KR20240037520A - Foul judgment system in badminton matches and method of operation of system - Google Patents

Abstract

A method of operating a foul decision system for a badminton game, comprising: receiving first use type information including user type information and foul decision type information; Obtaining a first image captured in real time of the progress of the first badminton game; Reading the first image based on a foul decision model that detects a violation of the rules of a badminton game through the image; and outputting decision information indicating a result of reading the first image based on the first use type information.

Description

Foul judgment system and method of operation of the badminton game {FOUL JUDGMENT SYSTEM IN BADMINTON MATCHES AND METHOD OF OPERATION OF SYSTEM}

This is about the foul decision system in badminton games.

The electronic system currently used in domestic professional sports badminton competitions (National Sports Festival, Divisional League, etc.) uses in-out decision video analysis technology.

Since the BWF (Badminton World Federation) also requires equipment and manpower, even in international competitions, it is only possible on one court and the rest must follow the referee's decision.

The Korea Badminton Association is implementing video review on all courts for the first time at the 2019 Miryang Wonwon Yonex Korea Junior Open International Badminton Championships. Even in major competitions such as the Olympics, an in-out video system (video review) is prepared, but the game is played with each line judge (lines judge) assigned.

Therefore, there is a need for a system that accurately determines fouls in real time in a badminton game. Additionally, there is a demand for a system that automatically judges fouls in real time even when there is no referee in a badminton game. Additionally, there is a need for a system that provides customized results according to the user based on referee decision information for badminton games.

Regarding the foul decision system for badminton games, we aim to provide customized decision information according to the type of user and the user's intended use.

The goal is to provide judgment information about badminton games in real time or provide judgment information for user learning.

According to one side, receiving first usage type information including user type information and foul decision type information; Obtaining a first image captured in real time of the progress of the first badminton game; Reading the first image based on a foul decision model that detects a violation of the rules of a badminton game through the image; and outputting decision information indicating a result of reading the first image based on the first use type information. A method of operating a foul determination system for a badminton game is provided.

According to another aspect, a communication unit that performs communication with an external device; output unit; processor; and a memory storing instructions executable by the processor, wherein the processor, by executing the instructions, receives first usage type information including user type information and foul decision type information, and first Acquire a first video captured in real time of the progress of the badminton game, read the first video based on a foul decision model that detects violations of the rules of the badminton game through the video, and based on the first use type information Thus, a foul decision system for a badminton game is provided that outputs decision information indicating the result of reading the first image.

For the foul decision system in a badminton game, customized decision information can be provided depending on the type of user and the user's intended use.

Judgment information for a badminton game can be provided in real time, or judgment information for user learning can be provided.

The present invention may be readily understood by combination of the following detailed description with the accompanying drawings, where reference numerals refer to structural elements.
Figure 1 is a conceptual diagram for explaining the operation of a badminton game foul decision system that provides results of reading real-time images of a badminton game according to usage type information, according to an embodiment.
Figure 2 is a flowchart showing a method of operating a foul decision system in a badminton game, according to an embodiment.
Figure 3 is a flowchart showing a method of operating a foul decision system in a badminton game when user type information is a player type, according to an embodiment.
Figure 4 is a flowchart showing a method of operating a foul decision system in a badminton game when user type information is a referee type, according to an embodiment.
FIG. 5 is a flowchart illustrating a method of learning a learning model for detecting foul actions in a foul decision system for a badminton game, according to an embodiment.
FIG. 6 is a diagram illustrating information on a decision result output as a predetermined image is input to a learning model, according to an embodiment.
Figure 7 is a block diagram showing the configuration of a foul decision system in a badminton game, according to one embodiment.

Hereinafter, various embodiments will be described in detail with reference to the drawings. The embodiments described below may be modified and implemented in various different forms. In order to more clearly explain the characteristics of the embodiments, detailed descriptions of matters widely known to those skilled in the art to which the following embodiments belong will be omitted.

