KR102495821B1 - Big data collection system using LiDAR sensor for football game and online football game offering system therewith - Google Patents

Abstract

The present invention collects the trajectory information (position, speed, etc.) of each soccer player and the trajectory information of the ball for each preset cube (S) using LiDAR sensors, and then acquires them using an artificial intelligence algorithm. Analyzing the collected big data, detecting the pattern information according to each player's ability level and situation, and applying the detected ability level and situational pattern information to the corresponding player character of the online soccer game, so that the player character and actual player for various variables Not only can it increase the realism and reality of online soccer games by maximizing the sense of coincidence, but also it is possible to provide user-centered services by improving user participation in online soccer games, thereby increasing user interest, participation, and fun. It relates to a big data collection system that can increase and an online soccer game providing system using the same.

Description

Big data collection system using LiDAR sensor for football game and online football game offering system therewith}

The present invention relates to a system for collecting big data of a soccer game using a lidar sensor and a system for providing an online soccer game using the same. By maximizing the matching of characters, the actual player's ability level and patterns according to various situations are reflected in the soccer game to maximize the reality and realism of the soccer game, as well as providing user-centered services to increase user interest and interest. It relates to a big data collection system for soccer games using lidar sensors that can increase participation by inducing fun, and an online soccer game providing system using the same.

LiDAR (Light Detection And Ranging) means measuring the distance by detecting the time difference (Time of flight) or phase difference of the received optical signal after irradiating an optical signal to an external object, and has spatial resolution compared to radar. Due to the advantages of excellent resolution and resolution, it was conventionally used only in special fields such as aviation and satellite fields, but recently, it has been used in various fields such as civil and national fields such as surveillance and reconnaissance, robots, unmanned surface ships, aircraft such as drones, and industrial security and safety fields. is being extended.

LiDAR is widely used for tracking and detecting the trajectory of a detected object because it can model a stereoscopic image in real time by utilizing a three-dimensional point cloud of x, y, and z.

On the other hand, with the recent expansion of communication infrastructure and the development of the game content industry, various online soccer games have been put on the market, and these online soccer games have been spotlighted as popular games due to the advantage of combining the fun of games and sports.

However, conventional online soccer games are made in such a way that the movements and abilities of soccer players and soccer teams are standardized, and the movements of the players are determined according to the ability of the gamer, so not only does the realism and realism decrease, but also fun and interest are lost. The problem of deterioration occurs.

Accordingly, in recent years, online soccer games have been developed and are gaining popularity in which actual soccer players' abilities such as shooting, passing, heading, speed, and tackle are given to corresponding characters of online soccer games, but these conventional online soccer games are popular for each player. Since the ability value is not detected based on the actual movement of the player, but the ability value is detected by referring to the player's records and characteristics, there is a gap between the actual player's movement and the online movement, resulting in insufficient improvement in realism and realism. have

1 is a flowchart showing a method for providing characters for online sports games disclosed in Korean Patent Registration No. 10-1188826 (Title of Invention: Method and Apparatus for Providing Characters for Online Sports Games).

The method for providing a character for an online sports game of FIG. 1 (hereinafter referred to as the prior art) (S900) includes the steps of generating a plurality of national player characters based on actual players hosted in a real sports event and preset in a character providing device ( S910) and, when the user enters the character trading screen through the user terminal, displaying a plurality of player characters including a predetermined number of national player characters on the character trading screen (S940), and a player reset item from the user terminal When the use signal is received, displaying a plurality of national team player characters in excess of a predetermined number on the character trading screen (S960).

In addition, the step of generating a plurality of national player characters (S910) includes obtaining information on player characters that can be provided in the game, and acquiring information on actual sports players corresponding to each of the obtained plurality of player characters. Step of extracting national players who participated in the national competition by analyzing the obtained information on actual sports players; It consists of a selection step.

Since the prior art (S900) configured as described above can generate a player character based on information of an actual soccer player, all soccer players and soccer teams have the same ability values applied in the prior art, which can solve the problem of lack of realism and interest have an advantage

However, in the prior art (S900), since the method for acquiring information on an actual soccer player remains at the level of simply obtaining information by referring to already existing records or characteristics, the actual soccer player's movement pattern, shooting pattern, or heading pattern Since the athletic ability of the back is not applied at all, there is a problem in that the match between the actual soccer player and the game character is poor.

