System for determining playing situations in sports games

Abstract

The present invention relates to a method and system (10) for determining a game scenario (18) in a sports game, the system (10) comprising: a player module (11) that can be attached to a player (11′); an equipment module (12) that can be attached to sports equipment (12′) to be used in the sports game; and a computing unit (13). The computing unit (13) can be configured to receive player-based data (14) from the player module (12) and equipment-based data (16) from the equipment module (12), the player-based data (14) representing an activity profile (14c) of the player (11′), and the equipment-based data (16) representing an action profile (16c) of the equipment (12′). The computing unit (13) can also be configured to determine one of a plurality of game scenarios (18) involving the player (11′) and the equipment (12′) based on the activity profile (14c) of the player (11′) and the action profile (16c) of the equipment (16).

Description

Technical Field

[0001] The present invention relates to determining at least one of a plurality of game scenes during a sports game. Embodiments of the present invention relate to a system for determining a game scene, and in particular to a system for performing motion analysis using cooperative sensor modules for athletes of a sports game and for sports equipment used in the sports game. Background Art

[0002] Sports games, especially team sports games such as soccer, football, baseball, basketball, etc., are very popular and attract a lot of people nowadays. Not only trainers, but also spectators and commentators, and of course the athletes themselves, may be particularly interested in the statistical data collected during the competition or even during the training phase. A lot of information can be obtained through such statistical analysis.

[0003] For example, in a soccer match, there may be many different game events and game scenarios, such as a pass from player A to player B, possession of the ball by a team, a duel between two players, a shot, etc. To address the above examples, these game scenarios may be statistically evaluated to collect information about, for example, the number of successful or unsuccessful passes, ball possession, duels won or lost, number of shots, etc. These statistics may be constructed for individual players or for an entire team.

[0004] In the present disclosure, a game event may be an event that occurs during a competition and involves one entity, such as an individual player or a single ball. A game scene may be a scene related to the game, which occurs during the game and generally involves at least two entities, specifically players of the game and sports equipment used in the game.

[0005] In order to construct the above-mentioned match-related statistics, it is usually necessary to determine the events and / or scenes related to the match and count them. In the prior art, there may be many concepts for determining a single match event or match scene.

[0006] For example, US2010 / 283,630A1 proposes to equip football players and the football itself with sensors for measuring, for example, the speed, acceleration and impact force of each player and / or the ball. In addition, the sensors allow tracking of the players and the ball in the field of play. Tracking can be performed by a real-time positioning system (RTLS) or a real-time sensing system (RTSS). The system also uses a video system for capturing the competition, in which the sensor data can be superimposed on the video footage. This allows visual allocation of sensor data to the corresponding players wearing the corresponding sensors. For example, the actual speed of the players running on the field of play can be graphically displayed on the screen in real time. The system can allow for convenient determination of game events and can be highly attractive to users due to its real-time imaging capabilities. However, such systems are very complex and therefore very expensive. In addition, since two different technical solutions, namely video imaging and RF communication, are used together, the failure of one of the two devices will lead to the failure of the entire system.

[0007] WO2015 / 169915A1 proposes a simpler method. It describes a method and device for detecting a ball kick by simply attaching a sensor device to the leg or foot of a player. The sensor device may include a rotation sensor and a microphone. The rotation sensor detects the rotation of the player's foot, thereby estimating that the ball has been kicked. However, it is difficult to distinguish between a ball kick and, for example, stepping on the ground or accidentally kicking the grass. Therefore, it is described to detect the sound that occurs when kicking the ball by means of a microphone. However, in a scenario where two or more players are close to each other, such as in a duel, it is difficult to accurately detect which of the two or more players has kicked the ball most recently.

[0008] Another prior art system is known under the name RedFIR®. This system is based on ultra-high precision position detection of the player and the ball. The player and the ball are equipped with transmitters, by means of which the position of the player and the ball can be detected with a resolution in the centimeter range. This can even allow different position detection between the left and right legs of the player. However, such systems using high precision position detection tend to be expensive, so this may be interesting only for professional sports rather than private or amateur athletes.

[0009] Therefore, it is an object to provide a system that overcomes the disadvantages of the prior art. In particular, it is desirable to provide a system that reliably detects playing scenes during competitions and which is simultaneously low cost and low complexity. Summary of the invention

[0010] According to the present invention, this problem is solved by a system according to an embodiment of the present application and a method according to an embodiment of the present application.

[0011] A first aspect relates to a system for determining a game scenario in a sports game. The system may include: an athlete module, which can be attached to an athlete; an equipment module, which can be attached to sports equipment to be used in the sports game; and a computing unit. The computing unit is configured to receive athlete-based data from the athlete module and equipment-based data from the equipment module. The athlete-based data can represent an activity profile of the athlete, and the equipment-based data can represent an action profile of the equipment. According to the present invention, the computing unit is configured to determine one of a plurality of game scenarios involving the athlete and the equipment based on the activity profile of the athlete and based on the action profile of the equipment.

[0012] A second aspect relates to a method for determining a game scene in a sports game. The method comprises the following steps: receiving athlete-based data from an athlete module attached to the athlete, wherein the athlete-based data may represent an activity profile of the athlete. The method further comprises the following steps: receiving equipment-based data from an equipment module attached to the sports equipment to be used in the sports game, wherein the equipment-based data may represent an action profile of the equipment. According to the present invention, the method comprises the following further steps: determining one of a plurality of game scenes involving the athlete and the equipment based on the activity profile of the athlete and based on the action profile of the equipment.

[0013] According to a third aspect, computer programs are provided, wherein each computer program is configured to implement the above method when executed on a computer or a signal processor, so that the above method is implemented by one of the computer programs.

[0014] Therefore, the inventive concept may suggest placing a module in a sports equipment (e.g., a ball), and also providing a player module. Compared to the prior art, a game scene in which at least sports equipment and players are involved can be reliably detected. For example, a game scene "a pass from a first player to a second player" in a soccer game can be determined, because it can be reliably detected that the first player kicked the ball, the ball has been kicked out by the first player and the second player receives the ball. It can even be determined whether the pass made is a high pass or a low pass, whether the ball has reached the player of the same team in order to be classified as a successful pass, whether the ball has reached the player of the opposing team in order to be classified as an unsuccessful pass, and so on. According to the present invention, this can be done by determining the player's activity profile and the motion profile of the ball in real time and linking the player's activity profile with the motion profile of the ball (e.g., combining or merging). By doing so, a very reliable and robust association between the ball and the player currently manipulating the ball becomes possible, while a simple and therefore inexpensive module can be deployed. In some examples of the present invention, the player's activity profile and the motion profile of the ball can be determined using only an acceleration sensor or an inertial measurement unit (IMU). Combining this information, for example, the estimation of a pass from a first player to a second player will have much higher accuracy than using, for example, only the player module, and will be much cheaper than using precise positioning. Even if more players and balls are involved, such as in a training scenario, it will be easier to distinguish between successful and unsuccessful passes. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In the following, embodiments of the present invention will be described in more detail with reference to the accompanying drawings, in which:

[0016] Figure 1 shows a schematic overview of the system of the invention according to an embodiment,

[0017] Figure 2 A schematic overview diagram showing an example of a communication infrastructure of the system of the invention according to an embodiment,

[0018] Figure 3 shows a schematic overview of a system according to the invention according to another embodiment,

[0019] Figure 4 shows a schematic high-level depiction of a player module and an equipment module according to an embodiment,

[0020] Figure 5 shows a schematic high-level description of a computing unit according to an embodiment,

[0021] Figure 6 A schematic block diagram of determining an activity profile, an action profile, and a game scenario based on the activity profile and the action profile according to an embodiment is shown.

[0022] Figure 7 shows an accelerometer diagram representing acceleration values ​​obtained from two player modules attached to one player according to an embodiment,

[0023] Fig. 8A shows a rotation measurement graph representing rotation values ​​obtained from an equipment module attached to a ball during a high pass according to an embodiment,

[0024] Figure 8B shows an accelerometer diagram representing acceleration values ​​obtained from an equipment module attached to a ball during a high pass according to an embodiment,

[0025] Figure 8C shows a motion profile of a ball according to an embodiment, the motion profile being determined based on rotation measurement data and acceleration measurement data,

[0026] Fig.9A shows a rotation measurement graph representing rotation values ​​obtained from an equipment module attached to a ball during a low / ground pass, according to an embodiment,

[0027] Fig. 9B shows an accelerometer graph representing acceleration values ​​obtained from an equipment module attached to a ball during a low / ground pass, according to an embodiment,

[0028] Fig. 9C shows a motion profile of a ball according to an embodiment, the motion profile being determined based on rotation measurement data and acceleration measurement data,

[0029] Fig.10 A schematic overview diagram showing another example of a communication infrastructure of the system of the invention including optional infrastructure modules according to an embodiment,

[0030] Fig.11 shows an example of the content of a message according to an embodiment, wherein athlete-based data and equipment-based data may be transmitted from the athlete module / equipment module to the computing unit,

[0031] Fig.12 A schematic overview diagram showing an example of a sports field in which the system of the invention may be deployed according to an embodiment,

[0032] Fig.13 shows an isotropic case of RSSI path loss measurements between a ball module and two infrastructure modules attached to two goal posts of a soccer goal as a ball is moving through the infrastructure module into the goal, according to an embodiment,

[0033] Fig.14shows the sectorization of RSSI path loss measurements between a ball module and two infrastructure modules attached to two goal posts of a soccer goal as a ball is moving through the infrastructure modules into the goal, according to an embodiment,

[0034] Fig.15 A schematic overview of a communication network including one or more athlete modules, one or more equipment modules and optionally one or more infrastructure modules according to an embodiment is shown. DETAILED DESCRIPTION

[0035] In the following description, the same or equivalent elements or elements having the same or equivalent functions are denoted by the same or equivalent reference numerals.

[0036] To exemplarily illustrate the inventive concept, a soccer ball may be described as a non-limiting example of a sports game. Thus, a soccer player may be described as a non-limiting example of an athlete, and a ball may be described as a non-limiting example of equipment. Thus, a ball module may be described as a non-limiting example of an equipment module. Furthermore, so-called tags may be described as non-limiting examples of modules (i.e., a player module and / or an equipment module).

[0037] Instead of soccer, various other sports games can be imagined in which the inventive concept can be employed, such as tennis, football, baseball, basketball, cricket, etc. The inventive concept can preferably be used in team sports involving multiple players. Each player can preferably be equipped with at least one player module.

[0038] However, the inventive concept may also be used for a single athlete, for example, during training. According to the inventive concept, an equipment module may be attached to an equipment or each of a plurality of equipment that may be used during a particular sporting event. For example, in racket sports, an equipment module may be attached to a racket (tennis racket, baseball bat, golf club, etc.) and / or an equipment module may be attached to a ball, a puck, etc.

[0039] The system of the present invention may be used to determine the playing scenario of a respective game during competition and / or during training.

[0040] The system of the present invention can provide each of the data described herein, such as athlete-based data and / or equipment-based data and / or RSSI measurement data in real time (i.e., delayed no more than three seconds, preferably delayed only a few milliseconds). This can allow real-time visualization of data that can be processed by the system of the present invention. For example, a trainer can watch data and / or game scenes determined by the system of the present invention in real time while the athlete is playing the game (whether during a competition or during a training phase).

[0041] Figure 1 A system 10 for determining a play scenario in a sports game is shown according to an exemplary embodiment.

[0042] The system 10 may include a player module 11 attachable to a player 11 ′ and an equipment module 12 attachable to sports equipment 12 ′ to be used in the sports game ESD.

[0043] System 10 may also include a computing unit 13 for receiving player-based data 14 from player module 11 and equipment-based data 16 from equipment module 12. Player-based data 14 may represent an activity profile of player 11' and equipment-based data 16 may represent an action profile of equipment 12'.