Meanwhile, in this specification, when a configuration “includes” another configuration, this means that other configurations may be further included rather than excluding other configurations, unless specifically stated to the contrary.

Additionally, terms including ordinal numbers such as 'first' or 'second' used in this specification may be used to describe various components, but the components should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another.

Figure 1 is a conceptual diagram for explaining the operation of a badminton game foul decision system 10 that provides a result of reading real-time video of a badminton game according to usage type information, according to an embodiment.

Referring to FIG. 1, the foul determination system 10 may receive usage type information 110. The usage type information 110 may include user type information 111 and foul decision type information 112. Additionally, the foul determination system 10 can acquire real-time video 120 of a badminton game.

The foul decision model 130 in the foul decision system 10 can detect foul decision information 140 including whether a rule is violated and rule violation information within the real-time first video 121 of a badminton game. The foul determination system 10 can output decision information about real-time images based on usage type information. The specific operation of the foul determination system 10 is explained in FIGS. 2 to 6.

FIG. 2 is a flowchart showing a method of operating the foul determination system 10 for a badminton game, according to an embodiment.

Referring to FIG. 2, in step S210, the foul determination system 10 may receive first use type information including user type information and foul decision type information.

For example, user type information may include player type and referee type. For example, the type information of the foul decision may include a real-time type and a learning type.

The foul determination system 10 may perform operations according to usage type information. For example, if the usage type information is a player type and a real-time type, the foul decision system 10 may operate to provide decision information representing the result of a reading that detects a rule violation in real time as the game between players progresses. You can.

For another example, if the usage type information is a player type and a learning type, the foul decision system 10 may operate to provide regulation information of a badminton game and feedback information corresponding to a foul action through a game between players. .

For another example, if the usage type information is a referee type and a real-time type, the foul decision system 10 may operate to provide decision information according to the referee's request.

For another example, if the use type information is a referee type and a learning type, the foul decision system 10 may operate to provide evaluation information about the referee decision information in which the referee judged the game.

In step S220, the foul determination system 10 may acquire a first image captured in real time of the progress of the first badminton game.

For example, as the start of the first badminton game is detected, the foul decision system 10 may acquire a first image in which the movements of the players participating in the first badminton game are captured in real time.

For example, the foul decision system 10 can obtain images of the server's actions among the players. Additionally, the foul determination system 10 can acquire images of the movements of receivers among players. Additionally, the foul determination system 10 can acquire an image of the shuttlecock's movement.

For example, a filming device may be installed around a badminton court to film the players' movements. For example, the photographing device may be a camera. As the first badminton match begins, the imaging device may capture the movements of players playing within the badminton court. Additionally, the imaging device can photograph an area within a certain range based on the badminton court. For example, the certain range may be a certain portion outside the outer line of the badminton court. Additionally, the certain range may be a certain height from the ground. Therefore, the imaging device can capture an area wider than a badminton court in three dimensions in real time. The photographing device may be a plurality of cameras, and the plurality of cameras may photograph a designated area. The image captured by the imaging device may be transmitted to the foul determination system 10.

For example, the foul determination system 10 may acquire at least one image received from a photographing device. Additionally, the foul determination system 10 may synthesize each of the images captured from a plurality of cameras. Additionally, the foul determination system 10 may extract an image corresponding to a predetermined section from an image captured by a predetermined camera.

In step S230, the foul determination system 10 may read the first image based on a foul determination model that detects a violation of the rules of a badminton game through the image.

For example, the foul decision model may be a learning model that detects foul actions indicating at least one of a fault and a let using real-time video of a certain game. The foul decision model may be a learning model created based on the results of learning the rules of the badminton game and the reference images corresponding to the rules of the badminton game. The foul decision model can detect actions that violate the rules of the badminton game in the video of the badminton game, based on the results of learning the rules of the badminton game and the reference video corresponding to the rules of the badminton game.

