CN103679712A - Human body posture estimation method and human body posture estimation system - Google Patents
Human body posture estimation method and human body posture estimation system Download PDFInfo
- Publication number
- CN103679712A CN103679712A CN201310634030.3A CN201310634030A CN103679712A CN 103679712 A CN103679712 A CN 103679712A CN 201310634030 A CN201310634030 A CN 201310634030A CN 103679712 A CN103679712 A CN 103679712A
- Authority
- CN
- China
- Prior art keywords
- human body
- attitude
- sensor
- human
- bayesian network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Abstract
The invention provides a human body posture estimation method and a human body posture estimation system. The method includes the following steps: A, building a human body structuralized model, and combining with information of a sensor to calculate posture of human body motion and position information in a geodetic coordinate system; B, building a Bayesian network model of human body motion, combining with the posture of human body motion and the position information in the geodetic coordinate system to solve the geodetic coordinate system, and predicting human body posture of a later frame by a belief propagation algorithm in the Bayesian network model through human body posture data of a former frame. The human body posture estimation method and the human body posture estimation system have the advantages that human body posture can be predicted, operation efficiency of the whole system is improved remarkably, and predicting instantaneity of the system is guaranteed.
Description
Technical field
The present invention relates to data processing field, relate in particular to human body attitude method of estimation and system.
Background technology
The existing method that human body attitude is estimated is mainly the technology based on Digital Image Processing at present, relates to computer vision field.
Wherein a kind of is with common camera collection human motion video sequence, then human motion video image input computing machine, use on computers Digital Image Processing correlation technique to process and (by background subtraction, divide and obtain human body silhouette human body image, extract human body contour outline, human body contour outline is carried out to thinning processing and obtain human bone stringing, in conjunction with some algorithms, obtain each articulation point position of human body again, thereby construct on computers manikin), then in conjunction with human body real motion situation, estimate the motion of manikin, and contrast with the video image collecting, and then the motion of manikin is corrected, make it more approach human body real motion.
Thisly by the visual meter attitude of articulated type object (as staff or human body) in image sequence of getting it right, estimate, be the important research content in computer vision field, its application comprises (Kinect) such as nature man-machine interaction, intelligent monitoring, virtual reality, robot motion's controls.
The motion analysis technique of the second based on vision can be called optical profile type and said method is similar, just when front end data obtains, what adopt is high-speed camera, the luminescent marking that post at each position of human body, high-speed camera is with higher frequency record human motion, on computers computing is processed in the locus of the luminous point on human body in video image again, thereby purely restore the motion process of human body, but the method does not relate to the estimation to human body attitude.This technology is widely used in the association areas (3D game) such as motion analysis, game making, film special efficacy, sports and scientific research training.
Another technology is the method based on inertial sensor, what be to utilize with optical profile type difference is that the inertial sensors such as acceleration transducer, gyroscope, magnetometer are the various information that gather human motion, and then extrapolate the motion state of human body, a main application of this technology is movement capturing technology (XSens), be widely used in too cartoon making, film special efficacy, the fields such as athletic training.
The problem of prior art: (common camera) 1, prior art is mainly to utilize video camera to gather the movable information of human body target, this has just determined that it can only carry out to the movement human target within the scope of cameras view the estimation of attitude, this can only be confined in laboratory or room prior art, and usable range is limited.2, prior art be camera acquisition to the basis of video sequence on use digital image processing techniques from picture frame, to extract body motion information, shooting angle problem due to video camera, overlapping and the eclipse phenomena at some position inevitably can occur in the human body image collecting, and this has just increased the difficulty of extracting body motion information.(high-speed camera) 3, when gathering the markd human motion of subsides, need to adopt non-common high-speed camera, although reconstruction human motion precision is higher, this video camera cost is higher, and late time data treatment capacity is too large, and can only in particular space, use.4. the achievement in research that human body attitude is estimated is having actual application value aspect intelligent monitoring, video conference, athletic training, medical analysis, cartoon making, augmented reality.No matter be optical profile type or inertial sensor formula, prior art mainly lays particular emphasis on the motion state of rebuilding human body, but aspect attitude estimation, also there is all poor features of accuracy and real-time in current achievement in research great majority, and most method needs the auxiliary of manual intervention and some hardware devices, apart from realizing the target that accurate tracking under simple condition and attitude estimate, also has a certain distance.Therefore, propose a kind of new human body attitude method of estimation and become very necessary.
Summary of the invention
In order to solve the problems of the prior art, the invention provides a kind of human body attitude method of estimation.
The invention provides a kind of human body attitude method of estimation, comprise the steps:
A. set up organization of human body model, and the information of combined sensor calculates the attitude of human motion and the positional information in earth coordinates;
B. set up the Bayesian network model of human motion, in conjunction with the attitude of human motion and the positional information in earth coordinates, Bayesian network model is solved the human body attitude of frame after predicting by belief propagation algorithm in Bayesian network model by frame human body attitude data before.
As a further improvement on the present invention, in described steps A, comprise the steps:
A1. set up organization of human body model, described organization of human body model comprises articulation point and limb segment;
A2. sensor, to the registration of human body, completes human body attitude and from sensor coordinates, is tied to the conversion of earth coordinates; During registration, allow human body overlap with earth coordinates, now the vector of each limb segment under earth coordinates is the known vG that is expressed as, the vector v S of limb segment below sensor coordinate system changes and obtains by the hypercomplex number of sensor output so, the vector of each moment limb segment under earth coordinates just obtained by hypercomplex number and the vS in this moment afterwards, namely the attitude in this moment of human body limb section;
A3. according to the acceleration information of foot's sensor, carry out touchdown point detection;
A4. human space displace analysis, according to the position of touchdown point, and the vector of each limb segment under earth coordinates obtain the position of each articulation point by the method for recursion, thereby determine the locus of human body.
As a further improvement on the present invention, in described step B, comprise the steps:
B1. according to organization of human body model, set up Bayesian network model, described Bayesian network model comprises at least two-layer human model, every one deck represents the residing state of a certain each node of frame human body, and layer interior nodes connects according to the annexation of human synovial, and interlayer node connects according to corresponding relation;
B2. in Bayesian network model, R is rotation matrix, and T is translation matrix; After surveying, before the human body attitude of frame, first by N frame and N-1 frame, obtain R and T, then with R corresponding in Bayesian network model and T weighted mean as new R and T, wherein N frame is present frame;
B3. upgrade after Bayesian network model, adopt belief propagation algorithm to calculate node to be asked; The direction of propagation of information is for being one way propagation between layers, and N layer only affects N+1 layer, and does not affect N-1 layer, and interlayer is two-way propagation; Iteration propagates until the value of all nodes convergence in Bayesian network model, and now the value of each node of N+1 layer is estimated value.
As a further improvement on the present invention, in described steps A 2, according to the acceleration collecting, calculate the coordinate of each node under terrestrial coordinate.
As a further improvement on the present invention, in described step B2, utilize rotation matrix R and translation matrix T to express the relation between two nodes, each parameter of rotation matrix R and translation matrix T all meets Gaussian distribution simultaneously; In described steps A 3, according to the acceleration information of foot's sensor, carry out the detection of touchdown point; In described step B3, by regulating the variance of each parameter Gaussian distribution, regulate the rate of convergence of belief propagation.
As a further improvement on the present invention, this human body attitude method of estimation also comprises:
C. current attitude and the form of prediction attitude by 3D rendering are shown in real time in terminal.
As a further improvement on the present invention, also comprise collecting method: at related parts of human body binding sensor, and open each sensor collection body motion information; Host node is connected to terminal, opens host node power supply, and start to receive the movable information that each sensor sends.
The invention provides a kind of human body attitude estimating system, comprising:
Set up organization of human body model unit, for setting up organization of human body model, and the information of combined sensor calculates the attitude of human motion and the positional information in earth coordinates;
Set up the Bayesian network model unit of human motion, for setting up the Bayesian network model of human motion, in conjunction with the attitude of human motion and the positional information in earth coordinates, Bayesian network model is solved the human body attitude of frame after predicting by belief propagation algorithm in Bayesian network model by frame human body attitude data before.
As a further improvement on the present invention, comprise data acquisition unit, in described data acquisition unit, be included in related parts of human body binding sensor, and open each sensor and gather body motion information, host node is connected to terminal, open host node power supply, and start to receive the movable information that each sensor sends.
As a further improvement on the present invention, described host node is wired or wireless connection with described terminal, and described terminal comprises host computer or mobile terminal.
The invention has the beneficial effects as follows: the present invention can predict human body attitude, and significantly improved the operation efficiency of whole system, guaranteed the real-time of system prediction.
Accompanying drawing explanation
Fig. 1 is human body attitude estimating system theory diagram of the present invention.
Embodiment
The invention discloses a kind of human body attitude method of estimation, comprise the steps:
A. set up organization of human body model, and the information of combined sensor calculates the attitude of human motion and the positional information in earth coordinates;
B. set up the Bayesian network model of human motion, in conjunction with the attitude of human motion and the positional information in earth coordinates, Bayesian network model is solved the human body attitude of frame after predicting by belief propagation algorithm in Bayesian network model by frame human body attitude data before.
In described steps A, comprise the steps:
A1. set up organization of human body model, described organization of human body model comprises articulation point and limb segment;
A2. sensor, to the registration of human body, completes human body attitude and from sensor coordinates, is tied to the conversion of earth coordinates; During registration, allow human body overlap with earth coordinates, now the vector of each limb segment under earth coordinates is the known v that is expressed as
g, the vector v S of limb segment below sensor coordinate system changes and obtains by the hypercomplex number of sensor output so, and the vector of each moment limb segment under earth coordinates is just by this hypercomplex number and v constantly afterwards
sobtain, namely the attitude in this moment of human body limb section;
A3. according to the acceleration information of foot's sensor, carry out touchdown point detection;
A4. human space displace analysis, according to the position of touchdown point, and the vector of each limb segment under earth coordinates obtain the position of each articulation point by the method for recursion, thereby determine the locus of human body.
In described steps A 1, think that human body is the rigid connecting rod model being coupled together by joint and bone here, so-called rigidity, just refers in motion process, the connecting rod that represents skeleton be length constant, can not be out of shape.Model comprises 8 articulation points and 7 limb segments altogether.Certainly different users, when first use system, carry out initialization to model according to self design parameter (such as bone length, hipbone width, foot length);
In described steps A 3, in order to determine spatial attitude and the position of human body, need to carry out the detection of touchdown point.In conjunction with the gait feature in human walking procedure, the acceleration information that utilizes sensor to export can be realized the detection of touchdown point.
In described step B, comprise the steps:
B1. according to organization of human body model, set up Bayesian network model, described Bayesian network model comprises at least two-layer human model, every one deck represents the residing state of a certain each node of frame human body, and layer interior nodes connects according to the annexation of human synovial, and interlayer node connects according to corresponding relation;
B2. in Bayesian network model, R is rotation matrix, and T is translation matrix; After surveying, before the human body attitude of frame, first by N frame and N-1 frame, obtain R and T, then with R corresponding in Bayesian network model and T weighted mean as new R and T, wherein N frame is present frame;
B3. upgrade after Bayesian network model, adopt belief propagation algorithm to calculate node to be asked; The direction of propagation of information is for being one way propagation between layers, and N layer only affects N+1 layer, and does not affect N-1 layer, and interlayer is two-way propagation; Iteration propagates until the value of all nodes convergence in Bayesian network model, and now the value of each node of N+1 layer is estimated value.
In described step B2, because of for people's motion has very large uncertainty, so the Bayesian network that we adopt is real-time update.Experimental results show that the mode that the motion process of each node of human body can add translation by rotation gives full expression to, thus the pass between the every docking point of this network be, node to be asked wherein, known node, R is rotation matrix, T is translation matrix.Before each prediction, first by N frame (present frame) and N-1 frame, obtain R, T, then with R corresponding in network, T weighted mean as new R and T.
In described steps A 2, according to the acceleration collecting and hypercomplex number, calculate the coordinate of each node under terrestrial coordinate.
In described step B2, utilize rotation matrix R and translation matrix T to express the relation between two nodes simultaneously, each parameter of rotation matrix R and translation matrix T all meets Gaussian distribution; In described steps A 3, according to the acceleration information of foot's sensor, carry out the detection of touchdown point; In described step B3, by regulating the variance of each parameter Gaussian distribution, regulate the rate of convergence of belief propagation.
This human body attitude method of estimation also comprises: C. passes through current attitude and prediction attitude form demonstration in real time in terminal of 3D rendering.
Also comprise collecting method: at related parts of human body binding sensor, and open each sensor collection body motion information; Host node is connected to terminal, opens host node power supply, and start to receive the movable information that each sensor sends.
As shown in Figure 1, the invention also discloses a kind of human body attitude estimating system, comprising:
Set up organization of human body model unit, for setting up organization of human body model, and the information of combined sensor calculates the attitude of human motion and the positional information in earth coordinates;
Set up the Bayesian network model unit of human motion, for setting up the Bayesian network model of human motion, in conjunction with the attitude of human motion and the positional information in earth coordinates, Bayesian network model is solved the human body attitude of frame after predicting by belief propagation algorithm in Bayesian network model by frame human body attitude data before.
Comprise data acquisition unit, in described data acquisition unit, be included in related parts of human body binding sensor, and open each sensor collection body motion information, host node is connected to terminal, open host node power supply, and start to receive the movable information that each sensor sends.
Described host node is wired or wireless connection with described terminal, and described terminal comprises host computer or mobile terminal.
The present invention partly uses a plurality of sensor nodes of tying up in related parts of human body at front end data acquisition, acceleration and the hypercomplex number of the corresponding site while gathering human motion, by being wirelessly transmitted to host node, the information exchange then host node being collected is crossed wire transmission to PC terminal or by being wirelessly transmitted to mobile terminal.
The present invention is mainly comprised of following several parts:
1. part of data acquisition: this part comprises a plurality of microsensors, and a host node.Microsensor is for gathering acceleration and the hypercomplex number in each joint of human body, and host node is for sending and receive data to host computer.
2. attitude estimating part: the data of utilizing host node to send, through host computer procedure background process, the attitude that obtains rear several frames is estimated.
3. motion reappearance part: map by 3D, be presented in host computer procedure during by the fructufy obtaining of attitude estimating part.
Human body attitude method of estimation of the present invention and system have following advantage:
1, be not subject to environmental restraint
Sensor assembly can be bundled in human body privileged site, the movable information collecting is processed by being wirelessly transmitted to host computer, this just frees human body target from video camera, has solved classic method moving target and can only be confined to the problem within the scope of video camera.
2, solved the occlusion issue of existing scheme
Due to new departure collection be the data of tying up at the sensor of related parts of human body, just there is not in existing scheme overlapping and occlusion issue when image is processed in this.
3, cheap
Along with the development of sensor technology and the progress of integrated circuit, the integrated sensor assembly of this height is more and more universal, and price is not high yet.
4, estimation effect is good
The present invention adopts Bayesian network, every pair of node in network connects with reference to the annexation of each node in real human body model completely, the numerical expression aspect of its relation, we have adopted rotation matrix and translation matrix, therefore whole system possesses the ability of the human motion of giving full expression to.Simultaneously our real-time update network parameter, utilizes statistical probability to make an estimate, therefore whole system has stronger robustness, still can make prediction comparatively accurately in the human motion of various complexity.Finally, we have adopted belief propagation algorithm to calculate predicted value, and this algorithm has the ability that quickly converges on actual value when solving Bayesian network, therefore adopt this algorithm to significantly improve the operation efficiency of whole system, have guaranteed the real-time of system prediction.
The present invention can estimate human body attitude, and for example, when someone is during in a certain posture, the present invention can estimate next attitude of this people, if this people will fall down, the present invention just can estimate, just can take appropriate measures so, has very wide application.
Above content is in conjunction with concrete preferred implementation further description made for the present invention, can not assert that specific embodiment of the invention is confined to these explanations.For general technical staff of the technical field of the invention, without departing from the inventive concept of the premise, can also make some simple deduction or replace, all should be considered as belonging to protection scope of the present invention.
Claims (10)
1. a human body attitude method of estimation, is characterized in that, comprises the steps:
A. set up organization of human body model, and the information of combined sensor calculates the attitude of human motion and the positional information in earth coordinates;
B. set up the Bayesian network model of human motion, in conjunction with the attitude of human motion and the positional information in earth coordinates, Bayesian network model is solved the human body attitude of frame after predicting by belief propagation algorithm in Bayesian network model by frame human body attitude data before.
2. human body attitude method of estimation according to claim 1, is characterized in that, in described steps A, comprises the steps:
A1. set up organization of human body model, described organization of human body model comprises articulation point and limb segment;
A2. sensor, to the registration of human body, completes human body attitude and from sensor coordinates, is tied to the conversion of earth coordinates; During registration, allow human body coordinate system overlap with earth coordinates, now the vector of each limb segment under earth coordinates is the known vG that is expressed as, the vector v S of limb segment below sensor coordinate system changes and obtains by the hypercomplex number of sensor output so, the vector of each moment limb segment under earth coordinates just obtained by hypercomplex number and the vS in this moment afterwards, namely the attitude in this moment of human body limb section;
A3. according to the acceleration information of foot's sensor, carry out touchdown point detection;
A4. human space displace analysis, according to the position of touchdown point, and the vector of each limb segment under earth coordinates obtain the position of each articulation point by the method for recursion, thereby determine the locus of human body.
3. human body attitude method of estimation according to claim 2, is characterized in that, in described step B, comprises the steps:
B1. according to organization of human body model, set up Bayesian network model, described Bayesian network model comprises at least two-layer human model, every one deck represents the residing state of a certain each node of frame human body, and layer interior nodes connects according to the annexation of human synovial, and interlayer node connects according to corresponding relation;
B2. in Bayesian network model, R is rotation matrix, and T is translation matrix; After surveying, before the human body attitude of frame, first by N frame and N-1 frame, obtain R and T, then with R corresponding in Bayesian network model and T weighted mean as new R and T, wherein N frame is present frame;
B3. upgrade after Bayesian network model, adopt belief propagation algorithm to calculate node to be asked; The direction of propagation of information is for being one way propagation between layers, and N layer only affects N+1 layer, and does not affect N-1 layer, and interlayer is two-way propagation; Iteration propagates until the value of all nodes convergence in Bayesian network model, and now the value of each node of N+1 layer is estimated value.
4. human body attitude method of estimation according to claim 3, is characterized in that, in described steps A 2, according to the acceleration collecting and hypercomplex number, calculates the coordinate of each node under terrestrial coordinate.
5. human body attitude method of estimation according to claim 4, it is characterized in that, in described step B2, utilize rotation matrix R and translation matrix T to express the relation between two nodes simultaneously, each parameter of rotation matrix R and translation matrix T all meets Gaussian distribution; In described steps A 3, according to the acceleration information of foot's sensor, carry out the detection of touchdown point; In described step B3, by regulating the variance of each parameter Gaussian distribution, regulate the rate of convergence of belief propagation.
6. according to the human body attitude method of estimation described in claim 1 to 5 any one, it is characterized in that, this human body attitude method of estimation also comprises:
C. current attitude and the form of prediction attitude by 3D rendering are shown in real time in terminal.
7. human body attitude method of estimation according to claim 6, is characterized in that, also comprises collecting method: at related parts of human body binding sensor, and open each sensor collection body motion information; Host node is connected to terminal, opens host node power supply, and start to receive the movable information that each sensor sends.
8. a human body attitude estimating system, is characterized in that, comprising:
Set up organization of human body model unit, for setting up organization of human body model, and the information of combined sensor calculates the attitude of human motion and the positional information in earth coordinates;
Set up the Bayesian network model unit of human motion, for setting up the Bayesian network model of human motion, in conjunction with the attitude of human motion and the positional information in earth coordinates, Bayesian network model is solved the human body attitude of frame after predicting by belief propagation algorithm in Bayesian network model by frame human body attitude data before.
9. human body attitude estimating system according to claim 8, it is characterized in that, comprise data acquisition unit, in described data acquisition unit, be included in related parts of human body binding sensor, and open each sensor and gather body motion information, host node is connected to terminal, opens host node power supply, and start to receive the movable information that each sensor sends.
10. human body attitude estimating system according to claim 9, is characterized in that, described host node is wired or wireless connection with described terminal, and described terminal comprises host computer or mobile terminal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310634030.3A CN103679712A (en) | 2013-11-29 | 2013-11-29 | Human body posture estimation method and human body posture estimation system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310634030.3A CN103679712A (en) | 2013-11-29 | 2013-11-29 | Human body posture estimation method and human body posture estimation system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103679712A true CN103679712A (en) | 2014-03-26 |
Family
ID=50317163
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310634030.3A Pending CN103679712A (en) | 2013-11-29 | 2013-11-29 | Human body posture estimation method and human body posture estimation system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103679712A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106600626A (en) * | 2016-11-01 | 2017-04-26 | 中国科学院计算技术研究所 | Three-dimensional human body movement capturing method and system |
| CN106846372A (en) * | 2017-02-13 | 2017-06-13 | 南京升渡智能科技有限公司 | Human motion quality visual A+E system and method |
| CN107545242A (en) * | 2017-07-25 | 2018-01-05 | 大圣科技股份有限公司 | A kind of method and device that human action posture is inferred by 2D images |
| CN108230429A (en) * | 2016-12-14 | 2018-06-29 | 上海交通大学 | Real-time whole body posture reconstruction method based on head and two-hand positions and posture |
| CN108965850A (en) * | 2018-07-05 | 2018-12-07 | 盎锐（上海）信息科技有限公司 | The acquisition device and method of human figure |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101579238A (en) * | 2009-06-15 | 2009-11-18 | 吴健康 | Human motion capture three dimensional playback system and method thereof |
| CN102567638A (en) * | 2011-12-29 | 2012-07-11 | 无锡微感科技有限公司 | Interactive upper limb rehabilitation system based on micro-sensor |
-
2013
- 2013-11-29 CN CN201310634030.3A patent/CN103679712A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101579238A (en) * | 2009-06-15 | 2009-11-18 | 吴健康 | Human motion capture three dimensional playback system and method thereof |
| CN102567638A (en) * | 2011-12-29 | 2012-07-11 | 无锡微感科技有限公司 | Interactive upper limb rehabilitation system based on micro-sensor |
Non-Patent Citations (1)
| Title |
|---|
| 谭丹: "基于统计模型估计人体行进跌倒姿态的研究", 《万方学位论文数据库》, 25 December 2012 (2012-12-25), pages 8 - 53 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106600626A (en) * | 2016-11-01 | 2017-04-26 | 中国科学院计算技术研究所 | Three-dimensional human body movement capturing method and system |
| CN106600626B (en) * | 2016-11-01 | 2020-07-31 | 中国科学院计算技术研究所 | Three-dimensional human motion capture method and system |
| CN108230429A (en) * | 2016-12-14 | 2018-06-29 | 上海交通大学 | Real-time whole body posture reconstruction method based on head and two-hand positions and posture |
| CN106846372A (en) * | 2017-02-13 | 2017-06-13 | 南京升渡智能科技有限公司 | Human motion quality visual A+E system and method |
| CN106846372B (en) * | 2017-02-13 | 2020-04-03 | 南京升渡智能科技有限公司 | Human motion quality visual analysis and evaluation system and method thereof |
| CN107545242A (en) * | 2017-07-25 | 2018-01-05 | 大圣科技股份有限公司 | A kind of method and device that human action posture is inferred by 2D images |
| CN107545242B (en) * | 2017-07-25 | 2020-05-26 | 大圣科技股份有限公司 | Method and device for deducing human body action posture through 2D image |
| CN108965850A (en) * | 2018-07-05 | 2018-12-07 | 盎锐（上海）信息科技有限公司 | The acquisition device and method of human figure |
| CN108965850B (en) * | 2018-07-05 | 2020-04-07 | 盎锐（上海）信息科技有限公司 | Human body shape acquisition device and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104700433B (en) | A kind of real-time body's whole body body motion capture method of view-based access control model and system thereof | |
| CN107833271B (en) | Skeleton redirection method and device based on Kinect | |
| CN110599540B (en) | Real-time three-dimensional human body shape and posture reconstruction method and device under multi-viewpoint camera | |
| CN104680582B (en) | A kind of three-dimensional (3 D) manikin creation method of object-oriented customization | |
| CN104167016B (en) | A kind of three-dimensional motion method for reconstructing based on RGB color and depth image | |
| CN108846867A (en) | A kind of SLAM system based on more mesh panorama inertial navigations | |
| CN103679712A (en) | Human body posture estimation method and human body posture estimation system | |
| WO2020054442A1 (en) | Articulation position acquisition method and device, and motion acquisition method and device | |
| JP7427188B2 (en) | 3D pose acquisition method and device | |
| CN110327048B (en) | Human upper limb posture reconstruction system based on wearable inertial sensor | |
| CN110544302A (en) | Human body action reconstruction system and method based on multi-view vision and action training system | |
| CN106815855A (en) | Based on the human body motion tracking method that production and discriminate combine | |
| CN111489392B (en) | Single target human motion posture capturing method and system in multi-person environment | |
| CN111539300A (en) | Human motion capture method, device, medium and equipment based on IK algorithm | |
| CN116206370B (en) | Driving information generation method, driving device, electronic equipment and storage medium | |
| KR20210058686A (en) | Device and method of implementing simultaneous localization and mapping | |
| CN203630822U (en) | Virtual image and real scene combined stage interaction integrating system | |
| CN113033369A (en) | Motion capture method, motion capture device, electronic equipment and computer-readable storage medium | |
| CN102479386A (en) | Three-dimensional motion tracking method of upper half part of human body based on monocular video | |
| CN113984068A (en) | Positioning method, positioning apparatus, and computer-readable storage medium | |
| Phan et al. | Towards 3D human posture estimation using multiple kinects despite self-contacts | |
| CN203630717U (en) | Interaction system based on a plurality of light inertial navigation sensing input devices | |
| Nonnarit et al. | Hand tracking interface for virtual reality interaction based on marg sensors | |
| CN113916221A (en) | Self-adaptive pedestrian track calculation method integrating visual odometer and BP network | |
| CN111435535A (en) | Method and device for acquiring joint point information |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20140326 |
|
| RJ01 | Rejection of invention patent application after publication |