CN108447077A - A kind of horsemanship jockey posture information acquisition analysis system - Google Patents
A kind of horsemanship jockey posture information acquisition analysis system Download PDFInfo
- Publication number
- CN108447077A CN108447077A CN201810256743.3A CN201810256743A CN108447077A CN 108447077 A CN108447077 A CN 108447077A CN 201810256743 A CN201810256743 A CN 201810256743A CN 108447077 A CN108447077 A CN 108447077A
- Authority
- CN
- China
- Prior art keywords
- jockey
- posture
- posture information
- node
- horsemanship
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/005—General purpose rendering architectures
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Graphics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention discloses a kind of horsemanship jockey posture information acquisition analysis systems, the system includes being placed in the sensing node motion information acquisition module of each characteristic portion of jockey, the aggregation node module of each joint movements information is handled, realizes the wireless data transmission module communicated with host computer and the interfaces 3D mans motion simulation PC that jockey's posture is reconstructed.The system can realize the real-time capture calculating to jockey's 3 d pose information by sensing node, and be realized to the Real-time data drive of three-dimensional (3 D) manikin by wireless communication mode, reach model and the athletic posture during jockey's horsemanship is reconstructed.The system has higher tracking accuracy and stronger real-time, the information capture and reconstruct to jockey's posture during horsemanship may be implemented, to provide good kinematic data for specialized coaches and jockey, deficiency existing for jockey is analyzed, the purpose for improving jockey's horsemanship level is finally reached.
Description
Technical field
The present invention relates to field of information processing, more particularly to a kind of horsemanship jockey posture information acquisition analysis system.
Technical background
Motion capture technology, refer to by the even abundant facial expression data of the limb motion data to experimenter into
Row acquisition and processing, and virtual three-dimensional (3 D) manikin is driven using these limb actions or facial expression data, realization pair
The analogue simulation of human motion.Motion capture technology is primarily referred to as the auxiliary by computer, all round computer graphics, electronics
The technologies such as, human engineering, optics, mechanics, using the methods of measurement, tracking, calculating each pass of captured object
The action data at key position acquires out, a series of processing such as is then filtered, merges to primary data, and pass through and transmit association
View by posture information for driving virtual three dimensional computer modeling, restore captured object a kind of acquisition of data action,
The novel human-machine interaction technology of processing, emulation, storage.This be it is a kind of can be accurate, fast by virtual three dimensional computer modeling
The new technique method that the real time kinematics state reduction of people in actual life or various objects is come out of speed.
Currently, common motion capture technology is different with the type of used sensor according to operating, can substantially divide
It is a kind of mainly to use video camera array for two classes, utilize multiple high-precision high sampling rates using the motion capture system of video camera
Camera capture the reflective marker on sporter joint, this kind of system is typically expensive, and data processing amount is big, and easily by outer
The influence of boundary's environment.Another kind of is the sensor group using wearing physically, by dressing all kinds of sensings with user
Device captures its limb motion, and such system is not generally by external environmental interference.And with microelectromechanical systems
(MEMS), the development of microsensor and wireless communication technique, this kind of motion capture system volume is smaller and smaller, one single chip body
Product is 3mm × 3mm × 1mm, and power consumption is lower and lower, is less than 2uA to reach standby current, is very suitable for making wearable structure.
In motion-captured model selection, there is also larger areas for its meaning for being showed of different models and mode
Not, currently, common manikin is broadly divided into stick model, physical model, surface model and multi-level fuzzy judgment.Stick model
It chooses limited rigid body segment and joint links, lack fidelity;Physical model simulates people by simple solid pattern
The structure of body, it is computationally intensive, and stability is poor;Multi-level fuzzy judgment includes casing play, muscle layer and skin layer, and complexity is high, counts
Calculation amount is relatively large;Surface model is made of casing play and skin layer, it is easy to accomplish, calculation amount is smaller.
Traditional movement capturing technology is mostly to expand to come on the basis of optical, passes through multiple high-resolution
After camera is shot and handled in special operating room, for driving virtual portrait model, achieve the purpose that reconstruct action.
Common optical system mainly has Vicon and Optitrack etc., this kind of system generally can provide very high accuracy, but its
For ordinary populace, price is more expensive.It is generally used for the mechanisms such as scientific research institutions and large-scale experiment room.With micro sensing
The development of technology and airmanship so that the motion capture technology based on motion-sensing is increasingly becoming possibility, at present by being worn on
Human body different parts, collection data acquisition, the motion capture system that more sensing nodes that processing and communication are integrated are constituted are gradual
It is universal.Compared to the features such as based on optical system, it has equipment simple, and cost is relatively low and fault-tolerance is strong.
At abroad, about in the 1970s, psychologist Johansson MLD have been put forward for the first time the general of motion capture
It reads;In the 1980s, professor Calvert of Simonfraser University (Simon Fraser University) et al. and
Carol of the Institute of Technology (MIT) et al. and scholar Robertson and Walter has carried out deep grind to movement capturing technology
Study carefully, has pushed the progress of motion capture technology;In the 1990s, utilizing computer skill with the scholar that Tardif etc. is artificially represented
The progress of art, sensor technology, network communication technology accelerates the development of motion capture technology so that the technology reaches its maturity.
The motion capture technology of early stage is all based on the timing camera work of Etinne Jules Marey (1830-1940)
On.The different frame that Marey is acted using the picture record of unified heliotype, then passes through the picture sequence that will shoot
It is arranged together the capture acted.And the motion capture technology of real meaning is to start from eighties of last century the seventies.
Disney company is in film《White Snow and the seven dwarfs》In attempt through the true stage business of " imitation " high-speed capture
Coherent photograph improves the quality of movement of cartoon role, although animation effect lacks dramatic and cartoon, action effect
But very true to nature.
With the further development of computer graphics, computer vision and motion capture technology, motion capture technology exists
Application in terms of computer animation is gradually ripe.The Rebecca Allen in New York computer graphics techniques laboratory are just devised
A kind of Optical devices of mercury mirror, the performance posture of true dancer is projected on computer screen, is waved as number
Step the reference of performer's animation key frame.The application of advanced optics motion capture system, and can be with considerably beyond the design of early stage
More trace point, higher speed carry out action intricately woven together capture.However since its limited camera covers model
The limitation enclosed, complicated tracking solution, and expensive price greatly limit it and use field.Hereafter in optical profile type
On the basis of motion capture, also occur mechanical, electromagnetic type, acoustics formula motion capture system in succession.
Currently, some scientific & technical corporation in American-European countries have been proposed different types of motion-captured product:Such as Holland
Xsens MVN and the 3Dsuit inertia actions that Xsens companies and 3Dsuit companies of the U.S. release capture suit;Britain Vicon and
The Vicon Motion System and Motion Analysis optical motion capture series sets of Qualisys companies of Sweden production
Dress.The research of domestic motion capture technology is started late, but has also been attracted many research institutions and scholar to carry out it and ground
Study carefully, typical represent has Chinese Academy of Sciences's sensing network and applies Joint Research Centre and NOITOM companies, technical merit,
Certain gap is still had in application range with external product.
