CN112822460B - Billiard game video monitoring method and system - Google Patents
Billiard game video monitoring method and system Download PDFInfo
- Publication number
- CN112822460B CN112822460B CN202110139041.9A CN202110139041A CN112822460B CN 112822460 B CN112822460 B CN 112822460B CN 202110139041 A CN202110139041 A CN 202110139041A CN 112822460 B CN112822460 B CN 112822460B
- Authority
- CN
- China
- Prior art keywords
- video
- target
- target monitoring
- monitoring object
- information
- 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.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/181—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a plurality of remote sources
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63D—BOWLING GAMES, e.g. SKITTLES, BOCCE OR BOWLS; INSTALLATIONS THEREFOR; BAGATELLE OR SIMILAR GAMES; BILLIARDS
- A63D15/00—Billiards, e.g. carom billiards or pocket billiards; Billiard tables
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/23—Recognition of whole body movements, e.g. for sport training
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Social Psychology (AREA)
- Psychiatry (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a billiard game video monitoring method and system. The method comprises the following steps: acquiring monitoring videos of all directions in the whole billiard game, and preprocessing the monitoring videos to obtain target monitoring videos; locking a target monitoring object based on the target monitoring video; extracting action information of the target monitoring object in each direction in the whole billiard game; and judging whether cheating behaviors exist in the action information of the target monitoring object in all directions of the billiard game based on the set game rule. By acquiring the competition monitoring video, analyzing the action information of each position of the competition personnel in the video in the competition and judging whether the competition personnel cheat in the whole competition process according to the analysis result, the accuracy of judging whether the competition personnel cheat in the competition process is improved, and the fairness and justness of the competition are ensured.
Description
Technical Field
The invention relates to the technical field of video monitoring, in particular to a billiard game video monitoring method and system.
Background
At present, with the rapid development of national economy and information technology and the rapid development of network technology, the application of a monitoring system in various industries is gradually widened, and the monitoring system is not only applied in the industries of communication, traffic, safety and the like, but also used for monitoring the action information of competitors in the competition process so as to be displayed for the second time when disputes occur.
However, the traditional video monitoring method and system for the billiard competition only simply retain the action information of the competitors and cannot judge whether the competitors cheat according to the monitored action information of the competitors, so that the invention provides the video monitoring method and system for the billiard competition.
Disclosure of Invention
The invention provides a video monitoring method and a video monitoring system for a billiard game, which are used for analyzing the obtained action information of each position of a competitor in the game and judging whether cheating behaviors exist in the competitor in the whole game process according to the analysis result.
The invention provides a video monitoring method for a billiard game, which comprises the following steps:
step 1: acquiring a monitoring video of each direction in the whole billiard game process, and preprocessing the monitoring video to obtain a target monitoring video;
step 2: locking a target monitoring object based on the target monitoring video;
and step 3: extracting action information of each direction of the target monitoring object in the whole course of the billiard game;
and 4, step 4: and judging whether cheating behaviors exist in the action information of the target monitoring object in all directions of the billiard game based on the set game rule.
Preferably, in step 1, the method for monitoring the billiard game video includes the following steps:
the server sends a monitoring video acquisition request and a transmission channel establishment request to the competition video monitoring platform;
after receiving the request for establishing the transmission channel, the competition video monitoring platform establishes a transmission channel with the server and searches for a monitoring video from each position monitoring device based on the monitoring video acquisition request;
packaging the monitoring video to obtain a target packaged monitoring video;
and transmitting the target packaging monitoring video to a server based on a transmission channel to obtain the monitoring video of each direction in the whole billiard game.
Preferably, in step 1, the method for monitoring the video of the billiard game includes:
acquiring the monitoring video, wherein the monitoring video comprises a plurality of collected frames of billiard game images;
removing noise data in the billiard game image in each frame according to a preset noise removal rule to obtain a first preprocessed image in each frame;
acquiring a standard image of a preset frame number corresponding to the first preprocessed image in each frame according to the first preprocessed image in each frame;
superposing the standard image of the preset frame number to the first preprocessed image in each frame by taking the first preprocessed image in each frame as a reference to obtain a second preprocessed image in each frame;
and carrying out continuous processing on the second preprocessed image in each frame to obtain a target monitoring video.
Preferably, in step 2, the method for monitoring the video of the billiard game, based on the target monitoring video, locking the target monitoring object, includes:
acquiring the target monitoring video and determining a target object appearing in the target monitoring video;
determining the core degree of a target object appearing in each target monitoring video;
comparing the core degree with a preset threshold value;
and determining the target object with the core degree larger than or equal to the preset threshold value as the target monitoring object.
