
Probability acquisition module, for obtaining the probability that pre-stored each feature occurs in various type of action;
Screening module is acted, for the screening from various type of action while there is the dynamic of some feature probability maximums
Make type;
Action recognition module, the recognition result for obtaining athletic performance according to the type of action filtered out.
In one embodiment, the action screening module includes:
Probability calculation unit, for appearing in the probability in various type of action according to each feature, calculates each action
There is the probability of some features simultaneously in type;
Unit is chosen in action, for there is the probability of some features, screening simultaneously according to each type of action
Go out the maximum type of action of probability.
In one embodiment, the action recognition module includes:
Whether judging unit, the type of action for judging to filter out is more than one kind, if it is not, what is then filtered out is described
Type of action as the athletic performance recognition result;
Matching unit, for by original motion signal sample motor message corresponding with each type of action filtered out
Waveform Matching is carried out, the sample motor message with the original motion Signal Matching is searched, the sample of the matching is moved and believed
Number corresponding type of action as the athletic performance recognition result.
Another further aspect, the disclosure has additionally provided a kind of electronic equipment, and the electronic equipment includes:
Processor;
Memory for storing processor-executable instruction;
Wherein, the processor is configured as performing the recognition methods of any one above-mentioned athletic performance.
In addition, the disclosure has additionally provided a kind of computer-readable recording medium, the computer-readable recording medium is deposited
Computer program is contained, the computer program can be completed the identification side of any one above-mentioned athletic performance by computing device
Method.
The technical scheme provided by this disclosed embodiment can include the following benefits:
The disclosure enters line translation by the original motion signal to Moving Objects, extracts some spies of original motion signal
Levy, then according to the probability that each feature occurs in known every kind of type of action, will occur some feature probabilities simultaneously maximum
Type of action as athletic performance recognition result.The program due to without by the waveform signal of action to be measured one by one with it is all
The waveform signal of known action is matched, and the probability occurred in known each action according to some features of unknown action is just
The identification of action can be realized, amount of calculation is smaller, recognition efficiency is high, further as action identification independent of movement locus
Matched with characteristic value, for the similar action in track or the close action of characteristic value, by some spies for calculating unknown action
The probability appeared in known action is levied, the identification of action can also be realized, thus the identification accuracy of athletic performance is high.
It should be appreciated that the general description of the above and detailed description hereinafter are only exemplary, this can not be limited
It is open.
Brief description of the drawings
Accompanying drawing herein is merged in specification and constitutes the part of this specification, shows the implementation for meeting the present invention
Example, and in specification together for explaining principle of the invention.
Fig. 1 is the schematic diagram of the implementation environment according to involved by the disclosure;
Fig. 2 is a kind of block diagram of device according to an exemplary embodiment;
Fig. 3 is a kind of flow chart of the recognition methods of athletic performance according to an exemplary embodiment;
Fig. 4 be Fig. 3 correspondence embodiment in step 310 details flow chart;
Fig. 5 is that a kind of complicated waveform signal is carried out forming tool after waveform conversion according to an exemplary embodiment
There is the schematic diagram of the waveform signal of obvious characteristic;
Fig. 6 be according to an exemplary embodiment to original motion signal according to specified rule carry out waveform become swap-in
The principle schematic of row feature extraction;
Fig. 7 be Fig. 3 correspondence embodiment in step 370 details flow chart
Fig. 8 is a kind of block diagram of the identifying device of athletic performance according to an exemplary embodiment;
Fig. 9 is the details block diagram of the action screening module of Fig. 8 correspondence embodiments;
Figure 10 is the details block diagram of the signal acquisition module of Fig. 8 correspondence embodiments.
Embodiment
Here explanation will be performed to exemplary embodiment in detail, its example is illustrated in the accompanying drawings.Following description is related to
During accompanying drawing, unless otherwise indicated, the same numbers in different accompanying drawings represent same or analogous key element.Following exemplary embodiment
Described in embodiment do not represent and the consistent all embodiments of the present invention.On the contrary, they be only with it is such as appended
The example of the consistent apparatus and method of some aspects be described in detail in claims, the present invention.
Fig. 1 is the schematic diagram of the implementation environment according to involved by the disclosure.The implementation environment includes:At least one movement is eventually
End 110 and smart machine 120.
Interrelational form between mobile terminal 1 10 and smart machine 120, including the network associate mode of hardware and/or association
View, and the data correlation mode come and gone therebetween.Smart machine 120 according to the difference of its type of athletic equipment that is applicable,
The position of installing is also different.
Smart machine 120 can be arranged in the racket handle of shuttlecock, tennis, table tennis, can also be arranged on baseball
In rod, football, basketball inside etc. can also be arranged on.Further, the smart machine 120 can also be applied to wall ball racket, plate
In racket, golf club etc. sports equipment.In other words, the smart machine 120 can be arranged on the sports equipment of activity
In, according to the activity of sports equipment, identification human body manipulates the action of sports equipment.As needed, the smart machine 120 can be with
Wrist or the ankle position of human body are worn on, according to the activity of wrist or ankle, action when identification human body performs physical exercises.
For example, waving the action of racket, the action for waving baseball rod, the action played soccer, action of shooting etc..
