CN103019024A - System for realtime and accurate observation and analysis of table tennis rotating and system operating method - Google Patents
System for realtime and accurate observation and analysis of table tennis rotating and system operating method Download PDFInfo
- Publication number
- CN103019024A CN103019024A CN2012104975637A CN201210497563A CN103019024A CN 103019024 A CN103019024 A CN 103019024A CN 2012104975637 A CN2012104975637 A CN 2012104975637A CN 201210497563 A CN201210497563 A CN 201210497563A CN 103019024 A CN103019024 A CN 103019024A
- Authority
- CN
- China
- Prior art keywords
- table tennis
- motion
- module
- camera
- cloud terrace
- 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
Images
Landscapes
- Length Measuring Devices By Optical Means (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention discloses a visual system for realtime and accurate observation and analysis of table tennis rotating and a system operating method, wherein the system is composed of the following modules: an image acquisition and table tennis target identification and location module, a table tennis space motion KALMAN filter module, a pan-tilt motion plan and control module, a table tennis mark recognition and three-dimensional reconstruction module, a table tennis rotating motion particle filter module and a three-dimensional virtual scene reproduction module. The biggest characteristic of the system for the realtime and accurate observation and analysis of the table tennis rotating disclosed by the invention is that the system can observe and analyze the rotating information of the table tennis in real time, to obtain a rotating state of the table tennis on the basis of the space motion of the table tennis, so that abundant and complete table tennis track information is obtained; and the system is still quite high in realtime performance, to provide abundant and accurate information for athlete training and teaching, game realtime judgment and live broadcast, and timely response of an online decision-making automatic system. By introduction of observation and analysis on internal motion of the target, the accuracy of the target motion state estimation and predication is improved, so that the system can be popularized in other operation projects, civil, military and other similar scenes in a relatively scientific manner.
Description
Technical field
The present invention relates to a kind of observation and analysis system to the table tennis rotary motion, be used in ping-pong contest and the state of flight of training to table tennis, mainly be real-time detection and the analysis of its rotation situation, thereby can obtain more complete table tennis ball flying path information, play ball flexibly for the style of playing, the servo ping-pong robot of analyzing the sportsman and establish better basis.
Background technology
In recent years, vision-aided system is existing increasing the application in competitive sports, " hawkeye " system in relaying such as tennis etc.In table tennis, 200910095291.6 Patent Application Publications accurate measurement and prediction table tennis track system and system operation method, the real time processing system of can accurately measure, prediction and three-dimensional reappearing the high-speed flight table tennis track; On this basis, 201010152129.6 Patent Application Publication the ping-pong contest vision support and match analytic system and system operation method, table tennis track is during the games analyzed, can for television relay provides abundanter race statistics data message, also can and train the characteristic analysis that players are provided for the sportsman.
The system that these two patents relate to has solved real-time table tennis track data acquisition and statistical information problem.But for the table tennis project, rotatablely moving of table tennis is to make it possess the key factor of dirigibility, challenge and sight, can get high-caliber screw and can judge correctly that the rotation status of adversary's ball also is a high performance athlete's Necessary Qualities.According to domestic and international gymnast's research, the screw that the paddler of specialty gets turns scooter more than 50 revolutions per seconds, under so high rotating speed, people's naked eyes can't be known rotation status by the observation table tennis, can only estimate by adversary's shot and experience.And by high-speed camera, then can capture in real time the tiny motion change of table tennis between the adjacent image frame, thereby derive in real time and analyze its motion state.
Utilize high-speed camera observation table tennis revolution of basketball, key issue wherein is how image should cover the entire motion track that whole table tennis field removes to obtain table tennis, clearly observe again the local detail (such as the ball mark) of table tennis thus rotation is calculated in observation.Under the limited constraint of resolution of video camera, only can't satisfy its demand with a cover vision system, need could reduce the true complete information of object of which movement in conjunction with the double data that overall situation observation and part are observed.
Summary of the invention
The objective of the invention is for the in-problem improvement of 200910095291.6 patents, propose vision system and a system operation method that improves feasible real-time accurate surveying and analyze table tennis rotary.Be used in ping-pong contest or training the rotation status of table tennis being carried out real-time monitored and analysis, and the movement locus of table tennis and 3 d pose are carried out three-dimensional virtual scene show in real time and the historical record playback.
To achieve the above object of the invention, the present invention proposes the burnt camera combination of a kind of length, the vision system theory of overall Vision and local visual combination, utilize the locus of the burnt overall Vision systematic observation of the weak point table tennis described in 200910095291.6 patents, the long focus video camera of drive motor cradle head control is followed the tracks of table tennis topography, utilize topography that the sign (ball mark) of table tennis is identified and three-dimensional reduction, obtain continuous table tennis 3 d pose change sequence, thereby analyze the rotation status of table tennis.Use video camera and adopt figure and data processed result storage, can carry out real-time follow-up and track record to table tennis, realize above-mentioned multi-functional.
Concrete technical scheme of the present invention is as follows:
The present invention is a kind of real-time accurate surveying and analyzes the table tennis rotary system, comprise the above motion The Cloud Terrace of plural at least short focus video camera, at least more than one long focus video camera, two-freedom, the synchronous device that the control multiple-camera is adopted figure synchronously, the image data acquiring transmitting device, include main control computer and the externally connected with display screen of the system environments of vision running software, described short focus video camera is fixed on the ping-pong table sky, visual field intersection covers the effective coverage of table tennis, carries out image acquisition; Long focus video camera is fixed on the motion The Cloud Terrace in the ping-pong table sky, makes the table tennis in its visual field pursuit movement obtain the local detail image by the motion of controlling The Cloud Terrace; Synchronous device produces the multichannel periodic pulse signal, connects video camera by control line, and the control multiple-camera is adopted figure synchronously; Video camera links to each other with Image Data Acquisition Card by data line; Data collecting card is by the bus of main control computer, and the processor that the data that collect is sent to the image data acquiring transmitting device is processed; Result outputs on the externally connected with display screen of main control computer and shows, the overall Vision system that is comprised of short focus video camera obtains the world coordinates movement locus of table tennis under the world coordinate system take the ball table as reference, drive The Cloud Terrace and follow the tracks of the local detail image that table tennis obtains table tennis, obtain the three dimensions attitude information of table tennis according to the local detail image, thereby analyze the state that rotatablely moves obtain table tennis, obtain the complete movable information of table tennis in conjunction with the information of rotatablely moving of global motion track and table tennis self.