For example, even if it is the same player in an actual game, the probability of a header goal is different depending on the pattern of the ball being passed (speed, rotation amount, height, etc.), but the prior art (S900) simply has many header goals by real soccer players. In this case, since it is operated in such a way that the heading ability of the corresponding player character is set higher than that of other player characters, not only the sense of realism and realism is lowered, but also has a structural limit that lowers the fun and interest of the game.

In general, kicks in soccer are classified into inside kicks, in-front kicks, in-foot kicks, out-front kicks, toe kicks, and non-rotating kicks according to the location of contact with the foot. It is important that the optimal kick be made in response to various variables such as the angle of the goalkeeper, the position of the goalkeeper, and the position of the opposing team's wall, but the prior art (S900) did not consider the type of kick at all, so There is a problem in that interest and realism of the game are reduced by applying only the degree of accuracy and speed of shooting.

In other words, after collecting and detecting movement patterns of actual soccer players under various conditions, research on online soccer games that can significantly increase the fun and interest of the game by maximizing the realism and reality of the game by applying it to the player character is urgently needed. The situation is.

The present invention is to solve this problem, and the problem of the present invention is to convert the trajectory information (position, speed, etc.) of each actual soccer player and the trajectory information of the ball using LiDAR sensors to a preset cube ( After collecting each S), the collected big data is analyzed using an artificial intelligence algorithm to detect the pattern information according to the ability level and situation of each player, and the pattern information according to the detected ability level and situation corresponds to the online soccer game. By applying it to the player character, it is possible to increase the realism and reality of the online soccer game by maximizing the match between the player character and the actual player for various variables, and to improve user participation in the online soccer game to create a user-centered It is to provide a big data collection system that can provide services and increase user interest, participation, and fun, and an online soccer game providing system using the same.

The solution of the present invention for solving the above problems includes a plurality of sensing means installed in a soccer stadium to detect an object (player or ball), and a controller for transmitting the sensing data sensed by the sensing means to the outside. a data collection system; Using object detection data as input data, motion information for each player, including at least one of movement speed, jumping ability, passing ability, shooting ability, tackling ability, physical fighting ability, centering ability, and trapping ability, and situational The machine learning algorithm that outputs the situation pattern information for each player, which is pattern information, stores the game analysis algorithm, and analyzes the detected data transmitted from the controller using the game analysis algorithm to analyze the motion information for each player and the situation for each player. An artificial intelligence analysis server that detects pattern information; An online soccer game operation server that produces online soccer games and provides game services at the same time, and reflects the movement information for each player and the situation pattern information for each player received from the artificial intelligence analysis server to the corresponding player character, The sensing means is a LiDAR sensor that receives a reflected signal after transmitting a LiDAR signal, and the controller connects the football field to first horizontal virtual lines in the longitudinal direction in the horizontal direction and in the width direction in the horizontal direction. When divided into second horizontal virtual lines and vertical virtual lines in the vertical direction, when the divided hexahedron-shaped area is called a cube, a memory for storing positional information of each cube; a lidar signal scanning module for scanning lidar signals received from the lidar sensors; Using the scanning information by the LIDAR signal scanning module, a Sobel filter is applied to generate an edge image, which is a 3D modeling image consisting of edge lines, which are pixels whose pixel change rate is greater than or equal to a threshold value, and utilizes the location information of each cube set in advance an edge image extraction module for each cube that extracts an edge image for each cube; A control module for transmitting the edge image for each cube extracted by the edge image extraction module for each cube to the artificial intelligence analysis server, wherein the controller controls the edge line of the object and the intersection of horizontal lines equally formed at regular intervals Further comprising a center line extraction module for each cube that extracts the center line of an object from the edge image for each cube extracted by the edge image extraction module for each cube when the line connecting the median pixels of is referred to as a center line and the control module transmits the center line information for each cube extracted by the center line extraction module for each cube to the artificial intelligence analysis server.

delete

delete

delete

In addition, in the present invention, the artificial intelligence analysis server includes an artificial intelligence analysis unit, a database unit; a center line merging unit for each cube that merges successive center lines of successive cubes (S) by utilizing the center line information for each cube transmitted from the controller; Using the game analysis algorithm, the merged center line information input from the center line merging unit for each cube and the edge images for each cube received from the controller are analyzed to obtain motion information for each player and situation for each player. Artificial intelligence analysis unit for generating pattern information; It is preferable to include a control unit for transmitting motion information for each player and situation pattern information for each player generated by the artificial intelligence analysis unit to the online soccer game operation server.