Preferably, in step 3, based on the target monitoring object, extracting the action information of the target monitoring object in each direction of the billiard game in the whole process includes:
acquiring a target monitoring video, and obtaining target information by passing the target monitoring video through a preset data conversion model;
the target information comprises structure type definition information and a serialization function signature of data corresponding to the target monitoring video;
generating a target serialization function through a preset rule according to structure type definition information and a serialization function signature of the data corresponding to the target surveillance video;
meanwhile, acquiring a target monitoring object in the target monitoring video, and determining the action attribute of the target monitoring object;
acquiring M video frames where the action attributes are located, and converting the M video frames into corresponding M text values;
based on the target serialization function, performing serialization processing on the M text values to obtain a target monitoring video sequence;
extracting a target area where the target monitoring object is located from each video frame in the target monitoring video sequence based on the target monitoring video sequence;
performing key point positioning on the target monitoring object in the target area to obtain target key points of the target monitoring object in each video frame and position information of the target key points in each video frame;
determining a dynamic characteristic value of the target monitoring object corresponding to the target key point in each video frame based on the position information of the target key point;
based on the dynamic characteristic value, selecting a video frame of which the dynamic characteristic value meets a preset threshold value from the target monitoring video sequence, and determining the video frame as a key video frame;
inputting the key video frame into a pre-trained two-dimensional coordinate extraction model to obtain two-dimensional image information of the target monitoring object in the key video frame;
obtaining three-dimensional image information of the target monitoring object in the key video frame based on the two-dimensional image information of the target monitoring object and a pre-trained three-dimensional coordinate extraction model;
and constructing a behavior acquisition model, inputting the three-dimensional image information of the target monitoring object into the behavior acquisition model, and obtaining the action information of the target monitoring object in each direction in the whole billiard game.
Preferably, in step 4, based on a set game rule, determining whether cheating behavior exists in the action information of the target monitoring object in each direction of the whole billiard game, includes:
acquiring competition videos of each position of the target monitoring object in the process of a billiard competition, and extracting position information of competition action information of the target monitoring object in the competition videos of each position;
determining the motion track of the target monitoring object according to the position information;
detecting skeleton key point information of the target monitoring object in each video frame in a video image sequence corresponding to the game video in each direction;
constructing a bone identification model, and inputting the bone key point information of the target monitoring object into the bone identification model to obtain the action category of the target monitoring object;
determining specific behavior information of the target monitoring object based on the motion track of the target monitoring object and the action category of the target monitoring object;
acquiring prohibited behavior information in the set competition rule;
comparing the specific behavior information of the target monitoring object with the behavior information forbidden in the set competition rule;
if the specific behavior information of the target monitoring object is consistent with the behavior information forbidden in the set competition rule, judging that the target monitoring object has a cheating behavior;
otherwise, judging that the target monitoring object has no cheating behavior.
Preferably, the billiard game video monitoring method for judging that the target monitoring object has cheating behavior includes:
obtaining a judgment result of the cheating behavior of the target monitoring object;
constructing a data coding model, and inputting the judgment result into the data coding model to obtain coded data corresponding to the judgment result;
transmitting the coded data corresponding to the judgment result to a judge terminal through a preset transmission channel;
the judging terminal grades the coded data corresponding to the judging result based on a preset cheating grading rule;
wherein the cheating ratings include minor cheating, moderate cheating, and severe cheating;
and processing the target monitoring object based on the grading result.
Preferably, in step 4, based on a set game rule, it is determined whether a cheating action exists in the action information of the target monitoring object in each direction of the billiard game in the whole course, and the method further includes:
acquiring action information of the target monitoring object in each direction in the whole billiard game, and eliminating irrelevant data in the action information;
calculating the rejection rate of irrelevant data in the action information, and calculating the accuracy rate of judging the action information of the target monitoring object according to the rejection rate, wherein the specific steps comprise:
calculating the rejection rate of irrelevant data in the action information according to the following formula:
wherein, α represents the rejection rate of the irrelevant data in the motion information, and the value range is (0.8,1); delta represents the amount of culled extraneous data; τ represents a total amount of irrelevant data included in the motion information; rho represents the effective data amount mistakenly removed in the removed irrelevant data;
representing a data culling factor; epsilon represents a misjudgment coefficient, and the value range is (0.25,0.56);
calculating the accuracy rate of judging the action information of the target monitoring object according to the following formula:
wherein beta represents the accuracy of the judgment of the target monitoring object motion information, the value range is (0.5,1), [ alpha ] represents the rejection rate of irrelevant data in the motion information, the value range is (0.8,1), [ mu ] represents the accuracy coefficient, the value range is (0.6,0.8), [ tau ] represents the total amount of the irrelevant data contained in the motion information, [ sigma ] represents the total amount of data corresponding to the target monitoring object motion information, [ epsilon ] represents the maximum error rate allowed to the judgment of the target monitoring object motion information, the value range is (0.12,0.25), [ omega ] represents the interference coefficient, and the value range is (0.2,0.4);
obtaining the rejection rate of irrelevant data in a plurality of groups of action information under ideal conditions and the judgment accuracy rate of the corresponding action information of the target monitoring object, and constructing a relation corresponding table for storage;
comparing the actual accuracy obtained by calculation with the corresponding ideal accuracy in the relation corresponding table;
if the actual accuracy is greater than or equal to the corresponding ideal accuracy in the relation corresponding table, judging that the judgment result of the target monitoring object action information is qualified;
otherwise, judging that the judgment result of the action information of the target monitoring object is unqualified, and removing irrelevant data in the action information of each position of the target monitoring object in the whole billiard game again until the actual accuracy is greater than or equal to the corresponding ideal accuracy in the relation corresponding table.
The invention provides a video monitoring system for billiard games, which comprises:
the video acquisition module is used for acquiring monitoring videos of all directions in the whole billiard game and preprocessing the monitoring videos to obtain target monitoring videos;
the object locking module is used for locking a target monitoring object based on the target monitoring video;
the action determining module is used for extracting action information of each direction of a target monitoring object in the whole billiard game based on the target monitoring object;
and the result judging module is used for judging whether cheating behaviors exist in the action information of the target monitoring object in all directions of the billiard game based on the set game rule.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
FIG. 1 is a flow chart of a video monitoring method for billiard games according to an embodiment of the present invention;
fig. 2 is a structural diagram of a billiards game video monitoring system in an embodiment of the invention.