[0044] The player module 11 may send the player-based data 14 to the computing unit 13 , and the equipment module 12 may send the equipment-based data 16 to the computing unit 13 .

[0045] The data 14 based on the athlete can include one or more data formats that can represent the activity profile of the athlete. For example, the data 14 based on the athlete can include the original measurement data 14a that can be provided by the sensor device 17, and / or one or more features 14b that can be extracted from the original measurement data 14a, and / or one or more physical athlete events that can be classified / estimated according to the extracted features 14b and / or the activity profile 14c itself of the athlete that can include one or more physical athlete events. Usually, the activity profile of the athlete 11 ' can be represented by the data 14 based on the athlete that can include different data formats.

[0046] The equipment-based data 16 may include one or more data formats that may represent the action profile of the equipment. For example, the equipment-based data 16 may include raw measurement data 16a that may be provided by the sensor device 19, and / or one or more features 16b that may be extracted from the raw measurement data 16a, and / or one or more physical equipment events that may be classified / estimated based on the extracted features 16b, and / or the action profile 16c itself that may include one or more physical equipment events. In general, the action profile of the equipment 12' may be represented by the equipment-based data 16 that may include different data formats.

[0047] Will refer to it later Figure 6 Further details are described regarding athlete-based data 14 representing an activity profile 14c of athlete 11 ' and equipment-based data 16 representing a motion profile 16c of equipment 12'.

[0048] The computing unit 13 may be configured to determine one of a plurality of play scenarios 18 involving the player 11 ′ and the equipment 12 based on the activity profile 14 c of the player 11 ′ and based on the motion profile 16 c of the equipment 12 ′.

[0049] As mentioned above, the player module 11 may include a sensor device 17. As the sensor device 17 may be attached to the player 11', preferably together with the player module 11, the sensor device 17 may also be referred to as a player sensor device (PSD).

[0050] Thus, according to an exemplary embodiment, the player module 11 may include a player sensor device 17. The player sensor device 17 may be configured to provide raw measurement data 14a relating to the player 11' to which the player module 11 is attached.

[0051] According to an exemplary embodiment, the athlete sensor device 17 may include at least one of the following items: a multi-axis accelerometer, an angular rate sensor, a magnetometer, and a pressure sensor, the multi-axis accelerometer being configured to provide raw measurement data representing the acceleration of the body part of the athlete 11' to which the athlete module 11 is attached, the angular rate sensor being configured to provide raw measurement data representing the rotation of the body part of the athlete 11' to which the athlete module 11 is attached, the magnetometer being configured to provide raw measurement data representing the current value of the earth's magnetic field, and the pressure sensor being configured to provide raw measurement data representing the pressure applied by or at the body part of the athlete 11' to which the athlete module 11 is attached.

[0052] The athlete module 11 may be configured to send raw measurement data 14a of the athlete sensor device 17 as athlete-based data 14 to the computing unit 13 , and the computing unit 13 may be configured to determine an activity profile 14c of the athlete 11 ′ from the received raw measurement data 14a .

[0053] Additionally or alternatively, the athlete module 11 itself may be configured to determine an activity profile 14c of the athlete 11 ′ from the raw measurement data 14a , and the athlete module 11 may be configured to send the activity profile 14c of the athlete 11 ′ as athlete-based data 14 to the computing unit 13 .

[0054] According to another exemplary embodiment, the athlete module 11 including the above-mentioned sensor device 17 can be configured to perform feature extraction of the raw measurement data 14a for estimating / classifying one or more physical athlete events that can be performed by the athlete 11 based on the extracted one or more features 14b. For example, one or more features 14b can be extracted from the raw measurement data 14a, and the features 14b can be classified, for example, by a feature or pattern classifier. According to this embodiment, the extracted features 14b can be classified into one or more physical athlete events that can be performed by the corresponding athlete 11' to which the athlete module 11 is attached. The physical athlete event can be, for example, kicking, running, sprinting, dribbling, etc.

[0055] The athlete module 11 can be configured to determine an activity profile 14c of the athlete 11' based on the extracted one or more features 14b and / or based on the determined one or more physical athlete events, and the athlete module 11 can also be configured to send the activity profile 14c as athlete-based data 14 to the computing unit 13. That is, the activity profile 14c can be determined in the athlete module 11.

[0056] Additionally or alternatively, the activity profile 14c may be determined in the computing unit 13. Thus, the athlete module 11 may, for example, be configured to send the extracted one or more features 14b and / or the determined one or more physical athlete events to the computing unit 13, which may then determine the activity profile 14c of the athlete 11 based on the received one or more features 14b and / or the received one or more physical athlete events.

[0057] As mentioned above, the equipment module 12 may also include a sensor device 19. Since the sensor device 19 may be attached to the equipment 12', preferably together with the equipment module 12, the sensor device 19 may also be referred to as an equipment sensor device (ESD).

[0058] Thus, according to an exemplary embodiment, the equipment module 12 may include an equipment sensor device 19 configured to provide raw measurement data 16a in conjunction with the equipment 12 ′ to which the equipment module 12 is attached.

[0059] According to an exemplary embodiment, the equipment sensor device 19 may include at least one of the following items: a multi-axis accelerometer, an angular rate sensor, a magnetometer and a pressure sensor, the multi-axis accelerometer being configured to provide raw measurement data 16a representing the acceleration of the equipment 12' to which the equipment module 12 is attached, the angular rate sensor being configured to provide raw measurement data 16a representing the rotation of the equipment 12' to which the equipment module 12 is attached, the magnetometer being configured to provide raw measurement data 16a representing the current value of the earth's magnetic field, and the pressure sensor being configured to provide raw measurement data 16a representing the pressure applied at the equipment 12' to which the equipment module 12 is attached.

[0060] The equipment module 12 may be configured to transmit raw measurement data 16a of the equipment sensor device 19 as equipment-based data 16 to the calculation unit 13 , and the calculation unit 13 may be configured to determine an action profile 16c of the equipment 12 ′ from the received raw measurement data 16a .

[0061] Additionally or alternatively, the equipment module 12 itself may be configured to determine an action profile 16 c of the equipment 12 ′ from the raw measurement data 16 a , and the equipment module 12 may be configured to send the action profile 16 c of the equipment 12 ′ as equipment-based data 16 to the calculation unit 13 .

[0062] According to another exemplary embodiment, the equipment module 12 including the above-mentioned equipment sensor device 19 can be configured to perform feature extraction of the raw measurement data 16a for classifying one or more physical events that can be performed by the equipment 12' or applied on the equipment 12' based on the extracted one or more features 16b. For example, one or more features 16b can be extracted from the raw measurement data 16a, and the features 16b can be classified, for example, by a feature or pattern classifier. According to this embodiment, the extracted features 16b can be classified into one or more physical equipment events, which can be performed by or applied to the corresponding equipment 12' to which the equipment module 12 is attached. The physical equipment event can be, for example, kicking, rolling on the ground, flying in the air, bouncing to the ground, etc.

[0063] The equipment module 12 may be configured to determine an action profile 16c of the equipment 12 based on the extracted one or more features 16b and / or based on the determined one or more physical equipment events, and the equipment module 12 may also be configured to send the determined action profile 16c as equipment-based data 16 to the computing unit 13. That is, the action profile 16c may be determined in the equipment module 12.

[0064] Additionally or alternatively, the action profile 16c may be determined in the computing unit 13. Thus, the equipment module 12 may be configured to send the extracted one or more features 16b and / or the determined one or more physical equipment events to the computing unit 13, which may then determine the action profile 16c of the equipment 12' based on the received one or more features 16b and / or the received one or more physical equipment events.

[0065] When the computing unit 13 has received athlete-based data 14 and equipment-based data 16 having at least one of the above-mentioned data formats, the computing unit 13 can determine one of multiple game scenarios 18 involving the athlete 11' and the equipment 12' based on the received athlete-based data 14 and based on the received equipment-based data 16.

[0066] For example, a soccer player 11' can run on a sports field and can kick a ball 12'. The ball 12' can fly in the air and bounce to the ground in response to the player's kick. The system 10 of the present invention can be configured to create an activity profile for the athlete 11', which includes one or more physical athlete events indicating that the athlete 11' ran and kicked the ball. In addition, the system 10 of the present invention can be configured to create an action profile for the ball 12', which includes one or more physical equipment events indicating that the ball is kicked out, flies in the air and bounces to the ground in response to the player's kick.

[0067] By fusing the activity profile of the player 11' with the motion profile of the ball 12', the system 10 of the present invention can determine a particular game scenario. In this example, the system 10 can determine that the particular player 11' made a high pass. For example, if the system 10 can include more than one player, a first game scenario indicating that a first player may have made a high pass to a particular second player can be determined, and a second game scenario indicating that the high pass may have been successfully or unsuccessfully received by the second player can be determined.

[0068] Generally speaking, a game scenario 18 may be a game-specific event in which at least one player 11' and at least one equipment 12' may be involved. The game scenario 18 may be determined by an activity profile 14c of at least one player 11' and an action profile 16c of at least one equipment 12'.

[0069] An activity profile 14c of the athlete 11 ′ may be determined in the athlete module 11 and / or the computing unit 13. An action profile 16c of the equipment 12 ′ may be determined in the equipment module 12 and / or the computing unit 13.

[0070] The activity profile 14c of a player 11' may include at least one activity that may be performed by the player 11'. The activity may also be referred to as a physical player event associated with the player 11', such as a "kick" performed by the player 11'.

[0071] The motion profile 16c of the equipment 12' may include at least one motion of the equipment 12'. The motion may also be referred to as a physical equipment event associated with the equipment 12'. For example, a sudden reduction in the spin of the ball 12' after it has been kicked may be interpreted as the ball 12' being "stopped" by the player.

[0072] By fusing the motion profile 16c of the equipment 12' with the activity profile 14c of the player 11', the computing unit 13 may determine a corresponding game scenario 18. In the above example, the computing unit 13 may determine a game scenario 18 in which a "high pass" is performed between the first player and the second player.

[0073] The system 10 of the present invention can further analyze any determined game scenario 18. For example, game-related statistics can be derived from the determined game scenario 18, such as how many high passes and / or low passes a particular player made during a competition or training session and optionally with what success rate, how many duels a particular player won or lost, etc. Further examples will be introduced in the following description.

[0074] Figure 2 A further overview of an exemplary information sharing architecture that may be deployed in the present inventive concept is shown.The system 10 may include a player tag as an example of a player module 11 attached to a player 11', and a ball tag as an example of an equipment module 12 attached to a ball 12'.

[0075] The player tag 11 may send the player-based data 14 to the computing unit 13 in one of the several ways and data formats described above. In this exemplary embodiment, the player-based data 14 may also be referred to as a player update message.

[0076] The ball tag 12 may send the equipment-based data 16 to the computing unit 13 in one of the several ways and data formats described above. In this exemplary embodiment, the equipment-based data 16 may also be referred to as an Equipment Update Message.

[0077] The computing unit 13 may include a graphical user interface (GUI) or may be connected to a graphical user interface (GUI). For example, the GUI may be implemented in a device 21, which may be a fixed or preferably handheld device, such as a smart phone, a tablet computer, a laptop, etc. The computing unit 13 may establish a one-way or two-way communication with the GUI device 21. For example, the GUI device 21 may display a game scene 18 that has been determined by the computing device 13. Additionally or alternatively, the GUI device 21 may display one or more of the update messages (e.g., based on the data 14 of the athlete or based on the data 16 of the equipment). The data 14 based on the athlete may be displayed as raw data 14a, and / or as one or more extracted features 14b, and / or as one or more classified physical athlete events, and / or as the activity profile 14c of the athlete 11'. The data 16 based on the equipment may be displayed as raw data 16a, and / or as one or more extracted features 16b, and / or as one or more classified physical equipment events, and / or as the action profile 16c of the equipment 12'.

[0078] The computing unit 13 may also be connected to a database 22. The database 22 may be an internal database implemented in the computing unit 13. Additionally or alternatively, the database 22 may be an external database to which the computing unit 13 may be connected via a corresponding interface. For example, the external database 22 may be connected via a local or remote network, in particular via the Internet.

[0079] The database 22 may, for example, include additional information for determining an activity profile 14c of a particular athlete 11' and / or a motion profile 16c of a sports equipment 12'. The database 22 may, for example, include additional information for classifying physical equipment / athlete events by features 14b, 16b extracted from measured raw sensor data 14a, 16a and other information (e.g., decision domains).

[0080] The additional information may include but is not limited to:

[0081] • Information about the athlete to which the athlete module is attached,

[0082] • Information about the sport equipment to which the equipment module is attached,

[0083] • Information about the player's position in the team formation,

[0084] • previous athlete-based data collected earlier during competition and / or during the training phase of the athletic competition,

[0085] • previous athlete-based data collected earlier during previous competitions and / or during previous training phases of athletic competition,

[0086] • previous equipment-based data collected earlier during competition and / or during the training phase of the athletic competition,

[0087] • previous equipment-based data collected earlier during previous competitions and / or during previous training sessions for athletic competitions,

[0088] • previous game scenarios identified earlier during competition and / or during the training phase of a sporting event,

[0089] The additional information from the database 22 can be used to generate higher level information. For example, the database 22 may include information about the playing position of the player 11 ' (e.g., goalkeeper, defense, offense, etc.), physical statistics, habits, history of events of the game or earlier events up to all classifications, or the development of the team.

[0090] Higher level information may be, for example, passes from a first player A to a second player B, unsuccessful / successful passes, possession time, first touch accuracy, unsuccessful / successful interceptions, duel wins / losses, and statistics and analysis based on such higher level information.

[0091] More generally, the database 22 may be configured to store additional information about at least one of: the athlete 11', the equipment 12' and a time history of a match, and the computing unit 13 may be configured to retrieve the additional information from the database 22 and determine one of a plurality of match scenarios 18 involving the athlete 11' and the equipment 12' based on the activity profile 14c of the athlete 11', the motion profile 16c of the equipment 12' and the additional information retrieved from the database 22.

[0092] like Figure 2 As further shown, the player tag 11 and the ball tag 12 can communicate with each other, as shown by arrow 23. Thus, the player tag 11 and the ball tag 12 can exchange data with each other, such as player-based data 14 and / or equipment-based data 16. Additionally or alternatively, the player tag 11 and the ball tag 12 can measure and transmit received signal strength RSSI values ​​between each other, which will be described in more detail later in this specification.

[0093] Figure 3 A schematic overview of an exemplary embodiment of a system 10 is shown, the system 10 including more than one athlete module (e.g., 1, 2, 3, ..., n athlete modules). In this example, the system 10 may include a first athlete tag 111 attached to a first athlete 11'1, a second athlete tag 112 attached to a second athlete 11'2, a second athlete tag 113 attached to an nth athlete 11' nThe nth athlete tag 11 n , and a ball tag 12 attached to the ball 12'. The system 10 may also include more than one equipment module 12, for example, 1, 2, 3, ..., n equipment modules.

[0094] The first athlete tag 111 may send data 141 based on the first athlete to the computing unit 13, the second athlete tag 112 may send data 142 based on the second athlete to the computing unit 13, and the nth athlete tag 113 may send data 141 based on the second athlete to the computing unit 13. n The data based on the nth athlete 14 n Send to the calculation unit 13.

[0095] Thus, according to an exemplary embodiment, the system 10 may include another (e.g., second) athlete module 112 attachable to another athlete 11'2, wherein the computing unit 13 may be configured to receive another athlete-based data 142 from the other athlete module 112, the other athlete-based data 142 representing another activity profile of the other athlete 11'2.

[0096] The computing unit 13 may be configured to determine one of a plurality of game scenarios 18 involving the player 11'1, the further player 11'2 and the equipment 12' based on an activity profile 14c of the first player 11'1, another activity profile of the further player 11'2 and a motion profile 16c of the equipment 12'.

[0097] For example, the ball 12' may be passed from the first player 11'1 to the second player 11'2. For example, the first player 11'1 may perform a low pass. The activity profile 14c of the first player 11'1 may indicate (e.g., via extracted features 14b or classified physical player events) that the first player 11'1 kicked the ball 12', the motion profile 16c of the ball 12' may indicate (e.g., via extracted features 16b or classified physical player events) that the ball 12' has been kicked and is rolling on the ground, and another activity profile of the second player 11'2 may indicate (e.g., via extracted features 142b or classified physical player events) that the second player 11'2 stopped the ball 12'.

[0098] The computing unit 13 may then be configured to fuse the activity profile 14c of the first player 11'1 with another activity profile of another player 11'2 and the action profile 16c of the equipment 12' to determine one of a plurality of game scenarios 18 involving the first player 11'1, the second player 11'2 and the equipment 12'. In this example, the determined game scenario 18 may be a successful low pass from the first player 111 to the second player 112.

[0099] According to another exemplary embodiment, the system 10 may be configured to synchronize the player-based data 14 in time with the equipment-based data 16. For example, the motion profile 12 of the equipment 12' may be synchronized in time with the activity profile 14 of the player 11'. For example, a physical player event (e.g., kicking a ball) may be associated in time with a physical equipment event (e.g., having been kicked).

[0100] For example, the athlete module 11 may be configured to generate the athlete-based data 14 during a predetermined time window, and the equipment module 12 may also be configured to generate the equipment-based data 16 during a predetermined time window. The computing unit 13 may receive the athlete-based data 14 and the equipment-based data 16, and the computing unit 13 may also be configured to synchronize the received athlete-based data 14 and the received equipment-based data 16 in time, so that the time windows of the athlete-based data 14 and the equipment-based data 16 at least partially overlap in the time domain.

[0101] For example, athlete module 11 can collect one or more athlete-based data 14 during the first time window between t1 and t2. The first time window can include a time length between a few milliseconds and a few seconds. A shorter time window can provide the higher accuracy of the real-time capability of system 10. For example, a time window within the millisecond range can provide a very good real-time experience to the user. The time length of the above-mentioned first time window can be between 0.05 seconds and 2.00 seconds, or between 0.25 seconds and 1.00 seconds, or between 0.40 seconds and 0.75 seconds. In some examples, the time length of the first time window can be about 0.50 seconds. During this first time window, athlete 11 'may have kicked a ball 12 ', which can be represented in activity profile 14c.

[0102] Equipment module 12 can collect one or more data 16 based on equipment during the second time window between t3 and t4, wherein t3 can be positioned between t1 and t2, so that the first time window and the second time window overlap at least in part in the time domain.The time length of the above-mentioned second time window can be between 0.05 second and 2.00 second, or between 0.25 second and 1.00 second, or between 0.40 second and 0.75 second.In some examples, the time length of the second time window can be about 0.50 second.Based on the data 14 of the athlete and based on the data 16 of equipment can be synchronized in time in the region overlapping at least in part.

[0103] Preferably, the time window of the equipment-based data 16 and the time window of the athlete-based data 14 may include the same window size and may include a consistent overlapping region in the time domain, for example, t1 = t3 and t2 = t4.

[0104] The athlete-based data 14 may include one or more athlete features 14b that may be extracted from raw sensor measurement data 14a of a sensor device 17 included by the athlete module 11. Thus, the one or more athlete features 14b may be extracted, for example, by using the time windowing scheme described above with general reference to the athlete-based data 14. For example, the athlete module 11 may extract one or more athlete features 14b from the raw measurement data 14a during a first time window, for example between t1 and t2.

[0105] The equipment-based data 16 may include one or more equipment features 16b that may be extracted from raw sensor measurement data 16a of a sensor device 19 included by the equipment module 12. Thus, the one or more equipment features 16b may be extracted, for example, by using the time windowing scheme described above with reference generally to the equipment-based data 16. For example, the equipment module 12 may extract one or more equipment features 16b from the raw measurement data 16a during a second time window, for example between t3 and t4.

[0106] By the above-mentioned time synchronization of the athlete-based data 14 and the equipment-based data 16, a temporal association between the action profile 16c of the equipment 12' and the activity profile 14c of the athlete 11' can be provided. For example, a physical athlete event can be associated with a physical equipment event, and the physical athlete event and the physical equipment event can occur roughly at the same time during a competition. Therefore, the possibility of correctly determining a game scene 18 involving the athlete 11' and the equipment 12' can be significantly improved.

[0107] If already referenced Figure 1 As described, each of the ball tag (i.e., equipment module 12) and the player tag (i.e., player module 11) may include at least one sensor device 17, 19. The respective sensor devices 17, 19 may include various sensors, such as accelerometers, magnetometers, gyroscopes / angular rate sensors, pressure sensors, which may be connected to the processing unit. In a particular embodiment, the player module 11 and / or the equipment module 12 may include sensor devices 17, 19, which sensor devices 17, 19 include at least a multi-axis accelerometer. The data retrieved from the respective sensor devices 17, 19 may optionally be stored in a buffer and may preferably be transmitted to the computing unit 13 via a radio interface.

[0108] exist Figure 4 A high-level functional description of the athlete module 11 is depicted in FIG. 1 , which may also be valid for the equipment module 12 accordingly.

[0109] The optional processing unit 41 inside the respective module 11, 12 may perform a signal analysis on the raw measurement signals 14a, 16a retrieved from the respective sensor device 17, 19 to extract relevant features 14b, 16b (see Figure 1 ). The features 14b, 16b extracted by the corresponding modules 11, 12 may include, but are not limited to, the rotation speed, impact strength and time, statistical torque, and action state of the corresponding modules 11, 12. Therefore, if the signal 14a originates from the athlete module 11, the activity profile 14c of the athlete 11' can be determined based on these extracted features 14b, or if the signal originates from the equipment module 12, the action profile 16c of the equipment 12' can be determined based on these extracted features 16b.

[0110] The player-based data 14 and the equipment-based data 16 can be sent to the computing unit 13 via the radio communication unit 42, respectively. The player-based data 14 and / or the equipment-based data 16 can be sent to the computing unit 13 as described above. Figure 1 At least one of the described data formats is sent to the computing unit 13. For example, the athlete tag 11 and / or the ball tag 12 can be configured to send the raw measurement data 14a, 16a as the data 14, 16 based on the athlete / equipment, and / or as the extracted features 14b, 16b as the data based on the athlete / equipment, and / or as the classified physical events as the data 14, 16 based on the athlete / equipment, and / or as the activity / action profile 14c, 16c itself as the data 14, 16 based on the athlete / equipment. Even if the raw measurement signal 14a, 16a can be sent optionally, this may also result in transmitting a higher data volume, for example, the three-axis acceleration data with a sampling rate of 300Hz. Therefore, it may be preferred to process the raw measurement data 14a, 16a already on the corresponding athlete / equipment module 11, 12.

[0111] Figure 5 A high level functional description of the computing unit 13 is shown. The computing unit 13 may receive athlete-based data 14 and / or equipment-based data 16 from the respective modules 11 , 12 via the radio communication unit 51 .

[0112] The received athlete-based data 14 and / or equipment-based data 16 may include a timestamp and a payload. The payload may include at least one of the following: a raw signal, extracted features, classified physical events, and action / activity profiles. In the synchronization stage 52, the received athlete-based data 14 and / or equipment-based data 16 may be synchronized in time between the different modules 11, 12. Figure 5In the example shown, physical events may be synchronized in synchronization level 52 as a non-limiting example of athlete-based data 14 and / or equipment-based data 16 .

[0113] The synchronized player-based data 14 and / or equipment-based data 16 may then be associated with each other in a data association stage 53, for example by using semantics or rules. Figure 5 In the example shown, the previously mentioned physical events may be associated with each other in an event association level 53 .

[0114] Figure 6 A schematic overview of an example of a possible data processing of the inventive system 10 according to an embodiment is shown.

[0115] The athlete module 11 may include a feature extraction stage 61, which may also be referred to as a parameter extraction stage. Raw and / or pre-processed sensor measurement data 14a (as an option for athlete-based data 14) generated by the sensor device 17 included in the athlete module 11 may be fed into the feature extraction stage 61. The feature extraction stage 61 may extract one or more features or parameters from the raw and / or pre-processed sensor measurement data 14a.

[0116] As a non-limiting example of feature extraction, we will briefly refer to Figure 7 , which shows a graph including raw measurement data that has been obtained by a first athlete module 111 and a second athlete module 112, both attached to one athlete 11'. Specifically, the first athlete module 111 is attached to the left shin guard of the athlete 11' and the second athlete module 112 is attached to the right shin guard of the athlete 11'.

[0117] Each of the athlete modules 111, 112 including an accelerometer delivers raw measurement data. Figure 7 The accelerometer data depicted in represents the respective norms of the respective accelerometer vectors. Specifically, the first graph 71 shows data for the first athlete module 111 attached to the left shin guard. The second graph 72 shows data for the second athlete module 112 attached to the right shin guard.

[0118] More specifically, the figure shows the time behavior of the norm of the acceleration vector obtained from the accelerometers on the left leg 71 and the right leg 72 of the athlete 11'. Relevant information is contained in the time instance where the norm deviates significantly from gravity at about 9.81 m / s². The parameter or feature represented by the time-constrained increase in the norm of the acceleration vector of the second athlete module 112 can be detected, as highlighted in the rectangular boxes 73, 74.

[0119] These two features (or parameters) can be detected in the aforementioned feature extraction stage 61 (see Figure 6 ). Therefore, the two peaks in the figure can be called features, specifically athlete features 14b1 and 14b2.

[0120] According to an example, at least one of the extracted features 14b1 , 14b2 may be sent as athlete-based data 14 to the computing unit 13 , wherein the computing unit 13 may determine an activity profile of the athlete 11 ′ based on the received extracted features 14b1 , 14b2 .

[0121] Additionally or alternatively, as in Figure 6 As shown in the example of , the activity profile of the athlete 11' can be determined by the athlete module 11. For example, the extracted features 14b1, 14b2 can be fed into the physical event classification / estimation stage 62. In the physical event classification / estimation stage 62, at least one of the extracted features 14b1, 14b2 can be classified as a physical event. The physical events of the athlete module 11 can also be referred to as physical athlete events.

[0122] Reference again Figure 7 , the depicted measurement portion shows a time portion of a scene where a pass is made between two players. Each of the above two player features 14b1, 14b2 highlighted in rectangular boxes 73, 74 may represent a kick that has been performed by one of the players 11' to which the first player module 111 and the second player module 112 are attached. Thus, the event classification / estimation stage 62 (see Figure 6 ) Each of these two features 14b1, 14b2 can be estimated or classified as the physical event "kicking the ball". Kicking the ball with the right leg can be visible and even classifiable.

[0123] Reference again Figure 6 , the physical event classification / estimation stage 62 may determine an activity profile 14c of the athlete 11' based on at least one classified physical athlete event. In other words, the activity profile 14c of the athlete 11' may include at least one physical athlete event. The activity profile 14c of the athlete 11' may include a plurality of physical athlete events in a sequential time sequence, e.g., running, then sprinting, then kicking a ball, etc. In this exemplary embodiment, the activity profile 14c of the athlete 11' may be determined by the athlete module 11 itself.

[0124] Additionally or alternatively, the physical event classification / estimation stage 62 may process one or more previous events 64 that may have been previously classified in the step of classifying a feature as a physical athlete event.

[0125] Additionally or alternatively, as a non-limiting example based on the athlete's data 14, the categorized physical athlete events may be sent to the computing unit 13, wherein the computing unit 13 may generate an activity profile for the athlete 11 ' based on the received physical athlete events.

[0126] like Figure 6 It can be further seen that the equipment module 12 may include a feature / parameter extraction level 65 and a physical event classification / estimation level 66, both of which have the same or similar functions as the above-mentioned target / parameter extraction level 61 and physical event classification / estimation level 62 included by the athlete module 11.

[0127] For example, raw and / or pre-processed sensor data 16a, which may be generated by sensor devices 19 included in equipment module 12, may be received by feature / parameter extraction stage 65. As exemplarily described above with reference to athlete module 11, one or more equipment features may be extracted from the signal.

[0128] Additionally or alternatively, as a non-limiting example of equipment-based data 16 , one or more extracted equipment features 16 b may be sent to the computing unit 13 , wherein the computing unit 13 may determine an action profile of the equipment 12 ′ based on the received one or more equipment features 16 b .

[0129] Additionally or alternatively, such as Figure 6 As shown, one or more extracted equipment features 16b can be fed into the physical event classification / estimation stage 66, in which the received one or more extracted equipment features 16b can be classified, as exemplarily illustrated above with reference to the athlete features 14b and the athlete module 11.

[0130] For example, equipment features representing an impact on ball 12' may be extracted from the signal of sensor device 19 by feature / parameter extraction stage 65. The impact may be classified as a physical equipment event "ball kicked" by physical event classification / estimation stage 66. These physical events may also be referred to as physical equipment events, or in this example, physical ball events.

[0131] The physical event classification / estimation stage 66 may generate an action profile 16c of the equipment 12' based on at least one classified physical equipment event. In other words, the action profile 16c of the equipment 12' may include at least one physical equipment event. The action profile 16c of the equipment 12' may include multiple physical equipment events in a sequential time sequence, for example, being kicked, then flying in the air, and then rolling on the ground. In this exemplary embodiment, the action profile 16c of the equipment 12' may be determined by the equipment module 12 itself.

[0132] Additionally or alternatively, the physical event classification / estimation stage 66 may process one or more previous events 67 that may have been previously classified in the step of classifying a feature as a physical ball event.

[0133] Additionally or alternatively, as a non-limiting example of equipment-based data 16, one or more categorized physical equipment events may be sent to computing unit 13, wherein computing unit 13 may determine an action profile of equipment 12' based on the received physical equipment events.

[0134] As a result, the computing unit 13 may receive at least the player-based data 14 from the player module 11 and the equipment-based data 16 from the equipment module 12 .

[0135] The athlete-based data 14 may be at least one of the following: raw and / or preprocessed sensor data 14a, which may be generated by a sensor device 17 included in the athlete module 11, and / or one or more extracted athlete features 14b, and / or one or more classified physical athlete events, wherein the one or more physical athlete events may generate an activity profile 14c of the athlete 11'. The equipment-based data 16 may be at least one of the following: raw and / or preprocessed sensor data 16a, which may be generated by a sensor device 19 included in the equipment module 12, and / or one or more extracted equipment features 16b, and / or one or more classified physical equipment events, wherein the one or more physical equipment events may generate an action profile 16c of the equipment 12'.

[0136] Regardless of whether the motion profile 16c of the equipment 12 and the activity profile 14c of the player 11 are determined by the respective equipment / player modules 11, 12 or by the computing unit 13, the computing unit 13 may determine the game scenario based on the motion profile 16c and the activity profile 14c.

[0137] like Figure 6 As shown, the computing unit 13 may include a play scenario classification / estimation stage 68 , wherein the motion profile 16 c of the equipment 12 ′ and the activity profile 14 c of the player 11 ′ may be classified into one of a plurality of play scenarios 18 .

[0138] For example, a plurality of game scenes may be stored in the computing unit 13. The computing unit 13 may compare both the activity profile 14c and the action profile 16c with corresponding reference profiles that may be associated with a category representing a typical game scene. Thus, the received activity profile 14c and action profile 16c may be classified into one category, for example, into one of the plurality of game scenes 18.

[0139] Additionally or alternatively, the play scenario classification / estimation stage 68 may process one or more previous data (eg, player-based data 14 and / or equipment-based data 16) in the step of classifying the activity profile 14c and motion profile 16c as a play scenario.

[0140] In the following, Figure 6 A brief overview of will give some examples of how features, physical events, and action or activity profiles can be distinguished:

[0141] Features may be, for example, parameters (e.g., vectors) obtained from raw measurement data of the corresponding sensor device. Thus, a feature may be, for example, the energy of a three-axis accelerometer signal. A sensor data set or a sequence of sensor data sets may be used to obtain features (e.g., energy of an accelerometer signal, spectral features, etc.). Features may be extracted from sensor data as a preprocessing stage for classification of the sensor data.

[0142] Events can be, for example, classification results obtained from raw measurement data or from other available information (e.g., extracted features). On a module basis, these events can be physical events of equipment or players, respectively (e.g., a ball is rolling, a player is running).

[0143] A higher level of abstraction is sports-related events, which may also be referred to as game scenarios (e.g., a pass between two players). Game scenarios may be determined based on the activity profiles of the players and the motion profiles of the equipment. The activity profiles of the players may include at least one physical player event or multiple (continuous) physical player events. The motion profiles of the equipment may include at least one physical equipment event or multiple (continuous) physical equipment events.

[0144] Physical events may be classified / estimated using features or feature sequences and / or knowledge about previous situations (features, physical events, sport-related events (including knowledge about all players / equipment)).

[0145] Sports-related events, i.e., game scenarios, can be estimated using physical events or sequences of physical events and / or knowledge about previous situations (physical events, sports-related events (including knowledge about all players / equipment)).

[0146] As mentioned earlier, for example, ref. Figure 3, the system 10 of the present invention may include one or more player modules 111, 112. As a non-limiting example, the concept of cooperation between two player modules 111, 112 and one equipment module 12 may be illustrated by detecting a typical scenario in a soccer match where a player 'A' passes a ball 12' to another player 'B'. At least one player module 111 may be attached to the player 'A', at least another player module 112 may be attached to the player 'B', and at least one equipment module 12 may be attached to the ball 12'.

[0147] In this example, player 'A' may kick ball 12', and ball 12' may fly in the air and be stopped by player 'B' before it hits the ground. In the system 10 of the present invention described herein, the physical player event "kick ball" from player 'A' can be identified by the player module 111 attached to player 'A'. Such physical events can be identified by various techniques (e.g., matching patterns, frequency analysis, etc.). Similarly, the equipment module 12 attached to ball 12' can also (e.g., simultaneously) identify one or more equipment features, such as impact, rotation rate, and impact frequency, and it can also estimate the flight parameters of ball 12', such as height, travel distance, rotation, etc. In addition, the equipment module 12 can generate physical equipment events based on the equipment features. By combining the information from the player module 111 attached to player 'A' and the equipment module 12 attached to ball 12', for example, by using the semantics that can associate the sequence of extracted features to actions (e.g., physical player events and activity profiles and / or physical equipment events and action profiles), it can be inferred that player 'A' has kicked ball 12' away from him. At the other end, when the ball 12' is stopped by a second player 'B', the player module 112 attached to player 'B' and the equipment module 12 attached to the ball 12' may recognize the corresponding event and transmit it to the computing unit 13. The computing unit 13 may then synchronize the received physical events (and / or activity profiles and / or action profiles) and correlate them to infer that the ball 12' was successfully passed from player 'A' to player 'B'.

[0148] Fig. 8A , Figure 8B and Figure 8C An example of equipment-based data that has been collected when a player kicks a ball and makes a high pass is shown. An equipment module is attached to the ball, wherein the equipment module includes a sensor device including a rotation sensor and an acceleration sensor.

[0149] Fig. 8A The raw measurement data of the rotation sensor is shown. Figure 8BThe raw measurement data of the acceleration sensor is shown. As described above, features are extracted and classified into physical ball events. These physical ball events are collected and combined into action profiles, which are Figure 8C shown.

[0150] As in Fig. 8A The rotation graph and Figure 8B As can be seen at the beginning of the time scale of the accelerogram, a first peak can be detected, which can be extracted and classified as a physical ball event 81 ( Figure 8C ) "The ball was kicked out."

[0151] Over time, the ball's spin decreases ( Fig. 8A ), the acceleration of the ball ( Figure 8B ) includes temporally consecutive peaks. This may occur when a ball flies in the air after having been kicked and then bounces (rebounds) on the ground continuously. These peaks can be extracted and classified as physical ball events 82 ( Figure 8C ) "High Pass / Bounce".

[0152] Figure 8C The last two physical events 83 shown are classified as "ground pass / touch". This may occur when the bounce of the ball only includes a low height at the end of a high pass. The low height bounce may also occur when a low pass is made. Therefore, it can be interpreted as a low pass.

[0153] However, multiple physics ball events can produce Figure 8C An exemplary description of the motion profile of the ball. The motion profile may indicate that a high pass was made.

[0154] The corresponding activity profile of the player who may have kicked the ball (not shown) can be compared with the motion profile of the ball ( Figure 8C ) are fused. Therefore, the computing unit 13 can determine the corresponding game scene involving the player and the ball, that is, the player may have made a high pass.

[0155] The computing unit 13 may process another activity profile of another player module attached to another player. The other activity profile may indicate that the other player may have received a pass. Therefore, the computing unit 13 may determine a corresponding game scenario involving both the player and the ball, i.e., a specific first player may have made a high pass to a specific second player.

[0156] Game-related statistics can be created based on the determined game scenario. For example, if the second player who receives the high pass is a member of the same team, then this would be a successful high pass. If the second player is a member of another team, then this would be an unsuccessful high pass.

[0157] Fig.9A , Fig. 9B and Fig. 9C An example of equipment-based data that has been collected when two players perform a low pass between each other is shown. A low pass may also be referred to as a ground pass hereinafter. An equipment module is attached to the ball, wherein the equipment module includes a sensor device including a rotation sensor and an acceleration sensor.

[0158] Fig.9A The raw measurement data of the rotation sensor is shown. Fig. 9B The raw measurement data of the acceleration sensor is shown. As described above, features are extracted and classified into physical ball events. These physical ball events are collected and combined into action profiles, such as Fig. 9C shown.

[0159] Can be obtained from Fig.9A As can be seen from the time scale of the rotation diagram in , the rotation of the ball increases steadily. This could happen if the ball is rolling on the ground after it has been kicked and steadily gains momentum, i.e. it spins faster and faster.

[0160] As in Fig. 9B As can be seen at the beginning of the time scale of the accelerogram in , a first peak can be detected which can be extracted and classified as a physical ball event 91 ( Fig. 9C ) "Ball kicked". A subsequent smaller peak may also be detected, which may occur when the ball contacts the ground quickly and repeatedly, which may be typical for a low pass. These consecutive peaks may be extracted and classified as physical ball events 92 ( Fig. 9C ) "Ground Pass / Touch".

[0161] At about 172.6 seconds, it can be seen in the acceleration graph ( Fig. 9B ) while another peak can be detected in the rotation diagram ( Fig. 9C ) in the rotation. This can happen when a player receives a ground pass and stops the ball. These features can be extracted and classified as physical ball events 93 ( Fig. 9C )"stop".

[0162] Afterwards, the two players performed a second ground pass. Fig. 9C As can be seen in the action profile shown, another physical ball event 94 "Ball kicked out" followed by a continuous physical ball event 95 "Ground pass / contact" followed by another physical ball event 96 "Stop" are classified and inserted into the action profile.

[0163] The action profile may indicate that a low pass (i.e., a ground pass) was made. The corresponding activity profile (not shown) of the player who may have kicked the ball may be compared to the action profile of the ball ( Fig. 9C ) are fused. Thus, the computing unit 13 can determine the corresponding game scene involving the player and the ball, that is, the player may have made a low pass.

[0164] The computing unit 13 may process another activity profile of another player module attached to another player. The other activity profile may indicate that the other player may have received a low pass. Therefore, the computing unit 13 may determine a corresponding game scenario involving both the player and the ball, i.e., a specific first player may have made a low pass to a specific second player.

[0165] exist Fig.10 Another embodiment of the system 10 of the present invention is schematically shown in FIG. Figure 2 Therefore, regarding the configuration contained in Figure 2 For the elements in , refer to the corresponding description above.

[0166] However, in addition to Figure 2 In addition, the system 10 may also include another module, namely an infrastructure module 113, also called an infrastructure tag. The infrastructure module 113 may be mobile or preferably a fixed module associated with the stadium where the sports game is played.

[0167] The infrastructure module 113 may include a communication interface for communicating in a unidirectional or bidirectional manner with at least one of the athlete module 11, the equipment module 12, and the computing unit 13. Additionally or alternatively, the infrastructure module 113 may be configured to perform RSSI measurements in cooperation with the equipment module 12 via the channel 111, and / or to perform RSSI measurements in cooperation with the athlete module 11 via the channel 112.

[0168] The infrastructure module 113 may generate infrastructure-based data 115, which may be sent to the computing unit 13. In addition to the received player-based data 14 and the received equipment-based data 16, the computing unit 13 may use the received infrastructure-based data 115 to determine a play scenario.

[0169] In this non-limiting example, infrastructure-based data 115 may be included in an infrastructure update message, while athlete-based data 14 may be sent from athlete module 11 to computing unit 13 in a athlete update message, and / or equipment-based data 16 may be sent from equipment module 12 to computing unit 13 in an equipment update message.

[0170] Athlete update messages, equipment update messages, and infrastructure update messages may each include the same data structure.

[0171] Fig.11 An example of a data structure of a corresponding update message is shown, which data structure may be valid for each of an athlete update message, an equipment update message, and an infrastructure update message. The update message may include a feature list, which includes one or more features 14b, 16b, which features 14b, 16b may have been described above with reference to Figure 6 Additionally or alternatively, it may include an event list, the event list including one or more physical events, which may have been described above with reference to Figure 6 Additionally or alternatively, it may include a raw data list, which includes one or more raw sensor measurement data 14a, 16a, which can be obtained by at least one of the sensor devices 17, 19 of the corresponding module 11, 12, 113.

[0172] Optionally, the corresponding update message may include an RSSI list including information about RSSI values ​​between two or more modules or tags, which may include one or more athlete modules 11, and / or one or more equipment modules 12 and / or one or more infrastructure modules 113. Further optionally, it may include a sender compression ID for identifying the sender of the message and a timestamp for synchronization purposes.

[0173] Since the tags 11, 12 may be equipped with a radio communication unit, they may additionally make temporal measurements of RSSI with each other. Since the RSSI value depends on both the distance and the position of the two tags 11, 12 between which the RSSI is measured, the tags 11, 12 may calculate a rough distance estimate to another tag in the system 10. The tags 11, 12 may report a list of other tags from which they received the signal with the strongest RSSI. The list of RSSI measurements may be fused with event detection at the tags 11, 12 themselves, for example to further improve the determination of ball possession.

[0174] The computing unit 13 can additionally manage a database 22 (see Figure 2 ), whereby the database 22 may contain information relating to a player's playing position, physical statistics, habits, history of events of a match or up to all earlier classified events, or the development of a team.

[0175] By fusing information about the timing of the ball, higher-level training statistics can be generated, such as successful or missed passes within a team. Other statistics may include, but are not limited to, pass frequency between two members of a team, contact time, first contact accuracy, etc.

[0176] As described above, the system 10 may include one or more infrastructure modules 113 . Fig.12 A non-limiting example of a system 10 is shown including twelve infrastructure modules 113 associated with a soccer playing field 121. Furthermore, as non-limiting examples, a ball 12' including at least one equipment tag 12, a first player 11'1 including at least one player module 111, and a second player 11'2 including at least one further player module 112 are depicted.

[0177] Depending on the embodiment, at least one of the infrastructure modules 113 may be arranged inside, outside, at, on, over, under, or around the sports field 121 .

[0178] Additionally or alternatively, at least one of the infrastructure modules 113 may be associated with a device belonging to the sports field 121. Such a device may be, for example, a goal post of a soccer or football goal, the edge of a basket for a basketball game, etc. At least one infrastructure module 113 may be arranged inside, outside, at, on, above, below or around the device of the sports field 121.

[0179] Again, the computing unit 13 may be configured to receive infrastructure-based data 115 from the infrastructure module 113, and the computing unit 13 may be configured to combine the infrastructure-based data 115 with the athlete-based data 14 and the equipment-based data 16 for determining a match scenario. In this example, the determined match scenario may be a match scenario in which an athlete, equipment, and optionally an infrastructure may be involved. Therefore, the computing unit 13 may determine a match scenario 18 based on the athlete-based data 14 representing the activity profile 14c of the athlete 11', the equipment-based data 16 representing the action profile 16c of the equipment 12', and the infrastructure-based data 115.

[0180] At least one of the infrastructure modules 113 may include a directional antenna including a directional antenna pattern, which may also be referred to as a sectorized antenna. Fig.12 , each infrastructure module 113 may include a directional antenna, wherein the directional antenna pattern is symbolized by a teardrop-shaped symbol 122 .

[0181] According to an embodiment, at least one of the infrastructure modules 113 may include a directional antenna 122 that is directed toward the sports field 121. For example, at least one infrastructure module 113 may be arranged such that the directional antenna pattern 122 is directed orthogonal to the spatially nearest boundary of the sports field 121. For example, Fig.12 The upper four infrastructure modules 113 and the lower four infrastructure modules 113 in the embodiment can be arranged along the side lines 123, 124 of the sports field 121. Therefore, with respect to these infrastructure modules 113, the corresponding side lines 123, 124 are considered to be the closest delimiting lines in space of the sports field 121.

[0182] As in Fig.12 As can be seen in FIG. 1 , the infrastructure modules 113 are arranged such that the directional antenna pattern 122 is oriented orthogonally to the respective side lines 123 , 124 of the sports field 121 .

[0183] As in Fig.12 It can also be seen in FIG. 1 that at least one infrastructure module 113 can be arranged in the goal area. According to this example, the at least one infrastructure module 113 can be arranged so that the directional antenna pattern 122 is oriented parallel to the spatially nearest boundary line of the sports field 121. In this example, the spatially nearest boundary line would be the goal lines 125, 126.

[0184] As in Fig.12 As can be seen in FIG. 1 , the infrastructure modules 113 are arranged such that the directional antenna pattern 122 is oriented parallel to the respective goal lines 125 , 126 of the sports field 121 .

[0185] Additionally or alternatively, at least one infrastructure module 113 may be arranged so that the directional antenna pattern 122 is oriented along the spatially nearest boundary line of the sports field 121. For example, at least one infrastructure module 113 may be arranged so that the directional antenna pattern 122 is oriented along the side lines 123, 124 or along the goal lines 124, 125. Each sports game may have its own boundary line, and the above-described embodiments may be applied to any other sports other than the exemplarily described English football game.

[0186] Therefore, the stadium 121 itself can be equipped with one or more of the above-mentioned infrastructure tags 113. The infrastructure tags 113 can differ from the player tags 11 and / or ball tags 12 in that they are fixed for the duration of the game. In particular, as described above, the infrastructure tags 113 can be equipped with directional antennas 122. The directional antennas 122 can be used for at least one of the following two purposes: first, they can be used to provide relative positioning of the player / ball tags 11, 12 in the stadium 121 with respect to the infrastructure tags 113, and second, they can be used to partially reduce interference from external radio (e.g. Bluetooth) devices to ensure communication, which external radio devices can be deployed nearby by spectators, for example, for their personal gadgets in the game.

[0187] Due to the directional (ie, sectorized) nature of the individual antennas 122, they can indicate whether the ball 12' is likely to be out of play. This information can be further combined with subsequent events (eg, throw-ins or free kicks) to generate statistics for the game.

[0188] Of particular interest here are infrastructure tags 113, which can indicate, for example, whether the ball 12' has crossed the goal lines 125, 126. The purpose of this technology is not primarily to make a "goal" decision at a competition level, but to automatically track the progress of the training phase. To this end, assuming that two infrastructure tags 113 can be placed on either side of the corresponding goal post, a decision can be made whether the ball 12' may have entered the goal post by estimating the RSSI of the ball 12' at the two infrastructure tags 113, or estimating the RSSI of the two tags 113 at the ball 12' itself due to channel reciprocity. The infrastructure tags 113 may also be referred to as anchors in this article.

[0189] Fig.13 and Fig.14 A rough depiction of the path loss from two infrastructure tags 1131, 1132 to the ball 12' is shown. The two infrastructure tags 1131, 1132 may also be referred to as Anchor 1 and Anchor 2. The two infrastructure tags 1131, 1132 are placed at the left goal post and the right goal post respectively.

[0190] Fig.13 are measured values ​​for the isotropic case, ie using an antenna comprising an isotropic antenna pattern. Fig.14 The measurements are shown for a sectorized case, i.e. using an antenna comprising a sectorized (i.e. directional) antenna pattern. In the latter case, the infrastructure tags 1131, 1132 are mounted such that their respective directional antenna patterns 122 are oriented parallel to the goal lines 124, 125, as shown in FIG. Fig.12 shown.

[0191] In both cases, the infrastructure tags 1131, 1132 are mounted at a height of 1.66 m, which is half the size of the goal post. Fig.13 The results shown in and for the sectorized case in Fig.14 The results shown in are valid for free space path loss and do not need to take lognormal shading into account.

[0192] The results show that in the isotropic case ( Fig.13 ) and in the case of sectorization ( Fig.14 ), the measured RSSI can be used to infer whether the ball 12' has passed the post. Using a probabilistic model, the RSSI information and the dynamics of the game can be fused to form a decision on whether a goal has been scored. However, using a sectorized antenna including a directional antenna pattern can reveal more accurate and reliable results. Therefore, sectorized antennas can be preferably used in the infrastructure module 113.

[0193] like Fig.15 As shown, RSSI measurements may be performed collaboratively between one or more athlete tags 111, 112, one or more equipment tags 12, and optionally one or more infrastructure tags 113. That is, infrastructure tag 113 may be configured to measure RSSI values ​​of signals received from at least one of athlete tags 111, 112 and equipment tag 12, as indicated by arrows 151, 152, 153 pointing to infrastructure tag 113. Furthermore, athlete tags 111, 112 and / or equipment tag 12 may be configured to measure RSSI values ​​of signals received from infrastructure tag 113, as indicated by arrows 151, 152, 153 pointing to athlete tags 111, 112 and equipment tag 12.

[0194] Fig.15 The simplified diagram of shows possible RSS measurements between all network members 111, 112, 12, 113, which can be represented by the two-way arrows 151, 152, 153, 154, 155, 156 on both sides, where the two sides highlight the fact that each tag can act as both a transmitter and a receiver.

[0195] Based on at least one of the above RSSI values, at least one of the athlete tags 111, 112, equipment tags 12, and infrastructure tags 113 may be configured to calculate the absolute position of the corresponding athlete tags 111, 112, and / or equipment tags 12 on the playing field 121. For example, the absolute position on the playing field within a decimeter or centimeter range may be calculated for at least one of the corresponding athlete modules 111, 112, and equipment modules 12.

[0196] Since the player 11', the ball 12' and the infrastructure tag 113 can all communicate via radio technology (such as but not limited to Bluetooth), the received signal strength (RSS) between each member of the network (i.e. between each tag 11, 12, 113) can be calculated at any time. If no fixed infrastructure tags 113 may be available, the method may be a swarm, such as a swarm where all other tags 11, 12 are mobile, as will be discussed further below. Since RSS propagation is related to distance, information about the relative positions of all members 11, 12, 113 of the network can be obtained in this way.

[0197] For those with A network with 11, 12, 113 members can be observed at any time. (Note that the RSS from network member A to network member B may not be the same as the RSS from member B to member A due to different antenna patterns). Fig.15 A simplified network of this type is shown, where a first player tag 111 may be attached to the player's left shin, a second player tag 112 may be attached to the same player's right shin, an equipment tag 12 may be attached to a ball 12', and an infrastructure tag 113 may be arranged at a playing field 121. Each of the first and second player tags 111, 112, the equipment tags 12, and the infrastructure tags 113 may be a network member.

[0198] For example, in a typical English soccer game (where each team has 11 players, with two tags for each player of each team, one ball tag, and four fixed infrastructure tags), there may be 2346 measurements available at one time. In contrast, the number of states (e.g., 46, the two-dimensional positions of all players and the ball) is relatively small, resulting in an overdetermined system of equations. The speed of the measured value from the position at P can be described, for example, in dBm by a The network member a (acting as the source / transmitter) sends a b The measured RSS RSS of the signal of another member b (acting as a receiver) ab An exemplary simplified model of , depending on their logarithmic Euclidean distance d ab :

[0199]

[0200] in represents the path loss parameter (in free space ), RSS0 represents the RSS at a reference distance, for example, 1 m, and Described in Pa The log-normal shadow fading component at , which is modeled as random but spatially correlated and thus also exhibits a dependence on the positions of the two tags (if tag a is fixed, this dependence is reduced to only dependence on b).

[0201] Now, these measurements or any subset thereof can be used to estimate a state of interest: the position of the ball 12' and / or the player 11' (or any subset of those) within the playing field 121, which can be defined by an infrastructure tag 113, or if no infrastructure tag 113 with a known position is available within the playing field 113, then by a relative position with respect to at least one of the tags 11, 12.

[0202] For example, this can be done using a Bayesian filter such as a Kalman or particle filter, the method of least squares and using the above model or another deterministic or stochastic model for RSS propagation (e.g., ray tracing). In addition, the following information can be used in filters such as those used for transformation modeling or prior probability distributions: additional information obtained by the system 10 (e.g., alignment of the positions of the ball 12' and the player 11' based on contact information of the ball, information obtained about the action state of the player 11' (running, dribbling, stopping), raw IMU data or kinematic information), and other available meta-information (e.g., the team position of the player (e.g., goalkeeper, defensive player, etc.), or input related to the player's technical form (e.g., 4-4-2), game events (corner kick, free kick, penalty kick, kick-off, throw-in, etc.), information related to the expected speed and movement of the English soccer player). In addition, observations of RSS at known locations (called "fingerprints") can be used in the scheme, for example as part of a measurement model, in particular to model the fading component.

[0203] According to a further exemplary embodiment, the sports field 121 may be divided into a plurality of segments, for example by a plurality of rectangles, wherein the absolute position may be calculated per segment for at least one of the respective player modules 111 , 112 and equipment module 12 .

[0204] As a non-limiting example, reference should again be made to Fig.12 It can be seen that the two infrastructure modules denoted by reference numerals 1131 and 1132 can be arranged relative to each other, for example, the first infrastructure tag 1131 can be arranged at the first side edge 123 and the second infrastructure tag 1132 can be arranged at the opposite second side edge 124. Each of the infrastructure tags 1131, 1132 can include a sectorized antenna including a directional antenna pattern, wherein the two antenna patterns can be oriented toward each other.

[0205] Thus, the sports field 121 may be divided into a plurality of sections. In this example, the first infrastructure module 1131 and the second infrastructure module 1132 may create a section 'C', which may be one of a plurality of sections 'A', 'B', 'C', and 'D'. For example, the ball 12' may be located in section 'C', the first player 11'1 may be located in section 'B', and the second player 11'2 may be located in section 'D'.

[0206] In any of the above-mentioned exemplary embodiments, at least one of the infrastructure module 113, the athlete module 11, the equipment module 12 and the computing unit 13 can be configured to calculate the relative spatial distance between the infrastructure module 113 and at least one of the athlete module 11 and the equipment module 12 based on the measured RSSI value.

[0207] Additionally or alternatively, in any of the above-mentioned exemplary embodiments, at least one of the infrastructure module 113, the athlete module 11, the equipment module 12 and the computing unit 13 can be configured to calculate the absolute position of at least one athlete module 11 and / or at least one equipment module 12 on the sports field 121 based on the measured RSSI value.

[0208] Thus, the above-mentioned positioning of at least one athlete 11 ′ and / or at least one equipment 12 ′ on the sports field 121 may be supported by one or more infrastructure tags 113 .

[0209] However, in yet another exemplary embodiment, it is possible that positioning of at least one athlete 11' and / or at least one equipment 12' may be performed without RSSI measurements associated with one or more infrastructure modules 113. In other words, it is possible that positioning of at least one athlete 11' and / or at least one equipment 12' may be performed using RSSI measurements associated with at least one athlete tag 11 and / or at least one equipment tag 12. According to such an embodiment, positioning of at least one athlete 11' and / or at least one equipment 12' may be performed even without any infrastructure tags 113.

[0210] Such an embodiment may correspond to Fig.15 In the illustrated embodiment, there may be no infrastructure tag 113. This method of RSSI positioning between movable one or more athlete modules 111, 112 and / or movable equipment modules 12 may be referred to as RSSI group positioning or RSSI collaborative positioning, wherein RSS may be calculated as described above, but without infrastructure tag 113 as a network member.

[0211] According to such an embodiment, at least one athlete module 111, 112 may be configured to measure a received signal strength indication RSSI value of a signal received from the equipment module 12. Additionally or alternatively, the equipment module 12 may be configured to measure a received signal strength indication RSSI value of a signal received from at least one athlete module 111, 112.

[0212] Based on the measured RSSI value, at least one athlete module 111 , 112 and / or equipment module 12 and / or computing unit 13 may be configured to calculate a relative spatial distance between at least one athlete module 111 , 112 and equipment module 12 .

[0213] Additionally or alternatively, based on the measured RSSI value, at least one athlete module 111, 112 and / or equipment module 12 and / or computing unit 13 may be configured to calculate the absolute position of at least one athlete module 111, 112 and equipment module 12 on the playing field 121.

[0214] For example, the ball module 12' and the player modules 111, 112 may have a radio (e.g. Bluetooth) data connection. With the radio (e.g. Bluetooth) capability, RSSI (signal strength) measurements may be made by the respective player / equipment modules 111, 112, 12 and used to estimate the distance between the modules 111, 112, 12, which may be the relative distance between the modules 111, 112, 12, from which the relative position of the modules 111, 112, 12 to each other may be determined / estimated. Combining all distance estimates, a picture of the situation may be created, including estimates of the positions of the players and the ball. This may also allow the absolute position of the modules 111, 112, 12 on the playing field 121 to be determined.

[0215] If in the best case RSSI measurement data is available on all modules 111, 112, 12, the relative distances between all modules 111, 112, 12 can be estimated and this information can be fused to improve the accuracy of the distance estimation and furthermore to locate the modules 111, 112, 12 relative to each other in a group-like manner, even if no infrastructure tags 113 may be available.

[0216] However, as described above, this can be improved by using modules (such as infrastructure modules 113) that are at fixed and known locations.

[0217] The computing unit 13 may receive one or more RSSI values ​​of the measurement from the athlete module 11 and / or the equipment module 12, and the computing unit 13 may use these one or more RSSI values ​​together with the athlete-based data 14 representing the activity profile of the athlete 11', the equipment-based data 16 representing the action profile of the equipment 12', and optionally the infrastructure-based data 155 (if available) to determine the game scenario 18.

[0218] In other words, according to an embodiment, the computing unit 13 may be configured to receive RSSI values ​​from at least one of the athlete module 11 and the equipment module 12, and determine one of a plurality of game scenarios 18 involving the athlete 11' and the equipment 12' based on the activity profile 14c of the athlete 11', the action profile 16c of the equipment 12', and the received RSSI values.

[0219] For example, the RSSI value may be used to determine which player on the field may have kicked the ball, for example, if two or more players are positioned close to each other. For example, in a matchup between a first player 11'1 including a first player module 111 and a second player 11'2 including a second player module 112, the first player 11'1 may kick the ball while the second player 11'2 may accidentally kick the grass. Both the player modules 111, 112 may generate acceleration signals that may be very similar. However, the RSSI value may indicate that the ball is closer to the first player 11'1 than the second player 11'2.

[0220] To this end, the RSSI value associated with the first player module 111 may be compared with the RSSI value associated with the second player module 112. For example, if the result of comparing the RSSI values ​​may reveal that the spatial distance from the first player module 111 to the ball 12' is smaller than the spatial distance from the second player module 112 to the ball 12', then this may indicate that the first player 11'1 is closer to the ball 12' than the second player 11'2. Therefore, there is a higher probability that the first player 11'1 may have kicked the ball 12'.

[0221] Based on this, the calculation unit 13 may determine the game scenario: the first player 11 ' 1 "wins the match" based on the player-based data 14 (activity profile), the equipment-based data 16 (motion profile) and the RSSI value of the first player 11 ' 1 , for example.

[0222] More generally, the computing unit 13 may be configured to receive a first RSSI value associated with the first athlete module 111 and a second RSSI value associated with the second athlete module 112, wherein the computing unit 13 may also be configured to compare the first and second RSSI values, and if the result of comparing the RSSI values ​​indicates that the first athlete module 111 has a smaller spatial distance to the equipment 12 than the second athlete module 112, then obtaining athlete-based data 14 of the first athlete module 111 for determining a game scenario.

[0223] The information of the above inventive concept may be fused, for example, in this context the dynamics of modules 111 , 112 , 12 may be helpful, for example, near zero speed, high speed movement, high dynamic movement, information about ball control, etc.

[0224] During the estimation process, channel parameters of the propagation channel (e.g. path loss coefficient, shadowing, log-normal fading, method Kriging) can be estimated in addition to the distance and position (e.g. RSSI measurement bias and nonlinearity).

[0225] Such RSSI collaborative positioning can have the following advantages:

[0226] • One of the cheapest positioning systems available

[0227] • Inexpensive: uses low-cost communications (Bluetooth or WiFi) compared to tracking systems with high-speed cameras or precise radio positioning

[0228] • Reuse of communications infrastructure

[0229] • No infrastructure installation required

[0230] • Hardware components will be used in the above system of the present invention

[0231] • Increased meaning of statistics in combination with game events from the system of the present invention, e.g. where on the field and in what context a pass was successful or unsuccessful

[0232] Hereinafter, the inventive concept will be briefly described again and summarized in other ways.

[0233] For coaches and the athletes themselves, objective data measuring the performance of athletes in a team sport setting such as football or soccer can be valuable information to be used to improve performance at the individual level and / or team level. To do this, statistics need to be generated, such as successful and missed passes from one athlete to another, pass speed, type of pass (high / low), ball possession, ball movement dynamics, athlete performance, etc.

[0234] To generate statistics, measurements can be taken by attaching modules 11, 12 (also referred to herein as tags, or depending on the location of attachment, as player modules 11 or equipment modules 12) preferably equipped with sensors 17, 19 and wireless communication capabilities with various entities in the game (e.g., players 11', balls 12', boundaries). The unsolved problem is to obtain and optimally combine these measurements to generate the required statistics. At the same time, the solution must also be portable, affordable, energy-efficient, scalable and resistant to external interference.

[0235] Therefore, the inventive concept may suggest placing the module 12 in the ball 12' and also having the player module 11. Compared to the prior art, it may be detected that the player 11' may have kicked the ball 12', that the ball 12' may have been kicked and later the player 11' may have accepted the ball 12', and that the ball 12' may have been accepted by another player 11'. Combining this information, estimating the pass from player A to player B will have much higher accuracy than using, for example, only the player module, and will be much cheaper than using precise positioning. Even if more players and balls are involved, for example in a training scenario, it will be easier to distinguish between successful and unsuccessful passes.

[0236] The proposed system may include Figure 2 The following components are shown:

[0237] 1) A valid tag (ball tag 12) attached to ball 12'

[0238] 2) At least one valid tag (athlete tag 11) attached to athlete 11'

[0239] 3) one or more computing units 13 (computing units) and

[0240] 4) At least one visualization unit 21 (APP + GUI)

[0241] 5) A database 22 containing relevant information for classifying events from the measured sensor data and other information, such as decision domains

[0242] The concept of the present invention can be briefly summarized as follows in the following points:

[0243] 1. Use of modules 11, 12 with motion sensors 17, 19 (e.g. inertial sensors, magnetometers and / or pressure sensors) and data connections between each other or with a central unit 13, and use of methods for analyzing data and e.g. estimating a pass from player A to player B, including feature extraction and decision trees and / or machine learning, while at least one module 12 can be integrated into a ball 12' and at least one module 11 can be attached to at least one player 11'.

[0244] ▫ Using the above method, events can be classified for each tag 11, 12 separately, and the obtained information can be fused between tags 11, 12 and using additional information (e.g., from a database 22) to generate higher-level information.

[0245] ▫ The database 22 may contain information about the player's playing position (goalkeeper, defense, offense, etc.), physical statistics, habits, history of events of the game or earlier events up to all classifications, or the development of the team.

[0246] ▫ Higher level information is things like passes from player A to player B, unsuccessful / successful passes, time of possession, first contact accuracy, unsuccessful / successful interceptions, duel wins / losses, and statistics and analysis based on these

[0247] 2. If RSSI or distance measurement data is available on at least some of the modules 11, 12 including the central unit 13, the mentioned higher level information can be improved by including RSSI measurement data, for example to estimate the distance between the tags 11, 12 (in particular between the player tag 11 and the ball tag 12).

[0248] 3. RSSI collaborative positioning

[0249] ▫ Idea: The ball module 12 and the player module 11 can have a Bluetooth data connection. With the help of the Bluetooth module, RSSI (signal strength) measurements can be made and used to estimate the distance between the modules 11, 12. Combining all distance estimates, a picture of the situation can be created, including estimates of the positions of the players and the ball.

[0250] ▫ Therefore, if RSSI measurement data is available on all modules 11, 12 in the best case, the distance between all modules 11, 12 can be estimated and this information can be fused to improve the accuracy of the distance estimation and furthermore to position the modules 11, 12 relative to each other in a group-like manner.

[0251] ▫This can be improved by using modules (such as infrastructure module 113) at fixed and known locations.

[0252] ▫The information of the above invention can be integrated:

[0253] The dynamics of modules 11, 12 can be helpful in this context, such as near-zero speed, high-speed motion, high-dynamic motion, etc.

[0254] Information about ball control in this article is helpful

[0255] ▫ During the estimation process, in addition to the distance and position (e.g., RSSI measurement bias and nonlinearity), channel parameters of the propagation channel can also be estimated (e.g., path loss coefficient, shadowing, log-normal fading, method Kriging).

[0256] Instead of a player tag 11 being fixed to a player 11 ', the tag 12 may be fixed to an object such as a bottle, a pole, a goal, a bouncer etc. The object may then be monitored on the tag 12 by analysing the sensor data if for example a target has been hit.

[0257] Furthermore, a system 10 of the present invention may be provided, the system 10 being used to determine a game scene 18 in a sports game, the system 10 comprising:

[0258] at least one player module 11, configured to be attached to a player 11' of the game,

[0259] at least one equipment module 12 configured to be attached to sports equipment (e.g., a ball, a puck) 12' to be used in the game, and

[0260] computing unit 13,

[0261] The athlete module 11 comprises: a first sensor device 17, comprising at least a multi-axis accelerometer; a signal processing unit for determining parameters (e.g. extracting features) 14b based on raw sensor measurement data 14a received from the first sensor device 17, and for classifying the features 14b as physical athlete events (for determining an activity profile 14c of the athlete 11'); and a radio communication device for sending the physical athlete events and / or the activity profile 14c to the computing unit 13,

[0262] The equipment module 12 comprises: a second sensor device 19, comprising at least a multi-axis accelerometer; a signal processing unit for determining parameters (e.g. extracting features) based on sensor measurement data 16a received from the second sensor device 19, and for classifying said features 16b as physical equipment events (for determining an action profile of the equipment 12'); and a radio communication device for sending the physical equipment events and / or the action profile 16c to the calculation unit 13,

[0263] The computing unit 13 includes: a radio communication device for receiving physical athlete events and / or activity profiles 14c from the athlete module 11 and physical equipment events and / or action profiles 16c from the equipment module 12, and also includes: a classifier 68 for classifying the physical athlete events and / or activity profiles 14c together with the physical equipment events and / or action profiles 16c into a game scene 18.

[0264] In the following, some non-limiting examples of game scenarios may be given, wherein the game scenarios may include at least one item from the group consisting of: a low pass performed by a player, a high pass performed by a player, a successful pass to another player from the same team, an unsuccessful pass to another player from a different team, a win or loss duel between two or more players, a penalty kick performed by a player, a corner kick performed by a player, a throw-in performed by a player, a volley performed by a player, a half-volley performed by a player, a header performed by a player, a dribble performed by a player, and the duration of a player's possession of the ball.

[0265] The system 10 of the present invention can provide the following advantages:

[0266] • Inexpensive: uses low-cost sensors (<50 €) compared to tracking systems with high-speed cameras or precise radio positioning

[0267] • Player and game equipment (e.g., balls) metrics and events, as tags can be integrated into balls and other game equipment

[0268] • Pre-processing of data on the tag (signal / data analysis and event detection and processing). As a result, the requirements for:

[0269] ▫ Delays due to radio communications

[0270] ▫ The amount of data to be transmitted via radio communication

[0271] ▫ Synchronization of tags

[0272] •Higher robustness and functionality than systems that evaluate only player tag metrics, as player events can be associated with and supplemented by ball events.

[0273] The system 10 of the present invention can be used to

[0274] • Motion analysis, metric investigation, state determination, state classification, information fusion, event recognition, event processing (localization, position, tracking, communication),

[0275] • Sports: All ball games, hockey, general sports with special equipment and one or more players interacting with it

[0276] • Training objectives and game analysis

[0277] • Digitization of training sessions and matches for later analysis or further processing

[0278] Although some aspects have been described in the context of an apparatus, it will be clear that these aspects also represent descriptions of corresponding methods, wherein a block or device corresponds to a method step or a feature of a method step. Similarly, aspects described in the context of a method step also represent descriptions of features of a corresponding block or item or a corresponding apparatus. Some or all of the method steps may be performed by (or using) a hardware device such as a microprocessor, a programmable computer, or an electronic circuit. In some embodiments, one or more of the most important method steps may be performed by such an apparatus.

[0279] According to certain implementation requirements, embodiments of the present invention may be implemented in hardware or software, or at least partially in hardware, or at least partially in software. Implementation may be performed using a digital storage medium (e.g., a floppy disk, DVD, Blu-ray, CD, ROM, PROM, EPROM, EEPROM, or flash memory) having electronically readable control signals stored thereon, which cooperate (or are capable of cooperating) with a programmable computer system to perform the corresponding method. Thus, the digital storage medium may be computer readable.

[0280] Some embodiments according to the invention comprise a data carrier having electronically readable control signals, which are capable of cooperating with a programmable computer system, such that one of the methods described herein is performed.

[0281] Generally, embodiments of the present invention can be implemented as a computer program product with a program code, the program code being operative for performing one of the methods when the computer program product runs on a computer.The program code may, for example, be stored on a machine readable carrier.

[0282] Other embodiments comprise the computer program for performing one of the methods described herein, stored on a machine readable carrier.

[0283] In other words, an embodiment of the inventive method is, therefore, a computer program having a program code for performing one of the methods described herein, when the computer program runs on a computer.

[0284] A further embodiment of the inventive method is, therefore, a data carrier (or a digital storage medium or a computer-readable medium) on which is recorded the computer program for performing one of the methods described herein. The data carrier, the digital storage medium or the recorded medium are typically tangible and / or non-transitory.

[0285] Therefore, another embodiment of the inventive method is a data stream or a sequence of signals representing the computer program for performing one of the methods described herein. The data stream or the sequence of signals may, for example, be configured to be transmitted via a data communication connection (eg, via the Internet).

[0286] A further embodiment comprises a processing means, for example a computer or a programmable logic device, configured to or adapted to perform one of the methods described herein.

[0287] A further embodiment comprises a computer having installed thereon the computer program for performing one of the methods described herein.

[0288] Another embodiment according to the invention comprises an apparatus or system configured to transmit a computer program to a receiver (e.g. electronically or optically), the computer program being used to perform one of the methods described herein. The receiver may be, for example, a computer, a mobile device, a storage device, etc. The apparatus or system may, for example, comprise a file server for transmitting the computer program to the receiver.

[0289] In some embodiments, a programmable logic device (e.g., a field programmable gate array) can be used to perform some or all of the functions of the methods described herein. In some embodiments, a field programmable gate array can cooperate with a microprocessor to perform one of the methods described herein. Generally, the method is preferably performed by any hardware device.

[0290] The devices described herein may be implemented using hardware devices, or using computers, or using a combination of hardware devices and computers.

[0291] The methods described herein may be performed using a hardware device, or using a computer, or using a combination of a hardware device and a computer.

[0292] The above embodiments are merely illustrative of the principles of the present invention. It should be understood that modifications and variations of the arrangements and details described herein will be apparent to other persons skilled in the art. Therefore, it is intended that the scope of the present invention be limited only by the scope of the appended patent claims and not by the specific details given by way of the description and explanation of the embodiments herein.

Claims

1. A system for determining a game scenario in a sports game, the system comprising: Athlete Module, which can be attached to an Athlete, an equipment module capable of being attached to sports equipment to be used in said sports game, and a computing unit for receiving athlete-based data from the athlete module and equipment-based data from the equipment module, the athlete-based data representing an activity profile of the athlete and the equipment-based data representing an action profile of the equipment, wherein the athlete module comprises: an athlete sensor device, comprising at least one acceleration sensor, the acceleration sensor being configured to provide raw measurement data associated with the athlete to which the athlete module is attached, the raw measurement data comprising a temporal behavior of a norm of an acceleration vector obtained from the acceleration sensor, wherein the athlete module is configured to perform feature extraction on the raw measurement data for determining one or more physical events that the athlete is capable of performing based on the extracted one or more features, wherein the feature extraction comprises determining a time-constrained increase in the norm of the acceleration vector, wherein the athlete module is configured to determine an activity profile of the athlete based on the extracted one or more features and / or based on the determined one or more physical events, and sending said activity profile as said athlete-based data to a computing unit, and / or sending the extracted one or more features and / or the determined one or more physical events to a computing unit for use in determining an activity profile of the athlete, and wherein the equipment module comprises: an equipment sensor device, comprising at least one of a multi-axis accelerometer and an angular rate sensor, the equipment sensor device being configured to provide raw measurement data associated with the equipment to which the equipment module is attached, wherein the equipment module is configured to perform feature extraction on the raw measurement data for determining a physical event that can be performed by the equipment or can be applied to the equipment based on one or more extracted features, wherein the feature extraction comprises determining one or more peaks of the raw measurement data of the multi-axis accelerometer and determining a time variation of the raw measurement data of the angular rate sensor, The equipment module is configured to: determine an action profile of the equipment based on the extracted one or more features and / or based on the determined one or more physical events, and sending the action profile as the equipment-based data to a computing unit, and / or sending the extracted one or more features and / or the determined one or more physical events to a computing unit for determining an action profile of the equipment, and Wherein, the computing unit comprises: a game scene classification stage, configured to combine the activity profile of the athlete and the motion profile of the equipment to classify the activity profile of the athlete and the motion profile of the equipment into one of a plurality of game scenes involving the athlete and the equipment.

2. The system according to claim 1, wherein: The athlete sensor device includes at least one of the following items: a multi-axis accelerometer, an angular rate sensor, a magnetometer and a pressure sensor, wherein the multi-axis accelerometer is configured to provide raw measurement data representing the acceleration of the athlete's body part to which the athlete module is attached, the angular rate sensor is configured to provide raw measurement data representing the rotation of the athlete's body part to which the athlete module is attached, the magnetometer is configured to provide raw measurement data representing the value of the earth's magnetic field, and the pressure sensor is configured to provide raw measurement data representing the pressure applied by or at the athlete's body part to which the athlete module is attached.

3. The system according to claim 1, wherein: The equipment sensor device includes at least one of the following items: a magnetometer and a pressure sensor, the multi-axis accelerometer is configured to provide the raw measurement data representing the acceleration of the equipment to which the equipment module is attached, the angular rate sensor is configured to provide the raw measurement data representing the rotation of the equipment to which the equipment module is attached, the magnetometer is configured to provide the raw measurement data representing the value of the earth's magnetic field, and the pressure sensor is configured to provide the raw measurement data representing the pressure exerted on the equipment to which the equipment module is attached.

4. The system according to claim 1, further comprising: A database configured to store additional information about at least one of the following items: the athlete, the equipment and the time history of the game, wherein the computing unit is configured to: retrieve the additional information from the database and determine one of a plurality of game scenarios involving the athlete and the equipment based on the activity profile of the athlete, based on the motion profile of the equipment and based on the additional information retrieved from the database.

5. The system according to claim 1, wherein: The athlete module is configured to determine the athlete-based data during a predetermined time window, and wherein the equipment module is configured to determine the equipment-based data during a predetermined time window, and wherein the computing unit is configured to synchronize the received athlete-based data and the received equipment-based data in time so that the time windows of the athlete-based data and the equipment-based data at least partially overlap in the time domain.

6. The system according to claim 1, wherein: The system comprises another athlete module attachable to another athlete, wherein the computing unit is configured to receive another athlete-based data from the other athlete module, the other athlete-based data representing another activity profile of the other athlete, The computing unit is configured to determine one of a plurality of game scenarios involving the athlete, the other athlete and the equipment based on an activity profile of the athlete, another activity profile of the other athlete and an action profile of the equipment.

7. The system according to claim 1, wherein: The system further comprises at least one infrastructure module, the at least one infrastructure module being associated with the sports field where the sports game is conducted and / or with the devices belonging to the sports field, the infrastructure module being arranged inside, outside, at, on, above, below or around the sports field and / or being arranged inside, outside, at, on, above, below or around the devices of the sports field, The computing unit is configured to receive infrastructure-based data from the infrastructure module and determine one of multiple game scenarios involving the athlete and the equipment based on the athlete's activity profile, the equipment's motion profile and the infrastructure-based data.

8. The system according to claim 7, wherein: The infrastructure module comprises a directional antenna which is directed towards the sports field and / or is directed parallel or orthogonal to a spatially nearest one of the boundaries of the sports field and / or is directed along the boundaries of the sports field.

9. The system according to claim 7, wherein: The infrastructure module is configured to measure a received signal strength indication RSSI value of a signal received from the athlete module and / or from the equipment module, and / or wherein at least one of the athlete module and the equipment module is configured to measure a received signal strength indication RSSI value of a signal received from the infrastructure module, and wherein at least one of the infrastructure module, the athlete module, the equipment module and the calculation unit is configured to calculate the following based on the measured RSSI value: the relative spatial distance between at least one of the athlete module and the equipment module and the infrastructure module, and / or a relative spatial distance between at least one of the athlete module and the equipment module on the playing field, and / or The absolute position of the player module and / or the equipment module on the playing field.

10. The system according to claim 1, wherein: The athlete module is configured to measure a received signal strength indication RSSI value of a signal received from the equipment module, and / or wherein the equipment module is configured to measure a received signal strength indication RSSI value of a signal received from the athlete module, and wherein at least one of the athlete module, the equipment module and the calculation unit is configured to calculate the following based on the measured RSSI value: the relative spatial distance between the athlete module and the equipment module, and / or The absolute positions of the player modules and the equipment modules on the playing field.

11. The system according to claim 1, wherein: The computing unit is configured to receive an RSSI value from at least one of the athlete module and the equipment module, and determine one of multiple game scenarios involving the athlete and the equipment based on the activity profile of the athlete, the action profile of the equipment, and the received RSSI value.

12. The system according to claim 10, wherein: The computing unit is configured to receive a first RSSI value associated with a first athlete module and a second RSSI value associated with a second athlete module, wherein the computing unit is further configured to compare the first RSSI value and the second RSSI value, and to obtain athlete-based data for the first athlete module if the result of comparing the RSSI values ​​indicates that the first athlete module has a smaller spatial distance to the equipment than the second athlete module.

13. A method for determining a game scene in a sports game, the method comprising: receiving athlete-based data from an athlete module attached to the athlete, the athlete-based data representing an activity profile of the athlete, receiving equipment-based data from an equipment module attached to athletic equipment to be used in the athletic competition, the equipment-based data representing an action profile of the equipment, Wherein, the step of receiving the athlete-based data from the athlete module includes: receiving raw measurement data from an athlete sensor device included in the athlete module, the athlete sensor device including at least one acceleration sensor, the acceleration sensor being configured to provide the raw measurement data, the raw measurement data being associated with the athlete to which the athlete module is attached, the raw measurement data including the time behavior of the norm of an acceleration vector obtained from the acceleration sensor; and performing feature extraction on the raw measurement data for determining one or more physical events that the athlete is capable of performing based on the extracted one or more features, wherein the feature extraction includes determining a time-constrained increase in the norm of the acceleration vector, determining an activity profile for the athlete based on the extracted one or more features and / or based on the determined one or more physical events, and sending said activity profile as said athlete-based data to a computing unit, and / or sending the extracted one or more features and / or the determined one or more physical events to a computing unit for use in determining an activity profile of the athlete, and Wherein, the step of receiving the equipment-based data from the equipment module comprises: receiving raw measurement data from an equipment sensor device included in the equipment module, the equipment sensor device comprising at least one of a multi-axis accelerometer and an angular rate sensor, the raw measurement data being associated with the equipment to which the equipment module is attached; and performing feature extraction on the raw measurement data for determining a physical event that can be performed by the equipment or applied to the equipment based on one or more extracted features, wherein the feature extraction comprises determining one or more peaks of the raw measurement data of the multi-axis accelerometer and determining a time variation of the raw measurement data of the angular rate sensor, determining an action profile of the equipment based on the extracted one or more features and / or based on the determined one or more physical events, and sending the action profile as the equipment-based data to a computing unit, and / or sending the extracted one or more features and / or the determined one or more physical events to a computing unit for determining an action profile of the equipment, and The activity profile of the player and the motion profile of the equipment are combined to classify the activity profile of the player and the motion profile of the equipment into one of a plurality of game scenarios involving the player and the equipment.

14. A computer-readable digital storage medium storing a computer program having a program code for executing the method of claim 13 when the computer program is run on a computer.

Images