System of the present invention comprises following composition module: image acquisition and table tennis target identification locating module, table tennis spatial movement KALMAN filter module, The Cloud Terrace motion planning and control module, table tennis landmark identification and three-dimensionalreconstruction module, table tennis rotary Motion Particles filtration module, three-dimensional virtual scene reappear module; The system architecture flow process is: the main control computer epigraph gathers and table tennis target identification locating module, through target identification, space orientation is sent into the KALMAN filter module to the data that generate in real time and is carried out filtering, obtains the motion state information such as the current locus of table tennis and speed; The Cloud Terrace motion planning and control module are rotated according to the real time kinematics information control horizontal stage electric machine of table tennis, make the topography that the video camera that is fixed thereon can the track up table tennis, obtain the three dimensions attitude information of table tennis by table tennis landmark identification and three-dimensionalreconstruction module, calculate the real time rotation situation of table tennis via table tennis rotary Motion Particles filtration module, at last the locus of table tennis and 3 d pose track are sent in real time three-dimensional virtual scene and reappeared module and carry out three-dimensional virtual scene and draw.
System of the present invention is divided into the motion state of table tennis the motion of global space position and self rotatablely moves, in the motion process of table tennis, observation and analysis is distinguished in two kinds of motions, again analysis result is carried out data fusion, obtain the entire motion information of table tennis.
The concrete steps of system of the present invention are as follows:
(1) builds hardware system, each camera interior and exterior parameter of semi-automatic demarcation;
(2) study table tennis spatial movement model and the model parameter that rotatablely moves, and online the correction;
(3) vision collecting table tennis space motion path, and obtain its spatial movement state with Kalman filtering;
(4) the control The Cloud Terrace is followed the tracks of table tennis, and calculates the pose parameter of video camera on it according to the motor movement situation;
(5) obtain the topography of table tennis by the The Cloud Terrace camera, identification trade mark feature is also carried out the 3 d pose reduction in conjunction with camera parameters to table tennis;
(6) obtain the state that rotatablely moves of table tennis with particle filter algorithm;
(7) the result data are inputted in real time three-dimensional virtual scene and reappeared module, the three dimensions attitude track of in real time broadcast or playback table tennis in external display screen.
Maximum characteristics of the present invention are can be in real time the rotation information of table tennis to be carried out observation and analysis, obtain its rotary state on the basis of table tennis spatial movement, thereby can obtain abundanter and complete table tennis track information, and still have high real-time, can instruct for training athlete, compete real-time judge and timely reflection live, the on-line decision automated system provide abundanter accurately information.Owing to introduced observation and analysis to target self internal motion, can Effective Raise to the accuracy of target state estimating and forecasting, thereby can be generalized to the similar scenes such as other operation project and civilian, military project more scientificly.
The beneficial effect that the present invention has is as follows:
(1) proposed a kind of can accurate surveying and analyze the System and method for of table tennis rotary, when the table tennis spatial movement is provided, also provide in real time the observation and analysis of himself rotating operation, can be rapidly provide the following constantly Flight Trajectory Prediction result of table tennis in the more accurate high-speed motion for user or automated system;
(2) method and system among the present invention can also be applied to paddler's training guidance, balling effect assessment, and in the automatic judge of ping-pong contest and the on-the-spot broadcasting.
Therefore, the present invention is a kind of very practical, effective high speed table tennis rotary real-time monitored and analytic system, has good application prospect.
Description of drawings
Fig. 1 system hardware structure block diagram;
Fig. 2 system architecture process flow diagram;
Fig. 3 The Cloud Terrace schematic three dimensional views;
Fig. 4 The Cloud Terrace camera coordinates system and cloud platform rotation synoptic diagram;
Fig. 5 horizontal stage electric machine movement angle calculates synoptic diagram;
Fig. 6 The Cloud Terrace camera target following design sketch;
Synoptic diagram is calculated in the reduction of Fig. 7 table tennis three dimensions attitude;
Fig. 8 table tennis rotary state estimation synoptic diagram;
The virtual scene display of Fig. 9 table tennis three dimensions attitude.
Among Fig. 1, the 1st, short focus video camera, the 2nd, long focus video camera, the 3rd, motion The Cloud Terrace, the 4th, synchronous device, the 5th, image data acquiring transmitting device, the 6th, main control computer, the 7th, the system environments of vision running software, the 8th, the externally connected with display screen of main control computer.
Embodiment
Real-time accurate surveying provided by the invention and analyze the table tennis rotary system, by three above video cameras, control multiple-camera adopt synchronously The Cloud Terrace, the The Cloud Terrace motion control device of synchronous device, the image data acquiring transmitting device of figure, two above degree of freedom, the main control computer that includes the system environments of vision running software forms; Wherein two above video cameras are fixed on the ping-pong table sky, and visual field intersection covers the effective coverage of table tennis, carries out image acquisition; A video camera is fixed on the The Cloud Terrace of ball table top, can be by the motion tracking table tennis of The Cloud Terrace motion control device control The Cloud Terrace; The synchronous device of image data acquiring transmitting device produces the multichannel periodic pulse signal, connects video camera by control line, and the control multiple-camera is adopted figure synchronously; Video camera links to each other with Image Data Acquisition Card by data line; Data collecting card is by the pci bus of main control computer, and the processor that the data that collect is sent to the image data acquiring transmitting device is processed; Result outputs on the externally connected with display screen of main control computer and shows.