Extend with the continuous development of motion capture technology, its main conditions faced has at present:In jockey's motion process
The acquisition of movable information and fusion problem, the cost problem of solution, the complexity issue of solution annual reporting law time.For above-mentioned
Problem, many researchers conduct a research to it using all kinds of methods.
In terms of the resolving of feature posture, current many relevant scholars expand a series of researchs, such as Luinge et al.
Using minisize gyroscopes and accelerometer, the inclined drift of horizontal plane is eliminated, but without eliminating the drift of vertical direction rotation
It moves, there are larger errors for the athletic posture calculated;
Roetenberg et al. proposes a kind of compensation Kalman filter algorithm, and compensation model is added on error model,
The influence that compensation changes of magnetic field is brought to evaluated error, but algorithm model calculating is more complicated, and do not account for high-speed motion
The case where middle acceleration of motion interference;Bachman proposes the Kalman filter algorithm based on quaternary number, uses and adds under low-speed situations
The low frequency part of speed and magnetic field intensity estimates direction, measures posture under high-speed case with angular speed, Yun et al. is used
QUEST algorithms obtain quaternary number to estimate posture by acceleration and magnetic field vector, are then further merged with angular speed again,
Processing has linearized observational equation in this way, but QUEST algorithms computation complexity is high, he used (factor later
Quaternion algorithm, FAQ) algorithm instead of QUEST, reduces algorithm computation complexity, but still does not have complete
Solve the problems, such as how to remove magnetic field and acceleration noise.
With the development of computer science, dimension display technologies reach its maturity, and human body motion capture technology is in game, video display
The fields such as design, medical science, athletic training have broad application prospects.In game design field, many large-scale movements
Human motion capture and virtual reality technology are combined by class game, keep the object of structure more life-like, vivid, such as current
《3D basketballs》、《NBA》Deng;Field, film are created in video display《A Fanda》、《Sound of chopping wood Adventures of Pei》Deng all widely having used human body
Capturing movement and virtual reality technology so that the picture of shooting is more rich in appeal;And in medical science field, the technology is wide
It is general be used for orthopaedics, surgical operation, the fields such as physical therapy, mainly by detect patients ' recovery before and after athletic posture and compared,
The case where judging rehabilitation helps doctor preferably to grasp the state of patient;Athletic training field is caught by the action of sportsman's wearing
Equipment (such as wearable sensing vest) is obtained, is shown by the real-time tracking at interface, it can be with the physical training condition of real-time monitoring sportsman.
It may be said that the application of movement capturing technology has larger market prospects, novel interaction mechanism is human-computer interaction
The another big innovation in field, greatly advances the development of field of human-computer interaction.If simultaneously can be used in conjunction at present miniature
Property sensing technology, wireless communication technique, real time data acquisition, in one, are developed with independent intellectual property rights in treatment technology
Micro-mechanical inertia motion capture system will greatly enrich the development of related field, while bring extremely considerable economy
Benefit and social benefit.
In the implementation of the present invention, inventor has found at least there is posture number in existing inertia action capturing technology
According to drift, the defects of long-range controllability is poor, and real-time is not strong.
Invention content
It is an object of the present invention to for traditional motion capture there are the problem of, propose a kind of to be based on micro- inertia sensing
The motion capture system of device, and on the basis of traditional algorithm, it is relatively high to design a kind of precision, the stronger posture feature of real-time
Strategy is resolved, realization more accurately captures posture information, while developing a kind of 3D based on three-dimensional bone storehouse and transporting in real time
Motion tracking man-machine interface, further improves effect of visualization.
Mentality of designing is as follows:This system is respectively by sensing node, aggregation node, wireless communication module, motor reorgan system
A few part compositions.Sensing node is made of the inertial sensor node for being worn on each characteristic portion, and distribution is set up between node
Formula body-sensing network can carry out real-time capture to the posture information at motion feature position, and is filtered to original posture information
And Data Fusion, calculate initial posture information;Aggregation node is communicated by bus mode and sensing node, real
When reception sensing node transmission signal, and the data of each sensing node are carried out to summarize packing;Wireless communication module is logical
It crosses wireless transmission protocol and the data that aggregation node is packaged is sent to the ends PC;And motor reorgan system then passes through wireless transmission protocol
The attitude data that reception slave computer is sent in real time, and resolve and demarcate by specific, it is translated into and is suitable for driving people
The real-time attitude signal of body Model, the tracking display that real human body is moved to drive manikin to realize.
Each sensing node is made of inertial sensor and a microcontroller core respectively in this system, microprocessor utensil
Body can be the singlechip chip, such as STM32 series etc. of ARM 32-bit Cortex series, and inertial sensor is by three axis
The 9 axis microsensor units that accelerometer, three-axis gyroscope and three axis magnetometers collectively constitute, inertial sensor and microprocessor
It is communicated by IIC or spi bus mode between device, microprocessor receives and processes the first of inertia sensing unit in real time
Beginning attitude data, and it is filtered fusion treatment.
It in system, is connected by bus between sensing node and aggregation node, can specifically show as CAN bus connection side
Formula, aggregation node are capable of the data of the reception sensing node of real-time stabilization by way of data outage, avoid losing for data
Packet and delay.
Wireless transport module described in system, is specifically as follows 2.4G wireless data transmission modules, Bluetooth communication modules or
WIFI module, specifically by taking WIFI module as an example, by establishing the TCP-IP agreements of WIFI hot spot, by the data transmission of aggregation node
To upper computer software.
In the system, motor reorgan part can specifically use Visual Studio in Directed3D graphic development libraries
On the basis of establish jockey's motion tracking software interface at the visual angles 3D, and TCP-IP agreements are called on basis of software, to connect
The attitude data of each node of slave computer is received, and by the data fusion of multisensor, calculated suitable for driving 3D manikins
Characteristic, finally realize the reconstruct moved to real human body.
In system, it is contemplated that the energy consumption of system, the rechargeable lithium battery that capacity can be selected larger is powered, and is directed to
Various pieces carry out power management respectively, improve the reliability of system.According to prototyping testing, the great Rong of 10000mah, 1A electric discharge
Amount lithium battery is probably capable of providing system worked well 2 hours or so, can reach practical standard.
The present invention uses for reference biotomy principle, and whole body bone is divided into:Left thigh, left leg, left foot, right thigh are right
Shank, right crus of diaphragm, pelvis, vertebra, chest, left hand, left elbow, left shoulder, the right hand, right elbow, right shoulder, 16, head etc. characteristic node, and
It is merged by the posture information to systemic features position, maps that on corresponding 3D manikins, ridden for reconstructing
The athletic posture of hand.