For example, faults can be classified as service faults and in-play faults. A fault can mean a foul. If a fault occurs, a point may be lost. A service fault may be a fault that occurs while the server is performing server operation.

For example, a service fault occurs when a service fails and the shuttlecock falls, when the server serves the service line brightly, when the service operation is performed twice rather than at once, when service is performed from a height of 1.15 m or higher, and when the shuttlecock is served. If the service is served with the entire hand held so that it cannot be seen, if the server does not keep both feet on the floor when serving, if the server's racket or body touches the net immediately after the service, or if the opposing receiver causes the shuttlecock to cross the net. This can be at least one of the following: if you hit it beforehand or your foot crosses the opponent's court, if the shuttle gets stuck on the net and stops, if the shuttle goes over the net and gets caught in the net, or if the receiver's partner hits the shuttle.

For example, a fault during a game is when the shuttlecock lands outside the court line, when the shuttlecock passes under or through the net, when the shuttlecock touches the ceiling or wall, when the shuttlecock touches the player's body or clothing, or when the shuttlecock touches the player's body or clothing. If the shuttlecock touches an object or person outside the court, if a stroke is made while the shuttlecock is resting on the racket, if the shuttlecock is hit twice in succession without passing it to the opponent's court, if the shuttlecock is hit by the racket and falls without passing to the opponent's court. In this case, if the racket, player or clothing touches the net or equipment supporting the net, if the player's body or racket passes over the net and into the opponent's court, or if the racket or person passes under the net, interfering with or distracting the opposing player. This can be at least one of the following: interfering with an opposing player's legitimate stroke at the point where the shuttlecock crosses the net, or a player intentionally shouting or making unnecessary gestures to distract the opposing player.

For example, a let can mean a stoppage or nullity of play. If a let occurs, the server serves again without conceding a point.

For example, a let occurs when the server serves before the receiver is ready, when both the receiver and the server fault during service, or when the shuttlecock stops on the net after the service is returned or goes over the net and gets caught in the net. , This may be at least one of the following cases: the base and feather of the shuttle are completely separated during a rally, the game is interrupted by a coach, an opposing player is interrupted, or an unpredictable accidental situation occurs.

For example, the foul determination system 10 can read whether there is an action that violates the rules of a badminton game in the first image based on the foul determination model.

In step S240, the foul determination system 10 may output decision information indicating the result of reading the first image based on the first use type information.

For example, the foul determination system 10 may output decision information indicating the result of reading the first image based on the user's type information and the foul decision type information.

FIG. 3 is a flowchart showing a method of operating the foul determination system 10 in a badminton game when user type information is a player type, according to an embodiment.

Referring to FIG. 3, in step S310, the foul determination system 10 may check whether the user's type information is the player type. If the user's type information is not the player type, the foul determination system 10 may perform the operation according to step S235. The operation according to step S235 is explained in FIG. 4.

On the other hand, if the user's type information is a player type, the foul decision system 10 may check whether the foul decision type information is a real-time type in step S320. If the type information of the foul decision is a real-time type, the foul decision system 10 may provide decision information indicating the result of reading the first image in real time in step S330. If the usage type information is player type and real-time type, even if the referee does not directly referee the game, the foul decision system 10 analyzes the video of the game between players in real time, determines whether there is a rule violation, and provides information on rule violations. Judgment information including may be provided.

If the type information of the foul decision is not a real-time type, the foul decision system 10 may check whether the type information of the foul decision is a learning type in step S325. If the type information of the foul decision is a learning type, the foul decision system 10 may provide information about the foul action and feedback information according to step S335.

For example, as the first badminton game ends, the foul determination system 10 may display an image of a foul action detected as a rule violation in the first video and regulation information related to the foul action. Specifically, when the first player makes a foul action corresponding to the rule for a service fault, the foul decision system 10 extracts an image of the first player's foul action from the video of the first badminton game, Extracted images can be displayed. Additionally, the foul decision system 10 may display regulatory information related to the first player's foul actions.