Accurate surveying and analysis table tennis rotary system mainly comprise following composition function module in real time: image acquisition and table tennis target identification locating module, table tennis spatial movement KALMAN filter module, The Cloud Terrace motion planning and control module, table tennis landmark identification and three-dimensionalreconstruction module, table tennis rotary Motion Particles filtration module, three-dimensional virtual scene reappear module.The system architecture flow process is: the main control computer epigraph gathers and table tennis target identification locating module, through target identification, space orientation is sent into the KALMAN filter module to the data that generate in real time and is carried out filtering, obtains the motion state information such as the current locus of table tennis and speed; The Cloud Terrace motion planning and control module are rotated according to the real time kinematics information control horizontal stage electric machine of table tennis, make the topography that the video camera that is fixed thereon can the track up table tennis, obtain the three dimensions attitude information of table tennis by table tennis landmark identification and three-dimensionalreconstruction module, calculate the real time rotation situation of table tennis via table tennis rotary Motion Particles filtration module; At last the locus of table tennis and 3 d pose track being sent in real time three-dimensional virtual scene reappears module and carries out three-dimensional virtual scene and draw.
The Cloud Terrace motion planning of the present invention and control module, can calculate the movement angle of following the tracks of needed each motor of table tennis according to the movable information of table tennis, and consider that the time-delay of whole system and motor movement accelerate to postpone to be compensated, and realize the real-time stabilization of rapid movement table tennis is followed the tracks of.
Table tennis landmark identification of the present invention and three-dimensionalreconstruction module, by color characteristic the special sign on the table tennis (mainly being trade mark) is identified, and combination is along with the real-time calculating of the pose of camera parameter of The Cloud Terrace motion is upgraded, the volume coordinate of trade mark and the centre of sphere is reduced, thereby obtain table tennis at the current figure of adopting three dimensions attitude information constantly.
Table tennis rotary Motion Particles filtration module of the present invention, set up the model that rotatablely moves of table tennis, count frame table tennis 3 d pose sequence (may table tennis self block cause observation information discontinuous) for front and back, adopt particle filter algorithm, calculate the information that rotatablely moves (rotating speed and direction) of table tennis.
Three-dimensional virtual scene of the present invention reappears module and reappears module for the three-dimensional virtual scene based on OpenGL: arbitrarily angled flight path of watching with the band table tennis spatial attitude information (by the trade mark indication) of replaying user appointment in three-dimensional virtual scene; And output on the display screen.
Real-time accurate surveying provided by the invention and analysis table tennis rotary system and system operation method, described operating procedure is as follows:
(1) builds hardware system, each camera interior and exterior parameter of semi-automatic demarcation;
(2) study table tennis spatial movement model and the model parameter that rotatablely moves, and online the correction;
(3) vision collecting table tennis space motion path, and obtain its spatial movement state with Kalman filtering;
(4) the control The Cloud Terrace is followed the tracks of table tennis, and calculates the pose parameter of video camera on it according to the motor movement situation;
(5) obtain the topography of table tennis by the The Cloud Terrace camera, identification trade mark feature is also carried out the 3 d pose reduction in conjunction with camera parameters to table tennis;
(6) obtain the state that rotatablely moves of table tennis with particle filter algorithm;
(7) the result data are inputted in real time three-dimensional virtual scene and reappeared module, the three dimensions attitude track of in real time broadcast or playback table tennis in external display screen.
Wherein following method is adopted in the motion of control The Cloud Terrace in the said method:
(1) the inside and outside ginseng parameter of the zero-bit of off-line calibration The Cloud Terrace and video camera;
(2) obtain online the spatial movement state of table tennis, and estimate the time-delay from the table tennis image acquisition to the drive motor locomotor system, predict the locus that table tennis is arrived behind this delay time;
(3) calculate that camera observes the needed horizontal stage electric machine angle in table tennis target location from the The Cloud Terrace;
(4) move by tripod head controlling device control The Cloud Terrace.
Wherein following method is adopted in the reduction of the 3 d pose of table tennis in the said method:
(1) utilize the video camera that carries on the The Cloud Terrace to obtain the local large image of table tennis;
(2) the ball cursor position by table tennis in the color characteristic recognition image;
(3) calculate the real-time pose parameter of video camera according to the current movement angle of The Cloud Terrace;
(4) obtain ping-pong racket target volume coordinate in conjunction with the volume coordinate of the table tennis centre of sphere, pose parameter and the image recognition result of video camera, utilize the centre of sphere and ball target relativeness to obtain the 3 d pose of table tennis.
The state of rotatablely moving that wherein obtains table tennis with particle filter algorithm in the said method adopts following method:
(1) off-line is set up the kinematics model of considering the table tennis rotary motion, study and Renewal model parameter;
(2) online with current table tennis 3 d pose input particle filter, the table tennis rotary motion state (comprising current 3 d pose, sense of rotation and speed) that real-time update is current;
Wherein three-dimensional virtual scene reappears the following method that adopts in the said method:
(1) builds the interior stationary body (ball table, room etc.) of virtual scene, the table tennis 3-D display of drawing initial pose by pinup picture with the OpenGL off-line;
(2) import table tennis spatial movement and rotation status information into the three-dimensional virtual scene display module, first by spatial translation mobile virtual table tennis to the designated space position, the posture information (ball mark normal line vector, ball mark direction vector) according to table tennis arranges the ball mark on table tennis surface is drawn again;
Below in conjunction with accompanying drawing, describing the present invention in detail utilizes length focus video camera machine to cooperate, overall Vision and local visual cooperate, realization is to the movement locus of table tennis, particularly the situation of rotatablely moving is carried out observation and analysis, obtain the entire motion information of table tennis flight, and reappear in the scene of virtual three dimensional space.Wherein obtain the table tennis three-dimensional space position and to the introducing in detail based on short focus video camera overall Vision system is existing in 200910095291.6 patents of table tennis spatial movement state filtering, no longer repeat in this instructions to introduce.