Each sensing node is embedded into tightly by the installation for system, specific as follows, the motion capture for whole body, consideration
In body uniform or special bandage, aggregation node and wireless transport module are integrated in one, and can be fixed on loins, so neither
There are redundancy issue, and it is easy to wear comfortable, there is portable, easy to wear, safe and reliable advantage.
In order to achieve the above object, horsemanship jockey posture information acquisition analysis system provided by the invention, including connect successively
The Posture acquisition sensing node for being placed in each characteristic portion of jockey that connects, each sensing node posture information is carried out System and
The data gathering node of forwarding and the 3D human bodies that posture information receive processing and carry out three-dimensionalreconstruction to jockey's posture
The interfaces motion tracking PC.
The interfaces 3D mans motion simulations PC carry out receiving processing and carry out Three-dimensional Gravity to jockey's posture to posture information
The detailed process of structure is as follows:
Step 100:In each limbs characteristic portion of jockey, including pelvis, chest, head, upper arm, underarm, hand, thigh, shank, foot
Posture acquisition sensing node is dressed in portion, and after the completion of wearing, sensing node is connected with aggregation node by USB interface, and uses
The USB power supply interfaces of the lithium battery connection aggregation node of 5V, complete the power supply to whole system;Hardware initialization is completed at this time,
The ends PC mans motion simulation interface is then turned on, after the WIFI hot spot that hardware end is sent out in connection, you can receive each acquisition section
The original posture information of point, including 3-axis acceleration, three axis angular rates and three-axle magnetic field intensity;
Step 200:After system is working properly, jockey faces due north 10 seconds, passes through formulaWithCalculate initial appearance
State information, whereinIndicate the acceleration of three axis,WithFormula can be passed throughAnd formulaCarry out table
Show, whereinWithIndicate the three-axle magnetic field intensity after calibration;Initial state information includes pitch angle, roll angle and boat
To angle, to complete the calibration of initial attitude, after the completion of initial alignment, jockey can freely do various actions;
Step 300:After obtaining the initial state information of jockey, initial Eulerian angles are converted into quaternary number, and at this time
Quaternary number as initial reference amount, it is inclined between sensor coordinate system and bone coordinate system and navigational coordinate system for correcting
Residual quantity, and as reference, the Extended Kalman filter data anastomosing algorithm declined by gradient calculates jockey and is being moved through
The posture information of each bone is mapped to the corresponding skeletal sites of threedimensional model by the athletic posture information of each bone in journey, to
The reconstruct of athletic posture is completed, and is shown the effect of capture by three dimensional constitution by computer graphics.
Preferably, each sensing node separately includes inertial sensor and corresponding microprocessor, and can be independent right
The posture information of each movement node analyze and handle in real time, and analyzing processing is mainly 9 axis informations exported to sensor:
3-axis acceleration, three axis angular rates and three-axle magnetic field intensity carry out data filtering and error correction, including to accelerometer
The zero shift rectifying of low-pass filtering and gyroscope and the ellipsoid fitting correction course for magnetometer;Aggregation node and each sensing
It is attached by high-speed CAN bus between node, after aggregation node is completed to the reception of the attitude data of each acquisition node, is led to
It crosses wireless WIFI module and attitude data is transmitted to the interfaces 3D mans motion simulation PC.
Preferably, the interfaces 3D mans motion simulations PC have embedded the high-precision Extended Kalman filter declined based on gradient
Data anastomosing algorithm, the algorithm pass through formulaCarry out the state of forecasting system, and passes through formulaCarry out the iteration into row coefficient, passes through formulaCarry out continuous update to ride
The posture information of hand during the motion, the posture information indicated by quaternary number, whereinFor predicted vector;
For the output of present filter;F indicates state-transition matrix;μ indicates iteration coefficient;Indicate current quaternary number state;δ t tables
Show the sampling period;Indicate the quaternary number currently measured;K indicates kalman gain.
Preferably, the interfaces 3D mans motion simulations PC, further pass through seat to the attitude data of each sensing node
Mark system depicts.
Preferably, unified concentration supply power is carried out by rechargeable lithium battery between each unit.
Preferably, it is used as with reference to action by upright posture of being specifically exposed to the north, to according to formulaWithIt calculates adaptive
Initial state information completes the calibration of initial attitude.
Preferably, the real-time friendship of hardware end each acquisition node and PC end datas is realized using the wireless transmission method of WIFI
Mutually, sampling rate is up to 100Hz.
Preferably, include mainly to predict and update two based on the Extended Kalman filter data anastomosing algorithm that gradient declines
Stage.Mainly predicted state changes forecast period on the basis of priori, and adoption status matrix F indicates state change;Wherein
Matrix F is:
Wherein ωx, ωy, ωyThree axis angular rates are corresponded to respectively, and δ t indicate the sampling time.The more new stage mainly utilizes calculated
Observed quantity updates predicted stateAnd final posture information is obtained, renewal process can pass through formulaAnd formula K=Pt|t-1·H·(H·Pt|t-1·HT+R)-1Shown in weighted sum carry out table
Show, wherein K indicates the gain of Kalman filtering, Pt|t-1To predict error, and Pt|t-1=(FQFT)+Q;H-matrix is will be true
Real state space is mapped to the observation model of observation space, can carry out H=I with formula4×4It indicates;R matrixes indicate observation covariance, can
With formula 0.05I4×4To indicate.
Preferably, the sensing node includes 9 axle sensor modules, data processing module and CAN bus communication module,
Aggregation node includes data processing module, CAN bus module and WIFI wireless communication modules, sensing node and aggregation node it
Between Star Network formed by CAN bus, and pass through WIFI module and realize data interaction between host computer and slave computer.
The further advantage of the present invention will be illustrated in the following description, below by attached drawing and example to the present invention
Technical solution carry out deep elaboration.
Description of the drawings
Attached drawing is further illustrated to of the invention herein, and a constitution instruction part, is used for explaining the present invention together, and not
It is construed as limiting the invention.In the accompanying drawings:
Fig. 1 is that equestrian sport process jockey's posture information real-time capture system body-sensing network based on inertia sensing of the present invention shows
It is intended to;
Fig. 2 is the power supply flow of whole system;Please the shade in figure is removed, Thanks for Your Cooperation
Fig. 3 is the whole flow chart of the present invention;
Fig. 4 is the coordinate system schematic diagram that the present invention defines;
Fig. 5 is effect display diagram before and after magnetometer ellipsoid fitting in the present invention;(a) design sketch before being fitted;(b) effect after being fitted
Fruit is schemed;
Fig. 6 is the Extended Kalman filter fusion process based on gradient descent method in the present invention;
Fig. 7 is the comparison diagram of the arithmetic accuracy and optical system of this embodiment;
Fig. 8 is each knot vector definition of human body and the displaying of three-dimensionalreconstruction effect of this embodiment;
Fig. 9 is the three-dimensionalreconstruction design sketch of horsemanship posture in this embodiment.