For example, the feedback information may include at least one of information on the cause of the player's foul action and guide information for preventing the player's foul action. For example, the foul determination system 10 may display guide information to prevent actions corresponding to service faults.

If the usage type information is the player type and the learning type, the foul decision system 10 can provide the player with foul actions and regulation information corresponding to the foul actions that occurred during the badminton game, and learning information and foul actions related to the foul actions. Guide information to prevent the behavior can also be provided.

FIG. 4 is a flowchart showing a method of operating the foul decision system 10 in a badminton game when user type information is a referee type, according to an embodiment.

Referring to FIG. 4, in step S410, the foul decision system 10 may check whether the user's type information is a referee type. If the user's type information is a referee type, the foul decision system 10 may check whether the foul decision type information is a real-time type in step S420.

If the type information of the foul decision is a real-time type, the foul decision system 10 may provide decision information indicating the result of reading the first image in real time in step S430. For example, the foul decision system 10 may receive a decision request from the referee's electronic device. The foul determination system 10 may output decision information indicating the result of reading the first image in real time in response to a decision request.

During the first badminton game, a movement may occur that is difficult for the referee to determine whether it is a foul movement. In this case, the referee can request the foul decision system 10 to judge actions between players in real time through an electronic device. Additionally, if there is an objection to the player's decision result, the referee may request a decision on the image from the foul decision system 10 through an electronic device.

For example, the decision information may include an image of a foul action detected as a rule violation or regulatory information related to the foul action.

On the other hand, if the type information of the foul decision is not a real-time type, the foul decision system 10 may check whether the type information of the foul decision is a learning type in step S425. If the type information of the foul decision is a learning type, the foul decision system 10 may obtain referee decision information indicating the result of the referee's decision in the first badminton game in step S435.

In step S445, the foul decision system 10 may display evaluation information indicating an evaluation of the referee's decision based on the referee decision information. For example, the foul determination system 10 may obtain decision information based on a foul action detected as a rule violation in the first video while the first badminton game is in progress. The foul decision system 10 may display evaluation information indicating an evaluation of the referee's decision based on the result of comparing the referee decision information and decision information obtained from the foul decision model.

Meanwhile, the foul decision system 10 can collect information on misjudgment made by the referee. The foul decision system 10 may provide learning information for referees based on misjudgment information. For example, based on the misjudgment information, the foul decision system 10 provides learning information including at least one of the following information: the player's action that caused the referee to make a misjudgment, the rule that led to the referee's misjudgment, and the cause of the referee's misjudgment. can be created. The foul decision system 10 may display learning information or transmit the learning information to the referee's electronic device.

The foul judgment system 10 can provide the referee with evaluation information that evaluates whether the referee has performed the refereeing duties in compliance with the badminton game regulations, and provides information on misjudgments made by the referee and learning information for learning about the misjudgments. , it can ensure that referees can accurately perform refereeing duties in badminton games.

FIG. 5 is a flowchart illustrating a method of learning a learning model for detecting foul actions in the foul determination system 10 for a badminton game, according to an embodiment.

The foul determination system 10 can learn a learning model that detects foul actions based on the badminton game regulations and reference images corresponding to the badminton game regulations.

Specifically, referring to FIG. 5 , in step S510, the foul determination system 10 may obtain information on the regulations of the badminton game. For example, the foul decision system 10 may obtain and store text information corresponding to the rules of a badminton game.

In step S520, the foul determination system 10 determines a preset value based on the time when a foul action corresponding to the rules of the badminton game occurs, based on the information on the rules of the badminton game and the video corresponding to the rules of the badminton game. A reference image corresponding to the viewpoint section can be obtained.

For example, the foul determination system 10 may acquire images corresponding to the regulations of a badminton game. The foul determination system 10 can match and classify images corresponding to the rules and regulations of a badminton game. The foul determination system 10 may obtain a reference image by extracting a foul action corresponding to a regulation from the image. The reference image may be an image corresponding to the section before and after the foul action occurred. For example, the foul determination system 10 may acquire images from before a foul action occurs to a time after the foul action occurs as a reference image.

In step S530, the foul determination system 10 can learn information about the rules of a badminton game and a reference image.

For example, the foul decision system 10 can learn what the badminton game rules mean and learn the correlation between reference images corresponding to the badminton game rules. The foul decision system 10 can learn the player's actions that violate the regulations within the reference video.

For example, the learning model may be a data recognition model that detects foul motion by reading a plurality of images. The data recognition model may be a model based on an artificial neural network. For example, learning models such as Deep Neural Network (DNN), Recurrent Neural Network (RNN), and Bidirectional Recurrent Deep Neural Network (BRDNN) may be used as data recognition models, but are not limited thereto.

For example, when learning the first rule of the badminton game in the foul decision system 10, the foul decision system 10 may view the image containing the player's movement that violates the first rule of the badminton game and the badminton game. Videos containing player movements that do not violate the first rule can be studied together. Here, an image containing a player's motion that does not violate the first rule of a badminton game may include a motion similar to the player's motion that violates the first rule, although the player does not violate the first rule. Videos containing player actions that do not violate the first rule may be preset by the user and used when learning the first rule.

The foul decision system 10 learns together images containing player movements that violate the rules of a badminton game and images containing player movements that do not violate the regulations, thereby accurately determining foul movements for a given input image. can be detected.

FIG. 6 is a diagram illustrating information on a decision result output as a predetermined image is input to a learning model, according to an embodiment.

The foul determination system 10 can learn a learning model that detects foul actions in real-time video of a predetermined game and store the learned results. Additionally, the foul determination system 10 may periodically learn a plurality of images corresponding to the regulations of a badminton game and update the learning model based on the learning results. In addition, when the rules of a badminton game are newly established or changed, the foul decision system 10 learns a plurality of images corresponding to the new or changed rules and updates the learning model based on the learning results. .

Referring to FIG. 6, the foul determination system 10 can learn images that violate rules and regulations. For example, the foul decision system 10 may learn the first rule 611 and the image 621 including the player's actions that violate the first rule 611. For example, the first regulation 611 may be a regulation for double operation. The double rule stipulates that it is a fault for the server to make a feint move that confuses the receiver when serving, or to stop midway and change direction. For example, actions such as turning the racket sharply to change direction, moving both arms at the same time, or stopping the racket after it has been pulled back can be double actions.

For example, the foul decision system 10 may learn the second rule 612 and the image 622 that includes the player's actions that violate the second rule 612. For example, the second rule 612 may be a rule for overhand. The overhand rule indicates that it is a fault if the head of the racket is similar to or higher than the wrist at the moment the server serves.

For example, the foul decision system 10 may learn the third rule 613 and the image 623 including the player's actions that violate the third rule 613. For example, the third rule 623 may be a rule for base over. The rule regarding base over is a rule indicating that it is a fault if the lower part of the shuttlecock does not hit the ball at the moment the server puts the server in.

For example, the foul decision system 10 may learn the fourth rule 614 and the image 624 that includes the player's actions that violate the fourth rule 614. The fourth regulation 614 may be a regulation for over-waist. The rule regarding over-waist is a rule indicating that it is a fault if the RBI is hit above the server's navel at the moment of serving.

For example, the foul decision system 10 may learn the fifth rule 615 and the image 625 that includes the player's actions that violate the fifth rule 615. The fifth regulation 615 may be a regulation for a foot fault. The rule for a foot fault is that it is a fault if the server touches the line with one of his two feet or lifts or drags the other foot.

For example, the foul determination system 10 may acquire the first image 630 of the first badminton game in real time as the first badminton game starts. The foul determination system 10 may receive the first image 630 as an input value for the learning model 600. Based on the learning model 600, the foul determination system 10 can read whether there is a player's action that violates the rules of the badminton game in the first image 630.

For example, the foul determination system 10 may detect a foul action corresponding to a service fault in the first image 630 based on the learning model 600. The foul determination system 10 may obtain decision result information 640 including a rule 641 matching the foul action and a hand signal image 642 corresponding to the rule 641. The foul determination system 10 may output at least one of a rule 641 matching the foul action and a hand signal image 642 corresponding to the rule 641.

FIG. 7 is a block diagram showing the configuration of a foul determination system 10 for a badminton game, according to an embodiment.

Referring to FIG. 7, the foul determination system 10 may include a communication unit 610, an output unit 620, a memory 630, and a processor 640. The foul determination system 10 may further include other components in addition to those shown in FIG. 7 . Additionally, foul determination system 10 may include fewer components than those shown in FIG. 7 . The components of the foul determination system 10 are not limited to those shown in FIG. 7 .

The communication unit 710 can communicate with an external device. For example, the external device may be a photographing device that films a badminton game, or an electronic device that transmits predetermined information to the foul determination system 10 and receives decision result information from the foul determination system 10. The communication unit 710 may include at least one of a Wi-Fi chip, a Bluetooth chip, an NFC chip, and a wireless communication chip. The processor 740 can communicate with various external devices using the communication unit 710.

The output unit 720 may display decision information obtained from the foul decision system 10. For example, the output unit 720 may include a display panel or a speaker.

The memory 730 may store software or a program that detects foul actions in a badminton game. Additionally, the memory 730 may store instructions for how to operate the foul determination system 10, which detects foul actions in a badminton game using a learning-based foul determination model. In addition, the memory 730, based on the usage type information, generates decision information corresponding to the usage type information based on the result of reading the video of the badminton game, and provides instructions for how to operate the foul determination system 10. You can save them.

The memory 730 is a flash memory type, hard disk type, multimedia card micro type, card type memory (for example, SD or XD memory, etc.), RAM (Random Access Memory), SRAM (Static Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), PROM (Programmable Read-Only Memory), magnetic memory, magnetic It may include at least one type of storage medium among disks and optical disks.

The processor 740 controls the overall operation of the foul determination system 10 and may include at least one processor such as a CPU. The processor 740 may include at least one processor specialized for each function, or may be an integrated processor.

The processor 740 may execute a program stored in the memory 730, read data or files stored in the memory 730, or store new data or files in the memory 730. The processor 740 executes instructions stored in the memory 730 to obtain a result of reading an image of a badminton game, generate decision information corresponding to the reading result and use type information, and provide decision information. there is.

The processor 740 may obtain first use type information including user type information and foul decision type information.

For example, user type information may include player type and referee type. For example, the type information of the foul decision may include a real-time type and a learning type.

The processor 740 may perform operations according to usage type information. For example, if the usage type information is a player type and a real-time type, the processor 740 may operate to provide decision information indicating a reading result of detecting a rule violation in real time as the game between players progresses. .

For another example, if the usage type information is a player type and a learning type, the processor 740 may operate to provide regulation information of a badminton game and feedback information corresponding to foul actions through a game between players.

For another example, if the usage type information is a referee type and a real-time type, the processor 740 may operate to provide decision information according to the referee's request.

For another example, if the usage type information is a referee type and a learning type, the processor 740 may operate to provide evaluation information about the referee decision information in which the referee judged the game.

The processor 740 may acquire a first image captured in real time of the progress of the first badminton game.

For example, as the start of the first badminton game is detected, the processor 740 may acquire a first image in which the movements of the players participating in the first badminton game are captured in real time.

For example, the processor 740 may obtain an image of the operation of a server among players. Additionally, the processor 740 can obtain images of the actions of receivers among the players. Additionally, the processor 740 may obtain an image of the shuttlecock's movement.

For example, a filming device may be installed around a badminton court to film the players' movements. For example, the photographing device may be a camera. As the first badminton match begins, the photographing device may capture the movements of players playing within the badminton court. Additionally, the imaging device can photograph an area within a certain range based on the badminton court. For example, the certain range may be a certain portion outside the outer line of the badminton court. Additionally, the certain range may be a certain height from the ground. Therefore, the imaging device can capture an area wider than a badminton court in three dimensions in real time. The photographing device may be a plurality of cameras, and the plurality of cameras may photograph a designated area. The image captured by the imaging device may be transmitted to the foul determination system 10.

For example, the processor 740 may acquire at least one image received from a photographing device. Additionally, the processor 740 may synthesize images captured from a plurality of cameras. Additionally, the processor 740 may extract an image corresponding to a predetermined section from an image captured by a predetermined camera.

The processor 740 may read the first image based on a foul decision model that detects a violation of the rules of a badminton game through the image.

For example, the foul decision model may be a model that detects foul actions indicating at least one of a fault and a let using real-time video of a certain game. The foul decision model may be a model created based on the results of learning the badminton game regulations and reference images corresponding to the badminton game regulations. The foul decision model can detect actions that violate the rules of the badminton game in the video of the badminton game, based on the results of learning the rules of the badminton game and the reference video corresponding to the rules of the badminton game.

For example, the processor 740 may read whether there is an action that violates the rules of a badminton game in the first image, based on the foul decision model.

The processor 740 may output decision information indicating the result of reading the first image based on the first usage type information. For example, the processor 740 may output decision information indicating the result of reading the first image based on the user's type information and the foul decision type information.

If the usage type information is a player type and a real-time type, the processor 740 can analyze the video of the game between players in real time and provide judgment information including information on whether or not there is a rule violation and information about the rule violation.

If the usage type information is a player type and a learning type, the processor 740 may provide information about foul actions and feedback information.

For example, as the first badminton game ends, the processor 740 may display an image of a foul action detected as a rule violation in the first video and regulation information related to the foul action. Specifically, when the first player makes a foul action corresponding to the rule for a service fault, the foul decision system 10 extracts an image of the first player's foul action from the video of the first badminton game, Extracted images can be displayed. Additionally, the foul decision system 10 may display regulatory information related to the first player's foul actions.

For example, the feedback information may include at least one of information on the cause of the player's foul action and guide information for preventing the player's foul action. For example, the foul determination system 10 may display guide information to prevent actions corresponding to service faults.

If the usage type information is a player type and a learning type, the processor 740 can provide the player with foul actions and regulatory information corresponding to foul actions that occurred during a badminton game, and provides learning information and foul actions related to the foul actions. Guidance information for prevention can also be provided.

If the usage type information is a judgment type and a real-time type, the processor 740 may provide decision information indicating the result of reading the first image in real time. For example, the processor 740 may receive a decision request from the referee's electronic device. The processor 740 may output decision information indicating the result of reading the first image in real time in response to a decision request.

On the other hand, if the use type information is a referee type and a learning type, the processor 740 may obtain referee decision information indicating the result of the referee's decision in the first badminton game. The processor 740 may display evaluation information indicating an evaluation of the referee's decision based on the referee decision information. For example, the processor 740 may obtain decision information based on a foul action detected as a rule violation in the first video while the first badminton game is in progress. The output unit 720 may display evaluation information indicating an evaluation of the referee's decision based on the result of comparing the referee decision information and decision information obtained from the foul decision model.

Meanwhile, the processor 740 may collect information on misjudgment made by the referee. The processor 740 may generate learning information about the referee based on the misjudgment information, and the output unit 720 may output the learning information. For example, based on the misjudgment information, the processor 740 may generate learning information that includes at least one of the following information: the player's action that caused the referee to make a misjudgment, the rule that caused the referee to misjudgment, and the cause of the referee's misjudgment. You can. The output unit 720 may display learning information, or the communication unit 710 may transmit the learning information to the referee's electronic device.

The processor 740 may learn a learning model that detects foul motion based on the badminton game regulations and reference images corresponding to the badminton game regulations. For example, the video corresponding to the regulations of a badminton game includes at least one of a foul action corresponding to the regulations for a service fault, a foul action corresponding to the regulations for a fault during the game, and a foul action corresponding to the regulations for a let. It may include video showing foul actions.

The processor 740 may learn a learning model that detects foul actions in real-time video of a predetermined game and store the learned results. Additionally, the processor 740 may periodically learn a plurality of images corresponding to the rules of a badminton game and update the learning model based on the learning results. Additionally, when the rules of a badminton game are newly established or changed, the processor 740 may learn a plurality of images corresponding to the new or changed rules and update the learning model based on the learning results.

Anyone skilled in the art related to this embodiment will know that other general-purpose components may be included in addition to the components shown in FIG. 7.

As described above, although the embodiments have been described with limited examples and drawings, various modifications and variations can be made by those skilled in the art from the above description. For example, the described techniques are performed in a different order than the described method, and/or components of the described system, structure, device, circuit, etc. are combined or combined in a different form than the described method, or other components are used. Alternatively, appropriate results may be achieved even if substituted or substituted by an equivalent.

Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined by the claims and equivalents thereof as well as the claims described later.

Claims (10)

In the method of operating the foul decision system in a badminton game,
Receiving first usage type information including user type information and foul decision type information;
Obtaining a first image captured in real time of the progress of the first badminton game;
Reading the first image based on a foul decision model that detects a violation of the rules of a badminton game through the image; and
An operating method comprising outputting decision information indicating a result of reading the first image based on the first use type information. According to paragraph 1,
The user's type information includes player type and referee type,
The operating method wherein the type information of the foul decision includes a real-time type and a learning type. According to paragraph 2,
When the first usage type information is the player type and the real-time type,
The step of outputting the decision information is,
As the first badminton game progresses, an operating method comprising outputting the decision information indicating a result of reading the first image in real time. According to paragraph 2,
When the first usage type information is the player type and the learning type,
The step of outputting the decision information is,
As the first badminton game ends, displaying an image of a foul action detected as a rule violation in the first video and regulation information related to the foul action; and
An operating method comprising displaying feedback information based on a result of analyzing the foul motion. According to paragraph 2,
If the first use type information is the referee type and the real-time type,
The step of outputting the decision information is,
Receiving a decision request from the referee's electronic device; and
An operating method comprising outputting the decision information indicating a result of reading the first image in real time in response to the decision request. According to paragraph 2,
When the first use type information is the referee type and the learning type,
As the first badminton game ends, further comprising obtaining referee decision information indicating the result of the referee's decision in the first badminton game,
The step of outputting the decision information is,
Obtaining the decision information based on a foul action detected as a rule violation in the first video while the first badminton game is in progress; and
An operating method comprising displaying evaluation information indicating an evaluation of the referee's decision based on a result of comparing the referee decision information with the decision information. According to clause 6,
Collecting information on misjudgments made by the referee; and
An operating method further comprising providing learning information for the referee based on the misjudgment information. According to paragraph 1,
The foul decision model is,
An operation method, characterized in that it is generated by learning the rules of the badminton game and a reference image corresponding to the rules of the badminton game. According to clause 8,
The video corresponding to the rules of the badminton game above is,
An operation method comprising an image showing at least one foul action among a foul action corresponding to a rule for a service fault, a foul action corresponding to a rule for a fault during a game, and a foul action corresponding to a rule for a let. A communication unit that communicates with external devices;
output unit;
processor; and
Includes a memory that stores instructions executable by the processor,
The processor, by executing the instructions,
Receive first usage type information including user type information and foul decision type information,
Obtain the first video captured in real time of the progress of the first badminton game,
Reading the first video based on a foul decision model that detects a violation of the rules of a badminton game through the video,
A foul judgment system for a badminton game that outputs judgment information indicating a result of reading the first image, based on the first use type information.