Fig. 1 has provided the hardware block diagram of system, whole system is mainly by the above motion The Cloud Terrace (3) of: short focus video camera (1) more than two, long focus video camera (2) more than, two-freedom, the synchronous device (4) that the control multiple-camera is adopted figure synchronously, image data acquiring transmitting device (5), includes main control computer (6) and externally connected with display screen (8) formation of the system environments (7) of vision running software.The system environments of described vision running software (7) comprises acquisition software, video image process software, polyphaser information fusion software, filtering software of the driving of video camera under different operating system, camera video etc., and all are hidden in Core Feature software intersection under the vision software users operation interface, and it is the bridge that hardware system exchanges with the user.The system environments of vision running software and hardware system have formed each functional module of whole vision system.
Short focus video camera (1) is fixed on the ping-pong table sky, and their visual field intersection covers the effective coverage of table tennis; Long focus video camera (2) is fixed on the The Cloud Terrace (3), can pass through the motion of control The Cloud Terrace (3) so that the table tennis of the image pursuit movement of long focus video camera (2); Synchronous device (4) produces the multichannel periodic pulse signal, connects short focus video camera (1) and long focus video camera (2) by control line, and the control multiple-camera is adopted figure synchronously; To lack the data that focus video camera (1) and long focus video camera (2) collect by image data acquiring transmitting device (5) and be sent to main control computer (6), and carry out data by vision software and process; Result outputs on the externally connected with display screen (8) of main control computer and shows.
Fig. 2 is the framework process flow diagram of system, and described system comprises following composition module: table tennis global position locating module (short burnt overall Vision), table tennis spatial movement state estimation module, The Cloud Terrace motion planning and control module, table tennis three-dimensional pose reconstructed module, table tennis rotary state estimation module, three-dimensional virtual scene reappear display module.
The framework flow process of system is: the burnt overall Vision of weak point system obtains the position of table tennis in three dimensions by high speed image acquisition and processing and target localization, then the locator data that calculates is in real time sent into spatial movement state estimation module and carry out Real-Time Filtering, obtain level and smooth stable table tennis locus and movement velocity.
The Cloud Terrace motion planning and control module are according to the spatial positional information of table tennis, calculate so that the image real-time follow-up of the long focus video camera that is fixed thereon is aimed at the required horizontal stage electric machine movement angle of table tennis, and drive these motor movements to required angle, obtain the real-time partial enlarged drawing picture of table tennis thereby can adopt figure by long focus video camera.
Table tennis three-dimensional pose reconstructed module at first obtains the real-time outer ginseng parameter of long burnt camera in conjunction with the initial alignment result of long burnt camera and the current movement angle of horizontal stage electric machine, then the topography of table tennis is processed, obtain the centre of sphere of table tennis and the image coordinate of the ball surface sign (such as trade mark); In conjunction with long burnt camera parameter and image recognition result the three dimensions attitude of table tennis is reduced afterwards.Table tennis rotary state estimation module is carried out filtering to the table tennis three dimensions attitude of several frames in front and back and is estimated that the state that rotatablely moves (direction of rotation and speed) of table tennis is upgraded in real time calculating.At last the spatial movement of table tennis and the information of rotatablely moving are combined, just can obtain the complete information of table tennis sporting flying, thus real-time servo motion mechanism or to the table tennis ball kicking style analysis better.
And reappear display module by three-dimensional virtual scene, and just can be from the arbitrarily angled in real time table tennis three dimensions posture information of demonstration tape identification feature three-dimensional virtual scene, be convenient to the user and observe more intuitively and understand the table tennis situation.
Fig. 3 is the schematic three dimensional views of motion The Cloud Terrace, this The Cloud Terrace has the degree of freedom of two (or more than), have fixedly to add and hold base and The Cloud Terrace can be fixed to assigned address, having the video camera mounting platform can fix video camera and The Cloud Terrace with device, by driving the motor movement of The Cloud Terrace, the picture centre of video camera can be aimed at the optional position in table tennis effective exercise zone.
Fig. 4 The Cloud Terrace camera coordinates system and cloud platform rotation synoptic diagram, the z axle of camera coordinates system from the camera photocentre along the lens direction forward on the The Cloud Terrace, the x axle is imaging plane lateral shaft (corresponding one-tenth image wide), and the y axle is the longitudinal axis (height of corresponding alleged image) of imaging plane.For two-degree-of-freedom cradle head shown in Figure 4, The Cloud Terrace is by two motor-driven, but a motor drives camera around vertical direction (parallel with world coordinate system z axle) rotation, we are referred to as beat; But the x axle rotation of another motor drives camera under The Cloud Terrace camera coordinates system, we are referred to as pitching.
Fig. 5 is that the horizontal stage electric machine movement angle calculates synoptic diagram,
Be camera position, the coordinate in world coordinate system is
Be the table tennis position, the coordinate in world coordinate system is
The true origin of world coordinate system, i.e. ray are pointed in the initial center, the visual field of camera
Rotate
After the angle (wherein
Be the angle of pitch,
Be deflection angle), the center, the visual field of camera becomes table tennis, i.e. ray
Be respectively
With
Projection on the O-X-Y plane,
Coordinate be
,
Coordinate be
Camera visual field central point becomes the needed beat of table tennis by the world coordinate system initial point and is shown in Figure 3
The angle.By geometric relationship as can be known,
The angle is
With
Angle, computing formula is:
When the camera beat
Be about among Fig. 4 after the angle
Plane Rotation has arrived
The plane, at this moment
In order to allow camera visual field central point aim at table tennis, must allow the camera pitching
Angle.