Specific implementation mode
The embodiment of the present invention is illustrated below in conjunction with the accompanying drawings, separately note that preferred embodiment example described herein
It is used only for explaining the present invention, not limit.
Embodiment according to the present invention provides a kind of equestrian sport process jockey's posture information based on inertia sensing
Real-time capture system.As shown in attached drawing 1- Fig. 8, this embodiment includes mainly:
(1) the athletic posture acquisition node based on inertial sensor, can be real by being worn on the characteristic node position of limb motion
When acquisition athletic posture raw information;
(2) it is used to obtain sensing node and carries out the data gathering node of data processing and forwarding, aggregation node is by microprocessor
It is integrated in one with WIFI module, compact can be forwarded attitude data in real time by WIFI module;
(3) real-time reception processing can be carried out to the data that WIFI is transmitted, and drives 3D manikins, during realizing to horsemanship
Upper computer software of jockey's posture into line trace.
In this embodiment, sensing node includes the inertial sensor unit and a microprocessor operation core of nine axis
The heart, nine axis inertial sensors include three axis accelerometer, three-axis gyroscope, three axis magnetometers, and nine axis inertia sensing units pass through
SPI or iic bus are connect with MCU microprocessor cores, and node and is affixed with limbs by being embedded in customization bandage, can be with
Human motion initial signal is parsed in real time.
In this example, aggregation node includes the calculation process core MCU of a data processing, is also integrated with a WIFI mould
Block is attached between aggregation node and sensing node by high-speed CAN bus, and the real-time reception by way of data outage
The data-signal of sensing node, and by the TCP-IP protocol stacks of burning inside WIFI after centainly handling, by data transmission
To host computer.Connection diagram between sensing node and aggregation node is as shown in Figure 1.
In this embodiment, upper computer software part is by Visual Studio in Directed 3D graphic development libraries
On the basis of write, first by 3 dimension module files of 3Dmax Software Create human bodies, by Visual Studio to model
File is loaded, and carries out covering, and the host computer model for meeting kinesiology is designed in the operations such as scene beautification, and on
TCP-IP protocol stacks are embedded in the machine of position, are capable of the signal of real-time reception WIFI transmission, at software end is to signal
The operations such as reason and correction, are converted into the characteristic that can be used for driving 3D manikins.
Equestrian sport process jockey's posture based on micro-inertia sensor of this example captures system, has the characteristics that:
(1) it is traditional based on it is optical motion-captured on the basis of, the acquisition interface using inertial sensor as posture information;
It (2) can be to acceleration during human motion, angular speed, whole body magnetic field, spatial movement displacement, spatial movement angle etc.
Information is measured in real time;
(3) tracking display of 3 dimensions can be carried out to the athletic posture of whole body and hand, and can be shown in real time in host computer
Every data target of human motion;
(4) WIFI module is applied to carry out the communication of data under TCP-IP agreements;
The inertia sensing motion capture system of this embodiment relates generally to micro-inertia sensor technology, wireless telecommunications skill
Art, the real-time acquisition technique of data, computer graphics techniques, human efficiency subject's technology and navigation technology etc. have one
Fixed practical value.It is primarily useful for the design of video display subject matter, game special makes, medical rehabilitation, athletic training, robot control
Equal fields.
The key technology that this embodiment focuses on solving mainly includes the design of sensing node, sensing node original motion letter
Number processing, the transmission mode of data-signal, the 3D model foundations of software end, resolving strategy of posture feature parameter etc..This reality
In example, by anatomy and human cinology's principle, Whole Body bone is divided into 15 nodes, and establish stratification bone mould
Type builds body-sensing network, the data of movable joint node are caught in realization by placing mini inertial sensor at characteristic node
It catches, and corresponding posture information is converted by filtering integration technology, host computer is transferred to by wireless transmission protocol, it is upper
Posture information is resolved with inverse kinematics algorithm in machine, is finally used for driving host computer 3D manikins, and export corresponding
Exercise data.
During this real-time example, Major Difficulties concentrate in the processing of data, at acquisition node end, pass through 9 axis inertia
Sensor can acquire the angular speed in 3 dimension spaces, acceleration and Geomagnetic signal in real time, how from these initial signals
Accurate posture information is extracted, is the difficult point part of the present invention, and the movement of human body in three dimensions, we can be by it
It is specifically described as the movement of relative angle between the displacement of space and bone, therefore with specific reference to Strapdown Inertial Units in this example
Navigation algorithm first does acceleration and angular speed the pretreatment of low pass and high-pass filtering, and then use is declined based on gradient
9 axis movable information of Extended Kalman filter core algorithm pair carries out posture fusion, calculates the quaternary number during human motion
Movable information.
Attitude data processing is related to the conversion of multiple coordinate systems in this embodiment, includes mainly sensor own coordinate
System, geomagnetic coordinate system and skeleton coordinate system and screen coordinate system.Gone out first by sensor measurement in data handling procedure
Signal under sensor own coordinate, and be transformed under human body coordinate system by transformational relation, finally it is transformed into navigation coordinate
Under system, it is used for the movement of driving model, the definition of coordinate system to be as shown in Figure 4.
It analyzes on the whole, the acquisition of athletic posture data, the solution of athletic posture feature is mainly included in this embodiment
It calculates, and shows that three parts, part of data acquisition include the design of sensing unit, microprocessor for the figure reconstruction of characteristic parameter
Core ARM microcontrollers design, power supply managing design etc.;It includes being resolved from original signal that athletic posture feature, which resolves mainly,
The parameter of jockey's motion feature can specifically be indicated by going out;Figure reconstruction display portion includes mainly the reception processing and 3 dimensions to data
The structure of model.Wherein the resolving of athletic posture is the part of whole system core the most again, determines the quality of whole system.
The flow of whole system can refer to Fig. 3.
Specifically, in equestrian sport process jockey posture capture system of this embodiment based on micro-inertia sensor,
Include mainly body-sensing network design, the design of motion capture algorithm, power management section design, upper computer software design, each portion
The setting divided can refer to as follows:
Body-sensing network design --- -- body-sensing network is mainly by being worn on the sensor node structure at each physical trait position
At, the exercise data for obtaining each characteristic node, the entire body-sensing network of this embodiment includes sensing node and aggregation node,
It is communicated using CAN bus between sensing node and aggregation node, aggregation node is packaged after receiving sensing node data, is passed through
WIFI is transmitted to host computer in real time.Sample frequency is up to 100HZ.In practical donning process, by human body be abstracted as by 16 according to
Secondary connected rigid body is constituted, respectively pelvis, chest, head, upper arm, underarm, hand, thigh, shank, 16 nodes of foot, and
A sensing node, to constitute body-sensing network, the movable information for acquiring each node in real time are arranged at every section of rigid body position.
The embodiment of wherein sensing node and aggregation node is as follows:
Sensing node:In this embodiment, sensing node uses the microcontroller of CortexTM-M3 kernels
STM32F103T8U6 is arithmetic core, each onboard three axis accelerometer of node, a three-axis gyroscope and three axis
Magnetometer, and CAN bus connecting interface and power supply chip are configured, realize cascade and communication between each node.
Aggregation node:In this embodiment, aggregation node uses the microcontroller of CortexTM-M4 kernels
STM32F407VGT6 is the development board of core, which is based on high-performanceCortexTM- M432 RISC cores, can
In the up to operation at frequencies of 168MHz.Be integrated with high speed in-line memory (flash memory is up to 1Mb, SRAM be up to 192Kb),
Up to 4Kb backups SRAM, and two APB buses of connection, three ahb bus and ahb bus matrix more than 32 it is a large amount of
Enhanced I/O and peripheral hardware.And support floating-point operation that can greatly improve the rate of operation using this chip.Converge section simultaneously
Point is also integrated with the EMW3162wifi modules of Shanghai Qing Ke companies, can the data of aggregation node be passed through WIFI moulds in real time
The TCP-IP protocol stacks of burning forward the data to upper computer software inside block.In addition aggregation node end and sensing node end connect
Mouth uses high-speed CAN bus, aggregation node to receive the data of each sensing node in real time by way of receiving and interrupting.
The solution process of system algorithm design --- -- posture in this embodiment includes mainly that coordinate defines, and sensor misses
Difference calibration, initial attitude prediction, the spreading kalman posture renewal declined based on gradient and athletic posture tracking, each section tool
Body is as follows:
Coordinate defines ----in this experiment, three kinds of coordinate systems are defined in total, and specific definition can refer to Fig. 4 institutes
Show.Respectively:
Navigational coordinate system (GCS):GCS defined in the present invention is north-east-ground coordinate system, and respectively X-axis is directed toward due north, Y
East is directed toward in the vertical equator of axis, and Z axis is perpendicularly oriented to ground.
Sensor coordinate system (SCS):Coordinate system as defined in sensor itself.
Body coordinate system (BCS):The coordinate system that each body part of jockey defines.
Sensor error is calibrated ----during predicting attitude of carrier, sensor measurement errors are that posture is caused to be lost
An accurate big factor, therefore before the posture of prediction carrier, need to calibrate the measurement error of sensor.It is real in this implementation
In example, following procedure is divided into the calibration of sensor:
Accelerometer:By doing spectrum analysis to the output of acceleration, it is found that the output of accelerometer contains more
High fdrequency component, data oscillation phenomenon is than more serious, and in order to filter off the high frequency section in signal, low-pass filtering is used in this example
Device carries out low-pass filtering to acceleration signal.
Gyroscope:The error of gyroscope is mainly manifested in drift, and it is defeated when static that drift can be understood as gyroscope
Go out fluctuating or fluctuation that signal surrounds zero.The drift for counting gyroscope by weighted average in this embodiment, with top
Difference operation is done in the output of spiral shell instrument, to influence of the removal drift to follow-up Attitude Calculation as far as possible.
Magnetometer:Due to the interference of device itself error and ambient enviroment magnetic field, the output of magnetometer usually exists centainly
Deviation, therefore different environment needs are targetedly calibrated.In the case of no deviation, three axis of magnetometer export
It is rendered as a Surface of Sphere in three dimensions.But often there are various interference in real process, for convenience of calculation, usually will
The model simplification of magnetic compass is at ellipsoidal model, the deviation of ellipsoid circle center distance origin (0,0,0)That is Hard Magnetic
It influences (sensor inherently interference) unrelated with earth's magnetic field, this is influenced and the ferromagnetic material of sensor local environment has
It closes, and soft magnetism influence (error generated by earth's magnetic field interference) is very small, generally ignores.Ellipsoid fitting equation such as following formula
It is shown:
WhereinIt is the offset for influencing to bring by Hard Magnetic;It is three axle magnetometer
Measurement data;A, b, c are that ellipsoid Semi is long;R is the long constant of earth's magnetic field mould.It is fitted the target of ellipsoid, exactly finds out center of circle offsetWith half axial length (a, b, c) of ellipsoid.The front and back effect of fitting is as shown in Figure 5.
Initial attitude is predicted ----in order to ensure the accuracy of attitude prediction, it is necessary to initial attitude is predicted, by
In in real process, our final purposes are the movement appearances for predicting the corresponding body part of sensor under navigational coordinate system
State, but since sensing node places problem (can not possibly just be placed on same position every time), we can not possibly uniquely determine biography
Transforming relationship between sensor and corresponding limbs.It is therefore necessary to the transforming relationships between sensor and corresponding limbs to carry out certainly
Adapt to calibration.In this embodiment, initial alignment process is as follows:Jockey makes system BCS shown in Fig. 4, to face due north standing
X-axis towards due north, Y-axis is directed toward positive east, and Z axis is downward perpendicular to the ground, final that BCS and GCS is kept to overlap, and keeps the process
Several seconds.And the initial aligned relationship of SCS to GCSIt can be acquired by the value of magnetometer and accelerometer, i.e.,:
Wherein, φin, θin,Roll angle, pitch angle and course angle are indicated respectively,After indicating calibration
3-axis acceleration,Indicate the three-axle magnetic field intensity after calibration.By mutual between quaternary number and Eulerian angles
Conversion, can obtain:
Wherein c and s indicates cos and sin functions.GCS is approximately considered at this time to overlap with BCS, thus from SCS to BCS and it
Between initial aligned relationshipDue to being fixed between sensor and limbs, soBecauseTherefore the initial aligned relationship between BCS to GCS can be found out
Based on gradient decline spreading kalman posture renewal ----Kalman filter, which is one kind, can be effectively estimated sight
Survey the optimization recursion filter of noise states.In real process, it is commonly used to carry out the prediction of position.In this embodiment
In we on the basis of Kalman filter, devise it is a kind of based on the extended Kalman filter of quaternary number come to three-dimensional space
Between in carrier positions predicted.Kalman filter is by establishing the kinetic model of system, in the input of known system
In the case of measurement parameter, to estimate the state of system so that its result estimated is better than other individual measurements
As a result.General general system model can be represented by the following formula:
WhereinFor state variable,For state observation noise.In the present invention
Filtering fusion process mainly include two steps:
Forecast period:Predicted state changes on the basis of priori, and adoption status matrix F indicates state change.
Renewal process:Pass through observed quantityTo be updated to state.
The filtering fusion process of this example is illustrated in fig. 6 shown below.In the actual implementation process, in order to reduce the complexity of algorithm
Degree, using quaternary number as state variable, quaternary number is defined as follows shown in formula:
Wherein q1For real part, q2, q3, q4For imaginary part.Extended Kalman filter process is as follows:
As reference by gyroscope, the form for being converted to quaternary number is shown below forecast period --- --:
In formula,For the quaternary number under previous moment,For current time
Angular velocity vector.Over time, it by constantly integrating come predicted state variable, can be represented by the following formula.