Detailed Description
The preferred embodiments of the present invention will be described in conjunction with the accompanying drawings, and it should be understood that they are presented herein only to illustrate and explain the present invention and not to limit the present invention.
Example 1:
the embodiment provides a billiard game video monitoring method, as shown in fig. 1, including:
step 1: acquiring monitoring videos of all directions in the whole billiard game, and preprocessing the monitoring videos to obtain target monitoring videos;
and 2, step: locking a target monitoring object based on the target monitoring video;
and 3, step 3: extracting action information of the target monitoring object in each direction in the whole billiard game;
and 4, step 4: and judging whether cheating behaviors exist in the action information of the target monitoring object in all directions of the billiard game based on the set game rule.
In this embodiment, the preprocessing includes denoising the surveillance video.
In this embodiment, the target surveillance video refers to the obtained surveillance video that is processed and does not contain other videos unrelated to the game.
In this embodiment, the target monitoring object refers to a competitor in a billiards game.
In this embodiment, the action information refers to the action that the player performs when hitting billiards in the billiards game.
In this embodiment, the set game rules are set in advance, for example, by not using anything other than a club.
The beneficial effects of the above technical scheme are: by acquiring the competition monitoring video, analyzing the action information of each position of the competition personnel in the video in the competition and judging whether the competition personnel cheat in the whole competition process according to the analysis result, the accuracy of judging whether the competition personnel cheat in the competition process is improved, and the fairness and justness of the competition are ensured.
Example 2:
on the basis of the foregoing embodiment 1, this embodiment provides a video monitoring method for a billiard game, where in step 1, obtaining a monitoring video image of each azimuth in the whole course of the billiard game includes:
the server sends a monitoring video acquisition request and a transmission channel establishment request to the match video monitoring platform;
after receiving a request for establishing a transmission channel, the competition video monitoring platform establishes a transmission channel with the server and searches a monitoring video from each position monitoring device based on the monitoring video acquisition request;
packaging the monitoring video to obtain a target packaged monitoring video;
and transmitting the target packaging monitoring video to a server based on a transmission channel to obtain the monitoring video of each direction in the whole billiard game.
In this embodiment, the game video monitoring platform refers to a computer terminal connected with a camera.
In this embodiment, each position monitoring apparatus refers to a camera which is provided in advance at each position in the field of the game.
In this embodiment, the target encapsulation of the monitoring video refers to the packing of videos in various directions, which is convenient for data transmission.
The beneficial effects of the above technical scheme are: the competition images are obtained from the monitoring equipment in each position in the competition field, so that the action information of the competition personnel in each position can be analyzed, and whether cheating behaviors exist in the competition process of the competition personnel can be accurately judged.
Example 3:
on the basis of the foregoing embodiment 1, this embodiment provides a method for monitoring a video of a billiard game, where in step 1, the monitoring video is preprocessed, including:
acquiring the monitoring video, wherein the monitoring video comprises a plurality of collected billiard game images;
removing noise data in the billiard game image in each frame according to a preset noise removal rule to obtain a first preprocessed image in each frame;
acquiring a standard image of a preset frame number corresponding to the first pre-processed image in each frame according to the first pre-processed image in each frame;
superposing the standard image of the preset frame number to the first preprocessed image in each frame by taking the first preprocessed image in each frame as a reference to obtain a second preprocessed image in each frame;
and carrying out continuous processing on the second preprocessed image in each frame to obtain a target monitoring video.
In this embodiment, the fact that the surveillance video includes a plurality of frames of collected images of the billiard game means that the surveillance video is composed of a plurality of frames, and each frame includes a still image.
In this embodiment, the preset noise removal rule is preset to remove noise in the surveillance video, and the removal rule may be frame-by-frame removal.
In this embodiment, the first preprocessed image refers to an image obtained by denoising an image in each frame of the monitored video.
In this embodiment, the standard image with the preset frame number is set in advance, and the image does not contain other abnormal data, so that the action information of the participants can be clearly displayed.
In this embodiment, the serialization processing is to link the processed images in each frame to obtain the final surveillance video.
The beneficial effects of the above technical scheme are: the images corresponding to each frame in the monitoring video are processed and are overlapped with the standard images after the processing is finished, the final image corresponding to each frame is obtained, the multiple images are overlapped, the processed monitoring video is obtained, convenience is brought to analysis of action information of participants, and the action judgment accuracy is improved.
Example 4:
on the basis of the foregoing embodiment 1, this embodiment provides a video monitoring method for a billiard game, and in step 2, a target monitoring object is locked based on a target monitoring video, including:
acquiring the target monitoring video and determining a target object appearing in the target monitoring video;
determining the core degree of a target object appearing in each target monitoring video;
comparing the core degree with a preset threshold value;
and determining the target object with the core degree greater than or equal to the preset threshold value as the target monitoring object.
In this embodiment, the target object refers to a person who appears in the billiard game surveillance video.
In this embodiment, the core degree of the target object refers to the importance of the person appearing in the video to the game, for example, the core degree of two players, the core degree of the player who is hitting the ball is 5, and the core degree of the referee is 2.
In this embodiment, the predetermined threshold is set in advance to measure the importance of the competitors in the race.
The beneficial effects of the above technical scheme are: by determining the target object appearing in the game monitoring video and determining the target monitoring object according to the target object, the target monitoring object can be conveniently tracked and locked, each action chip of the target monitoring object can be obtained in time, and the accuracy of analyzing the target monitoring object is improved.
Example 5:
on the basis of the foregoing embodiment 1, this embodiment provides a video monitoring method for a billiard game, where in step 3, based on a target monitoring object, extracting motion information of the target monitoring object in each direction in the whole course of the billiard game includes:
acquiring a target monitoring video, and obtaining target information by passing the target monitoring video through a preset data conversion model;
the target information comprises structure type definition information and a serialization function signature of data corresponding to the target monitoring video;
generating a target serialization function through a preset rule according to structure type definition information and a serialization function signature of the data corresponding to the target surveillance video;
meanwhile, acquiring a target monitoring object in the target monitoring video, and determining the action attribute of the target monitoring object;
acquiring M video frames where the action attributes are located, and converting the M video frames into corresponding M text values;
based on the target serialization function, performing serialization processing on the M text values to obtain a target monitoring video sequence;
extracting a target area where the target monitoring object is located from each video frame in the target monitoring video sequence based on the target monitoring video sequence;
performing key point positioning on the target monitoring object in the target area to obtain target key points of the target monitoring object in each video frame and position information of the target key points in each video frame;
determining a dynamic characteristic value of the target monitoring object in each video frame corresponding to the target key point based on the position information of the target key point;
based on the dynamic characteristic value, selecting a video frame of which the dynamic characteristic value meets a preset threshold value from the target monitoring video sequence, and determining the video frame as a key video frame;
inputting the key video frame into a pre-trained two-dimensional coordinate extraction model to obtain two-dimensional image information of the target monitoring object in the key video frame;
obtaining three-dimensional image information of the target monitoring object in the key video frame based on the two-dimensional image information of the target monitoring object and a pre-trained three-dimensional coordinate extraction model;
and constructing a behavior acquisition model, inputting the three-dimensional image information of the target monitoring object into the behavior acquisition model, and obtaining the action information of the target monitoring object in each direction in the whole billiard game.
In this embodiment, the target information refers to converting data corresponding to the target surveillance video into a corresponding information form, so as to facilitate processing of the target surveillance video.
In this embodiment, the serialization function signature refers to identification information used to match the data type when the serialization process is performed on the data. A (c)
In this embodiment, the action attribute refers to the strength, type, and the like of an action made by the target monitoring object in the process of a match.
In this embodiment, the text value refers to converting each video frame in the target monitoring video into a text form, so as to facilitate processing of the action information of the target monitoring object.
In this embodiment, the target surveillance video sequence refers to a byte stream obtained by decomposing a target surveillance video, so as to extract motion information of a target surveillance object in the target surveillance video from the byte stream.
In this embodiment, the target area refers to the movement space area of the target monitoring object when the target monitoring object strikes the billiards at various orientations during the game.
In this embodiment, the target key point refers to a specific station point when the target monitoring object strikes a billiard ball in the target area.
In this embodiment, the position information of the target key point refers to the orientation of the target monitoring object when hitting the billiard ball, the distance from the reference point, and the like.
In this embodiment, the dynamic characteristic value refers to a key data segment in the hitting action information, which can represent the action of the target monitoring object when the target monitoring object hits the billiards ball.
In this embodiment, the key video frames refer to video segments of the target monitoring video when the target monitoring object strikes a billiard ball.
The beneficial effects of the above technical scheme are: the target monitoring video is serialized, so that the activity area of the target monitoring object can be determined in each serialized byte stream, the position information of the target monitoring object when the billiards is hit can be accurately locked, the action characteristic value of the target monitoring object can be extracted according to the position information, the action characteristic value of the target monitoring object can be accurately analyzed, the action information of the target monitoring object when the billiards is hit in each direction can be accurately obtained, the action information of the target monitoring object can be conveniently compared with the rules set by a game, whether the target monitoring object cheats or not can be accurately judged, and the accuracy of result analysis can be improved.
Example 6:
on the basis of the foregoing embodiment 1, this embodiment provides a video monitoring method for a billiard game, where in step 4, based on a set game rule, it is determined whether cheating behavior exists in the action information of the target monitoring object in each direction in the whole course of the billiard game, and the method includes:
acquiring competition videos of each position of the target monitoring object in the process of a billiard competition, and extracting position information of competition action information of the target monitoring object in the competition videos of each position;
determining the motion track of the target monitoring object according to the position information;
detecting skeleton key point information of the target monitoring object in each video frame in a video image sequence corresponding to the competition video of each direction according to the position information of the target monitoring object when the target monitoring object acts in the competition video of each direction;
constructing a skeleton identification model, and inputting skeleton key point information of the target monitoring object into the skeleton identification model to obtain the action category of the target monitoring object;
determining specific behavior information of the target monitoring object based on the motion track of the target monitoring object and the action category of the target monitoring object;
acquiring forbidden behavior information in a set competition rule;
comparing the specific behavior information of the target monitoring object with the behavior information forbidden in the set competition rule;
if the specific behavior information of the target monitoring object is consistent with the behavior information forbidden in the set competition rule, judging that the target monitoring object has cheating behaviors;
otherwise, judging that the target monitoring object has no cheating behavior.
In this embodiment, the movement locus of the target monitor object refers to the information of the movement of the target monitor object to and fro around the cue ball when hitting the billiard ball.
In this embodiment, the skeleton key point information refers to the movement information of the skeleton point of the arm when the target monitoring object strikes the billiard ball.
The beneficial effects of the above technical scheme are: the specific action information of the target monitoring object when the billiards is hit is determined, and the specific action information of the target monitoring object is compared with the action information forbidden in the preset competition rule, so that whether the target monitoring object has cheating behavior in the competition process is accurately judged, and the accuracy of result judgment is improved.
Example 7:
on the basis of the foregoing embodiment 6, this embodiment provides a video monitoring method for a billiard game, which determines that a cheating action exists in the target monitored object, and includes:
obtaining a judgment result of the cheating behavior of the target monitoring object;
constructing a data coding model, and inputting the judgment result into the data coding model to obtain coded data corresponding to the judgment result;
transmitting the coded data corresponding to the judgment result to a judge terminal through a preset transmission channel;
the judging terminal grades the coded data corresponding to the judging result based on a preset cheating grading rule;
wherein the cheating ratings include minor cheating, moderate cheating, and severe cheating;
and processing the target monitoring object based on a grading result.
In this embodiment, the processing of the target monitoring object includes a certain time length of a forbidden competition and a lifelong forbidden competition.
The beneficial effects of the above technical scheme are: by grading the judgment result of the cheating behavior of the target monitored object, the method is favorable for accurately judging the severity of the cheating behavior of the target monitored object and improving the fairness and fairness of the competition.
Example 8:
on the basis of the foregoing embodiment 1, this embodiment provides a video monitoring method for a billiard game, where in step 4, based on a set game rule, it is determined whether cheating behavior exists in the action information of the target monitoring object in each direction in the whole course of the billiard game, and the method further includes:
acquiring action information of the target monitoring object in each direction in the whole billiard game, and eliminating irrelevant data in the action information;
calculating the rejection rate of irrelevant data in the action information, and calculating the accuracy rate of judging the action information of the target monitoring object according to the rejection rate, wherein the specific steps comprise:
calculating the rejection rate of irrelevant data in the action information according to the following formula:
wherein, α represents the rejection rate of the irrelevant data in the motion information, and the value range is (0.8,1); δ represents the amount of culled extraneous data; τ represents irrelevant data contained in the motion informationTotal amount; rho represents the effective data amount mistakenly removed in the removed irrelevant data;
representing a data culling factor; epsilon represents a misjudgment coefficient, and the value range is (0.25,0.56);
calculating the accuracy rate of the target monitoring object action information judgment according to the following formula:
wherein beta represents the accuracy of the judgment of the target monitoring object motion information, the value range is (0.5,1), [ alpha ] represents the rejection rate of irrelevant data in the motion information, the value range is (0.8,1), [ mu ] represents the accuracy coefficient, the value range is (0.6,0.8), [ tau ] represents the total amount of the irrelevant data contained in the motion information, [ sigma ] represents the total amount of data corresponding to the target monitoring object motion information, [ epsilon ] represents the maximum error rate allowed to the judgment of the target monitoring object motion information, the value range is (0.12,0.25), [ omega ] represents the interference coefficient, and the value range is (0.2,0.4);
obtaining the rejection rate of irrelevant data in a plurality of groups of action information under ideal conditions and the judgment accuracy rate of the corresponding action information of the target monitoring object, and constructing a relation corresponding table for storage;
comparing the actual accuracy obtained by calculation with the corresponding ideal accuracy in the relation corresponding table;
if the actual accuracy is larger than or equal to the corresponding ideal accuracy in the relation corresponding table, judging that the judgment result of the target monitoring object action information is qualified;
otherwise, judging that the judgment result of the action information of the target monitoring object is unqualified, and removing irrelevant data in the action information of each position of the target monitoring object in the whole course of the billiard game again until the actual accuracy rate is greater than or equal to the corresponding ideal accuracy rate in the relation corresponding table.
In this embodiment, the unrelated data refers to the sports data corresponding to the action information, which has no influence on the game, such as the sports data of the player walking at the billiard table.
In this embodiment, the range of values of the data culling factor is (0.5,0.85).
The beneficial effects of the above technical scheme are: the method comprises the steps of obtaining action information of a target monitoring object in each direction of the whole billiard game, rejecting irrelevant data in the action information, calculating rejection rate of the irrelevant data in the action information, calculating accuracy rate of judgment of the action information of the target monitoring object according to the rejection rate, considering rejected irrelevant data quantity and mistakenly rejected effective data quantity in the rejected irrelevant data when calculating the rejection rate, and considering eight factors influencing the rejection rate to enable a calculation result to be accurate and reliable.
Example 9:
the present embodiment provides a video monitoring system for billiard games, as shown in fig. 2, including:
the video acquisition module is used for acquiring monitoring videos of all directions in the whole billiard game and preprocessing the monitoring videos to obtain target monitoring videos;
the object locking module is used for locking a target monitoring object based on the target monitoring video;
the action determining module is used for extracting action information of each direction of the target monitoring object in the whole billiard game based on the target monitoring object;
and the result judging module is used for judging whether cheating behaviors exist in the action information of the target monitoring object in all directions of the billiard game based on the set game rule.
The beneficial effects of the above technical scheme are: by acquiring the competition monitoring video, analyzing the action information of each position of the competition personnel in the video in the competition and judging whether the competition personnel cheat in the whole competition process according to the analysis result, the accuracy of judging whether the competition personnel cheat in the competition process is improved, and the fairness and justness of the competition are ensured.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (8)
1. A video monitoring method for billiard games is characterized by comprising the following steps:
step 1: acquiring a monitoring video of each direction in the whole billiard game process, and preprocessing the monitoring video to obtain a target monitoring video;
step 2: locking a target monitoring object based on the target monitoring video;
and step 3: extracting action information of the target monitoring object in each direction in the whole billiard game;
and 4, step 4: judging whether cheating behaviors exist in the action information of the target monitoring object in all directions of the billiard game on the basis of set game rules;
the method for extracting the action information of the target monitoring object in each direction of the whole billiard game comprises the following steps:
acquiring a target monitoring video, and obtaining target information by passing the target monitoring video through a preset data conversion model;
the target information comprises structure type definition information and a serialization function signature of data corresponding to the target monitoring video;
generating a target serialization function through a preset rule according to structure type definition information and a serialization function signature of the data corresponding to the target surveillance video;
meanwhile, acquiring a target monitoring object in the target monitoring video, and determining the action attribute of the target monitoring object;
acquiring M video frames where the action attributes are located, and converting the M video frames into corresponding M text values;
based on the target serialization function, performing serialization processing on the M text values to obtain a target monitoring video sequence;
extracting a target area where the target monitoring object is located from each video frame in the target monitoring video sequence based on the target monitoring video sequence;
performing key point positioning on the target monitoring object in the target area to obtain target key points of the target monitoring object in each video frame and position information of the target key points in each video frame;
determining a dynamic characteristic value of the target monitoring object in each video frame corresponding to the target key point based on the position information of the target key point;
based on the dynamic characteristic value, selecting a video frame of which the dynamic characteristic value meets a preset threshold value from the target monitoring video sequence, and determining the video frame as a key video frame;
inputting the key video frame into a pre-trained two-dimensional coordinate extraction model to obtain two-dimensional image information of the target monitoring object in the key video frame;
obtaining three-dimensional image information of the target monitoring object in the key video frame based on the two-dimensional image information of the target monitoring object and a pre-trained three-dimensional coordinate extraction model;
and constructing a behavior acquisition model, inputting the three-dimensional image information of the target monitoring object into the behavior acquisition model, and obtaining the action information of the target monitoring object in each direction in the whole billiard game.
2. The billiard game video monitoring method of claim 1, wherein in step 1, the obtaining of the monitoring video images of all directions of the billiard game in the whole process comprises:
the server sends a monitoring video acquisition request and a transmission channel establishment request to the match video monitoring platform;
after receiving a request for establishing a transmission channel, the competition video monitoring platform establishes a transmission channel with the server and searches a monitoring video from each position monitoring device based on the monitoring video acquisition request;
packaging the monitoring video to obtain a target packaged monitoring video;
and transmitting the target packaging monitoring video to a server based on a transmission channel to obtain the monitoring video of each direction in the whole billiard game.
3. A billiard game video surveillance method as claimed in claim 1, wherein in step 1, the pre-processing of the surveillance video includes:
acquiring the monitoring video, wherein the monitoring video comprises a plurality of collected billiard game images;
removing noise data in the billiard game image in each frame according to a preset noise removing rule to obtain a first preprocessed image in each frame;
acquiring a standard image of a preset frame number corresponding to the first preprocessed image in each frame according to the first preprocessed image in each frame;
superposing the standard image of the preset frame number to the first preprocessed image in each frame by taking the first preprocessed image in each frame as a reference to obtain a second preprocessed image in each frame;
and carrying out continuous processing on the second preprocessed image in each frame to obtain a target monitoring video.
4. The billiard game video monitoring method of claim 1, wherein in step 2, locking the target monitoring object based on the target monitoring video includes:
acquiring the target monitoring video and determining a target object appearing in the target monitoring video;
determining the core degree of a target object appearing in each target monitoring video;
comparing the core degree with a preset threshold value;
and determining the target object with the core degree greater than or equal to the preset threshold value as the target monitoring object.
5. The video monitoring method for billiard games as claimed in claim 1, wherein in step 4, based on the set rules of the game, determining whether the action information of the target monitoring object in each direction of the billiard game is cheating or not includes:
acquiring competition videos of each position of the target monitoring object in the process of a billiard competition, and extracting position information of competition actions of the target monitoring object in the competition videos of each position;
determining the motion track of the target monitoring object according to the position information;
detecting skeleton key point information of the target monitoring object in each video frame in a video image sequence corresponding to the competition video of each direction;
constructing a bone identification model, and inputting the bone key point information of the target monitoring object into the bone identification model to obtain the action category of the target monitoring object;
determining specific behavior information of the target monitoring object based on the motion track of the target monitoring object and the action category of the target monitoring object;
acquiring prohibited behavior information in the set competition rule;
comparing the specific behavior information of the target monitoring object with the behavior information forbidden in the set competition rule;
if the specific behavior information of the target monitoring object is consistent with the behavior information forbidden in the set competition rule, judging that the target monitoring object has cheating behaviors;
otherwise, judging that the target monitoring object has no cheating behavior.
6. The video monitoring method for billiard games as claimed in claim 5, wherein determining that cheating is occurring in said target monitored object comprises:
obtaining a judgment result of the cheating behavior of the target monitoring object;
constructing a data coding model, and inputting the judgment result into the data coding model to obtain coded data corresponding to the judgment result;
transmitting the coded data corresponding to the judgment result to a judge terminal through a preset transmission channel;
the judging terminal grades the coded data corresponding to the judging result based on a preset cheating grading rule;
wherein the cheating ratings include slight cheating, moderate cheating, and severe cheating;
and processing the target monitoring object based on the grading result.
7. A billiard game video monitoring method as recited in claim 1, wherein in step 4, based on the set game rules, it is determined whether there is cheating action in the action information of each orientation of the target monitoring object during the whole course of the billiard game, further comprising:
acquiring action information of the target monitoring object in each direction in the whole billiard game, and removing irrelevant data in the action information;
calculating the rejection rate of irrelevant data in the action information, and calculating the accuracy rate of judging the action information of the target monitoring object according to the rejection rate, wherein the specific steps comprise:
calculating the rejection rate of irrelevant data in the action information according to the following formula:
wherein, α represents the rejection rate of the irrelevant data in the motion information, and the value range is (0.8,1); δ represents the amount of culled extraneous data; τ represents a total amount of irrelevant data included in the operation information; rho represents the effective data amount mistakenly removed in the removed irrelevant data; θ represents a data culling factor; epsilon represents a misjudgment coefficient, and the value range is (0.25,0.56);
calculating the accuracy rate of the target monitoring object action information judgment according to the following formula:
wherein beta represents the accuracy of the judgment of the target monitoring object motion information, the value range is (0.5,1), [ alpha ] represents the rejection rate of irrelevant data in the motion information, the value range is (0.8,1), [ mu ] represents the accuracy coefficient, the value range is (0.6,0.8), [ tau ] represents the total amount of the irrelevant data contained in the motion information, [ sigma ] represents the total amount of data corresponding to the target monitoring object motion information, [ epsilon ] represents the maximum error rate allowed to the judgment of the target monitoring object motion information, the value range is (0.12,0.25), [ omega ] represents the interference coefficient, and the value range is (0.2,0.4);
obtaining the rejection rate of irrelevant data in a plurality of groups of action information under ideal conditions and the judgment accuracy rate of the corresponding action information of the target monitoring object, and constructing a relation corresponding table for storage;
comparing the actual accuracy obtained by calculation with the corresponding ideal accuracy in the relation corresponding table;
if the actual accuracy is greater than or equal to the corresponding ideal accuracy in the relation corresponding table, judging that the judgment result of the target monitoring object action information is qualified;
otherwise, judging that the judgment result of the action information of the target monitoring object is unqualified, and removing irrelevant data in the action information of each position of the target monitoring object in the whole billiard game again until the actual accuracy is greater than or equal to the corresponding ideal accuracy in the relation corresponding table.
8. A video monitoring system for a billiard game, comprising:
the video acquisition module is used for acquiring monitoring videos of all directions in the whole billiard game and preprocessing the monitoring videos to obtain target monitoring videos;
the object locking module is used for locking a target monitoring object based on the target monitoring video;
the action determining module is used for extracting action information of each direction of the target monitoring object in the whole billiard game based on the target monitoring object;
the result judging module is used for judging whether cheating behaviors exist in the action information of the target monitoring object in all directions of the billiard game based on the set game rule;
the method for extracting the action information of the target monitoring object in each direction of the billiard game in the whole process comprises the following steps:
acquiring a target monitoring video, and obtaining target information by passing the target monitoring video through a preset data conversion model;
the target information comprises structure type definition information and a serialization function signature of data corresponding to the target monitoring video;
generating a target serialization function through a preset rule according to structure type definition information and a serialization function signature of the data corresponding to the target surveillance video;
meanwhile, acquiring a target monitoring object in the target monitoring video, and determining the action attribute of the target monitoring object;
acquiring M video frames where the action attributes are located, and converting the M video frames into corresponding M text values;
based on the target serialization function, performing serialization processing on the M text values to obtain a target monitoring video sequence;
extracting a target area where the target monitoring object is located from each video frame in the target monitoring video sequence based on the target monitoring video sequence;
performing key point positioning on the target monitoring object in the target area to obtain target key points of the target monitoring object in each video frame and position information of the target key points in each video frame;
determining a dynamic characteristic value of the target monitoring object in each video frame corresponding to the target key point based on the position information of the target key point;
based on the dynamic characteristic value, selecting a video frame of which the dynamic characteristic value meets a preset threshold value from the target monitoring video sequence, and determining the video frame as a key video frame;
inputting the key video frame into a pre-trained two-dimensional coordinate extraction model to obtain two-dimensional image information of the target monitoring object in the key video frame;
obtaining three-dimensional image information of the target monitoring object in the key video frame based on the two-dimensional image information of the target monitoring object and a pre-trained three-dimensional coordinate extraction model;
and constructing a behavior acquisition model, inputting the three-dimensional image information of the target monitoring object into the behavior acquisition model, and obtaining the action information of the target monitoring object in each direction in the whole billiard game.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110139041.9A CN112822460B (en) | 2021-02-01 | 2021-02-01 | Billiard game video monitoring method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110139041.9A CN112822460B (en) | 2021-02-01 | 2021-02-01 | Billiard game video monitoring method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112822460A CN112822460A (en) | 2021-05-18 |
| CN112822460B true CN112822460B (en) | 2023-02-03 |
Family
ID=75861235
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110139041.9A Active CN112822460B (en) | 2021-02-01 | 2021-02-01 | Billiard game video monitoring method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112822460B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015135357A1 (en) * | 2014-03-10 | 2015-09-17 | 北京瑞盖科技有限公司 | Assisted-training system and method for billiards |
| JP2020031406A (en) * | 2018-08-24 | 2020-02-27 | 独立行政法人日本スポーツ振興センター | Determination system and determination method |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9159195B2 (en) * | 2005-07-14 | 2015-10-13 | Ag 18, Llc | Interactive gaming among a plurality of players systems and methods |
| EP2227299B1 (en) * | 2007-09-07 | 2012-07-18 | Edh Us Llc | Methods and processes for detecting a mark on a playing surface and for tracking an object |
| US9889367B2 (en) * | 2013-03-18 | 2018-02-13 | Zvi Minkovitch | Sports match refereeing system |
| CN108229294B (en) * | 2017-09-08 | 2021-02-09 | 北京市商汤科技开发有限公司 | Motion data acquisition method and device, electronic equipment and storage medium |
| CN116052280A (en) * | 2019-02-01 | 2023-05-02 | 沛勒尔维珍公司 | System and method for monitoring player performance and events in sports |
| CN112183153A (en) * | 2019-07-01 | 2021-01-05 | 中国移动通信集团浙江有限公司 | Object behavior detection method and device based on video analysis |
| CN210042036U (en) * | 2019-07-31 | 2020-02-07 | 南京铁鸽网络科技有限公司 | GPS intelligent foot ring device for preventing racing pigeon cheating in sports |
| CN111586356B (en) * | 2020-05-08 | 2021-07-20 | 国家邮政局邮政业安全中心 | Violation monitoring method, device and system, electronic equipment and storage medium |
| CN111589167B (en) * | 2020-05-14 | 2024-01-12 | 腾讯科技（深圳）有限公司 | Event sightseeing method, device, terminal, server and storage medium |
| CN111756746B (en) * | 2020-06-24 | 2022-03-25 | 国家计算机网络与信息安全管理中心 | Network attack and defense competition dynamic Flag anti-cheating application method |
| CN111860418A (en) * | 2020-07-29 | 2020-10-30 | 重庆道吧网络有限公司 | Intelligent video examination and consultation system, method, medium and terminal for athletic competition |
-
2021
- 2021-02-01 CN CN202110139041.9A patent/CN112822460B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015135357A1 (en) * | 2014-03-10 | 2015-09-17 | 北京瑞盖科技有限公司 | Assisted-training system and method for billiards |
| JP2020031406A (en) * | 2018-08-24 | 2020-02-27 | 独立行政法人日本スポーツ振興センター | Determination system and determination method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112822460A (en) | 2021-05-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lames et al. | On the search for reliable performance indicators in game sports | |
| CN111444890A (en) | Sports data analysis system and method based on machine learning | |
| CN114360209B (en) | Video behavior recognition security system based on deep learning | |
| CN113486757B (en) | Multi-person linear running test timing method based on human skeleton key point detection | |
| CN109740019A (en) | A kind of method, apparatus to label to short-sighted frequency and electronic equipment | |
| CN113440789B (en) | Intelligent counting method and system for multi-person rope skipping test | |
| CN111860418A (en) | Intelligent video examination and consultation system, method, medium and terminal for athletic competition | |
| CN115546688A (en) | Long jump evaluation method, device, electronic equipment and storage medium | |
| CN112822460B (en) | Billiard game video monitoring method and system | |
| CN113312840B (en) | Badminton playing method and system based on reinforcement learning | |
| CN110180158B (en) | Running state identification method and system and terminal equipment | |
| CN114264239B (en) | Motion platform laser calibration system | |
| CN115346161A (en) | Football game situation evaluation and prediction method, device and electronic equipment | |
| CN112734793B (en) | Intelligent billiard referee anti-cheating method and system | |
| CN115845349A (en) | General training method for ball game items for moving target detection based on deep learning technology and auxiliary referee system | |
| CN109045662A (en) | The shuttlecock service judge auxiliary system of view-based access control model | |
| CN115116087A (en) | Action assessment method, system, storage medium and electronic equipment | |
| CN113082680A (en) | Automatic judgment system for ball out of bound of ball game based on machine vision | |
| CN116416551A (en) | Video image multi-person self-adaptive rope skipping intelligent counting method based on tracking algorithm | |
| CN118097516B (en) | Auxiliary detection method and auxiliary detection system for table tennis ball service penalty based on video identification | |
| CN117333947B (en) | Badminton action analysis method and system | |
| CN118506453A (en) | Method for automatically judging long jump process of athlete | |
| CN112883862B (en) | Shooting identification method, shooting identification device, computer equipment and storage medium | |
| CN114241603B (en) | Shuttlecock action recognition and level grade evaluation method and system based on wearable equipment | |
| CN116384951A (en) | Property inspection anti-cheating method and device and electronic equipment |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |