CN106991358A - The algorithm that automatic identification football match based on panoramic video is scored - Google Patents
The algorithm that automatic identification football match based on panoramic video is scored Download PDFInfo
- Publication number
- CN106991358A CN106991358A CN201610037554.8A CN201610037554A CN106991358A CN 106991358 A CN106991358 A CN 106991358A CN 201610037554 A CN201610037554 A CN 201610037554A CN 106991358 A CN106991358 A CN 106991358A
- Authority
- CN
- China
- Prior art keywords
- football
- track
- goal
- scored
- region
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
- G06V20/42—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items of sport video content
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Closed-Circuit Television Systems (AREA)
Abstract
The invention discloses the algorithm that a kind of automatic identification football match based on panoramic video is scored, including the markup information of generation scene, the detection of candidate's football, the generation of football track, the detect and track of goalkeeper, the step of automatic judgement 5 scored.The algorithm that the automatic identification football match based on panoramic video that the present invention is provided is scored, analyzed by football race video, realize that automatic identification football match is scored by generating the markup information of scene, the detection of candidate's football, the generation of football track, the detect and track of goalkeeper, the step of automatic judgement 5 scored, overcome and scored the drawbacks of judging to bring using goal detection in race video and close-up shot detection in the prior art.
Description
Technical field
Aphorama is based on the present invention relates to a kind of algorithm of section of football match video identification, more particularly to one kind
The algorithm that the automatic identification football match of frequency is scored.
Background technology
Commonly the video of large-scale football race is analyzed at present, would generally be used in these videos
5 camera lenses illustrate event of once scoring, and this 5 camera lenses are the remote camera lens of shooting panorama, shooting sportsman
Close-up shot, crowd shots, the middle camera lens comprising some sportsmen and playback camera lens.Therefore, it is logical at present
Cross and carry out goal judgement using to carrying out goal detection and close-up shot detection etc. in race video.So
And when for carrying out ball match unsupervised recording using common camera, feature mirror is had no in video
Head, therefore above method often fails.In addition, the area of football is smaller under panning mode, this makes
Obtain the problem more difficult.For these defects of the prior art, the present invention proposes one kind and is based on
The algorithm that the automatic identification football match of panoramic video is scored.
The content of the invention
The present invention is to provide one kind based on panorama to solve the technical scheme that above-mentioned technical problem is used
The algorithm that the automatic identification football match of video is scored, wherein specifically including following steps：
1) markup information of scene is generated；The method learnt automatically using artificial mark or algorithm, is obtained
The prior information of scene is obtained, according to goal size and location, the shooting of match each side is generated and enters
The interest region of ball；
2) detection of candidate's football；
(1) from the interest region for taking out shooting and goal in panoramic video, background model is set up,
And therefrom it is partitioned into prospect；
(2) size is utilized, shape, the information such as color, to being filtered at prospect, obtains some times
The suspicious football region of choosing；
(3) result being tracked using early stage to football, predicts the position of present frame football, if
There is prospect on the position, be then added to the suspicious football region of candidate；
(4) to suspicious ball region, the prototype soccerballs come out using offline machine learning carry out football
Detection, regard the region for detecting football as candidate's football region；
3) generation of football track；
(1) using currently detected candidate's football as region to be tracked, the region is tracked (real
Border tracking go up+detect be combined obtain tracking result), by tracking result formed track.If
Do not track, with the replacement that predicts the outcome；If but continuous a period of time does not track, then terminates this
The tracking of track；
(2) time sequencing and spatial relationship are utilized, short track is connected as long track, is formed more complete
Whole football pursuit path；
4) detect and track of goalkeeper；
(1) to the prospect in the interest region shot and scored, carry out human testing and (be proposed with head
Shoulder pattern)；
(2) human body detected is matched and tracked, and using multiframe tracking result information,
Determine general sportsman and goalkeeper；
(3) goalkeeper is persistently tracked；
5) the automatic judgement scored；Football track is analyzed, when its y-coordinate is in the upper of goal
Lower boundary, and x coordinate between the right boundary at goal when, the judgement followed the steps below.
The algorithm that the above-mentioned automatic identification football match based on panoramic video is scored, wherein, generate field
In the markup information step of scape, prior information includes extracting goal from video；Interest region is y
Direction respectively extends toward goal bottom prescription to top-direction, and x directions are respectively expansion toward goal right boundary
Exhibition.
The algorithm that the above-mentioned automatic identification football match based on panoramic video is scored, wherein, goal
Automatic decision steps include：(1) judge the football track within former seconds before present frame, judge
Whether it intersects with goal line, intersecting intersecting with dotted line comprising solid line, if non-intersect, does not score,
Otherwise, labeled as suspicious shooting, and next step is turned；(2) motion state to track makes decisions,
If it was found that the mutation more than 90 degree occurs for the track direction of motion, or track movement velocity is successively decreased and approached
Zero, then it is assumed that suspicious to score；Otherwise, do not score.
The algorithm that the above-mentioned automatic identification football match based on panoramic video is scored, wherein, judging
Go out after suspicious goal, to being analyzed in subsequent 2 seconds the motion state of goalkeeper, if at this section
Interior goalkeeper's displacement is less than 1 meter, then it is assumed that score, otherwise it is assumed that not scoring.
The present invention has the advantages that relative to prior art：
Analyzed by football race video, by generating the markup information of scene, candidate's football
Detection, the generation of football track, the detect and track of goalkeeper, the step of automatic judgement 5 reality scored
Existing automatic identification football match is scored, overcome in the prior art using goal detection in race video and
The drawbacks of close-up shot detection goal judges to bring.
Brief description of the drawings
The algorithm that Fig. 1 scores for the automatic identification football match based on panoramic video that the present invention is provided
Schematic diagram.
Embodiment
The invention will be further described with reference to the accompanying drawings and examples.
The algorithm that the automatic identification football match based on panoramic video that the present invention is provided is scored, specific side
Case includes：
First, the markup information of scene is generated.
The method learnt automatically using artificial mark or algorithm, acquires some prior informations of scene
(goal is such as extracted from video).
According to goal size and location, the shooting of generation match each side and interest region (the y directions of goal
Certain proportion can be respectively extended toward goal bottom prescription to top-direction, x directions can be toward goal or so
Border is respectively extension certain proportion).
2nd, the detection of candidate's football.
(1) from the interest region for taking out shooting and goal in panoramic video, background model is set up,
And therefrom it is partitioned into prospect；
(2) size is utilized, shape, the information such as color, to being filtered at prospect, obtains some times
The suspicious football region of choosing；
(3) result being tracked using early stage to football, predicts the position of present frame football, if
There is prospect on the position, be then added to the suspicious football region of candidate；
(4) to suspicious ball region, the prototype soccerballs come out using offline machine learning carry out football
Detection, regard the region for detecting football as candidate's football region.
3rd, the generation of football track.
(1) using currently detected candidate's football as region to be tracked, the region is tracked (real
Border tracking go up+detect be combined obtain tracking result), by tracking result formed track.If
Do not track, with the replacement that predicts the outcome；If but continuous a period of time does not track, then terminates this
The tracking of track；
(2) time sequencing and spatial relationship are utilized, short track is connected as long track, is formed more complete
Whole football pursuit path.
4th, the detect and track of goalkeeper.
(1) to the prospect in the interest region shot and scored, carry out human testing and (be proposed with head
Shoulder pattern)；
(2) human body detected is matched and tracked, and using multiframe tracking result information,
Determine general sportsman and goalkeeper；
(3) goalkeeper is persistently tracked.
5th, the automatic judgement scored.
Football track is analyzed, when its y-coordinate is in the up-and-down boundary at goal, and x coordinate is in ball
When between the right boundary of door, the judgement followed the steps below：
(1) judge the football track within former seconds before present frame, judge its whether with goal bottom
Line is intersecting (intersecting intersecting with dotted line comprising solid line), if non-intersect, does not score, otherwise, mark
For suspicious shooting, and turn next step；
(2) motion state to track makes decisions, if finding, the track direction of motion occurs more than 90 degree
Mutation, or track movement velocity successively decreases and close to zero, then it is assumed that suspicious to score；Otherwise, do not score.
After suspicious goal is judged, in subsequent shorter a period of time (in 2 seconds), to goalkeeper
Motion state analyzed, if goalkeeper's displacement is smaller in this time (be less than 1 meter),
Then think to score, otherwise it is assumed that not scoring.
Although the present invention is disclosed as above with preferred embodiment, so it is not limited to the present invention, appoints
What those skilled in the art, without departing from the spirit and scope of the present invention, when a little modification can be made
With it is perfect, therefore protection scope of the present invention is when by being defined that claims are defined.
Claims (4)
1. the algorithm that a kind of automatic identification football match based on panoramic video is scored, it is characterised in that
Comprise the following steps：
1) markup information of scene is generated；The method learnt automatically using artificial mark or algorithm, is obtained
The prior information of scene is obtained, according to goal size and location, the shooting of match each side is generated and enters
The interest region of ball；
2) detection of candidate's football；
(1) from the interest region for taking out shooting and goal in panoramic video, background model is set up,
And therefrom it is partitioned into prospect；
(2) size is utilized, shape, the information such as color, to being filtered at prospect, obtains some times
The suspicious football region of choosing；
(3) result being tracked using early stage to football, predicts the position of present frame football, if
There is prospect on the position, be then added to the suspicious football region of candidate；
(4) to suspicious ball region, the prototype soccerballs come out using offline machine learning carry out football
Detection, regard the region for detecting football as candidate's football region；
3) generation of football track；
(1) using currently detected candidate's football as region to be tracked, the region is tracked, will
Result formation track in tracking, with the replacement that predicts the outcome if not tracking；If but continuous one section
Time does not track, then terminates the tracking of the track；
(2) time sequencing and spatial relationship are utilized, short track is connected as long track, is formed more complete
Whole football pursuit path；
4) detect and track of goalkeeper；
(1) to the prospect in the interest region shot and scored, carry out human testing and (be proposed with head
Shoulder pattern)；
(2) human body detected is matched and tracked, and using multiframe tracking result information,
Determine general sportsman and goalkeeper；
(3) goalkeeper is persistently tracked；
5) the automatic judgement scored；Football track is analyzed, when its y-coordinate is in the upper of goal
Lower boundary, and x coordinate between the right boundary at goal when, the judgement followed the steps below.
2. the calculation that the automatic identification football match based on panoramic video is scored as claimed in claim 1
Method, it is characterised in that in the markup information step of generation scene, prior information includes carrying from video
Take out goal；Interest region is that y directions respectively extend toward goal bottom prescription to top-direction, x side
It is respectively extension to yearn for goal right boundary.
3. the calculation that the automatic identification football match based on panoramic video is scored as claimed in claim 2
Method, it is characterised in that the automatic decision steps of goal include：(1) before judging before present frame
Football track in several seconds, judges whether it intersects with goal line, intersects comprising solid line and dotted line phase
Hand over, if non-intersect, do not score, otherwise, labeled as suspicious shooting, and turn next step；(2) it is right
The motion state of track makes decisions, if finding, the mutation more than 90 degree occurs for the track direction of motion,
Or track movement velocity is successively decreased and close to zero, then it is assumed that suspicious to score；Otherwise, do not score.
4. the calculation that the automatic identification football match based on panoramic video is scored as claimed in claim 3
Method, it is characterised in that after suspicious goal is judged, to the motion in subsequent 2 seconds to goalkeeper
State is analyzed, no if goalkeeper's displacement is less than 1 meter in this time, then it is assumed that score
Then, it is believed that do not score.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610037554.8A CN106991358A (en) | 2016-01-20 | 2016-01-20 | The algorithm that automatic identification football match based on panoramic video is scored |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610037554.8A CN106991358A (en) | 2016-01-20 | 2016-01-20 | The algorithm that automatic identification football match based on panoramic video is scored |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106991358A true CN106991358A (en) | 2017-07-28 |
Family
ID=59414264
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610037554.8A Pending CN106991358A (en) | 2016-01-20 | 2016-01-20 | The algorithm that automatic identification football match based on panoramic video is scored |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106991358A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107633241A (en) * | 2017-10-23 | 2018-01-26 | 三星电子（中国）研发中心 | A kind of method and apparatus of panoramic video automatic marking and tracking object |
| CN108882003A (en) * | 2018-07-25 | 2018-11-23 | 安徽新华学院 | A kind of electronic software control system that can detect excellent race automatically |
| CN109126135A (en) * | 2018-08-27 | 2019-01-04 | 广州要玩娱乐网络技术股份有限公司 | Virtual shooting method, computer storage medium and terminal |
| CN110543856A (en) * | 2019-09-05 | 2019-12-06 | 新华智云科技有限公司 | Football shooting time identification method and device, storage medium and computer equipment |
| CN110753267A (en) * | 2019-09-27 | 2020-02-04 | 珠海格力电器股份有限公司 | Display control method and device and display |
| CN111147889A (en) * | 2018-11-06 | 2020-05-12 | 阿里巴巴集团控股有限公司 | Multimedia resource playback method and device |
| CN111797812A (en) * | 2020-07-20 | 2020-10-20 | 深圳市云数链科技有限公司 | Method, system, terminal and medium for automatically recording effective goal in football match |
| CN113537168A (en) * | 2021-09-16 | 2021-10-22 | 中科人工智能创新技术研究院(青岛)有限公司 | Basketball goal detection method and system for rebroadcasting and court monitoring scene |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100026801A1 (en) * | 2008-08-01 | 2010-02-04 | Sony Corporation | Method and apparatus for generating an event log |
| CN101645137A (en) * | 2009-07-17 | 2010-02-10 | 中国科学院声学研究所 | Method for automatically detecting location of a football in long shot of football video |
| CN101794451A (en) * | 2010-03-12 | 2010-08-04 | 上海交通大学 | Tracing method based on motion track |
| WO2014066779A2 (en) * | 2012-10-25 | 2014-05-01 | Nike International Ltd. | Athletic performance monitoring systems and methods in a team sports environment |
| CN104881882A (en) * | 2015-04-17 | 2015-09-02 | 广西科技大学 | Moving target tracking and detection method |
-
2016
- 2016-01-20 CN CN201610037554.8A patent/CN106991358A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100026801A1 (en) * | 2008-08-01 | 2010-02-04 | Sony Corporation | Method and apparatus for generating an event log |
| CN101645137A (en) * | 2009-07-17 | 2010-02-10 | 中国科学院声学研究所 | Method for automatically detecting location of a football in long shot of football video |
| CN101794451A (en) * | 2010-03-12 | 2010-08-04 | 上海交通大学 | Tracing method based on motion track |
| WO2014066779A2 (en) * | 2012-10-25 | 2014-05-01 | Nike International Ltd. | Athletic performance monitoring systems and methods in a team sports environment |
| CN104881882A (en) * | 2015-04-17 | 2015-09-02 | 广西科技大学 | Moving target tracking and detection method |
Non-Patent Citations (1)
| Title |
|---|
| 张有为 等著: "《人机自然交互》", 30 September 2004, 北京：国防工业出版社 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107633241A (en) * | 2017-10-23 | 2018-01-26 | 三星电子（中国）研发中心 | A kind of method and apparatus of panoramic video automatic marking and tracking object |
| CN107633241B (en) * | 2017-10-23 | 2020-11-27 | 三星电子（中国）研发中心 | Method and device for automatically marking and tracking object in panoramic video |
| CN108882003A (en) * | 2018-07-25 | 2018-11-23 | 安徽新华学院 | A kind of electronic software control system that can detect excellent race automatically |
| CN109126135A (en) * | 2018-08-27 | 2019-01-04 | 广州要玩娱乐网络技术股份有限公司 | Virtual shooting method, computer storage medium and terminal |
| CN111147889A (en) * | 2018-11-06 | 2020-05-12 | 阿里巴巴集团控股有限公司 | Multimedia resource playback method and device |
| CN111147889B (en) * | 2018-11-06 | 2022-09-27 | 阿里巴巴集团控股有限公司 | Multimedia resource playback method and device |
| CN110543856A (en) * | 2019-09-05 | 2019-12-06 | 新华智云科技有限公司 | Football shooting time identification method and device, storage medium and computer equipment |
| CN110543856B (en) * | 2019-09-05 | 2022-04-22 | 新华智云科技有限公司 | Football shooting time identification method and device, storage medium and computer equipment |
| CN110753267A (en) * | 2019-09-27 | 2020-02-04 | 珠海格力电器股份有限公司 | Display control method and device and display |
| CN110753267B (en) * | 2019-09-27 | 2020-12-29 | 珠海格力电器股份有限公司 | Display control method and device and display |
| CN111797812A (en) * | 2020-07-20 | 2020-10-20 | 深圳市云数链科技有限公司 | Method, system, terminal and medium for automatically recording effective goal in football match |
| CN113537168A (en) * | 2021-09-16 | 2021-10-22 | 中科人工智能创新技术研究院(青岛)有限公司 | Basketball goal detection method and system for rebroadcasting and court monitoring scene |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106991358A (en) | The algorithm that automatic identification football match based on panoramic video is scored | |
| JP6625220B2 (en) | Method and system for detecting the action of an object in a scene | |
| CN106991357A (en) | The shooting of automatic identification Basketball Match and the algorithm scored based on panoramic video | |
| Sha et al. | End-to-end camera calibration for broadcast videos | |
| CN110929596A (en) | Shooting training system and method based on smart phone and artificial intelligence | |
| JP4956273B2 (en) | Throwing ball type discriminating device, discriminator generating device, throwing ball type discriminating program and discriminator generating program | |
| CN112183355B (en) | Effluent height detection system and method based on binocular vision and deep learning | |
| CN106991359B (en) | Algorithm for tracking basketball in ball game video in panoramic mode | |
| Pervaiz et al. | Artificial neural network for human object interaction system over Aerial images | |
| KR101428922B1 (en) | Golf swing classification apparatus and method using fuzzy system | |
| Teachabarikiti et al. | Players tracking and ball detection for an automatic tennis video annotation | |
| CN106991685B (en) | Algorithm for tracking football in ball game video in panoramic mode | |
| Conaire et al. | Tennissense: A platform for extracting semantic information from multi-camera tennis data | |
| Gao et al. | Markerless racket pose detection and stroke classification based on stereo vision for table tennis robots | |
| CN113362374A (en) | High-altitude parabolic detection method and system based on target tracking network | |
| Connaghan et al. | Recognition of tennis strokes using key postures | |
| Hayashi et al. | Head and upper body pose estimation in team sport videos | |
| JP2020107071A (en) | Object tracking device and program thereof | |
| Chotimanus et al. | Real swing extraction for video indexing in golf practice video | |
| Yang | Research on basketball sports neural network model based on nonlinear classification | |
| Li et al. | Automatic video-based analysis of athlete action | |
| Guo et al. | Adaptive Object Tracking Discriminate Model forMulti-Camera Panorama Surveillance in Airport Apron. | |
| Mao | Tracking a tennis ball using image processing techniques | |
| Feng et al. | Complex behavior recognition based on convolutional neural network: A survey | |
| Poliakov et al. | Physics based 3D ball tracking for tennis videos |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170728 |