Wherein δ t indicate the time of integration, are generally indicated with the sampling period.At this point, state-transition matrix F can be expressed as:
In order to establish passing noiseModel, the variance on each axis of gyroscope is had estimated in this embodiment, and count
Having calculated covariance matrix Q is:
Wherein A=- ωx-ωy-ωz, B=ωx-ωy+ωz, C=ωx+ωy-ωz,
D=- ωx+ωy+ωz。
E [ω are assumed at this timei]=0 (gyro drift has been removed, and error maintains 0.1/0.2deg/s), and E [ωi·
ωj]=0,So covariance matrix can be expressed as:
Wherein At this point, prediction process can be represented by the following formula:
WhereinFor predicted vector,For the output of present filter.The observation of system is by four of quaternary number
Component forms, and as the state variable of system, is realized by the signal of accelerometer and gyroscope.Pass through acceleration and magnetometer
Carry out calculating observation quaternary number, can usually be attributed to an optimization problem, even if the error of observation quaternary number is substantially equal to
0.We calculate the observation quaternary number of optimization using gradient descent method at this time, and this method returns and currently measures relevant four
First numberWherein coefficient μ is iterated in a manner of such as following formula:
The more new stage --- --, filter updated predicted state using calculated observed quantity in this stageMore
New process can indicate that wherein K indicates the gain of Kalman filtering by weighted sum shown in following formula.
K=Pt|t-1·H·(H·Pt|t-1·HT+R)-1
Wherein Pt|t-1To predict error, and Pt|t-1=(FQFT)+Q.H-matrix is by time of day space reflection to sight
Survey the observation model in space.H=I in this embodiment4×4.R matrixes indicate observation covariance, due to accelerometer and magnetic force
The signal of meter is not directly relevant to model, in real process, we attempted it is a variety of different follow the example of, to obtain optimal knot
Fruit.In common document, R generally takes 0.001I4×4, however R is related to the calculating of kalman gain, in this embodiment
Take 0.05I4×4, the optimum point but small oscillations of appearance to avoid too small due to variance.
In order to compare the precision of data anastomosing algorithm in this embodiment, We conducted optical contrast experiments, and give
The comparing result for having gone out this embodiment and optical system arranges optics in thigh and shank and lumbar regions respectively in experiment
Marker, while whole body dresses sensing node.Fig. 7 gives the donning process and angleonly tracking curve graph and mistake of experimentation
Difference comparison frequency substep histogram.By the Comparative result of optical device and the method in the present invention, it can be found that this implementation is real
Algorithm disclosed in example can accurately track the reference value of optical system output substantially, and the error of quaternary number is leaned on mostly
Nearly zero is in normal state substep, it can thus be seen that the precision of the algorithm provided in the present invention disclosure satisfy that requirement substantially, has one
Fixed reference significance.
Athletic posture tracks ----in order to which the athletic posture to jockey captures, this embodiment, which establishes, meets human body
The 3 D human body vector model of dynamical structure, as shown in figure 8, each bone section is indicated with defined direction vector, every section
It is connected by the artis of setting between bone vector.Whole section of movements of parts of the body can easily pass through combination as a result,
The movement of each body part is expressed.More precisely, the posture of human body can be by during capturing human motion
It is obtained in conjunction with the vector of each human body parts.
Describe the vector definition of each limbs joint in Fig. 8, O in figureBIndicate the reference zero of limbs coordinate system, each limbs
Section bone vector definition be:
Wherein η and γ indicates the definition of corresponding limbs and bone section respectively,Indicate arrow of each bone section at BCS
Measure length, lη,γIndicate the length of the limbs of definition, νη,γIndicate unit vector of the corresponding limbs at BCS.
Determine the posture of each bone section, the most important unit vector for being just to determine each bone section, to according to definition
Bone length, to determine the vector length of each bone section.Bone unit vector is as shown in table 1.By each bone section in the present invention
Regard straight rigid structure as, when human body is in upright state, and when towards due north, each limbs coordinate system BCS for defining at this time with
Navigation is that GCS is overlapped, then can be in the hope of the transforming relationship of each body part and binding sensor according to initial alignment processAnd according to data fusion process, the rotation relationship that can be relative to navigation in the hope of sensorSo each bone section from
Transforming relationship under BCS to GCS can be defined as:
Posture of so each bone vector at navigational coordinate system GCS can be expressed as:
WhereinIndicate the bone vector in t moment,Indicate that each corresponding bone section of t moment turns from BCS
Change the spin matrix under GCS.
1 each bone section unit vector of table defines
Once we obtainPosition of so every section of bone at GCS can be expressed by following formula:
WhereinIndicate corresponding η, position of the bone section under γ relative to his father's bone η, γ -1 bone section.OG
(t) position of t at the time of reference point pelvis is at GCS is indicated.At this point, the position and direction of each bone section just can be obtained in we,
To reconstruct the posture of entire jockey.The three-dimensionalreconstruction design sketch of horsemanship posture in this embodiment is given in Fig. 9.
Power module design --- in-this embodiment, by the test to node, the electric current of each node is about
20mA, aggregation node end include that the electric current of MCU and WIFI module is about 50mA, therefore has done place at power management end in this example
Reason, has chosen the linear voltage regulator TPS79633DCQR of TI, and maximum current can reach 1A or so, is fully able to provide maximum
The operating current of 17 nodes, and power supply is powered using the high-capacity lithium battery of 3.7V, which exists
10000mah or so can about power 2 hours or so to whole system, disclosure satisfy that requirement substantially.
Upper computer software end design --- the main function at-upper computer software end be the signal of slave computer is received, and
The data of each node are parsed, it is made to be converted into the posture feature parameter that can drive three-dimensional motion model, and by its
As the interface of data-driven, it is used for the movement of reconstruction model, enables to realize the three-dimensional tracking for moving real human body.This
Constructed upper computer software end uses Visual Studio2012 in the environment of Direct3D graphic development libraries in embodiment
Under build, first by 3dMAX Software Create manikins, and save as support software end call model file, such as
The formats such as 3ds, OBJ .X, thereafter through Direct3D shape libraries to 3ds, the compatibility of the 3D files of the formats such as OBJ passes through
Visual Studio load 3D model files, and adjust position of the model relative to screen space, certainly can also be
Dermatoglyph is made on the basis of this, and model is made to seem more vivid.Structure to this model is basically completed.In addition,
Software inhouse is embedded in TCP-IP protocol stacks, to complete the reception to WIFI signal in slave computer, and the letter that the ends WIFI are received
It number is handled and is calculated, obtain the quaternary number information that can be used in driving model, and map that corresponding model position,
Realize the data-driven to model.
This embodiment elaborates that equestrian sport process jockey's posture based on micro-inertia sensor captures the design of system
Principle and its realization process, the motion capture system construct the motion capture of whole body respectively using Whole Body joint as object
Sensing node network, and using WIFI Radio Transmission Technologys as medium, realize the wireless transmission to signal, finally by 3 dimension of structure
Motor reorgan platform, real-time tracking of the implementation model to jockey's athletic posture.The system is easy to wear, can be by being embedded in
The wearing of system is completed in tight or customization bandage, in addition such system application prospect is than neck that is wide, being currently known
Just there are the making of video display subject matter, game design, athletic training, robot control etc., therefore, with larger economy and academic ginseng in domain
Examine value.
Finally it should be noted that described above is presently preferred embodiments of the present invention, but the present invention should not be limited to
The embodiment and attached drawing disclosure of that.So every do not depart from the equivalent of the lower completion of spirit disclosed in this invention or repair
Change, both falls within the scope of protection of the invention.
Claims (9)
1. horsemanship jockey's posture information acquisition analysis system, which is characterized in that be placed in each feature of jockey including sequentially connected
The Posture acquisition sensing node at position, each sensing node posture information is carried out the data gathering node of System and forwarding with
And posture information is carried out to receive processing and to the interfaces 3D mans motion simulation PC of jockey's posture progress three-dimensionalreconstruction;
The interfaces 3D mans motion simulations PC carry out receiving processing and carry out three-dimensionalreconstruction to jockey's posture to posture information
Detailed process is as follows:
Step 100:In each limbs characteristic portion of jockey, including pelvis, chest, head, upper arm, underarm, hand, thigh, shank, foot
Posture acquisition sensing node is dressed in portion, and after the completion of wearing, sensing node is connected with aggregation node by USB interface, and uses
The USB power supply interfaces of the lithium battery connection aggregation node of 5V, complete the power supply to whole system;Hardware initialization is completed at this time,
The ends PC mans motion simulation interface is then turned on, after the WIFI hot spot that hardware end is sent out in connection, you can receive each acquisition section
The original posture information of point, including 3-axis acceleration, three axis angular rates and three-axle magnetic field intensity;
Step 200:After system is working properly, jockey faces due north 10 seconds, passes through formulaWithCalculate initial appearance
State information, whereinIndicate the acceleration of three axis,WithPass through formulaAnd formulaCarry out table
Show, whereinWithIndicate the three-axle magnetic field intensity after calibration;Initial state information includes pitch angle, roll angle and boat
To angle, to complete the calibration of initial attitude, after the completion of initial alignment, jockey can freely do various actions;
Step 300:After obtaining the initial state information of jockey, initial Eulerian angles are converted into quaternary number, and at this time
Quaternary number as initial reference amount, it is inclined between sensor coordinate system and bone coordinate system and navigational coordinate system for correcting
Residual quantity, and as reference, the Extended Kalman filter data anastomosing algorithm declined by gradient calculates jockey and is being moved through
The posture information of each bone is mapped to the corresponding skeletal sites of threedimensional model by the athletic posture information of each bone in journey, to
The reconstruct of athletic posture is completed, and is shown the effect of capture by three dimensional constitution by computer graphics.
2. horsemanship jockey posture information acquisition analysis system according to claim 1, which is characterized in that each sensing node
Inertial sensor and corresponding microprocessor are separately included, and the independent posture information to each movement node of energy carries out in fact
When analyze and processing, analyzing processing are mainly 9 axis informations that are exported to sensor:3-axis acceleration, three axis angular rates and three
Axis magnetic field intensity carries out data filtering and error correction, includes the zero shift rectifying of the low-pass filtering and gyroscope to accelerometer
And the ellipsoid fitting correction course for magnetometer;Between aggregation node and each sensing node by high-speed CAN bus into
Row connection, after aggregation node is completed to the reception of the attitude data of each acquisition node, by wireless WIFI module by attitude data
It is transmitted to the interfaces 3D mans motion simulation PC.
3. horsemanship jockey posture information acquisition analysis system according to claim 1, which is characterized in that 3D human motions with
The interfaces track PC have embedded the high-precision Extended Kalman filter data anastomosing algorithm declined based on gradient, which passes through formulaCarry out the state of forecasting system, and passes through formulaCarry out the iteration into row coefficient, passes through formulaConstantly to update the posture information of jockey during the motion, the posture information is logical
Quaternary number is crossed to indicate, whereinFor predicted vector;For the output of present filter;F indicates state-transition matrix;μ
Indicate iteration coefficient;Indicate current quaternary number state;δ t indicate the sampling period;Indicate the quaternary number currently measured;K tables
Show kalman gain.
4. horsemanship jockey posture information acquisition analysis system according to claim 1, which is characterized in that the 3D human bodies fortune
Motion tracking PC is further depicted the attitude data of each sensing node by coordinate system at interface.
5. horsemanship jockey posture information acquisition analysis system according to claim 1, which is characterized in that its each unit it
Between unified concentration supply power is carried out by rechargeable lithium battery.
6. horsemanship jockey posture information acquisition analysis system according to claim 1, which is characterized in that pass through specific court
Northern upright posture is used as with reference to acting, to according to formula WithAdaptive initial state information is calculated, the calibration of initial attitude is completed.
7. horsemanship jockey posture information acquisition analysis system according to claim 1, which is characterized in that utilize the nothing of WIFI
Line transmission mode realizes the real-time, interactive of hardware end each acquisition node and PC end datas, and sampling rate is up to 100Hz.
8. horsemanship jockey posture information acquisition analysis system according to claim 3, which is characterized in that declined based on gradient
Extended Kalman filter data anastomosing algorithm mainly include predict and update two stages, forecast period is mainly in the base of priori
Predicted state changes on plinth, and adoption status matrix F indicates state change;Wherein matrix F is:
Wherein ωx, ωy, ωyThree axis angular rates are corresponded to respectively, and δ t indicate the sampling period;The more new stage mainly utilizes calculated
Observed quantity updates predicted stateAnd final posture information is obtained, the Extended Kalman filter based on gradient decline
Renewal process can pass through formulaAnd formula K=Pt|t-1·H·(H·Pt|t-1·HT+R)-1
Shown in weighted sum indicate that wherein K indicates the gain of Kalman filtering, Pt|t-1To predict error, and Pt|t-1=(F
Q·FT)+Q;H-matrix is that the observation model of time of day space reflection to observation space can be carried out H=I with formula4×4It indicates;R squares
Matrix representation observes covariance, can use formula 0.05I4×4To indicate.