In the present invention, the artificial intelligence analysis unit includes a data input module that receives merged center line information input from the center line merging unit for each cube and edge images for each cube transmitted from the controller; an analysis module which executes the game analysis algorithm by using the merged center line information input by the data input module and the edge images of each cube as input data of the game analysis algorithm; an object recognition module for recognizing a ball or body object for each cube through the game analysis algorithm; a situational awareness module recognizing the current state of each player and ball by the game analysis algorithm; a motion information generation module for each player for detecting motion information of each player by the game analysis algorithm; It further includes a situation pattern detection module for each player that detects each player's movement pattern according to the current state recognized by the situation awareness module, and the current state applied to the situation awareness module is 'attack' and 'attack' in the widest range. It is classified as 'defense', and the 'attack' category is classified into subcomponents including at least one of 'counterattack', 'delay', 'tactic type', 'free kick', 'corner kick', and 'penalty kick' , 'Defense' category is a sub-component including at least one of 'counterattack defense', 'pressure defense', 'local defense defense', 'dedicated mark defense', 'corner kick defense', and 'penalty kick defense' It is classified into , and the smallest is each player's own position and motion vector (direction and speed of movement), the position and motion vector of players on the same team, the position and motion vector of players on the opposing team, and the position and motion vector of the ball. It is desirable to recognize it in detail.

In addition, in the present invention, the artificial intelligence analysis unit detects the movement of the foot, the part of the foot in contact with the ball, and the part of the ball in contact with the foot when the player and the ball are in contact, and after the kick is made, each cube It is preferable to further include a matching information generation module for generating matching information by matching the speed and direction of the ball with reference to the trajectory of the ball.

In addition, in the present invention, the online soccer game operating server receives a gamer's selection of either a manual mode or an automatic mode, and when the manual mode is selected, the game is operated based on the gamer's manipulation, but when the automatic mode is selected, artificial intelligence It is desirable to proceed with the game by automatically determining the attack tactics, player position, penalty kick player, corner kick player, and free kick player using the movement information, matching information, and situation pattern information of the player characters.

According to the present invention having the above problems and solutions, after collecting the trajectory information (position, speed, etc.) of each actual soccer player and the trajectory information of the ball for each preset cube (S) using LiDAR sensors , Analyzing big data acquired and collected using artificial intelligence algorithms, detecting pattern information according to each player's ability level and situation, and applying the detected ability value and situational pattern information to the corresponding player character in the online soccer game. Not only can the realism and reality of online soccer games be enhanced by maximizing the match between player characters and actual players for variables, but also user-centered services can be provided by improving user participation in online soccer games. increase interest, engagement and fun.

1 is a flowchart showing a method for providing characters for online sports games disclosed in Korean Patent Registration No. 10-1188826 (Title of Invention: Method and Apparatus for Providing Characters for Online Sports Games).
2 is a block diagram showing an online soccer game providing system according to an embodiment of the present invention.
Figure 3 is an exemplary view showing the lidar sensors of FIG.
FIG. 4 is a plan view for explaining a cube applied to the object detecting unit of FIG. 3 .
Figure 5 is a side view of Figure 4;
6 is an exemplary view for explaining a 3D image for each cube of a trajectory of a ball acquired by an object detecting unit according to the present invention.
FIG. 7 is a block diagram illustrating the controller of FIG. 2 .
8(a) is an exemplary view showing an edge image for each cube extracted by the edge image extraction module of FIG. 7, and (b) is an example view showing a conventional LIDAR image.
FIG. 9 is an exemplary view showing the center line of each cube extracted by the center line extraction module for each cube of FIG. 7 .
10 is a block diagram illustrating an artificial intelligence analysis server of FIG. 2 .
11 is a block diagram illustrating an artificial intelligence analysis unit of FIG. 10 .
12 is a diagram for detecting a hitting part of a soccer ball according to the present invention using the analysis module of the present invention.
13 is a front view of a hitting part of a foot hitting a soccer ball according to the present invention.
Figure 14 is a side view of the hitting part of the foot hitting the soccer ball according to the present invention.

Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

2 is a block diagram showing an online soccer game providing system according to an embodiment of the present invention.

An online soccer game providing system 1 according to an embodiment of the present invention uses LiDAR sensors to transfer trajectory information (position, speed, etc.) of each soccer player and trajectory information of a ball to a preset cube (S). After collecting each item, the acquired big data is analyzed using an artificial intelligence algorithm to detect the pattern information according to each player's ability level and situation, and the detected ability value and situational pattern information is converted into the corresponding player character of the online soccer game. By applying it to the game, it is possible to increase the realism and reality of the online soccer game by maximizing the match between the player character and the actual player for various variables, and to improve user participation in the online soccer game to provide user-centered services. This is to increase the user's interest, participation, and fun.

In addition, as shown in FIG. 2, the online soccer game providing system 1 of the present invention includes a data collection system 3, an artificial intelligence analysis server 5, an online soccer game operating server 7, and a communication network 10. made up of

The communication network 10 provides a data movement path between the data collection system 3, the artificial intelligence analysis server 5 and the online soccer game operating server 7, and in detail, a wired and wireless network such as a wide area network (WAN) It can be composed of networks, 3G, 4G, 5G, LTE, etc.

3 is an exemplary view showing lidar sensors of FIG. 2 , FIG. 4 is a plan view illustrating a cube applied to the object sensing unit of FIG. 3 , and FIG. 5 is a side view of FIG. 4 .

As shown in FIGS. 2 to 5, the data acquisition system 3 is installed in an actual soccer stadium to transmit a LiDAR signal, and then receives the reflected LiDAR sensor 31-1, ... , (31-N) and lidar sensors (31-1), ..., by analyzing lidar signals transmitted and received by (31-N), each object (player or ball) for each preset cube (S) It consists of a controller 30 for detecting the edge image and center line of ).

LiDAR sensors (31-1), ..., (31-N) are installed spaced apart from each other on the side and top of the soccer stadium to transmit LiDAR signals and receive reflected signals.

At this time, Lidar (Light Detection And Ranging) fires a laser and measures the distance, concentration, speed, shape, etc. It is similar to RADAR (Radio Detection And Ranging), which measures the properties and calculates the distance by observing the round-trip time to the object using microwaves, but has a difference in that it uses light, unlike radar that uses radio waves. It is also called 'imaging radar'.

In addition, lidar signal information transmitted and received by the lidar sensors 31-1, ..., (31-N) is transmitted to the controller 30.

The controller 30 analyzes and scans the LIDAR signal information transmitted and received by the LIDAR sensors 31-1, ..., (31-N), and uses the point where the light is reflected to form a 3D modeling image. LiDAR images can be acquired.

In addition, the controller 30 utilizes the location information of the preset cubes (S) of FIGS. 4 and 5 to obtain LiDAR images of objects for each preset cube (S).

At this time, the cubes (S) mean each area of the divided hexahedral shape when the soccer stadium is divided into horizontal imaginary lines L1 and vertical imaginary lines L2.

That is, the controller 30 obtains a lidar image by scanning the lidar signal transmitted and received by the lidar sensors 31-1, ..., 31-N, and simultaneously acquires the position of the lidar image. LIDAR image for each cube (S) is obtained by utilizing information and location information of each cube (S).

6 is an exemplary view for explaining a 3D image for each cube of a trajectory of a ball acquired by an object detecting unit according to the present invention.

In general, actual kicks of soccer players are classified into various types such as inside kicks, in-front kicks, instep kicks, out-front kicks, and toe kicks, depending on the part of the foot that is in contact and the position of the ball at the time of hitting. Depending on this, the ball forms various trajectories.

However, since the trajectory of the ball is conventionally detected using only the initial and final positions of the ball, the ball's trajectory is detected with a line (L) connecting the initial and final positions of the ball, resulting in significantly lower precision and accuracy. do.

The present invention can solve this problem, and since the controller 30 of the present invention can acquire lidar images of the ball for each cube, tracking of the trajectory of the ball for each cube is possible, so that the kick It is possible to precisely detect the trajectory of various balls according to the type of the ball.

For example, as shown in FIG. 6, after the ball is rotated and bent to the right, when it heads toward the goal post, conventionally, a line (L) simply connecting the initial position (A) and the final position (B) of the ball is detected as the trajectory of the ball, but since the controller 30 of the present invention can detect objects by dividing each cube, it is possible to accurately detect the ball trajectory of a -> b -> c -> d .

In other words, the controller 30 of the present invention installs a plurality of LiDAR sensors 31 installed on the sides and top of the soccer stadium, and at the same time, the vertical virtual lines (L2) connecting the soccer field in the vertical direction Based on the plurality of cubes (S) formed by the horizontal virtual lines (L1) connected in the horizontal direction and the LIDAR image for each cube is obtained.

In addition, when the 3D image for each cube is obtained, the controller 30 analyzes the acquired 3D image to extract an edge image of the object (human body or ball), and then extracts a center line forming the center line of the object from the extracted edge image. detect the line

FIG. 7 is a block diagram showing the controller of FIG. 2, FIG. 8 (a) is an exemplary view showing an edge image for each cube extracted by the edge image extraction module of FIG. 7, and (b) is a conventional lidar. FIG. 9 is an example view showing a center line for each cube extracted by the center line extraction module for each cube in FIG. 7 .

As shown in FIG. 7, the controller 30 includes a control module 300, a memory 301, a data transmission/reception module 302, a LIDAR signal scanning module 303, and an edge image extraction module 304 for each cube. ), and a center line extraction module 305 for each cube.

The control module 300 is an O.S (Operating System) of the controller 30, and manages and controls the control targets 301, 302, 303, 304, and 305.

In addition, the control module 300 transmits the lidar signal information received from the lidar sensors 31-1, ..., (31-N) through the data transmission/reception module 301 to the lidar signal scanning module 303. output as

In the memory 301, installation location information of each lidar sensor 31 and location information for each cube are preset and stored.

In addition, the memory 301 temporarily stores edge images for each cube generated by the edge image extraction module 304 for each cube.

In addition, center line information for each cube generated by the center line extraction module 305 for each cube is temporarily stored in the memory 301 .

The data transmission/reception module 302 transmits and receives data with the lidar sensors 31-1, ..., (31-N) and the artificial intelligence analysis server 5.

The lidar signal scanning module 303 scans lidar signals transmitted and received by lidar sensors 31-1, ..., 31-N.

The edge image extraction module 304 for each cube applies a Sobel filter commonly used to detect an edge to the lidar image of FIG. An edge image 910 for each cube composed of edge lines as shown in FIG.

At this time, assuming that the controller 30 generates a lidar image that is not a separate edge image, as shown in (b) of FIG. Although this excessively increases, in the present invention, the edge image extraction module 304 for each cube of the controller 30, as shown in FIG. It is possible to significantly reduce the amount of calculation processing and maximize calculation processing efficiency.

At this time, the edge images 910 for each cube generated by the edge image extraction module 304 for each cube are temporarily stored in the memory 301 under the control of the control module 300 and at the same time through the data transceiver 302. It is transmitted to the artificial intelligence analysis server (5).

The center line extraction module 305 for each cube extracts the center line 101 of the object from the edge image 910 for each cube extracted by the edge image extraction module 304 for each cube.

In this case, as shown in FIG. 9 , the center line 101 means a line connecting the edge line of the object and the middle value pixel of the intersection of horizontal lines equally formed at regular intervals.

This center line 101 information for each cube is used to determine the position of the head (A), hands (B, C), and feet (D, E) of the body by the artificial intelligence analysis server 5 described later.

In addition, the center line information for each cube extracted by the edge image extraction module 305 for each cube is temporarily stored in the memory 301 under the control of the control module 300 and at the same time artificial intelligence through the data transceiver 302 It is transmitted to the analysis server (5).

10 is a block diagram illustrating an artificial intelligence analysis server of FIG. 2 .

As shown in FIG. 10, the artificial intelligence analysis server 5 includes a control unit 50, a database unit 51, a communication interface unit 52, an artificial intelligence learning unit 53, and a central line merging unit for each cube. (54), an artificial intelligence analysis unit (55), and a player-by-player ability value generation unit (56).

The control unit 50 is the O.S (Operating System) of the artificial intelligence analysis server 5, and manages and controls the control objects 51, 52, 53, 54, 55, and 56. .

In addition, when the control unit 50 receives edge images for each cube and center line information for each cube from the controller 30 of the data collection system 3 through the communication interface unit 52, it stores them in the database unit 51. At the same time as saving, it is input to the center line merging unit 54 for each cube.

In addition, the control unit 50 executes the artificial intelligence learning unit 53 every preset first cycle T1.

In addition, the control unit 50 stores the ability value information for each player generated by the ability value information generation unit 56 for each player, the matching information detected by the artificial intelligence analysis unit 55, and the situation pattern information for each player in the database unit 51. and at the same time control the communication interface unit 52 to transmit to the online soccer game operating server 7.

In the database unit 51, identification information of each soccer field, identification and location information of each of the lidar sensors 31 installed in the soccer field, and location information of each cube of the soccer field are preset and stored.

In addition, the data analysis unit 51 stores a game analysis algorithm, which is a machine learning algorithm learned in the artificial intelligence learning unit 53.

At this time, the game analysis algorithm is a machine learning algorithm that uses the edge image of each cube and the center line information of each cube as input data and the motion information of each player according to the situation as output data.

In addition, the database unit 51 stores edge images for each cube and center line information for each cube received from the controller 30 through the communication interface unit 52 .

In addition, the database unit 51 stores the ability value information for each player generated by the ability value information generation unit 56 for each player, matching information generated by the artificial intelligence analysis unit 55, and situation pattern information for each player.

The communication interface unit 52 transmits and receives data with the controller 30 and the online soccer game operating server 7 under the control of the controller 50 .

The artificial intelligence learning unit 53 is executed every preset first cycle T1 under the control of the control unit 50, and refers to the edge image of each cube and the center line information of each cube stored in the database unit 51 and It is used to learn the game analysis algorithm, which is a machine learning algorithm.

At this time, the game analysis algorithm learned by the artificial intelligence learning unit 53 is applied to the artificial intelligence analysis unit 55.

The center line merging unit 54 for each cube merges the center lines consecutive to adjacent cubes among the center line information for each cube received from the controller 30 by referring to preset location information for each cube.

11 is a block diagram illustrating an artificial intelligence analysis unit of FIG. 10 .

As shown in FIG. 11, the artificial intelligence analysis unit 55 includes a data input module 551, an analysis module 552, an object recognition module 553, a situation recognition module 554, and a matching information generation module 555. ), motion information detection module 556 for each player, situation pattern information detection module 557 for each player.

The data input module 551 receives the merged center line information from the center line merging unit 54 for each cube, and under the control of the control unit 50, the edge images for each cube received from the controller 30 and each Center line information for each cube is input.

The analysis module 552 converts the merged center line information input by the data input module 551, the edge images of each cube, and the center line information of each cube into a game learned by the artificial intelligence learning unit 53. The game is analyzed using the input data of the analysis algorithm.

At this time, the game analysis algorithm is defined as a machine learning algorithm that uses the edge image of each cube and the center line information of each cube as input data and the motion information of each player according to the situation as output data.

In addition, the analysis module 552 detects the movement of the foot, the part of the foot in contact with the ball, and the part of the ball in contact with the foot when the recognized player is in a kicking state in contact with the ball, and at the same time, after the kick is made, each cube Match the speed and direction of the ball by referring to the trajectory of the star ball.

12 is a diagram for detecting a hitting part of a soccer ball according to the present invention using the analysis module of the present invention.

As shown in FIG. 12, in the case of the soccer ball 70, the spin occurs in the soccer ball 70 depending on which part of the soccer ball the player 20 hits, and the spin of the soccer ball 70 The trajectory will change.

As shown in (a) of FIG. 12, when the player 20 hits the soccer ball 70, the body part of the player 20 does not always hit a certain portion of the soccer ball 70. Therefore, as shown in (b) of FIG. 12, it is possible to determine which part of the divided area the player 70 hits by dividing the outside of the soccer ball into predetermined areas.

13 is a front view of a foot hitting a soccer ball according to the present invention, and FIG. 14 is a side view of a foot hitting a soccer ball according to the present invention.

As shown in FIGS. 13 and 14, in soccer, various trajectories occur depending on which part of the foot of the player 20 hitting the soccer ball 70 is used to hit the soccer ball 70.

Among the body parts of the player in contact with the soccer ball 70, the foot may be divided into one or more areas to determine the contact area.

In the present invention, the foot is divided into five areas such as the toe 21, the in-front of the foot 22, the out-front of the foot 23, the instep 24, and the inside of the foot 25, but the contact area of the foot increases. or may be reduced.

The object recognition module 553 recognizes an object of a player or a ball by a game analysis algorithm and simultaneously detects positional information of the recognized object.

In addition, when the recognized object is a player, the object recognition module 553 uses the merged center lines as shown in FIG. , feet (D, E), etc. are detected.

At this time, in the drawings, for convenience of explanation, the body part is described as being composed of a head, hands, and feet as an example, but it is natural that the body part can be more subdivided and detected.

In addition, object information detected (recognized) by the object recognition module 553 is input to the situation recognition module 553, motion information generation module 555 for each player, and situation pattern information detection module 556 for each player.

The context recognition module 554 recognizes the current state of each team and player by a game analysis algorithm.

At this time, the current state can be detected as 'attack' and 'defense' in a wide range of categories, and the 'attack' category is 'counterattack', 'delay', 'tactic type', 'free kick', 'corner kick', and 'penalty' It can be classified into sub-components such as 'kick', and the 'defense' category includes 'defense against counterattack', 'pressure defense', 'local defense defense', 'dedicated mark defense', 'corner kick defense', 'penalty' It can be classified into sub-components such as 'kick defense', and to a smaller extent, each player's own position and motion vector (movement direction and speed), the position and motion vector of players on the same team, the position and Recognition can be subdivided into motion vector, ball position, and motion vector.

The matching information generation module 555 detects the motion of the foot, the part of the foot in contact with the ball, and the part of the ball in contact with the foot when the analysis module 552 is in contact with the ball. After the kick is made, matching information matching the speed and direction of the ball is generated by referring to the trajectory of the ball for each cube.

The motion information generation module 556 for each player detects motion information for each player.

At this time, the motion information includes the moving speed, jumping ability, passing ability, shooting ability, tackling ability, physical fighting ability, centering ability, trapping ability, etc. of the corresponding player.

The situation pattern detection module 557 for each player detects a movement pattern of each player according to the current state recognized by the situation recognition module 554 .

Returning to FIG. 10 again, looking at the ability value information generation unit 56 for each player, the ability value information generation unit 56 for each player refers to the movement information of each player generated by the movement information generation module 556 for each player. Generate ability information.

At this time, the control unit 50 transmits the matching information generated by the artificial intelligence analysis unit 55 and situation pattern information for each player, and the ability value information for each player generated by the ability value generation unit 56 for each player to an online soccer game operation server. Controls the communication interface unit 52 to transmit to (7).

The online soccer game operating server 7 is a server that produces a normal online soccer game and provides services at the same time.

At this time, the online soccer game operating server 7 includes player characters matched with actual players, and reflects matching information, situation pattern information for each player, and ability value information for each player transmitted from the artificial intelligence analysis server 5 to the corresponding player character. let it

In addition, the online soccer game operation server 7 allows a gamer to select either manual mode or automatic mode to play the game, and when the manual mode is selected, the soccer game is played based on the gamer's operation, When selecting a mode, it may be configured to play a soccer game based on artificial intelligence.

In addition, when the online soccer game operation server (7) selects the automatic mode, artificial intelligence itself determines the attack strategy, player position, penalty kick player, corner kick player, and free kick player using the information on the ability level and situation pattern of the player characters. You can proceed with the game.

As such, the online soccer game application method (S1), which is an embodiment of the present invention, uses LiDAR sensors to store trajectory information (position, speed, etc.) of each actual soccer player and trajectory information of the ball in a preset cube ( After collecting each S), the collected big data is analyzed using an artificial intelligence algorithm to detect the pattern information according to the ability level and situation of each player, and the pattern information according to the detected ability level and situation corresponds to the online soccer game. By applying it to the player character, it is possible to increase the realism and reality of the online soccer game by maximizing the match between the player character and the actual player for various variables, and to improve user participation in the online soccer game to create a user-centered It is possible to provide services to increase user interest, participation, and fun.

1: Online soccer game provision system 3: Data collection system
5: AI analysis server 7: Online soccer game operating server
10: communication network 30: controller
31-1, ..., 31-N: lidar sensors 50: control unit
51: database unit 52: communication interface unit
53: artificial intelligence learning unit 54: center line merging unit for each cube
55: artificial intelligence analysis unit 56: player ability information generation unit
300: control module 301: memory
302: data transmission/reception module 303: lidar signal scanning module
304: Edge image extraction module for each cube 305: Center line extraction module for each cube

Claims (8)

A data collection system including a plurality of sensing means installed in a soccer stadium to sense an object (player or ball) and a controller that transmits the sensed data sensed by the sensing means to the outside;
Using object detection data as input data, motion information for each player, including at least one of movement speed, jumping ability, passing ability, shooting ability, tackling ability, physical fighting ability, centering ability, and trapping ability, and situational The machine learning algorithm that outputs the situation pattern information for each player, which is pattern information, stores the game analysis algorithm, and analyzes the detected data transmitted from the controller using the game analysis algorithm to analyze the motion information for each player and the situation for each player. An artificial intelligence analysis server that detects pattern information;
An online soccer game operation server that produces online soccer games and provides game services at the same time, and reflects the movement information for each player and the situation pattern information for each player received from the artificial intelligence analysis server to the corresponding player character,
The sensing means is a LiDAR sensor that receives a reflected signal after transmitting a LiDAR signal,
The controller
When a football stadium is divided into first horizontal virtual lines in the longitudinal direction in the horizontal direction, second horizontal virtual lines in the width direction in the horizontal direction, and vertical virtual lines in the vertical direction, the divided hexahedron-shaped area is called a cube. memory for storing positional information of each cube when doing so;
a lidar signal scanning module for scanning lidar signals received from the lidar sensors;
Using the scanning information by the LIDAR signal scanning module, a Sobel filter is applied to generate an edge image, which is a 3D modeling image consisting of edge lines, which are pixels whose pixel change rate is greater than or equal to a threshold value, and utilizes the location information of each cube set in advance an edge image extraction module for each cube that extracts an edge image for each cube;
A control module for transmitting the edge image for each cube extracted by the edge image extraction module for each cube to the artificial intelligence analysis server,
The controller
When the center line is a line connecting the edge line of the object and the median pixel of the intersection of the horizontal lines formed evenly at regular intervals, from the edge image for each cube extracted by the edge image extraction module for each cube, the object Further comprising a center line extraction module for each cube that extracts the center line,
The control module transmits the center line information for each cube extracted by the center line extraction module for each cube to the artificial intelligence analysis server. delete delete delete The method of claim 1, wherein the artificial intelligence analysis server comprises an artificial intelligence analysis unit,
database unit;
a center line merging unit for each cube that merges successive center lines of successive cubes (S) by utilizing the center line information for each cube transmitted from the controller;
Using the game analysis algorithm, the merged center line information input from the center line merging unit for each cube and the edge images for each cube received from the controller are analyzed to obtain motion information for each player and situation for each player. Artificial intelligence analysis unit for generating pattern information;
The online soccer game application system comprising a control unit for transmitting motion information for each player and situation pattern information for each player generated by the artificial intelligence analysis unit to the online soccer game operating server. The method of claim 5, wherein the artificial intelligence analysis unit
a data input module for receiving the merged center line information input from the center line merging unit for each cube and the edge images for each cube transmitted from the controller;
an analysis module which executes the game analysis algorithm by using the merged center line information input by the data input module and the edge images of each cube as input data of the game analysis algorithm;
an object recognition module for recognizing a ball or body object for each cube through the game analysis algorithm;
a situational awareness module recognizing the current state of each player and ball by the game analysis algorithm;
a motion information generation module for each player for detecting motion information of each player by the game analysis algorithm;
Further comprising a situation pattern detection module for each player for detecting a movement pattern of each player according to the current state recognized by the situation recognition module;
The current state applied to the situational awareness module is classified into 'attack' and 'defense' in the widest range, and the 'attack' category is 'counterattack', 'delay', 'tactic type', 'free kick', and 'corner kick'. , 'Penalty Kick' is classified into sub-components including at least one of them, and the 'defense' category includes 'counterattack defense', 'pressure defense', 'regional defense defense', 'dedicated mark defense', ' It is classified into sub-components including at least one of 'corner kick defense' and 'penalty kick defense', and the smallest is each player's own position and motion vector (moving direction and speed), the position and An online soccer game application system characterized in that recognition is subdivided into motion vectors, positions and motion vectors of opponent team players, and position and motion vectors of the ball. The method of claim 6, wherein the artificial intelligence analysis unit
When the player is in contact with the ball, the movement of the foot, the part of the foot in contact with the ball, and the part of the ball in contact with the foot are detected, and the speed of the ball is determined by referring to the trajectory of each cube after the kick is made. and a matching information generation module for generating matching information matching directions. The method of claim 7, wherein the online soccer game operation server
Either the manual mode or the automatic mode is selected by the gamer, and when the manual mode is selected, the game is operated based on the gamer's operation, but when the automatic mode is selected, the artificial intelligence automatically provides movement information and matching information of the player characters. And an online soccer game application system characterized in that the game is played by determining attack tactics, player positions, penalty kick players, corner kick players, and free kick players using situational pattern information.

Images