CN107945206A - A kind of mobile object track determination methods of view-based access control model sensor - Google Patents
A kind of mobile object track determination methods of view-based access control model sensor Download PDFInfo
- Publication number
- CN107945206A CN107945206A CN201710991072.0A CN201710991072A CN107945206A CN 107945206 A CN107945206 A CN 107945206A CN 201710991072 A CN201710991072 A CN 201710991072A CN 107945206 A CN107945206 A CN 107945206A
- Authority
- CN
- China
- Prior art keywords
- phy
- axis
- mobile object
- information
- coordinate
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/223—Analysis of motion using block-matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/50—Depth or shape recovery
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Toys (AREA)
Abstract
The invention discloses a kind of mobile object track determination methods of view-based access control model sensor, pass through the depth information of the image acquisition of camera, the training pattern of different mobile objects is loaded into detection algorithm function, finally it is fitted track, the present invention can realize the detection of movement locus of object, with easy to use, the advantages of having a wide range of application.
Description
Technical field
The present invention relates to a kind of mobile object track determination methods of view-based access control model sensor.
Background technology
Nowadays billiard movement has worldwide been becoming increasingly popular as current emerging sports item.
From preceding billiard ball as a kind of graceful gentleman's movement, but at present based on sports and life & amusement.With " billiard ball infant genius "
Ding Junhui is known as Hengdeli and the acquisition of " billiard ball emperor " having been defeated in world's Snooker occupation Ranking Tournament in the world
The champion of match.Snooker and billiard movement start upsurge at home again.But influenced by Chinese macroeconomic, the operation of billiard-room
Cost is increasing always, mainly rent and manually, including coach etc..The quantity and scale of billiard-room are to increase, and in the market is competing
Strive also more and more fierce, the experience requirements of consumer groups are also being lifted.
Original billiard ball only have artificial match judgement and training mode, can not pass through the tennis intelligence based on machine vision
Robot system completes the roles such as coach, judge, assiatant to meet that game of billiards judge's law enforcement function and billiard training project are commented
Valency function, associated user check competition data statistical analysis and level of training assay not side after match and after training
Just, human cost is high.And can not achieve the main reason for machine judges automatically at present, it exactly can not accurately judge the movement of object
Track.
The content of the invention
In view of the drawbacks described above of the prior art, the technical problems to be solved by the invention are to provide a kind of more accurately thing
Body movement locus determination methods.
To achieve the above object, the present invention provides a kind of mobile object track determination methods of view-based access control model sensor,
By the depth information of the image acquisition of camera, it is determined in z-axis depth information within the vision, according to camera image information
(x, y-axis) and depth information (z-axis) carry out three-dimensional modeling, and using the center of objective plane as coordinate origin, length direction is as x
Axis, width is as y-axis, perpendicular to the direction of plane as z-axis;
The training pattern of different mobile objects is loaded into detection algorithm function, by the algorithm of convolutional neural networks,
Convolutional calculation is carried out to image information, the matching degree of full articulamentum is compared, when matching degree is higher than threshold value k, then it is assumed that herein
The mobile object of position is matched with the mobile object in training pattern, so as to recognize the position of different mobile objects in the picture
Information and its said tag;
Converted by built-in algorithm, positional information of the different mobile objects under image coordinate system is scaled
Wherein pixel is the pixel value of coordinate points, and f is focal length, and dis is the distance measured by depth camera
Value, phy are the positional value under space coordinates;Positional information (x_phy, y_phy, z_phy) under real space coordinate system, knot
Close the coordinate length model that place is divided into x, y, z three-dimensional shafts, i.e. origin (0,0,0) and three axis (w, h, k) by 3 d space coordinate system
Enclose, obtain actual position information of the mobile object under secondary coordinate system;
The accumulation of passage time section, coordinate data (x_phy, y_phy, z_phy) can form tired at t (0)-t (n) moment
Product, and then run fitting function according to these coordinate datas and fit rail of this mobile object in t (0)-in t (n) periods
Mark.
The beneficial effects of the invention are as follows:The present invention can realize the detection of movement locus of object, have easy to use, application
The advantages of scope is wide.
Brief description of the drawings
Fig. 1 is the principle schematic of the embodiment of the invention.
Embodiment
The invention will be further described with reference to the accompanying drawings and examples:
Sensed as shown in Figure 1, moving-object trace approximating method of the present invention is applied in a kind of view-based access control model
In the billiard ball intelligent robot system of device, including client, server, cloud storage, robot system and man-machine interactive system;
The client sends instructions to the server, and the server sends instructions to the robot system, institute
State the real-time uploaded videos of robot system to cloud storage, the cloud storage and carry out data interaction with client;The machine
People's system is connected with the man-machine interactive system to carry out data interaction;
The server system:For message scheduling, the information exchange between client is arranged, is such as matched into administrative staff,
Place election, initiates game match, online to initiate the functions such as challenge, and arranges user into the movement under line on the spot, comprising but
It is not limited to match, training, coach, assiatant etc..
The robot system:Different balls pictorial information in the case of diverse location and rotational angle is gathered, in picture
The positional information of middle extraction ball, and stamps the label of corresponding ball, i.e. (x, y, with, height) and label, then inputs depth
The algorithm frame of study is trained, draw different balls training pattern ball.model (training pattern be data acquisition system one
A little forms), label and number mark including the training data of different balls, such as yellow1, red2 etc..
The depth information for the image acquisition that the robot system passes through camera, determines that it is deep in z-axis within the vision
Information is spent, three reconstructions carry out ball desktop according to camera image information (x, y-axis) and depth information (z-axis), that is, demarcate origin
Information (0,0,0) and x, y, the measurement range and positive direction of z-axis.
The training pattern of different balls is loaded into detection algorithm function fuction_ball by the robot system, is passed through
The algorithm of convolutional neural networks, carries out convolutional calculation to image information, the matching degree of full articulamentum is compared, when matching degree is higher than
During threshold value k, then it is assumed that the ball of position is matched with the ball in training pattern herein, so as to recognize the position of different balls in the picture
Information localtion (x, y, z) and its said tag label
The robot system is converted by built-in algorithm, and positional information of the different balls under image coordinate system is converted
For the positional information (x_phy, y_phy, z_phy) under real space coordinate system, place is divided into reference to 3 d space coordinate system
The coordinate length range of x, y, z three-dimensional shaft, i.e. origin (0,0,0) and three axis (w, h, k), obtains reality of the ball under secondary coordinate system
Positional information.
The accumulation of the robot system passage time section, coordinate data (x_phy, y_phy, z_phy) is in t (0)-t
(n) moment can form accumulation, and then run fitting function according to these coordinate datas and fit this ball in t (0)-t (n) times
Track in section.
The robot system vision system is switched by built-in algorithm records video camera, shoots video and then captures ball
Member's action, the position of ball, movement locus and ball speed, rotary speed data, three-dimensional reconstruction and rail for sportsman's action and ball movement
Mark is fitted.
Man-machine interactive system described in the present embodiment includes the touch-screen and loudspeaker that can be interacted, display screen;Loudspeaker is used
In prompting billiard ball event information, touch-screen is used to select to train, and related content of competing, the display screen allows user's Real Time Observation ratio
Match the data message that must grade.There was only artificial match judgement and training mode for original billiard ball, by based on machine vision
Tennis intelligent robot system complete coach, judge, the role such as assiatant, meets that game of billiards judge's law enforcement function and billiard ball are instructed
Practice project appraisal function, while client is provided, facilitate associated user after match and check competition data statistical after training
Analysis and level of training assay, complete the function such as coach and judge, save human cost, strengthen interactivity and recreational, carry
Rise user experience.
Preferred embodiment of the invention described in detail above.It should be appreciated that those of ordinary skill in the art without
Need creative work to conceive according to the present invention and make many modifications and variations.Therefore, all technologies in the art
Personnel are available by logical analysis, reasoning, or a limited experiment on the basis of existing technology under this invention's idea
Technical solution, all should be in the protection domain being defined in the patent claims.
Claims (1)
- A kind of 1. mobile object track determination methods of view-based access control model sensor, it is characterised in that:By the depth information of the image acquisition of camera, it is determined in z-axis depth information within the vision, according to camera image Information (x, y-axis) and depth information (z-axis) carry out three-dimensional modeling, using the center of objective plane as coordinate origin, length direction As x-axis, width is as y-axis, perpendicular to the direction of plane as z-axis;The training pattern of different mobile objects is loaded into detection algorithm function, by the algorithm of convolutional neural networks, to figure Picture information carries out convolutional calculation, the matching degree of full articulamentum is compared, when matching degree is higher than threshold value k, then it is assumed that position herein Mobile object matched with the mobile object in training pattern, so as to recognize the positional information of different mobile objects in the picture With its said tag;Converted by built-in algorithm, positional information of the different mobile objects under image coordinate system is scaled Wherein pixel is the pixel value of coordinate points, and f is focal length, and dis is the distance value measured by depth camera, and phy is space coordinates Under positional value;Positional information (x_phy, y_phy, z_phy) under real space coordinate system, will with reference to 3 d space coordinate system Place is divided into the coordinate length range of x, y, z three-dimensional shafts, i.e. origin (0,0,0) and three axis (w, h, k), obtains mobile object secondary Actual position information under coordinate system;The accumulation of passage time section, coordinate data (x_phy, y_phy, z_phy) can form accumulation at t (0)-t (n) moment, into And run fitting function according to these coordinate datas and fit this mobile object in the track of t (0)-in t (n) periods.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710991072.0A CN107945206A (en) | 2017-10-23 | 2017-10-23 | A kind of mobile object track determination methods of view-based access control model sensor |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710991072.0A CN107945206A (en) | 2017-10-23 | 2017-10-23 | A kind of mobile object track determination methods of view-based access control model sensor |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107945206A true CN107945206A (en) | 2018-04-20 |
Family
ID=61935546
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710991072.0A Pending CN107945206A (en) | 2017-10-23 | 2017-10-23 | A kind of mobile object track determination methods of view-based access control model sensor |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107945206A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109520415A (en) * | 2018-09-18 | 2019-03-26 | 武汉移动互联工业技术研究院有限公司 | The method and system of six degree of freedom sensing are realized by camera |
CN110728649A (en) * | 2018-06-28 | 2020-01-24 | 北京京东尚科信息技术有限公司 | Method and apparatus for generating location information |
CN111157998A (en) * | 2020-01-02 | 2020-05-15 | 自然资源部第三海洋研究所 | Dolphin underwater positioning system and method |
-
2017
- 2017-10-23 CN CN201710991072.0A patent/CN107945206A/en active Pending
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110728649A (en) * | 2018-06-28 | 2020-01-24 | 北京京东尚科信息技术有限公司 | Method and apparatus for generating location information |
CN109520415A (en) * | 2018-09-18 | 2019-03-26 | 武汉移动互联工业技术研究院有限公司 | The method and system of six degree of freedom sensing are realized by camera |
CN111157998A (en) * | 2020-01-02 | 2020-05-15 | 自然资源部第三海洋研究所 | Dolphin underwater positioning system and method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11941915B2 (en) | Golf game video analytic system | |
Li et al. | Application of artificial intelligence in basketball sport | |
CN106131469B (en) | Ball intelligent robot coach and judgment system based on machine vision | |
CN106843460B (en) | Multiple target position capture positioning system and method based on multi-cam | |
US20100151942A1 (en) | System and method for physically interactive board games | |
CN110739040A (en) | rehabilitation evaluation and training system for upper and lower limbs | |
CN102918489A (en) | Limiting avatar gesture display | |
CN102947777A (en) | User tracking feedback | |
US20230009354A1 (en) | Sporting sensor-based apparatus, system, method, and computer program product | |
CN107945206A (en) | A kind of mobile object track determination methods of view-based access control model sensor | |
JP2016182339A (en) | Method of operating screen baseball system | |
CN107886537A (en) | A kind of billiard ball intelligent robot system of view-based access control model sensor | |
US20240245995A1 (en) | Mixed reality golf competitions involving real and virtual participants | |
Yang et al. | Research on face recognition sports intelligence training platform based on artificial intelligence | |
Fang et al. | Technical and tactical command decision algorithm of football matches based on big data and neural network | |
Yu | Evaluation of training efficiency of table tennis players based on computer video processing technology | |
Echeverria et al. | Punch Anticipation in a Karate Combat with Computer Vision | |
Takano et al. | A multimedia tennis instruction system: Tracking and classifying swing motions | |
CN109126100A (en) | Ball controlling percentage calculation method, device, equipment and storage medium | |
CN115475373B (en) | Display method and device of motion data, storage medium and electronic device | |
Zhang et al. | The application of artificial intelligence technology in sports competition | |
Zhang | Construction of sports recognition system based on sports visual image technology under the background of information technology | |
Dabnichki | Computers in sport | |
CN104050346A (en) | Scheme for assisting in catching an object in a computer simulation | |
Yang | Smart Motion Capture and Scoring System for Taekwondo Training based on Camera Network Technology |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180420 |