9. horsemanship jockey posture information acquisition analysis system according to claim 1, which is characterized in that the sensing section
Point include 9 axle sensor modules, data processing module and CAN bus communication module, aggregation node include data processing module,
CAN bus module and WIFI wireless communication modules, Star network is formed between sensing node and aggregation node by CAN bus
Network, and the data interaction between host computer and slave computer is realized by WIFI module.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810256743.3A CN108447077A (en) | 2018-03-27 | 2018-03-27 | A kind of horsemanship jockey posture information acquisition analysis system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810256743.3A CN108447077A (en) | 2018-03-27 | 2018-03-27 | A kind of horsemanship jockey posture information acquisition analysis system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108447077A true CN108447077A (en) | 2018-08-24 |
Family
ID=63197418
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810256743.3A Withdrawn CN108447077A (en) | 2018-03-27 | 2018-03-27 | A kind of horsemanship jockey posture information acquisition analysis system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108447077A (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109745205A (en) * | 2019-01-31 | 2019-05-14 | 西安交通大学 | A kind of multiple information acquisition system of wearable lower limb exoskeleton |
CN109785416A (en) * | 2018-12-18 | 2019-05-21 | 武汉西山艺创文化有限公司 | A kind of three-dimensional motion capture system and method |
CN111078489A (en) * | 2018-10-19 | 2020-04-28 | 宏达国际电子股份有限公司 | Electronic device and attitude correction method thereof |
CN111693019A (en) * | 2020-05-20 | 2020-09-22 | 西安交通大学 | Attitude sensing device and data fusion and attitude calculation method |
CN112037312A (en) * | 2020-11-04 | 2020-12-04 | 成都市谛视科技有限公司 | Real-time human body posture inverse kinematics solving method and device |
CN114234963A (en) * | 2021-12-20 | 2022-03-25 | 北京华如科技股份有限公司 | Device and method for recognizing individual posture in individual training confrontation device |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102252676A (en) * | 2011-05-06 | 2011-11-23 | 微迈森惯性技术开发(北京)有限公司 | Method and related equipment for acquiring movement attitude data and tracking human movement attitude |
CN103136912A (en) * | 2013-03-05 | 2013-06-05 | 广西师范大学 | Moving posture capture system |
CN103135765A (en) * | 2013-02-20 | 2013-06-05 | 兰州交通大学 | Human motion information capturing system based on micro-mechanical sensor |
CN202974288U (en) * | 2012-12-05 | 2013-06-05 | 南京理工大学 | Miniature strapdown navigation attitude system |
CN104523282A (en) * | 2015-01-09 | 2015-04-22 | 大连理工大学 | Wearable person and horse action monitoring method and system |
-
2018
- 2018-03-27 CN CN201810256743.3A patent/CN108447077A/en not_active Withdrawn
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102252676A (en) * | 2011-05-06 | 2011-11-23 | 微迈森惯性技术开发(北京)有限公司 | Method and related equipment for acquiring movement attitude data and tracking human movement attitude |
CN202974288U (en) * | 2012-12-05 | 2013-06-05 | 南京理工大学 | Miniature strapdown navigation attitude system |
CN103135765A (en) * | 2013-02-20 | 2013-06-05 | 兰州交通大学 | Human motion information capturing system based on micro-mechanical sensor |
CN103136912A (en) * | 2013-03-05 | 2013-06-05 | 广西师范大学 | Moving posture capture system |
CN104523282A (en) * | 2015-01-09 | 2015-04-22 | 大连理工大学 | Wearable person and horse action monitoring method and system |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111078489A (en) * | 2018-10-19 | 2020-04-28 | 宏达国际电子股份有限公司 | Electronic device and attitude correction method thereof |
CN111078489B (en) * | 2018-10-19 | 2023-03-10 | 宏达国际电子股份有限公司 | Electronic device and attitude correction method thereof |
CN109785416A (en) * | 2018-12-18 | 2019-05-21 | 武汉西山艺创文化有限公司 | A kind of three-dimensional motion capture system and method |
CN109745205A (en) * | 2019-01-31 | 2019-05-14 | 西安交通大学 | A kind of multiple information acquisition system of wearable lower limb exoskeleton |
CN111693019A (en) * | 2020-05-20 | 2020-09-22 | 西安交通大学 | Attitude sensing device and data fusion and attitude calculation method |
CN111693019B (en) * | 2020-05-20 | 2021-04-20 | 西安交通大学 | Attitude sensing device and data fusion and attitude calculation method |
CN112037312A (en) * | 2020-11-04 | 2020-12-04 | 成都市谛视科技有限公司 | Real-time human body posture inverse kinematics solving method and device |
CN112037312B (en) * | 2020-11-04 | 2021-02-09 | 成都市谛视科技有限公司 | Real-time human body posture inverse kinematics solving method and device |
CN114234963A (en) * | 2021-12-20 | 2022-03-25 | 北京华如科技股份有限公司 | Device and method for recognizing individual posture in individual training confrontation device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108447077A (en) | A kind of horsemanship jockey posture information acquisition analysis system | |
CN101579238B (en) | Human motion capture three dimensional playback system and method thereof | |
CN103488291B (en) | Immersion virtual reality system based on motion capture | |
CN106153077B (en) | A kind of initialization of calibration method for M-IMU human motion capture system | |
CN201431466Y (en) | Human motion capture and thee-dimensional representation system | |
CN108939512A (en) | A kind of swimming attitude measurement method based on wearable sensor | |
CN107898466A (en) | A kind of limb motion based on inertial sensor catches system and method | |
CN110327048B (en) | Human upper limb posture reconstruction system based on wearable inertial sensor | |
CN107330967A (en) | Knight's athletic posture based on inertia sensing technology is caught and three-dimensional reconstruction system | |
CN203673431U (en) | Motion trail virtual device | |
CN102323854A (en) | Human motion capture device | |
CN109284006B (en) | Human motion capturing device and method | |
CN109785416A (en) | A kind of three-dimensional motion capture system and method | |
CN110609621B (en) | Gesture calibration method and human motion capture system based on microsensor | |
CN109846487A (en) | Thigh measuring method for athletic posture and device based on MIMU/sEMG fusion | |
CN101246601A (en) | Three-dimensional virtual human body movement generation method based on key frame and space-time restriction | |
CN109000633A (en) | Human body attitude motion capture algorithm design based on isomeric data fusion | |
CN109166181A (en) | A kind of mixing motion capture system based on deep learning | |
CN103886588A (en) | Feature extraction method of three-dimensional human body posture projection | |
CN108269309B (en) | Parachute jumping operation action quantitative evaluation method and system | |
WO2024094227A1 (en) | Gesture pose estimation method based on kalman filtering and deep learning | |
Tao et al. | Human modeling and real-time motion reconstruction for micro-sensor motion capture | |
CN108734762A (en) | Movement locus emulation mode and system | |
Lin et al. | Using hybrid sensoring method for motion capture in volleyball techniques training | |
CN106112997B (en) | Ectoskeleton clothes |
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 | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20180824 |
|
WW01 | Invention patent application withdrawn after publication |