After smart machine 120 is connected with the foundation of mobile terminal 1 10, the recognition result of athletic performance can be uploaded to movement
Terminal 110 is shown and stored.Bluetooth connection can be used between smart machine 120 and mobile terminal 1 10.
Fig. 2 is a kind of block diagram of device 200 according to an exemplary embodiment.For example, device 200 can be Fig. 1
Smart machine 120 in shown implementation environment.
Reference picture 2, device 200 can include following one or more assemblies:Processing assembly 202, memory 204, power supply
Component 206, multimedia groupware 208, audio-frequency assembly 210, sensor cluster 214 and communication component 216.
The integrated operation of the usual control device 200 of processing assembly 202, such as with display, call, data communication, phase
Operation that machine is operated and record operation is associated etc..Processing assembly 202 can include one or more processors 218 to perform
Instruction, to complete all or part of step of following methods.In addition, processing assembly 202 can include one or more modules,
It is easy to the interaction between processing assembly 202 and other assemblies.For example, processing assembly 202 can include multi-media module, with convenient
Interaction between multimedia groupware 208 and processing assembly 202.
Memory 204 is configured as storing various types of data supporting the operation in device 200.These data are shown
Example includes the instruction of any application program or method for operating on the device 200.Memory 204 can be by any kind of
Volatibility or non-volatile memory device or combinations thereof realization, such as static RAM (Static Random
Access Memory, abbreviation SRAM), Electrically Erasable Read Only Memory (Electrically Erasable
Programmable Read-Only Memory, abbreviation EEPROM), Erasable Programmable Read Only Memory EPROM (Erasable
Programmable Read Only Memory, abbreviation EPROM), programmable read only memory (Programmable Red-
Only Memory, abbreviation PROM), read-only storage (Read-Only Memory, abbreviation ROM), magnetic memory, flash
Device, disk or CD.Also be stored with one or more modules in memory 204, and one or more modules are configured to by this
One or more processors 218 are performed, to complete all or part of step in any shown method of following Fig. 3, Fig. 4, Fig. 7
Suddenly.
Power supply module 206 provides electric power for the various assemblies of device 200.Power supply module 206 can include power management system
System, one or more power supplys, and other components associated with generating, managing and distributing electric power for device 200.
Multimedia groupware 208 is included in the screen of one output interface of offer between described device 200 and user.One
In a little embodiments, screen can include liquid crystal display (Liquid Crystal Display, abbreviation LCD) and touch panel.
If screen includes touch panel, screen may be implemented as touch-screen, to receive the input signal from user.Touch panel
Including one or more touch sensors with the gesture on sensing touch, slip and touch panel.The touch sensor can be with
The not only border of sensing touch or sliding action, but also the detection duration related to the touch or slide and pressure
Power.Screen can also include display of organic electroluminescence (Organic Light Emitting Display, abbreviation OLED).
Audio-frequency assembly 210 is configured as output and/or input audio signal.For example, audio-frequency assembly 210 includes a Mike
Wind (Microphone, abbreviation MIC), when device 200 is in operator scheme, such as call model, logging mode and speech recognition mould
During formula, microphone is configured as receiving external audio signal.The audio signal received can be further stored in memory
204 or sent via communication component 216.In certain embodiments, audio-frequency assembly 210 also includes a loudspeaker, for exporting
Audio signal.
Sensor cluster 214 includes one or more sensors, and the state for providing various aspects for device 200 is commented
Estimate.For example, sensor cluster 214 can detect opening/closed mode of device 200, the relative positioning of component, sensor group
Part 214 can be with the position change of 200 1 components of detection means 200 or device and the temperature change of device 200.At some
In embodiment, the sensor cluster 214 can also include Magnetic Sensor, pressure sensor or temperature sensor.
Communication component 216 is configured to facilitate the communication of wired or wireless way between device 200 and other equipment.Device
200 can access the wireless network based on communication standard, such as WiFi (WIreless-Fidelity, Wireless Fidelity).Show at one
In example property embodiment, communication component 216 receives broadcast singal or broadcast from external broadcasting management system via broadcast channel
Relevant information.In one exemplary embodiment, the communication component 216 also includes near-field communication (Near Field
Communication, abbreviation NFC) module, to promote junction service.For example, radio frequency identification (Radio can be based in NFC module
Frequency Identification, abbreviation RFID) technology, Infrared Data Association (Infrared Data
Association, abbreviation IrDA) technology, ultra wide band (Ultra Wideband, abbreviation UWB) technology, Bluetooth technology and other skills
Art is realized.
In the exemplary embodiment, device 200 can be by one or more application specific integrated circuit (Application
Specific Integrated Circuit, abbreviation ASIC), it is digital signal processor, digital signal processing appts, programmable
Logical device, field programmable gate array, controller, microcontroller, microprocessor or other electronic components are realized, for performing
Following methods.
Fig. 3 is a kind of flow chart of the recognition methods of athletic performance according to an exemplary embodiment.The motion is moved
The scope of application and executive agent of the recognition methods of work, for example, this method is used for the smart machine 120 of implementation environment shown in Fig. 1.
As shown in figure 3, the recognition methods, can be performed by smart machine 120, may comprise steps of.
Step 310, the motion process that moving object tracking is performed, obtains original motion signal.
Wherein, Moving Objects can refer to be steered the sports equipment for performing motion process or manipulation locomotory apparatus
The body parts such as wrist, the ankle of the human body of material.When the sports equipment that presence can be moved, smart machine 120 can be with
The inside of sports equipment is arranged on, by the motion process of pursuit movement equipment, original motion signal is obtained.For motionless fortune
Dynamic equipment (such as horizontal bar) or the situation (for example running) without using sports equipment, can be worn on human body by smart machine 120
Wrist or ankle position, original motion signal is obtained by the activity for tracking wrist or ankle.Specifically, smart machine
Sensor can be set in 120, and in Moving Objects active procedure, the kinematic parameter of continuous collecting Moving Objects is obtained original
Motor message.
Optionally, as shown in figure 4, step 310 is specifically included:
Step 311, the 3-axis acceleration component of the sensor continuous collecting Moving Objects on Moving Objects and angle speed are passed through
Degree;
Wherein, sensor is carried on Moving Objects, follows Moving Objects generation activity in the lump, timing acquiring Moving Objects
3-axis acceleration component.The sensor can include three axis accelerometer and gyroscope.Three axis accelerometer can be adopted persistently
Collect the 3-axis acceleration component of Moving Objects, i.e. component of the acceleration on X-axis, Y-axis and Z axis.It can then be adopted by gyroscope
Collect the angular speed of Moving Objects.
Step 312, according to the 3-axis acceleration component and angular speed of collection obtain the Moving Objects posture and
Position;
Wherein, the posture of Moving Objects refers to the inclination angle of Moving Objects in three dimensions.By merging 3-axis acceleration
Component and angular velocity measurement result, can obtain the angle of pitch, yaw angle and the roll angle of smart machine 120 in three dimensions.
MCU (microprogram control unit) in smart machine 120 can be using existing secondary expansion Kalman filtering algorithm according to three axles
The measurement result of accelerometer and gyroscope resolves the angle of pitch, yaw angle and the roll angle of Moving Objects.Wherein, Moving Objects exist
In three dimensions around x-axis rotation angle θ be the angle of pitch, around the y-axis anglec of rotationIt is yaw angle around z-axis anglec of rotation γ for roll angle.Fortune
The angle of pitch, yaw angle and the roll angle of dynamic object form the posture of the moment Moving Objects.
In addition, smart machine 120 since inactive state from upper electric or a certain trigger point (being such as changed into motion state), it is assumed that
Initial position is the origin of coordinates, component of the acceleration obtained according to three axis accelerometer measurement on X-axis, Y-axis and Z axis, profit
Use formulaCoordinate of the smart machine 120 on X-axis, Y-axis and Z axis can be calculated respectively, so as to obtain intelligence
The each moment of equipment 120, in the coordinate of three dimensions, the position at the moment is just formed in the coordinate of three dimensions.
Step 313, accelerated according to the 3-axis acceleration component of Moving Objects, angular speed, posture and the change of position
Degree change waveform signal, angular speed change waveform signal, attitudes vibration waveform signal and change in location waveform signal.
Wherein, the 3-axis acceleration component and angular speed of sensor timing acquiring Moving Objects, is moved according to each moment
3-axis acceleration component, angular speed, posture and the change of position of object, can form acceleration change waveform signal, angle speed
Degree change waveform signal, attitudes vibration waveform signal and change in location waveform signal.It should be noted that Moving Objects are in difference
The location of moment constitutes the change in location waveform signal by line.These waveform signals can be stored in smart machine
In 120 RAM (random access memory).It should be noted that the component of acceleration on X-axis, Y-axis and Z axis can by synthesis
To obtain linear acceleration change wavy curve.
Step 330, original motion signal is converted by default change scaling method, extracted from the signal after conversion
Some features of the original motion signal;
It is to be understood that original motion signal carries out various conversion by default change scaling method, can be formed has
The waveform signal of obvious characteristic.The default change scaling method refers to original motion signal can be made to be formed with obvious by conversion
The change scaling method of characteristic waveform signal.For example, as shown in figure 5, left figure is the wavy curve of original motion signal, due to very
Hardly possible describes the feature of the wavy curve, therefore can convert the wavy curve by time domain frequency domain, obtains right figure waveform.It is very aobvious
So, right figure wave character is obvious, so as to obtain some features of original motion signal from the waveform signal after conversion.Example
Such as maximum, minimum value.
Optionally, step 330 is specifically included:By the acceleration change waveform signal, angular speed change waveform signal, appearance
State changes waveform signal and change in location waveform signal and carries out waveform conversion by a variety of operation rules, and waveform is believed after being converted
Number some features.
Wherein, a variety of operation rules can include it is conventional add, subtract, multiplication and division computing, integrate, differentiate, Tai Lezhan
Open, Fourier transformation, time domain, frequency-domain transform etc..As shown in fig. 6, acceleration change waveform signal, angular speed change waveform letter
Number, attitudes vibration waveform signal and change in location waveform signal by specifying after operation rule, can export with obvious characteristic
Waveform signal, some features can be obtained according to the waveform signal with obvious characteristic of output, these features are exactly original
Some features of motor message.
Which kind of it is not limited it should be noted that specifically being handled using operation rule waveform signal, it is only necessary to ensure
The feature of waveform signal after conversion is obvious.For example, acceleration has an obvious maximum by derivation, then
Acceleration can be using derivation conversion, and the wave character obtained after derivation conversion can be designated as feature x.By that analogy, acceleration
Change waveform signal, angular speed change waveform signal, attitudes vibration waveform signal and change in location waveform signal pass through a variety of fortune
Calculate rule to carry out after waveform conversion, feature a, feature b, feature c, feature d ... can be obtained.
Step 350, obtain and be pre-stored the probability that each feature occurs in various type of action,
It should be noted that can store each feature in the memory cell of smart machine 120 in advance appears in various move
Make the probability in type.The probability that each feature occurs under each type of action can constitute statistical matrix.It the following is statistical matrix
A kind of form:
Wherein, each row represent a kind of type of action respectively, represent a kind of feature respectively per a line, for example, probability
k11It is the probability that feature a occurs in type of action 1, k21Represent type of action 1 and the probability of feature b, k occur34Represent type of action 4
There is feature c probability, by that analogy.
For different motions, it can separately count and occur the probability of each feature in each type of action.To play badminton
Exemplified by motion, above-mentioned steps 310 and 330 are referred to, first by obtaining in motion process of playing badminton under different actions
Sample motor message, then carries out waveform conversion to the sample motor message under different actions, is converted into obvious characteristic
Waveform signal, and then obtain some features of sample waveform signal under different actions.Can be by under multi collect difference action
Sample motor message and carry out feature extraction, finally count the probability that each feature under every kind of action occurs, build above-mentioned
Statistical matrix.If subsequently to recognize newly-increased action or need newly-increased feature, the statistical matrix only need to be improved, can
Autgmentability is stronger.
Step 370, there is the maximum type of action of some feature probabilities simultaneously in screening from various type of action;
Specifically, doing n action 1, the probability that then feature a occurs in statistics action 1 is designated as k11, similarly, complete each special
The probability occurred in each action is levied, statistical matrix is formed.For the action newly produced, if this new element existing characteristics
There is the maximum probability that feature a, b, c occur simultaneously under action an x, action x in a, b, c, and statistical matrix, then this
Action x can consider to be exactly new element.
Optionally, as shown in fig. 7, step 370 is specifically included:
Step 371, the probability in various type of action is appeared according to each feature, calculates each type of action and deposit simultaneously
In the probability of some features;
Specifically, assuming that the original motion signal of the action newly produced obtains existing characteristics by step 310 and step 330
A, b, c, and then the probability in each type of action is appeared according to each feature, each type of action is calculated respectively while going out
Existing feature a, b, c probability.Assuming that feature a, b, c occur independently of one another, then feature a, b, c probability occur simultaneously in acting 1
For k11×k21×k31.Similarly, it can calculate respectively in every kind of type of action while there is feature a, b, c probability.
Step 372, there is the probability of some features simultaneously according to each type of action, filter out probability maximum
Type of action.
While there is some features (such as feature a, b, c) probability in above-mentioned steps 371 calculate every kind of type of action
Afterwards, the maximum type of action of probability can be filtered out.Using the type of action as the athletic performance newly produced recognition result.
Illustrate the technical scheme to be clear and concise, it is assumed that action 1,2,3 is only existed in a certain motion, for what is newly produced
Athletic performance, existing characteristics a, b, c are obtained by step 310 and step 330, then can be filtered out by following calculating process
Occurs the maximum type of action of some feature probabilities simultaneously:
Q1=k11*k21*k31
Q2=k12*k22*k32
Q3=k13*k23*k33
P1=Q1/(Q1+Q2+Q3)
P2=Q2/(Q1+Q2+Q3)
P3=Q3/(Q1+Q2+Q3)
max(P1,P2,P3)
Wherein, Q1The product for the probability that feature a, b, c occur, Q in representative action 12Feature a, b, c occur in representative action 2
The product of probability, Q3The product for the probability that feature a, b, c occur, max (P in representative action 31,P2,P3) represent to take P1、P2、P3In most
Big person.If P1Maximum, represents that the action newly produced is action 1, if P2Maximum, represents that the action newly produced is action 2, such as
Fruit P3Maximum, represents that the action newly produced is action 3.
Assuming that the action only one of which feature d newly produced, as can be seen that feature d occurs in each action from statistical matrix
Probability be k41, k42, k43..., wherein, the type of action for feature d maximum probabilities occur may be considered the action newly produced.
Step 390, the recognition result of athletic performance is obtained according to the type of action filtered out.
Specifically, the athletic performance newly produced has some features, on the basis of above-mentioned steps, it can filter out simultaneously
There is the maximum type of action of the probability of some features, and then can be moved the type of action filtered out as the motion newly produced
The recognition result of work.
Optionally, step 390 is specifically included:Judge whether the type of action filtered out is more than one kind, if it is not, then sieving
The type of action selected as the athletic performance recognition result;
If so, original motion signal sample motor message corresponding with each type of action filtered out is carried out into waveform
Matching, searches the sample motor message with the original motion Signal Matching, and the sample motor message of the matching is corresponding
Type of action as the athletic performance recognition result.
If it should be noted that filtering out more than one type of action (possibility very little), that is, filtered out at least two kinds of dynamic
Make type, the probability that these type of action have some features simultaneously is identical, then can be according to filtering out for having prerecorded
Each type of action sample motor message (such as sample waveform signal), using these sample waveform signals be used as ATL, will
The original motion signal (such as Raw waveform signals) of new element is matched one by one with the sample waveform signal in ATL, is searched
With Raw waveform signals matching degree highest sample waveform signal, the corresponding action of the sample waveform signal is regard as new element
Recognition result.Certainly, if only filtering out a kind of type of action, this type of action filtered out is as the motion newly produced
The recognition result of action.
In other embodiments, when the type of action more than one filtered out, can also occur according in current kinetic
The priority of each type of action, is ranked up to the type of action filtered out, regard the high type of action of priority as motion
The recognition result of action.Wherein, can be according to there is each type of action in the priority of each type of action in current kinetic
Accounting is determined.
As an example it is assumed that in a badminton, action 1 accounts for 39%, and action 2 accounts for 35%, and action 3 is accounted for
20%...... (action 1, action 2, action 3 ... represent different type of action respectively), then can consider the preferential of action 1
Level is higher than action 2, and the priority of action 2 is higher than action 3.
It is further to note that the mode that the disclosure does not first pass through Waveform Matching directly carries out the knowledge of athletic performance
Not, but the mode that first passes through probability statistics filters out type of action, when filtering out type of action more than one, just enter traveling wave
Shape is matched.It is more due to there is action in a motion, therefore there is more sample motor message in ATL, for new dynamic
The original motion signal of work, if carrying out Waveform Matching, data processing amount with sample motor message all in ATL one by one
Larger, athletic performance recognition efficiency is low.And the mode based on probability statistics, type of action is screened, amount of calculation is smaller, transported
Dynamic action recognition efficiency high.
The disclosure enters line translation by the original motion signal to Moving Objects, extracts some spies of original motion signal
Levy, then according to the probability that each feature occurs in known every kind of type of action, will occur some feature probabilities simultaneously maximum
Type of action as athletic performance recognition result.The program due to without by the waveform signal of action to be measured one by one with it is all
The waveform signal of known action is matched, and the probability occurred in known each action according to some features of unknown action is just
The identification of action can be realized, amount of calculation is smaller, recognition efficiency is high, further as action identification independent of movement locus
Matched with characteristic value, for the similar action in track or the close action of characteristic value, by some spies for calculating unknown action
The probability appeared in known action is levied, the identification of action can also be realized, thus the identification accuracy of athletic performance is high.
In addition, as needed, according to these data of angular speed, acceleration, position and posture, needed for can also obtaining other
Parameter.For example, it is also possible to calculate the speed and strength of athletic performance.
Assuming that smart machine 120 is in t1And t2The position coordinates at moment is respectively (x1, y1, z1) and (x2, y2, z2), then t1Extremely
t2Moment smart machine occur displacement be
Derivation can obtain the speed at each moment.
By swing the bat or bat swing exemplified by, it is known that acceleration is a, and the quality of racket or bat is m, according to formula F=m*
A, can also calculate each moment to racket or the active force of bat.
Wherein, for athletic performance recognition result and the speed and strength of athletic performance can be stored in smart machine
In 120 flash (memory cell), mobile terminal 1 10 can read the data in smart machine 120 by software APP, then
Shown and stored, and be synchronized to high in the clouds.When software APP is again mounted or with changing mobile terminal 1 10, account can be passed through
Number data synchronous in advance are obtained from high in the clouds, then stored and shown in mobile terminal 1 10.
Following is disclosure device embodiment, and the motion that can be used for performing the above-mentioned execution of smart machine 120 of the disclosure is moved
The recognition methods embodiment of work.For the details not disclosed in disclosure device embodiment, disclosure athletic performance refer to
Recognition methods embodiment.
Fig. 8 is a kind of block diagram of the identifying device of athletic performance according to an exemplary embodiment, the athletic performance
Identifying device can be used in the smart machine 120 of implementation environment shown in Fig. 1, perform Fig. 3, Fig. 4, Fig. 7 it is any shown in fortune
The all or part of step of the recognition methods of action.As shown in figure 8, the identifying device includes but is not limited to:Signal acquisition
Module 810, characteristic extracting module 830, probability acquisition module 850, action screening module 870 and action recognition module 890.
Wherein, signal acquisition module 810, the motion process being performed for moving object tracking obtains original motion letter
Number;
Characteristic extracting module 830, for the original motion signal to be converted by default change scaling method, from change
Some features of the original motion signal are extracted in signal after changing;
Probability acquisition module 850, for obtaining the probability that pre-stored each feature occurs in various type of action;
Screening module 870 is acted, for the screening from various type of action while there are some feature probability maximums
Type of action;
Action recognition module 890, the recognition result for obtaining athletic performance according to the type of action filtered out.
The function of modules and the implementation process of effect specifically refer to the identification side of above-mentioned athletic performance in said apparatus
The implementation process of correspondence step, will not be repeated here in method.
Signal acquisition module 810 such as can be some physical arrangement sensor cluster 214 in Fig. 2.
Characteristic extracting module 830, probability acquisition module 850, action screening module 870 and action recognition module 890
Can be functional module, the corresponding step in recognition methods for performing above-mentioned athletic performance.It is appreciated that these modules can
To be realized by hardware, software or the two combination.When realizing in hardware, these modules may be embodied as one or
Multiple hardware modules, such as one or more application specific integrated circuits.When being realized with software mode, these modules may be embodied as
The one or more computer programs performed on the one or more processors, such as storage performed by the processor 218 of Fig. 2
Program in memory 204.
Optionally, as shown in figure 9, the action screening module 870 includes but is not limited to:
Probability calculation unit 871, for appearing in the probability in various type of action according to each feature, calculates each dynamic
Make type while there is the probability of some features;
Screening unit 872 is acted, for there is the probability of some features, sieve simultaneously according to each type of action
Select the maximum type of action of probability.
Optionally, the action recognition module 890 includes but is not limited to:
Whether judging unit, the type of action for judging to filter out is more than one kind, if it is not, what is then filtered out is described
Type of action as the athletic performance recognition result;
Matching unit, for when the type of action filtered out is more than a kind of, according to it is pre-stored with filter out it is each
The corresponding sample motor message of type of action, the sample motion searched by Waveform Matching with the original motion Signal Matching is believed
Number, using the corresponding type of action of sample motor message of the matching as the athletic performance recognition result.
Optionally, as shown in Figure 10, the signal acquisition module 810 includes:
Data acquisition unit 811, for passing through Moving Objects described in the sensor continuous collecting on the Moving Objects
3-axis acceleration component and angular speed;
Posture position determining unit 812, described in being obtained according to the 3-axis acceleration component and angular speed of collection
The posture of Moving Objects and position;
Waveform determining unit 813, for 3-axis acceleration component, angular speed, posture and the position according to the Moving Objects
The change put obtains acceleration change waveform signal, angular speed change waveform signal, attitudes vibration waveform signal and change in location
Waveform signal.
Further, the characteristic extracting module 830 includes:
Waveform converter unit, for the acceleration change waveform signal, angular speed to be changed into waveform signal, attitudes vibration
Waveform signal and change in location waveform signal carry out waveform conversion by a variety of operation rules, if waveform signal after being converted
Dry feature.
Optionally, the disclosure also provides a kind of electronic equipment, the electronic equipment can as implementation environment shown in Fig. 1 intelligence
Can equipment, all or part of step of the recognition methods of any shown athletic performances of execution Fig. 3, Fig. 4, Fig. 7.The electronics
Equipment includes:
Processor;
Memory for storing processor-executable instruction;
Wherein, the processor is configured as performing the recognition methods of the athletic performance described in above-described embodiment.
The concrete mode of the computing device operation of device in the embodiment is in the identification about the athletic performance
Detailed description is performed in the embodiment of method, explanation will be not set forth in detail herein.
In the exemplary embodiment, a kind of storage medium is additionally provided, the storage medium is computer-readable recording medium,
For example can be to include the provisional and non-transitorycomputer readable storage medium of instruction.The storage medium is stored with computer
Program, the computer program can be completed the recognition methods of the athletic performance in above-described embodiment by computing device.
It should be appreciated that the invention is not limited in the precision architecture for being described above and being shown in the drawings, and
And various modifications and changes can be being performed without departing from the scope.The scope of the present invention is only limited by appended claim.
Claims (10)
1. a kind of recognition methods of athletic performance, it is characterised in that including:
The motion process that moving object tracking is performed, obtains original motion signal;
The original motion signal is converted by default change scaling method, extracts described original from the signal after conversion
Some features of motor message;
Obtain and be pre-stored the probability that each feature occurs in various type of action;
There is the maximum type of action of some feature probabilities simultaneously in screening from various type of action;
The type of action according to filtering out obtains the recognition result of athletic performance.
2. according to the method described in claim 1, it is characterised in that the screening from various type of action occurs described simultaneously
The maximum type of action of some feature probabilities, including:
The probability occurred according to each feature in various type of action, calculates each type of action and there is some spies simultaneously
The probability levied;
There is the probability of some features simultaneously according to each type of action, filter out the maximum type of action of probability.
3. according to the method described in claim 1, it is characterised in that the type of action that the basis is filtered out is moved
The recognition result of action, including:
Judge whether the type of action filtered out is more than one kind, if it is not, the type of action then filtered out is as described
The recognition result of athletic performance;
If so, original motion signal sample motor message corresponding with each type of action filtered out is carried out into waveform
Match somebody with somebody, search the sample motor message with the original motion Signal Matching, the sample motor message of the matching is corresponding dynamic
Make type as the recognition result of the athletic performance.
4. according to the method described in claim 1, it is characterised in that the motion process that the moving object tracking is performed, obtain
Original motion signal is obtained, including:
Pass through the 3-axis acceleration component and angular speed of Moving Objects described in the sensor continuous collecting on the Moving Objects;
Posture and the position of the Moving Objects are obtained according to the 3-axis acceleration component and angular speed of collection;
Acceleration change ripple is obtained according to the 3-axis acceleration component of the Moving Objects, angular speed, posture and the change of position
Shape signal, angular speed change waveform signal, attitudes vibration waveform signal and change in location waveform signal.
5. method according to claim 4, it is characterised in that the original motion signal is passed through into default change scaling method
Enter line translation, some features of the original motion signal are extracted from the signal after conversion, including:
By the acceleration change waveform signal, angular speed change waveform signal, attitudes vibration waveform signal and change in location ripple
Shape signal carries out waveform conversion, some features of waveform signal after being converted by a variety of operation rules.
6. a kind of identifying device of athletic performance, it is characterised in that including:
Signal acquisition module, the motion process being performed for moving object tracking obtains original motion signal;
Characteristic extracting module, for the original motion signal to be converted by default change scaling method, after conversion
Some features of the original motion signal are extracted in signal;
Probability acquisition module, for obtaining the probability that pre-stored each feature occurs in various type of action;
Screening module is acted, for the screening from various type of action while there is the maximum action class of some feature probabilities
Type;
Action recognition module, the recognition result for obtaining athletic performance according to the type of action filtered out.
7. device according to claim 6, it is characterised in that the action screening module includes:
Probability calculation unit, for appearing in the probability in various type of action according to each feature, calculates each type of action
There is the probability of some features simultaneously;
Unit is chosen in action, for there is the probability of some features simultaneously according to each type of action, filters out several
The maximum type of action of rate.
8. device according to claim 6, it is characterised in that the action recognition module includes:
Whether judging unit, the type of action for judging to filter out is more than one kind, if it is not, the action then filtered out
Type as the athletic performance recognition result;
Matching unit, for original motion signal sample motor message corresponding with each type of action filtered out to be carried out
Waveform Matching, searches the sample motor message with the original motion Signal Matching, by the sample motor message pair of the matching
The type of action answered as the athletic performance recognition result.
9. a kind of electronic equipment, it is characterised in that the electronic equipment includes:
Processor;
Memory for storing processor-executable instruction;
Wherein, the processor is configured as the recognition methods that perform claim requires the athletic performance described in 1-5 any one.
10. a kind of computer-readable recording medium, it is characterised in that the computer-readable recording medium storage has computer journey
Sequence, the computer program can as computing device complete claim 1-5 any one described in athletic performance identification side
Method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710463245.1A CN107273857B (en) | 2017-06-19 | 2017-06-19 | Motion action recognition method and device and electronic equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710463245.1A CN107273857B (en) | 2017-06-19 | 2017-06-19 | Motion action recognition method and device and electronic equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107273857A true CN107273857A (en) | 2017-10-20 |
CN107273857B CN107273857B (en) | 2021-03-02 |
Family
ID=60069067
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710463245.1A Active CN107273857B (en) | 2017-06-19 | 2017-06-19 | Motion action recognition method and device and electronic equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107273857B (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110377159A (en) * | 2019-07-24 | 2019-10-25 | 张洋 | Action identification method and device |
CN111803903A (en) * | 2019-04-10 | 2020-10-23 | 深圳先进技术研究院 | Body-building action recognition method and system and electronic equipment |
CN114259720A (en) * | 2020-09-15 | 2022-04-01 | 荣耀终端有限公司 | Action recognition method and device, terminal equipment and motion monitoring system |
WO2022083388A1 (en) * | 2020-10-21 | 2022-04-28 | 歌尔股份有限公司 | Racket ball game method and apparatus based on head-mounted device, and device |
CN115019240A (en) * | 2022-08-04 | 2022-09-06 | 成都西交智汇大数据科技有限公司 | Grading method, device and equipment for chemical experiment operation and readable storage medium |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1705956A (en) * | 2002-10-29 | 2005-12-07 | 索尼株式会社 | Gait waveform feature extracting method and individual identification system |
CN1756938A (en) * | 2003-01-10 | 2006-04-05 | 江苏千鹏诊断工程有限公司 | State judging method, and state predicting method and device |
CN101394787A (en) * | 2006-03-14 | 2009-03-25 | 索尼株式会社 | Body movement detector, body movement detection method and body movement detection program |
CN102930250A (en) * | 2012-10-23 | 2013-02-13 | 西安理工大学 | Motion recognition method for multi-scale conditional random field model |
CN103156617A (en) * | 2011-12-08 | 2013-06-19 | 浩华科技实业有限公司 | Sensor for thorax-abdomen respiratory movement wave shapes |
CN103377542A (en) * | 2013-07-16 | 2013-10-30 | 中国科学院深圳先进技术研究院 | Human body fall-down preventing early warning method and human body fall-down preventing early warning device |
US20140207401A1 (en) * | 2013-01-18 | 2014-07-24 | Postech Academy-Industry Foundation | Apparatus for recognizing motion feature of user, method for generating orthogonal non-negative matrix factorization (onmf)-based basis matrix, and method for generating orthogonal semi-supervised non-negative matrix factorization (ossnmf)-based basis matrix |
US20140354550A1 (en) * | 2013-05-29 | 2014-12-04 | Microsoft Corporation | Receiving contextual information from keyboards |
CN106203484A (en) * | 2016-06-29 | 2016-12-07 | 北京工业大学 | A kind of human motion state sorting technique based on classification layering |
CN106289309A (en) * | 2016-10-26 | 2017-01-04 | 深圳大学 | Step-recording method based on 3-axis acceleration sensor and device |
-
2017
- 2017-06-19 CN CN201710463245.1A patent/CN107273857B/en active Active
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1705956A (en) * | 2002-10-29 | 2005-12-07 | 索尼株式会社 | Gait waveform feature extracting method and individual identification system |
CN1756938A (en) * | 2003-01-10 | 2006-04-05 | 江苏千鹏诊断工程有限公司 | State judging method, and state predicting method and device |
CN101394787A (en) * | 2006-03-14 | 2009-03-25 | 索尼株式会社 | Body movement detector, body movement detection method and body movement detection program |
CN103156617A (en) * | 2011-12-08 | 2013-06-19 | 浩华科技实业有限公司 | Sensor for thorax-abdomen respiratory movement wave shapes |
CN102930250A (en) * | 2012-10-23 | 2013-02-13 | 西安理工大学 | Motion recognition method for multi-scale conditional random field model |
US20140207401A1 (en) * | 2013-01-18 | 2014-07-24 | Postech Academy-Industry Foundation | Apparatus for recognizing motion feature of user, method for generating orthogonal non-negative matrix factorization (onmf)-based basis matrix, and method for generating orthogonal semi-supervised non-negative matrix factorization (ossnmf)-based basis matrix |
US20140354550A1 (en) * | 2013-05-29 | 2014-12-04 | Microsoft Corporation | Receiving contextual information from keyboards |
CN103377542A (en) * | 2013-07-16 | 2013-10-30 | 中国科学院深圳先进技术研究院 | Human body fall-down preventing early warning method and human body fall-down preventing early warning device |
CN106203484A (en) * | 2016-06-29 | 2016-12-07 | 北京工业大学 | A kind of human motion state sorting technique based on classification layering |
CN106289309A (en) * | 2016-10-26 | 2017-01-04 | 深圳大学 | Step-recording method based on 3-axis acceleration sensor and device |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111803903A (en) * | 2019-04-10 | 2020-10-23 | 深圳先进技术研究院 | Body-building action recognition method and system and electronic equipment |
CN110377159A (en) * | 2019-07-24 | 2019-10-25 | 张洋 | Action identification method and device |
CN110377159B (en) * | 2019-07-24 | 2023-06-09 | 张洋 | Action recognition method and device |
CN114259720A (en) * | 2020-09-15 | 2022-04-01 | 荣耀终端有限公司 | Action recognition method and device, terminal equipment and motion monitoring system |
CN114259720B (en) * | 2020-09-15 | 2022-10-18 | 荣耀终端有限公司 | Action recognition method and device, terminal equipment and motion monitoring system |
WO2022083388A1 (en) * | 2020-10-21 | 2022-04-28 | 歌尔股份有限公司 | Racket ball game method and apparatus based on head-mounted device, and device |
CN115019240A (en) * | 2022-08-04 | 2022-09-06 | 成都西交智汇大数据科技有限公司 | Grading method, device and equipment for chemical experiment operation and readable storage medium |
CN115019240B (en) * | 2022-08-04 | 2022-11-11 | 成都西交智汇大数据科技有限公司 | Grading method, device and equipment for chemical experiment operation and readable storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN107273857B (en) | 2021-03-02 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107273857A (en) | The recognition methods of athletic performance and device, electronic equipment | |
US10600334B1 (en) | Methods and systems for facilitating interactive training of body-eye coordination and reaction time | |
Blank et al. | Sensor-based stroke detection and stroke type classification in table tennis | |
CN107281709B (en) | A kind of extracting method and device, electronic equipment of sport video segment | |
US11210855B2 (en) | Analyzing 2D movement in comparison with 3D avatar | |
CN104056441B (en) | Information processing system, information processing method, and storage medium | |
CN103970271B (en) | The daily routines recognition methods of fusional movement and physiology sensing data | |
DE102015207415A1 (en) | Method and apparatus for associating images in a video of a person's activity with an event | |
CN105797319B (en) | A kind of badminton data processing method and device | |
US11819734B2 (en) | Video-based motion counting and analysis systems and methods for virtual fitness application | |
CN102222342A (en) | Tracking method of human body motions and identification method thereof | |
CN106648118A (en) | Virtual teaching method based on augmented reality, and terminal equipment | |
CN109376705A (en) | Dance training methods of marking, device and computer readable storage medium | |
CN110443190A (en) | A kind of object identifying method and device | |
Sun et al. | Tracknetv2: Efficient shuttlecock tracking network | |
CN106730723A (en) | A kind of soldier based on Intelligent worn device pang ball training method and system | |
Johansen et al. | Combining video and player telemetry for evidence-based decisions in soccer | |
Lou et al. | Real-time monitoring for manual operations with machine vision in smart manufacturing | |
CN111617464A (en) | Treadmill body-building method with action recognition function | |
CN104050454B (en) | A kind of motion gesture track acquisition methods and system | |
US20230206697A1 (en) | Action recognition method and apparatus, terminal device, and motion monitoring system | |
CN110083742A (en) | A kind of video query method and device | |
CN113873100A (en) | Video recording method, video recording device, electronic equipment and storage medium | |
US20160314352A1 (en) | Analysis device, recording medium, and analysis method | |
Shahar et al. | Wearable inertial sensor for human activity recognition in field hockey: Influence of sensor combination and sensor location |
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 |