Will
Coordinate substitution (3) (4) (5)
Obtain this moment
The angle and
The angle all is absolute value, in order accurately to appear at the center in the camera visual field by table tennis by the control motor, also must determine its sign symbol.Known camera enters that center, the visual field is the initial point of world coordinate system after the state in place, sets this moment
Value be 0.When beat motor forward (angle value
) time, the center, the visual field of camera can be moved to y axle negative sense; Otherwise the center, the visual field of camera can be moved to y axle forward; When pitching motor forward (angle value
) time, the center, the visual field of camera can be moved to z axle positive dirction; Otherwise the center, the visual field of camera can be moved to z axle negative direction.
According to the image-forming principle of camera, be a ray in world coordinate system as planar central point position correspondence, and table tennis center point just under world coordinate system in order to unifiedly calculate, is arranged on same plane with comparison point, as all at x=
This plane.The intersection point that we establish camera visual field central ray and this plane is
, then
Obtain the coordinate of table tennis in world coordinate system by outside vision
, obtain the position of camera in world coordinate system by camera calibration
, utilize these two coordinates namely can uniquely determine and calculate the angle of pitch that cloud platform rotation is done to be needed
And deflection angle
, and then realize that The Cloud Terrace is to the position tracking of table tennis.
The confidential reference items of camera can not change in the The Cloud Terrace motion process, and the Camera extrinsic on the initial position after the startup also can obtain by off-line calibration, are assumed to be
, thus, when horizontal stage electric machine has moved
The angle and
After the angle, can obtain the post exercise Camera extrinsic by the Douglas Rodríguez conversion and be:
Wherein, the z axle of rotation former world coordinate system (namely
The angle rotating shaft) vector under camera coordinates system is
, and
Fig. 6 is the design sketch by the table tennis of control The Cloud Terrace motion tracking, the table tennis local detail image that obtains by growing burnt camera, can clearly the ball mark feature of table tennis be identified and locate, hsv color space characteristics coupling and the Contour searching method of continuing to use the overall Vision system can extract the image-region of table tennis, obtain the image coordinate location at the table tennis centre of sphere and ball mark center.
Fig. 7 is that synoptic diagram is calculated in the reduction of table tennis three dimensions attitude, according to the camera calibration result, can extend two straight lines that penetrate from the camera photocentre, i.e. centre of sphere straight line among Fig. 7 and ball mark straight lines with parameter and the image recognition result of current camera.Centre of sphere straight line is the line of camera photocentre and the table tennis centre of sphere.Owing to have error between outside vision system and the The Cloud Terrace camera vision system, may be not on the extended centre of sphere straight line by the The Cloud Terrace camera by Table Tennis World's coordinate position that outside vision system provides, because need to provide by The Cloud Terrace camera vision the 3 d pose of table tennis, therefore the sphere centre coordinate that outside vision system is provided is mapped on the centre of sphere straight line in the The Cloud Terrace vision system, with this subpoint as new sphere centre coordinate point in order to calculating.Point centered by new sphere centre coordinate take the radius (2cm) of table tennis as radius, can obtain the spheroid space equation on ball surface:
Thereby according to ball mark straight-line equation and spheroid space equation, can calculate ping-pong racket and be marked on coordinate position in the space.Can determine thus the three dimensions attitude of table tennis.
Adopt at a high speed in the process of figure at video camera, fast image processing and data processing algorithm by vision system, can be adopted the current location of figure table tennis constantly at every turn, in conjunction with the data processed result in front and back cycle, can obtain level and smooth table tennis space motion path and motion state by Kalman filtering.Similarly, the rotation status that obtain table tennis is estimated, also needs to calculate in conjunction with the front and back cycle data.
Fig. 8 is table tennis rotary state estimation synoptic diagram, supposes the rotational speed availability vector of table tennis
Expression, the direction of vector is parallel with the rotor shaft direction of table tennis rotary, and the mould of vector is the rotational speed size
, sense of rotation is determined by right-hand screw rule.Ball mark central point is m, takes up an official post from rotating shaft and gets a bit (acquiescence is got the centre of sphere) to m point work vector
, the linear velocity of m point place rotation is
, then got by geometric relationship
Can get:
Wherein
Can obtain by visual identity result of calculation, because camera is adopted at a high speed figure, all period interval are little,
Can provide by the variable quantity of direct p of two cycle of front and back.
And because the observation of the three dimensions attitude of table tennis can be subject to blocking of ball self and cause observation data discontinuous, also need to carry out filtering just can obtain stable rotation status estimated result, adopted particle filter algorithm among the present invention.
The table tennis revolution of basketball is set up state equation:
Wherein,
For t constantly the rotation status of table tennis (when forecourt cursor position and rotational speed
),
Be controlled quentity controlled variable (air resistance),
With
Be plant noise and observation noise.By the sampling repeatedly of a plurality of cycles, estimate, to proofread and correct and the resampling process, particle filter algorithm can obtain more accurately table tennis rotary state estimation result.
Fig. 9 is the virtual scene display of table tennis three dimensions attitude.The three-dimensional of table tennis shows by increasing the ball mark and intuitively shows, and ball target 3-D display is by the texture mapping realization of OpenGL.After the three dimensions attitude of table tennis had been determined, position that just can be by being controlled at table tennis surface pinup picture and angle control that the ball target is corresponding to be shown, and from obtaining the viewing effect with current actual table tennis state consistency.Control again the frequency of 3-D display according to the high-frequency observed result of vision system, just can in three-dimensional scenic, dynamically show movement locus and the rotary course of table tennis.
Claims (4)
1. a real-time accurate surveying and analyze the table tennis rotary system, comprise the above motion The Cloud Terrace of plural at least short focus video camera, at least more than one long focus video camera, two-freedom, the synchronous device that the control multiple-camera is adopted figure synchronously, the image data acquiring transmitting device, include main control computer and the externally connected with display screen of the system environments of vision running software, described short focus video camera is fixed on the ping-pong table sky, visual field intersection covers the effective coverage of table tennis, carries out image acquisition; Described long focus video camera is fixed on the motion The Cloud Terrace in the ping-pong table sky, makes the table tennis in its visual field pursuit movement obtain the local detail image by the motion of controlling The Cloud Terrace; Described synchronous device produces the multichannel periodic pulse signal, connects video camera by control line, and the control multiple-camera is adopted figure synchronously; Described video camera links to each other with Image Data Acquisition Card by data line; Described data collecting card is by the bus of main control computer, and the processor that the data that collect is sent to the image data acquiring transmitting device is processed; Result outputs on the externally connected with display screen of main control computer and shows, it is characterized in that, the overall Vision system that is comprised of short focus video camera obtains the world coordinates movement locus of table tennis under the world coordinate system take the ball table as reference, drive The Cloud Terrace and follow the tracks of the local detail image that table tennis obtains table tennis, obtain the three dimensions attitude information of table tennis according to the local detail image, thereby analyze the state that rotatablely moves obtain table tennis, obtain the complete movable information of table tennis in conjunction with the information of rotatablely moving of global motion track and table tennis self.
2. real-time accurate surveying according to claim 1 and analyze the table tennis rotary system, it is characterized in that, comprise following composition module: image acquisition and table tennis target identification locating module, table tennis spatial movement KALMAN filter module, The Cloud Terrace motion planning and control module, table tennis landmark identification and three-dimensionalreconstruction module, table tennis rotary Motion Particles filtration module, three-dimensional virtual scene reappear module; Described system architecture flow process is: the main control computer epigraph gathers and table tennis target identification locating module, through target identification, space orientation is sent into the KALMAN filter module to the data that generate in real time and is carried out filtering, obtains the motion state information such as the current locus of table tennis and speed; The Cloud Terrace motion planning and control module are rotated according to the real time kinematics information control horizontal stage electric machine of table tennis, make the topography that the video camera that is fixed thereon can the track up table tennis, obtain the three dimensions attitude information of table tennis by table tennis landmark identification and three-dimensionalreconstruction module, calculate the real time rotation situation of table tennis via table tennis rotary Motion Particles filtration module, at last the locus of table tennis and 3 d pose track are sent in real time three-dimensional virtual scene and reappeared module and carry out three-dimensional virtual scene and draw.
3. real-time accurate surveying according to claim 1 and analyze the table tennis rotary system, it is characterized in that, the motion state of table tennis is divided into the motion of global space position and self rotatablely moves, in the motion process of table tennis, observation and analysis is distinguished in two kinds of motions, again analysis result is carried out data fusion, obtain the entire motion information of table tennis.
4. real-time accurate surveying according to claim 1 and analyze the table tennis rotary system is characterized in that concrete steps are as follows:
Build hardware system, each camera interior and exterior parameter of semi-automatic demarcation;
Study table tennis spatial movement model and the model parameter that rotatablely moves, and online the correction;
Vision collecting table tennis space motion path, and obtain its spatial movement state with Kalman filtering;
The control The Cloud Terrace is followed the tracks of table tennis, and calculates the pose parameter of video camera on it according to the motor movement situation;
Obtain the topography of table tennis by the The Cloud Terrace camera, identification trade mark feature is also carried out the 3 d pose reduction in conjunction with camera parameters to table tennis;
Obtain the state that rotatablely moves of table tennis with particle filter algorithm;
The result data are inputted three-dimensional virtual scene in real time reappear module, the three dimensions attitude track of in real time broadcast or playback table tennis in external display screen.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210497563.7A CN103019024B (en) | 2012-11-29 | 2012-11-29 | Real-time accurate surveying and analysis table tennis rotary system and system operation method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210497563.7A CN103019024B (en) | 2012-11-29 | 2012-11-29 | Real-time accurate surveying and analysis table tennis rotary system and system operation method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103019024A true CN103019024A (en) | 2013-04-03 |
CN103019024B CN103019024B (en) | 2015-08-19 |
Family
ID=47967784
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210497563.7A Active CN103019024B (en) | 2012-11-29 | 2012-11-29 | Real-time accurate surveying and analysis table tennis rotary system and system operation method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103019024B (en) |
Cited By (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103364579A (en) * | 2013-07-02 | 2013-10-23 | 北京理工大学 | Method for predicting ping-pong spin angle velocity of ping-pong robot |
CN103595992A (en) * | 2013-11-08 | 2014-02-19 | 深圳市奥拓电子股份有限公司 | Court LED display screen system and advertisement insertion method thereof capable of achieving precise advertisement delivery |
CN104469325A (en) * | 2014-12-26 | 2015-03-25 | 安徽寰智信息科技股份有限公司 | Video collecting and processing method used for weight throw teaching and training |
CN104469326A (en) * | 2014-12-25 | 2015-03-25 | 安徽寰智信息科技股份有限公司 | Video collecting and processing method used for cricket teaching and training |
CN104486591A (en) * | 2014-12-26 | 2015-04-01 | 安徽寰智信息科技股份有限公司 | Video acquisition and processing method for table tennis teaching and training |
CN104486592A (en) * | 2014-12-26 | 2015-04-01 | 安徽寰智信息科技股份有限公司 | Video acquisition and processing method for javelin teaching and training |
CN104486594A (en) * | 2014-12-26 | 2015-04-01 | 安徽寰智信息科技股份有限公司 | Video acquisition and processing method for squash teaching and training |
CN104506830A (en) * | 2014-12-25 | 2015-04-08 | 安徽寰智信息科技股份有限公司 | Video acquisition and processing device for cricket teaching and training |
CN104539900A (en) * | 2014-12-26 | 2015-04-22 | 安徽寰智信息科技股份有限公司 | Video collecting and processing method for shot-put teaching and training |
CN104581065A (en) * | 2014-12-29 | 2015-04-29 | 安徽寰智信息科技股份有限公司 | Video acquisition and processing method for hockey ball hitting instruction and training |
CN105225269A (en) * | 2015-09-22 | 2016-01-06 | 浙江大学 | Based on the object modelling system of motion |
CN107315349A (en) * | 2017-07-31 | 2017-11-03 | 中科新松有限公司 | The batting motion control method of robot |
CN108568086A (en) * | 2018-04-24 | 2018-09-25 | 上海中医药大学 | A kind of badminton game automatic judge platform and its application method based on Internet of Things |
CN109045661A (en) * | 2018-08-06 | 2018-12-21 | 深圳博特创新科技有限公司 | The automatic method and system for carrying out regular judge are moved for unmanned plane during flying |
CN109711246A (en) * | 2018-09-30 | 2019-05-03 | 鲁东大学 | A kind of dynamic object recognition methods, computer installation and readable storage medium storing program for executing |
CN109872372A (en) * | 2019-03-07 | 2019-06-11 | 山东大学 | A kind of small-sized quadruped robot overall Vision localization method and system |
CN111465434A (en) * | 2017-12-12 | 2020-07-28 | 高尔纵株式会社 | Sensing device for moving ball and method for calculating motion parameter of ball by using same |
CN111939541A (en) * | 2020-06-23 | 2020-11-17 | 北京瑞盖科技股份有限公司 | Evaluation method, device, equipment and system for table tennis training |
CN112702481A (en) * | 2020-11-30 | 2021-04-23 | 杭州电子科技大学 | Table tennis track tracking device and method based on deep learning |
WO2021100516A1 (en) * | 2019-11-22 | 2021-05-27 | ソニーグループ株式会社 | Information processing device, information processing method, and program |
CN114612522A (en) * | 2022-05-09 | 2022-06-10 | 广州精天信息科技股份有限公司 | Table tennis sport parameter detection method and device and table tennis training auxiliary system |
WO2022141197A1 (en) * | 2020-12-30 | 2022-07-07 | 深圳市大疆创新科技有限公司 | Method and device for controlling cradle head, movable platform and storage medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4241716A (en) * | 1979-02-12 | 1980-12-30 | Arco Industries Ltd. | Automatic toy gun for ping pong balls |
JPH08285686A (en) * | 1995-04-11 | 1996-11-01 | Fuji Denshi Kogyo Kk | Color discriminating device for ping-pong ball |
CN101049536A (en) * | 2006-09-21 | 2007-10-10 | 上海大学 | Intellective ping-pong wall for household bodybuilding, and operation method |
CN101458434A (en) * | 2009-01-08 | 2009-06-17 | 浙江大学 | System for precision measuring and predicting table tennis track and system operation method |
CN101833232A (en) * | 2010-04-20 | 2010-09-15 | 浙江大学 | Visual support and match analysis system for ping-pong match and method for running same |
CN102489016A (en) * | 2011-12-28 | 2012-06-13 | 哈尔滨师范大学 | Internal service type table tennis service and swing speed monitoring device |
-
2012
- 2012-11-29 CN CN201210497563.7A patent/CN103019024B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4241716A (en) * | 1979-02-12 | 1980-12-30 | Arco Industries Ltd. | Automatic toy gun for ping pong balls |
JPH08285686A (en) * | 1995-04-11 | 1996-11-01 | Fuji Denshi Kogyo Kk | Color discriminating device for ping-pong ball |
CN101049536A (en) * | 2006-09-21 | 2007-10-10 | 上海大学 | Intellective ping-pong wall for household bodybuilding, and operation method |
CN101458434A (en) * | 2009-01-08 | 2009-06-17 | 浙江大学 | System for precision measuring and predicting table tennis track and system operation method |
CN101833232A (en) * | 2010-04-20 | 2010-09-15 | 浙江大学 | Visual support and match analysis system for ping-pong match and method for running same |
CN102489016A (en) * | 2011-12-28 | 2012-06-13 | 哈尔滨师范大学 | Internal service type table tennis service and swing speed monitoring device |
Cited By (30)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103364579B (en) * | 2013-07-02 | 2015-04-15 | 北京理工大学 | Method and device for predicting ping-pong spin angle velocity of ping-pong robot |
CN103364579A (en) * | 2013-07-02 | 2013-10-23 | 北京理工大学 | Method for predicting ping-pong spin angle velocity of ping-pong robot |
CN103595992B (en) * | 2013-11-08 | 2016-10-12 | 深圳市奥拓电子股份有限公司 | A kind of court LED display screen system and realize advertisement accurately throw in inserting method |
CN103595992A (en) * | 2013-11-08 | 2014-02-19 | 深圳市奥拓电子股份有限公司 | Court LED display screen system and advertisement insertion method thereof capable of achieving precise advertisement delivery |
CN104469326A (en) * | 2014-12-25 | 2015-03-25 | 安徽寰智信息科技股份有限公司 | Video collecting and processing method used for cricket teaching and training |
CN104506830A (en) * | 2014-12-25 | 2015-04-08 | 安徽寰智信息科技股份有限公司 | Video acquisition and processing device for cricket teaching and training |
CN104469325A (en) * | 2014-12-26 | 2015-03-25 | 安徽寰智信息科技股份有限公司 | Video collecting and processing method used for weight throw teaching and training |
CN104486591A (en) * | 2014-12-26 | 2015-04-01 | 安徽寰智信息科技股份有限公司 | Video acquisition and processing method for table tennis teaching and training |
CN104486592A (en) * | 2014-12-26 | 2015-04-01 | 安徽寰智信息科技股份有限公司 | Video acquisition and processing method for javelin teaching and training |
CN104486594A (en) * | 2014-12-26 | 2015-04-01 | 安徽寰智信息科技股份有限公司 | Video acquisition and processing method for squash teaching and training |
CN104539900A (en) * | 2014-12-26 | 2015-04-22 | 安徽寰智信息科技股份有限公司 | Video collecting and processing method for shot-put teaching and training |
CN104581065A (en) * | 2014-12-29 | 2015-04-29 | 安徽寰智信息科技股份有限公司 | Video acquisition and processing method for hockey ball hitting instruction and training |
CN105225269B (en) * | 2015-09-22 | 2018-08-17 | 浙江大学 | Object modelling system based on motion |
CN105225269A (en) * | 2015-09-22 | 2016-01-06 | 浙江大学 | Based on the object modelling system of motion |
CN107315349A (en) * | 2017-07-31 | 2017-11-03 | 中科新松有限公司 | The batting motion control method of robot |
CN107315349B (en) * | 2017-07-31 | 2020-08-21 | 中科新松有限公司 | Ball hitting motion control method of robot |
CN111465434B (en) * | 2017-12-12 | 2022-05-31 | 高尔纵株式会社 | Sensor device for moving ball and method for calculating motion parameter of ball using the same |
CN111465434A (en) * | 2017-12-12 | 2020-07-28 | 高尔纵株式会社 | Sensing device for moving ball and method for calculating motion parameter of ball by using same |
CN108568086B (en) * | 2018-04-24 | 2021-01-26 | 上海中医药大学 | Badminton match automatic referee platform based on Internet of things and use method thereof |
CN108568086A (en) * | 2018-04-24 | 2018-09-25 | 上海中医药大学 | A kind of badminton game automatic judge platform and its application method based on Internet of Things |
CN109045661A (en) * | 2018-08-06 | 2018-12-21 | 深圳博特创新科技有限公司 | The automatic method and system for carrying out regular judge are moved for unmanned plane during flying |
CN109711246A (en) * | 2018-09-30 | 2019-05-03 | 鲁东大学 | A kind of dynamic object recognition methods, computer installation and readable storage medium storing program for executing |
CN109872372A (en) * | 2019-03-07 | 2019-06-11 | 山东大学 | A kind of small-sized quadruped robot overall Vision localization method and system |
WO2021100516A1 (en) * | 2019-11-22 | 2021-05-27 | ソニーグループ株式会社 | Information processing device, information processing method, and program |
CN111939541A (en) * | 2020-06-23 | 2020-11-17 | 北京瑞盖科技股份有限公司 | Evaluation method, device, equipment and system for table tennis training |
CN112702481A (en) * | 2020-11-30 | 2021-04-23 | 杭州电子科技大学 | Table tennis track tracking device and method based on deep learning |
CN112702481B (en) * | 2020-11-30 | 2024-04-16 | 杭州电子科技大学 | Table tennis track tracking device and method based on deep learning |
WO2022141197A1 (en) * | 2020-12-30 | 2022-07-07 | 深圳市大疆创新科技有限公司 | Method and device for controlling cradle head, movable platform and storage medium |
CN114612522A (en) * | 2022-05-09 | 2022-06-10 | 广州精天信息科技股份有限公司 | Table tennis sport parameter detection method and device and table tennis training auxiliary system |
CN114612522B (en) * | 2022-05-09 | 2023-01-17 | 广东金融学院 | Table tennis sport parameter detection method and device and table tennis training auxiliary system |
Also Published As
Publication number | Publication date |
---|---|
CN103019024B (en) | 2015-08-19 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103019024B (en) | Real-time accurate surveying and analysis table tennis rotary system and system operation method | |
CN101458434B (en) | System for precision measuring and predicting table tennis track and system operation method | |
CN103442773B (en) | The sensing apparatus of virtual golf analogue means and use thereof and method for sensing | |
US8994826B2 (en) | Portable wireless mobile device motion capture and analysis system and method | |
CN103458977B (en) | Sensing device and sensing method used for virtual golf simulation device | |
US9320957B2 (en) | Wireless and visual hybrid motion capture system | |
CN101231752B (en) | Mark-free true three-dimensional panoramic display and interactive apparatus | |
CN101836237B (en) | Method and system for selecting the viewing configuration of a rendered figure | |
CN1425171A (en) | Method and system for coordination and combination of video sequences with spatial and temporal normalization | |
CN102576466A (en) | Systems and methods for tracking a model | |
CN102004840A (en) | Method and system for realizing virtual boxing based on computer | |
CN111612826A (en) | High-precision three-dimensional motion track acquisition positioning and motion process reproduction method based on binocular video sensor | |
JPH06105231A (en) | Picture synthesis device | |
CN108434698B (en) | Sports ball game teaching system | |
CN113018827B (en) | Auxiliary training system, method and terminal for accurately collecting and analyzing ping-pong ball tracks | |
CN116703968A (en) | Visual tracking method, device, system, equipment and medium for target object | |
TW202146084A (en) | A portable apparatus, method, and system of golf club swing motion tracking and analysis | |
US20230100572A1 (en) | Methods and systems to track a moving sports object trajectory in 3d using multiple cameras | |
CN116523962A (en) | Visual tracking method, device, system, equipment and medium for target object | |
Andersson | A real experiment in virtual environments: A virtual batting cage | |
CN114241602A (en) | Multi-purpose rotational inertia measuring and calculating method based on deep learning | |
CA2633197A1 (en) | Method and system for creating event data and making same available to be served | |
CN113274733A (en) | Golf ball top-placing type detection method, system and storage medium | |
Myint et al. | Tracking a table tennis ball for umpiring purposes using a multi-agent system | |
KR102518635B1 (en) | Vision AI and IOT based interactive sports system and method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |